Cassirer Substance and Function Einsteins Relativity

SUBSTANCE AND FUNCTION
AND
EINSTEIN S THEORY OF RELATIVITY

AND

BY
ERNST CASSIRER
AUTHORIZED TRANSLATION
BY
WILLIAM CURTIS SWABEY,

AND
PH.D.
PH.D.
MARIE COLLINS SWABEY,
CHICAGO LONDON
THE OPEN COURT PUBLISHING COMPANY

1923
COPYRIGHT 1923 BY THE OPEN COURT PUBLISHING COMPANY CHICAGO
Made
in United States of America
80
aai
COMPOSED AND PRINTED AT THE
WAVERLY PRESS
BT THE WILLIAMS & WILKINS COMPANV
BALTIMORE, U.
S.
A.
PREFACE
The
by
investigations contained in this volume were first prompted In the course of an studies in the philosophy of mathematics.
attempt to comprehend the fundamental conceptions of mathematics from the point of view of logic, it became necessary to analyse more closely the function of the concept itself and to trace it back to its
presuppositions. Here, however, a peculiar difficulty arose: the traditional logic of the concept, in its well-known features, proved inadequate even to characterize the problems to which the theory
mathematics led. It became increasingly evident that exact science had here reached questions for which there existed no precise correlate in the traditional language of formal logic. The
of the principles of
form
content of mathematical knowledge pointed back to a fundamental of the concept not clearly defined and recognized within logic
In particular, investigations concerning the concepts of the series and of the limit, the special results of which, however, could not be included in the general exposition of this book, con
itself.
firmed this view and led to a renewed analysis of the principles of the construction of concepts in general.
The problem thus defined gained more general meaning when it became clear that it was in no way limited to the field of mathe
matics, but extended over the whole field of exact science. The systematic structure of the exact sciences assumes different forms
according as
it
is
regarded in different logical perspectives.
Thus
an attempt had to be made to advance from this general point of view to the forms of conceptual construction of the special disciplines, It did not accord of arithmetic, geometry, physics and chemistry.
with the general purpose of the enquiry to collect special examples from the particular sciences for the support of the logical theory, but it was necessary to make an attempt to trace their systematic structures as wholes, in order that the fundamental unitary relation by which these structures are held together might be revealed more I did not conceal from myself the difficulty of carrying distinctly.
out such a plan; I finally resolved to make the attempt only because the value and significance of the preliminary work already accom plished within the special sciences became increasingly evident to me.
IV
PREFACE
Particularly in the exact sciences, the investigator has turned from the special problems to the philosophical foundations with ever clearer consciousness and energy. Whatever one may judge in detail
of the results of these researches, there can be no doubt that the logi cal problem has thereby been greatly and directly advanced. I
have, therefore, sought to base the following exposition upon the historical development of science itself and upon the systematic
presentation of cannot consider
its all
content by the great scientists. Although we the problems that arise here, nevertheless, the
special logical point of view which they represent must be carried through and verified in detail. What the concept is and means in its general function can only be shown by tracing this function
through the most important

representing
it
fields
of scientific investigation
and
in general outline.
The problem
logical
receives
new meaning when we advance from purely

conception of knowledge of
reality.
considerations to the
original opposition of thought and being breaks up into a number of different problems, which are, nevertheless, connected and held in
intellectual unity by their common point of departure. Whenever, in the history of philosophy, the question as to the relation of thought
The
and being, of knowledge and reality, has been raised, it has been dominated from the first by certain logical presuppositions, by cer tain views about the nature of the concept and judgment. Every change in this fundamental view indirectly produces a complete change in the way in which the general question is stated. The system of knowledge tolerates no isolated "formal" determination without consequences in all the problems and solutions of knowledge. The conception, therefore, that is formed of the fundamental nature
(
of the concept
is
which are generally considered under

("
directly significant in judging the questions of fact "Criticism of Knowledge,"
Erkenntniskritik"}
or
"Metaphysics."
The transformation which
these questions undergo when regarded from the general point of view that is gained by criticism of the exact sciences and the new
show.
direction which their solution takes, Part II of the book attempts to Both parts, though seemingly separate in content, are united,
nevertheless, in a philosophical point of view; both attempt to repre sent a single problem which has expanded from a fixed center, drawing
ever wider and more concrete
fields into its circle.
ERNST CASSIRER.
TRANSLATORS PREFACE
was thought that there was need for some comprehensive work on the philosophy of the exact sciences which w ould do full justice to the newer developments in mathe matical and physical speculation while showing at the same time It seemed that the the historical connections of these tendencies. of Professor Ernst Cassirer herewith presented fulfilled two works
It
in the English language

r
these requirements best of

tive
all.
The reader
will find here a construc
and systematic survey of the whole field of the principles of the exact sciences from the standpoint of a logical idealism, which is historically derived from Kant, but which lacks the fatal rigidity As Professor Cassirer develops his logical or of the latter s system. His critical idealism it becomes a doctrine of creative intelligence.
doctrine is neither idealism, pragmatism nor realism as these terms are understood in our English-speaking philosophy; it is rather a positivistic and non-static rationalism, which seeks to preserve the spirit which unites Plato, Descartes, Leibniz and Kant and to
show how this spirit reaches its fulfillment in the modern develop ment of mathematical and physical theory. The first part of the present book, Substanzbegriff und Funktionsbegriff was published in 1910, while the second part, which we have called the Supplement, Zur Einstein schen Relativitdtstheorie> ap peared in 1921. The intervening period was, of course, one of immense importance for the philosophy of physics, since it marked the development of the new and revolutionary theory of relativity. In accordance with the fundamental maxim of his critical method
of science, of space
Professor Cassirer based his analysis in 1910 on the historical state which was still dominated by the Newtonian conceptions
and time. On the ground of the same maxim, he has since taken account of the new theory of relativity and has, with good
logical justification, seen in the latter the relative completion and realization of the historical tendency which he had described in his
earlier
works.
Professor Cassirer
philosophy
may
be regarded
as a fundamental epistemological "theory of relativity" which sets forth a general philosophical standpoint from which Einstein s theory
is
seen to be only the latest and most radical fulfillment of the motives
VI
TRANSLATORS PREFACE
which are inherent in mathematical and physical science as such. While Professor Cassirer has had his fundamental principles confirmed rather than disproved by this recent development, his discussion in Chapter IV, Section VI, of Substance and Function must be taken in connection with his later statements. With regard to the translation, the translators are aware that a good deal of the vigor and savor of the original has escaped in the process of substituting correct but colorless terms for the more vivid language of the original. Accuracy and clarity have been their
They alone are responsible for the italicized paragraph headings, which were inserted because it was thought that the book might be used as a text or reference work in connection with an
chief aim.
advanced course
in the
Theory
of
Knowledge and that perhaps these
guide-posts might help the student in finding his difficult material.
way through
the
Professor Cassirer himself kindly read the entire work in manuscript . and, in a friendly letter, states that "nach der Gesamteindruck besteht fur mich kein Zweifel dass der Sinn des Ganzen richtig getroffen
.
We wish herewith to express our hearty thanks to Professor Cassirer for permitting us to translate his works as well as for his trouble in reading the manuscript of the translation and for his courtesy in the whole transaction.
und wiedergegeben
ist."
WILLIAM CURTIS SWABEY, MARIE COLLINS SWABEY.
CONTENTS
PART
I
THE CONCEPT OF THING AND THE CONCEPT OF RELATION

CHAPTER
I
ON THE THEORY OF THE FORMATION OF CONCEPTS

I.
New
II.
developments in logic. The concept in Aristotelian logic. Purpose and nature of the generic concept. The problem of abstraction. The metaphysical presuppositions of Aristotelian The concept of substance in logic and metaphysics logic. The psychological criticism of the concept (Berkeley). The psy chology of abstraction. Mill s analysis of mathematical concepts.
The
forms of
III.
defect of the psychological theory of abstraction. The The place of the thing-concept in the system series.
of logical relations
The negative process of "abstraction." The mathematical con cept and its "concrete universality." The criticism of the theories of abstraction. Objects of the and "second" orders. The variety of objective "intentions." The serial form and the members of the series
"first"
18
CHAPTER
II
THE CONCEPT OF NUMBER

I.
The
sensationalistic deduction of number.
Frege
foundations of
presenta
.
arithmetic.
tion.
The system
of arithmetic.
Number and
of presentation and the act of presentation. II. The logical foundations of the pure concept of number (Dedekind). The logic of relations. The concept of progression. Number
The content
27
as ordinal number.
The
theories of Helmholtz
s
and Kronecker.
35
Criticism of the nominalistic deduction III. Number and the concept of class. Russell
theory of cardinal
numbers.
cept.
of the zero
The logical definition theories." The presupposition of the class con 44 The generic concept and the relational concept
Criticism of
"class
and
of unity.
IV. Extension of the concept of number.
Gauss theory
of the negative
and imaginary numbers.
The
irrational numbers.
Dedekind
ill
CONTENTS
explanation of the irrational numbers. The problem of the The concept of "power." The production transfinite numbers. The second "principle of generation" of of transfinite numbers.
numbers (Cantor)
54
CHAPTER
III
THE CONCEPT OP SPACE AND GEOMETRY

I.
Concept and form. The method of ancient geometry. The con cept of space and the concept of number. The fundamental
principle of analytic geometry. Magnitudes and functions
The
infinitesimal geometry.
68
II. Intuition
tion.
and thought in the principles of the geometry of posi Steiner and Poncelet. The concept of "correlation" and
the principle of continuity. The transference of relations dis tinguished from induction and analogy. Projection and the imaginary in geometry. Metrical and projective geometry, and quadrilateral construction of Staudt. Projective metric (Cayley
and Klein. The concept of space and the concept of order. Geometry and the group theory. The concepts of constancy and change in geometry
III. Characteristic
76
(Kombinatorik) as pure "doctrine of forms" (Leib niz). Geometry as pure "doctrine of relations" (Hilbert). The syntheses of generating relations. Grassmann s Ausdehnungslehre and its logical principles. The forms of calculus, and the concept of the Source IV. The problem of metageometry. The attempt at an empirical
91
grounding of geometry (Pasch). Ideal objects in empirical geometry. Veronese s modification of empiricism. Rationalism and empiricism. Mathematical space and sensuous space. Ob jections to the Kantian theory of geometry. Real space and experiment. The conceptual principles of pure space. Eu clidean space and the other forms of mathematical space.
Geometry and
reality
100
CHAPTER IV
THE CONCEPTS OF NATURAL SCIENCE
I.
The constructive concepts and the concepts of nature. The con cept of traditional logic and the scientific ideal of pure descrip tion. The apparent logical ideal of physics. Is this the true
ideal of physics?
112
II.
Numbering and measuring as presuppositions. Mechanism and the concept of motion. The "subject" of motion. The "limiting concept" and its significance for natural science (Karl Pearson). The P. du Bois-Reymond s theory of the limiting concept.
CONTENTS
IX
III.
problem of existence. The existence of the limiting point. Logical idealism on the problem of existence. Consequences of the confusion of truth and reality. The "idealization" of pres 115 entations. The relation of the ideal and reality The problem of the physical method and its history. The problem
The sceptical theory of knowledge (Pro The concepts of nature and purpose (Plato). Mathematics and teleology (Plato, Aristotle, Kepler). The concept of hypothesis (Kepler and Newton). The logical and
of
knowledge (Plato).
tagoras, etc.).
ontological "hypotheses." IV. Robert Mayer s methodology of natural science.
130
Hypotheses and
The presuppositions of physical "measurement." The physical and the physical "theory." Units of measurement. The verification of physical hypotheses. The motive of serial construction. The physical concepts of series. V. The concept of substance in the Ionian philosophy of nature. The hypostatization of sensuous qualities (Anaxagoras). The hypostatization of sensuous qualities (Aristotle). Atomism and num ber. The impact of atoms. The postulate of continuity, and the The concept of the "simple" atom of Boscovich and Fechner. atom and the application of differential equations. The changes in the concept of the atom. The concept of the ether. The and logical form of the concept of the physical object. The "not real" elements in the concepts of the physical object. concept of non-being. Matter and idea and Galileo s concept
natural laws.
"fact"
.
139
"Real"
of inertia
151
VI. The concepts of space and time. Newton s concepts of absolute space and absolute time. The system of reference of pure mechanics. The substitution of the fixed stars for absolute space. The "intellectual experiment" and the law of inertia.
Streintz s concept of the "fundamental body." The theory of C. Neumann: the body alpha. Space and time as mathematical
ideals.
Hertz
system of mechanics.
Construction and con

170
vention
VII.
The concept of energy. Energy and the sense qualities. Energy and the concept of number. The concept of the measure of work. The formal presuppositions of energism. Rankine s de ductions of energetics. Criticism of the method of physical "abstraction." The problem of abstraction in modern logic.
Energy as a relational concept.
Energetics and mechanics. 187 Physics as a science of qualities VIII. The problem of the construction of concepts in chemistry. The chemistry of sensuous qualities and Richter s law of definite pro portions. Dalton s law of multiple proportions. The atom as a relational concept. The "regulative" use of the concept of the atom. The concept of valency and the theory of types. Logical aspects of the concept of type. The chemical concept as a rela-
CONTENTS
tional concept. The concept of the "radical" and the theories The reconstruction of the systematic of "composite radicals." form of chemistry. The periodic system of the elements.
Chemistry and mathematics IX. The concept of natural science and
203
"reality."
Rickert s theory of the scientific construction of concepts. Criticism of Rickert s theory. Word-meanings and mathematical concepts. Rickert s
of
"meanings"
and
"presentations."
The concept
as
(confusion
relations
the expression of individual relations. The problem of the con stants of natural science. Magnitudes and other forms of
220
PART
II
THE SYSTEM OF RELATIONAL CONCEPTS AND THE PROBLEM OF REALITY

CHAPTER V
ON THE PROBLEM OP INDUCTION
I.
The metaphysical tendency
in
induction and deduction.
The
empirical theory of judgment. Mach s "thought-experiment." Criticism of Mach s theory. Locke s theory of empirical judg ment. The "element of eternity" in all empirical judgment. The postulate of necessary determinateness. Judgments of
perception and judgments of experience. Experience as aggre gate and as system. Discrete and continuous "wholes." In duction and the theory of invariants. Induction and analogy. 237 II. Induction and analysis, "compositive" and "resolutive" methods. Experiment as the means of analysis. The relation of "uni
.
versal"
position."
relation
III.
and "particular" relations. "Isolation" and "super Laws and rules. The concept of the "fundamental" and the relation of mathematical necessity. The two
252
of laws of nature.
fundamental types of knowledge
The problem
ants of
Laws and constants.

of the a priori
The general
"invari
form of experience.
experience."
The concept
and the
265
CHAPTER VI
THE CONCEPT OF REALITY
I.
The separation
of "subjective" and "objective" reality. The development of the concepts of objectivity and subjectivity. Changing and constant elements of experience. The subjectivity of the sensuous qualities. The series of degrees of objectivity. The logical gradations of the contents of experience. The prob-
CONTENTS
lem of transcendence. The meaning of judgment. The "tran of sensuous experience. The concept of "represen Transformation of the concept of representation and tation."
progress to the
"whole
XI
scending"
of
experience."
Association as a prin
271
ciple of explanation II. The concept of objectivity
of projection
and
its
and the problem of space. The theory defects. Concept and perception distin
guished (Helmholtz).
"Projection"
The
division into circles of objectivity.
III.
and "selection." 286 The function of judgment; permanence and repetition. The prob lem of the "transsubjective." The correlation of the conscious ness of the ego and the consciousness of the object. The sepa ration of thought and experience. The concept of the object in
The objectivity within pure mathematics. The unity of the physical world 293 The historical transformation of the Helmholtz theory of signs. The logical and the jyitoLogical conceptions of rela The unity of the scientific views 302 "world. T.TTT.77"T tivity.
critical idealism.
"thing."
o?"th"e
CHAPTER VII
SUBJECTIVITY AND OBJECTIVITY OF THE RELATIONAL CONCEPTS
l|
II.
The problem of the subjectivity and objectivity of relational con The universal functions of rational and empirical knowl cepts. and "matter" of edge. The reciprocal relation of the knowledge. The existence of the "eternal truths." The con 309 cept of truth of modern mathematics The relational concepts and the activity of the ego. Constancy and change in knowledge. The independence of logical truths of the thinking subject. The problem of pragmatism. Truth and the "practical." The critical concept of truth. The reconcili ation of permanence and change. The double form of the
"form"
concept
314
CHAPTER VIII
ON THE PSYCHOLOGY OF RELATIONS
I.
Logical relations and the problem of self-consciousness. Plato psychology of relations. Aristotle s Doctrine of the ~K.oi.vovb
"Thoughts of relation" in
modern psychology.
"form-quality"
substance.
The doctrine
of the
in
The concept of modern psy
chology. Ebbinghaus s physiological account of relations. Criticism of the physiological explanation of relational concepts. 326 II. Meinong s theory of "founded contents." "Objects of a higher
order." The conflict between empiricism and nativism. The 337 psychology of the idea of space. The psychology of thought.
. .
Xll
CONTENTS
SUPPLEMENT
EINSTEIN
I.
THEORY OF RELATIVITY CONSIDERED FROM THE

EPISTEMOLOGICAL STANDPOINT
351
of the theory of rela
Concepts of measure and concepts of things
II.
The empirical and conceptual foundations

tivity
367
.
III.
The philosophical concept of truth and the theory of relativity. 387 394 IV. Matter, ether and space V. The concepts of space and time of critical idealism and the theory
of relativity
VI. Euclidean and non-Euclidean geometry VII. The theory of relativity and the problem of reality
Bibliography
Index...
.
409 430 445 457

461
PARTI
THE CONCEPT OF THING AND THE CONCEPT
OF RELATION
CHAPTER
ON THE THEORY
OF THE FORMATION OF CONCEPTS
is developing in the foundations of theoretical contemporary philosophy regarding knowledge is manifested perhaps nowhere as clearly as in the trans
New
developments in
logic.
The new view that
formation
it
In logic alone, of the chief doctrines of formal logic. seemed to have gained a firm foundation; in philosophical thought
a field seemed to be marked off that was assured against all the doubts aroused by the various epistemological standpoints and hy- ,] The judgment of Kant appeared verified and confirmed ^ potheses. that here the steady and secure way of science had finally been jj reached. The further consideration that, as logic since the time of Aristotle had not had to retrace a single step, so also it had not
single step, seemed from this point of view a confirmation of its peculiar certainty. Undisturbed by the continuous trans formation of all material knowledge, it alone remained constant and
advanced a
without variation.
If
we
follow the evolution of science within the last decades
more
closely, however, a different picture of formal logic appears. Every where it is occupied with new questions and dominated by new tendencies of thought. The work of centuries in the formulation of
fundamental doctrines seems more and more to crumble away; while on the other hand, great new groups of problems, resulting from the general mathematical theory of the manifold, now press to the fore ground. This theory appears increasingly as the common goal toward which the various logical problems, that were formerly investigated separately, tend and through which they receive their ideal unity. Logic is thereby freed from its isolation and again led to concrete tasks and achievements. For the scope of the modern theory of the manifold is not limited to purely mathematical problems, but involves a general view which influences even the special methods of the natural sciences and is therein verified. But the systematic connection into which logic is thus drawn compels renewed criticism
of its presuppositions.
disappears criticism
;
The appearance of unconditional now begins to be applied to those very

3
certainty doctrines
4:
that have persisted unchanged historically in the face of profound changes in the ideal of knowledge.
The concept in Aristotelian logic. The Aristotelian logic, in its general principles, is a true expression and mirror of the Aristotelian metaphysics. Only in connection with the belief upon which the latter rests, can it be understood in its peculiar motives. The conof the nature and divisions of being predetermines the con ception In the further ception of the fundamental forms of thought. of logic, however, its connections with the Aristotelian development ontology in its special form begin to loosen; still its connection with
the basic doctrine of the latter persists, and clearly reappears at definite turning points of historical evolution. Indeed, the basic
|T
which is ascribed to the theory of the concept in the structure of logic, points to this connection. Modern attempts to reform logic have sought in this regard to reverse the traditional
significance,
order of problems by placing the theory of the judgment before the theory of the concept. Fruitful as this point of view has proved to
be, it has, nevertheless, not been maintained in its full purity against the systematic tendency which dominated the old arrangement. The intellectual tendency still shaping these new attempts revealed itself in that features crept into the theory of judgment itself, which
could only be understood and justified- by_the traditional theory of the generic concept (Gattungsbegriff).\ The primacy of the concept, which they sought to lay aside, was once more implicitly ac
knowledged. The actual center of gravity of the system had not been changed but merely the external arrangement of its elements. Every attempt to transform logic must concentrate above all upon this one point: all criticism of formal logic is comprised in
I
criticism of the general doctrine of the construction of concepts
this doctrine are
Purpose and nature of the generic concept. The chief features of well-known and do not need detailed exposition.
r
\
Its presuppositions are simple and clear; and they agree so largely with the fundamental conceptions, which the ordinary view of the world consistently uses and applies, that they seem to offer no foothold for criticism. Nothing is presupposed save the existence of
things in their inexhaustible multiplicity, and the power of the mind to select from this \vealth of particular existences those features that are common to several of them. When we thus collect objects
characterized
by possession
of
some common property
into classes.
THEORY OF THE FORMATION OF CONCEPTS

and when we repeat this process upon higher levels, there gradually an ever firmer order and division of being, according to the
arises
u series of factual similarities running through the particular things. r The essential functions of thought, in this connection, are merely those of comparing and differentiating a sensuously given manifold. J Reflection, which passes hither and thither among the particular objects in order to determine the essential featuresîn-wliich
l j
Abstraction la/s hojd^jion and agree, leads of itself to abstraction. raises to clear consciousness these related features,v pure, t)y them
selves, freed
from
all
admixture of dissimilar
elements."
Thus the
peculiar merit of this interpretation seems to be that it never destroys or imperils the unity of the ordinary view of the world. The c_oncept
does not appear as something foreign to sensuous reality, but forms a part of this reality; it is a selection from what is immediately contained in it. In this respect, the concepts of the exact mathematical
sciences stand
sciences,
-^
\
\
upon the same plane as the concepts
of tbe descriptive
ordering^mcT""""
which are merely concerned with a superficial Just as we form the concept of a tree classification .of what is given. by selecting from the totality of oaks, beeches and birch trees, the group of common properties, so, in exactly the same way, we form
the concept of a plane rectangular figure by_igolating the common properties which are found in the square, the right angle, the
rhomboid, the rhombus, the symmetrical and asymmetrical trapezium and trapezoid, and which can be immediately seen and pointed out The well-known guiding principles of the concept follow of themselves from thesr foundations. Every series of comparable objects has a
supreme generic concept/ which comprehends within itself all the determinations in which these objects agree, while on the other hand,
within this supreme genus, the sub-species at various levels are defined by properties belonging only to a part of the elements. In the same
way that we ascend from
the species to the higher genus by abandon-^ a certain characteristic, thereby, drawing a larger range of objects ing into the circle, so by a reverse process, the specification of the genus
takes place through the progressive addition of new elements 6f content. \ Hence, if we call the number of properties of a concept
magnitude of
1
its content, this
magnitude increases as we descend from

der Logik, Ed. 4, Leipzig, 1875, 51 ff.
16
ff.;
Cf. e.g., Drobisch,
Neue Darslellung
1857,
Uberweg, System der Logik, Bonn,

2
Cf. intension. (Tr.)
the higher concepts to the lower, and thus diminishes the number of pecies subordinate to the concept; while, when we ascend to the higher genus, this content will diminish as the number of species is increased. This increasing extension of the concept corresponds to a progressive diminution of the content; so that finally, the most general
concepts
we can reach no
conceptual pyramid, which we form

in the abstract representation of
longer possess any definite content. The in this way, reaches its summit
"something"
under the
all-inclusive
being of which every possible intellectual content falls, but at the same time is totally devoid_pf specific meaning.
of abstraction. of the concept logical theory
its
which
The problem
universal
At
this
point
first
in
the
traditional
arise
the
this
method
validity and applicability. of forming concepts is entirely

it
jckmbtsx concerning If the fin^l goal of
empty,
Such"
process leading to
is
must arouse
suspicion.
the whole an outcome
unintelligible
if
the individual steps have

to
fulfilled
the require
fruitful
ments,
which we are accustomed
make
of
every
and concrete process of construction of scientific concepts. What we demand and expect of a scientific concept, first of all, is this: that, in the place of original indefiniteness and ambiguity of ideas, it shall institute a sharp and unambiguous determination; while, in this case, on the contrary, the sharp lines of distinction seem the more effaced, the further we pursue the logical process. And in fact, from the
standpoint of formal logic itself, a new problem arises here. If all construction of concepts consists in selecting from a plurality of objects before us only the similar properties, while we neglect the
rest, it is clear
that through this sort of reduction what
is
merely a
This part, ijparf has taken the place of the original sensuous whole. The concept ffihowever, claims to characterize and explain the whole. \ would lose all value if it meant merely the neglect of the particular
I
and the annihilation of their peculiarity. is meant to be the expression of a thoroughly positive process; what remains is not to be merely an arbitrarily chosen part but an "essential" moment by which the
cases
it starts,
from which
The
act of negation, on the contrary,
whole
is
determined.
The higher concept
is
to
make
the lower
intelligible by setting forth in abstraction the ground of its special form. The traditional rule, however, for the formation of the generic concept contains in itself no guarantee that this end will be actually
achieved.
In fact, there
is
nothing to assure us that the
common

properties,
which we
truly
select
include
the
from any arbitrary collection of objects, which characterize and typical features,
determine the total structures of the members of the collection. We may borrow a drastic example from Lotze: If we group cherries
and meat together under the attributes
red, juicy
and
edible,
we do
not thereby attain a valid logical concept but a meaningless combina tion of words, quite useless for the comprehension of the particular
cases.
Thus
it
does not
suffice;
becomes clear that the general formal rule in itself that on the contrary, there is always tacit reference
logic.
to another intellectual criterion to supplement it. The metaphysical presuppositions of Aristotelian
In the
system of Aristotle, this criterion

are
left in logic
is
are
filled in
and made
plainly evident; the gaps that good by the Aristotelian metais
the special link that binds the two fields together. For Aristotle, at least, the concept is no mere subjective schema in which we collect the common elements of
physics.
The
doctrine of the concept
an arbitrary group of things. The selection of what is common -remains an empty play of ideas if it is not assumed that what is thus gained is, at the same time, the real Form which guarantees the causal and teleological connection of particular things. The real and ulti mate similarities of things are also the creative forces from which they spring and according to which they are formed. The process
of comparing things and of grouping them together according to similar properties, as it is expressed first of all in language, does not lead to what is indefinite, but if rightly conducted, ends in the dis-
"Specific
co very of the real essences of things. Thought only isolates the type; tEiiTlatteTTs contained as an active factor in the indi
special forms.
vidual concrete reality and gives the general pattern to the manifold The biological species signifies both the end toward
which the living individual strives and the immanent force by which
its
evolution
is
guided.
The
logical doctrine of the construction of
~"\
the concept and_of_definition can only be built up with reference to ihese fundamental relations of the real. The determination of concept according to its next higher genus and its specific difference
reproduces the process by which the real substance successively unfolds itself in forms of being. Thus it is this basic jits special of substance to which the purely logical theories of Aristotle inception
constantly have reference.
The complete system
of scientific defini-
8
tions
would also be a complete expression of the substantial forces which control reality. 3 The concept of substance in logic and metaphysics. An understand ing of Aristotle s logic is thus conditioned by an understanding of his conception of being. Aristotle himself clearly distinguishes the various sorts and meanings of being from each other; and it is the essential problem of his theory of the categories to trace through
and make clear this division of being into its various subspecies. Thus he also expressly distinguishes the existence, which Is indicated by mere relations in judgment, from existence after the fashion of a ingi the being of a conceptual synthesis, from that of a concrete
subject.
In
all
these quests for a sharper division, however, the
Only logical primacy of the concept of substance is not questioned. in given, existing substances are the various determinations of being
Only in a fixed thing-like substratum, which must first be given, can the logical and grammatical varieties of being in general find their ground and real application. Quantity and quality, and time determinations, do not exist in and for themselves, space but merely as properties of absolute realities which exist by them selves. The category of relation especially is forced into a dependent
thinkable.
and subordinate position by

of Aristotle.
it
Relation
is
this fundamental metaphysical doctrine not independent of the concept of real being
;
can only add supplementary and external modifications to the In this way the latter, such as do not affect its real "nature." Aristotelian doctrine of the formation of the concept came to have a characteristic feature, which has remained in spite of all the mani fold transformations it has undergone. The fundamental categorical relation of the thing to its properties remains henceforth the guiding
point of view; while relational determinations are only considered in
so far as they can be transformed, by some sort of mediation, into This view is in properties of a subject or of a plurality of subjects. evidence in the text-books of formal logic in that relations or con
among the "non-essential" proper a concept, and thus as capable of being left out of its defi nition without fallacy. Here a methodological distinction of great
ties of
3
nections, as a rule, are considered
On
cially, Prantl, Geschichte der
the metaphysical presuppositions of the Aristotelian logic, cf. espe LogikimAbendlande, I; Trendelenburg, Geschichle
der
Kategorienlehre; H. Maier, Tubingen, 1900, pp. 183 ff.
Die
Syllogistik
des
Aristoteles,
II,
2,

significance
The two chief forms of logic, which are appears. to each other in the modern scientific develop especially opposed ment, are distinguished as will become clear by the different
value which
is
placed upon thing-concepts and

ii
The psychological
this general criterion,
supposition, upon In fact the whole special doctrines of the peripatetic metaphysics. struggle against the Aristotelian "concept realism" has been without
effect upon this decisive point. The conflict between nominalism and realism concerned only the question of the metaphysical reality of concepts, while the question as to their valid logical definition was not considered. The reality of "universals" was in question. But what was beyond all doubt, as if by tacit agreement of the conflict ing parties, was just this: that the concept was to be conceived as a
If we accept recognize further that the essential pre which Aristotle founded his logic, has survived the
criticism of the concept (Berkeley).
we
universal genus, as the
common
element in a series of similar or
this mutual assumption, resembling particular things. all conflict as to whether the common element possessed a separate factual existence or could only be pointed out as a sensuous moment in the individuals,
Without
would be
essentially unintelligible.
Moreover,
the psychological criticism of the "abstract" concept, radical as it seems at first sight, introduces no real change here. In the case of Berkeley, we can follow in detail how his skepticism as to the worth
and
fruitfulness of the abstract concept implied, at the sarae time, a dogmatic belief in the ordinary definition of the concept. That
the true scientific concept, that in particular the concepts of mathe matics and physics, might have another purpose to fulfill than is
ascribed to
them
4
in this scholastic definition,
this
thought was not
the psychological deduction of the fact, concept, the traditional schema is not so much changed as carried over to another field. While formerly it had been outer things that
comprehended.
In
in
common element was selected, merely transferred to presentations as psychi cal correlates of things. The process is only, as it were, removed to another dimension, in that it is taken out of the field of the physical
were compared and out of which a
here the same process
is
For greater detail c/. my Das Erkenntnisproblem in der Philosophic Wissenschaft der neuern Zeit, Vol. II, Berlin, 1907, pp. 219 ff.
und
10
remain the same.
into that of the psychical, while its general course and structure When several composite presentations have a of their content in common, there arises from them, according part
to the well-known psychological laws of simultaneous stimulation and fusion of the similar, a content in which merely the agreeing
determinations are retained and
all
the others suppressed. 5
In this
way no new, independent and special structure is produced, but only a certain division of presentations already given, a division in which
certain
tion,
To
are emphasized by a one-sided direction of atten way raised more sharply out of their surroundings. the "substantial forms," which, according to Aristotle, represent
moments
in this
and
the final goal of this comparing activity, there correspond certain fundamental elements, which run through the whole field of percep
And it is now asserted still more emphatically that these absolute elements alone, existing for themselves, constitute the real kernel of what is given and Again the role of relation is
tion.
"real."
limited as
much
as possible.
Hamilton, with
all
his recognition
of the Berkeleian theory,
showed the
characteristic function of relat-
/ s ing thought. Against him, J. Stuart Mill emphasizes that the true [/ positive being of every relation lies only in the individual members
which are bound together by it, and that hence, since these members can only be given as individuals, there can be no talk of a general
The concept does not exist save as a part of a concrete presentation and burdened with all the attributes of presen tation. What gives it the appearance of independent value and
meaning
of relation. 6
underived psychological character is merely the circumstance that our attention, being limited in its powers, is never able to illumine the whole of the presentation and must of necessity be narrowed to a
mere
selection of parts. The consciousness of the concept is re solved for psychological analysis into consciousness of a presentation or part of a presentation, which is associatively connected with some
word or other sensuous

The psychology of
according
to
this
sign.
abstraction.
The
"psychology
of
abstraction,"
key to the logical of every form of concept. This meaning is derived from meaning the simple capacity of reproducing any given content of presenta
view,
furnishes the
real
tion.
5
Abstract objects arise in every perceiving being in

tlberweg, op.
cit.,
whom
like
Cf., e.g.,
51.
s
Cf. Mill, 1865, p. 319.
An
Examination of Sir William Hamilton
Philosophy, London,
11
sentations.
determinations of the perceived have been given in repeated pre For these determinations are not confined to the
particular moment of perception, but leave behind them certain Since these traces of their existence in the psycho-physical subject. which must be thought of as unconscious during the time traces,
between the real perception and the recall, are again aroused by newly occurring stimuli of a similar sort, a firm connection is gradually formed between the similar elemeats of successive percep That which differentiates them tends more and more to tions. disappear; it finally forms only a shadowy background on which
the
constant
features
stand
of these
out
the
more
clearly.
The
their
pro
fusion agree, gressive into a unitary, indivisible whole, constitutes the psychological nature of the concept, which is consequently in origin as in function merely a
solidification
features that
totality of memory-residues, which tions of real things and processes. shown in that they exert a special
have been
left in
us by percep
The
and
reality of these residues is independent influence upon
newly occurring content and transformed according to them. Thus we stand apprehended and this is sometimes emphasized by the advocates of this here, view themselves, at a point of view closely akin to that of medieval "conceptualism;" real and verbal abstracta can only be taken from
the act of perception
is
itself,
in so far as every
the content of perception because they are already containedjrui^-ê common elements. The difference between the ontological and
psychological views
is
merely that the
"things"
of scholasticism
were
the beings copied in thought, while here the objects are meant to be nothing more than the contents of perception.
may appear from the standpoint of meaning and content of the logical problem are, If we remain in the sphere of this nevertheless, not affected by it. latter problem, we find here a common, fundamental belief regarding the concept, which has remained apparently unassailable throughout all the changes of the question. Yet precisely where there seems no
Weighty as
this distinction
metaphysics, the
conflict of opinion, the real methodological difficulty begins.* Is the theory of the concept, as here developed, an adequate and faithful
and characterize
picture of the procedure of the concrete sciences? Does it include all the special features of this procedure; and is it
able to represent
7
them
in all their
mutual connections and

2,
specific
Cf. especially B.
Erdmann, Logik, Ed.
pp. 65
ff.,
88
ff.
12
characters?
With regard to the Aristotelian theory, at least, this must be answered negatively. The concepts, which are question Aristotle s special object and interest, are the generic concepts of of the the descriptive and classifying natural sciences. The the horse, the lion, is to be ascertained and established. olive-tree, Wherever he leaves the field of biological thought, his theory of the concept at once ceases to develop naturally and freely. From the
"form"
beginning, the concepts of geometry, especially, resist reduction to the customary schema. The concept of the point, or of the line, or of the surface cannot be pointed out as an immediate part of physicallypresent bodies and separated from them bysimple "abstraction." Even in this example, which is the simplest offered by exact science,
logical technique faces a
(
new problem.
definition,
which
arise
through genetic
Mathematical concepts, through the intellectual

f
establishment of a constructive connection, are different

cal concepts,
rom empiri-
which aim merely to be copies of certain factual charac While in the latter case, the teristics of the given reality of things. multiplicity of things is given in and for itself and is only drawn together for the sake of an abbreviated verbal or intellectual expres sion, in the former case we first have to create the multiplicity which is the object of consideration, by producing from a simple act of construction (Setzung}, by progressive synthesis, a systematic con nection of thought-constructions (Denkgebi Ideri) There appears here in opposition to bare "abstraction," an act of thought itself, a free production of certain relational systems. It can easily be understood that the logical theory of abstraction, even in its modern forms, has frequently attempted to obliterate this opposition, for
.
at this point that questions as to the value and inner unity of the theory of abstraction must be decided. But this very attempt leads at once to a transformation and disintegration of the theory,
it is
in
whose favor
it
was undertaken.
The
doctrine of abstraction
loses either its universal validity or the specific logical character
that originally belonged to it. Mill s analysis of mathematical concepts. Thus Mill, for instance, in order to maintain the unity of the supreme principle of experi
ence, explains mathematical truths and concepts as also mere expres sions of concrete physical matters of fact. The proposition that
1
upon
= 2 merely describes an experience which has been forced us by the process of joining things together. In another sort
13
tion of
lose
of world of objects, in a world, for example, in which by the combina two things, a third always came into being of itself, it would
significance
and
validity.
The same
"round
is
true of the axioms
concerning spatial relations; a dictory concept for us because
it
is only a contra has been shown in experience

square"
without exception, that a thing loses the property of having four corners the moment it assumes the property of roundness, so that the beginning of one impression is inseparably connected with the cessa
According to this mode of explanation, geometry and arithmetic seem again resolved into mere statements concerning But this interpretation fails when certain groups of presentations. Mill further attempts to justify the value and peculiar significance, inherent in these special experiences of numbering and measuring, in
tion of the other.
the whole of our knowledge. Here, first of all, reference is made to the accuracy and trustworthiness of the images, which we retain of
spatial
and temporal
relations.
The reproduced presentation
is,
a varied experience has shown* the image that the geometrician frames corresponds In this way it can be conceived perfectly to the original impression.
in this case, similar to the original in all details, as
new geometrical or arithmetical truths, we do not need each time renewed perceptions of physical objects; the
that, in order to reach
memory-image, by virtue
of its clarity
itself.
and
distinctness,
is
able to
is
supplant the sensible object once crossed by another.
However,
peculiar
this explanation
at
The
"deductive"
certainty,
which we ascribe to mathematical propositions, is now traced back to the fact that in these propositions we are never concerned with statements about concrete facts, but only with relations between There are no real things which precisely agree hypothetical forms. with the definitions of geometry; there are no points without magni tude, no perfectly straight lines, no circles whose radii are all equal. Moreover, from the standpoint of our experience, not only the actual reality, but the very possibility of such contents must be denied; it is at least excluded by the physical properties of our planet, if not by those of the universe. But psychical existence is denied no less than physical to the objects of geometrical definitions. For in our mind we never find the presentation of a mathematical point, but always only the smallest possible sensible extension; also we never "conceive" a line without breadth, for every psychical image we can
14

lines of a certain breadth. 8 It is evident at
form shows us only
once
that this double explanation destroys itself. On the one hand, all emphasis is laid upon the similarity between mathematical ideas
and the
original impressions;
on the other, however,
it is
seen that
this sort of similarity does
not and cannot
exist, at least for
those
forms which alone are defined and characterized as "concepts" These forms cannot be attained by bare in the mathematical sciences. selection from the facts of nature and presentation, for they possess no concrete correlative in all of these facts. "Abstraction," as it has litherto been understood, does not change the constitution of con sciousness and of objective reality, but merely institutes certain
and divisions in it; it merely divides the parts of the sensempression but adds to it no new datum. In the definitions of pure mathematics, however, as Mill s own explanations show, the world of sensible things and presentations is not so much reproduced as
imits
transformed and supplanted by an order of another sort. If we trace the method of this transformation, certain forms of relation, or rather an ordered system of strictly differentiated intellectual
functions, are revealed, such as cannot
less
even be characterized, much
justified,
is
result
also
by the simple schema of "abstraction." And this confirmed if we turn from the purely mathematical
concepts to those of theoretical physics. For in their origin the same process is shown, and can be followed in detail, of the trans
formation of the concrete sensuous reality, a process which the traditional doctrine cannot justify. These concepts of physics also not intended merely to produce copies of perceptions, but to are
put in place of the sensuous manifold another manifold, which agrees with certain theoretical conditions. 9 The defect of the psychological theory of abstraction. Neglecting the nature of abstract concepts, however, we find that the naive view of
the world, to which the traditional logical conception especially appeals and upon which it rests, conceals within itself what is ulti mately the same problem. The concepts of the manifold species
and genera are supposed to
arise for us
by the gradual predominance
of the similarities of things over their differences, i.e., the similarities alone, by virtue of their many appearances, imprint themselves upon
8
Cf. Mill,
System of Logic, Ed.
7,
London, 1868, Book
II,
Ch.
5,
and Book
III,
9
Ch.
Cf.
24.
more particularly Ch. IV.
15
*
the mind, while the individual differences, which change from case The similarity of to case, fail to attain like fixity and permanence.
things, however, can manifestly only be effective and fruitful, if it is understood and judged as such. That the "unconscious" traces left in us by an earlier perception are like a new impression in point of fact, is irrelevant to the process implied here as long as the elements are not recognized as similar. By this, however, an act of identifica tion is recognized as the foundation of all "abstraction." A charac
function is ascribed to thought, namely, to relate a present content to a past content and to comprehend the two as in some re spect identical. This synthesis, which connects and binds together the two temporally separated conditions, possesses no immediate
teristic
sensible
correlate
in
the contents compared.
According to the
manner and
direction in which the synthesis takes place, the
same
sensuous material can be apprehended under very different concep tual forms. The psychology of abstraction first of all has to postulate
that perceptions can be ordered for logical consideration into "series of similars." Without a process of arranging in series, without running through the different instances, the consciousness of their
and consequently of the abstract object could This transition from member to member, however, manifestly presupposes a principle according to which it takes place, and by which the form of dependence between each member and the
generic connection
arise.
never
succeeding one, is determined. Thus from this point of view also it appears that all construction of concepts is connected with some definite form of construction of series. We say that a sensuous
-
manifold is conceptually apprehended and ordered, when its members do not stand next to one another without relation but proceed from
a definite beginning, according to a fundamental generating rela

tion, in necessary sequence.
It is
the identity of this generating rela
maintained through changes in the particular contents, which constitutes the specific form of the concept. On the other hand, whether from the retention of this identity of relation there finally evolves an abstract object, a general presentation in which similar
tion,
features are united, is merely a psychological side-issue and does not affect the logical characterization of the concept. The appearance
of a general
image of that sort
may
be excluded by the nature of the
generating relation, without the definitive moment in the clear deduction of each element from the preceding being thereby removed.
16
h
I
ij
In this connection, the real weakness of the theory of abstraction is apparent in the one-sidedness of its selection, from the wealth of possible principles of logical order, of merely the principle of similarIn truth, it will be seen that a series of contents in its conceptual ity.
V/l
may be arranged according to the most divergent points of but only provided that the guiding point of view itself is view; maintained unaltered in its qualitative peculiarity^ Thus side by side with series of similars in whose individual members a common element uniformly recurs, we may place series in which between each member and the succeeding member there prevails a certain degree of difference. Thus we can conceive members of series
ordering
ordered according to equality or inequality, number and magnitude,
The relation relations, or causal dependence. thus produced is in each case decisive; the concept is of necessity merely the expression and husk of it, and is not the generic presenta
spatial
and temporal
tion,
which may arise incidentally under special circumstances, but which does not enter as an effective element into the definition of the
concept.
The forms of series. Thus analysis of the theory of abstraction leads back to a deeper problem. The "comparison" of contents, here referred to, is primarily only a vague and ambiguous expression,
which hides the

tive
difficulty of the
problem.
categorical functions are here united
under what
In truth, very different is merely a collec
name.
And
the real task of logical theory with regard to any
definite concept consists in setting forth these functions in their
and in developing their formal aspects. The theory of abstraction obscures this task since it confuses the categori cal forms, upon which rests all definiteness of the content of percep
essential characteristics
with parts of this very content itself. And yet even the most simple psychological reflection shows that the "likeness" between any contents is not itself given as a further content that similarity or
tion,
;
dissimilarity does not appear as a special element of sensation side by ide with colors and tones, with sensations of pressure and touch.
The ordinary schema

calls for
for in
it
of the construction of concepts, therefore, a thorough-going transformation, even in its outer form; the qualities of things and thepurejispect of relation are
and"
^placed on the âme~level"
fused without distinction.
Once
this
it can indeed appear as if the work of were limited to selecting from a series of perceptions aa, a/3, thought
identification has taken place,

In truth, however, the con the possession of a common by a special example of logically possible connections "property" is only The connection of the members is in every case pro in general. duced by some general law of arrangement through which a thorough
ay
the
common
element
a.
nection of the
members
of a series
together is not itself a new element, that was factually blended with them, but it is the rule of progression, which remains the same, no matter in which member it is represented.
. .
going rule of succession of the series a,b,c,
is
established.
.
.
That which binds the elements
which determines the The function F(a, 6), F(b, c), sort of dependence between the successive members, is obviously not to be pointed out as itself a member of the series, which exists
.
and develops according

can thus be
it. The unity of the conceptual content out of the particular elements of its-exten sion only in the sense that it is in connection with them that wo become Conscious of the specific rule, according to which they are
to
"abstracted"
relate*];
but not
in the sense that
we construct
this rule out of
them
through either bare summation or neglect of parts. What lends the theory of abstraction support is merely the circumstance that it does not presuppose the contents, out. of which the .concept is to develop, as disconnected particularities, but that it tacitly thinks them in the form of an ordered manifold from the first. The con
cept,
lf|
^|
we
however, is not deduced thereby, but presupposed; for when ascribe to a manifold an order and connection of elements, we
have already presupposed the concept, if not in its complete form, yet in its fundamental function. The place of the thing-concept in the system of logical relations.
There are two different
lines of consideration in
which
this logical
presupposition is plainly evident. On the one side, it is the category of the whole and its parts on the other, the category of the thing and its attributes, of which application is made in the customary doctrine
;
of the origin of the generic concept. That objects are given as organizations of particular attributes, and that the total groups of such are divided into parts and sub-parts, which are attributes common tfco several of Ithem, is here taken as the self-evident, basic
principle. \
described, \but contrast,

thafe
put we stand here
In truth, tyowever, the "given" is not thereby merely is judgecl and shaped according to a certain conceptual as soonWs this is recognized it must become evident
before a mere beginning that points beyond
18
SUBSTANCE AND FUNC
itself.
The
categorical acts (Akte),
which we characterize by the
concepts of the whole and its parts, and of the thing and its attributes, are not isolated but belong to a system of logical categories, which moreover they by no means exhaust. After we have conceived the
from
plan of this system in a general logical theory of relations, we can, On the other hand, it is this standpoint, determine its details.
not possible to gain a view of all possible, forms of connection from the in the naive view limited standpoint of certain relations emphasized The category of the^thingyehows itself unsuited for of the world.
purpose in the very fact that we have in pure mathematics _a field of knowledge, in which things and their properties are disre garded in principle, ajidTn^hose fundamental concepts therefore, no general property of things can be contained.
this
\
The negative process of
m
At
this point, a
"abstraction."
new and
more general
difficulty
arises to
threaten the traditional logical
If we merely follow the traditional rule for passing from doctrine. / the particular to the universal, we reach the paradoxical result that / thought, in so far as it mounts from lower to higher and more inclu-
moves in mere negations. The essential act here that we drop certain determinations, which we had hitherto held; that we abstract from them and exclude them from consideration as irrelevant. What enables the mind to form con
sive
concepts,
presupposed
is
cepts is just its fortunate gift of forgetfulness, its inability to grasp the individual differences everywhere present in the particular
If all the memory images, which remained with us from cases. previous experiences, were fully determinate, if they recalled the vanished content of consciousness in its full, concrete and living
nature, they would never be taken as completely similar to the new impression and would thus not blend into a unity with the latter.
Only the inexactness of reproduction, which never retains the whole renders possible __pf the earlier impression but merely its hazy outline,
this unification of elements- that are in themselves dissimilar.
all
Thus
formation of concepts begins with the substitution of a generalized image for the individual sensuous intuition, and in place of the actual 10 perception the substitution of its imperfect and faded remainder.
10 Cf., Sigwart, Logik, Ed. 2, p. 50 f., also, H. Maier, Psychologic emotionalen Denkens, Tubingen, 1908, pp. 168 ff.
des

If
19
strictly to this conception, we reach the strange result the logical labor which we apply to a given sensuous intuition Instead of reach serves only to separate us more and more from it. ing a deeper comprehension of its import and structure, we reach
we adhere
all
that
only a superficial schema from which lar case have vanished.
all
peculiar traits of the particu
The mathematical concept and its "concrete universality." But from any such conclusion we are once more safeguarded by considera tion of that science in which conceptual definiteness and clarity have
reached their highest
level.
It is at this point, indeed, that the
mathematical concept appears most sharply distinguished from the

In the methodological struggle over the limits ontological concept. of mathematics and ontology, which took place in the philosophy of the eighteenth century, this relation was incidentally given happy
and
In his criticism of the logic of the v it was the exclusive Wolffian school, merit of mathematical "general concepts" not to cancel the determinations of the special cases, but in all strictness fully to retain them. When a mathematician makes his formula more general,
significant
expression.
Lambert pointed out that
this
means not only that he is to retain all the more special cases, but also be able to deduce them from the universal formula. The possibility of deduction is not found in the case of the scholastic concepts, since these, according to the traditional formula, are formed by neglecting the particular, and hence the reproduction of the par
ticular
./
seems excluded. Thus abstraction "philosopher," but on the other hand, the deter mination of the particular from the universal so much the more difficult; for in the process of abstraction he leaves behind all the particularities in such a way that he cannot recover them, much less reckon the transformations of which they are capable. 11 This remark contains, in fact, the germ of a distinction of great simple consequence. The ideal of a scientific concept here appears in opposi tion to the schematic general presentation which is expressed by a
moments
of the concept
is
very easy for the
mere word. The genuine concept does not disregard the peculiarities and particularities which it holds under it, but seeks to show the necessity of the occurrence and connection of just these particularities.
Lambert, Anlage zur Architektonik oder Theorie des Einfachen und des Erslen inderphUosophischenundmathematischenErkenntnis, Riga, 1771, 193 ff. Cf. my Erkenntnisproblem in der Philosophic und Wissenschaft der neuern Zeit, Vol. II, p. 422 f.
11
20

it
What
gives
is
a universal rule for the connection of the particulars
themselves.
Thus we can proceed from a general mathematical

from the formula of a curve
of the second
formula,
for example,
to the special geometrical forms of the circle, the ellipse, order, etc., by considering a certain parameter which occurs in them and
permitting it to vary through a continuous series of magnitudes. Here the more universal concept shows itself also the more rich in content; whoever has it can deduce from it all the mathematical relations which concern the special problems, while, on the other hand, he takes these problems not as isolated but as in continuous con
nection with each other, thus in their deeper systematic connections. The individual case is not excluded from consideration, but is fixed
and retained
change.
It is evident
is
as a perfectly determinate step in a general process of anew that the characteristic feature of the
"universality"
concept
not the
of
a;
presentation, hnf
tfoft
universal
alidity of a principle of serial order..
We do not isolate any abstracT

of
part whatever from the manifold before us, but
members a
definite relation
by thinking
them
as
we create for its bound together
by an inclusive law. And the further we proceed in this and the more firmly this connection according to laws is established, so much
the clearer does the unambiguous determination of the particular stand forth. Thus, for example, the intuition of our Euclidian
three-dimensional space only gains in clear comprehension when, in modern geometry, we ascend to the "higher" forms of space; for in
this
way
the total axiomatic structure of our space
is first
revealed in
full distinctness.
The
criticism of the theories of abstraction.
Modern
expositions
have attempted to take account of this circumstance by opposing, in accordance with a well-known distinction of Hegel s,
of logic
the abstract universality of the concept to the concrete universality of the mathematical formula. Abstract universality belongs to the
genus in so fat
if
as,
considered in and for
itself, it
neglects all specific
fercnces; concrete universality,
systematic whole (Gesamtbetfrrfi") liarities of all the species and develops

en, e.g., algebra solves the
on the contrary, belongs to the which takes up into itself the pecu
them according
to a rule,
problem of finding two whole numbers, whose sum is equal to 25, and of which one is divisible by 2 and the other by 3, by expressing the second by the formula 6z + 3, in which z can only have the values 0, 1, 2, 3, and from which of itself 22-6z follows as a formula of the first, these formulae possess concrete

universality.
They
z
is
which determines
because,
all
are universal because they represent the law the numbers sought; they are also concrete
when
given the four above-mentioned values, the
numbers sought for follow from these formulae as species of them. The same is true in general of every mathematical function of one or more variables. Every mathematical function represents a univer sal law, which, by virtue of the successive values which the variable can assume, contains within itself all the particular cases for which it
once recognized, a completely new In opposition to the logic opened investigation of the generic concept, which, as we saw, represents the point of view and influence of the concept of substance, there now appears However, the field of logic of the mathematical concept of function. of this form of logic is not confined to mathematics alone. application
holds."
12
If,
however, this
is
is
field of
for logic.
the contrary, it extends over into the field of the knowledge of nature; for the concept of function constitutes the general schema
On
and model according to which the modern concept molded in its progressive historical development.
Before
of nature has
been
to trace the construction of functional conceps and thus to verify our new conception of the concept by concrete examples, we may indicate the meaning of the problem by citing a characteristic turn recently taken by the theory of ab straction. Everywhere a new motive is apparent, which, if systemati cally developed and carried through, will raise logical questions that
in science itself
we proceed
extend beyond the traditional point of view. An indication of this motive is to be found in the first place in Lotze s skeptical comments
of abstraction. As he explains, the real practice of thought in the formation of concepts does not follow the course prescribed by this doctrine; for it is never satisfied to advance
on the traditional doctrine
to the universal concept by neglecting the particular properties without retaining an equivalent for them. When we form the concept of metal by connecting gold, silver, copper and lead, we cannot indeed
ascribe to the abstract object that thus comes into being the particu lar color of gold, or the particular luster of silver, or the weight of
sible
copper, or the density of lead however, it would be no less inadmis if we simply attempted to deny all these particular determina
;
tions of
it.
tion of metal, that

12
For the idea obviously does not suffice as a characteriza it is neither red nor yellow, neither of this or that
p. 22.
Drobisch, Neue Darstellung der Logik,
22
hardness or resisting power; but the positive thought must be added that it is colored in some way in every case, that it is of some degree of hardness, density and luster. And analogously, we would not retain the general concept of animal, if we abandoned in it all thought of the aspects of procreation, of
specific weight, neither of this or that
movement and
of respiration, because there is no form of procreation, of breathing, etc., which can be pointed out as common to all animals. It is not, therefore, the simple neglect of the "marks" pip2, qiQ2,
that are different in the different species, which is the rule of abstrac tion but always, in the place of the neglected particular determina
;
"marks" P and Q must be set up, the particular which are pip2 and qiQ2. The merely negative procedure, on the contrary, would lead in the end to the denial of all determina tion, so that our thought would find no way of return from the logical 13 We see here how "nothing" which the concept would then signify. Lotze, on the basis of psychological considerations, approaches the problem which Lambert had clearly and definitely formulated, using the example of the mathematical concept. If we carry through the
tions, the general
species of
jabove
lar
rule to the end,
it
obliges us to retain, in place of the particu
^4
;
which are neglected in the formation of the concept, the systematic totality (Inbegriff) to which those marks belong as We can abstract from the particular color special determinations. only if we retain the total series of colors in general as a fundamental schema, with respect to which we consider the concept determined,
"marks"
represent this systematic totality substitute for the constant particular "marks," variable terms, such as stand for the total group of possible values which the different "marks" can assume. Thus it becomes evident
diich
we
are
forming.
We
(Iribegriff]
when we
is only in a purely negative process. In truth, what seems to be appearance cancelled in this way is maintained in another form and under a differ ent logical category. As long as we believe that all determinateness
that the falling aside of the particular determinations
consists in constant
"marks"
in things
and
their attributes, every
process of logical generalization must indeed appear an impoverish ment of the conceptual content. But precisely to the extent that
the concept is freed of all thing-like being,ots peculiar functional character is revealed. Fixed properties are replaced by universal rules that permit us to survey a total series of possible determinations
18
Lotze, Logik, Ed.
2,
Leipzig, 1880, p. 40
f.
23
This transformation, this change into a new form at a single glance. ofjpgical being, constitutes the real positive achievement of abstrac
proceed from a series aai/3i, 00:2/82, aa^ 3 .... constitutive a, but replace the totality of individual members a by a variable expression x, the totality of
tion.
We
do not
directly to their
common
individual members by a variable expression y. In this way we unify the whole system in the expression a x y ... which can be changed into the concrete totality (Allheit) of the members of the
by a continuous transformation, and which therefore perfectly represents the structure and logical divisions of the concept. and "second" orders. This turn of thought Objects of the
series
"first"
can be traced even in those expositions of logic that, in fundamental tendency, retain the traditional theory of abstraction. It is signifi
cant of this tendency, for instance, that Erdmann, after completing
his psychological theory of the concept, finds himself forced by his consideration of the mathematical manifold to the introduction of a
point of view and a new terminology. The first phase of every construction of concepts, he now teaches, does indeed involve the separating out of a certain universal on the basis of the uniformity
new
with which
*
its
uniformity of the given,
content recurs amid varying particulars: but^thig.. though perhaps the original, is not the sole
condition which enables us to mark off the objects of our presentations. In the progress of thought, the consciousness of uniformity is rather
supplemented by the consciousness of necessary connection; and this supplementation goes so far that ultimately we are not dependent upon number of repetitions to establish a concept. "Wherever in developed presentation a composite object is found in our perception, which takes its place as a well-defined member of a series of presenta tions, such as a new shade in the series of colors, a new chemical
compound
in a series of
known compounds
of similar constitution,
there a single occurrence suffices to fix it in its definite character as a member of the series, even in case we never perceive it again." 14
In contrast to objects of sense-perception, which we can designate as "objects of the first order," there now appearj/objeets of the second
whose logical character is determined solely by the form of connection from which they proceed. In general, wherever we unify the objects of our thought into a single object, we create a new
order,"
"object
of the second
order,"
whose total content
is
expressed in the
14
B. Erdmann, Logik, Ed.
2, p. 158. f .
24
relations established
unification.

between the individual elements by the act of
This type of thought, to which Erdmann declares he was led by the problems of the modern theory of groups, breaks through the old schema of the formation of concepts; for instead of the community of "marks," the unification of elements in a concept And this criterion, is decided by their "connection by implication." here only introduced by way of supplement and as a secondary aspect,
proves on closer analysis to be the real logical prius; for we have already seen that "abstraction" remains aimless and unmeaning if it
does not consider the elements from which
it takes the concept to be from the first arranged and connected by a certain relation. The variety of objective "intentions" In general, as the purely logical aspect of the concepts of relation and of the manifold becomes If clearer, a greater need is felt for a new psychological foundation. the objects with which pure logic deals are not identical with the
individual contents of perception, but possess their own structure and "essence," then the question must arise as to how this peculiar
comes to our consciousness and by what acts it is grasped. the mere sensuous experiences, however much we heap them up and however much we complicate them, can never suffice for this purpose. For sensuous experience is concerned
"essence"
It is clear that
exclusively with a particular object or with a plurality of such objects; no summation of individual cases can ever produce the specific unity
which is meant in the concept. The theory of attention, therefore, as the truly creative faculty in the formation of concepts, loses all
application in a deeper phenomenology of the pure thought processes. For attention only separates or connects elements already given in
can give these elements no new meaning and invest It is such a change of function, logical function. however, which first transforms the contents of perception and pres
perception;
it
them with no new
entation into concepts in the logical sense. From the standpoint of purely descriptive analysis of conscious process also, it is something
different
when
for
as
its
when
grasp this or that particular property of a thing, example I select from the perceptual complex of a house
I
"the"
There
red as a species. special red color, than when I contemplate is a difference between making valid mathematical judgments
"four,"
concerning the number thereby placing it in an objective connection of relations, and directing consciousness upon a con crete group of things or presentations of four elements. The logical

determination of
"four"
25
(in the first case) is given by its place in an and therefore timelessly valid whole of relations, by its place in a mathematically defined number-system; but sensuous presentation, which is necessarily limited to a particular here and now, is unable Here the psychology of to reproduce this form of determination. thought strives to make a new advance. By the side of what the content is in its material, sensuous structure, there appears what it means in the system of knowledge; and thus, its meaning develops out of the various logical which can be attached to the content. which differentiate the sensuously unitary content by These
ideal
"acts"
"acts,"
it different objectively directed "intentions," are psychologically completely underived; they are peculiar forms of consciousness, such as cannot be reduced to the consciousness of
imprinting upon
If we are still to speak of abstraction as sensation or perception. that to which the concept owes its being, nevertheless its meaning is
now
from that of the customary sensationalistic doc no longer a uniform and undifferentiated attention to a given content, but the intelligent accomplishment of the most diversified and mutually independent acts of thought, each of which involves a particular sort of meaning of the content, a
totally different
trine; for abstraction is
15 special direction of objective reference.
The
serial
form and
the
members
of the series.
Thus the
circle of
complete, since we are led, from the side of "subjective" analysis, from the pure phenomenology of consciousness, to the same fundamental distinction, the validity of which has been shown in the
our subject
is
"objective"
logical
investigation.
While the empiristic doctrine
regards the "similarity" of certain contents of presentation as a selfevident psychological fact which it applies in explaining the forma tion of concepts, it is justly pointed out in opposition that the similarity ot-eeftain elements can only be spoken of significantly when
been established from which the^,eleThis identity of refer ence, of point of view, under which the comparison takes place, is, however, something distinctive and new as regards the compared contents themselves. The difference between these contents, on the one hand, and the conceptual "species," on the other, by which we unify them, isjyj_iixeducible fact; it is categorical and belongs to the
a certain
QK>int
of
view"/nas
mnts-can be designated
as like or unlike.
15
On
this
No.
II, Halle, 1901
whole subject cf. Husserl, Logische Untersuchungen, Vol. II, Die ideale Einheit der Species und die neuern Abstractions:
theorien.
26

of consciousness."
"form
In fact,
it
is
new
characteristic contrast between the

\
member
of the series
expression of the and the
form
The content of the concept cannot be dissolved of the series. into the elements of its extension, because the two do not lie on the same plane but belong in principle to different dimensions. The
meaning
of the law that connects the individual
members
is
not to
be exhausted by the enumeration of any number of instances of the law; for such enumeration lacks the generating principle that enables
us to connect the individual members into a functional whole.! If I know the relation according to which a b c are ordered,
.
I can deduce
them by reflection and
isolate
them as objects of thought;
it is impossible, on the other hand, to discover the special character of the connecting relation from the mere juxtaposition of a, b, c in In this conception there is no presentation. (Cf. above pp. 16 ff .)
danger of hypostasizing the pure concept, of giving it an independent reality along with the particular things. The serial f orm F(a, b,c. ) which connects the members of a manifold obviously cannot be
.
.
thought after the fashion of an individual a or 6 or

losing its peculiar character.
Its
"being"
c,
without thereby
in
consists exclusively
.; and this determination can only be expressed by a synthetic act of definition, and not by a simple sensuous intuition. These considerations indicate the direction of the following inves
. .
.
the logical determination by which other possible serial forms $, ^
it is
clearly differentiated
from
tigation.
The
totality
and order
of pure
"serial
forms"
lies
before
us in the system of the sciences, especially in the structure of exact science. Here, therefore, the theory finds a rich and fruitful field,
which can be investigated with respect to its logical import inde pendently of any metaphysical or psychological presuppositions as to the "nature" of the concept. This independence of pure logic, however, does not mean its isolation within the system of philosophy.
at the evolution of "formal" logic would show the dogmatic inflexibility of the traditional forms begins to And the new form that is beginning to take shape, i.; also a yield. form for a new content. Psychology and criticism of knowledge,
Even a hasty glance
how
the problem of consciousness and the problem of reality, both take part in this process. For in fundamental problems there are no absolute divisions and limits; every transformation of the genuinely
"formal"
that
is
characterized
concept produces a new interpretation of the whole and ordered by it.
field
CHAPTER
II
Among
It is in
number stands
the fundamental concepts of pure science the concept of in the first place, both historically and systematically.
it
connection with
that consciousness of the meaning and value
In the thought of of the formation of concepts first develops. number all the power of knowledge seems contained, all possibility If there were no of the logical determination of the sensuous.
number, nothing could be understood

or in their relations to each other.
in things, either in themselves
remains unchanged in
philosophical thought.
in
its
real import
This Pythagorean doctrine through all the changes in
The claim to grasp the substance of things number has indeed been gradually withdrawn; but at the same time the insight has been deepened and clarified that in number is rooted the substance of rational knowledge. Even when the meta
physical kernel of the object
number remains the

in general.
first
is no longer seen in it, the concept of and truest expression of rational method
In
it
are directly mirrored the differences in principle
between the fundamental interpretations of knowledge. Through number the general ideal of knowledge gains a more definite form, in which for the first time it is defined with full clarity. The sensationalistic deduction of number. Thus it is quite intelligible that we should meet upon the threshold of algebra the same typical opposition that was traceable in the field of logic. If we accept the traditional logical view, we should expect to find certain fundamental The properties of objects revealed in the numerical concepts. theory of abstraction provides, strictly speaking, for no other point of view. Just as objects are differentiated according to size and form, according to smell and taste, so also, on the theory of abstrac tion, they must have a certain property which gives them their numerical character. The concept of or would be abstracted from a plurality of objective groups, just as the concept of a certain color arises from the comparison of colored perceptual It is consistent, from this standpoint, to regard all asserthings.
"two"
"three"
27
28
tions concerning
numbers and numerical
relationships as expressive
of certain physical properties of objects.
In the modern develop
ment come
2
of empiricism, this latent consequence has for the first time fully
to light. Thus, according to J. S. Mill, the proposition that = 3 represents no mere definition, no mere fixation of the
meaning which we are to connect with the concepts of two and three, but it reports an empirical matter of fact which our spatial perception has hitherto always presented in the same way. We have always been able when we have seen three things before us in a certain arrangement, for example, in the form o%, to analyse them into Three pebbles do not make the same partial groups of the sort 00, 0. impression on our senses when they lie before us in two separate Hence the asser piles as when they are collected into a single pile. tion, that the perception that arises in the first case can always be transformed through mere spatial re-arrangement of its parts into the second perception, is no mere identical proposition that says noth ing, but an inductive truth learned through early experience and which has since been continually confirmed. Such truths constitute the foundation of the science of number. The appearance of ideality
that attaches to this science must, therefore, disappear. The propo sitions of arithmetic lose their former exceptional position; they
come to be on the same plane as other physical observations that we have made concerning separations and combinations in the world For how can there be significant and valid judgments of bodies. that have no reference to sensible facts? The concept of ten either means nothing or it means a certain uniform total impression that is
always found in groups of ten bodies, ten tones or ten pulse-beats. And that the various impressions thus gained from objects form a system among themselves, in which certain constant relations prevail,
likewise a proposition possessing merely empirical certainty. reality of another sort, a new physical environment into which
is
A
we
= 5 just as were thrust, could make the proposition that familiar and self-evident to us as it now seems unintelligible and
absurd. 1
2X2
foundations of arithmetic. With this first step into the exact scientific problems, we can already see very clearly what real meaning and importance may be contained in what appear
Frege
s
field of
Cf. Mill,
s
Hamilton
System of Logic, Bk. II, Ch. 6; Philosophy, pp. 67 ff.
An
Examination of Sir William

TO
29
be purely formal logical differences. For, however we may judge Mill s theory of the fundamental arithmetical principles, we must recognize that it follows with convincing necessity from his
is it
general interpretation of the concept. So much the more significant that the theory, when carried through, leads to a direct conflict
Wherever an attempt has itself. been made in modern mathematics to analyse and explain the fact of scientific arithmetic, the logical structure of pure number has been sharply distinguished from Mill s arithmetic of "pebbles and ginger bread nuts," thereby guarding against a possible illusion. Indeed, if Mill s deduction were correct, the arithmetical concepts would be deprived of the very determinateness which constitutes their peculiar value and import. The logical differences of numbers would be
with the fact of scientific arithmetic
tion
limited and restricted to whatever psychological power of differentia we had attained in the apprehension of given groups of objects.
The absurdity of this consequence can, however, be easily seen. The number 753,684 is just as definitely and clearly differentiated from the number that immediately precedes or follows it as three is from two or four; but who could point out the "impression" which
differentiates the sensuous intuitions of the
each other?
And
in the
two concrete groups from same way that the characteristic content of
the numerical concepts is lost here, so, on the other hand, we lose The the scope and freedom of application that is essential to it. of numbering can only take place, according to Mill, synthesis
where the combining and separating instituted by it can actually be carried out with physical objects, where things themselves can be The vary collected and separated into perceptible spatial groups. which arise in us from the different groups, constitute ing images, the real and indispensable basis of all assertions concerning numerical
Hence beyond the field of spatial sensuous intuition, wherein these actual combinations and separations alone are possible, the real foundation of the numerical concepts would be lacking. But in truth, we speak not merely of the number of seeds in a pile,
relations.
but of the number of categories, of the number of Kepler s laws, or of the number of factors of energy: all objects which cannot be
arranged side by side and separated from each other like pebbles. would indeed be strange," remarks Frege, in his drastic and
"It
pertinent criticism of Mill s doctrine," if a property abstracted from outer things could be carried over without change of meaning to
30
It would be precisely as experiences, presentations and concepts. if one were to speak of a fusible experience, of a blue presentation, It is absurd that what is of a salty concept or a sticky judgment.
by nature
sible.
sensible should present itself in connection with the insen
see a blue surface, we have a peculiar impression which corresponds to the word blue; and this we recognize again when we look at another blue surface. If we assume that in the same way in looking at a triangle there is something sensuous which then this sensuous element must corresponds to the word also be found in three concepts something non-sensuous would have something sensuous as a property. It can be granted that there is a
When we
"three"
kind of sensuous impression corresponding to the word "triangular" We do not see the three in it in case we take the word as a whole.
immediately, but we see something to which an intellectual activity can attach, which leads to a judgment in which the number 3
2
appears."
of arithmetic. If the absurdities inevitably implied in the sensationalistic interpretation of number do not come to light directly in the first deduction, the reason lies in the fact that intellectual
The system
activities, processes of
judgment, are not entirely excluded but are
According to this theory, only the first truths of tacitly assumed. the most elementary formulae, are to be the result arithmetic, only of the immediate observation of physical facts, while the scientific system of algebra is not to rest upon the continually renewed influx of facts of perception, but upon a "generalization" of the original sensuous facts. But this conception involves all the riddles which the theory promised to solve. When we attempt to give such a
conception clear and definite meaning, we find that it directly implies a plurality of different intellectual functions which partake in the construction of number. If it is to be possible to carry over observa tions that we have made with smaller complexes of objects progres sively to larger and ever larger complexes, and to determine the
assume that some form
properties of the latter after the analogy of the former, we must of connection and dependence exists between
the compared cases, by virtue of which the one can be deduced from the other. We would not have the right to extend any determina
tion,
2
which appears to us in any individual group, to groups of a

Breslau, 1884, p. 31
f.
matter
Frege, Die Grundlagen der Arithmetik. cf. especially pp. 9 ff., 27 ff.
For the whole
31
larger or smaller number of elements, if we did not comprehend them all as similar in "nature." This similarity, however, means nothing more than that they are connected by a definite rule, such as permits
us to proceed from one manifold to another by continued identical application of the same fundamental relation. In fact, without the
assumption of such a connection we should have to be prepared for

the possibility that any unity, added to or substracted from a given group, would alter the total character of the group so that no infer
ence would be possible from the relations of one group to those of any other. The new unities would then act like as many physical
conditions or forces, which could completely transform the whole and cancel its fundamental characteristics. No generally applicable laws, no thorough-going relation would bind together the members of the realm of number; on the contrary, every arithmetical proposition would have to be specially proved by observation or perception for every individual number. The sensualistic theory is only able to avoid this consequence by an unnoticed deviation into another line
The demand for a generalization of primitive experiences of number contains, although concealed, that very function of the universality of number concepts, which was to have been set aside by the explanation. The road to a purely deductive
of consideration.
construction of the realm of
the insight
suffices,
number is thereby reopened; for this that the same intellectual procedure, which is
indispensable to every theory in proceeding to the higher arithmeti cal forms, constitutes the necessary and adequate foundation for the
determination of the elements themselves. In this consequence, which the sensualistic theory finally admits against its will, is offered the first view of a unified methodical deduction, deriving both its foundation and superstructure from a common principle. Number and "presentation." At first, however, there seems to be another way to establish the desired relation between numerical propositions and the empirical existence of things. If we renounce the view that all arithmetical judgments are directed upon physical things and dependent upon them for their validity, there remains, nevertheless, another class of realities, in which we now seem to grasp the true original of the numerical concepts. Not external things, but "consciousness" itself in its peculiar and irreducible mode of being, is the source of these concepts; what they seek to encompass and represent is not a material but a mental being. The scope and
32
universality of the numerical concepts seems here to be explained. Number, as presentation, as psychical reality, remains free from all the limitations that had to be placed upon it as long as it was taken
as an expression of particular material existences
and
their relations.
We
can recognize
how
here, in connection with a special problem,
the same intellectual transformation takes place that we met earlier The attempt to have the in the field of general logical theory.
concept copy outer reality directly

this outer reality there appears its
abandoned; but in place of phenomenal form in our mind.

is
act of enumeration does not give the relations of things in themselves, but only the way in which they are reflected in the com
The
prehension of our ego.
However much
there
still
this transformation
it
may advance
the problem,
remains an element which
shares with the sensationalistic
deduction.
The
doctrine of
independent logical
p.
number here again fails to attain an foundation; it now forms an appendix to psychol
ogy, just as formerly
last
it was a special case of physics. (Cf. above For psychology, however, "presentation" means in the ff.). analysis nothing else than a definite psychical content, which
arises in the individual subject according to special circumstances,
and can be destroyed again
in the
same way.
Such a content
is
different in different individuals and, furthermore, with respect to the
same subject, once it has disappeared, never returns in precisely the same form. Thus what is here given is always only a temporally limited and determined reality, not a state which can be retained in
unchanging logical identity. It is the fulfillment of the demand for this latter, however, which constitutes all the meaning and value of the pure numerical concepts. The proposition that 7 + 5 = 12 reports no connection of presentational experiences, either as it has
occurred in the past or will occur in the future in thinking individuals; but it establishes a connection which, according to the Platonic
expression, binds the seven
and the
five in
themselves with the
twelve in
object upon which this judgment is directed has, in spite of its ideality, a fixity and definiteness which sharply differentiates it from the changing contents of presentation. The
itself.
The
psychological image of two may, in the case 6f one person, be con nected with an accompanying spatial presentation and, in the case of another, be without it; it may now be vividly grasped, now dimly;
nevertheless, the arithmetical
meaning
of
two
is
not affected by
all

\
3
"is"
33
and means can only be What the concept these differences. ascertained through understanding it as the bearer and starting-point \ of certain judgments, as a totality of possible relations. Concepts are identical when they can be substituted for each other in all the \
assertions into which they enter; when every relation which holds of the one can also be carried over to the other. If we apply this crite rion, the whole difference between the logical meaning of the concept
of
once revealed.
number and the psychological conception of presentation is at The characteristic relations which prevail in the
series of numbers are not thinkable as properties of the given contents of presentation. Of a presentation it is meaningless to say that one is larger or smaller than another, the double or triple of it, that one
is
divisible
by another,
etc.
And no
less
does the postulate of an
infinity of numbers exclude any such conception, for all the "being" of a presentation is exhausted in its immediate givenness, in its
actual occurrence.
If
the numbers are realities in individual con

"given"
sciousness, they can only be
in finite groups,
i.e.,
realized in
this consciousness as particular elements.
The content of presentation and the act of presentation. Yet in establishing this opposition between the pure numerical concepts
and the psychological contents

not seem to have grasped the
significance
is
of presentation, this criticism does field of psychical being in its full
and scope. It may with justice be objected that what characteristic of number cannot be pointed out in any particular
isolated content of consciousness only because there
is
and
here a
universal presupposition, which controls and guides the origin formulation of contents in general. The act, by which we define
and any
unity and the synthesis by which we join together such unities into new forms, constitute the only condition under which we can speak
and their connection. The activity of and connection alone, not any particular content subsequently resulting from it, can be the desired psychological
of a manifold of elements
differentiation
correlate of the numerical concepts. It is not with objects, either those of inner or outer reality, but with acts of apperception, that the numerical determination is connected, and to which it goes back
The "universality" of the pure numerical concepts can thus be understood and grounded from a new stand point. Even sensationalism recognizes this universality; but it underfor its real meaning.
3
Cf. again Frege, op.
cit., p.
37.
34
stands

it,
its fundamental theory, as a thing-like "mark," uniformly found in a group of particular objects. "All numbers," says Mill, "must be numbers of something: there are no
according to
which
is
such things as numbers in the abstract. But though numbers must be numbers of something, they may be numbers of anything. Propo sitions, therefore, concerning numbers, have the remarkable peculiar
they are propositions concerning all things whatever; all existences of every kind, known to our experience." 4 objects, The mathematical property of the enumerability of things is thus
ity that
all
ascertained here in the

as
all
same manner as any physical property.
Just
we
learn
by thorough-going comparison
of individual cases that
bodies are heavy, so, by an analogous method, we discover their numerical determinateness. We recognize, however, that the asser
tion of the universality of number, in so far as it rests upon such a procedure as this, has in truth been gained surreptitiously; for noth
ing assures us that the cases beyond our experience show the same properties as those actually observed, and thus fall under the arith metical laws. A new standpoint regarding the foundation of number
reached through the deeper and more mature psychological deduction of the numerical concepts from the fundamental act of From this apperceptive connection and separation in general. number is to be called universal not because it is con standpoint,
is first
tained as a fixed property in every individual, but because it repre sents a constant condition of judgment concerning every individual
as
an individual.
The
consciousness of this universality

indefinite plurality of cases,
;
is
not gained
is already presupposed in the apprehension of every one of them for the arrange ment of these individuals into an inclusive whole is only rendered possible by the fact that thought is in a position to recognize and
by running through an
but
maintain a
rule, in conceptual identity, in spite of all differences peculiarities of application.
and
In this attempted deduction also, which goes back from the finished contents of presentation to the acts by which they are formed, the real logical problem of number is not so much solved as rather pushed back a step. For whatever constructive value we ascribe to the
pure acts of thought, they remain, nevertheless, in their purely psychological sense, always occurrences which come and go in time. They thus belong to a certain individual stream of consciousness as
4
Mill,
System of Logic, Bk.
II,
Ch.
6, 2.

it
35
of
runs
off
here
and now under the particular conditions
Here, however, the earlier question recurs. In the arithmetical judgments it is not the relation of temporally limited
the
realities which is expressed and established; here thought reaches out beyond the whole field of thought-processes to a realm of ideal objects, to which it ascribes a permanent and unchanging form.
moment.
fundamental form that every element of the connected with every other according to a fixed systematic rule. But a psychological analysis of the acts of forming presentations cannot disclose how one is connected with two, or two with three, and how the entire logical complex of propositions con
It is
by virtue
of this
is
numerical series
tained in pure arithmetic arises according to this connection. The construction and objective foundation of this systematic connection belongs to a totally different method. (Cf. below, especially Ch.
In the beginning, indeed, this method is a mere postulate, VIII.) the fulfillment of which must appear entirely problematical. For what way remains of grounding a concept, if we are to regard it
neither as a copy of inner nor of outer being, neither of the psychical or the physical? This question, however, which constantly presses
to the fore,
is only an expression of a certain dogmatic view of the nature and function of the concept. The system of arithmetical
concepts and propositions is not to be estimated in terms of this view; but considerations of formal logic, on the contrary, find a limit and standard in this system, which has gradually evolved out
of its independent
and immanent presuppositions.

ii
The
logical
foundations of the pure concept of number.
(Dedekind.)
The development of scientific arithmetic in the last decades is charac terized by the increasing demand for the deduction of the concept of number, in its full import, from purely logical premises. The science
of space
perception.
all
intuition, or perhaps even to empirical the other hand, the thought gains acceptance that determinations of number are to be grounded, without any appeal
seems to belong to
On
any dependence upon concrete measurable magnitudes, "by a finite system of simple steps of thought." In this deduction of arithmetic out of logic, however, the latter is pre
to sensible objects or
supposed
in
a new form.
"If
we
trace
exactly,"
Dedekind says
in

beginning his deduction of the concept of number, "what we do in counting a group or collection of things, we are led to the considera tion of the power of the mind to relate thing to thing, to let one thing
general without which thought
"
correspond to another thing, one thing copy another, a capacity in is impossible. Upon this one, but
absolutely inevitable, foundation the whole science of number must 5 be erected The starting-point here seems to be the traditional logical doctrine of a plurality of things and the power of
the mind to copy them; nevertheless it becomes evident with deeper understanding that the old terms have here gained a new import and
which are spoken of in the further deduc assumed as independent existences present anterior to any relation, but they gain their whole being, so far as it comes within the scope of the arithmetician, first in and with the relations which are predicated of them. Such are terms of relations, and as such can never be in isolation but only in ideal
meaning.
"things,"
The
tion, are not
"things"
"given"
community with each other. The procedure of "copying" has also mdergone a characteristic transformation. For we are no longer
concerned to produce a conceptual copy of outer impressions, such as
correspond to them in some particular feature; for copying means nothing else than the intellectual arranging by which we bind otherwise totally diverse elements into a systematic unity.
shall
Here the question concerns merely the unification of members of a series by a serial principle, not their agreement in some factual con stitutive part. After a certain starting-point has been fixed by an original assumption, all further elements are given by the fact that a relation (R) is given, which in continued application generates all the members of the complex. Thus arise systems and groups of
systems in stj^"ctfrieptual division without its being necessary that one elemenjcbe connected with another by any sort of factual similarThe xo^ying does not produce a new thing, but a new necesity.
"
operations of thought and objects of thought. In his work, Was sind und was sollen die Zafden, Dedekind has shown how the complete construction of arith metic and the exhaustive exposition of its scientific content is possible
sary order
among
The
logic of relations.
on the basis of
tiiese
simple principles.
We
shall not trace the

details,
2,
mathematical dev&opment of this thought

6
in its
but shall
Dedekind, Was siyd und was sollen die Zahlen?
Ed.
Braunschweig,
1893, p. VIII.

content ourselves merely with emphasizing its essential tendency, much as we are not interested in the concept of number for its
in as
own sake but only as an example of the structure of a pure "functional The presuppositions of the deduction of the concept of concept." number are given in the general logic of relations. If we consider
the totality of possible relationsâcconling^toVEicn a series of intellec tual constructions (Denksetzungeri) can be arranged, there emerge in the
first
place certain fundamental formal determinations, such as
uniformly belong to certain classes of relations and differentiate them from other classes of different structure. Thus if any relation
between two members a and b is given, which we can symbolically represent by the expression aRb, it can in the first place be so con stituted that it also holds in the same way between b and a, so that
from the validity of aRb follows that of bRa. In this case we call the relation "symmetrical" and distinguish it, on the one hand, from the non-symmetrical relations, in which the validity of aRb indeed permits that of bRa but does not necessarily imply it, and on the other hand, from the asymmetrical relations, in which this sort of reversal is not possibfe, in which therefore aRb and bRa cannot
exist together.
relation, furthermore, is called transitive

it
when
and
c,
from the fact that

its
holds of each pair of
members a and
6, b
be non-transitive when this extension is not necessary, and intransitive when it is excluded 6 These determina by the nature of the relation under consideration. which have far-reaching application in the calculus of rela tions,
validity follows for a
and
c.
It is said to
/
*
come in for consideration here chiefly in so far as upon them more exact definition of what we are to understand as the order of a given whole. It is, in fact, a na ive prejudice to regard the order which holds between the members of a manifold as something
tions,
rests the
self-evident, as if it were immediately given through the bare exist ence of the individual members. In truth, it is not attached to the elements as such but to the serial relation by which they are con
nected,
and
all its
determinate character and specific peculiarity are
8 Russell, to whom these distinctions are due, illustrates them by different family relationships; the relation involved in "brothers and sisters" (Geschwister) is symmetrical and transitive; the relation "brother" is nonsymmetrical and transitive; the relation "father" asymmetrical and intransi
tive, etc.
I,
On this and what
follows see Russell, The Principles of Mathematics,
Cambridge, 1903; cf. also Kant Studien XII, p. 1 ff.
my
essay
"Kant
und
die
moderne
Mathematik,"
38

Closer investigation shows that
is it
derived from this serial relation.

is
always some transitive and asymmetrical relation that to imprint upon the members of a whole a determinate
a
necessary
order. 7
The concept of progression. If, now, we consider a series which has first member, and for which a certain law of progress has been established, of such a sort that to every member there belongs an immediate successor with which it is connected by an unambiguous transitive and asymmetrical relation, that remains throughout the whole series, then, in such a "progression," we have already grasped the real fundamental type with which arithmetic is concerned. All
the propositions of arithmetic, all the operations that it defines, are related solely to the general properties of a progression; hence they are never directed primarily upon "things" but upon the ordinal
which prevails between the elements of certain systematic The definitions of addition and subtraction, multiplication and division, the explanation of positive and negative, whole and fractional numbers, can be developed purely on this basis, and with out especially going back to the relations of concrete measurable objects. According to this deduction, the whole "certitude" (Bestand) of numbers rests upon the relations, which they show within themselves, and not upon any relation to an outer objective reality. They need no foreign (Substrai) but mutually sustain and support each other in so far as the position of each in the system is clearly determined by the others. "When," says Dedekind in
relation,
wholes.
"basis"
definition,
"in
arranged by the
the consideration of a simple infinite system N, we totally abstract from "copying" (Abbildung)
</>,
the particular properties of the elements, retain merely their dis tinctness, and attend only to the relations in which they are placed
to each other
called the natural
by the ordering "copying" $, then these elements are numbers or the ordinal numbers or also simply
numbers, and the fundamental element 1 is called the fundamental number of the numerical series N. With reference to this liberation of the elements from every other content (abstraction), we can The correctly call numbers the free creation of the human mind.
relations or laws,
which
....
in all ordered simple infinite
systems are always the same, whatever names may accidentally be given to the individual elements, form the primary object of the
7
For more exact treatment
cf.
Russell, op.
cit.,
Chs. 24 and 25.
THE CONCEPT Of NUMBER

science of
39
it is
number
or
arithmetic."
From
The
a logical standpoint,
act of abstraction
is
of special interest
that here the concept and term
"abstraction"
is
obviously applied in a
directed
new
sense.
not
upon the separating out of the quality of a thing, but its to bring to consciousness the meaning of a certain relation independently of all particular cases of application, purely in itself.
aim
is
The function
of
"number"
is,
in its
meaning, independent of the
factual diversity of the objects which are enumerated; this diversity must therefore be disregarded when we are concerned merely to
develop the determinate character of this function. Here abstrac tion has, in fact, the character of a liberation; it means logical con
centration on the relational connection as such with rejection of all psychological circumstances, that may force themselves into the
subjective stream of presentations, but which form no actual con***-~stitutive aspect of this connection.
Number
as ordinal number.
It
has occasionally been objected to
Dedekind s deduction that according to it there remains in principle no distinguishing content for number, to mark its peculiarity as opposed to other serially ordered objects. Since in the determina
tion of
concept only the general moment of "progression" is retained, everything that is here said of number is valid with regard to every progression in general; it is thus only the serial form itself
its
If the defined and not what enters into it as material. numbers in general are to exist then they must, so it seems, have some nature and property; they must be distinguished in the same way that from other entities by some absolute 9 But this points are different from instants, or tones from colors. objection mistakes the real aim and tendency of Dedekind e deter
that
is
ordinal
"inner"
"mark,"
minations.
What
is
here expressed
"essence"
is
?ystem of ideal objects whose whole content
just this: that there is a is exhausted in their
mutual
relations.
The
of the
10 expressed in their positions.

8
And
numbers is completely the concept of position must,

"Abbildung"
Dedekind, op.
Cf. Russell, op.
cit.,
6.
On
the concept of
system" cf.
cf.
above.
5
On
6.
the definition of a
9
"simple
infinite
Dedekind, op.
cit.,
and
cit.,
242.
10
On
the deduction of
number
as pure
"serial
number," cf.
especially the
exposition of G. F. Lipps (Philosoph. Studien, ed. by Wundt, Vol. Ill), and also the latest discussions of Natorp, which carry through these thoughts with special clarity and penetration. (Die logischen Grundlagen der exakten Wissenschaften, Leipzig, 1910, Chs. Ill
and IV.)
40
first

of
all, be grasped in its greatest logical universality and scorjg. distinctness required of the elements rests upon purely concep The intuition of pure time, tual and not upon perceptual conditions^.
The
upon which Kant based the concept of number, is indeed unnecessary here. True, we think the members of the numerical series as an
ordered sequence; but this sequence contains nothing of the concrete character of temporal succession. The "three" does not follow the
"two"
as lightning the thunder, for neither of
them
possesses
any
temporal reality but a merely ideal logical constitution. The meaning of the sequence is limited to the fact that two enters as a premise into the determination of three, that the meaning of one concept can only be explained by that of the other. The lower number is "presupposed" by the higher number; but this does not imply a physical or a psychological earlier and later, but a pure rela
sort of
tion
of
systematic
position
the
"later"
conceptual dependence. What characterizes is the circumstance that it issues from the
fundamental unity through a more complex application of the generating relation, and consequently takes up the elements that
precede it as logical constitutive parts and phases into itself. time (if we understand by it the "concrete form" of the "inner
time.
Thus,
sense")
presupposes number, but number does not, conversely, presuppose Arithmetic can be defined as the science of pure time only
of time (as Hamilton does, for instance), all special determination of character, and merely retain the moment of "order in progression." 11 It is just this that now
when we remove from the concept
proves to be the merit number: that in it the

sive connection
is
in
fundamental method of the science of

of the elements of a certain progres
"how"
"what"
of this connec disregarded and merely the tion is taken into account. Here we meet for the first time a general procedure, which is of decisive significance in the whole formation of mathematical concepts. Wherever a system of conditions is given
that can be realized in different contents, there we can hold to the form of the system itself as an invariant, undisturbed by the differ ence of the contents, and develop its laws deductively. In this way we produce a new "objective" form, whose structure is independent of all arbitrariness; but it would be uncritical naivet6 to confuse the
11
On
William Hamilton
progression"
s definition its
of algebra as
"science
of
pure time or
order in
essay
and
relation to the
Mathematik,"
"Kant
und
die
moderne
Kantian concept of time, c/. my Kant Studien, Vol. XII, p. 34 f.

object which thus arises with sensuously real
41
We
cannot read
for its
form
is
and effective things. empirically, nor do we need to, revealed in all determinateness as soon as we have
off its
"properties"
grasped in
its
The
theories of
purity the relation from which Helmholtz and Kronecker.

it
it
develops.
is
Fundamental as
the
A new aspect appears as soon as number, which has hitherto been deduced as a purely logical sequence of intellectual constructs, is understood and applied as an expres sion of plurality. This transition from the pure ordinal number to
,
conceptual moment of order, yet tent of the concept of number.
does not exhaust the whole con
is made with general agreement by the various ordinal theories of arithmetic, as they have been developed by
the cardinal
number
Dedekind and in particular by Helmholtz and Kronecker. Given any finite system, we can relate it to the previously developed totality of numbers in a clear and definite way by letting every element of the system correspond to one and only one position in this totality.
In this way we are finally able, by following the fixed prescribed order of the positions, to coordinate the last member of the system with a certain ordinal number, n. This act of coordination, however, which concludes the process, contains in itself all the previous phases;
for since the progress
from 1 to n can only take place in one way, number which we reach reproduces the total operation in its The number n, which was primarily gained as a specific character.
the
characteristic of the last element, can thus be regarded, from another point of view, as a character of the total system: we call it the
system considered, and now say of the latter n elements. 12 Here it is presupposed above all that there can be one and only one cardinal number of a given group, that thus the "position" we finally reach is independent of the order in which we successively regard and emphasize the members of the
cardinal
it
number
of the
that
consists of
group. This presupposition, however, as Helmholtz in particular has shown, can be proved with all strictness from the premises of the ordinal theory without the assumption of any new postulate, pro
vided
definitions of the fundamental arithmetical can also be transferred without difficulty to the new sort operations of numrJer. Thus the formation of the sum (a b) means, with the pure ordinal numbers, that starting from a, we "count regard
we maintain the The finite system.
condition that the manifold considered forms
t^>
12
Cf. especially,
Dedekind, Was sind und was sollen die Zahlen,
161, p. 54.
42

place in the series which
b steps, that is, that we determine the we reach when we coordinate the numbers
on"
following
a,
member
for
This member, 1, 2, remains valid without modification when we go over explanation to the addition of the cardinal numbers; it is evident that from the combination of the elements of two groups, which fall under the cardinal numbers a and b, there results a new group C, the number of
series
3, ... 6.
is given by the number (a b) in the previously determined meaning. The consideration of the "cardinal numbers" thus occasions the discovery of no new property and no new relation, which could not have been previously deduced from the bare ele
with the
elements of the
whose members
of order. The only advantage is that the formulae developed the ordinal theory gain a wider application, since they can hence by forth be read in two different languages. 13
ment
Even though no actual new mathematical content is produced through this transition, nevertheless in the formation of the cardinal number a new logical function is unmistakably at work. As in the
theory of the ordinal number the individual steps as such are estab lished and developed in definite sequence, so here the necessity is
comprehending the series, not only in its successive elements, but as an ideal whole. The preceding moment is not to be merely set aside by its successor, but is to be retained in its entire logical
felt of
import in the latter, so that the final step of the procedure contains in itself at once all the preceding steps and the law of their mutual connection. It is first in this synthesis that the bare sequence of the ordinal numbers is developed into a unitary, self-inclosed system, in which no member exists merely for itself alone but, on the contrary, represents the structure and formal principle of the whole series. Criticism of the nominalistic deduction. If once these two funda mental logical acts at the basis of all differentiation and connection of
numbers are recognized, then no further

needed to determine the
field of
special presupposition is the operations of arithmetic. The demand is therewith realized for a purely rational deduction, which renounces all dependence on the empirical relations of physical
objects.
quently misunderstood in estimating the
Indeed precisely this distinctive character has been fre "ordinal" theory of number. The explanation of the theory, as given by Helmholtz for example,
13
Cf. Helmholtz, Zdhlen
und Messen,
erkenntnistheoretisch betrachtet (Philo-
soph. Aufsalze, Ed. Zeller gewidmet, Leipzig, 1887, p. 33).
43
leads necessarily to the view that first of all concrete groups of ob presupposed as given, and that the whole work of thought is exhausted in introducing a diversity of symbols corresponding to
jects are
this diversity of things.
However
"symbols"
are nothing
them
selves but groups of perceptible objects, visibly distinguishable from each other by shape and position. Hence we seem to be able to
abstract from the immediate properties of things in assertions con cerning numerical relations, only because we have substituted for
the reality of things their sensible "copies." Thus the true beginning of the formation of numbers would be not an abstraction from physi
cal objects, but on the contrary a solidification and concentration of their sensuous import. Every such interpretation, which different
mathematicians have seemed to approach at times in their expositions of the ordinal theory of number, contradicts its essential and deeper The symbols produced would cease to be symbols, logical tendency. would lose their characteristic function, if they were judged merely
according to what they sensuously are, and not according to what they intellectually mean. What would remain, in fact, would be only certain "images," which we could investigate as to form and But no mathematical "nominalism," size, position and color. however extreme, has ever actually attempted to transform the import of valid judgments about numbers into assertions of this
kind. It is only the ambiguity in the concept of symbol, only the circumstance that under it can be understood, now the bare exist ence of a sensuous content, and now the ideal object symbolized by the latter, which makes possible this reduction to the nominalistic
schema. Leibniz, whose entire thought was concentrated upon the idea of a "universal characteristic," clearly pointed out in opposition
to the formalistic theories of his time, the fact which is essential here. The "basis" of the truth lies, as he says, never in the symbols but in
the objective relations between ideas. If it were otherwise, we would have to distinguish as many forms of truth as there are ways of symbolizing. Among modern mathematicians, Frege especially has
shown
itself.
in
is
symbols
penetrating, detailed criticism that the arithmetic of only able to keep itself in existence by being untrue to
In place of the empty symbols the meaning of the arithmeti
cal concepts appears unnoticed in the course of the argument. 14

14
Frege, Grundgesetze der Arithmetik, Vol. II, Jena, 1903, pp. 69
ff.,
139, etc.
44
In the theory of the pure ordinal numbers also, the nominalistic interpretation forms only an outer husk, which must be stripped off
to reach the real kernel of the logical and mathematical thought. Once this is done, what we retain is the purely rational moment; for
not something which can be immediately pointed out in sense-impressions but is rather something which belongs to them only
"order"
is
virtue of intellectual relations. Thus the theory in its pure form does not require the assumption of a group of physically given, particular things, as has been urged against it. 15 The manifolds which it makes fundamental are, not empirically present, but ideally defined,
by
which are progressively constructed from an assumed beginning according to a constant rule. In this rule also are rooted
totalities,
all
genuinely
"formal"
properties, such as distinguish the series of
numbers and make
it
the fundamental type of a conceptually intel-
Jigible connection in general.
in
actual
Number and the concept of class. modern development of the

if
If,
however, we survey the

it
principles of mathematics,
the previous treatment of the subject had overlooked the moment which alone completes the logical characterization essential
seems as
of
number.
Wherever the attempt has been made to resolve the
concept of
into pure "logical constants," the concept of class has been regarded as a necessary and sufficient presupposition.
number
jj
number seems completed only when all its special has been derived from the general function of the concept. import But according to the dominant logical theory, here again, the formaanalysis of
The
|1
means nothing but the collection of objects into and genera by virtue of subsumption under general attributes. species Thus, in order to understand the concept of number, everything must first be removed from it which does not fit into this schema. But a fundamental difficulty here arises for the theory. If we are
tion of concepts
number in general, but the concept of number, we are not dealing with a logically universal concept but with an individual concept. Our concern is not with giving a species, which can be found in any number of
considering not the thought of
this or that determinate
individual examples, but with the determination of a certain definite

15
Cf. Couturat,
De
lnfini mathematique, Paris, 1896, p. 318
ff.
45
position within a total system. There is only one "two," only one and both of them possess certain mathematical properties "four,"
and
characteristics
which they share with no other object.
If
the
reduction of the concept of
to the concept of class is to be To in spite of this, then another direction must be taken. possible determine what a number according to its pure essence, we
"is,"
number
do not attempt to analyse it directly into simpler constitutive parts, but we ask primarily what the equality of numbers means. As soon as it is established under what conditions we shall regard two groups as of the same numerical value, the peculiarity of the "mark," which we assume to be identical in both, is thereby indirectly determined. The criterion of the numerical equality of two groups, however, con sists in that a certain relation can be given, by virtue of which the members of the two groups can be mutually coordinated one to one.
By
virtue of this process of coordination,
we
r
establish
among
the
infinitely
numerous
possible classes of objects, certain connections
by
uniting into a total complex those groups which can be coordinated In other words, we unite into a species all the manifolds in this way.
which there exists such a relation of "equivalence" or of one to one coordination; while we regard groups in which this condition is not fulfilled as belonging to different species. When this has been
for
done, any individual group, by virtue of the character of equivalence, can be regarded as a perfect representative of the whole species: For since it can be shown that two groups which are equivalent to a
third are also equivalent to each other, it is enough to prove of a that it can be coordinated member for member with given total
any group of the whole complex in order to establish the certainty that the same is true of all groups of the complex in question. Now
if
we abstract the common relation, which all the wholes of such a complex have to each other and consider it as a possible object of
thought, we gain the moment which, in ordinary language, we call which belongs to a the number of these wholes. "The number
concept
to
F,"
F,"
mental features,
We
says Frege, to whom we owe this deduction in its funda the extension of the concept: numerically equal the number of a concept when we do not regard the grasp
"is
objects which fall under it merely for themselves alone, but at the same time include all those classes whose elements stand in the rela
tion of one to one coordination with those of the whole under
consideration.
46

Russell s theory of cardinal numbers.
Thus
it
is
characteristic of
this conception that it emphasizes what appears in the ordinary view simply as the criterion of numerical equality as the peculiar consti
of
upon which rests the whole content of the concept number. While the traditional view presupposes the individual numbers as given, as known, and decides on the basis of this knowledge as to their equality or inequality, here the reverse process holds. Only the relation, which is asserted in the equation, is known while the elements entering into this relation are at first undetermined and are only determinable by the equation. Frege thus describes
tutive character,
;
the general procedure, "Our purpose is to shape the content of a r judgment, w hich can be interpreted as an equation, in such a manner that upon each side of this equation there is a number. We thus
to reach by means of the already known concept wish Here a methodologi of equality that which is to be regarded as equal." cal tendency is clearly defined, which is fundamental in all construc tion of mathematical concepts:: the construct" is to gain its total
.
constitution from the relations in which

"
it
stands.
(Cf.
above
p.
40
of
The only question that remains is whether, in the relation the equivalence of classes, we really grasp a relation which is
f.).
logically simpler than the totality of functions which, in the ordinal An theory, leads to the systematic series of the ordinal numbers. advance in analysis would necessarily result in our being able to
abstract entirely from all these functions, and yet accomplish in a new way the complete construction of the realm of number and its laws.
It is
upon
this point, therefore, that all critical investigation
must be
concentrated: Is the deduction of the series of numbers out of the

circle
concept of class really successful, or does this deduction move in a by tacitly presupposing concepts from the very field which it
undertakes to deduce? 16
The
conflict
theory, which has here been developed, although in sharp with the empirical interpretation of the nature of number,
The problem, which is involved here, has been discussed in lively fashion modern logico-mathematical literature; for positive exposition of the theory cf. especially the writings of Frege, Russell and Peano; for criticism cf. B. Kerry, Uber Anschauung und ihre psychische Verarbeitung, Vierteljahrsschr.
16
in the
f.
wissensch. Philos. XI, 287

ff.;
ff.
1891, p. 129 1908, p. 158
ff Husserl, Philosophic der Arithmetik, I, Halle, Jonas Cohn, Voraussetzungen und Ziele des Erkennens, Leipzig,
.
THE CONCEPT OF NUMBEK

:
47
agrees with it in one formal characteristic it also understands number as a "common property" of certain contents and groups of contents.
The
sized, is
basis of numerical assertions, however, as is especially empha not to be sought in sensuous physical things themselves but
wholly in the concepts of things.
Every judgment about numerical
relations ascribes certain attributes, not to objects, but to their con cepts, by which they are divided into classes with peculiar properties.
Venus has moons, there is in fact no moon or aggre there of which anything could be said but a property is thereby ascribed to the concept moon of Venus/ namely, that of including nothing under itself. When I say the coach of the emperor
"When
say:
gate of
moons
is
drawn by four
concept
horses, I thereby ascribe the number four to the This fact horse that draws the coach of the emperor.
alone explains the universal applicability of numerical assertions, which equally covers the material and the immaterial, outer and inner phenomena, things as well as experiences and deeds. This apparent diversity of the field of what can be numbered proves upon closer consideration to be strict uniformity; for the enumeration never concerns the heterogeneous contents themselves but the concepts under which they are comprehended, and thus is always concerned with the same logical nature. The previous exposition has shown how this can be more precisely understood; a certain numerical determination is to be ascribed to concepts when they are collected into classes with other concepts with which they stand in the relation of possible one to one coordination of their elements in extension.
Criticism of "Class theories." Against this explanation, however, one objection necessarily arises. The theory, which is here defended, is by no means concerned to excogitate the concept of number arbitrarily, but to indicate the real function which number possesses in the actual whole of knowledge. Precisely this is emphasized in to the interpretation which proceeds from the ordinal opposition numbers, that the "logical" properties of number here deduced are the very ones that are definitive and essential in its "use in daily A natural deduction, such as shall do justice to the concrete applications of number, is to be opposed to the technical deduction, which merely keeps in view the purposes of scientific arithmetic. But a closer examination shows that this goal is not attained; for what is logically deduced here is in no way identical with the real meaning that we connect with numerical judgments in actual knowllife."
48
edge.
If
we limit ourselves merely to the previous postulates, we are indeed enabled by means of them to place different groups of elements together and to regard them as similar from a certain point of view;
but no sufficient determination
is
thereby gained of their
"number"
in the ordinary sense of the word. Our thought could run number of "equivalent" groups and consider their any
through
relations without
any
characteristic consciousness of the pure
mutual numeri
of
cal concepts resulting

"four"
from the process.

"four"
The
specific
meaning
or
"seven"
any number groups must first be determined as ordered sequences thus as numbers in the sense of the ordinal theory. The
of
could never result from the bare placing together of groups of or "seven" elements; the individual
of elements,
"how many"
of the elements, in the ordinary sense, can be changed by no logical transformation into a bare assertion concerning "just as many;" that remains an independent problem of knowledge. Considera
tion of this problem, however, leads back to a deeper opposition in method between the two interpretations of number. It is a funda
mental characteristic of the ordinal theory that in it the individual number never means anything by itself alone, that a fixed value is only ascribed to it by its position in a total system. The definition of the individual number determines at once and directly the relation in which it stands to the other members of the field and this relation cannot be eliminated without losing the entire content of the particu lar numerical concept. In the general deduction of cardinal numbers, which we are considering, this connection is eliminated. It is also
;
necessary that this deduction erect and logically derive a fixed principle of arrangement of the individual numbers; however the
meaning of the elements is to be established before this arrangement and independently of it. The members are determined as the
common properties of certain classes before anything whatever has been established as to their relation of sequence. Yet, in truth, it is precisely in the element here at first excluded that the peculiar
numerical character
basis of
is
rooted.
The conceptual
construction at the
to abstract similarities, as must be the case according to the traditional doctrine of abstraction, but to sepa rate out and maintain diversity. The consideration of groups, which
number does not tend
can be mutually coordinated member for member, can only lead to the separating out of the identical "mark" in them; but this "mark" is in itself not yet "number" but is merely a logical property not

further defined.
49
separates
itself
Such a property only becomes number when it from other "marks" of the same logical character by
"later,"
or "more" appearing in a relation to them of "earlier" or or Those very thinkers, who have carried through the explanation of number by equivalent classes most strictly and con
"less."
sistently, emphasize therefore that this explanation is irrelevant to the methods of pure mathematics. What the mathematician con siders in number are merely the properties on which rest the order of
Number in itself may be what it will; for analysis and algebra it only comes into consideration in so far as it can be 17 purely and completely developed in the form of a "progression." Strictly speaking, once this is admitted, the dispute as to the method For where can ological precedence of the ordinal number is over. more certain information be gained as to the "nature" of number, in the sense of the criticism of knowledge, than in its most general
the positions.
scientific use?
The appeal
pre-scientific
to the
meaning attached
to the concept of
number
in
thought also does not withstand criticism. Psychologi cal analysis at least offers no support to the theory. All reflection on the actual state of thought, on the contrary, shows clearly the inner difference between the thought of equivalence and that of number. If number were what it alone should be according to this deduction, then it would be a peculiarly complicated and difficult task to point out the process by which such a concept would arise and be main tained in consciousness. For number means here a relation between two classes wholly heterogeneous in content, which are connected by no further moment than the mere possibility of mutual coordination. But what intellectual motive would there be in general to relate such dissimilar groups with each other; what meaning would there be,
for example, in placing the class of moons of Jupiter beside that of seasons of the year, the group of pins in nine-pins beside that of the muses! Such a comparison is intelligible after the numerical value
of these classes
and an
indirect
agreement between them has been
established in another way; but here, on the other hand, where this value is not presupposed but is to be gained from the comparison, the
comparison itself lacks any fixed guide or standard. It has been urged against the theory of equivalence that it leads to an "extreme relativism" in so far as the determination of number is to be a
17
Cf. Russell, op.
cit.,
230.
On
the concept of progression,
cf.
above
p. 49.
50

itself,
property, which does not belong to a group in
but merely
in.
relation to other groups. This charge is at least ambiguous; for the of number can, in fact, be nothing in any form of deduction concept
but a pure relational concept.
Only here the
field
and thus the
logical place of the relation is shifted; for while
we
are concerned in
the ordinal theory with ideal constructions, which are mutually related to each other, here each individual construction is deduced from the relation of given "classes."
The logical definition of the zero and of unity. The presuppositions, that are assumed here, come to light very clearly as soon as we pro ceed to give a strictly logical definition of the individual numerical values from this point of view, and to determine the conditions under
which we are to regard two of these numerical values as immediately In fact, in the explanation of null there appear grave successive. difficulties; for there is obviously no meaning in still speaking of the mutual one to one coordination of the members of different classes in the case in which these classes by definition contain no members. But, even if this difficulty could be removed by complicated logical
transformations of the concept of equivalence, 18 the circle in the explanation again becomes clear as soon as we advance to the defini
tion of
"one."
here assumed to be
equality
What it means to apprehend an object as is known from the very beginning; for the numerical of two classes is known solely by the fact that we coordinate
"one"
with each element of the first class one and only one of the second. This remark, simple and even trivial as it seems, has been frequently
It has been objected that it is something different take the number one in its strictly mathematical meaning or merely in the vague sense represented by the indefinite article:
controverted.
I
whether
it is
merely this
"That
last sense,
which
is
presupposed,
relate
it
when
am requested
of a class in a certain
to take
v.
any member
of a class
u and
"is
to a
member
is,
every individual or every
member
of a class
sense,
says Russell, e.g. naturally incontestable, but it does not follow from this that the concept of is presupposed, when we speak of an individual. We can rather conversely regard the concept of the individual as the fundamental concept, from
one,"
"one"
p. 113,
p.
on this point: Frege, Grundlagen der Arithmetik, p. 82 ff.; Russell, along with the criticism of Kerry, Vierteljahrsschr. f. wiss. Philos. XI, 287 ff., and also of Poincar<, Science et Methode, Paris, 1908, Bk. II. For
Cf.
cf.
18
criticism of Frege,
Natorp, op.
cit.,
p. 112
ff.
51
which the concept of one is derived." From this point of view, the member meaning of the assertion that a class u possesses the arithmetical sense) is determined by the fact that this class is (in not null and that if x and y are u s, then x is identical with y. A similar determination is then to fix the meaning of the concept of mutual one to one correlation between terms: R is such a relation that in the case that x and x have the relation R to y, and x possesses the relation R to y and y x and x and also y and y are identical. 19
"one"
It
is
nevertheless, easy to see that here the logical function of number not so much deduced as rather described by a technical circumlocu
is,
tion.
least necessary that
In order to understand the explanations given here, it is at we comprehend a term x as identical with itself,
while at the same time we relate it to another term y and judge the former as agreeing with or differing from the latter, according to Now if we take this process of positing special circumstances. and differentiation as a basis, we have done nothing but (Setzung]
presuppose number in the sense of the ordinal theory.

the class of two objects
it
Thus,
e.g.,
is
defined
by Russell by the
terms and that, if x is class, y, which is different from x; while further, if x, y are different terms of the class u, and z is different from x and y, every class that belongs to z is also different from u. We see how here in order to complete the explanation, the elements x, y, z,
in general possesses
conditions, that one of its terms, there is
another term of the
are produced in progressive
differentiation
and must,
third
therefore,
be
indirectly
distinguished
as
first,
second,
....
members.
definitely
In general, in order to bring the different numbers into the form of a ordered "progression" (and it is upon this form their
scientific
meaning and
number.
is
use primarily rest),

is
that enables us,
when any number n
we must have a principle given, to define the succeeding

"being
This relation between two numbers
neighbors"
determined, according to the theory, by comparing the corre sponding classes u and v, with each other, by coordinating their
now
elements member for member. If it is found here that in class (v) a member remains, which possesses no corresponding member in the other (u), then we designate v in relation to u as the "next higher"
class.
part of
19
Here also it is postulated that we first grasp as a whole the v, which can be coordinated member for member with u,
124-126,
496.
Cf. Russell,
Frege, Grundlagen, p. 40
ff.
52
in order to grasp the
member
that remains unconnected in this form
Thus here the progress from unity of connection, as a "second." to what next follows it is founded on the same intellectual synthesis
in principle as that upon which it rests in the theory number; and the only difference in method consists of the ordinal
in that these
syntheses appear, in the ordinal theory, as free constructions while here they depend upon given classes of elements. 20 The presupposition of the class concept. It follows, however, from a
final
inverted in this interpretation.
decisive consideration that the logical order of concepts is The determination of number by the
equivalence of classes presupposes that these classes themselves are given as a plurality. The concept of the "similarity" of classes, on which the meaning of the cardinal numbers is to be based, requires
at least the consideration of
tion.
It
two wholes connected by a certain
rela
has been emphasized that, for the establishment of this one to one relation, it is not necessary that the members of the two mani
numbered but that, on the contrary, the general any element of the first manifold is placed in con nection with any element of the second. But even if we give up, in
folds be previously
is satisfied if
law
accordance with this point of view, the prior enumeration of the individual classes which we compare, nevertheless the circumstance
remains that we must oppose the classes as wholes to each other, and different units. It may be thereby understand them as
"two"
immediately given by the purely and thus neither needs nor is We would therewith be led from the classes themselves back to the generating relations, upon which they rest and to which they owe their definition and character. The
objected that
this
difference
is
logical difference of the class concepts, capable of any further deduction.
series of natural
In order to explain the relation, in which two neighboring members of the numbers stand to each other, Frege, for example, starts from the proposition: "there is a concept F and an object x falling under it of such a sort that the number which belongs to the concept F is n, and the number this is explained which belongs to the concept falling under F but notx is m as equivalent to the proposition that n is the immediate successor of m in the natural number series. Op. cit., p. 89. Thus here a distinction is drawn within the totality Fin which a single member x is selected and opposed to the others: all these others are then used in the definition of the neighboring "just lower" number. We thus have here also only a circumlocution of the "popu lar" view, according to which each member of the series of numbers is dis tinguished from its neighbors by the addition or absence of a "unity."
20
;"
53
difference in the systematic wholes reduces to the difference in the
however,
conceptual law, from which they have proceeded. From this point, it is evident that the system of the numbers as pure ordinal
numbers can be derived immediately and without the circuitous route through the concept of class; since for this we need assume nothing but
the possibility of differentiating a sequence of pure thought construc tions by different relations to a certain fundamental element, which
serves as a starting-point.
The theory
of the ordinal
number thus
represents the essential minimum, which no logical deduction of number can avoid; at the same time, the consideration of equivalent classes is of the greatest significance for the application of this concept,
yet does not belong to
its original
content.
The generic concept and the relational concept. Thus the conflict of mathematical theories here again combines with the general questions of logical principle that were our starting-point. In the different
interpretations of the concept of number, there is repeated the general conflict between the logic of the generic concept and the nf fhft rftlntinnnl Jofrjc concept. If the attempt to derive the concept
of class were successful, the traditional form of would gain a new source of confirmation. The ordering of individuals into the heirarchy of species would be, now as before,
of
number from that
logic
the true goal of all knowledge, empirical as well as exact. In the attempts to ground the logical theory of the cardinal numbers, this connection has occasionally come clearly into view. According to
Russell, if I grasp the thought "two men" I have thereby formed the logical product of the concepts "man" and the concept "couple;" and the proposition that there are two men says only that a complex
is
class
given which simultaneously belongs to the claaaJlman! and the 21 It becomes evident at this point that the theory "couple."
has not carried through the fundamental critical ideas from which it started. Frege and Russell regard as the decisive merit of their
it number does not appear as a property of physical but as an assertion concerning certain properties of classes;^ things that in it therefore objects as such do not form the basis of numerical judgments but rather the concepts of these objects. (Cf. above p. 29 ff .) It is incontestable that, compared with the sensationalistic
doctrine, that in
interpretation, an extraordinary liberation of thought and increase of depth is gained by this transformation. Nevertheless it does not
21
Cf. Russell, op.
cit.,
111.
54
suffice to
I

emphasize the purely conceptual character of numerical
J(
and functional concepts are placed on the same plane. Number appears, according to this view, not as the expression of the fundamental condition which first renders possible every plurality, but as a "mark," that belongs to the given plurality of classes and can be separated from the latter by compari
assertions as long as thing-concepts
son.
Thus the fundamental

is
:
deficiency of the whole doctrine
of
repeated an attempt is made to view what guides and controls the formation of concepts, i.e., a purely "categorical" point of view, as, in some wayâr^ristitutive part of the compared
abstraction
(Cf. above p. 24 ff .) /The theory finally shows itself to be a objects. subtle and extended attempt to deaL/oy means of th^gejiataLachejna
/of the generic concept, with a problem which belongs in its mean ing and scope to a new field and presupposes another concept of 22 knowledge.
IV
Extension of the concept of number. The previous attempts to estab the character of number and the principle of its formation, how ever, have not grasped the question in that universality and breadth,
lish
which
is
it
to
number
has gained in the development of modern mathematics. It in its most primitive form and meaning that the
22 Indeed not merely logical views but also more special mathematical reasons led to the explanation of number through the equivalence of classes. Only on this foundation did it seem possible to produce a theory, which would not be limited from the very beginning to the finite numbers, but would include and characterize the "finite" and "infinite" numbers in a single deduction.
The aspect
of
significance, for it
mutual one to one coordination of groups seemed of fundamental remained when one abstracted from the finitude and there
with the enumerability of the groups, according to the ordinary interpreta tion of enumeration as the successive advance from unity to unity. Fruit ful as the general point of view of "power" (Machtigkeit), which arises in this connection, has shown itself to be, it has nevertheless in no way been proved to be identical with the concept of number. The purely mathematical sig nificance of the concept of "power" is obviously unaffected by whether we
regard it as the original principle of number or as only a derived result that The properties, which are presupposes another explanation of number. common to the finite and transfinite numbers, by no means contain as such the essential element for the construction of number in general: the "summum genus" in the sense of the logic of the generic concept is here also not identical with the conceptual origin of knowledge. (On the problem of the transfinites, cf. below p. 80 ff.)
THE COXCEPT OF NUMBER

attempted deduction of the
55
class theory as well as that of the ordinal In principle, the standpoint of the Pythagoreans is theory applies. not yet left behind; number, in the narrower sense of the whole
forms the only real problem. The scientific system of arithmetic, however, is first completed in the extensions, which the concept of number undergoes through the introduction of the opposi tions of positive and negative, whole and fractional, rational and irrational numbers. Are these extensions, as prominent mathemati cians have asserted, merely technical transformations, which can only
number,
still
be explained and justified as applications, or are they expressions of the same logical function that characterizes the first institution
numbers? Gauss theory of the negative and imaginary numbers. The difficul ties encountered in the introduction of every new type of number,
of
1
of the negatives and the irrationals as well as of the imaginaries, are easily explained if we consider that, in all these transformations, the real basis of numerical assertions seems more and more to disap
pear.
diately
valid.
Enumeration, in shown to be
its
"real"
most fundamental sense, could be imme by means of sensible objects and therefore
"four"
The meaning of or forms, as it seems, no serious problem, for the empirical world of things everywhere offers us groups of two and four things. With the first generalization and extension
"two"
number, however, this reference to things, upon which the naive interpretation especially rests, disappears. The concept and designation of the "imaginary" number is the expression of a thought, that is effective in principle in each of the new types of number and gives them their characteristic stamp. Judgments
of the concept of
and assertions concerning the "unreal" here lay claim to a definite, indispensable cognitive value. This connection and the general principle, to which all the different methods of the extension of
number go back,
is
stated
by Gauss with complete
clarity
and
dis
tinctness in a passage in which he sets himself the task of grounding "Positive and negative the true "metaphysics of the imaginary."
numbers," it is
here stated,
"can
only find application where what

is
is
enumerated has an opposite, to be united with which

annihilation.
identical to
In strictness, this presupposition is only realized where not substances (things conceivable by themselves) but relations between two objects are enumerated. It is thereby postulated that these objects are ordered in some way into a series, e.g., A, B,
56

. . .
and that the relation of A to B can be regarded as the C, D, same as that of B to C. Now in this case nothing more belongs to
the concept of opposition than the reversal of the relation; thus, if the relation (and thus the transition) between A and B is represented 1. In so far as such 1, the relation of B to A is represented by by
a series
is
sents the relation of
unlimited in both directions, every real whole number repre any member arbitrarily chosen as a beginning to
some definite member of the series." The deduction of imaginary numbers rests, further, upon the fact that the objects we are inves tigating are no longer regarded as ordered in a single series, but as requiring for their arrangement consideration of a series of series, and
from
thus the introduction of a new unity (+ i, i). Here, if we abstract all details of the deduction, the dominating logical view stands
out clearly. The meaning of the generalized concepts of number cannot be grasped as long as we try to indicate what they mean with regard to substances, with regard to objects conceivable in themselves. But the meaning at once becomes intelligible when we regard the
concepts as expressing pure relations through which the connections in a constructively produced series are governed. A negative sub stance, which would be at once being and not-being, would be a contraadjecto; a negative relation is only the necessary logical correlate of the concept of relation in general, for every relation of to B can also be represented and expressed as a relation of B to A.
dictio in
If
we
the transition from a
consider, therefore, the generating relation (R), on which rests member of the series of numbers to that imme
diately following, we thereby postulate also a relation of the follow ing member to the preceding, and thus define a second direction of
progress, which we can understand as the converse of the first, or as the inverse relation (R). The positive and negative numbers (+ a, a) now appear merely as expressions of progress (Fortgang) in these a two directions of the relations (R a From this fundamental ).
,
the operations of calculation within the thus extended conception, field of numbers can be deduced, for all these operations are founded on the character of pure number as relational number and express
all
with increasing clarity. 23 The irrational numbers. Once more we shall not trace the develop ment in all its particular phases but shall merely study typical
this
23
Cf. here especially the penetrating exposition

cit.,
and proof
of this connection
given by Natorp, op.
Chs. Ill and IV.

examples
in
57
tendency of the thought is most clearly principle is verified, above all, in the deduction expressed. There seem at first two ways in which a of the irrational numbers. deduction of the irrationals can be attempted. We could start either from the relations between given geometrical extensions or from the
logical
which the
The new
postulate of the solubility of certain algebraic equations.
The
first
method, which was almost exclusively accepted up to the time of Weierstrass and Dedekind, bases the new number upon space and thus upon relations found between measurable objects. Here again it seems to be experiences of physical-spatial objects which control the process of the formation of mathematical concepts and prescribe its direction. Nevertheless it becomes evident that at least the appeal
to the relations of concrete empirical things must fail at this point. The relative magnitudes of things are known to us only through observation, and thus only within the limits imposed by the personal
equation (Beobachtungsfehkr). To demand an absolutely exact determination in this field is to mistake the nature of the question
itself.
Thus the ordinary system
of fractional
numbers
is
obviously
adequate in every respect to accomplish For within this system there all the tasks that can arise in this field. is no smallest difference, for between any two elements, however near, there can always be given a new element belonging to the system;
thus a conceptual differentiation is offered here, which is never reached in the observable relations of things, to say nothing of being surpassed. The determinations of size, to which we are led by exter
nal experience, can thus never force us to the concept of the irrational On the contrary, this con in its strict mathematical significance.
arise and be grounded from within the circle of the postu upon which rests the systematic connection of mathematical In any case, not the bodies of physical reality but the cognitions.
an
intellectual instrument
cept
must
lates,
purely ideal extensions of geometry can afford the desired basis for the derivation of the irrationals. The new problem does not develop
out of the apprehension of given, factually presented magnitudes, but out of the laws of certain geometrical constructions. Once this
is
recognized, the further
demand must
arise that the construction,
indispensable in
itself
of
any attempted deduction, shall proceed and justify from the fundamental principle of number itself. The shifting the question from number to space would destroy the unity and
itself.
completeness of the system of algebra
58
The ordinary algebraical method, which introduces the irrational values as the solutions of certain equations is indeed inadequate, for For even it confuses the erection of a postulate with its fulfillment.
if
we
abstract from the fact that there are infinitely
many
irrational
values such as cannot be represented as the roots of algebraical equations, in any case, such an explanation does not decide whether the object produced by them is definitely determined or whether there are several different values which satisfy the stated conditions.
An
adequate definition must not characterize the ideal object upon which it is directed merely by some particular "mark" that belongs to it, but must comprehend and determine it in its full characteristic This individuality, by which it is distinguished from all others. individuality, however, is completely determined for any numerical value when its position in the total system is given with its deduction, and its relation thus determined to all the other known members of the realm of number. From the first, this relative position includes within it all other properties that can ever be ascribed to the indi vidual numbers, for all these properties follow from it and are based on it. Dedekind s explanation of the irrational numbers. This guiding thought appears in its purest form in Dedekind s well-known explana
tion of the irrational
numbers as
"cuts."
If
we
first
take as given the
totality of rational fractions, a fraction being defined as a proportion (Verhaltniszahl) and derived without appeal to measurable and
divisible
relations,
24
magnitudes from the consideration of the pure ordinal then every individual element a, which we can select
from this totality, divides the totality itself into two classes 21 and 33. The first of these classes includes all numbers which are smaller than a (i.e., which precede a in the systematic order of the whole); the second, all numbers which are than a (i.e. which follow a). If, however, the designation of any individual
"larger"
fraction implicitly involves such a division of the total system, the converse of this proposition does not hold; for not every strictly
defined,
unambiguous
division,
that can be
made
intellectually,
corresponds to a definite rational number. For example, if we consider any positive whole number D, which however is not the
square of a whole number, then

so that a positive whole
14
it
will
number
always lie between two squares can be pointed out such that
cit.,
More
particularly
cf., e.g.,
Russell, op.
144
ff.,
230.

/1
2
<D<(/1
59
I)
If
now we
into
unite
all
are smaller than
a class
21,
while
the numbers whose squares we think of all the
numbers whose squares are greater than D as united in a class 33, then any possible rational value belongs to one of these classes,
so that the division here
rational numbers.
made completely
is
exhausts the system of
evidently no element in this system, which produces this separation, and which would thus be greater than all the numbers of the class 21, and smaller than all numbers of the class 33. We have thus by means of a
Nevertheless, there
conceptual rule (by the side of which any number of others could be placed) reached an entirely sharp and clear relation between classes of numbers, which is nevertheless represented by no individual
numerical value in the manifold as hitherto denned. It is this cir cumstance which now occasions the introduction of a new "irrational"
element; an element which has no other function and meaning than
to
represent
The new number,
conceptually this determinateness of division itself. in this form of derivation, is thus not arbitrarily
conceived, nor is it introduced as a mere "symbol;" but it appears as the expression of a complex whole of relations, which were first deduced with strict logic. It represents, from the beginning, a
system of relations and into such can be again resolved. objection has frequently been raised against Dedekind s deduction, both from the side of philosophy and from that of mathe The exist matics, that it involves an indemonstrable assumption. ence of one and only one numerical element, in every case of complete division of the system of rational numbers, is not proved but merely asserted on the ground of a general postulate. In fact, Dedekind s
definite logical
The
exposition suggests this consideration in so far as it takes its start from geometrical analogies, with the purpose of clarifying the funda
The continuity of the straight line, it is explained, the fact that when all the points of a straight line are expressed by divided into two classes, in such a manner that every point of the
mental thought.
is
first class lies
to the left of every point of the second class, there exists one and only one point of the straight line which produces this division
of all the points, this cutting of the line into two pieces. 25 The assumption of this property of a line is characterized by Dedekind
himself as an axiom,
25
by which we
first
recognize the continuity of the

2,
Dedekind, Stetigkeit und irrationals Zahlen, Ed.
Braunschweig, 1892,
p. 9ff.
60
in general, space has a into it. not necessarily continuous; innumerably many of its properties w ould remain the same even if it were dis continuous. And even if we were sure that space were discontinuous, there would be nothing to hinder us from making it continuous by intellectually filling in its holes, if we chose to do so; this filling in,
line,
"If,
by which we read continuity

it
real existence, then
is
r
however, would consist in the creation of new individual points and would be done according to the above principle." 25 Such an opposi can indeed occasion the thought that no and tion of
"ideal"
"real"
conceptual determination, which forces itself
upon us
in
understanding
the realm of number, need thereby involve a determination of being. The advance from an ideal systematic connection to the existence
of a
new element seems
to involve a
jueTct/3a<ris
ets
aXXo
ykvos.
In
truth, however, our concern here is not with an unjustified transition, for in the realm of number, at least, the whole dualistic separation of ideal
and
real being, of
"essence"
and
"existence,"
is
irrelevant.
Even
if,
in the case of space, such a separation
between the product of

"is"
free geometrical construction
and what space
in the nature of
things can possibly be maintained, nevertheless in the field of pure number it loses all meaning. No number (the whole numbers as
little
as the fractional
it is
and
irrational
numbers)
"is"
anything other
than
made
in certain conceptual definitions.

is
"cut"
The assumption
that wherever there

rational
a complete division or (Schnitte) of the there "exists" one and only one number, which corresponds to it, thus implies no questionable meaning. What is here determined with absolute definiteness is primarily
number system
the division
classes
21
itself.
When
by any
the rational system
is
divided into two
sort of conceptual rule, we can decide with absolute certainty regarding any of its elements whether it belongs to one or the other class; and further show that this alterna
and
33
tive leaves
is
no member out of account, i.e., that the resulting division as such has thus indubitable complete and exhaustive. The logical "reality," which does not have to be conferred upon it by a
"cut"
postulate. Furthermore, the order in which the different "cuts" follow each other is not arbitrary but is definitely prescribed for them by their original concept. call the first of the two "cuts"
We
(H, 33)
larger than the second
(|[, 33)
21
when an element a can be

division
indicated that belongs to class

"
of the first
and
class
Idem, p.
12.

33 of
61
criterion exists for
"cuts."
the second.
Thus a
fixed
and universal
determining the serial order of the individual Thereby the forms thus produced receive the character of pure number. For number, in its original meaning, possesses no specific character but is merely the most general expression of the form of order and series in general; thus wherever such a form exists the concept of
number
finds application. The "cuts" may be said to numbers, since they form among themselves a strictly ordered manifold in which the relative position of the elements is determined according to
"be"
a conceptual rule. Thus in the creation of the
new
irrational elements,
"between"
we
the
are not
concerned with the notion that somehow
known
members
is
system supposed or assumed. But we are concerned fact, in itself meaningless and unintelligible. with the fact that there arises, on the basis of the originally given totality, another more complex system of serially arranged deter minations. This system includes the previous totality and takes it up into itself; for the characteristic mark of succession which belongs
to the
"cuts"
of the
of rational
numbers the being of other elements This way of stating the question is, in
holds directly of the rational numbers themselves,

"cuts."
which can all be understood and represented as Thus here an inclusive point of view is found from which the relative position
of all
members
of the old, as well as of the
new system,
is
determined.
We see how the fundamental thought of the ordinal theory of number

is here verified. The notion must be given up that number arises from the successive addition of unities and that its true conceptual nature is based on this operation. Such a process contains indeed one principle from which ordered wholes result, but in no way contains
the only principle for the production of such wholes. The introduc tion of the irrationals is ultimately nothing but the general expres sion of this thought it gives to number the whole freedom and scope
:
method for the production of order in general, by virtue of which members can be posited and developed in ordered sequence, without of the The conceptual limiting it to any special relation. individual number disappears gradually and plainly in its peculiar
of a
"being"
On the ordinary interpretation, with which "function." Dedekind s deduction is at first connected, although a certain number, in a system, none the less given and at hand, produces a definite the process is finally reversed, for this production comes to be the
conceptual
"cut"
62
necessary and number at all.

sufficient condition of
our speaking of the existence of a
complex, for
it
it
The element cannot be separated from the relational means nothing in itself aside from this complex, which
it
brings to expression, as
were, in concentrated form.
The problem of the transfinite numbers. The general thought, on which the formation of number rests, takes a new turn when we pass from the field of finite number to that of transfinite number. Here
problems accumulate; for the concept of the here the center of discussion, has always belonged no less to the domain of metaphysics than to that of mathematics.
specifically philosophical
infinite,
which
is
Thus, when Cantor, in the course of his decisive investigations, created the system of transfinite numbers, he conjured up again all the scholastic oppositions of the potential and actual infinite, of the
and the indefinite. 27 We seem here finally to be forced from the question of the pure meaning of the concept for knowledge to the problem nf aV>gnlnfp he^ng and its properties. The concept of the infinite seems to mark out the limits of logic and the point at
infinite
which
it
comes
in contact with another field that lies outside of its
sphere.
The concept of "power." Nevertheless the problems, which lead to the creation of the realm of transfinite numbers, issue with absolute
necessity from purely mathematical presuppositions. They arise as soon as we generalize the fundamental concept of "equivalence," which from the first forms the criterion for the numerical equality of
groups, in such a way as to make it applicable to infinite wholes. wholes, whether the number of their elements is limited or are equivalent or of the same "power" (Machtigkeit), unlimited,
finite
Two
when
their
members can be mutually coordinated one
to one.
The
application of this criterion can manifestly not proceed, in the case of infinite groups, by coordinating the elements individually with each other but rather presupposes that a general rule can be given,
by which a complete correlation is established that can be surveyed at a glance. Thus we are sure that for every even number 2n there corresponds an odd number, 2n + 1, and that, if we let n assume all
the possible values of whole numbers, the two groups, of even and odd numbers, are exhaustively correlated in a one to one manner. The concept of "power," which is thus introduced, first gains a
27
Cf. especially Cantor,

f.
Zur Lehre vom
Transfiniten.
Gesammelte Abhandl.
a. S.
aus der Zeitschr.
Philosophic u. philos. Kritik.
Halle
1890.
THE COXCEPT OF NUMBER

specifically
63
seen that
it is
mathematical interest when
it
is
itself
capable of differences and degrees. If we say that all wholes whose elements can be coordinated in one to one fashion with the series of
natural numbers belong to the first "power," then the question arises whether the totality of possible manifolds is exhausted in them or whether there are groups of another character, as regards the specified property. The latter is the case, in fact, as can be proved. For, while the advance from the positive whole numbers to the totality of rational numbers produces no change in "power," and the same is true when we go from the system of rational numbers to the system of algebraic numbers, nevertheless, the system assumes a new character
when we add the

extend
it
totality of transcendental
numbers to
it
and thus
to the group of real numbers in general. This manifold thus represents a new level, which rises above the former level; for,
on the one hand, it includes systems of the first "power" within itself, while, on the other hand, it goes beyond them, since when we attempt
to coordinate
its
elements with those of the series of natural numbers
there always remains an infinity of unconnected elements. 28 introduction of the transfinite numbers ai and a is merely to
tain this characteristic difference.
The main The new number here means
nothing else than a new point of view, according to which infinite systems can be arranged. A more complex group of distinguishing
characteristics results
when we
place along side of the transfinite
which are limited to giving the "powers" of infinite groups, the coordinate system of ordinal numbers, which arises when we no longer compare the groups in question merely with regard to the number of their elements, but also with regard to the
cardinal numbers,
position of the elements in the system.
We
ascribe to the well-
ordered groups 29
of
order"
and
the same ordinal

of
number
or the
same
both can be mutually co "type ordinated one to one with each other, the sequence which holds for both being retained. Thus if E and F are elements of M, and EI
when the elements
and FI are the corresponding elements of N, the

28
relative position of
For more detailed exposition, cf. my essay, "Kant und die moderne (Kant-Studien XII, 21 ff.); for all particulars, the literature there cited should be consulted well as Cantor s presentation in the Mathemat. Annalen. 29 For the definition of "well-ordered groups," cf. Cantor, Grundlagen einer
Mathematik"
as"
allgemeinen Afannigfaltigkeitslehre,
2.
64
SUBSTANCE AND FUNCTION and
in the succession of the first
relative position of
other words, if E 30 FI in the second group. Thus, while in the comparison of the of two manifolds use can be made of any possible arrange "powers"
group is in agreement with the EI and FI in the succession of the second. In precedes F in the first group, then EI must precede
ment of their members, in establishing their "type bound to a certain prescribed kind of succession.
that
all series,
of
order"
If
we are now we say
that can be correlated in one to one fashion according
to this condition with the sequence of natural numbers, belong to the "type of order" co, then we can, by adding to such series in their
2 or 3 members, form series of the types co 2, 1, 3; and furthermore by uniting two or more systems of the nco, so that type o} we can create the type of order #co, &o, by further application of this procedure, we can go on to the produc w co *, or indeed to co And these tion of types co 2 co 3 etc.
totality
o)
1,
<o
cow",
by no means introduced here as mere arbitrary symbols but are signs of conceptual determinations and differences, that are actually given and can be definitely pointed out in the field of infinite groups. The form of enumeration also is only an expression of a necessary logical differentiation, which first gains clear and adequate
are
interpretation in this form.
The production of transfinite numbers. In this type of deduction, the metaphysical problems of the actual infinite fall completely into the background. It has been rightly emphasized 31 that our concern
with the new forms of number is not so much with "infinite numbers" as with "numbers of something infinite;" that is, with mathematical
expressions, which we create in order to grasp certain distinctive characteristics of infinite totalities. The conflicts that result from
the connection of the concepts "infinity" and "reality" are not involved here, where we move entirely in the realm of ideal con
These conflicts may be represented in two-fold form, according as they are regarded from the side of the object or from that of the subject, from that of the world or of the activity of the
structions.
knowing
ego.
From
is
is
the
first
actual infinite
shown by the
directed
point of view, the impossibility of the fact that the objects, upon which the
act of enumeration
30
31
and which, as
it
seems,
it
must pre-
S.
Cantor, op. cit., 2, p. 5. Kerry, System einer Theorie der Grenzbegriffe, Leipzig und Wien, 1890,
f.
p. 68
65
suppose, can only be given in finite numbers. No matter what breadth and scope we may ascribe to abstract number, whatever is
for
counted must always be thought as enclosed within certain limits, it is not accessible to us otherwise than by experience, which advances from individual to individual. Viewed from the other
it is
point of view,
tion,
the psychological synthesis of the act of enumera

;
which would exclude the actual infinite no "finite understanding" can actually survey an unlimited number of unities and add them But both objections are unjustified with successively to each other.
reference to the
"transfmite,"
when we
limit the latter to its strictly
mathematical meaning.
disposal
is
The
"material"
of
enumeration at our
unlimited, for it is not of an empirical but of a logicoconceptual nature. It is not assertions concerning things that are to
be collected, but judgments concerning numbers and numerical con cepts; thus the "material," which is presupposed, is not to be thought
of as outwardly given but as arising by free construction. Just as little is there demand for the psychological processes of particular,
isolated acts of presentation and their subsequent summation. The of the transfinite implies rather the opposite thought: it concept
from
represents the independence of the purely logical import of number Even in the "enumeration" in the ordinary sense of the word.
grounding of the irrational numbers, we could not avoid considering infinite classes of numbers, such as could be represented and surveyed in the totality of their elements only by a general conceptual rule and could not be counted off member for member. The new cate gory of number gives this fundamental distinction its most general Cantor expressly distinguishes the "logical function," significance. on which the transfinite is based, from the process of successive construction and synthesis of unities. The number co is not the result of such a perpetually renewed addition of particular elements, but is meant to be merely an expression for the fact that the whole unlimited system of the natural numbers, in which there is no "last
given in its natural succession according to its law." indeed permissible to think of the newly created number co as the limit toward which the numbers 1, 2, 3, 4, ... v tend, if we thereby understand nothing else than that co is to be the first whole number that follows upon all the numbers v; that is, is to be
member,"
"It
"is
is
called greater
tion,
The logical func than any of the numbers v which gives us co is obviously different from the first principle of
.
66
generation; I call
it

the second principle of generation of real whole same more closely thus: when any definite
numbers and
exists
define the
succession of defined real whole
numbers
is
given,
is
among which
there
no one that
is largest,
new number
produced, on the basis
of this second generating principle, which is to be regarded as the limit of those numbers, i.e., as the first number greater than all of them." 32
The second
"second
"principle
of
generation"
of
number (Cantor}.
This
principle of
generation" is
ultimately only permissible and
fruitful
because it does not represent an absolutely new procedure, but merely carries further the tendency of thought that is unavoid able in any logical basing of number. Consideration of the properties of external things, such as those of particular psychical contents and
acts of presentation, proved them to be incapable not only of con structing the series of the "natural" numbers in their lawful order but of rendering them intelligible. Even here it was not the bare
addition of unity to unity, which controlled the formation of concepts; it rather appeared that the individual members of the series of
numbers could only be deduced in their total extent by considering one and the same generating relation, held to be identical in content through all variations in particular application. It is this same thought, which now attains sharper formulation. Just as the endless multiplicity of natural numbers is ultimately posited by one concept, i.e., according to a universal principle, so its content can be drawn together again in a single concept. For mathematical thought, the fundamental relation, that includes within itself all the members that proceed from it, becomes itself a new element, a kind of fundamental unity, from which a new form of number-construction takes its start.
The whole
by a
law,"
endless totality of natural numbers, in so far as it is "given i.e., in so far as it is to be treated as a unity, becomes the
starting-point for a
arise other
new
construction.
From
the
first
order there
and more complex orders, which use the former as material basis. Once more we see the liberation of the concept of number from the concept of collective unity. To seek to understand and represent the "number" co as an aggregate of individual unities would be non On the other hand, sensical, and would negate its essential concept. the ordinal view is verified here for in the concept of a new construc
:
tion following
32
upon
all
elements of the series of natural numbers,

11, p. 33.
Cantor, Grundlagen,

there
is
67
is no contradiction, as long as it is remembered that this totality to be logically surveyed and exhausted in a single concept. The problem of the infinity of time can also be left out of con
sideration at
first.
For the meaning
of
"succession"
in a series
is
independent of the concrete temporal succession. Thus three does not follow upon two in the sense of a succession of events, but the
relation merely points out the logical circumstance that the definition of three "presupposes" that of two; the same holds in a still stricter
sense for the relation between the transfinite and the finite numbers.
That the number
co is
"after"
all
the
finite
numbers
of the series of
natural numbers, means ultimately merely this same sort of concep tual dependence in the order of grounding. The judgments, into which the transfinite enters, prove to be complex assertions, which are
reduced by analysis to relative determinations of infinite systems In this sense, a thorough-going conceptual "natural" numbers. between the two fields. The new constructions continuity prevails are "numbers" in so far as they possess in themselves a prescribed
of
serial
form, and hence obey certain laws of connection for purposes of calculation, which are analogous to those of the finite numbers,
33 though they do not agree with them in all points. Thus the new forms of negative, irrational and transfinite numbers are not added to the number system from without but grow out of the continuous unfolding of the fundamental logical function that was effective in the first beginnings of the system. A new direction in principle comes in, however, as soon as we advance from the complete and closed system of the real numbers to the more complex
According to the "metaphysics of the imaginary," which Gauss founded and developed, we are no longer concerned here with the establishment of the most general laws of the order of a single series, but rather with the unification of a plurality of series, of which each one is given by a definite generating relation. With this
to a multidimensional manifold, there appear logical problems, which find their complete formulation beyond the limits of the pure doctrine of numbers in the field of general geometry.
number systems.
transition
33
For the arithmetic of the

286, 294.
transfinite, cf.
more particularly
Russell, op.
cit.,
CHAPTER
III
THE CONCEPT OF SPACE AND GEOMETRY
The
logical
transformations
of
the
concept
of
number
are
dominated by a general motive, which constantly gains more definite The meaning of number is first grasped completely expression. when thought has freed itself from seeking in concrete experience a In its most general mean correlative for each of its constructions.
number reveals itself as a complex intellectual determination, which possessgyjoîmmediate sensible copy in the properties of the Necessary as it is in modern analysis and algebra physical object. to carry through this development, still it might appear that it repre sents only a technical detour of thought and not the original and
ing,
natural principle of the scientific construction of concepts. This principle seems only to come to light in its purity where thought does
realm of number, act exclusively according to selfits value and support in intuition. Precisely here then lies the critical decision for every logical theory of the concept. A conceptual construction may be spun out ever so subtly and consistently from presuppositions; nevertheless it seems empty and meaningless so long as it does not deepen and enrich our intuition.
not, as in the
created laws but finds
But
if
we adhere
to the criterion of intuition, the opposition of the
logical theories to each other appears in a
new
light.
The model,
which theory must follow, lies henceforth not exclusively in algebra Not number, but the concepts of but, in purer form, in geometry. space, because of their immediate relation to concrete reality, must
serve as the type.
Concept and form. In the historical beginnings of logic this fact is most evident. Concept and form are synonyms; they unite without
meaning of eidos. The sensuous manifold is ordered and divided by certain spatial forms, which appear in it and run through all diversity as permanent features. In these forms we possess the fixed schema by which we grasp in the flux of sensible
distinction in the
things a system of unchanging determinations, a realm of "eternal Thus the geometrical form becomes at once the expression being."
68

and the confirmation
of the logical type.
is
69
The principle of the logic confirmed from a new angle and this time it is neither the popular view of the world nor the grammatical struc ture of language, but the structure of a fundamental mathematical
of the generic concept
;
science
upon which
it
rests.
Just as
we
recognize as identical the

size it
outline of the visible
form
in
whatever sensuous material or
may appear, so in general we are able to grasp the highest genera, to which existence owes its uniform structure and the constant recur
rence of
its definite
features.
The method of ancient geometry. These relations have not been exclusively significant with regard to logical problems they have also been decisive in the scientific evolution of geometry. The synthetic
;
geometry of antiquity
that
is
is
dominated by the same fundamental view
given general expression in formal logic. The genera of existence can only be clearly grasped when they are strictly dis tinguished from each other and limited to a fixed circle of content.
form possesses an isolated and unchanging first is not directed so much upon the unity of the forms as upon their strict differentiation. The view, indeed, that the problem of change was in general alien to the mathe matical spirit of the Greeks has been more and more completely
also each geometrical
Thus
character.
The proof
at
by investigation of the historical sources. Not only did they grasp the concept of number, including that of the irrational, in all sharpness, but the Ephodion of Archimedes also shows with great
refuted
clarity how completely Greek thought, where it advanced freely in methodic discovery, was penetrated by the concept of continuity and thus anticipated the procedure of the analysis of the infinite
itself.
1
But
precisely
when we
realize
this,
the difference that
remained between the method of discovery and that of scientific The scientific exposition is exposition becomes the more manifest. influenced by certain logical theories, from which it cannot entirely free itself. As circle and ellipse, ellipse and parabola do not belong to the same visible type, it seems that they cannot, in a strict sense, fall under the unity of a concept. However much the geometrical judgments, which we can make about the two fields, meet and corre
spond
in content, nevertheless there is here only a
secondary similar-
1 Cf. the exposition of Max Simon, Geschichte der Mathematik im Altertum in Verbindung mil antiker Kulturgeschichte, Berlin, 1909, especially pp. 256,
274
ff.,
373.
70
and not an original logical identity. The proofs of the two types of assertion are in each case strictly separate they gain their validity
ity
;
and necessity only because they are individually derived from the concept in question and its specific structure. Every difference in the arrangement of the given and sought lines of a problem presents a
new problem in regard to the proof; to every difference in the total sensuous appearance of a figure corresponds a difference in interpreta A problem, which modern synthetic geometry tion and deduction.
solves by a single construction, was analysed by Apollonius into more than eighty cases, differing only in position. 2 The unity of the constructive principles of geometry is hidden by the specialization of its particular forms of which each is conceived as irreducible. The concept of space and the concept of number. The transformation of geometry in modern times begins with an insight into the philo It was not by chance that the new sophical defect of this procedure. form of geometry, although anticipated, especially by Fermat, The reform first gained definitive formulation through Descartes. of geometry could only be carried out after a new ideal of method had been clearly conceived. The method of Descartes is everywhere directed toward establishing a definite order and connection among all particular expressions of thought. It is not the content of a given that determines its pure cognitive value, but the necessity by thought which it is deduced from ultimate first principles in unbroken sequence. The first rule of all rational knowledge is then that cognitions be so arranged that they form a single self-contained series within which there are no unmediated transitions. No member can be introduced as an entirely new element, but each must issue step by step from the earlier members according to a certain rule. What ever can be an object of human knowledge is necessarily subject to this condition of continuous connection, so that there can be no question, however remote, which we are not able to master fully by this methodical progress from step to step. This simple thought, on which the Discours de la Methode is founded, at once demands and conditions a new conception of geometry. Geometrical knowl edge in the strict sense is not found where the particulars are studied
2 See Reye, Die synthetische Geometric im Altertum und in der Neuzeit (Jahresberichte der Deutschen Mathematik. Vereinigung. XI. (1902) p. 343 ff. Cf. also my work, Leibniz System in seinen wissensch. Grundlagen, Marburg,
1902, p. 220
ff.
71
as isolated objects, but only where the totality of these objects can e constructively generated according to a given process. Ordinary synthetic geometry violates this postulate; for its object is the isolated
spatial form whose properties it grasps in immediate sensuous intui tion but whose systematic connection with other forms it can never
completely represent. At this point, we are led with inner philosophi cal necessity to the thought of the completion of the concept of space through the concept of number. The notebook of Descartes, which
reveals the development of his fundamental thought, contains a "The sciences in their present characteristic expression of this:
condition are masked and will only appear in full beauty when we remove their masks; whoever surveys the chain of the sciences will find them no more difficult to hold in mind than the series of 3 numbers." This is the goal of the philosophical method: to con ceive all its objects with the same strictness of systematic connection
as the system of numbers. From the standpoint of the exact sciences in the time of Descartes, this is the only manifold which is built up
from a self-created beginning according to immanent laws, and thus can conceal within itself no question in principle insoluble for thought. The demand that spatial forms be represented as forms of number and
be wholly expressed in the latter, may appear strange when regarded from the standpoint of the Cartesian ontology; for in this, "extension" signifies the true substance of the external objects and is thus an But here the analysis original and irreducible condition of being. of being must be subordinated to the analysis of knowledge. We can only bring space to exact intelligibility by giving it the same Number logical character as hitherto belonged only to number. is not understood here as a mere technical instrument of measure ment. Its deeper value consists in that in it alone is completely fulfilled the supreme methodological postulate, which first makes
knowledge knowledge. The conversion of spatial concepts into numerical concepts thus raises all geometrical enquiry to a new intellectual level. The substantial form-concepts of ancient geome which remain opposed to each other in bare isolation, are changed try, by this into pure "serial concepts" which can be generated out of each other by a certain fundamental principle. The scientific
discovery of analytic geometry rests, in fact, upon a true philosophi cal "revolution." Traditional logic seems impregnable as long as
*
Descartes, Oeuvres inedites, pub. by Foucher de Careil, Paris, 1859, p.
4.
72
is at its side as an immediate confirmation and embodiment of its principles; the extension of geometry first makes room for a new logic of manifolds that extends beyond the
ancient synthetic geometry
limits of the syllogism.
The fundamental principle of analytic geometry. This connection appears even more clearly when we consider the special develop
ment which
analytic geometry received from Descartes.
Here
it
appears that the apparently individual form of exposition contains features of universal significance which, in another guise, run through
the whole philosophical history of geometry. The fundamental concept upon which Descartes founds his considerations is the concept of movement. From the standpoint of the older interpretation, there
a problem even in this. For only the individual figure, which stands before us in fixed, closed limits, seems accessible to exact intellectual treatment, while the transition from one form into
is
another seems to force us back into the chaos of mere perception, yînto the sensuous realm of "becoming." It might at first appear
that recognition of the concept of motion introduced a not entirely rational element into the Cartesian geometry, contrary to its real
Motion leads at once to the question of the moving but does not this subject presuppose a material body, "subject;" thus a purely empirical element? This doubt vanishes, however,
tendency.
when we analyse in detail the function here ascribed to the concept of The various forms of plane curves arise by our prescribing to a given point, which we make a fundamental element, various kinds of movement relative to a vertical and a horizontal axis. From the unification of these types of movement must be completely deduced the characteristics of lines generated in this way as of points. Here, as we see, movement does not signify
motion.
"paths"
a concrete, but merely an ideal process; it is an expression of the synthesis by which the successive manifold of positions, that are connected by any law, are brought into the unity of a spatial form.
Here the concept of motion, as previously the concept of number, serves simply as an example of the general concept of series. The individual point of the plane is first determined by its distance from
two
fixed lines,
totality of possible positions.
and gains hereby its fixed systematic place within the These point-individualities, which are
characterized by definite numerical values, do not merely stand next to each other but are differently related, according to various complex
73
rules of arrangement, and are thus brought to unified forms. The of the "movement" of the points is only the sensuous representation
symbol
of this logical act of arrangement.
The
intuitive geometri
cal line is thus resolved into a
connected with each other by

sensible properties
pure succession of numerical values, a certain arithmetical rule. All the
by which we distinguish one line from another, as for example, constancy or change in direction and curvature, must, in so far as they can be given exact conceptual expression, be expres
The concept of motion sible as peculiarities of these series of values. thus does not serve the purpose of pictorial representation but rather that of progressive rationalization; the given form is destroyed that
it
may
rise
anew from an arithmetical
serial law.
How
strictly this
general postulate is maintained is seen in characteristic fashion in Descartes exposition, for it is this postulate that defines and limits the field of geometry itself. "Transcendental" curves are excluded
because in their case, with the technical means at Descartes the postulated deduction from the relations of purely disposal, numerical rules, seems impossible. These curves, which in their intuitive construction have no exceptional position, are ruled out of
geometry because they can not be brought under the new definition of the geometrical concept, by which the concept is ultimately
reduced to a system of elementary operations of calculation. The infinitesimal geometry. Here, however, appear the limits of
Cartesian geometry, which had to be transcended in later historical development. A new ideal of knowledge was affirmed; but this ideal was not able to include all the scientific questions that had
hitherto been united under the
tual construction
name
of geometry.
Rigor of concep
had to be purchased through the exclusion of important and far-reaching fields. The path of logical progress was
resolution of spatial concepts into serial concepts remained the guiding standpoint; but the system of serial concepts had to be deepened and refined so that not merely a
thus clearly prescribed.
The
narrow selection but the whole field of possible spatial forms could be surveyed and mastered. Because of this demand, Cartesian geome try develops with inner necessity into infinitesimal geometry. Here in infinitesimal geometry first appears in more perfect form the new conceptual construction which can be recognized in its general outlines in analytic geometry. The procedure here also starts from
a fundamental series
xi
x2
...
xn
which
is
coordinated,
ac-
74
cording to a definite rule, with another series of values y\ jz But this coordination is no longer limited to the yn ordinary algebraic procedures, to the addition and subtraction, multiplication and division of numbers or groups of numbers, but
.
. . .
all possible forms of the dependency of magnitudes accord ing to law. The concept of number is fulfilled in the general concept of function, which permeates it; and their cooperation first enables
includes
us to develop the whole content of geometry in logical perfection. In the advance to differential geometry there appears, however, a new decisive moment. It is an infinite manifold of coordinations from
whose unification the curve
first results
method
of infinitestimal analysis first clearly
as a conceptual whole. The shows that this infinity
of determinations does not destroy all determinateness, but that it is possible rather to unify the determinations again in a geometrical
As in analytic geometry the individual point of the plane is essentially determined by the numerical values of its co ordinates x and y, so, by the differential equation f(x, y, y ) = 0,
conception.
to every point thus given is co6rdinated a certain direction of move ment, and the problem consists in reconstructing from the system of
these directions the whole of a given curve with all peculiarities of its geometrical course. The integration of the equation signifies the synthesis of these infinitely numerous determinations of only
and connected structure. In the same way, = co a differential equation of the second order f(x, y, y/ ordinates to each point and its direction of movement a certain
direction into a unitary
y")
form
radius of curvature, whereby the problem arises of deducing the of the curve as a whole from the totality of values of curvature
thus gained. 4 The elements, which are here indicated geometrically by the concept of direction and curvature, are in their most general
expression obviously nothing else than simple serial principles, which we comprehend in their totality and in their transformation accord
ing to law.
If
we
measured by a moving body as represented by the integral

velocities, the process that
consider, in the sense of infinitesimal analysis, the space of its
we apply
consists in reading into every
moment
of the actually progressing movement a certain law of progress, by which the transition to the following points of space
4 Compare, F. Klein, Einleitung in die hohere Geometric, Autographierte Vorlesung, Gottingen, 1893, I, 143 ff.
THE CONCEPT OF SPACE AXD GEOMETRY
75
is to be exactly determined. The "velocity/ which a body possesses at a given point of its path at a given moment of time, can only be conceived and represented by the comparison and reciprocal rela
and time values. Logically considered, no absolute property of the moving thing but merely an expression of this reciprocal relation of dependence. We assume that the body, if at a given point all outer influence upon it ceased, would
tion of a series of space values
velocity
is
move forward uniformly in a definite way, that is, that after a certain time ti had elapsed the distance Si would have been traversed, after the time t 2 = 2ti had elapsed the distance 2s i would have been In all this we are not concerned with indicating the traversed, etc.
real
movement
it
of the
body by giving the
particular places through
passes, but with construing its path purely ideally according to the various laws of the possible coordination of points of space and
which
points of time. The individual values within the various series are never factual, for the uniformity of movement is never actually
realized; nevertheless,
thought necessarily uses these hypothetical values and series of values in order to render the complex whole,
The same holds true of i.e., the real path, completely intelligible. the procedure used in the analysis of the infinite in geometry. The curve is here also conceived as a certain order of points; but this order,
which, as immediately given, is a highly involved serial form, is conceptually analysed as a manifold of simpler laws of serial order,
that mutually determine each other. The concrete form is analysed into a system of virtual grounds of determination, which are assumed
to be different from point to point. The geometrical form, which, from the standpoint of direct intuition and that of elementary syn
seems to be something absolutely known and imme The diately comprehensible, appears here as a mediated result. form is as if resolved into manifold strata of relations, which are superimposed upon each other and which, by the definite type of dependence among them, finally determine a single whole. Magnitudes and functions. Herewith, however, is revealed a prob lem of comprehensive significance. The construction of curves out of
thetic geometry,
the totality of their tangents, as shown in infinitesimal geometry, is only an example of a procedure of more general applicability. All
mathematical conceptual construction sets itself a double task, in fact, the task of the analysis of a certain relational complex into
elementary types of relation and the synthesis of these simpler types
76

of construction into relations of higher orders.
is
and laws
The
analysis
and complete expression of this For even here mathematical investigation intellectual tendency. advances beyond the mere consideration of magnitudes and turns to
of the infinite
logically a first
a general theory of functions. The "elements" here joined into new unities are themselves not extensive magnitudes which are combined of a whole, but are forms of function which reciprocally as
"parts"
determine each other and unite into a system of dependencies. How ever, before we can follow this development, which gives mathematics
its
peculiar character,
we must turn back
to the special problems of
geometry; for in the philosophical struggles concerning the geometri cal methods, there clearly appear the beginnings of a new and univer sal formulation of the logical question.
ii
Intuition
and thought in
first
Modern geometry
field
the principles of the geometry of position. attains a strictly logical construction of its
and true freedom and universality of method in advancing from the geometry of measure to the geometry of position. In relation to the analytic geometry of Descartes, this step seems to signify a reaction. Intuition again asserts its claim as in ancient synthetic
geometry.
seek
to
It is
not when we limit intuition as

it
replace
by mere operations
truly logical and strictly deductive of space results, but when we restore
as possible and calculation, that a construction of the science

of
much
intuition
to
its
full
Thus the development leads back from the abstract concept of number to the pure concept of form. That in this there lies a new motive, in the philosophical He saw in sense, Descartes himself discovered and asserted. the methods of Desargues, which contained the first approach to a projective treatment and conception of spatial forms, an in If we follow dication of a general "metaphysics of geometry." 5
scope
and
independence.
this
"metaphysics"
further,
it
seems
in
immediate
conflict
with his
tendencies and deductions. In fact, the new interpretation could be carried through only by a stubborn struggle against the supremacy of the analytic methods. Criticism of these methods
5
own
See Descartes
letter to
Mersenne
of Jan. 9, 1639, Correspondence, ed.
Adam-Tannery,
II, 490.
THE CONCEPT OF SPACE AXD GEOMETRY
77
begins with Leibniz and is brought to a first conclusion with his founding the analysis of position. It is charged that analysis is not able to establish the universal principle of order, upon which it prides itself, within the field to be ordered, but that it is obliged to
have recourse to a point of view external to the object considered. The reference of a spatial figure to arbitrarily chosen coordinates introduces an element of subjective caprice into the determination; the conceptual character of the form is not established on the basis of properties purely within itself, but is expressed by an accidental relation, which may be different according to the choice of the assumed system of reference. Whether from among all the various equations which can be applied, according to this process, in the
expression of a spatial figure, the relatively simplest is chosen depends skill of the calculator, and thus upon an element which the strict progress of method seeks to exclude. If this defect
upon the individual

is
to be avoided, a procedure
must be found which
is
equal to the
analytic methods in conceptual rigor but which accomplishes the rationalization wholly within the field of geometry and of pure space. The fundamental spatial forms are to be grasped as they are
"in
themselves"
and understood
Poncelet.
in their
own
laws without translation
into abstract numerical relations. 6
and standpoint, there is no philosophically characteristic and significant, possibility of return to the point of view of ancient elementary
Steiner
is
and
Nevertheless from this
this
geometry. The reversion to the intuitive aspect of the figure pro duces only an apparent connecting link; for the content itself, that is now understood under geometrical "intuition," has been deepened and transformed. If, in order to gain a fixed criterion in the philo
sophical conflict of opinion, we enquire of the scientific founders of modern geometry concerning the meaning they attach to the concept and term "intuition," there appears at first a peculiarly conflicting result. On the one hand, Jakob Steiner, following his teacher and model Pestalozzi, is never tired of praising the logical justification and Steiner and his disciples see the fruitfulness of pure intuition.
defect of ordinary synthetic
intuition only in a limited sense,
geometry in that it teaches us to use and not in the whole freedom and
situs"
6 For further particulars regarding Leibniz sketch of Leibniz System in s. wiss. Grundlagen, Ch. III.
"Analysis
cf.
78
scope of Poncelet
its

meaning.
find
we
value of the
new
On the other hand, in the chief work of the opposite logical tendency expressed. The method is found in that through it geometrical
7
deduction can proceed entirely unhindered, in that, without being narrowed to the limits of possible sensuous representation, it especially takes account of imaginary and infinitely distant elements that possess no individual geometrical "existence," and thus first
attains completeness of rational deduction.
The opposition here
in
removed as soon as we thought In the case where follow the exposition on both sides more closely. the geometry of position is founded purely on intuition, the meaning
the
formulations
of
the
is
is not that of adherence to the sensuously given figure but is the free constructive generation of figures according to a definite unitary
principle.
The various sensuously
possible cases of a figure are not,
as in
Greek geometry, individually conceived and investigated, but all interest is concentrated on the manner in which they mutually proceed from each other. In so far as an individual form is con
it
sidered,
never stands for

it
itself
alone but as a symbol of the system
to which
belongs and as an expression for the totality of forms into which it can be transformed under certain rules of transforma
tion.
Thus
"intuition"
is
never concerned with the particular

is,
figure with its accidental content, but directed to the mediation of the
forms upon each other. 8

7
according to Jakob Steiner, dependency of geometrical The particular terms are here also subor-
C/., for
example, B. Reye, Die synthetische Geometric im Altertum und in
der Neuzeit, Op.

8
cit., p. 347. Cf. the preface to J. Steiner s
work:
keit geometrischer Gestalten voneinander, Berlin, 1832:
Systemat. Entwicklung der Abhangig"The present work seeks
to reveal the organization by which different kinds of phenomena are bound together in the spatial world. There are a small number of simple funda mental relations, wherein the schema is expressed from which the remaining
propositions develop logically and without any difficulty. Through proper appropriation of a few fundamental relations, we master the whole subject; order comes out of chaos, and we see how all the parts naturally fit together, forming into series in the most perfect order, and the allied parts united into
well-defined groups. In this way we succeed, as it were, in gaining posses sion of the elements from which nature starts, by which, with the greatest
possible economy and in the simplest way, innumerable properties can be ascribed to figures. In this, neither the synthetic nor the analytic method constitutes the essence of the matter, which consists in the discovery of the
of forms on each other and the manner in which their properties are continued from the simpler figures to the more complex."
dependency
79
dinated to the systematic relation that unites them. The deduction the fundamental forms expresses this in so far as, for example, the particular straight line is not defined in itself, but as an element in a pencil of rays, or the particular plane as an element in a sheaf of
itself of
rays.
In general, we see that the fundamental methodological
standpoint, which led to the discovery of analytic geometry, is not set aside here but retained and brought to a new fruitful application
The motive of number is excluded; but the general motive of series stands forth the more clearly. We saw how, with Descartes, number was not characterized as the fundamental principle because of its own content, but was only retained because it represented the purest and most perfect type of a logically ordered manifold in general. Strictness of deductive
in the field of the spatial itself.
connection seemed only able to be carried over to space by the mediation of number. (Cf. above p. 71 f .) We can understand that henceforth there must arise a new, important task continuous with
the achievements of analytic geometry.
The construction
of the
spatial forms from original fundamental relations remains as an in violable postulate, but this postulate must now be satisfied by
purely geometrical means and without
the
introduction
of
the
concepts of measure and of number.
The concept of "correlation" and the principle of continuity. The evolution, which begins here, is characterized and guided in detail by logical views. This is especially manifest in the case of Poncelet
who, in the
foundations.
conflict in
which he engaged regarding the principles of
his discipline, refers with increasing definiteness to the philosophical
In opposition to the criticism which the Parisian
especially Cauchy, made on the philosophical presupposi tions of his work, Poncelet urges with emphasis that in these pre
Academy,
suppositions we are not dealing with a secondary matter, but with the real root of the new view. He appropriates the words of Newton
that in geometry the method of discovery
means everything,
so that
once found and established, results occur of themselves as the fruit of the method. 9 The theory of the protective properties is to be no mere material extension of the field of geometry, but is to
this
is
9
when
1865,
Cf. Poncelet, Traite des proprietes projectives des figures, I, p. 365; II, p. 357.
2nd
edition, Paris,
80
introduce a
first

10 The principle of investigation and discovery. necessary step is to free geometrical thought from the
new
and
its close adherence to the particu the directly given, individual figure. Descartes charged ancient mathematics with not being able to sharpen the intellect
narrowness of the sensuous view with
larities of
without tiring the imagination by close dependence on the sensuous form, and Poncelet maintains this challenge throughout. The true It can only show synthetic method cannot revert to this procedure. itself equal in value to the analytic method, if it equals it in scope and
universality, but at the same time gains this universality of view from purely geometrical assumptions. This double task is fulfilled as soon as we regard the particular form we are studying not as itself the concrete object of investigation but merely as a startingpoint, from which to deduce by a certain rule of variation a whole system of possible forms. The fundamental relations, which charac terize this system, and which must be equally satisfied in each particu
constitute in their totality the true geometrical object. the geometrician considers is not so much the properties of a given figure as the net-work of correlations in which it stands with other allied structures. We say that a definite spatial form is correla
lar form,
What
tive to another
when
it is
deducible from the latter

its
by a continuous
transformation of one or more of
elements of position: yet in which
the assumption holds that certain fundamental spatial relations, which are to be regarded as the general conditions of the system,
remain unchanged. The force and conclusiveness of geometrical proof always rests then in the invariants of the system, not in what is
peculiar to the individual
members
as such.
It is this interpretation,
which
Poncelet
of the
is
principle of continuity,
principle
by the expression and which he formulates more precisely as the permanence of mathematical relations. The only
characterizes
philosophically
postulate that
10
involved can be formulated by saying that
it
is
"La doctrine des propriete s projectives, celle de la perspective-relief, le principe ou la loi de continuity, enfin la the orie des polaires reciproques et la the orie des transversales 6tendue aux lignes et surfaces courbes, ne forment pas simplement des classes plus ou moins etendues de problemes et de theo-
remes, mais constituent proprement, pour la Geometric pure des principes, des me thodes d investigation et d invention, des moyens d extension et d exposition, dans le genre de ceux qu on a nomme principes d exhaustion, methode des infmiment petits, etc." Op. cit., p. 5.
THE COXCEPT OF SPACE AXD GEOMETRY
81
for
possible to maintain the validity of certain relations, defined once all, in spite of a change in the content of the particular terms,
of the particular relata.
i.e.,
We thus begin by considering the figure

and do not analyse
it
in a general connection (Lags ),
in the beginning
into all its individual parts, but permit changes of them within a certain sphere defined by the conditions of the system. If these
changes proceed continuously from a definite starting-point, the systematic properties we have discovered in a figure will be trans ferable to each successive "phase," so that finally determinations,
which are found in an individual case, can be progressively extended all the successive members. The transference of relations distinguished from induction and There is clearly manifest in these elucidations of Poncelet analogy. a tendency toward an exact and universal expression of the new thought. Above all, he is concerned to guard the transference of relations, which he assumes as basic, from any confusion with merely Induction proceeds from the analogical or inductive inference.
to
particular to the universal it attempts to unite hypothetically into a whole a plurality of individual facts observed as particulars without necessary connection. Here, however, the law of connection is not subsequently disclosed, but forms the original basis by virtue of which the individual case can be determined in its meaning. The conditions of the whole system are predetermined, and all specializa tion can only be reached by adding a new factor as a limiting deter mination while maintaining these conditions. From the beginning, we do not consider the metrical and protective relations in the manner in which they are embodied in any particular figure, but take them with a certain breadth and indefiniteness, which gives them room for
;
It may seem at first surprising and paradoxical development. that this indefiniteness of the starting-point is held to be the ground of the fruitfulness of the new procedure and of its superiority over
11
the ancient methods. However, it soon appears that the expression of the thought here suffers from the ambiguity of the traditional logical terminology, in which concept and image are not strictly
distinguished, and in which the identical and clearly defined meaning of a conceptual rule always threatens to dissolve into an abstract
and schematic generic image.

11
What
it
the standpoint of image, because

Traite des proprictes projectives, p.
appears as indefiniteness from neglects the individual features

f.,
XIII
XXI
f.
82
of the picture (Bild), appears from the standpoint of the concept as the basis of all exact determination since in it is contained the univer
sal rule for the construction of the individual.
sal"
Between the
"univer
and
"particular"
there subsists the relation which characterizes
true mathematical construction of concepts; the general case does not absolutely neglect the particular determinations, but it
all
reveals the capacity to evolve the particulars in their concrete totality As Poncelet empha (See above p. 18 f .) entirely from a principle.
never the mere properties of the particular kind but the properties of the genus, from which the projective treatment of a figure takes its start; the "genus," however, here signifies merely a
sizes, it is
connection of conditions by which everything individual is ordered, not a separated whole of attributes which uniformly recur in the The inference proceeds from the properties of the individuals.
connection to those of the objects connected, from the serial princi ples to the members of the series. The peculiarity of the Projection and the imaginary in geometry.
method appears most clearly in its fundamental procedure. The most important form of correlation, by which different figures are connected, is found in the procedure of projection. The essential problem consists in separating out those "metrical" and "descriptive"
elements of a figure which persist unchanged in its projection. All the forms, that can issue in this way from each other, are considered
as an indivisible unity; they are, in the sense of the pure geometry of position, only different expressions of one and the same concept.
it is immediately evident that to belong to a concept does not depend on any generic similarities of the particulars, but merely presupposes a certain principle of transformation, which is main tained as identical. The forms, which we unite in this way into a can belong to totally different in their sensuous "group," intuitive structure; indeed, they can be deprived of any reference to such a type in so far as there is no geometrical existence, in the sense of direct intuition, corresponding to them. The new criterion
"types"
Here
of the geometrical construction of concepts is shown here in its general significance; for it is this criterion upon which the admission of the imaginary into geometry ultimately rests. In general, accord to Poncelet, three different forms of the procedure of "correla ing
can be distinguished. We can transform a certain figure, which we choose as a starting-point, into another by retaining all
tion"

its
83
parts as well as their mutual arrangements, so that the difference
consists exclusively in the absolute magnitude of the parts. In this we can speak of a direct correlation, while in the case in which case,
exchanged or reversed in the of an "indirect" correlation. figure, and this is methodologically the most interesting and Finally, important case the transformation can proceed in such a manner that certain elements, which could be indicated in the original form
is
the order of the individual parts
deduced
we can speak only
If as real parts, entirely disappear in the course of the process. for example, a circle and a straight line that intersects consider,
we
it,
then we can transform this geometrical system by continuous dis placements in such a manner that the straight line finally falls wholly outside of the circle, so that the intersections and the directions of
the radii corresponding to them are to be expressed by imaginary values. The coordination of the deduced figure with the original
figure
no longer connects elements that are actually present and

it
observable, but merely intellectual elements; into a purely ideal correlation.
has resolved
itself
But
it is
geometry
is
precisely these ideal correlations that cannot be spared if The to be given a unitary and self-contained form.
defect of the ancient methods consists in that they neglect this funda mental logical instrument, and thus consider only magnitudes of an absolute and quasi-physical existence. The new view is obliged to break with this procedure, since from the beginning it defines as the real object of geometrical investigation not the individual form in its sensuous existence, but the various species of dependency that can From this point of view, real subsist between forms. (Cf. p. 77.)
and imaginary elements are
essentially similar; for the latter also are
the expression of perfectly valid and true geometrical relations. That under definite conditions certain elements of a figure disappear and
cease to exist, this contains a fruitful
is
in itself no merely negative knowledge but and thoroughly positive geometrical insight.
Further, the imaginary intermediate members always serve to make possible insight into the connection of real geometrical forms, which
without this mediation would stand opposed as heterogeneous and

unrelated.
It is this ideal force of logical connection, that secures
them
full
right
to
"being"
in
it
imaginary
subsists, in so far as
a logico-geometrical sense. The fulfills a logically indispensable func
tion in the system of geometrical propositions.
The only
it
"reality,"
which we can
intelligibly
expect and demand
of
consists in the truth
84
it

contains, in its relation to the valid propositions and judgments it brings to expression. Here in the realm of geometry is
which
repeated the same process that we are able to follow in the realm of number; from the retention of definite relations arise new "elements,"
essentially similar and equal to the earlier, since these also have no deeper or firmer basis than consists in the truth of relations. (Cf.
above
If
p.
60
ff.)
consider a simple example from ordinary geometry, viz., of two circles in a plane, then in the case in which they intersect,
we
is
given in the straight line, that connects their two points of The points of intersection, a new structure of definite properties.
there
this straight line,
which we
call
the
"common
chord"
of the
two
circles, are distinguished by the fact that the tangents drawn from them to the circle are equal to each other. The geometrical rela
tion thus established can also be traced
and expressed
in the case in
which the circles no longer intersect, but fall completely outside of each other. In this case, also, there always exists a straight line, the so-called "radical axis" of the two circles, which satisfies the essential condition previously mentioned, and in this sense can be called the ideal common chord of the two circles, which contains their
two
element
points of intersection. Thus a certain intuitive expressed here and completely replaced by certain con ceptual properties belonging to it and this logical determination is also retained after the substratum, in connection with which it was
"imaginary"
is
first
of the relation
proceed from the persistence discovered, has disappeared. and create by definition in the imaginary "points"
of
We
The fruitfulness is predicated. seen in that a systematic connection between forms is thereby established, such as permits us to carry over proposi tions, which are discovered and proved in connection with one
the
"subjects,"
which the relation

is
of this procedure
form, to another in connection with which they were not immediately obvious. 12 Along with the particular relations of content, there are
12 Thus, e.g., if any three circles in a plane are given, and we construct for each two of them the "radical axes," it can easily be shown that the three lines so arising intersect in one point; from this it results further that the same also holds, in the special case of three common chords of really inter secting circles, etc. The real properties of real common chords are thus dis covered and grounded by reference to the chords. Cf. Chasles, Apergu historique sur I origine et le developpement des methodes en Geometrie, 2nd. ed., Paris, 1875, p. 205 ff.; also Hankel, Die Elemente der projektivischen
"ideal"
Geometrie, Leipzig, 1875, p. 7
ff.

above
real"
85
all
certain universal formal determinations
by which the
"un
"real"
elements, which geometry produces, are connected with the The principle of the "permanence of formal laws," points.
it
even before
was used
in algebra as the justification of the general
number, was introduced and founded by Poncelet from a purely geometrical standpoint. The infinitely distant point in which, according to the projective theory of space, two parallels intersect, and the infinitely distant straight line in which two parallel
ized concepts of
planes intersect, are logically justified conceptual constructions, not only because they represent concentrated assertions about definite
relations of position, but because these
new
constructions are
com
pletely subjected to the geometrical axioms, as can be shown, in so far as they do not refer to relations of measure. Here a higher point of view is found, which is equally just to the and "unreal"
"real"
distinctness, are in their object but they are nevertheless thoroughly paradoxical logical in their structure, in so far as they lead to strict and incon testable truths. 13
points.
The new elements, Poncelet says with
Metrical and projective geometry, and the quadrilateral construction
The development of projective geometry, which cannot be of Staudt. followed here in detail, has thus brought the philosophical principles upon which it is founded to more and more explicit expression. To
the extent that the geometry of position is built up from independent assumptions, the general logical character and meaning of the new method becomes evident. The constructive process, by which we
generate in strict deduction the whole of projective space from the simple concepts of the point and the straight line, begins with the consideration of harmonic pairs of points. Thus in the first phase of
projective geometry, the harmonic position of four points on a straight line is at first introduced exclusively by means of the concept of the
when the
13
double proportion: the points a, 6, c, d form a harmonic sequence relation of the distances a 6 to 6 c is the same as that of the
For the whole matter, compare Poncelet, Considerations philosophiques
le et
techniques sur
principe de continuite dans les lois geometriques, section III. (Applications d Analyse et de Geometrie, Paris 1864, p. 336 ff .), as well as Traite des proprietes projectives I, p. XI ff ., 66 ff For the designation of the principle
.
of continuity as the "permanence of geometrical relations, "cf. Applic. p. 319; Traite II, 357; the same thought is expressed by Chasles in his "Principe des
relations
contingentes"
with another turn of expression.
(Apcrcu historique
p.204ff.,357ff.,368ff.)
86
distances a d to
c d.

This explanation obviously presupposes the distances, and is thus in
measurement and comparison of certain
essence of purely metrical nature; that it is, nevertheless, made the basis of the geometry of position rests On the fact that it represents a
metrical relation, which remains unchanged in every projective transformation of a given figure. The concept of measure is here not excluded, but is taken up into the foundations as an underived
element.
unified exposition
Projective geometry only gains an independent and strictly when this last limit is also set aside, when the
determination, which is characterized metrically as the double pro The decisive method portion, is derived in a purely descriptive way. for this is given in the known quadrilateral construction of Staudt.
We
points a 6 c
determine the fourth harmonic point d to three given co-linear by constructing a quadrilateral of such a sort that two
opposite sides pass through a, the diagonal through 6, and the two other opposite sides through c; the point of intersection of the second diagonal of the quadrilateral with the straight line a b c is the desired
d, which is definitely determined by this method, since it can be proved that the construction indicated always gives the same result no matter what quadrilateral is taken as a basis, so long as it
point
satisfies
Thus without any application of metrical concepts, a fundamental relation of position is established by a proce dure which uses merely the drawing of straight lines. The logical
the conditions. 14
ideal of a purely projective construction of geometry is thus reduced to a simpler requirement; it would be fulfilled by showing the possibil ity of deducing all the points of space in determinate order as members
of a systematic totality,
by means merely
of this
fundamental
rela
tion
and
its
repeated application.
The demonstration of this Projective metric (Cayley and Klein). furnished in the formulation which projective geometry has gained through Cayley and Klein. We here gain a general procedure that
is
enables us to coordinate all the points of space that can be generated from a given starting-point by progressive harmonic constructions, with certain numerical values, and thus to give them a fixed position
within a general serial order. If we start with three points a, 6, c, in a straight line to which we coordinate the values 0, 1, then by
means
14
of the Staudtian quadrilateral construction,
we can
43
ff.
;
find their
C/. Staudt, Geometric der Lage, Niirnberg 1847, Geometric der Lage, 4th. ed., Leipzig 1899, I, p. 5.
8, p.
Reye, Die
87
fourth harmonic point to which we let the number 2 correspond, and we can further determine a new point, which forms with the points a harmonic fourth, and give this the value 3, until finally by 1, 2,
virtue of this method we gain an infinite manifold of simple deter minations of position, to each of which a whole number is coordinated. This manifold can be further completed by becoming a universally
group in which every element corresponds to a definite number. The transition to the pointcontinuum takes place on the basis of a further intellectual postulate, which is analogous to the postulate by which Dedekind in his theory
"dense"
rational positive or negative
introduces the irrational numbers as
"cuts"
(Schnitte).
We
thus
gain a complete scale on the basis of which a unified projective metric can be evolved in which the elementary operations, such as addition and subtraction, multiplication and division of distances, are defined purely geometrically. Also the advance to structures of
higher dimensions offers no difficulty in principle; it results when we take into consideration, instead of the points of one straight line,
those of two or more straight

this
lines. 15
The concept of space and the concept of order. The working out of thought is chiefly of technical mathematical interest; but a general philosophical result, anticipated from the beginning of modern
geometry,
also
is
evident
here.
The
inclusion
of
the
spatial
concepts in the schema of the pure serial concepts is here finally accomplished. The designation of the individual points of space by corresponding numbers might indeed at first cause the illusion that concepts of magnitude, of length and distance, are applied in this deduction. In truth, however, number is only used here in its most
general logical meaning: not as an expression of the measurement and comparison of magnitudes but as the expression of an ordered
sequence. We are not concerned with the addition or division of distances and angles, but only with the differentiation and gradation
of the
members
of a certain series, the elements of
which are defined
as pure determinations of position. Here we find verification of the fact that, in our general logical deductions, number was evolved as
15 For all particulars regarding this method, of which only the principle can be suggested here, cf. F. Klein, Vorlesungen uber Nicht-Euklidische Geometric, 2nd. impression, Gottingen 1893, p. 315 ff ., 338 ff. as well as Math. Annalen, IV, On projective metric, see also Weber-Wellstein, Encyklopadie der p. 573 ff. Elementar -Mathematik, Vol. II, 18.
;
pure ordinal number and kept free from

able magnitudes.
all
connection with measur
The demand, which Descartes made, is thus The order of points of space is conceived in satisfied in a new way. the same manner as that of numbers. True, the two fields remain
strictly separated in essence; the
"essence"
immediately reduced to that of number.
of the figure cannot be But precisely in this
relative independence of the elements, as in the independence of their fundamental relation, is manifest the connection in general deduc
tive
method.
As
in the case of
number we
start
from an original
unit (ursprungliche Einheitssetzung) from which, by a certain generat ing relation, the totality of the members is evolved in fixed order, so
here
we first postulate a plurality of points and a certain relation of position between them, and in this beginning a principle is dis covered from the various applications of which issue the totality of
"a
In this connection, projective geome possible spatial constructions. priori" science try has with justice been said to be the universal of space, which is to be placed beside arithmetic in deductive rigor and
purity.
16
Space
is
here deduced merely in

coexistence"
its
most general form as
in general, while no decision is made concerning its special axiomatic structure, in particular concerning the validity of the axiom of parallels. Rather it can be shown that
the
"possibility
of
by the addition of special completing conditions, the general projec tive determination, that is here evolved, can be successively related
to the different theories of parallels and thus carried into the special 17 "parabolic," "elliptical" or "hyperbolic" determinations.
Geometry and
the
group theory.
Thus
in contrast to the multiplicity of the geometrical concept

Its logical character persists
of geometrical methods, the single
form
stands out with increasing clarity.
through all change in particular application. This character can be brought to mind by considering the most general interpretation which the modern concept of geometry has attained. The addition of geometry to the theory of groups forms the final and decisive step for the whole interpretation. The very definition of "group" contains a new and important logical aspect, in so far as through it
brought to intellectual unity, not so much a whole of individual elements or structures, as a system of operations. A totality of operations forms a group, when witp any two operations their
is
,
16
Cf. Russell,
17
Cf. F. Klein,
The Foundations of Geometry, ambridge Mathem. Annalen IV, 575 ff.
1897, p. 118.

combination
89
is also found in the group, so that the successive appli cation of different transformations belonging to the totality leads
only to the operations originally contained in it. In this sense, a group is formed by all the geometrical transformations which result
when we permit the elements any movements whatever
in ordinary three-dimensional space; for the result of two successive movements can here always be represented by a single movement. 18 In this
concept of the group, a general principle of classification is gained by which the different possible kinds of geometry can be unified under a
single point of
raise the question as to
view and their systematic connection surveyed. If we what in general is to be understood as a
"geometrical" property, we find that we consider as geometrical only such properties as remain unaffected by certain spatial trans formations. The propositions, which geometry evolves about .a,
certain structure, persist unchanged when we vary the absolute position of this structure in space, when we increase or decrease the
absolute magnitude of its parts proportionately, or when finally we reverse the arrangement of the individual parts, as when we sub/stitute for the original figure another, which is related to it as its / image in a mirror. The thought of independence from all these trans formations must be added to the intuition of the individual form, that
serves as a starting-point, to give this form true universality and therewith true geometrical character. "Geometry is distinguished
from topography by the fact that only such properties of space are geometrical as remain unchanged in a certain group of opera If we adhere to this explanation, we gain a view of verytions." \called
diverse possibilities for the construction of geometrical systems, all equally justified logically. For as we are not bound in the choice of the group of transformations, which we take as the basis of our
investigation, but can rather
broaden this group by the addition of
conditions, a way is opened by which we can go from one form of geometry to another by changing the fundamental system to which all assertions are related. For instance, if we take the ordinary
new
metrical geometry as characterized by the appropriate group of spatial transformations (i.e., by the specific operations of movement,
of similarity-transformation can broaden it to projective
and
of
mirroring
(Spiegelung))
we
geometry by adding the system of all projective transformations to this group, and considering the con18
Cf. F. Klein, Einleitung in die hohere Geometric II, p.
1 ff.
90
stant properties in this broadened sphere of transformations. As F. Klein has shown in detail, the most diverse kinds of geometry can
be methodically grounded and similarly deduced, by proceeding from a given group to a more inclusive system by means of a definite rule. In general, in each of these geometries where a manifold con
taining a transformation-group is given, the problem is to develop the invariant-theory applicable to the group. 19 The concepts of constancy and change in geometry. This universal
procedure throws a bright light on the essential relation of the con cepts of constancy and change in the foundations of geometry. saw how from the beginnings of Greek mathematics, the philosophical
question constantly reverts to this relation.
defined, in Platonic language, as
if it
We
If
geometry were
"eternal
being,"
what possesses
were true that exact proof were only possible of that which always maintains itself in the same form, then change could be tolerated as an auxiliary concept, but could not be used as an independent logical The field of becoming marked out a region within which principle. pure mathematical thought possessed no force and which thus seemed given over to the indeterminateness of sensuous perception. This emphasis on permanence, which was intended to exclude all sensuous elements from the foundations of pure mathematical knowledge,
finally
proved to work within geometry in the opposite direction.
requisite rigid constancy of the intuitive spatial form narrowed the freedom of geometrical deduction; thought remained entangled in the particular figure instead of directing itself to the ultimate
The
grounds of the connection of figures according to law. A new development begins only after the concept of change has been This development critically tested and confirmed by analysis.
systematic conclusion in the theory of groups; for here change recognized as a fundamental concept, while, on the other fixed logical limits are given it. The Platonic explanation is Lhand,
its
is
reaches
jlnow
.
confirmed
in
new
sense.
Geometry, as the theory of
invariants, treats of certain unchangeable relations; but this unchangeableness cannot be defined unless we understand, as its ideal
background, certain fundamental changes in opposition to which it gains its validity. The unchanging geometrical properties are not
19
For
gramm"
all particulars, we must refer again to F. Klein s Erlanger Proof 1872, Vergleichende Betrachtungem iiber neuere geometrische For"
schungen (reprinted Math. Annalen
43, 1893, p. 63 ff.)
91
such in and for themselves, but only in relation to a system of possible transformations that we implicitly assume._jCojistancy and change
thus appear as thoroughly correlative moments, definable only through each other. The geometrical "concept" gains its identical and determinate meaning only by indicating the definite group of \changes with reference to which it is conceived. The permanence in question denotes no absolute property of given objects, but is jhere /valid only relative to a certain intellectual operation, chosen as a Here already a change appears in the meaning / system of reference.
F
/
of the general category of substantiality, that must constantly grow clearer in the course of the enquiry. Permanence is not related to
/ the duration of things and their properties, but signifies the relative independence of certain members of a functional connection, which
\
prove in comparison with others to be independent moments.

in
The evolution of modern mathematics has approached the ideal, which Leibniz established for it, with growing consciousness and success. Within pure geometry, this is shown most clearly in the
development of the general concept of space. The reduction of metrical relations to projective realizes the thought of Leibniz that, before space is defined as a quantum, it must be grasped in its original
qualitative peculiarity as an "order of coexistence" (ordre des coexist ences possibles ). The chain of harmonic constructions, by which
the points of projective space are generated, provides the structure of this order, which owes its value and intelligibility to the fact that
it is
a^
not sensuously presented but is constructed by thought through succession of relational structures. 20 We can still take the elemen20
It is of historical interest that the logical problem of a metrical geometry based on pure projective relations was, as a matter of fact, grasped by Leibniz. Against Leibniz s definitions of space as an order of coexistence and time as an order of succession, Clarke, who advocated Newton s theory of absolute space and absolute time, raised the objection that they did not touch the essential import of the two concepts. Space and time are first of all quantities, which position and order are not. Leibniz replied that also within pure deter minations of order, determinations of magnitude are possible, in so far as a preceding member is distinguished from a succeeding member and the "dis tance" between them can be conceptually defined. Relative things have their
magnitudes just as well as absolute things; thus, e.g., in mathematics, relations or proportions have magnitudes, which are measured by their logarithms;
92
tary contents of geometry: the point, the straight line and the plane, from intuition but all that refers to the connection of these contents must be deduced and understood conceptually. In this sense,
;
to free a relation, such as the general rela which at first_seems_to possess an irreducible ^sensuous existence, from this restriction and to raise it to free logical The meaning of this relation must be determined by application. definite axioms of connection in abstraction from the changing sensuous material of its presentation; for from these axioms alone is gained the meaning in which it enters into mathematical deduction. By this extension, we can make the concept of bet ween" independ ent of its original perceptual content and apply it to series in which the relation of "between" possesses no immediate intuitive correlate. 21
modern geometry seeks

tion of
"between/^
"
Characteristic (Kombinatorik) as pure
"doctrine
of forms
(Leibniz).
This interpretation, however, advances still further when it attempts to subsume the specific order of spatial externality under a universal system of possible orders in general. Again we are led to the Leibnizian
mathematics
conception of mathematics. According to this conception, is not the general science of magnitude but of form, not
the science of quantity but of quality. Characteristic (Kombina torik) thereby becomes the fundamental science; we do not compre hend under it the doctrine of the number of combinations of given
elements, but the universal exposition of possible forms of connection
22 Wherever a definite general and their mutual dependency. form of connection is given, which we can express in certain rules and is defined in the mathematical "axioms, there an identical
in
"object"
nevertheless they are and remain relations." Leibniz, Hauptschriften zur Grundlegung der Philosophic, I, Philos. Bibl. 107, p. 189 f. We see here a reference to a question, which has been repeated in the modern grounding of
projective metrical geometry: for in the latter, in fact, the "distance" between two points is defined and measured by the logarithm of a certain doublerelation. Cf. Klein, Vorlesungen iiber Nicht-Euklidische Geometric, p. 65 ff.
21
More
"Hinc
especially see Pasch, Vorles. uber neuere Geometric, 1 and 9. etiam prodit ignorata hactenus vel neglecta subordinatio Algebrae
ad artem Combinatoriam, seu Algebrae Speciosae ad Speciosam generalem, seu scientiae de formulis quantitatem significantibus ad doctrinam de formulis, seu ordinis similitudinis relationis, etc., expressionibus in universum, vel scientiae generalis de quantitate ad scientiam generalem de qualitate, ut adeo speciosa nostra Mathematica nihil aliud sit quam specimen illustre Artis Combinatoriae seu speciosae generalis. Leibniz, Math. Schriften, Gerhardt,
VII, 61.
93
The relational structure as such, not the absolute property sense. of the elements, constitutes the real object of mathematical investiga Two complexes of judgments, of which the one deals with tion.
straight lines and planes, the other with the circles and spheres of a certain group of spheres, are regarded as equivalent to each other on this view, in so far as they include in themselves the same content of
conceptual dependencies along with a mere change of the intuitive of which the dependencies are predicated. In this "subjects,"
with which ordinary Euclidian geometry deals "points" can be changed into spheres and circles, into inverse point-pairs of a hyperbolic or elliptical group of spheres, or into mere number-trios without specific geometrical meaning, without any change being produced in the deductive connection of the individual propositions, which we have evolved for these points. 23 This deductive connection constitutes a distinct formal determination, which can be separated from its material foundation and established for itself in its systematic
sense, the
The particular elements in this mathematical construc character. tion are not viewed according to what they are in and for themselves, but simply as examples of a certain universal form of order and
connection; mathematics at least recognizes in them no other than that belonging to them by participation in this form. For
"being"
it is
only this being that enters into proof, into the process of inference, and is thus accessible to the full certainty, that mathematics
gives
its
objects.
Geometry as pure "doctrine of relations" (Hilbert). This interpreta tion of the methods of pure mathematics receives its clearest expres
sion in the procedure which Hilbert has applied in the exposition and deduction of the geometrical axioms. In contrast to the Euclidian
which take the concepts of the point or the straight line as immediate data of intuition, from which fixed content they proceed, the nature of the original geometrical objects is here exclusively
definitions,
by the conditions to which they are subordinated. The begin ning consists of a certain group of axioms, which we assume, and their compatibility has to be proved. From these rules of connection, that we have taken as a basis, follow all the properties of the elements.
defined
The point and the straight line signify nothing but structures which stand in certain relations with others of their kind, as these relations
23 Cf. the very instructive examples and explanation given by Wellstein, Encyklopadie derElem. Mathematik., Vol. II, Bk. I, 2nd. sect.
94
are defined by certain groups of axioms. Only this systematic "complexion" of the elements, and not their particular characters, is taken here as the expression of their essence. In this sense,
tions. 24
Hilbert s geometry has been correctly called a pure theory of rela In this, however, it forms the conclusion to a tendency of
we can trace in its purely logical aspects from the beginnings of mathematics. At first, it might seem a circle to define the content of the geometrical concepts exclusively by their axioms: for do not the axioms themselves presuppose certain concepts
thought, which
first
in their formulation?
This difficulty
is
disposed of
when we
clearly
It distinguish the psychological beginning from the logical ground. is true that, in the psychological sense, we can only present the
meaning
of a certain relation to ourselves in connection with
some
given terms, that serve as its "foundations." But these terms, which we owe to sensuous intuition, have no absolute, but rather a change
able existence.
We
all
take them only as hypothetical starting-points;
but we look for
closer determination
from their successive inser
tion into various relational complexes. It is by this intellectual proc ess that the provisional content first becomes a fixed logical object.
The law
<pv<T6L,
of connection, therefore, signifies the real irpbrpov TJJ while the elements in their apparent absoluteness signify
-n-pos fnj.as.
only a Trporepov
|
1
"
Intuition seems to grasp the content as
isolated self-contained existence; but as soon as we go on to characterize this existence in judgment, it resolves into a web of related
an
structures which reciprocally support each other. Concept and know the individual only as a member, as a point in a judgment
systematic manifold; here as in arithmetic, the manifold, as opposed to all particular structures, appears as the real logical prius. (Cf. above p. 68). The determination of the individuality of the
elements
is
ment;
it is
not the beginning but the end of the conceptual develop the logical goal, which we approach by the progressive
connection of universal relations. The procedure of mathematics here points to the analogous procedure of theoretical natural science,
which it contains the key and the justification. (Cf. Ch. V.) The syntheses of generating relations. From this point of view, we can understand how the center of gravity of the mathematical system
for
has
moved
in
a definite direction in
its
historical
is
The
24
circle of objects to
which mathematics
applicable
development. is extended;
Wellstein, op.
cit.,
p. 116.
95
becomes clear that the peculiarity of the method is bound and limited by no particular class of objects. The "mathesis universalis" in the philosophical sense, which it had for Descartes, was to form the fundamental instrument for all problems relating to order and measure. Leibniz replaced this mere conjunction of two aspects
finally it
by a
relation of logical subordination; the doctrine of the different possible types of connection and arrangement was made the pre
25 supposition of the science of measurable and divisible magnitudes.
The development of mathematical exposition projective geometry independently of the instrumentalities of measure and the compari son of magnitudes. Metrical geometry itself is deduced from purely qualitative relations, that merely concern the relative position of the points of space. The extension of mathematics beyond traditional bounds is still more striking in the case of the theory of groups; there the immediate object is not determinations of magnitude or position, but a system of operations, which are investigated in their mutual dependency. In the theory of groups, for the first time the supreme and universal principle is reached, from which the total field of mathematics can be surveyed as a unity. In its general meaning, the task of mathematics does not consist in comparing, dividing or compounding given magnitudes, but rather in isolating the generating relations themselves, upon which all possible deter mination of magnitude rests, and in determining the mutual connec
clarifies this
Modern mathematics
conception.
realizes the ideal of
tion of these relations.
The elements and
all their
derivatives appear
as the result of certain original rules of connection, which are to be examined in their specific structure as well as in the character that
The various results from their composition and interpenetration. forms of calculus of modern mathematics, Grassmann s Ausdehnungslehre, Hamilton s theory of quaternions, the projective calculus of distances, are only different examples of this logically universal
25
Cf.
above
p. 121,
Leipzig, 1904, p.
5, p. 50, p. 62.
note 2; also Leibniz Hauptschriften (Phil. Bibl. Vol. 107), For the modern interpretation, see Russell,
Principles of Mathematics, p. 158 and 419: "Quantity, in fact, though philoso phers appear still to regard it as very essential to Mathematics, does not occur in pure Mathematics, and does occur in many cases not at present amenable to mathematical treatment. The notion which does occupy the place tradi
tionally assigned to quantity is order." Cf. Gregor Itelson s definition of mathematics as the science of ordered objects. (Revue de Metaphysique, XII, 1904.)
96
procedure.
of all these procedures con here achieves completely free and independent activity (Betdtigung) in that it no longer remains limited to the compounding of quantities, but is directly applied to the
sists precisely in
The methodological merit

that the
"Calculus"
,
synthesis of relations. We were able to trace this synthesis as the real goal of mathematical operations in the field of magnitude itself in the development of the
analysis of the infinite. sphere of consideration
(Cf.
is
above
p.
73
f.)
Now, however, the
ment
widened; for any arbitrarily chosen ele can serve as a foundation in so far as a new structure can be
to issue
It is
made
from
it
by repeated application
its
of a certain defined
relation.
merely
this possibility of determination,
retained in the calculus and constitutes

condition.
The certainty of element. We can deal with products of points or of particular vectors, as in Grassmann s geometrical characteristic and in the
theory of quaternions; we can have points characterized not only by their different positions in space, but by different mass-values, as in
which is and sufficient necessary the deductive structure is bound to no
Mobius barycentric
establish the result
lar surfaces, forces or pairs of forces in
we can compound distances or triangu any way with each other and In all these cases, we are not calculation. 26 by
calculus;
concerned in analyzing a given "whole" into parts similar to it, or in compounding it again out of these, but the general problem is to combine any conditions of progression in a series in general into a
If an initial element is defined and a principle given which we can reach a manifold of other elements by a regular progression, then the combination of several such principles will be an operation, which can be reduced to fixed systematic rules. Wherever such a transition from simple to complex series is possible, there a new field for deductive mathematical treatment is defined. Grassmann s Ausdehnungslehre and its logical principles. It seems to have been this general thought, as it evolved in strict sequence from the philosophical ideal of Descartes and Leibniz of "mathesis universalis," which led also to one of the most weighty and fruitful con
unified result.
by means
of
ceptions of
Ausdehnungslehre.
modern mathematics, viz., to Hermann Grassmann s The general considerations, which Grassmann
26 Cf. more particularly regarding these methods of calculation, Whitehead, Universal Algebra I, Cambridge, 1898, as well as H. Hankel, Theorie der komplexen Zahlensysteme, Leipzig, 1867.
97
prefaced to his work, if regarded as mathematical definitions, might occasionally seem unsatisfactory and obscure; nevertheless, they
signify a clear methodological project, whose significance is explained and confirmed by the further development of the problems. 27 The goal, which Grassmann set himself, was to raise the science of space
to the rank of a universal science of form.

science of form
is
-The character of a pure
beyond thought
it proof does not go another sphere, -but remains entirely in the combination of different acts of thought. This postulate is fulfilled
defined
by the
fact that in
itself into
numbers for all details in the field of number can be from the system of ordered postulations, to which the number series itself owes its being. But as "immediate" a "beginning" must now be gained for geometry as is already given and assured within arithmetic. 28 For this purpose, we must here also go back from the given extensive manifold to its simple "manner of generation," by virtue of which the manifold is first surveyed and
in the science of
;
entirely deduced
grasped. In the ordinary account of the geometrical elements, we are accustomed to speak of the generation of the line out of the point, of the surface out of the line but what is here meant as a mere picture,
;
must receive a
strictly conceptual interpretation in order to serve
as the starting-point of the new science. The intuitive spatial relations may offer the first occasion for grasping pure conceptual
relations; but they do not exhaust the real content of the latter. Instead of the point (i.e., the particular place), we now assume the element, by which is meant only a pure particular grasped as different
A specific content is thus not yet assumed: particulars. can be no thought here as to what sort of particular this really for it is the particular absolutely, without any real content; nor is can there be question as to in what relation one particular differs
from other
"there
from another, for the particular is defined as different absolutely, without the assumption of any real content with reference to which it 29 is different." In the same way, we expressly abstract from all characters of the changes, which we think of the fundamental special element as undergoing, and merely retain the abstract thought of an
27
j.
Mathem.
28
Cf. especially V. Schlegel, Die u. Physik, Vol. 41, 1896.
Grassmannsche Ausdehnungslehre.
Ztschr.
Grassmann, Die
lineale Ausdehnungslehre: ein neuer

p. 47.
Zweig der Mathematik

I,
(1844).
29
Ges. mathemat. u. physikal. Werke, Leipzig 1894,

cit.,
p. 10, p. 22.
Ausdehnungslehre, op.
98
original beginning,

from which there
issues
is primarily concerned with definite kinds of transformation, nevertheless, the total scheme from the beginning reaches further. Here we are occupied only with
of one operation a multiplicity of members. working out of Grassmann s Ausdehnungslehre
by continuous repetition Thus if the concrete
that aspect, which stands out as the most general function of the mathematical concept: with giving some qualitatively definite and
unitary rule that determines the form of the transition between the members of a series. "The Different must evolve according to a The simple form of extension is law, if the result is to be definite.
thus that which arises by a transformation of the generating element according to one law; the totality of the elements, which can be gener ated according to the same law, we call a system or a field." 30 Simi
larly, there arise systems on higher planes when we combine different transformations in such a way, that first a manifold evolves by a certain transformation from the initial element, and then the totality of its members is subjected to a new transformation. In as much as
fields, which we consider, are not given us from elsewhere, but are merely known and defined by the rule of their construction, it is clear that this rule must suffice to represent exhaustively all their
the
properties.
These general considerations gain a more precise mathematical meaning, when Grassmann goes on to develop the various possible forms of connection in detail and to limit them from each other by the formal conditions to which they are subjected. There results a
developed doctrine of the
"addition"
and
"subtraction"
of similar
or dissimilar transformations, a theory of external and internal All these operations multiplication of distances and points, etc.
agree with the algebraic procedures of the
same names merely
in
certain formal peculiarities, such as subjection to the associative or the distributive law; but in and for themselves they represent entirely
independent processes, by which a new structure can be definitely determined from any given elements. We advance from the rela
tively simple forms of
"generation,"
which we have established by
definition, to ever more complex ways of constructing a manifold out of certain fundamental relations. If an initial member a is assumed
and a plurality of operations Hi R2 Rs same time, which successively transform

30
indicated at the
it
into the different values
Ausdehnungslehre, p. 28.

a\
99
2 az as etc., then the result of the compounding of these 3, oil operations and the various possible types of this compounding is deductively determined. The considerations by which Grassmann
introduces his work thus create a general logical schema under which the various forms of calculus, which have evolved independ ently of the Ausdehnungskhre, can also be subsumed; for they only show from a new angle that the real elements of mathematical calculus are not magnitudes but relations. The forms of calculus, and the concept of the Source. If we survey the whole of these developments, we recognize at once how the fundamental thought of logical idealism has been progressively confirmed and deepened in them. More and more the tendency of modern mathematics is to subordinate the elements as such and to allow them no influence on the general form of proof. Every concept and every proposition, which is used in a real proof and is not merely related to pictorial representation, must be fully grounded and understood in the laws of constructive connection. The logic of mathematics, as Grassmann understands it, is, in fact, in a strict
"given"
der reinen Erkenntnis
(Logik des Ursprungs). Cohen s Logik developed its fundamental thought of the Source in connection with the principles of the infinitesimal calculus. 31
sense
"logic
of the
source"
is the first and most striking example of the general of view, which leads from the concept of magnitude to the point concept of function, from "(quantity" to "quality" as the real founda
Here, in fact,
In advancing to the other fields of modern mathematics, the However logical principle here established gains new confirmation. different these fields may be in content, in structure they all point
tion.
back to the fundamental concept of the Source. The postulate of this concept is fulfilled wherever the members of a manifold are deduced from a definite serial principle and exhaustively represented by it. The most diverse forms of "calculus," in so far as they satisfy this condition, belong to one logical type, as also they agree in their Thus fruitfulness for the problems of mathematical natural science. Mobius applied his universal calculus to a strictly rational construc tion of statics, while Maxwell evolved the elements of mechanics from the fundamental concepts of vector analysis. 32 The systematic
Cohen, Logik der reinen Erkenntnis, especially p. 102 ff. Mobius, Lehrbuch der Statik (T.I, 1837); cf. especially Hankel, Theorie der komplexen Zahlensysteme, VII; Maxwell, Matter and Motion.
32
31
100
connection of operations, once deduced, remains unchanged when we substitute forces for straight lines, pairs of forces for certain distanceproducts, and thus relate every geometrical proposition directly to a
mechanical proposition.
The subordination
of
the
infinitesimal
analysis to the more inclusive system of "analysis of relations" as such serves also to fix and limit its own problem. In spite of the protests of idealistic logic, the concept of the "infinitely small" has
continually led to the misunderstanding that here magnitudes are not understood from their conceptual principle, but rather compounded from their disappearing parts. Thereby, however, the real question
is
mistaken and displaced; for we are not concerned with pointing
out the ultimate substantial constitution of magnitudes, but merely with finding a new logical point of view for their determination. This point of view comes out sharply, however, when we place the other
possible forms of mathematical "determination" by the side of the procedure of the infinitesimal calculus. For example, it would be
nonsense to ascribe the ordinary
"arithmetical"
meaning to opera
tions in the barycentric calculus, such as when simple points are added or the sum of two distances with direction is represented by
the diagonal of the parallelogram constructed out of them, or when we speak of the product of two or three points or of the product
of a point
and a
is
distance.
The
relation of the
"whole"
to
its
com
here excluded and replaced by the general relation ponent "parts" the conditioned to the individual moments, which conceptually of
constitute
it.
The
distinction,
clearly
emphasized by Leibniz,
is
unavoidable: in contrast to the "analysis into parts," there appears everywhere the "resolution into concepts," which as the univer
sal
instrument guarantees the certainty and the progress of pure

f
IV
deduction.
The problem of metageometry. The extension, which the system of Euclidean geometry has undergone through metageometrical investi
gations and speculations, falls in point of content outside of the sphere of our enquiry. For we are not concerned with presenting the results
of mathematics, significant and fruitful as they may be from the standpoint of the critique of cognition, but merely with determining the principle of the mathematical construction of concepts. But
even from this limited point of view, we cannot avoid taking up the
101
problem of metageometry; for it is the special distinction of this problem that it has not merely transformed the content of mathemati cal knowledge, but also the interpretation of its basis and source. The question necessarily arises as to whether the view, which has previously been gained of the mathematical concept, can be main
tained in the face of the
new problems
arising in this connection.
That there is here a

is
justified extension of the original field of
geometry
now beyond question with philosophers as with mathematicians; so much the more necessary is it to discover whether the new con
tent breaks through the logical form of geometry or confirms
it.
The attempt at an empirical grounding of geometry (Pasch). The answer of mathematics itself seemed for a time definitive; in general it was the empirical character of the geometrical concepts, that was deduced from the metageometrical researches. Veronese s Fondamenti di
critical
geometria, the first complete historical survey of all attempts to reform the theory of the principles of geometry, affirms as a common conviction of scientific investi gators, that at least the ordinary geometry of tridimensional 33 If we examine the motives space is founded merely on experience.
which have led individual investigators to this decision, we soon recognize that the agreement of interpretation is only apparent. It is as if geometry, on entering the field of philo sophical speculation, had lost its characteristic privilege of applying its concepts in a strictly unambiguous sense. The whole indefiniteness
closely,
and reasons more
belonging to the concept of experience in popular usage al once comes to light. An empirical grounding of the mathematical con
cepts would only be given, in the strict sense, where proof was adduced that their entire content was rooted in concrete perceptions, and deducible from them. Thus the one consistent empiristic system of mathematics has been constructed by Pasch, in so far as he
attempts to introduce the elementary structures, such as the point and the straight line, not in exact conceptual form but merely in the meaning which they can possess for sensation. The fruitful applica
tion,
which geometry continually receives in natural science and in practical life, Pasch explains, can only rest on the fact that its con
cepts originally correspond exactly to the actual objects of observa Only secondarily is this original content overlaid with a net33 Veronese. Grundzuge der Geometric von mehreren Dimensionen und mehreren Arten geradliniger Einheiten, German ed., Leipzig, 1894, p. VIII, Note 1.
tion.
102

of technical abstractions,
is
work
by which indeed
is
its
theoretical con
struction
furthered, but nothing

propositions.
If
truth of
its
added to the fundamental we abandon these abstractions, if we
turn resolutely back to the real psychological beginnings, geometry retains the character of a natural science and is only distinguished from the other natural sciences by the fact that it only needs to take a very limited number of concepts and laws directly from experi ence, and gains all the rest by the development of this once assumed material. The "point," according to this conception, is nothing but a
material body, which proves to be not further divisible within the given limits of observation, while distance is compounded out of a
The validity of the geometrical princi accordingly subject to certain limitations, which are demanded by the nature of the geometrical objects as mere objects of percep tion. Thus to the proposition, that between two points we can
finite
number
of such points.
ples
is
draw one and only one straight line, the reservation is to be added that the points considered must not be too close to each other. The theorem, that between two given points a third can always be
inserted,
when we go beyond
34
remains in force only for these cases, while it loses its validity certain limits, which cannot indeed be clearly
assigned. Ideal objects in empirical geometry. All these developments are consistent with the chosen starting-point; but it soon appears
impossible to reach the structure of the total historical system of geometry from it. In order to give the proofs true rigor and universality, we are forced from the assumption of points,
scientific
"real"
that represent actual objects of observation, to the assumption of "unreal" structures, which are ultimately nothing but a result of
(those ideal constructions, that
we
originally sought to exclude.
The
concepts of perfectly determinate points, straight lines and planes are used also, and serve as a basis for the definitions of those elements
which the geometrical idea is only imperfectly and approximately Every geometry of approximation is obliged to operate with presuppositions taken from geometry; it cannot serve for the deduction of methods, of which it is rather only a special
in
realized.
"pure"
application.
34
35
35
Pasch, Vorlesungen iiber neuere Geometrie, p. 17 f. Cf. the criticism of Pasch s system by Veronese, p. 655
cit.,
ff.,
and by Well-
stein, op.
p. 128
f.

Veronese s modification of empiricism.
103
The search
for
an empirical
foundation of geometry is thus led into a new path. Veronese, who at first approves the search, gives the thought a new turn, when he
urges that geometrical "possibility" is not to be based merely on direct external observation, but also on "mental facts." The geomet
axioms are not copies of the real relations of sense perception, but they are postulates by which we read exact assertions into inexact intuition. The raw material of sense impressions must be worked over by our mind before it can be useful as a starting-point
rical
mathematical considerations; and it is this "subjective" element which in pure mathematics, geometry and rational mechanics asserts its superiority over the "objective" element. Although geometry is defined here also as an exact experimental science, nevertheless, the logical role of experience has become entirely different. We start from "empirical considerations," from certain facts of sensuous intuition; but these facts serve, in Platonic language, only as the "spring-board" from which we ascend to the conception of universal systems of conditions with no sensuous correlate. The sensuous contents form indeed the first occasion, but express neither the limit nor the real meaning of the mathematical construction of concepts. They serve as the first incentive, but as such do not enter into the system of deductive proof, which is to be formed in strict independ ence. But in establishing this, the issue is already decided from the standpoint of the critique of knowledge; for such critique does not ask as to the origin of concepts, but only what they mean and are worth as elements of scientific proof. Rationalism and empiricism. Thus we are obliged ultimately to
for
appeal to a specific function of the intellect in the deduction of the geometries of more than three dimensions. In the system of Pasch, as Veronese remarks, multiple-dimensional geometry is not excluded
a posteriori but a priori, i.e., not factually but methodically. For the data of observation negate every attempt to enter a field, which lies beyond the possibilities of our spatial intuition. For this always
demands a pure act of construction, a possible "intellectual activity" in which we go beyond the given, and in which the generated element is determined from the beginning by the fact that we subject it to certain general laws of relation. As the axioms, propositions and
tion,
proofs of geometry cannot contain any undefined element of intui when we abandon intuition in general, there must at least remain
104

is
a purely hypothetical connection of abstract truths, which

sible
"If
acces
we are called," Veronese to intellectual investigation. adds, "Rationalists or Idealists because of these ideas, we accept the title in distinction from those who would unjustifiably deny the
and geometri and who would enquire whether each new hypothesis possesses a possible perceptual representation, e.g., in geometry a
greatest possible logical freedom to the mathematical
cal intellect,
purely external perceptual representation. We accept the title, however, only under the condition that no really philosophical mean The "really philosophical" meaning, that is ing be attached to here guarded against, is, as the reference to P. du Bois-Reymond
it."
only the hypostatization of mathematical ideals into a absolute existences; their purely intellectual value as 3 hypotheses is not being thereby affected.
shows,
sort
of
3 *5
Mathematical space and sensuous space.
The
logical
freedom here
sought for geometrical concepts cannot, however, merely relate to those that apply to more than three-dimensional spaces; in so far as a true unity of principles is sought, it must be recognized in the of this methods of ordinary Euclidean geometry. If the
"point"
geometry were only the image of an object existing outside of thought, "because there are outer objects which directly (!) present or arouse in us the perception of a point without which there are no real so-called 38 the continuity of the system of geometry would be broken; points," for what conceptual analogy and affinity subsists between elements, which are copies of presented things and elements that entirely result from "intellectual activities"? And conversely, if those intellectual procedures suffice to constitute the element of an n-dimensional manifold, what difficulty is there in gaining the element in the special
In fact, it is precisely when we compare case of three dimensions? Euclidean space with other possible "forms of space" that its peculiar conceptual character stands forth sharply. If from the standpoint of
metageometry, Euclidean geometry appears as a mere beginning, as given material for further developments, nevertheless, from the stand point of the critique of knowledge, it represents the end of a compli
cated series of intellectual operations. The psychological investiga tion of the origin of the idea of space (including those which were
36
Cf.
more particularly Ch. IV,

cit.,
p. 162
ff.
37
38
Veronese, op. Veronese, op.
p.
VIII
VII.
ff.,
XIII
ff., f.
p. 658, 687, etc.
cit.,
p.
cf.
p. 225
105
undertaken with a purely sensatiorialistic tendency) have indirectly confirmed and clarified this. They show unmistakably that the space of our sense perception is not identical with the space of our geometry, but is distinguished from it in exactly the decisive con stitutive properties. For sensuous apprehension, every difference of place is necessarily connected with an opposition in the content of
sensation.
"Above"
and
"below,"
"right"
and
"left"
are here not
equivalent directions, which can be exchanged with each other with out change, but they remain qualitatively distinct and irreducible
determinations, since totally different groups of organic sensations correspond to them. In geometrical space, on the contrary, all these oppositions are cancelled. For the element as such possesses
specific content, but all its meaning comes from the relative position it occupies in the total system. The principle of the absolute homogeneity of spatial points denies all differences, like the difference of above and below, which merely concern the relation of
no
object.
outer things to our bodies, and thus to a particular, empirically given 39 Points are what they are only as starting-points of possible
constructions, in which the postulate holds that the identity of these constructions can be recognized and retained through all diversity
of the initial elements.
The
further
such as
its
continuity and infinity, rest
they are in no
way given in spatial completions, which we assume in the continuity of space is a sensuously phenomenal property has been definitely set aside by the deeper mathematical analysis of the conil
moments of geometrical space, upon a similar foundation; sensations, but rest upon ideal them. The appearance that
manifold.
[I
||
tinuum, which has been carried out through the modern theory of the The concept of the continuum used by the mathematician in his deductions is in no way to be gained from the indefinite image This image can of space, that is offered us by sensuous intuition. never represent precisely the ultimate deciding difference by which
continuous manifolds are distinguished from other infinite totalities; no sensuous power of discrimination, however sharp, can discover any difference between a continuous and a discrete manifold in so far as
30 Concerning the differentiation of "homogeneous" geometrical space from inhomogeneous and "anisotropic" physiological space, cf. more particularly Mach, Erkenntnis u. Irrtum, Leipzig, 1905, p. 331 ff. Cf. especially the exposi tion of Stumpf, Zur Einteilung der Wissenschaften (Abhandl. der Berliner Akademie d. Wiss., 1906, p. 71 ff.).
106

"everywhere
dense," i.e.,
the elements of the latter are
where between
itself.
40
any two members, however
close
we may choose them, another

Just
member can be
discovered belonging to the assemblage
broadened by gradual steps of into the continuous totality of real numbers, so by a series of thought
as the field of rational
numbers
is
intellectual transformations, does the space of sense pass into the homogeneous and conceptual space of geometry.
infinite,
continuous,
to the
Kantian theory of geometry. It is thus a strange anomaly, when from the possibility of metageometry the empirical character of Euclidean space is inferred. Euclidean geometry does
Objections
not cease to be a purely rational system of conditions and conse quences, when it is shown that along with it other systems can be thought, which are capable of the same logical strictness of connec It is to be noted that two opposite objections, based upon the tion.
same premises which are taken from metageometry, have been expressed against the Kantian theory of geometry. On the one hand, the pure apriority of space is contested on the basis of these
own
and
premises, while on the other hand, it is objected that in Kant s exposition, the a priori freedom of the mathematical concept
all sensuous representation is not view that the axioms were satisfactorily expressed. in "pure intuition" can only be explained "by that residuum of sensualism which still attached to the Kantian idealism." 41 Of these two opposed objections, only the last possesses an entirely clear and consistent meaning. Not the empirical but the logical character of the fundamental concepts is confirmed and illumined in a new way
its
possible separation
from
s
Kant
"given"
by the modern extension

which we can
\
of the field of mathematics.
The
role,
ascribe to experience, does not lie in founding the particular systems, but in the selection that we have to make among
still
them.
reasoned that, as all the systems are equally valid in we need a principle that guides us in their applica tion. This principle can be sought only in reality, since we are not here concerned with mere possibilities, but with the concept and the problem of the real itself; it) short, it can be sought only in observa
It is
logical structure,
tion
40
and
scientific
experiment.
Experience thus never serves as a
uum
my
41
For explanation and examples, cf. especially Huntington, "The Contin as a type of order," Annals of Mathematics, 2ser., VI and VII; compare "Kant und die moderne Mathematik," Kant Studien, XII, 15 ff.
Wellstein, op.
cit..
p. 146.
107
proof or even as a support of the mathematical system of conditions, for such a system must rest purely in itself; but it points the way from the truth of concepts to their reality. Observation closes the gap
logical determination; it leads from the many geometrical space to the one space of the physical object.
left
by purely
forms of
Real space and experiment. This connection leads, however, beyond the bounds of pure mathematics and results in a problem, which can only be adequately solved by a critical analysis of the
procedure of physics.
The question
itself
of the
method and value
of the
If of central importance. physical experiment one looks to experiment for the confirmation or refutation of a certain system of mathematical hypotheses, experiment is essentially under
now becomes
stood in the Baconian sense of the "experimentum crucis" Experi ence and hypotheses belong accordingly to separate fields; each exists for itself and can function by itself. experience, which is conceived as separated from any conceptual presupposition, is
"Pure"
appealed to as a criterion of the value or lack of value of a certain theoretical assumption. The critical analysis of the concept of on the contrary, that the separation here assumed experience shows,
involves an inner contradiction.
Abstract theory never stands on
one
it
side, while on the other side stanjds the material of observation as
r"h
\
1
and without any conceptual interpretation. Rather we are to ascribe to it any definite character at all, must always bear the marks of some sort of conceptual shaping. We can never oppose to the concepts, which are to be tested, the ll empirical data as naked but ultimately it is always a certain of connection of the empirical, which is measured by a logical system similar system and thus judged. 42 But if the measuring experiment vjs always bound in this way to a system of presuppositions, which include both purely geometrical assumptions concerning space and
is
in itself
this material,
if
"facto";
then
concrete physical assumptions concerning the relations of bodies, it is clear that we can expect no clear decision from it with regard
to the conflict of geometrical systems. Wherever a value gained by experiment contradicts the value demanded by deductive theory, the alternative is left open to us whether we shall restore the agree
ment
and observation by changing the mathematical part And thought would undoubtedly avail itself of this latter procedure. The possible
of concept
or the physical part of our abstract hypothesis.
42
Cf. here the detailed
grounding in Ch. IV, especially Sect. IV.
108
SUBSTANCE AND FUNCTION and

is
variation of the conditions follows certain rules
bound
to a
Before we would proceed on the basis of the certain sequence. results of astronomical measurements, to change from the geometry
geometry of Lobatschefski, we would first have to as to whether we could take account of the new result investigate by an altered conception of the system of physical laws, for example, by revising the assumption of the strictly rectilinear propagation of
of Euclid to the
This state of affairs has been continually emphasized from light. the philosophical side in the controversies about the principles of geometry; but it seems that it was first through the expositions of
Poincare, which were decisive in this connection, that
clear
it
became
and gained wide recognition within mathematics.
As
Poincare"
justly emphasizes, all our experiences are related only to the relations of bodies to each other and their physical interactions, but never to the relation of bodies to pure geometrical space, or the parts of this
space to each other.

"essence"
It is thus vain to expect instructions
of space
from a procedure which, according to
about the its whole
tendency and disposition, is directed upon entirely different questions. Since the objects with which experience deals are of an entirely differ ent sort from the objects of which the assertions of geometry hold,
since the investigation of material things never directly touches the ideal circle or straight line, we never gain in this way a decision
the different systems of geometry. 43 The conceptual principles of pure space. Thus, if the choice between the various systems is not to be surrendered entirely to subjective caprice, we must face the problem of discovering a rational criterion of difference. Logical consistency, such as belongs to all these systems, is merely a negative condition, which they all share among
among
themselves. But within the group thus established the differences in fundamental structure and in relative simplicity of structure are not extinguished. While from the standpoint of the principles of identity and contradiction, the thought of the heterogeneity of -space may be
equivalent to that of homogeneity, nevertheless, there can be no doubt that within the rational system of knowledge the concept of
uniformity, in the most diverse fields, always precedes that of nonuniformity. The non-uniform is always gained from the uniform in the process of constructive synthesis by the addition of a new condition, and thus represents a more complex intellectual structure.
41
Cf. Poincarg,
La Science
et
VHypothese, Chs. 3-5.
THE CONCEPT OF SPACE AND GEOMETRY The form

form
of Euclidean space
109
of space, in the
is thus in fact "simpler" than any other same sense that within algebra a polynomial
of the first degree is simpler than a polynomial of the second degree. 44 In the order of knowledge, at least, there is here a necessary and
definite sequence;
but
it is
this order of
knowledge by which, in the

of objects.
critique of cognition,
we determine the order
The
differ
ences between Euclidean space and the space represented in the hypothesis of Lobatschefski or of Riemann first become manifest
when we compare
parts of these spaces, which transcend a certain
magnitude, with each other. If we limit ourselves, on the contrary, to the generating element of all these spaces, the difference disap The Euclidean standard holds without modification for pears. measurements in infinitesimals, which thus proves it in principle It represents the first and fundamental schema really fundamental. with which all other constructions are connected and from which they
are distinguished.
The uniformity
of Euclidean space
is
really only
an expression of the fact that it is conceived merely as a pure rela tional and constructive space and that all further determination of content, which might lead to a difference in absolute magnitude and in absolute direction, is eliminated from it. 45 In so far as absolute
determinations of magnitude as such are permissible in pure geometry, they always rest on a universal system of relations, which have been previously developed independently and which is only more closely determined in details by the addition of particular conditions. Euclidean space and the other forms of mathematical space. Thus Euclidean space remains, indeed, a conceptual hypothesis in a system
of possible hypotheses; but within this system, nevertheless, it From a possesses a peculiar advantage in value and significance. a manifold system of pure logico-mathematical forms, we select
that corresponds to certain rational postulates, and attempt with the help of this manifold to render the character of the real intelligible.
We
the
in
do not thereby exclude that, along with the fundamental system,
more complex systems
also possess a certain sphere of application

significance.
which they also gain concrete

44
45
In the
first
place,
Cf. Poincare, op.

Cf., e.g.,
cit.,
p. 61.
22: "The simplicity 1844, of space is expressed in the principle: space has the same properties in all places and in all directions, that is, in all places and in all directions, the
Grassmann, Ausdehnungslehre of
same constructions can be
produced."
HO
SUBSTANCE AXD FUNCTION*
the results of these systems are often themselves capable of an inter pretation and translation which brings them, at least indirectly, to As Beltrami has shown, the relations of intuitive representation.
Lobatschefskian geometry find their exact correlate and copy in the geometry of pseudo-spherical surfaces, itself a particular section of ordinary Euclidean geometry; while the "elliptical geometry" of planes, as developed by Riemann, corresponds to the geometry of
spherical surfaces within the Euclidean space of three dimensions. And also, when we go over to systems of higher dimensions, this
can again select possibility of referring back does not cease. within our intuitive space itself structures that are subject in all
their reciprocal relations to the abstract rules, which are deduced and proved for any manifold of more than three dimensions. Thus the
all spheres forms a linear manifold of four dimensions, the form of which can be investigated and established in universal 46 But even where we lack this reduction to known spa geometry.
We
manifold of
tial relations
and problems, the
possibility
is
not excluded of inter
preting the propositions of non-Euclidean geometry so that a definite concrete "meaning" corresponds to them. For all these propositions
only express a system of relations, while they make no final deter mination of the character of the individual members, which enter into
these relations.
The points, with which they are concerned, are not independent things, to which in and for themselves certain properties are ascribed, but they are merely the assumed termini of the relation
and gain through it all their character. (Cf. above p. 94 ff.) Hence, where any system is found in accordance with the rules of connection of any of these general theories of relation, a field of
itself
application for the abstract propositions is indicated and defined, no matter whether the qualitative character of the elements of the
system can be pointed out or whether

in space.
it can be intuitively represented In so far as physics offers us systems, which require a plurality of means of determination for their complete exposition, we can speak of a manifold of several "dimensions," to be judged and
treated according to the previously evolved deductive laws of these manifolds, regardless of whether or not these means of determina
tion permit a spatial interpretation.
Geometry and
rational
46
form
In any case, the result is that the purely reality. of the geometrical construction of concepts, as the
Wellstein, op.
cit.,
Cf.
more particularly
p. 102.

latter
111
has been gradually established, is not threatened by the metageometrical considerations but is rather confirmed. Even if
one heeds
all
the doubts that

still
may
be awakened through these
these doubts never concern the real ground of considerations, the concepts, but only the possibility of their empirical application. That experience in its present scientific form gives no occasion to go
beyond Euclidean space is expressly admitted even by the most radical empiristic critics. 47 From the standpoint of our present knowledge, they also conclude, we are justified in the judgment that
physical space
"is
we must not exclude the
to be regarded as positively Euclidean." Only possibility that in a distant future perhaps
changes will take place here also. If any firmly established observa tions appear, which disagree with our previous theoretical system of nature, and which cannot be brought into harmony with it even by
far-reaching changes in the physical foundations of the system, then, all conceptual changes within the narrower circle having been
tried in vain, the
query
may
arise
whether the
of
lost
unity
is
not to be
if
reestablished
by a change in the
"form
space"
itself.
But even
we take
into account such possibilities, the proposition would only be thereby strengthened that, as soon as we enter the field of the deter mination of reality, no assertion, however indubitable it may appear,
can lay claim to absolute certainty. It is only the pure system of con ditions, which mathematics erects, that is absolutely valid, while the
assertion, that there are existences corresponding to these conditions in all respects, possesses only relative and thus problematic meaning.
of universal geometry shows that this sphere of problems does not affect the logical character of mathematical knowledge as such. It shows that the pure concept on its side is prepared and fitted for all conceivable changes in the empirical character of percep
tions; the universal serial
The system
form
is
of the empirical
47
is
to be understood
the means by which every order and logically mastered.

ff.
of the
Enriques. Problemi della Scienza, Bologna, 1906, p. 293 Supplement on Einstein s theory of relativity. Tr.)
(See Ch.
VI
CHAPTER
IV
The constructive concepts and the concepts of nature. The logical nature of the pure functional concept finds its clearest expression and most perfect example in the system of mathematics. Here a field of free and universal activity is disclosed, in which thought transcends The objects, which we consider and into all limits of the "given." nature we seek to penetrate, have only an ideal whose objective being; all the properties, which we can predicate of them, flow exclu
from the law of their original construction. But precisely where the productivity of thought unfolds most purely, its here, The constructive con characteristic limit seems to come to light. of mathematics may be fruitful and cepts (Konstruktionsbegriffe) indispensable in their narrow field; but they seem to lack an essential element for serving as an example for the whole circle of logical problems, as typical of the properties of the concept in general. For however much logic limits itself to the "formal," its connection with the problems of being is never broken. It is the structure of being with which the concept and the logically valid judgment and inference are concerned. The Aristotelian conception and foundation of the syllogism assume this at all points ontology gives the basic plan for the construction of logic. (Cf. above p. 4 ff.) If this, however, is the case, mathematics can no longer serve as the type and model, for since it remains strictly within the field of its self-created structures, it has in principle no concern with being. The difference between the "generic concept" in the sense of traditional logic and the constructive mathematical concept may be freely granted; but one might be tempted to explain this difference by the fact that within mathematics the final and conclusive function of the concept is not sought, and is accordingly not found. The voluntary limitation, which we assume in it, is justified; but it would be a failure in method, if we were to attempt to solve all logical problems from the narrow standpoint that we have here defined for ourselves. The decision as to the direction of logic cannot be gained
sively
:
112

by a type
Rather
of consideration
113
which remains one-sidedly in the ideal. the genuine concepts of being, the assertions concerning things and their real properties, which must constitute the true standard. The question as to the meaning and function of the con
it is
cept gains nature.
its final
and
definitive formulation only in the concepts of
The concept of traditional logic and the scientific ideal of pure descrip If we proceed, however, from this conception of the problem, the solution seems to turn at once in favor of the traditional logical view. The concepts of nature know and can know no other task than to copy the given facts of perception, and to reproduce their content in abbreviated form. Here truth and certainty of judgment rest only on observation; there remains no creative freedom and arbitrariness of thought, but the character of the concept is from the
tion.
we
beginning prescribed by the character of the material. The more free ourselves from our own constructions, from the
"idols"
the image of outer reality presented to us. It is passive surrender to the object, which here seems to secure to the concept its force and effectiveness. We thus stand
of the mind, the
is
more purely
again wholly within the general view, that has found its logical expres The concept is only the copy of the sion in the theory of abstraction.
given; it only signifies certain features, which are present and can be The concep indicated in the perception as such. (Cf. above p. 5.)
meaning and task of natural science also corresponds completely to this view. The whole meaning and certainty of the concept as found in natural science depends accordingly on the condition, that it contain no element which does not possess its In order to represent ade precise correlate in the world of reality.
tion of the
quately a certain group of phenomena, theory may indeed assume and apply certain hypothetical elements; but in this case, also, the postu
late holds that these elements
must
at least be validated in a possible
perception.
it
hypothesis signifies only a gap in our knowledge; means the assumption of certain data of sensation, that have
An
hitherto been accessible to us in no direct experience, but which are nevertheless regarded as thoroughly homogeneous in their properties with the really perceived elements. Perfect knowledge could
abandon
asylum ignorantiae: for it, reality would be clearly and completely given as a whole in actual perception.
this
114

logical ideal of physics.
The whole modern philoso at first glance merely as the increasingly rigor phy of physics appears ous and consistent working-out of this view. In this view alone, does
The apparent
the possibility seem given of sharply separating experience and speculative philosophy of nature; and in it there seems to be indi
is first
cated a necessary condition, by which the scientific concept of physics In opposition to the metaphysical defined and completed.
ideal of the explanation of nature, there now appears the more modest task of describing the real completely and clearly. We no longer reach beyond the field of the sensible in order to discover the inex-
perienceable, absolute causes and forces, upon which rest the multi The content of plicity and change of our world of perception.
physics is rather constituted merely by the phenomena in the form in which they are immediately accessible to us. Colors and tones, smell and taste sensations, sensuous muscle-feelings and perceptions of pressure and contact are the only material out of which the world of the physicist is constructed. What this world seems to contain in addition, what is added in concepts, as atom or molecule, ether or energy, is in truth no fundamentally new element, but only a peculiar guise in which the data of sense appear. Complete logical
these concepts to their significance, when it them as symbols for certain impressions and complexes of The unity of the physical method seems thereby to impressions. be secured for the first time for now it is no longer compounded from
analysis
reduces
recognizes
heterogeneous elements, but in the general concept of sensation the
common denominator is
ity
fixed, to
which
all
must be ultimately
reducible.
Whatever
assertions concerning real resists this reduction
thereby shows itself to be a factor arbitrarily introduced, which must disappear in the final result. The goal of this philosophy of physics would be reached, if we resolved every concept, which enters into physical theory, into a sum of perceptions, and replaced it by this sum; that is, if we retraced the path from the intellectual abbre viation (which is what all concepts reveal themselves to be) to the concrete fullness of the empirical facts. The exclusion of all ele
ments, which possess no direct sensuous correlate in the world of perceptible things and processes, would be, accordingly, the true
logical ideal of physics.
Is this the true ideal of physics? ing the justification of this ideal,
However we may judge concern

its
very conception contains an

ambiguity, which must
first
115
be set aside. The description of the is confused with a general demand Which of the two elements is the Is it merely the actual procedure of original and determining one? science itself that is here brought to its simplest and shortest expres sion, or, on the contrary, is this procedure measured by a general theory of knowledge and of reality, which decides concerning its value? In the latter case, whatever the final result might be, the method of consideration would not be changed in principle. Again it would be a certain metaphysics of knowledge which sought to point out the way to physics. The answer to this question can only
actual status of physical theories that is made of these theories.
be won by following the course of physical investigation itself and considering the function of the concept that is involved directly in The same impartiality, that is demanded by the its procedure. positivistic critic with regard to the facts of sense perception, must also be demanded with respect to the more complex facts of knowl
edge.
scientific
also the first task is to grasp the actual side" of theory in its purity, before we decide as to the value or lack of value of the view of reality which it contains. Is this theory,
"f
Here
as it is historically presented, really only a collection of observations strung together as if on a thread, or does it contain elements, which belong to another logical type and therefore demand another
foundation?
ii
Numbering and measuring as presuppositions. The first and most striking characteristic which forces itself upon us with regard to any scientific theory, involves a peculiar difficulty when we consider it from the standpoint of the general logical demand for
The theories of physics gain their definiteness from the mathematical form in which they are expressed. The function of numbering and measuring is indispensable even in
description of the given.
order to produce the raw material of "facts," that are to be repro duced and unified in theory. To abstract from this function means
to destroy the certainty
and
clarity of the facts themselves. this connection
How
ever self-evident, indeed
may seem, it is highly paradoxical in principle when we look back over our general estimate of the principle of mathematical conceptual construction. It has become increasingly clear that all content belonging to the
trivial,
116
mathematical concept rests on a pure construction. The given of intuition forms merely the psychological starting-point; it is first
known mathematically when it is subjected to a transformation, by which it is changed into another type of manifold, which we can pro duce and master according to rational laws. Every such transforma tion, however, must obviously be abandoned where we are merely
concerned with grasping the given as given in its specific individual structure and properties. For the purposes of knowledge of nature, in the positivistic sense of the word, the mathematical concept is not
so
much a justified and necessary instrument to be applied along with experiment and observation as a constant danger. Does it not falsify the immediate existence, revealed to us in sensation, to subject this existence to the schema of our mathematical concepts,
and thus to let the empirical determinateness of being disappear into the freedom and caprice of thought? And yet this danger, however clearly it may be envisaged, is never to be avoided or set aside. No matter how penetratingly the physicist
portray it as empirical philosopher, he directly falls into it again as soon as he sets to work as scientific investigator. There is no exact establishment of a time-space fact, which does not involve
may
the application of certain numbers and measures. One might over look the difficulty in this, if it were merely a matter of the elementary
concepts and structures of mathematics. Although the first of Kepler s laws of planetary motion makes use of the purely geometrical
definition of the ellipse as a conic section,
and the third of the
arith
metical concepts of the square and the cube, at first no epistemological problem might be seen in this; to the naive comprehension, number
and form themselves appear as a sort of physical property, inhering in things precisely as do their color or their lustre and hardness. The more this appearance is destroyed in the (Cf. above p. 28.) advance of mathematical conceptual construction, the more strikingly the general question is thrown into relief. For it is precisely the complex mathematical concepts, such as possess no possibility of
direct sensuous realization, that are continually used in the construc tion of mechanics and physics. Conceptions, which are completely
alien to intuition in their origin and logical properties, and transcend it in principle, lead to fruitful applications within intuition itself.
This relation finds

infinite,
yet
is
its most pregnant expression in the analysis of the not limited to the latter. Even so abstract an intel-
117
lectual creation as the system of complex numbers offers a new example of this connection; Kummer, for instance, has developed the thought that the relations, which prevail within this system, possess their concrete substratum in the relations of chemical combina tion Chemical combination corresponds to the multiplication of the complex numbers; the elements, or more exactly the atomic weights of the same, correspond to the prime factors; and the chemical formulae for the analysis of bodies are exactly the same as the formulae for the analysis of numbers. Even the ideal numbers of our theory are found in chemistry, perhaps only too often, as hypo thetical radicals, which have hitherto not been analysed, but which, like the ideal numbers, have their reality in compounds The analogies here indicated are not to be regarded as a mere play of wit, but have their justification in the fact that chemistry, as well as the part of the number theory here considered, have both the same fundamental concept as their principle, namely, that of com
.
position, although in different spheres of
being."
The
real
problem
is, however, precisely this transference of structures, whose whole content is rooted in a connection of purely ideal constructions, to
Even here it appears that it is and elements, As soon as we take one step beyond the first naive observation of isolated facts, as soon as we ask about the connection and law of the real, we have transcended the
the sphere of concrete factual being.
upon a peculiar interweaving of that every scientific theory rests.
"real"
"not-real"
strict limits prescribed
indicate this connection
by the positivistic demand. In order even to and law clearly and adequately, we must go
back to a system that develops only universal hypothetical connec

tions of grounds and consequences, and which renounces in principle the "reality" of its elements. That form of knowledge, whose task
and lay bare its finest threads, begins by aside from this very reality and substituting for it the turning symbols of number and magnitude.
is
to describe the real
Mechanism and
scientific
the concept of motion.
The
first
phase of the
The exact clearly expresses rooted in the thought of mechanism, and concept can only be reached on the basis of this thought. The explanation
theory
of
nature
this.
of nature
is
of nature, in its later development,

Crelle s Journal, Vol. 35, p. 360. Zahlensysteme, p. 104.
1
may
attempt to free
itself
from
Cited by Hankel, Theorie der komplexen
118
schema and to replace it by one broader and more universal; nevertheless, motion and its laws remain the real problem in connec tion with which knowledge first becomes clear regarding itself and
this first
its task. Reality is perfectly understood as soon as it is reduced to a system of motions. This reduction, however, can never be accomplished as long as consideration remains in the sphere of mere data of perception. Motion, in the universal scientific sense, is
nothing but a certain relation into which space and time enter. Space and time themselves, however, are assumed as members of
immediate, psychological and "phenomenal" meaning. As long as we understand by space nothing else than a sum of various visual and tactual impressions, qualitatively different from each other according to the special physiological conditions under which they come into existence, no "motion" is possible in it in the sense of exact physics.
this relation not in their
properties, but in their strict mathematical
This latter demands the continuous and homogeneous space of pure geometry as a foundation; continuity and uniformity, however,
never belong to the coexistence of the sensuous impression itself, but only to those forms of manifold, into which we constructively
transform
it
by
certain intellectual postulates.

itself is also
(C/.
above
Thus, from the beginning, motion
drawn
p. 105.) into this circle
of purely conceptual determinations. It is only in appearance that it forms a direct fact of perception, indeed the fundamental fact,
first presents us. Perhaps the change of the qualitative difference of successive presentations sensations, i.e., may be conceived in this way; but this aspect alone is in no sense sufficient to ground the strict concept of motion, that is needed by
which
all
outer observation
Here unity is demanded along with diversity, identity along with change and this identity is never provided by mere obser The vation, but involves a characteristic function of thought.
mechanics.
;
individual positions of Mars, which Kepler took as a basis, following the observations of Tycho de Brahe, do not in themselves alone contain the thought of the orbit of Mars; and all heaping up of partic
ular positions could not lead to this thought, if there were not active from the beginning ideal presuppositions through which the gaps of
actual perception are supplemented. What sensation offers is and remains a plurality of luminous points in the heavens; it is only the pure mathematical concept of the ellipse, which has to have been previously conceived, which transforms this discrete aggregate into

a continuous system.
119
Every assertion concerning the unitary path involves the assumption of an infinity of possible moving body places; however, the infinite obviously cannot be perceived as such, but first arises in intellectual synthesis and in the anticipation of a
of a
Motion is gained as a scientific fact only after we produce by this law a determination that includes the totality of the space and time points, which can be constructively generated, in so
universal law.
far as this determination coordinates to every moment of continuous time one and only one position of the body in space. Thus, from a new angle it is revealed that even the first approach
to mechanics depends upon presuppositions, which go beyond possible sensuous experience. The well-known definition of Kirchoff, which defined the task of mechanics as the complete and unambiguous
movements taking place in nature, may be entirely meaning which its author connected with it, yet without the philosophical consequences that are ordinarily drawn from it being thereby justified. Kirchoff himself leaves no doubt that the "description," at which he aimed, has the exact mathemati cal equations of motion as a presupposition, and in them is involved the concepts of the material point, of uniform and variable velocity
description of the
justified in the
as well as uniform acceleration. All these concepts may justifiably serve the mathematical physicist as fixed and immediate data; but they are in no way such for the epistemologist. For the latter, a
"nature" exists in which movements are found as describable objects only as a result of a thorough intellectual transformation of the given. This mathematical transformation, which the physicist assumes to have taken place, constitutes the real and original problem. If the
thought of the continuity and uniformity of space, as well as the exact concepts of velocity and acceleration, are grasped and grounded, then with the help of this logical material the totality of possible phenomena of motion can be completely surveyed and mastered in its form; but the question arises all the more urgently as to the intel lectual means by which this result is reached. The "subject" of motion. This ideal dependence stands out most sharply when we pass from the process of movement to the subject of movement. Again it seems as if this subject could be
directly pointed out in perception; tangible and visible qualities, to
it
is
body,
it
is is
which motion
a complex of ascribed as a
property.
Even
at this point, however, sharper conceptual analysis
120
meets peculiar difficulties. In order to serve as the subject of move ment, the empirical body must be definitely determined, distinguished and limited from all other structures. As long as it is not enclosed in fixed and unchanging limits, by which it is separated from its sur roundings and recognized as a whole of individual form, so long it is
unable to furnish a constant point of reference for change. Yet the bodies of our world of perception never satisfy this condition. They owe their determinateness merely to a first and superficial unifica tion, wherewith we unite into a whole parts of space that seem to
Where such a properties. determined with absolute exacti unity begins and ends can never be tude; a keener faculty of sensuous discrimination would show us, at the point where two different bodies seem to be in contact, a con
possess approximately the
same sensuous
stant reciprocal exchange of parts and thus a continuous movement of the limiting surfaces. Only when we ascribe a strict geometrical
form to the body and thus
raise
it
from the sphere of the merely
perceptible to the determinateness of the concept, does it attain that And identity, which makes it useful as the "bearer" of motion.
body is required with respect to all elements on the other hand, it is to be demanded that it represent a strict unity in itself. As soon as we think of its individual parts as movable with relation to each other, the supreme condition of the definiteness of the point of reference is again aban doned in place of the one movement have been substituted as many different movements as there are independently moving particles. Thus a system must be taken as a basis, which is closed off from the outside and also is in itself incapable of further differentiation and
as exact limitation of the
of its outer surroundings, so
;
disintegration into a plurality of independently moving subjects. The "rigid" body of pure geometry has to be substituted for the
perceptible
body and
its limitless
is
the exact theory of motion
changeability, if the grounding of to be accomplished.

its
The
"limiting
concept"
and
significance for
natural
science.
(Karl Pearson). In fact, the necessity of such a transformation of the problem is recognized and emphasized by the adherents of the theory of "description" themselves. It is Karl Pearson above all, who has described this process with clearness and emphasis in his
work, The Grammar of Science. As he explains, it is never the contents of perceptions as such that we can use as foundations for the judgments of pure mechanics, as points of application in the
121
expression of the laws of motion. Rather, all these laws can only be asserted with meaning of the ideal limiting structures which we
conceptually substitute for the empirical data of sense-perception. Motion is a predicate that is never immediately applicable to the of the surrounding sense-world, but holds solely of that other "things"
which the mathematician substitutes for them in his Motion is not a fact of sensation, but of thought; not of "perception" but of "conception." "Startling as it may,
class of objects,
free construction.
when
first
stated, appear,
it is
nevertheless true that the
mind strug
motion of anything which is neither a geometrical point nor a body bounded by continuous surfaces; the mind absolutely rebels against the notion of anything moving but these conceptual creations, which are limits unrealizable, as we have seen, in the field of perception." Groups of sensuous impres sions can change, can lose old parts and gain new, can form into new groups; but these changes in no way signify the real object of mechanics. is in the field of conception solely that we can talk of the motion of bodies; it is there, and there only, that properly
gles in vain to clearly realize the
"It
geometrical forms change their position in absolute time that is, move." The contradictions, in which mechanics often becomes
involved and which have come to light especially in the attempts to apply the general mechanical laws to the movements of the ether,
can be explained for the greater part by the fact, that the two spheres knowledge here opposed have not been sharply and definitely separated from each other. These contradictions disappear as soon
of
learn not to confuse immediately sensuous with conceptual elements, as soon as we give up trying to conceive an intellectual
as
we
construction for the establishment of a scientific order of phenomena,
What we can alone as itself a particular phenomenal existence. in physics is the construction of a world of geometrical accomplish
forms; yet these, in the multiplicity of movements we ascribe to them, reproduce and represent with wonderful exactness the complex phases of our sensuous experience. As soon as we read this whole thought-world directly into the sense-world again, as soon as we transform its logical assumptions directly into parts of reality, which
would thus be apprehended by sensation, we
fall
once more into
all
the antinomies that necessarily inhere in every type of dogmatism, 2 All this exposition of Pearson s physical as well as metaphysical.
2
Cf. Pearson,
ff
.,
The Grammar of Science, Second edition, London

282, 325 etc.
1900, p.
198
p. 239
ff.,
122
admirable but we ask in vain how, on these assumptions, mechanics can still be conceived as purely descriptive science. Can it be called a description of perceptual contents to substitute in place of them a
is
;
system of geometrical
ideals, such as are necessarily foreign to the world of our perceptions? If the task of a true "objective" descrip tion is to conceive the given as faithfully as possible, neither adding
it is
nor subtracting anything: then, on the contrary, sort of transformation of the initial experience,
precisely that
which constitutes
the character and value of the intellectual procedure of physics. Instead of a mere passive reproduction, we see before us an active
which transports what is at first given into a new logical It would be a strange way of describing what is presented, sphere. if for this purpose we concerned ourselves with bare concepts, which can themselves in no way be "presented." P. du Bois-Reymond s theory of the limiting concept. The question
process,
as to the character of the fundamental concepts of natural science merges here into a more general problem. We saw how the first step
in the formation of these scientific
place of the
members
of a certain
limit of this manifold.
The
concepts was to introduce, in sensuous manifold, the ideal justification of such construction of a
limit cannot be perfectly

it
remains purely within
is slight as long as logic and epistemology themselves have not reached clarity on this point. Here more than anywhere else they both seem entangled in insoluble difficulties; and the only way out for clear thought seems to be not so much in resolving the antinomies, which appear at this point,
general logical reduction of the question, however,
demonstrated by natural science, as long as its sphere; yet the construction rests on The advantage to be gained from this principles.
as rather in understanding and recognizing

fact, this decision is expressly cian in more recent times.
them
in their insolubility.
In
represented by a noted mathemati According to him, the consideration of
the concept of the mathematical limit leads back to a fundamental
metaphysical problem, which, like all problems of this species, is not to be solved according to strict objective criteria, but according to the subjective inclination of the individual investigator. The "general
theory of functions," as Paul du Bois-Reymond develops it, illumines this dualism on all sides; but disclaims from the beginning any
exists
attempt to remove it. When we raise the question whether there an exact limit to a definite given sequence of presentations, as

for
123
example to the
figures of a decimal fraction, such that the limit
possesses the same existence as the members of the sequence them selves, the answer we give cannot be clearly determined by logical and mathematical considerations alone. The simple mathematical
problem leads us into the conflict of two general views of the world, which stand irreconcilably opposed. We must choose between these two views of the world: either with empiricism we must assume as existent only what can be pointed out as an individual in the real presentation, or with idealism, affirm the existence of structures, which constitute the intellectual conclusion of certain series of presentations, but which can never themselves be directly presented. The mathematician is not in a position to grant the victory to either one of these fundamental views all that he can and must do,
;
in order to bring clarity into the foundations of analysis, is to follow them to their ultimate intellectual roots. The solution of the riddle is
that
of
it
remains and
observation of our
its
relations to
"The most persistent du Bois-Reymond, "and thought-process," says perception does not go beyond showing that there
will
remain a
riddle.
are two completely different interpretations, that have equal claim to serve as the foundations of exact science, because no absurd
consequence of them is found, at least as long as we are concerned with pure mathematics However, it remains a very phenomenon that, after the removal of everything which strange
might conceal the truth and when at last one might expect to behold its image clearly and definitely, it appears before us in double form. He, who first noticed through a transparent crystal the double image of a single object, could not have showed it to his friends more moved, than I today, at the end of most careful and
eager -reflection,
am
forced to develop before the reader the double

3
Interpretation of the foundations of our science." he problem of existence. It is, in fact, worth while to seek out the
mgin
of this .peculiar result
for here
we stand
at a point, which
represents the decisive turning-point of all critique of knowledge. The old question as to the relation of concept and existence, of idea
and
form.
meets us once more hi a characteristic and original Indeed the suspicion must at once arise as to whether the opposition found here between "empiricism" and "idealism" reprereality, here
3
Paul du Bois-Reymond, Die allgemeine Funktionstheorie, Tubingen 1882,

f.
p. 2
124
sents a complete disjunction, whether it contains within itself all In this case only, would the antinomy possible manners of thinking. be insoluble while if it could be shown that there are problems wholly
;
removed from the opposition, which here serves as a starting-point, and which are thus wholly independent of its solution in their logical In structure and validity, the antinomy would lose its sharpness. truth, it appears even in the first arguments of du Bois-Reymond, that it is not the mathematician but the philosopher and psychologist,
What in the world could "the persistent is speaking here. observation of our thought-process and its relations to perception"
who
problem?
that
it
contribute to the solution of any particular, specifically mathematical Pure mathematics is precisely characterized by the fact
abstracts completely from
all
such investigation of the thought-
process and its subjective conditions, and merely directs itself upon the objects of thought as such and their objective logical connection.
The manner,
in
which the concept of existence alone appears within
mathematics, confirms this exclusive direction of interest. The student of algebra, who speaks of the "existence" of the numbers e and TT, undoubtedly intends to signify no fact of outer physical reality; but just as little is it the presence of certain contents of presentation in any perceiving and thinking subjects, that is to be
thereby affirmed. If this were the meaning of the assertion, the mathematical standpoint would lack any means of testing and verifying it; for only experiment and generalizing induction warrant us in making a decision concerning real events in the psychic life of
number e means nothing else than number system, one and only one position is determined definitely and with objective necessity by the series, which we apply in its definition. If we assume the general rule
individuals.
The
existence of the
that, within the ideal
T2
is
+ ro +
(in inf.)
then by
it
the system of rational
numbers
contains
analysed into two strictly divided classes, of which one elements, that are ever exceeded by the series when it is carried far enough, while the other contains all those elements, with
all
which
it
this
is
not the case.
By virtue
effects in the field of rational
of this complete division, which numbers, the series gains a definite

it
relation to the relation of
members
and
of this field, since
stands to
them
in the
"before"
"after,"
and thus
of
"smaller"
and
"greater."
The
"number"
validity of all these relations alone justifies us in speaking of a e, and constitutes the entire "being," the complete and

self-contained existence of this number.
(Cf.
125
61
above
p.
f.)
The
determination, which arises in this way, is, although purely ideal, nevertheless in principle fif no other sort than the whole numbers and
is just as strictly and sharply dis from that of any other number, however near e it may tinguished Here lie, as the value of 1 is distinguished from the value of 1,000. we do not appeal in any way to the faculty of separating presentations and similar particular contents of perception in consciousness; we are concerned on both sides with pure concepts, which are sufficiently divided from each other by the logical conditions, which their defini tion imposes upon them. The existence of the limiting point. It appears, indeed, to be other wise when we turn from the algebraic meaning of limit to its geomet rical meaning. The existence of a point seems to be verifiable, in fact, only by a procedure, which allows us to point it out in intuition and to distinguish it from other positions. Here, however, certain limits to further advance are felt, on the basis of the psychological If we remain at the principle of the threshold of discrimination.
fractions: since the value of e
standpoint of the "empiricist," if we hold to the belief that we are only justified in assuming a particular "thing," where there is a particular presentation for its representation at our command, then
we
see that, according to this assumption, the existence of a limiting point for any definite converging sequence of points can never be proved from the consideration of the sequence itself. For example,
we think of the individual numbers of a convergent series as repre sented by points on the abscissa, then all these points, as we advance further in the series, will move nearer and nearer to each other, until
if
finally
our intuition
is
unable further to separate them.
After a
into
certain
member, the terms become indistinguishable and flow
each other; we are, accordingly, not in a position to decide finally whether that point, which corresponds to the algebraic limit of the series, exists as a particular geometrical individual, or whether only
those positions possess reality, which are algebraically expressed by the members of the series itself. "We demand, in fact, what is
impossible,"
remarks du Bois-Reymond, we demand that the sequence of points abstracted from the given points shall determine a point not belonging to the given points. I hold this to be so incon ceivable that I affirm, th^t no intellectual labor will extort from a brain such a proof for the existence of the limiting point, even if it
"if
126 united
Newton
s gift of
divination, Euler s clarity

Gauss."
4
and the crushing
power F
of the intellect of
It is entirely correct Logical idealism on the problem of existence. that all these powers would not suffice to produce the desired proof;
for with the mere question under investigation we have already set :; ourselves outside the field of pure mathematics. To "prove" the
is here taken, will never be attempted by anyone who has ever made fully clear to himself even the critical refutations of the ontological argument. The deeper ground of all the misunderstandings and contradictions, however, lies here also in the indefmiteness and ambiguity, in which
existence of points, in the sense in which existence
the concept of being is understood. The "being" of the geometrical point is not different in principle from that of the pure numbers,
The construction of the logical sphere. manifold takes place, it was seen, according to laws geometrical thoroughly analogous to those of the systematic development of the system of numbers. Here as there we start from an ideal postulation of unity, and here as there intellectual progress consists in our
taking up into the system
original
all
and belongs to no other
elements, that are connected with the
by an unambiguous conceptual relation or a chain of such relations. We saw how the paradoxes of imaginary and infinitely distant points were solved from this standpoint little as these points
:
could claim for themselves any sort of mysterious "reality" in space, they proved themselves, on the other hand, an expression of valid
5 spatial relations.
Their being
is
exhausted in their geometrical
meaning and necessity. (Cf. above p. 83 ff.). It is this necessity which the true "idealism" can alone demand and claim for the struc
tures of pure mathematics. On the other hand, the idealist in the sense of du Bois-Reymond, goes far beyond such a demand. "The fundamental view of the idealistic system," we read here, thus the real existence not only of that which is presented, but of the intuitions necessarily following from the presentations
"is
The idealist believes in some sort of existence (Vorhandenseiri) of unpresentable, verbal conclusions of sequences of presentations,
generated by our thought
4 6
process."
Here speaks an
"idealist,"
Allgemeine Funktionentheorie,
Cf.
p. 66
f.
pertinent criticism of the doctrine of P. du Bois-Reymond, System einer Theorie der Grenzbegriffe, Lpz. and \ v ien, 1900, p. 175 ff
.
Kerry
Funktionentheorie, p. 87. Cf. Ju Bois-Reymond s work, I ber die Grundlagen der Erkenntnis in den exakten fcissenschaften, Tubingen.
^
Allgemeine
1890, p. 91.
127
who has permitted

the
"empiricist,"
his conception to be perverted by his opponent as we can easily see, when as here he only recognizes
the existent (Vorhanderi) as true. The whole antinomy unfolded in the Allgemeine Funktionentheorie disappears as soon as we destroy this confusion of truth and reality, which is common to the advocates
of both these.
The consequences reality. even more sharply in the interpretation of appear the fundamental concepts of natural science than in the purely mathematical discussion. These concepts are drawn into the same conflict they also continually go beyond the given, but this unavoid able process cannot be critically justified and grounded. We cannot
Consequences of the confusion of truth and
of this confusion
;
abandon the concepts
of the absolutely rigid body, of the
atom
or of
force at a distance, although, on the other hand, we must give up all hope of finding a direct verification of them in any part of the outer
world of perception. The consciousness of the limits set to all our knowing by its nature and essence is felt increasingly here. Ever anew we find ourselves led to unpresentable elements, which lie
behind the known and accessible world of sensuous appearance, and ever again it appears, when we attempt to grasp and analyse them, that no intelligible meaning can be gained. "Our thought is as if
paralyzed, and
makes no
us.
"We
progress."
The organ
for reality
is
and
remains denied to
are enclosed in the box of our perceptions

blind.
and
for
what
is
beyond them born
We
cannot have a glim
of light, for a glimmer is already light: but what corresponds in the real to light?" 7 This radical scepticism, in which the expo sition of the foundations of exact knowledge here results, is a
mer
consistent
and
significant
consequence.
"organ"
On
the
basis
of
this
for the real; for the necessary possess no which form the real organs for the logical interpretation concepts, and mastery of the manifold of sensations, are transformed into mysterious realities behind the phenomena. The "idealization" of presentations. If this transformation is once understood, however, the mist is again dispelled, which threatened to settle more thickly around the image of scientific reality. Indeed, this image arises first through a process of idealization, in which the
view, in fact,
we
indefinite data of sensation are supplanted

7
by
their strict conceptual
P,
du Bois-Reymond.
Uber die Grundlagen der Erkenntnis in den exakten
Wissenschaften, Abschn. VIII.
128
limits.
But the assertion of the objective validity of this process is not the same as the assertion of a new class of objects. "Our field of affirms du Bois-Reymond s "idealist," "contains not thought,"
only the mosaic of the perceptible and the images and concepts deduced from it by the process of thought, thus by transformation and combination, but there dwells within us the indestructible
conviction
....
of the
8
the system of
presentations."
presence of certain things outside This proposition is undoubtedly
correct, in so far as by "system of presentations" nothing else is understood than the mass of given perceptions, than the system of
colors
tions.
and tones, of tastes and odors, of pressure and contact sensa But the completion of this "mosaic of the perceptible" cannot
take place by our simply inserting new "insensible" things into this first empirical reality; for the parts of the mosaic would thereby, indeed, be moved together more closely and densely, but in spite of
no new form of connection, no deeper relation would be gained. The aggregate of sensuous things must be related to a system of necessary concepts and laws, and brought to unity in this relation. This process of thought, however, demands really more than the mere combination and transformation of parts of presentations; it pre supposes an independent and constructive activity, as is most clearly manifest in the creation of limiting structures. The "empiricist" also must accept this form of idealization; for, without it, the world of
this,
perception would not be merely a mosaic but a true chaos. It is a mere misunderstanding when he affirms that he does not recognize that the absolutely straight line and the absolutely exact plane exist, but only more or less straight lines, more or less exact planes. For
this
very discrimination of different stages of exactitude presupposes comparison with the exact idea, whose fundamental function is thus here throughout confirmed. The "being" of the idea, however,
and needs no other support and no other Also the ideal concepts of natural science affirm nothing regarding a new realm of separate absolute objects, but they would
consists in this function
proof.
only establish the inevitable, logical lines of direction, by which alone complete orientation is gained within the manifold of
phenomena. They only go beyond the given, in order to grasp the more sharply the systematic structural relations of the given.
8
Allgemeine Funktionentheorie, p. 110
f.

As soon
129
as the empiricist, along with du Bois-Reymond, character izes idealization as throughout justifiable, and explains that he only
9 refuses to accept the ideal itself, all conflict is removed in principle. For the existence of the ideal, which can alone be critically affirmed
and advocated, means nothing more than the objective logical neces That we are not here concerned with such a sity of idealization. not with an arbitrary play of phantasy, becomes clearer necessity, the more deeply the concept of the object is analysed into its condi
tions.
It is
vain to interpret the ideal limits, ascribed to certain
sequences on the basis of conceptual criteria, as mere verbal conclu "The sions with no real or logical meaning corresponding to them. it is affirmed, "can in no way be grasped as a pictorial perfect," presentation. Nevertheless, as it enters our thought and finds and as our thought consists in the application there, succession of presentations, so it must be somehow a presentation, and it is, namely, as a word. The sequence of objective presenta tions of what is exact have, therefore, as their conclusion, a word for 10 This nominalism, however, something that cannot be presented."
....
fails in
the explanation of the concept of limit, as
it
has already failed

p.
in the explanation of the pure
numbers.
(Cf.
above
43
ff.)
For
here precisely the characteristic meaning and the real function of the concept of limit is obviously excluded. Between the limit and the
members
of the series, certain relations hold,
which are mathe
The "number" arbitrarily changed. numerical relations to the other numbers, that are gained from the partial sums of the defining series it takes its place with them in a series, in which the position of each element, its
matically fixed
and cannot be
e stands in certain
Is there any meaning in later, is unalterably ascribed. that there are such relations of order in the sequence, in asserting the greater and smaller of elements, where one is taken as an actual, psychologically significant image, while its correlate is made to consist
earlier
and
There can only be valid mathematical relations between ideas and words. From this connection with the relation of the ideal and reality. /The logic of mathematics, we can explain and understand better why any attempt to interpret the concepts of natural science as mere aggreof a
ideas, not
9
mere sound? between ideas and
10
Allgemeine Funktionentheorie, p. 118. Grundlagen der Erkenntnis, p. 80;
cf.
Allgemeine Funktionentheorie,
p. 95.
130
No scientific gates of facts of perception must necessarily fail. but is related to the ideal limits, theory is directly related to these facts,
We investigate the intellectually. bodies by regarding the masses, which affect each other, impact of as perfectly elastic or inelastic; we establish the law of the propagation of pressure in fluids by grasping the concept of a condition of perfect
which we substitute for them
ture and
investigate the relations between the pressure, tempera of gas by proceeding from an "ideal" gas and com a hypothetically evolved model to the direct data of sensation. paring "Such extrapolations," says so convinced a "positivist" as Wilhelm
fluidity;
we
volume
a procedure very generally applied in science; and a very large part of the laws of nature, especially all quantitative laws, i.e., such as express a relation between measurable values, only hold
Ostwald,
"are
exactly for the ideal case. We thus stand before the fact that many and among them the most important laws of nature are asserted and hold of conditions, which in reality in general are never found." 11
The problem here

formulation.
If
raised reaches further than appears in this initial
the procedure of natural science only consisted in substituting the ideal limiting cases for the directly observable phenomena, then we could attempt to do justice to this method by a
simple extension of the positivistic schema. which the theoretical consideration of nature
For the objects with

is
concerned, although
beyond the real field of empirical perception, seem to lie on the same line with the members of this field and the laws, that we assert, do not seem to represent a transformation so much as a mere extension of certain perceptible relations. Yet, in sooth, the between the theoretical and factual elements at the basis of physics cannot be described in this simple way. It is a much more /relation complex relation, it is a peculiar interweaving and mutual interpenetration of these two elements, that prevails in the actual structure of science and calls for clearer expression logically of the relation between principle and fact.
they
fall
;
in
The problem of the physical method and its history. In epistemological discussion of the foundations of natural science, we often meet the view that the ideal of pure description of the facts is a specifically
modern achievement.
11
It is
thought that here for the
first
time
Ostwald, Grundriss der Natur philosophic (Reclam), p. 55.
131
physics has reached true clarity regarding its proper goal and intel lectual instruments; while before, in spite of all the wealth of results, the way to these results remained in darkness. The separation of
"physics" and "metaphysics," the exclusion in principle of all factors which cannot be empirically confirmed, is thought to be the decisive product of the critically philosophical and most modern research.
signifies a misapprehension of the continuous which physics has reached its present form. From the by first scientific beginnings of physics, the problem of method has been continually and vitally important, and it was only in struggling
This view, however,

course
with this problem that physics gained full mastery over its field of Reflection and productive scientific work have never been strictly separated here, but have mutually assisted and illumined each other. And the further back one follows this reflection, the
facts.
more
in
it.
clearly
is a fundamental opposition of viewpoints discoverable This opposition persists in the modern expositions un weakened
;
but
its
it
gains full sharpness and definiteness,
when we
trace
it
back to
general systematic and historical sources. The problem of knowledge. (Plato.) As modern investigation has more and more destroyed the prejudice that the scientific use of experiment was unknown to the Greeks, so we can also indubitably
recognize in ancient philosophy the theoretical controversy over the The conflict, that begins here, principles of empirical knowledge.
whole speculative view. It is expressed in an incompar and unforgettable picture in the Platonic metaphor of the cave. For the human mind, there are two types of consideration and judgment regarding the phenomena of the world of sense, phe nomena which pass across the mind like shadows. The one is satisfied
affects the
able
merely with grasping the sequence of the shadows, with fixing their
before and after, their earlier and later. Custom and practice gradually enable us to distinguish certain uniformities in the sequence of phenomena, and to recognize certain connections between them as
uniformly recurrent, without the grounds of this connection being Common understanding and the view of the intelligible to us. world based upon it do not need these reasons; for both it is sufficient
they are able to predict one phenomenon from another by means of the empirical routine which they have made their own, and to draw the phenomenon into the circle of practical calculation. Philosophi cal insight, however, begins with a withdrawal from every such
if
132
manner of consideration; it presupposes the "turning" of the soul to another ideal of knowledge. Not the phenomena in their bare succession, but the eternal and unchanging rational grounds, from
which they proceed, are the unique object of knowledge. To grasp these rational grounds, this realm of \6yoi in the phenomena them selves was, indeed, according to Plato, denied to thought. Whoever,
had once grasped the nature of insight into the necessary, only turned back under compulsion to the con sideration of a field in which, owing to the flowing and indefinite character of the objects, the same rigor of connection could never be
in the field of mathematics,
reached.
In this sense, empirical knowledge of the sequence of phenomena is not the completion and fulfillment of the pure knowl edge of ideas, but only serves as a dark background, against which
the clarity of purely conceptual investigation and knowledge stands out the more strongly.
it
Furthermore, sceptical theory of knowledge. (Protagoras, etc.) highly probable that this opposition signifies no intellectual construction (of Plato s), but that it represents with radical sharp
is
The
ness a concrete historical opposition, which was already developed at the time of Plato. 12 In any case, the whole later development of
scientific investigation in
antiquity
is
dominated by
this Platonic
everywhere echoed in the controversy between the "empirical" and the "rational" physicians, that runs through Greek medicine. But the more investigation was applied to the discovery and establishment of individual facts, the more the value and order
distinction.
It is
was changed. Scientific empiricism expressed itself in the sceptical theory of knowledge, and affirmed as its positive signifi cance and distinction the very feature, which Plato regarded as the
of knowledge
It is, indeed, not given to lasting defect of all empirical knowledge. knowledge to comprehend the essence of things from a universal
What remains for us is only the observation of principle of reason. the customary sequence of phenomena, which enables us to use one phenomenon as the sign of another. The task of science is fulfilled
grouping and sifting such signs, each one of which awakens in us a memory, and thus directs our expectation of the future into fixed paths. The real causes of occurrences remain unknown to us; but we do not need to know them, as the real and final goal of all
in
certain
12
Cf. Natorp,
Forschungen zur Geschichte des Erkenntnisproblcms im Alterff.
tum, Berlin 1884, p. 146
133
theory lies in the practical consequences of our action. These consequences remain essentially the same, whether we logically comprehend how one event issues from another, or merely accept the fact of a certain empirical coexistence or succession, and rest in it. The concepts of nature and purpose. (Plato.} We recognize,
however, in the case of Plato, that the division he draws between
rational
tion of
is
and empirical knowledge, 13 produces no complete disjunc the entire field of knowledge. Empirical knowledge, which
satisfied with the sequence of "shadows," is sharply characterized; but there remains an indefiniteness in the characterization of its ideal counterpart. This fact is the more significant since it has
constantly recurred in the historical development of the problem. The real balance and division of the matter is rendered obscure as
is exactly defined, while the other falls into two between which opinion varies. At first, Plato opposes to the bare sequence of phenomena, insight into their ideo We do not possess true knowledge of natural logical connection. as long as we simply permit them to run off before us as processes before an indifferent spectator, but we first have true knowledge
long as one
member
different meanings,
when we survey the

ordered whole.
total
movement
of the process as a purposively
We
must understand how one element demands
another; how all the threads are mutually interwoven finally into one web, to form a single order of the phenomena of nature. The ethical idealism of Socrates lives in this view of nature. As little as the continuance of Socrates in prison can be explained by describing the position and relation of his muscles and cords, without con sidering the ethical reasons that determined him to obey the law, just as little can an individual event be truly understood, as long as its place in the total plan of reality is not clearly distinguished. For example, if we attempt to explain the fact that the earth moves
freely in the center of the universe, no sensuous connection, no mechanical vortex of bodies or any other cause of the same sort, can
satisfy us;
to be pointed out as the Sensuous being must be reduced to its ideal reasons; the conclusion of the world of ideas is the Idea of the Good, into which all concepts ultimately merge. Another view, however, is found in Plato opposed to this deduction
"the
but
good and
right"
alone
is
ultimate and decisive basis of the fact. 14
13
14
Cf. especially Republic, 509 Cf. Phaedo 99 f., 109.
ff.
134
of natural
phenomena out
of purposes.
It
is
rooted in Plato
interpretation of mathematics, which is for him the "mediator" between the ideas and the things of sense. The transformation of
empirical connections into ideal ones cannot take place without this middle term. The first and necessary step throughout is to trans
form the sensuous
indefinite,
which as such cannot be grasped and
enclosed in fixed limits, into something that is quantitatively definite, that can be mastered by measure and number. It is especially the
later Platonic dialogues, as for
example the Phikbus, which most
clearly developed this postulate.
must be confined
quantity, before
in strict limits,
The chaos of sense perception by applying the pure concepts of

"less,"
it
rest with the indefinite

"weaker,"
can become an object of knowledge. We cannot or "more" with the "stronger" or
which we think we discern in sensation, but we must strive throughout for exact measurement of being and process. In this measurement, being is grasped and explained. 15 Thus we stand before a new ideal of knowledge, one which Plato himself recognized as in immediate harmony with his teleological thought, and combining with it in a unified view. Being is a cosmos, a purposively ordered whole, only in so far as its structure is characterized by strict mathematical laws. The mathematical order is at once
the condition and the basis of the existence of reality; it is the nu merical determinateness of the universe, that secures its inner selfpreservation.
Mathematics and teleology. (Plato, Aristotle, Kepler.} In Aristotle the two lines of thought have already separated, which were insepar ably connected for Plato. The mathematical motive recedes into
the background and thus only teleology, the doctrine of final causes, remains as a conceptual foundation of physics. The outer process and its quantitative order according to law merely mirrors the dynamic process, by which the absolute substances maintain and
;
develop
themselves.
The
empirical-physical
relation
of
bodies
from their essence, from the immanent purpose given them by their nature, and which they progressively strive to fulfill. Thus the elements are arranged in the cosmos according to
results ultimately
the degree of their affinity, while those that agree with each other in any quality lie next each other; thus each body retains a tendency
to
its "natural
15
place,"
prescribed to
it
by
its
property, even after
it
Cf. Philebus
16,
24
f.
135
has been forcibly deprived of this place. Here the true, inner causes of every physical connection are revealed; while the mathematical mode of consideration, on the other hand, does not attain to the causes but only to the quantities of being, and is limited to the
"accidents" and their sphere. Thus a new opposition becomes henceforth effective in history. The unity of the teleological and the mathematical methods of consideration, which still existed in
Plato
system of nature,
is
relation of superordination
destroyed, and its place is taken by a and subordination. The line of division
has moved; now not only is the sensuous observation of contingent, empirical uniformities excluded from the highest ideal knowledge of the supreme causes, but also the exact representation of the processes in pure concepts of magnitude is excluded from this ideal knowledge.
Here the opposition between the empirical and the speculative views of nature is first sharply defined. In modern times, mathematical physics first seeks to prove its claims and independence by going back from the philosophy of Aristotle to that of Plato. Above all, 16 it is Kepler of whom this reversion is characteristic. With energy and clearness he repudiates a conception that would reduce the mathematician to a mere calculator and exclude him from the community of philosophers and the right to decide as to the total structure of the universe. Absolute substances and their inner forces are indeed unknown to the mathematical physicist and must remain so, in so far as he simply pursues his own task, free from all extraneous problems; but his abstraction from this problem in no way signifies his persistence in the ordinary empirical method, which is satisfied with the mere collection of individual facts. The mathe matical hypothesis establishes an ideal connection among^ these facts; it creates a new unity to be tested and verified by thought, but which cannot be directly given by sensation. Thus the true hypothesis limits the field of mathematical physics to two different directions. It expands immediate experience inta theory by filling up the gaps left by direct observation, and by substituting a continuous connec
tion of intellectual consequences for isolated sense-data. On the other hand, it limits itself to representing this system of conse
quences merely as a system and dependence of

16
magnitudes.
At
The more exact evidence

Zeit, I 258
ff .,
in
my work Dos
Erkenntnisproblem in der Philosophic

308
ff.,
for the following historical exposition is given u. Wissenschaft der

ff .
neuen
II 322
136
the same time, the mathematical expression of the hypothesis, its algebraic-geometrical form, is the whole of its meaning. In defend
ing the legitimacy of hypothesis, Kepler places its characteristic function elsewhere than does the ordinary speculative philosophy He is not concerned with a transition from the mathe of nature.
matically characterized phenomenon to its absolute causes, but with a transition to a quantitative "understanding" of reality from the
facts of perception (before they have been conceptually worked The scientific physicist can simply leave alone the ques over). tion as to the ultimate "forces," which shape being; yet all the more
first
he must seek to advance from a mere collection of observations to a universal "statics of the universe," to a mastery of the all-inclusive
harmonious order prevailing in it. This order is not directly seized and understood by the senses, but only by the mathematical intellect.
According to this view, the legitimate function of the concept does not consist in revealing a path to a new non-sensuous reality; but it plays its role in the conception of reality of mathematical empiricism (Empiric), and gives it definite logical form. The concept of hypothesis. (Kepler and Newton.} Physics, how ever, did not reach this conception of its problem without vacil lation and difficulty. The particular historical conditions, under which modern natural science developed, forced the negative rather than the positive part of the new task into the center of considera
tion.
First of
all,
and
this
warding
off
the theory had to ward off metaphysical claims; could only be accomplished by bringing to light
the empirical foundations of exact science. The logical factors, on the other hand, remained in the background as long as all philosophi cal power was concentrated upon protecting pure experience from
the incursions of metaphysics.
From this, one understands the fundamental change in view between Kepler and Newton. Kepler,
although he strongly defends the claims of empirical investigation against the metaphysics of substantial forms, nevertheless reverts to the mathematical teleology of Plato in his final conception of the
world.
types,"
The mathematical ideas are the eternal patterns and "arche according to which the divine architect ordered the cosmos. Thus the more deeply we penetrate into the exact structure and exact
more danger there is that the strict between experience and speculation will again be Newton s regulae philosophandi seek especially to meet
presuppositions of physics, the

line of division
effaced.

this danger.
137
Induction is very definitely characterized as the only source of certainty in physical matters. Observation and scientific investigation teach us that those properties, which can neither be
increased nor diminished and which are
common
stitute in their totality the essence of body.
to all bodies, con This expression can

of certain
In this sense, but only in this sense, we can conceive weight as an "essential" property of matter; i.e., we can grasp it as such only in so far as we know of no experiment w hich might occasion us to doubt its universal empirical presence. The
r
thus signify nothing facts of perception.
more than an empirical generalization
question as to the cause of the reciprocal attraction of cosmic masses cannot occupy the real physicist as such and lure him to speculative hypotheses; for attraction is to him nothing but a certain numerical
value, which contains the measure of the acceleration which a body undergoes at each point of its path. The law of the change of this value from point to point contains the answer to all questions, that can be raised with scientific justification regarding the "nature" of
weight. It is Newton s first disciples and pupils, who generalize these explanations and extend them to the whole field of natural
science.
distinctly with them,
The demand for a physics without hypotheses first appears and the technical expression of "description of
It is now recognized as a phenomena" is first formulated here. fundamental failure in method to attempt to form physical explana tions on the model of logical definitions; we make such a failure if, instead of proceeding from the observation and collection of indi vidual cases, we proceed from the hierarchy of concepts and species. Definitions, which claim to discover the ground and essence of any natural process, must be excluded from physics; they form no instru ment of knowledge, but merely hinder the unprejudiced under standing of phenomena, on which depends the whole value of physics
as a science.
The logical and ontological "hypotheses." Further development, however, even within the Newtonian school, showed very clearly what was problematical in this apparently fial conclusion of methodology. If physics should be forbidden the use of hypothesis in every sense, then all elements would have to be removed, which
realization of this
had no immediate correlate in the field of perception. But the demand would mean nothing less than the destruc tion of the Newtonian mechanics and its systematic conception.
138
of absolute space and absolute time, which Newton takes as the starting-point of his deduction, lose every legitimate
The concepts
meaning when measured with the logical criteria, which Newton s methodology alone permits. And, nevertheless, it is precisely on these concepts that the possibility of distinguishing between real and apparent motion rests, and thus the very concept of physical_
This antinomy is insoluble within the limits of the reality itself. Newtonian system; its deeper ground lies in the indefiniteness with which the concept of the hypothesis is here interpreted. Aristotle and Descartes, the metaphysics of substantial causes and the first,
although imperfect, plan of a complete mechanical explanation of the There is a lack of certainty in world are here alike condemned.
distinguishing the assumption of some sort of "dark qualities" of things from the fundamental theoretical thought, which is assumed as a basis for defining and limiting the problem and empirical field of
physics.
And this same ambiguity remains to some extent in modern
discussion, in spite of all attempts at sharper epistemological defini tion of the problem. The .most striking expression of such ambiguy-îty is in the concept of description itself.
For
this
term serves to
unite investigators,
speculative
who merely agree with each other in opposing metaphysics, but who entirely disagree in their positive
Such an investiga interpretations of the logical structure of physics. tor as Duhem, for instance, forcibly and clearly develops the idea that every mere establishment of a physical fact involves certain theoretical presuppositions and thus a system of physical hypotheses,
directly on the side of an "empiricism," which upon a misapprehension of this fundamental double relation. Thus the difficulty involved in its historical development still The necessary and warranted persists undiminished in physics. struggle against ontology leads to an obscuring of the simple logical facts. Philosophical criticism must seek here for a strict separation of these two materially heterogeneous questions, long inseparably connected in history. The relation of physics and logic is still always described by distinguished scientific investigators as if we still stood in the midst of the conflict between Newton and Wolff, which stamped its form on the philosophy of the eighteenth century.
stands
rests
here
new,
This dispute, however, may be regarded as settled; for critical form, has given up metaphysical claims.
standpoint, however,
it is
logic, in its
From
this
new
clear that the
"phenomenalism"
of a
139
Newton is not on the same plane intellectually with that developed and advocated by ancient scepticism. The problem arises to inves tigate the fundamental differences of these two views, which both The concept "phenomena." according as it is applied to an phenomenon indefinite object of sense perception or to the theoretically con structed object of mathematical physics; and it is precisely the
agree in limiting physics to the
field of
of the
itself differs,
conditions of this construction which give rise ever
anew
to the
epistemological question.
IV
methodology of natural science. The discoverer of the fundamental law of modern natural science agrees entirely
Robert
Mayer
in his methodological views with the series of great
investigators
Robert Mayer begins with the of the problem of physics as is found in same theoretical definition The logical Galileo and Newton in the most diverse applications.
that starts with the Renaissance.
continuity appears unbroken, in all the material remodeling of "The most impor physics introduced with the principle of energy.
tant, not to say the only, rule for the true investigation of nature
this: to
is
remain persuaded that
it is
our task to learn to
know
the
phenomena before we seek explanations or ask about higher causes. Once a fact is known on all sides, it is thereby explained and the work of science is ended. This assertion may be pronounced by some as trivial, or combated by others with ever so many reasons;
yet certain it is, not only that this rule has been too often neglected down to the most recent times, but that all speculative attempts, even of the most brilliant intellects, to raise themselves above facts
up till now borne only such language as Kepler used precisely against the alchemists and mystics of his time, or as Galileo used against the Peripatetic philosophy of the school. The question as to how heat arises from diminishing motion or how heat is again changed into motion, is declined by Robert Mayer, just as Galileo do not know what avoided the question as to the cause of weight.
instead of taking possession of them, have
barren
fruit."
17
This
is
"I
heat, electricity
etc.
are in their inner essence
just as little as I
know
;
the inner essence of a material substance or of any thing in general
17 Robert Mayer, Bemerkungen uber das mechanische Aequivalent derWdrme, (Mechanik der Warme, ed. by von Weyrauch, 3rd ed., Stuttgart 1893, p. 236).
140
I know this, however, that I see the connection of many phenomena much more clearly than has hitherto been seen, and that I can give
a clear and good notion of what a force
is."
This, however,
is all
"The sharp that can be required of an empirical investigation. definition of the natural limits of human investigation is a task of practical value for science, while the attempt to penetrate the
depths of the world-order by hypotheses is a counterpart of the In the light of this conception, only numbers, efforts of adepts."
only the quantitative determinations of being and process ultimately remain as the firm possession of investigation. A fact is understood when it is measured: single number has more true and permanent
"a
value than a costly library of
18
hypotheses."
of permanent Hypotheses and natural laws. with the rejection of the false problem. significance is indicated, along A problem is held to be explained, when it is known (bekannt) perfectly
Here a new problem
This definition must, indeed, be accepted without but back of it arises the further question, as to under limitation; what conditions a phenomenon is to be taken as known in the sense The "knowledge" of a phenomenon, which exact science of physics. brings about, is obviously not the same thing as the bare sensuous
and on
all sides.
cognizance of an isolated fact. A process is first known, when it is added to the totality of physical knowledge without contradiction; when its relation to cognate groups of phenomena is clearly estab
lished,
and
finally to the totality of facts of experience in general.
Every
assertorical affirmation of a reality, at the
same time, implies

i.e.,
an assertion concerning certain relations of law,

validity of universal rules of connection.
is
implies
the
When
the phenomenon
brought to a fixed numerical expression, this logical relativity becomes most evident. The constant numerical values, by which
we
but
characterize a physical object or a physical event, indicate nothing its introduction into a universal serial connection. The indi
vidual constants
established
mean nothing in themselves; their meaning is first by comparison with and differentiation from other values. Thereby, however, reference is made to certain logical presuppositions
which lie at the basis of all physical enumeration and measurement, and these presuppositions form the real "hypotheses," that can no longer be contested by scientific phenomenalism. The "true hypoth18
See
Mayer
s letter
to Griesinger (Kleiner e Schriften u.
Briefe, ed. by
Weyrauch, Stuttgart
1893, p. 180, 226 etc.)

esis"
141
nothing but a principle and means of measurement. not introduced after the phenomena are already known and ordered as magnitudes, in order to add a conjecture as to their
signifies
It
is
this
by way of supplement, but it serves to make possible order. It does not go beyond the realm of the factual, in very order to reach a transcendent beyond, but it points the way by which
absolute causes
we advance from the sensuous manifold

tual manifold of measure
of sensations to the intellec
and number.
Ostwald, in his of physical "measurement" against the use of hypotheses, has laid great emphasis on polemic the difference between the hypothesis as a formula and the hypothesis
The presuppositions
Formulae contain merely algebraic expressions; they only express relations between magnitudes, which are capable of direct measurement and thereby of immediate verification by obser
as a picture.
vation.
all
In the case of physical pictures (Bilderri), on the contrary, such means of verification are lacking. Often, indeed, these
pictures themselves appear in the guise of mathematical exposition, so that the given criterion of differentiation seems, at first glance, insufficient. But in every case there is a simple logical procedure,
which always leads to a clear discrimination. "When every magni tude appearing in the formula is itself measurable, then we are con cerned with a lasting formula or with a law of nature; if, on the contrary, magnitudes, which are not measurable, appear in the formula, then we are concerned with a hypothesis in mathe matical form, and the worm is in the fruit." 19 While this postulate of measurability is justified, it is erroneous to regard measurement itself as a purely empirical procedure, which could be carried out by mere perception and its means. The answer given here signifies only the repetition of the real question; for the numbered and meas ured phenomenon is not a self-evident, immediately certain and given starting-point, but the result of certain conceptual operations, which must be traced in detail. In fact, it soon appears that the bare attempt to measure implies postulates that are never fulfilled in the field of sense-impressions. We never measure sensations as such, but only the objects to which we relate them. Even if we
....
grant to psychophysics the measurability of sensation, this insight remains unaffected; for even granting this assumption, it is clear that the physicist at least never deals with colors or tones as sensuous
19
Ostwald, Vorlesungen
iiber
Naturphilosophie, Leipzig 1902, p. 213
f.
142
experiences and contents, but solely with vibrations; that he has noth ing to do with sensations of warmth or contact, but only with temperature and pressure. None of these concepts, however, can
be understood as a simple copy of the facts of perception. If we consider the factors involved in the measurement of motion, the
general solution
definition
is
already given; for
it is
evident that the physical
motion cannot be established without substituting the geometrical body for the sensuous body, without substituting the "intelligible" continuous extension of the mathematician for sensuous extension. Before we can speak of motion and its exact measurement in the strict sense, we must go from the contents of It is perception to their conceptual limits. (C/. above p. 119 ff.) no less a pure conceptual construction, when we ascribe a determinate
of
velocity to a non-uniformly moving body at each point of its path; such a construction presupposes for its explanation nothing less than the whole logical theory of infinitesimal analysis. But even
where we seem to stand closer to direct sensation, where we seem

guided by no other interest than to arrange its differences as pre sented us, into a fixed scale, even here theoretical elements are requi
and clearly appear. It is a long way from the immediate sensa tion of heat to the exact concept of temperature. The indefinite stronger and weaker of the impression offers no foothold for gaining
site
fixed numerical values.
In order even to establish the schema of
measurement, we
are obliged to pass from the subjective perception to an objective functional correlation between heat and extension. If we give to a certain volume of mercury the value of degrees,
and to another volume
of mercury the value of 100 degrees, then in order to divide the distance between the two points thus signified into further divisions and subdivisions, we must make the assump
tion that the differences of temperature are directly proportional to the volume of the mercury. This assumption is primarily nothing
but an hypothesis suggested by empirical observation, but
is
in
no
If we go from solid absolutely forced upon us by it alone. bodies to fluid, from the mercury thermometer to the water thermom eter, then, for purposes of measurement, the simple formula of
way
proportionality must be replaced by a more complex formula, accord ing to which the correlation between temperature values and volume values is established. 20 In this we see how the
example,
simple
p.
20
C/. the pertinent exposition of
G. Milhaud, Le Rationnel, Paris 1898,
47
ff.
143
quantitative determination of a physical fact draws it into a network of theoretical presuppositions, outside of which the very question as
to the measurability of the process could not be raised.
The physical
cal insight
"fact"
and
the physical
"theory."
has been increasingly
clarified
This epistemologithe philosophical work by

It is
of the physical investigators themselves.
Duhem
above
fact
all,
who has brought

physical theory to
the mutual relation between physical

its
and
clearest expression. He gives a of the contrast between the naive living portrayal sensuous observation, which remains merely in the field of the con
simplest
and
convincing and
crete facts of perception, and the scientifically guided and controlled experiment. Let us follow in thought the course of an experimental
investigation; imagine ourselves, for instance, placed in the laboratory where Regnault carried out his well-known attempt to test the law of Mariotte; then, indeed, we see at first a sum of direct observations,
tions in
which we can simply repeat. But the enumeration of these observa no way constitutes the kernel and essential meaning of
s results.
Regnault
before
of
him
What the physical investigator objectively sees are certain conditions and changes in his instruments
But the judgments he makes
are
measurement.
not related
to these
by
that
instruments, but to the objects, which are measured them. It is not the height of a certain column of mercury
is
reported, but a value of

it is
"temperature"
that
is
estab
lished;
not a change which takes place in the manom eter, but a variation in the pressure, under which the observed gas stands, that is noted. The peculiar and characteristic function
of the scientific concept is found in this transition from what is directly offered in the perception of the individual element, to the form, which
the elements gain finally in the physical statement. The value of the volume which a gas assumes, the value of the pressure it is under, and the degree of its temperature, are none of them concrete objects
properties, such as we could coordinate with colors and tones; but they are "abstract symbols," which merely connect the physical theory again with the actually observed facts. The apparatus, by which the volume of a gas is established, presupposes not only the principles of arithmetic and geometry, but also the abstract
and
principles of general mechanics definition of pressure requires,
and
for
celestial
its
complete
mechanics; the exact understanding,
insight into the deepest
and most
difficult theories of hydrostatics,
144
Between the phenomena actually observed in of electricity, etc. the course of an experiment, and the final result of this experiment as the physicist formulates it, there lies extremely complex intellec
tual labor; and it is through this, that a report regarding a single instance of a process is made over into a judgment concerning a law
This dependence of every practical measurement on certain fundamental assumptions, which are taken as universally valid, appears still more clearly when we consider, that the real out
of nature.
an investigation never comes directly to light, but can only be ascertained through a critical discussion directed upon the exclu In truth, no physicist experi sion of the "error of observation." ments and measures with the particular instrument that he has sensibly before his eyes; but he substitutes for it an ideal instrument in thought, from which all accidental defects such as necessarily belong to the particular instrument, are excluded. For example, if we measure the intensity of an electric current by a tangent-compass, then the observations, which we make first with a concrete apparatus, must be related and carried over to a general geometrical model,
come
of
before they are physically applicable. substitute for a copper wire of a definite strength a strictly geometrical circle without breadth; in place of the steel of the magnetic needle, which has a
We
magnitude and form, we substitute an infinitely small, which can be moved without friction around a vertical axis; and it is the totality of these transformations, which permits us to carry the observed deflection of the magnetic needle into the general theoretical formula of the strength of the The correc current, and thus to determine the value of the latter. tions, which we make and must necessarily make with the use of every physical instrument, are themselves a work of mathematical
certain
horizontal magnetic axis,
theory; to exclude these latter,

its
21 meaning and value.
is
to deprive the observation itself of
Units of measurement. This connection appears from a new angle, realize that every concrete measurement requires the estab lishment of certain units, which it assumes as constant. This
when we
constancy, however,
tible as such,
21
is
but
is
first
never a property that belongs to the percep conferred upon the latter on the basis of
Cf. the excellent exposition of Duhem, in which this connection is ex plained in its particulars and illuminated on all sides. (La Theorie Physique, son objet et sa structure, Paris 1906.)

intellectual postulates
145
and definitions. The necessity of such pos seen especially in the fundamental physical problem of measurement, viz., in the problem of the measurement of time. From
tulates
is
the beginning, the measurement of time must forego all sensuous helps, such as seem to stand at the disposal of the measurement of
space.
We
cannot
and compare them
stretch of time to the place of another, in direct intuition, for precisely the characteristic
move one
element of time is that two parts of it are never given at once. Thus there only remains a conceptual arrangement made possible by re course to the phenomena of movement. For abstract mechanics, those times are said to be equal, in which a material point left to
itself
traverses equal distances.
Here again we
find the concept of
the mass-point, and thus a purely ideal concept of a limit; and once more it is the hypothetical assumption of a universal principle, which first makes possible the unit of measure. The law of inertia
enters as a conceptual element into the explanation of the unit of time. might attempt to eliminate this dependence by going
We
over from rational mechanics to
its
ing to establish a strictly phenomena. The daily revolution of the earth, as it seems, offers the required uniformity in all the perfection that could ever be taken
into account for purposes of measurement.
uniform motion
empirical applications, and seek in the field of the concrete
The
lies directly given us by the interval, culminations of the same star. More exact consideration, however, renders apparent the difference, which always remains between the
which
unit here would be between two successive
and empirical measure of time. The inequality of the stellar days is what is rather demanded now, on the basis of theoretical The friction, considerations, and is confirmed by empirical reasons. which arises from the continuous change of the ebb and flow of the
ideal
tides,
of the earth,
produces a gradual diminution of the velocity of the rotation and thus a lengthening of the stellar day. The desired
exact measure again eludes us, and we are forced to more remote intellectual assumptions. All of these gain their meaning only by
some physical law, which we tacitly assume with them. time, in which the emanation of radium loses its radio has recently been proposed as an exact unit of measurement activity, in this, the law of exponents, in accordance with which the diminution
relation to
Thus the
of effect takes place, serves as a foundation. It is analogous to this, that the principles and theories of optics are presupposed in order to
146
introduce the wave-lengths of certain rays of light as the foundation What guides us in the choice of of the measurement of distance. units is thus always the attempt to establish certain laws as universal.
assume the empirically entirely "equal" stellar days to be unequal, in order to maintain the principle of the conservation of energy.
We
The real constants are thus fundamentally, it has been justly urged, not the material measuring-rods and units of measurement, but these
very laws, to which they are related and according to whose model 22 they are constructed. The verification of physical hypotheses. The naive view, that measurements inhere in physical things and processes like sensuous properties, and only need to be read off from them, is more and more
superseded with the advance of theoretical physics. Nevertheless, the relation of law and fact is thereby altered. For the explanation, that we reach laws by comparing and measuring individual facts, is
now revealed as a logical circle. The law can only arise from measurement, because we have assumed the law in hypothetical form in measurement itself. Paradoxical as this reciprocal relation
may
appear,
it
exactly expresses the central logical problem of
physics.
The
intellectual anticipation of the
law
is
not contradictory,
because such anticipation does not occur in the form of a dogmatic assertion, but merely as an initial intellectual assumption; because
it
does not involve a final answer, but merely a question. The value and correctness of this assumption are first shown when the totality of experiences are connected into an unbroken unity on the
it. The correctness of the assumption, on the other hand, cannot indeed be assured by our verifying every hypothesis and every theoretical construction directly in an individual experience, in a particular sensuous impression. The validity of the physical concept does not rest upon its content of real elements of existence, such as can be directly pointed out, but upon the strictness of con-
basis of
nection,
it
which
(Cf.
it
makes
p.
possible.
In this fundamental character,
constitutes the extension
concept.
above
and continuation of the mathematical 83 f.) Thus the individual concept can
22 Cf. Henri Poincare, La mesure du temps, Revue de Metaphysique et de Morale VI, 1898. Concerning the theoretical presuppositions of the determina tion of units of measurement, cf. especially Lucien Poincar6, La Physique moderne, dtsch. v. Brahn, Lpz. 1908; and also Wilbois, L Esprit positif, Revue de Metaph. IX (1901).
147
never be measured and confirmed by experience for itself alone, but it gains this confirmation always only as a member of a theoretical complex. Its "truth" is primarily revealed in the consequences it leads to; in the connection and systematic completeness of the
Here each element needs explanations, which it makes possible. the other for its support and confirmation; no element can be sepa rated from the total organism and be represented and proved in
this
isolation.
We
do not have physical concepts and physical
facts in pure separation, so that

first
we could
select
member
of the
sphere and enquire whether it possessed a copy in the second; but we possess the only by virtue of the totality of concepts, just as, on the other hand, we conceive the concepts only with refer
"facts"
error of
ence to the totality of possible experience. It is the fundamental Baconian empiricism that it does not grasp this correlation; as isolated entities existing for them that it conceives the
"facta"
which our thought has only to copy as faithfully as possible. Here the function of the concept only extends to the subsequent inclusion and representation of the empirical material; but not to
selves,
the testing and proving of this material. 23 Although this conception has been obstinately maintained within the epistemology of natural
science, nevertheless there are
many
signs that physics itself in its

it.
modern form has
definitely
overcome
Those thinkers,
also,
who
urge strongly that experience in its totality forms the highest and ultimate authority for all physical theory, repudiate the naive
Baconian thought of the "experimentum crucis." experi ence, in the sense of a mere inductive collection of isolated observa tions, can never furnish the fundamental scaffolding of physics; The intellec for it is denied the power of giving mathematical form. work of understanding, which connects the bare fact systemati tual cally with the totality of phenomena, only begins when the fact is 24 represented and replaced by a mathematical symbol. The motive of serial construction. Yet if we conceive the final result of the analysis of physical theory in this fashion, there remains a paradox. Of what value is all the intellectual labor of physics, if we
"Pure"
must ultimately recognize that all the complication of investigation and its methods simply removes us more and more from the concrete
fact of intuition in its sensuous
23 24
immediacy?
Is all this
ff. ff.
expenditure of
Cf.
more particularly Erkenntnisproblem
II, 125
Cf. in particular
Duhem, La Theorie Physique,
p. 308
148
scientific
means worth while, when the final outcome is and can be but the transformation of facts into symbols? The reproach nothing
raised by modern physics, in its inception, against scholasticism, that scholasticism replaced the consideration of facts by that of names, now threatens to fall upon physics itself. Nothing but a new nomenclature seems to be gained, by which we are more and
more alienated from the true
reality of sensation.
In
fact,
this
consequence has occasionally been pointed out: the necessity, to which physical theory tends, has been opposed to the evidence and truth, which comes to consciousness in the experience of individual This separation, however, rests upon a false abstraction; it is facts. based in the attempt to isolate two moments from each other, which
by the very presuppositions of has been shown, in opposition to the traditional logical doctrine, that the course of the mathematical construction of concepts is defined by the procedure of the construction
are inseparably connected together
the formation of concepts.
It
We have not been concerned with separating out the element from a plurality of similar impressions, but with establishing a principle by which their diversity should appear.
of series.
common
The unity
properties,
been found in a fixed group of which represents the mere diversity as a A con sequence of elements according to law. (Cf. above 14 ff .) sideration of the fundamental physical concepts has confirmed and broadened this view. All these concepts appear to be so many means of grasping the in series, and of assigning it a fixed
of the concept has not
in the rule,
but
"given"
Scientific investigation accomplishes this place within these series. last definitely; but in order for it to be possible, the serial principles
themselves, according to which the comparison and arrangement of the elements takes place, must be theoretically established. The
individual thing is nothing for the physicist but a system of physical constants; outside of these constants, he possesses no means or In order possibility of characterizing the particularity of an object.
to distinguish an object from other objects, and to subsume it under a fixed conceptual class, we must ascribe to it a definite volume and a
definite mass, a definite specific gravity, a definite capacity for heat, a definite electricity, etc. The measurements, which are necessary for
however, presuppose that the aspect, with respect to which the comparison is made, has been previously conceived with conceptual This aspect is never given in the original rigor and exactitude.
this,
149
impression, but has to be worked out theoretically, in order to be then applied to the manifold of perception. The physical analysis of
the object into the totality of its numerical constants is thus in no sense the same as the breaking-up of a sensuous thing into the group of
its
must be introduced,
sensuous properties; but new and specific categories of judgment in order to carry out this analysis. In this
judgment, the concrete impression first changes into the physically determinate object. The sensuous quality of a thing becomes a
^physical object,
The
.cal
.-v.
"thing"
when it is transformed into a serial determination. now changes from a sum of properties into a mathematiwhich are established with reference to some
system of values, scale of comparison.

such a
scale,
Each
of the different physical concepts defines
and thereby renders possible an increasingly intimate connection and arrangement of the elements of the given. The chaos of impressions becomes a system of numbers; but these numbers first gain their denomination, and thus their specific meaning, from the system of concepts, which are theoretically established as universal standards of measurement. In this logical connection, we first
see the
"objective"
into the mathematical
value in the transformation of the impression It is true that, in the symbolic "symbol."
designation, the particular property of the sensuous impression is but all that distinguishes it as a member of a system is retained and brought out. The symbol possesses its adequate correlate in the
lost;
connection according to law, that subsists between the individual members, and not in any constitutive part of the perception; yet it is
this connection that gradually reveals itself to
be the real kernel of
the thought of empirical
"reality."
The physical concepts of series. The relation, that is fundamental here, can be illumined from another standpoint by connecting it with the ordinary psychological theory of the concept. In the
language of this theory, the problem of the concept resolves into the problem of "apperceptive connection." The newly appearing im
is first grasped as an individual, and first gains conceptual comprehension through the apperceptive interpretation and arrange
pression
ment that
it
undergoes.
If this reference of
the individual to the

of consciousness"
totality of experience
were lacking, the
"unity
the impression would no longer belong itself would be destroyed, to world of reality. In the sense of this conception, we can characterize the various physical concepts of measurement evolved
"our"
by
scientific
theory as the real and necessary apperceptive concepts
150
for all empirical
knowledge in general. Without them, indeed, it shown there is no arrangement of the factual in series, and has been thus no thorough reciprocal determination among its individual members. We would only possess the fact as an individual subject, without being able to give any predicate for determining it more It is only when we bring the given under some norm of closely. measurement, that it gains fixed shape and form, and assumes clearly Even before its individual value has defined physical "properties."
been empirically established within each of the possible comparative
series, the fact is recognized, that it necessarily belongs to some of these series, and an anticipatory schema is therewith produced for The preliminary deductive work furnishes its closer determination.
a survey of the possible kinds of exact correlation; while experience determines which of the possible types of connection is applicable to the case in hand. Scientific experiment always finds several
ways before it, which theory has prepared, and between which a selection must be made. Thus no content of experience can ever
appear as something absolutely strange; for even in making it a content of our thought, in setting it in spatial and temporal relations
with other contents, we have thereby impressed it with the seal of our universal concepts of connection, in particular those of mathe
The material of perception is not merely subse quently moulded into some conceptual form but the thought of this form constitutes the necessary presupposition of being able to predi
matical relations.
;
cate any character of the matter itself, indeed, of being able to assert any concrete determination and predicates of it. Now it can no longer seem strange, that scientific physics, also, the further it seeks
to penetrate into the "being" of its object only strikes new strata of numbers, as it were. It discovers no absolute metaphysical
qualities;
but
process
it is
seeks to express the properties of the body or of the investigating by taking up into its determination new
it
Such a parameter is the mass, which we ascribe to "parameters." an individual body in order to render rationally intelligible the totality of its possible changes and its relations with respect to exter nal impulses to motion, or the amount of energy, which we regard as
characteristic of the
momentary condition
of a given physical system.
The same holds

26
of all the different magnitudes,
by which physics
of the real world. 25
and chemistry progressively determine the bodies

nis,
Cf. the striking developments of G. F. Lipps, Mythenbildung und ErkenntLpz. 1907, p. 211 ff.
151
enter into this procedure, the more clear becomes the character of the scientific concept of the thing and its difference from the metaphysical concept of substance. Natural science in
The more deeply we
development has everywhere used the form of this latter; yet in its progress it has filled this form with a new content, and raised it to a
its
new
level of confirmation.
The the Ionian philosophy of nature. of substance stands at the pinnacle of the scientific logical conception view of the world in general; it is the concept of substance, which
The concept of substance in
historically
marks the
line of distinction
between investigation and
myth. Philosophy has its own origin in this achievement. The attempt to deduce the multiplicity of sensuous reality from a single ultimate substance contains a universal postulate which, however
imperfectly
istic
it
may
be
fulfilled at first,
nevertheless
is
the character
For of thought and a new question. the first time, being is conceived as an ordered whole, not guided by alien caprice but containing within itself its principle of being.
expression of a
new mode
At first, however, this new conception seeks its confirmation only in the sphere of sensuous things, as they alone seem to constitute the The conceptual and critical work fixed positive content of reality.
of investigation not having yet been conceived or begun, percep tion offered the only fixed limit separating reality from myth
and poetic fantasy. Thus it is some empirically given, particular material, which constitutes the meaning of "substance." But even within the Ionic philosophy of nature, tendencies begin to Anaximander s principle arise that transcend this conception. of the aTTupov already rises in logical freedom above the sphere
ological
of immediately perceptible reality. It contains indication of the that what constitutes the origin of sensuous being cannot thought
have the same properties that it has. It cannot be clothed with any particular material quality, for all particular qualities must develop out of it. It thus becomes a being without definite, sensuous, differentiating properties, in whose homogeneous structure the oppo sitions of the warm and the cold, the wet and the dry still lie together unseparated. The realm of the material in general is not abandoned;
it is
rather precisely the pure abstraction of matter
itself
that
first
gains clear expression in the infinite and indeterminate substance v of Anaximander.
152
The hypostatization of sensuous qualities. (Anaxagoras .) The problem of the particular qualities and properties, however, is not answered in this first attempted solution, but is rather raised for the While the opposites are to evolve by "separating out" first time. from the homogeneous ultimate principle, the manner and occasion of The question their differentiation remains at first wholly obscure. in this gives the motive to the further development of the involved speculative philosophy of nature. The unity, assumed by Anaxi-
mander
merely a logical which lacks exact foundation. In order to reach clarity anticipation, on this point, thought must apparently take the opposite course. The true infinity of the ultimate substance is found not so much in its homogeneous and undifferentiated structure, as rather in the unlimited fullness and multiplicity of qualitative differences con It is the cosmology of Anaxagoras, which draws cealed within it.
in his principle of the infinite, represents
out this tendency to
its
conclusion.
With Anaxagoras, not only
is
universal principle of motion first established, but also the physical explanation of the particular qualities enters on a new phase. It is vain to seek to deduce the particular from the universal, if the former
is
not already contained in the latter in some form.
Thus the mani
apparent properties of bodies, of whose existence and differences the senses inform us, point back to permanent and absolute properties
fold,
The wet and the dry, the bright and the cold, the thick and the thin are all fundamental properties of things; and on the manner and quantita tive proportion in which these properties are mixed with each other, rest all the differences and contrasts between the composite sensuous
of
matter as their real origin.
the dark, the
warm and
substances, such as air and water, ether and earth. All the elemen tary properties enter in their totality into every composition, and are
to be thought of as always contained in even the smallest material What gives the particular part, however far we pursue the analysis. substances their distinguishing form is not that they contain any one
of the qualitative elements in isolation, but that one of them pre dominates in the composition, so that in ordinary popular considera tion the remaining factors can be practically left out of account,
is in all;" although they are really never lacking. In this sense, every particle, however small, even every physical point, represents a
"all
The totality of infinitely many qualities, which interpenetrate in it. detailed working-out of this theory is merely of historical interest;
153
but even aside from this, it contains an element of typical meaning, which has frequently come to light in the advance of physics. The analysis of Anaxagoras aims to go back of the concrete sensuous
object, as it is first offered in intuition, to its conceptual principles; but it defines the content of these principles in expressions, which
are throughout taken from sensuous perception. Properties and of sensation are immediately transformed into substantial oppositions
causes,
which
exist in themselves
and can operate
for themselves,
although only in connection with other causes of the same sort. The varied multiplicity of sensible qualities is thus retained; indeed,
it is
which seem
To each of these properties, consciously increased to infinity. in appearance to go hither and thither, to appear and
disappear, there corresponds in truth an unchanging and substantial For any sensuous property to arise newly in a subject or to being.
ficial
disappear is a mere deception, with which we are misled by a super consideration of things; on the contrary, each of these properties
and is only temporarily, as it were, concealed from view by which seem to take its place. Here we have, in fact, the others, characteristic attempt to construe the permanent being, which thought demands, without going beyond the sphere of the "given." It is no longer, as in the Ionic philosophy of nature, a particular
persists
empirical material like air or water, that represents the permanent existence of things but this function is transferred to the totality of properties, from which each body results, and which can be discovered
;
by perception. The hypostatization of these properties leaves unchanged; it is true they gain thereby a changed metaphysical significance, but in principle they do not go beyond the
in
it
their nature
character of the sensuous.
The hypostatization
of sensuous qualities.
(Aristotle.)
Aristotelian
physics also represents no intrinsic change in this respect. The fundamental qualities are here reduced again to a small number;
instead of the infinitely numerous "germs" of things, there are merely the properties of warm and cold, wet and dry, from the combination
of
which are to
fire.
The nature
motions
cosmos.
water and earth, air and of these elements determines the peculiarity of the they undergo, and thereby the total plan and order of the
arise the four elements:
Thus the structure
of this physics also rests
upon the same
procedure of converting the relative properties of sensations into absolute properties of things. The view at the basis of this comes out with especial force and distinctness in its historical consequences.
154

of natural science, especially the
The whole
whole of chemistry and
ages, is first intelligible, when considered in connection with the logical presuppositions of the Aristotelian system.
alchemy of the middle
The elevation of qualities into separate essences, which are different from the being of the body, and therewith at least in principle trans The ferable from one body to another, is here the dominant view. properties, which are common to a class of things, and which thus
furnish the foundation for the construction of a certain generic concept, are separated off as physical parts and raised to independent
solid bodies are distinguished from the fluid and by the presence of a certain absolute and separable property, that is immanent in them; transition to another state of aggregation means the loss of this quality and the assumption of a new substantial nature. Thus mercury, for example, can be changed into gold, if we successively remove from it the two "elements," on which its fluidity and volatility rest, and substitute other proper
existence.
The
volatile bodies
ties for
them.
a position to imprint them successively on matter. of metals into each other is conceived and repre sented according to this fundamental view. We take from the partic ular body its particular properties, which are conceived as so many
is in
general way that one
it is sufficient
In order to transform any body into another, in to master the different "natures," in such a
The transmutation
independent substances in
creaking,
silver,
its fusibility
it;
far example,
and
its softness, in
we separate from tin its order to make it approach
from which it is at first separated by all these properties. The whole view, on which this conception of nature rests, appears Bacon s clearly even in modern times in the physics of Bacon. theory of forms goes back to the axiom, that what constitutes the generic common element of a group of bodies must be somehow present in them as a separable part. The form of heat exists as a peculiar somewhat, that is present in all warm things, and by its presence calls forth certain effects in them. The task of physics is
exhausted in reducing the complex sensuous thing to a bundle of abstract and simple qualities, and explaining it from them. The hypothesis of a heat substance, like the assumption of a special
electric or
magnetic
fluid,
shows how slowly

science. 28
replaced also in
28
modern
this conception has been Especially in the concepts of
1908, p. 300
Cf. the excellent presentation by E. Meyerson, Identite el Rfalite, Paris, ff.; also Berthelot, Les origines se I Alchimie, Paris, 1885, p. 206 ff.,
279
ff., etc.

chemistry
it
155
chemistry Thus sulphur is the expression of the combustibility of bodies; salt, the expression of their solubility; while mercury comprehends and It is always certain expresses the totality of metallic properties.
reactions according to law, for which a substantial substrate is directly assumed. The property of combustibility, which we seem to perceive
is
at once a bearer
reappears in diverse forms. Every element of the older and a type of a certain striking property.
transformed by the assumption mixed with bodies and from this assumption, the whole structure of chemistry before Lavoisier follows with inner necessity.
of bodies, is
sensuously in a
number
of phlogiston into a particular substance, that is
Atomism and number.

traced here in
its
Along with the development we have

first
general features, there exists from the
another
fundamental view of physical being and process. Ancient science gave a perfect expression to this view in the system of atomism. Atomism, through the mediation of the Eleatic system, goes back in its historical presuppositions to the fundamental concept of the Pythagorean doctrine. The concept of empty space, from which Democritus starts, is directly taken from the KCVOV of the Pythago reans. Here we face a change in the direction of thought. Being is no longer sought directly in the sensuous perceptible qualities, nor in what corresponds to them as an absolute correlate and counterpart; but being is resolved into the pure concept of number. On number, rests all the connection and inner harmony of things; precisely for that reason number is to be characterized as the substance of things; for it alone gives them a definite, knowable character. The mystical exaggeration, with which this thought was first grasped, gradually ceases with the advance of Greek science, till it finally yields to a purely methodological and rational explanation. This change is completed in the atomistic system; what was an abstract postulate for the Pythagorean, is here embodied in a concrete construction of mechanics. The sensuous properties of things are banished from
the scientific
universe; it is only according to to the untested "subjective" view, that "opinion," only according there is a sweet and a bitter, a colored and a colorless, a warm and a
picture
of
the
For the representation of objective reality, on the contrary, these properties are to be cast aside, for no one of them is capable of exact quantitative determination, and thus of a truly unambigu
cold.
all
ous definition.
Thus only those
properties of things, that are deter-
156

"real"
minable in the sense of pure mathematics, remain as the The abstract number-schema of the Pythagoreans is properties. now, however, supplemented by a new element, which enables it to develop to its full effect. In order to advance from number to mate rial physical existence, it is necessary to have the mediation of the
Space, however, is here taken in a sense that makes the pure image of number. It represents all the properties, and Its characteristic feature, fulfills all the conditions of number.
concept of space.
it
accordingly, is the absolute uniformity of its parts; all inner differ ences are resolved into mere differences of position. The differences existing in the immediate space of perception are wholly removed, so that each particular point represents merely an equivalent startingpoint for geometrical relations and constructions. Now if the real
determined from this point of view, nothing remains of it but what makes it a numerical order, a quantitative whole. Precisely in this is rooted the meaning and justification of the concept of the
is
atom.
The world
of
atoms
is
nothing but the abstract representation
of physical reality, in so far as nothing is retained in this but pure determinations of magnitude. It was in this sense that Galileo,
at the threshold of
modern
physics, understood
and founded atomism.
He explains,
save
its
that in the concept of matter nothing else is fundamental conception as of this or that form, as in this or that place,
as large or small, as in motion or rest. From all other properties, on the other hand, we can abstract without thereby destroying the
thought of matter itself. No logical necessity forces us to conceive matter as either white or red, sweet or bitter, pleasant or evil-smell ing rather all these characterizations are mere names to which (since they cannot be reduced to exact numerical values) there corresponds no fixed, objective correlate. The substance of the physical body
:
is
try and the pure theory of motion (which goes back to can discover and establish in it.
exhausted in the totality of properties, which arithmetic and geome them both)
The impact of atoms. However, with this acceptance of atomism, the problem is only raised in general terms, but is in no way solved completely. For the atom signifies no fixed physical fact, but a
logical postulate; it is thus itself not unchanging, but rather a vari able expression. It is interesting to trace how, in the transformations which the concept of the atom undergoes in the course of time, the intellectual motive, to which it owes its origin, continually works
157
out with greater clarity. In the atom of Democritus, the dissolution If we take of sensuous determinations is not fully carried through. the well-known report of Aristotle as our authority, the atoms are
distinguished, according to Democritus, not only in their position in place, but also in their magnitude and form; they possess different extension and different form, although no reason for this difference
shown. Yet to the extent that the dynamical interaction of the atoms becomes a real problem, the logical necessity appears above all of endowing each atom with an absolute hardness, by means of which it excludes all others from its position in space. The opposition of the hard and the soft, like that of the light and the heavy, is thus
is
again taken up directly into the objective consideration of nature; a residuum of the sensible properties of body is retained and put on
a level with the determinations discerned by mathematical thought. The consequences of this dualism soon become clear in the develop ment of the doctrine. They concentrate into a real antinomy, when
we
being
consider the relation that results between the physical concept of and the physical law of process. This law, if we merely consider
its applications to mechanics, demands that the sum of energy remain unchanged in every transference of motion from one body to another. Yet if we attempt to apply this point of view to the impact of atoms, a peculiar difficulty results. If we consider the atoms as perfectly hard bodies, then their properties and modes of action are to be deter mined according to the relations we can directly observe empirically
every collision of wholly or par appears a certain loss of energy. In order to remove this contradiction of the law of the conservation of energy, the theory must assume that a portion of the energy has
tially inelastic bodies, there
in inelastic masses; however, with
gone from the masses to their parts, that "molar" energy has been changed into "molecular." But this explanation is obviously not available for the atoms themselves, since these, according to their conception, are simple subjects of motion, and lack all possibility of being further analysed into parts and subparts. The postulate of continuity, and the "simple" atom of Boscovich and Fechner, Kinetic atomism has sought in various ways to remove this contradiction in its foundations, without ever having fully succeeded. 27 And a second problem no less difficult results, when
17
lose of energy, that
For criticism of the attempted solution of Secchi, according to which the must occur in the impact of absolutely hard bodies, is
158
the mechanics of the atoms with the demands which from the postulate of the continuity of process. The change of velocity, which two absolutely hard bodies undergo in the moment of their impact, can only consist in a sudden transition, in a jump from one magnitude to another, that is separated from it by a fixed, If a slowly moving body is overtaken by a more finite amount. rapidly moving one, and they both advance after the impact with a common velocity, which is determined by the law of the conservation of the algebraic sum of their motions, then this result can only be expressed by our ascribing to the one body an abrupt loss, to the other an abrupt increase in velocity. This assumption, however, leads to the discovery, that we can establish no definite value for the velocity of the two masses at the moment of impact, and that a gap remains in the mathematical determination of the whole process. 28 The advocates of extended atoms have occasionally replied to objec
result
we compare
tions of this sort, that a false standard is applied here to the thetical structure on which mechanics should rest. The contradic
tion merely comes from the fact, that the atoms, which are meant to be nothing but rational constructions of thought, have certain properties ascribed to them, which properties are only deduced from analogy with the sensuous bodies of our world of perception. This
very analogy, however, is to be abandoned from the standpoint of the theory of knowledge. The norm for the formation of the con tent of the atom, is not the behaviour of the empirical bodies of
our environment, but the universal laws and principles of mechanics. Thus we are not directed for this norm to a mere vague comparison with directly observable phenomena, but we determine the conditions
of motion, on the basis of by the "subject" conceptual postulates. Thus we need not ask whether it is possible or impossible for absolutely rigid bodies in impact to satisfy the law of the conservation of energy, but conversely we assume the
"real"
to be satisfied
validity of this law as an axiom, to which cal construction of the atoms and their
we are bound
movements.
in the theoreti
The compati-
counterbalanced by a part of the rotary movement of the atoms being trans formed into progressive movement, cf. Stallo, The Concepts and Theories of Modern Physics, German edition, Lpz. 1901, p. 34 ff.; For general criticism of the concept of the atom, cf. especially Otto Buek, Die Atomistik und Faraday s
Begriff der Materie, Berlin, 1905. Cf. more particularly, Erkenntnisproblem, II, 394
"
ff.
159
bility of this construction with the other assumptions of rational mechanics is the rule that alone can guide us, and not the similarity of atomic motions with any processes of known physical reality. 29 This reply is, in principle, thoroughly satisfactory; yet when we think it through, we find ourselves forced also from the logical side to that transformation of the concept of the atom, which natural In place of the extended science since Boscovich has carried out. there now appears the absolutely simple but indivisible particle,
point of force.
We
see
how
the reduction of the sensible properties,
which was already characteristic of Democritus, has here advanced another step. The magnitude and form of the atoms have now disappeared; what differentiates them is merely the position, that they mutually determine for each other in the system of dynamic
actions
and
reactions.
The negation
of extension
is
joined to the
negation of the sensuous qualities; therewith there is the negation of every determination of content in general, by which one empiri
All independent, selfcal "thing" is distinguished from another. existent attributes are now completely effaced; what remains is merely the relation of a dynamic coexistence in the law of the recip
and repulsion of the points of force. Boscovich urges energetically, and Fechner after him, that force itself, as it is here understood, resolves into the concept of law and that it is meant to be merely the expression of a functional dependence of magnitudes. The atom, which in its origin goes back to the pure
rocal attraction
its origin after manifold trans but the member of a systematic formations; signifies nothing manifold in general. All content, that can be ascribed to it, springs from the relations of which it is the intellectual center. The concept of the atom and the application of differential equations. The scientific development, which the concept of the atom has undergone in the most recent modern physics, confirms this view throughout. In the conflict between atomism and energism, Boltzmann has sought to deduce the necessity of the atomic hypothesis from the fundamental method of theoretical natural science, from the procedure of applying differential equations. If we do not deceive ourselves concerning the meaning of a differential equation, he explains, we cannot doubt that the scheme of the world, that is assumed with it, is in essence and structure atomistic. "On closer
concept of number, here reverts to

it
29
See Lasswitz, Geschichte der Atomistik,
II,
380
ff.
160
inspection, a differential equation is only the expression of the fact, that we have first to conceive a finite number; this is the first condi
tion,
then this number must increase until
its
further increase
is
no
longer of influence.
What
conceiving a great number the differential equation the value expressed by the same has been Thus he, who believes he can free himself defined by that demand?" from atomism by differential equations, does not see the wood for the
to suppress the demand for of individuals, when in the explanation of
use
is
it
This type of explanation is of high interest from the stand point of the critique of knowledge; for the necessity of the atom is here meant to be deduced not from the facts of the empirical con
trees. 30
sideration of nature, but from the conditions of the method of exact If, however, this is the case, then it is clear indeed that physics.
the
"existence"
assured to the atom in this
way can be no
other than
belongs in general to the pure mathematical concepts. Thus Boltzmann expressly guards against the assumption that the absolute
existence of the
for the exact representation Precisely in this assumption, however, the necessity finally appears of going from the extensive corpuscle to the simple mass-point, if the "image" is to gain its full sharpness and exactitude.
of
to be understood
atoms is to be proved by and applied as images
his deduction; they are only
phenomena.
31
The procedure
itself
of the infinitesimal calculus, to

If
which Boltzmann
we start with the presentation appeals, urges this transition. of certain finite magnitudes, and then permit the latter to diminish continuously in order to gain a point of application for the differen
tial
if
equation, this process only reaches
its
we permit the magnitudes under consideration
mathematical conclusion to converge toward
the limiting value null; while, according to atomism, a constant value could always be given, beyond which the ideal procedure could not go without involving itself in contradictions with the
As long as we remain with magnitudes of a we reach no definite logical determination, no matter how small we may choose these magnitudes. Along with
reality of
phenomena.
certain extension,
whatever physical
divisibility
one
may
.
assume, there always remains
30 Boltzmann, Uber die Unentbehrlichkeit der Atomislik, Annalen der Physik und Chemie N. F. Vol. 60, p. 231 ff (Populare Schriften, Lpz. 1905,
p. 141
31
ff.)
Cf. Boltzmann, p. 129 ff.)
Bin Wort der Mathematik an
die Energetik (Pop. Schriften
161
the intellectual possibility of analysing the body further, and of ascribing different velocities to the many, distinguishable sub groups. It is only when we advance to the material point, that this
indefiniteness
duced.
is done away with, and a fixed subject of motion pro above p. 119 ff). The changes in the concept of the atom. It has been urged from the
(Cf.
Boltzmann, that the concept of the material assumed by mechanics, does not develop out of that of body by our abstracting from extension practically or entirely, but we have to consider by our abstracting from rotary movement. we analyse the body into other than purely forward movements, have absolutely nothing to do with atoms, parts, which into elements of volume, by which we can approach the material 32 point that only moves forward, to any degree of approximation." Here an important logical moment is indicated the simplicity of the The point is assumed for the sake of simplicity of movement.
side of energism against
point, as
it is
"If
....
assumption of the simple, not further reducible body is only a methodological device to advance to the abstraction of simple move ment. In this sense the "atom," according to its fundamental physical meaning, is not defined and postulated as a part of matter, but as a subject of possible changes. It is only considered as an
intellectual point of application for possible relations.
We
analyse
into elementary processes, for which latter we then introduce the atoms as hypothetical substrata. Thus we are
complex movements
not primarily concerned with separating out the ultimate elements of things, but with the establishment of certain simple, fundamental
from which the variety of processes can be deduced. Thus we understand how the atom, in its modern physical application, loses the aspect of materiality more and more how it is resolved into vortex movements in the ether, which, however, in accordance with
processes,
;
their character
indivisibility.
is
fulfill
The
the conditions of indestructibility and physical postulate of identity, which is of course inevitable,
here satisfied not by any kind of material substratum, but by permanent forms of motion. It appears that, in general, as soon as any sort of physical process, which hitherto passed as simple, is regarded from a new point of view, by which it appears as the result of a plurality of conditions, the substratum, which we took as its
82
Cf.
Helm, Die Energetik und ihre
geschichtl. Entwicklung, Leipzig, 1898,
p. 215.
162
basis, at
once divides. Thus when inertia no longer appears as an absolute property of bodies, but a way is offered to deduce it from the laws of electrodynamics, the hitherto material atom breaks up and But the new unity gained in resolves into a system of electrons.
this in
manner can
itself
only be conceived as relative, and as thus
Sharper analysis of physical relations principle changeable. leads constantly to new determinations and differentiations within
their subject.
the
Thus we can say that the content of the concept of atom may be regarded as changeable, while the function, which belongs to it, of defining the condition of knowledge at any given time and of bringing it to its most pregnant intellectual expression,
Only the point of application changes; but the procedure of postulating a unity remains constant. The "simplicity" of the atoms is fundamentally a purely logical predicate; it is deter mined with reference to the intellectual analysis of the phenomena of nature, and not by reference to our sensuous capacity of discrimina
remains unchanged.
tion or with regard to technical-physical
means
of analysis.
Every
advance of analysis, every process of bringing a great field into a new connection, (which is especially possible in modern physics on the basis of phenomena of radio-activity) changes our view of the "constitution" of matter and of the elements out of which it is The new unity, which we define, is always only the expres built up.
sion of the relatively highest
and most comprehensive standpoint of which we grasp the totality of physical things and judgment, by
The concept of
the ether.
processes in general.
An
when we proceed from the concept
analogous development is presented, of matter to the second principal
concept of natural science, the concept of the ether. The difficulties, that arise here at first, spring from the fact that, in order to give this concept a definite content we have to utilize certain properties
originally
tion.
won through comparison with the objects of sense percep Ether accordingly appears as a perfect fluid, which, however,
At first no two aspects;
fully unified picture results
possesses certain properties of perfectly elastic bodies. from the combination of these
the limiting case
itself
presents a different appearance, according as
we approach
to reach
it
from one or the other direction, according as we seek through progressive idealization from different empirical
it
starting-points.
in
principle only
The conflict, that develops here, finds its solution when we resolve to abandon all direct sensuous

representation of the ether, and to use 33 symbol of certain physical relations.
point of space, while
it
163
merely as a conceptual discover a physical as for example, a certain effect of light at a certain phenomenon,
We
we
are obliged to locate
its
"cause"
at a point of
space remote from it. In order to establish a continuous connection between these two conditions, we postulate a medium for them by conceiving the space between them continuously filled with certain The qualities, which can be expressed by pure numerical values. totality of such numerical determinations is really the fundamental conception which we express in the thought of the ether. The unitary and strictly homogeneous space is progressively differentiated through our inscribing upon it, as it were, a web of numbers. This gradation of the individual positions and their arrangement into different mathematical-physical series is what gives them a new content. space, which only represents one principle of "Empty" arrangement, is now in a certain sense covered over with a wealth of other determinations; these, however, are all held together by the fact that certain functional dependencies subsist between them. of the ether can, in fact, be All that physics teaches of the ultimately reduced to judgments about such connections. When,
"being"
according to the electro-magnetic theory of light, the identity is affirmed of the ether of light with that ether in which electro-magnetic
effects are propagated, this is
because the equations, to which we are led in the investigation of light vibrations, are identical in their form with those, which result for dielectric polarization, and because
further the numerical constants, especially the constants for the 34 The assumption of the velocity of propagation, mutually agree.
same substratum
is
here also only another expression of the thorough
going analogy of the mathematical relations: for the connections, that subsist between the values of the optical and electrical constants.
inclusively and consciously physics makes use of the con of the ether, the more clearly it appears that the object thus cept signified cannot be understood as an isolated, individual thing of
The more
perception, but only as a unification

valid, measurable relations.
33 34
and concentration
of objectively
Cf., e.g.,
Pearson, The Grammar of Science, p. 178 ff., 262 ff. See Henri Poincare Elektrizitat und Optik, German trans., Berlin, 1891,
1
p. 159
ff.
164

logical
The
form
of the concept of the physical object.
Now
if
we
again survey the changes, which the scientific concept of substance has undergone from its first speculative beginnings, the unitary
tendencies stands out clearly. True, it must appear as a genuine impoverishment of reality that all existential qualities of the object are gradually stripped off; that the object loses not only its color, its taste, its smell, but gradually also its form and extension
goal of
its
and shrinks to a mere
35 of wax," which Descartes The "point." took as the basis of his well-known analysis of the concept of the object, changed from a fixed, warm, bright, odorous thing into a
"bit
mere geometrical
figure of certain outlines
and dimensions.
And
even with this reduction, the intellectual process did not stop; it did not come to rest until extension itself had been dissolved into the mere appearance of simple and indivisible centres of force. This progres
sive transformation
must appear
unintelligible, if
we
place the goal

of outer
of natural science in gaining the

reality.
most perfect possible copy
Every new theoretical conception, that science introduces, would then separate science further from its real task; owing to this peculiar method, empirical existence, which should be retained and verified unfalsified, would on the contrary threaten to vanish away. Here, in fact, no reconciliation is possible; the exactitude and perfect rational intelligibility of scientific connections are only purchased with a loss of immediate thing-like reality. This recipro cal relation between reality and the concepts of science, however,
furnishes the real solution of the problem. It is only owing to the fact that science abandons the attempt to give a direct, sensuous copy of reality, that science is able to represent this reality as a
necessary
connection
of
grounds and
"electron,"
consequents.
It
is
only
thus of the fundamental element p. 249: "Thus the elec tron must be regarded as a simple electric charge without matter. Our first investigations caused us to ascribe to it a mass one thousand times less than that of an atom of hydrogen; a more careful study now shows us that his mass was only a fiction the electro-magnetic phenomena, which occur when we set the electron in motion, or permit its velocity to vary, resemble inertia to a certain extent, and this inertia depending on its charge, led us astray. The electron is thus simply a definite small volume at a point of the ether, which possesses special properties, and this point is propagated with a velocity, which cannot exceed the velocity of light." Cf. also E. Meyerson, Identite el Realite,
Cf., e.g. the description of the
of
"matter,"
by Lucien Poincar6, La Physique moderne,
p. 228
ff

through going beyond the circle of the given, that science creates the intellectual means of representing the given according to laws. For the elements, at the basis of the order of perceptions according to
If the law, are never found as constituent parts in the perceptions. of natural science consisted simply in reproducing the significance
reality that is given in concrete sensations, then it would indeed be a r vain and useless work; for what copy, how ever perfect, could equal
the original in exactness and certainty? Knowledge has no need for such a duplication, which would still leave the logical form of the
perceptions unchanged. Instead of imagining behind the world of perceptions a new existence built up out of the materials of sensation,
it traces the universal intellectual schemata, in which the relations and connections of perceptions can be perfectly represented. Atom and ether, mass and force are nothing but examples of such schemata, and fulfill their purpose so much the better, the less they contain of
direct perceptual content. "Real" and "not real" elements in the concepts of the physical
object.it
Thus we have two separate
fields,
two
different dimensions as
y^were, of the concept; opposed to the concepts that represent an existence, are the concepts that merely express a possible form of connection. Yet there is no metaphysical dualism here; for although
is no direct similarity between the members of the two fields, they reciprocally refer to each other. The concepts of order of mathematical physics have no other meaning and function than to
there
serve as a complete intellectual survey of the relations of empirical If this connection with empirical being is destroyed, a being.
double antinomy arises. Behind the world of our experience arises a realm of absolute substances, which are themselves a kind of thing, yet which remain inaccessible to all the intellectual means for grasping the things of experience. The "genuinely real" of physics, the
system of atoms and forces acting at a distance, remains in principle There presses upon us the inevitable idea of some unintelligible.
thing existing
since
eternally
beyond presentation, something which,
we cannot reach into this "extra-phenomenal beyond," we can never attain. Thus the world of immediate experience pales into a shadow; while, on the other hand, that for which we exchange it
remains before us as an eternally incomprehensible riddle. "The manifold forms of the absolute are not windows in our system of presentations, that afford a view into the extra-phenomenal world;
166
they only show how impenetrable are the walls of our intra-phenomenal prison." Physics itself, in its continuous and necessary progress, leads to a field permanently closed to research: to a "terra nunc et in aeternum incognita." 35 On the other side of the question, it is unin
telligible
how, with our physical concepts formed through transcend

of presentations,"
we could ever return to exactly could hope to master it on the basis of concepts produced in conscious contradiction to its real content. All these doubts are resolved, however, as soon as we consider the physical
ing the
"system
this system, or
how we
concepts no longer for themselves but, as it were, in their natural In genealogy, in connection with the mathematical concepts.
fact, the physical
begun
clarity.
in
concepts only carry forward the process that is the mathematical concepts, and which here gains full The meaning of the mathematical concept cannot be
comprehended, as long as
for
it
weêk any
sort of presentational correlate
in the given; the
meaning only appeals when we recognize the
oncept as the expression of\a pure relation,.upon which rests the unity nd continuous connection of~~the members of a manifold. The
unction of the physical concept also is first evident in this interpreThe more it disclaims every independent perceptible con ation.
tent and everything pictorial, the more clearly matic function is shown. (Cf. above p. 147 ff .)
of the popular
;
its logical
and syste
"thing"
All that the
view of the world loses in properties, it gains in rela tions for it no longer remains isolated and dependent on itself alone, but is connected inseparably by logical threads with the totality of Each particular concept is, as it were, one of these experience. threads, on which we string real experiences and connect them with
future possible experiences. The objects of physics: matter and force, atom and ether can no longer be misunderstood as so many new realities for investigation, and realities whose inner essence is to
be penetrated, when once they are recognized as instruments pro duced by thought for the purpose of comprehending the confusion of
phenomena as an ordered and measurable whole. Thus there is only one reality, which is given to us, but it comes to conscious ness in different ways thus at one time we consider it in its sensuously
;
intuitive character
but
in its
sensuous isolation, while from the

retain those elements in
it,
standpoint of science
at the basis of
39
we merely
which are
its intellectual
connection and
"harmony."
Cf. P. du Bois-Reymond, Uber die Grundlagen der Erkenntnis in den exaklen Wissenschaften, p. 112 ff. (Cf. above p. 122 ff.)

The concept
167
how
In the history of physics, we can see of non-being. this characteristic interpenetration of the sensuous by the intel
lectual has
insight
first
by the great empirical investigators created the universal schema of the
been expressed with growing consciousness as a logical also. Democritus, who

scientific
conception of
nature, also grasped the philosophical problem latent in it. Motion requires the void for its representation; empty space itself, however, is no sensuous "given," no thing-like reality. Thus it is impossible to
relate scientific thought merely to being, as the Eleatic idealism had attempted; non-being is just as necessary and unavoidable a concept. Without this concept, intellectual mastery of empirical reality is not The Eleatics in their denial of non-being not only to be attained. robbed thought of one of its fundamental instruments, but they
destroyed the phenomena themselves by giving up the possibility of
understanding them in their multiplicity and mutability. The thought of non-being is thus no dialectical construction; but, on the contrary, it is taken as the sole means of protecting physics from the extravagances of a speculative idealism. Even when the facts them selves are regarded as the supreme standard for all intellectual con ceptions, and when no other purpose of the concept is recognized than
that of rendering the fact of motion, and thus of nature, intelligible, even so it has to be granted that in this fact a moment is involved, which is absent from direct intuition. Empty space is necessary for the phenomenon, even though it does not have the same sensuous form of existence as the concrete, particular phenomenon. In the conception of the real, this sensuous "nothing" has the same place and
same
TO
inviolable
37
.
validity as the
"something:"
/xi)
fj,d\\ov TO 8ev
f)
The being, that belongs to the scientific principle in /jTjdei contrast to any concrete thing, gains clear definition here historically for the first time. 38 The physical concept is defined through a
37
Cf. Aristoleles de generatione

e
et
corruptione
A
.
Apxaiuv e5oe TO ov
Kivridijvai, 5 oiiK
AydyKTjs
p.rj
elvai nod, &K.LVTJTOV.

Kexcopur/uej ou
8,325 a: &/ICHS y&p TUV TO ntv y&v Kfvov OVK ov,

. .
av ovvacrdai
cWos Kfvov
K nev ovv
TOVTWV TUV
Xoyuv
virepf3a.VTs
T^V
ai<rdr)<riv
Kai &Klvr)Tov TO irav elval

ovTf
(petal
Kal irapioovTts avrrjv cos Xoycp okov &KO\ovdelv Iv Kal &Treipov Ivioi. , . AevKnriros 5 k\eiv cbi^tfy Xo-yous
T<$
OLTIVCS irpos TT]v alo 9i)ou> 6/j.o\oyovnfvct \kyovTts OVK

tptiopav
Kivr](nt>
&i>aiprj<Tovcri.v
OVTC yevtaiv
TO.VTO.
oijTe
Kai TO
irXJ7!?os
TUV
OVTUV.
cos
bpoXoyrjcFas
Klvrjcriv
6i
n^v TOLS
tpaivo^ tvois, Tols 5

Kevov
38
/J.TI
TO \v KaTavntvaôvaiv
ipijcriv
OVK av
ovaav &vev Kevov, TO Te
bv Kal TOV &VTOS ovdtv ny ov
elvai.
TO yap Kvpiws ov iranirXripfs ov.
see
For the historical and systematic significance of the concept of the ov, Cohen, Platans Ideenlehre und die Mathematik, Marburg 1879; Logik der
etc.
reinen Erkenntnis, p. 70
168
double opposition: an opposition, on the one hand, to metaphysical sensuous perception. speculation, and, on the other, to unmethodical Geometrical space serves here as an example and type of a pure Since it connects the atoms in a unity and relational concept. renders possible motion and interaction among them, it can serve in
which the connection of general as a symbol of those principles, on the real and given rests, without these principles themselves forming part of the intuitive reality. The senses are entangled in the "con
ventional"
and subjective opposition of the warm and the cold, the sweet and the bitter; they do not exhaust the whole of objectivity. For this whole is completed only in the mathematical functional dependencies, which are inaccessible to the senses, since these are
limited to the individual.
Matter and idea and Galiko
concept of inertia.
The physics
of
modern times has retained these principles unchanged; thus Galileo, as an experimental investigator, directly joins on to Archimedes, Like while, in his philosophical view, he goes back to Democritus.
Democritus, he describes and completes the conception of nature by the conception of necessity; only "the true and necessary things, which could not be otherwise," belong in the sphere of scientific
investigation.
distinct
The concept of truth, however, remains for him also from the concept of reality. Just as the propositions of Archimedes concerning the spiral remain true, even in case there is no body in nature with a spiral movement, so in the same way in found ing dynamics we can proceed from the presupposition of a uniformly accelerated movement towards a definite point, and conceptually deduce all the resulting consequences. If empirical observation
agrees with these consequences, so that in the movement of a body with weight the same relations are found, which the theory evolved from hypothetical assumptions, then without danger of error we can
regard the conditions, at first established purely intellectually, as realized in nature. But even if the latter were not the case, our
propositions would lose nothing of their validity, since in and for themselves they contain no assertions about existence, but only connect certain ideal consequences to certain ideal premises. This
general thought gains significant application in Galileo s exposition and defence of his supreme dynamic principle. For him the law of
inertia has
and even
if
throughout the character of a mathematical principle, its consequences are applicable to relations of outer

reality,
169
way signify a direct copy of any conditions, of which it speaks, are actually never realized; they are only gained by means of the "resolutive method." Thus, at one point in the "Dialogue concerning the
empirically given fact.
they nevertheless in no
The
two Systems of the World," when Simplicio is ready to grant the unlimited persistence of the motion of a body left to itself on a horizontal plane, in so far as the body itself is of sufficiently lasting
material, Salviati-Galilei points out to him that this assumption is without significance for the real meaning of the principle of inertia; the material constitution of the particular body is merely an acciden
tal and external circumstance, which is in no way used in the deduction and proof of the principle. Inertia is for Galileo, as empty space was for Democritus, a postulate that we cannot do without in the scientific exposition of phenomena, but which is not itself a concrete sensible process of external reality. It denotes an idea, conceived for the purpose of ordering the phenomena, yet not standing on the same plane methodologically with these phenomena. Hence this motion needs no real, but only a conceived substratum; the real
subjects for the exact expression of the principle are the "material points" of mechanics, not the empirical bodies of our world of per
We see how modern science has here retained the funda mental thought of Democritus only in order, in a certain sense, to pass beyond it; for what was there developed for the concept of the
ception.
void,
ov.
here carried over to the concept of matter, to the Tra^TrX^pco-s Matter also in the sense of pure physics is no object of percep
is
but rather of construction. The fixed outlines and geometrical we have to give it are only possible because we go beyond the field of sensations to their ideal limits. The matter, with which exact science is alone concerned, never exists then as a "perception,"
tion,
definiteness
but always only as a "conception." "When we consider space as says a modern objective and matter as that which occupies of strictly "empiristic" tendency, "we are forming a con physicist struct largely based on geometrical symbols. We are projecting the form and volume of conception into perception, and so accustomed have we got to this conceptual element in the construct, that we confuse it with a reality of perception itself. It is the conceptual volume or form which occupies space, and it is this form, and not the
it,"
sense-impressions, which
39
we conceive
to
.
move."
39
Thus the con-
Pearson, The
Grammar
of Science, p. 250 f
170
cept of matter follows the same law that is universally characteristic The sensuous of the logical development of scientific principles. properties no longer constitute any essential part of its meaning.
Even the element
of
"weight,"
which at
first
seems an inseparable
part, is subordinated and excluded in the transition from the concept From mass, we attain to a of matter to the pure concept of mass.
mere mass-point, which

value,
by a
is distinguished by a certain numerical certain coefficient. Matter itself becomes idea, being
increasingly limited to ideal conceptions, that are produced
and
confirmed by mathematics.
vr
concepts of space
and
time.
The
structure of pure mechanics
logically be derived in different ways, according to the kind and number of fundamental concepts from which one starts. While
the classical mechanics, that reaches a first conclusion in Newton s is built upon the concepts of space and time, mass and force, in modern expositions the concept of energy appears in place
Principia,
The Prinzipien der Mechanik of Heinrich Hertz developed a new view; they rest merely on the establish ment of three independent concepts: space, time and mass, and from these it is sought to deduce the totality of the phenomena of motion as an intelligible whole governed by law, through introducing invisible masses along with the sensuously perceptible masses.
of this last concept.
have
finally
Even
the
in this plurality of possible starting-points, it is evident that "picture" that we form of the reality of nature is not dependent
on the data of sense perception alone, but upon the intellectual views and postulates that we bring to it. Among them, it is espe cially space and time, that uniformly recur in the different systems and thus form the unchanging part, the real invariant, for every
theoretical founding of physics.
of space
Owing
to
this
unchangeableness
glance to be of sensuous content; since sensation never appears outside these forms and, conversely, since these forms are never given separated from sensation, the psychological unity and interpenetration of the two moments leads at once to their logical identification. The very beginning of theoretical physics with Newton, however, destroys this apparent
first
and time, both concepts seem at
unity.
different
It is
when we grasp them
emphasized here that space and time are something after the fashion of immediate sensa-

tion,
171
and when we grasp them after the fashion of mathematical is merely in the latter interpretation that their Absolute, motionless space and absolute, uni formly-flowing time are the true reality, while the relative space and the relative time offered us by outer and inner observation signify only sensuous, and thus inexact, measures for empirical movements. It is the task of physical investigation to advance from these sensuous measures, which are satisfactory for practical purposes, to the If there is objective realities indicated and expressed through them. knowledge of nature, it must present the time-space order of the cosmos not only as it appears to a sentient individual from his relative
And it concepts. truth is affirmed.
standpoint, but also as it is in itself in absolutely universal form. The pure concept alone gives this universality and necessity, because
abstracts from all the differences, which are based on the physiologi cal constitution and the particular position of the individual subject.
it
From the concepts of absolute space and absolute time. of view, the first scientific determination of the epistemological point
Newton
s
problem of objectivity in general is in the definition of space and time, and in the opposition of the sensuous and the mathematical meanings of the two concepts. This problem cannot yet be surveyed in its
whole extent; but the decisive preliminary consideration
here. 40
is
reached
It is clear that the philosophical oppositions in the
funda
mental conception of physics are expressed more strongly in this question than in any other. The struggle over principles has con stantly referred back to the Newtonian theory of space and time in
order to decide the general problem of the foundation of physics. What do the concepts of absolute space and absolute time mean,
if
experience can never give us true examples of these concepts?
Can a conception claim any sort of physical significance, when on principle we have to refuse to apply it definitely to the reality acces sible to us? It must appear as a barren intellectual game that we laws for absolute movements in pure mechanics, so long as develop we have no infallible mark for deciding regarding the absolute or The abstract rule in relative character of an actual movement. itself means nothing, if the conditions are not also known for its concrete application, by which we can subsume empirical cases
under
it.
formulation.
40
contradiction remains here, however, in the Newtonian The laws of natural science, which are to be especially
"objectivity,"
For the problem of
see
more particularly Chs. VI and VII.
172
understood as inductions from given facts, are ultimately related to objects, which (like absolute space "and absolute time) belong to another world than that of experience, and are conceived as the This metaphysical eternal attributes of the infinite divine substance. determination recedes into the background in the further develop
ment
based,
of natural science; but the logical opposition,

is
on which
it is
not thereby removed. The question continually arises, whether in the foundation of mechanics we have to assume only such concepts as are directly borrowed from the empirical bodies and their
perceptible relations, or whether we must transcend the sphere of empirical existence in any direction in order to conceive the laws of
this existence as a perfect, closed unity.
The system of reference of pure mechanics. Henceforth the real The epistemological difficulty is concentrated in this problem. discussion of the mechanical concepts has failed to mark this diffi culty with sufficient sharpness; following the course of history, it has placed the opposition of the "absolute" and "relative" in the center of consideration. But this opposition, which comes from the field of ontology, does not give adequate expression to the method
ological questions pressing for solution.
It is
easy to
see, of course,
that
space and "absolute" time, if they are to be regarded with Newton as mathematical conceptions, do not exclude every
"absolute"
sort of relation.
all
Precisely this, in fact,
is
the essential character of
in
mathematical constructions, that no one of them means anything itself alone, but that each individual is to be understood only in
thorough-going connection with all the others. Thus it is nonsensical to seek to conceive a without at the same time relating it to another distinguished from it, or to seek to establish a moment of
"place"
time without regarding

"here"
it as a point in an ordered manifold. The gains its meaning only with reference to a "there," the "now" only with reference to an earlier or later contrasted with it. No
we subsequently take up into our concepts of space and time, can impugn this fundamental logical character. in the sense that They are and
physical determination, which
remain^/^tems^rekôns
every particular construction in themdenoteTalways only an indi vidual position, that gains its full meaning only through its connec tions with the totality of serial members. Moreover, the conception of absolute motion only apparently contradicts this postulate. No physical thinker has ever interpreted this concept as excluding
173
regard for any system of reference at all. The conflict is only with regard to the kind of system of reference, only with regard to whether
it is
or as an ideal construction.
to be taken as material or as immaterial, as empirically given The postulate of absolute motion does
mean the exclusion of any correlate, but rather contains an assumption as to the nature of this correlate, which is here determined as space separated from all material content. Thereby the problem first loses its vague dialectical form and gains a definite physical meaning. The "relativity," that is inseparably connected with any scientific construction in general, can be left entirely out of account for it constitutes the general and self-evident presupposi tion, which is without significance for the solution of any special But it is such special questions, which are here to be question. decided. It is above all necessary to become clear as to whether and time, in the sense in which physics takes them, are only space aggregates of sensuous impressions, or whether they are independent intellectual "forms;" as to whether the system, to which the funda mental equations of the Newtonian mechanics refer, can be pointed
not
"pure"
;
out as an empirical body, or whether it possesses an "intellectual" being. As soon as we decide in favor of the latter view, the further problem arises of mediating between the ideal beginnings of physics
and
at
its real results.
The sensuous and
intellectual elements stand
abstract opposition and require unification under a general of view^ to determine their part in the unitary concept of point
first in
objectivity. The substitution of the fixed stars for absolute space. At first glance, it might seem that the answer to these questions required the media
no complicated, logical terms. The answer that empiricism has ready avoids all difficulties by resolving the problems into mere The principle of inertia would indeed be meaningless if illusions.
tion of
did not tacitly conceive it with reference to some system of co ordinates, in connection with which the persistence of uniform motion in a straight line could be pointed out. But this unavoidable sub
we
stratum we do not need to establish by wearisome conceptual deduc

tion, since experience of itself definitely presses
it
upon
us.
The
system of reference, in connection with which the phenomenon of movement according to inertia can always be demonstrated with the exactitude of which empirical judgments in general are capable. It is a mistaken desire to ask more than this;
fixed stars offer us a
174
it is

idle to
seek to gain a notion as to what form the law of inertia if we excluded reference to the fixed stars and replaced them by another system. What laws of motion would hold if the fixed stars did not exist, or if we lost the power of orientating our
would assume
observations on them, we lack all possibility of judging, because we are here concerned with a case which is never realized in actual experi
ence.
The world is not given to us twice: once in reality and once in thought; but we must take it as it is offered in sensuous perception, without enquiring how it would appear to us under other conditions, which we logically imagine. 41 In this solution of the problem
by Mach, the consequence
of the empiristic
offered
view
is
drawn with
According to this view, every scientifically valid judgment gains its meaning only as an assertion concerning a con Thought can merely follow the crete, factually present existence. indications of sensation, which reveal this being to us, but it can in
great energy.
no place go beyond them and take into account merely possible, hitherto not given cases. But as has already appeared on all sides, this inference, though unavoidable from the presupposition assumed, contradicts the known fact of scientific procedure itself. The funda
mental theoretical laws of physics throughout speak of cases that are never given in experience nor can be given in it; for in the formula of the law the real object of perception is replaced by its ideal limit. The insight gained through them never issues (Cf. above p. 129 f.) from consideration of the real alone, but from the possible conditions
and circumstances;
"virtual"
it
process. virtual velocities, which, since Lagrange, forms the real foundation of analytic mechanics. The movements of a material system, as here
This
includes not only the actual, but also the is very clearly expressed in the principle of
considered, do not need to be able to be actually carried out; their means merely that we can intellectually formulate "possibility"
them without thereby being brought into contradiction with the conditions of the system. The further development of the principle
in physics has allowed this methodological aspect to appear with In the development of modern thermo increasing distinctness.
dynamics, the principle of virtual changes is freed from its initial limitation to mechanical processes and is transformed into a universal principle, that includes all the fields of physics equally. By a virtual change of a system is now understood not only an infinitesimal
41 Cf. Mach, Die Mechanik in ihrer Entwicklung; Die Geschichte und die Wurzel des Satzes von der Erhallung der Arbeit, p. 47 ff.
175
spatial displacement of its individual parts, but also an infinitely small increase or decrease of temperature, an infinitely small change in
the distribution of electricity on the surface of a conducting body; in short, any elementary increase or decrease of one of the variable
magnitudes that characterize the total state of the system, in so far as it is in accord with the general conditions which the system has to satisfy. Whether the transformation involved can be physically carried out is here of no importance; for the truth of our theoretical deductions is entirely independent of this possibility of an immediate realization of our intellectual operations. in the course of the Duhem remarks, "the magnitudes, to which the theory deductions," refers, are subjected to certain algebraic transformations, then we need not ask whether these calculations have a physical meaning, whether the particular methods of measurement can be directly
"If
translated into the language of concrete intuition, and correspond in this translation to real or possible facts. To raise any such ques
tion
rather to form a totally erroneous conception of the nature of a 42 The discovery and first formulation of the physical theory."
is
Galileo, principle of inertia thoroughly confirms this conception. at least, leaves no doubt that the principle, in the sense that he takes
has not arisen from the consideration of a particular class of He certainly would have made the empirically real movements.
it,
same answer
to the objection, that the truth of the law of inertia
presupposes the permanent existence of the fixed stars, that he made to Simplicio in a similar case: viz., that the reality of the fixed stars
belongs, like that of the moving body itself, only to the "accidental and external" conditions of the investigation on which its real
Into that "mente concipio" theoretical decision does not depend. with which Galileo begins his general exposition, the existence of the
fixed stars does not enter.
straight line is here
by geometry and arithmetic. Whether the which we deduce from such ideal conceptions, are applicable to the world of perception must be ultimately decided by experiment; the logical and mathematical meaning of the hypothetical laws is 43 independent of this form of verification in the actually given.
laws,
42
43
sense; it is ideal schemata offered
The concept of uniform motion in a introduced purely in an abstract phoronomic not related to any material bodies, but merely to the
Cf.
Duhem, L
above
p.
Cf.
evolution do la Mtcanique, Paris 1903, p. 211 ff. 123 ff.; on the whole matter cf. now especially
Natorp
exposition, op.
cil. p.
356
ff.
176

"intellectual experiment"
The
and
the
law of inertia.
In order to
Galileo,
s
justify logically the deduction here actually need ultimately only to appeal to Mach himself.
made by
In
we
Mach
develop
ment
of the general
methods
of physics, the
"intellectual experiment"
He emphasizes that all really fruitful have intellectual experiments as their neces physical investigations sary condition. We must anticipate (at least in its general features) the consequence which a definite arrangement of the experiment
takes an important place.
promises; we must compare the possible determining factors with each other and intellectually vary them, in order to give the obser vation itself a definite direction. It is this procedure of intel lectual variation of the determining factors for a certain result
which first gives us an entirely clear survey of the field of facts. Here the meaning of each individual moment is first made clear; here the perception is first analysed into an ordered complex, in which we clearly grasp the significance of each part in the structure The essential features, on which its action according of the whole. to law depends, are separated from the accidental, which can vary 44 arbitrarily without thereby affecting our real physical deduction. We need only apply all these considerations to the discovery and
expression of the principle of inertia in order to recognize that the real validity of this principle is not bound to any definite material
system of reference. Even if we had found the law at first verified with respect to the fixed stars, there would be nothing to hinder us from freeing it from this condition, by calling to mind that we can
allow the original substratum to vary arbitrarily without the meaning and content of the law itself being thereby affected. For the assump
tion on which
Mach s initial objection rests, viz., that thought must never look beyond the circle of given facts, is now abandoned. The method of the "intellectual experiment" involves a characteristic
activity of thought, by which we go from the real to the possible cases and also undertake to determine them. In fact, the logical
meaning of the law

even
if,
of inertia
in the course of experience, reasons
would obviously remain unchanged were found for us to

the fixed stars themselves.
ascribe
certain
movements
to
The
principles of pure
insight,
44
mechanics would lose none of their validity by this but would be completely maintained in the new system of
1905,
"Uber
Cf.
Mach, Erkenntnis und Irrtum, Leipzig

p. 180
ff.
Gedanken-experi-
mente,"
177
Such a transference orientation, which we would then have to seek. would be impossible even in thought, if these principles merely
reproduced the relations of moving bodies relative to a particular empirical system of reference. Mach himself must, according to his whole assumption, regard the fixed stars not as an element, which enters into the conceptual formulation of the law of inertia, but must conceive them as one of the causal factors on which the law of inertia 45 In a formula that merely expresses the relation and is dependent. interaction of definite physical objects, the one of the two factors can
obviously not be replaced by another without the relation itself thus gaining an entirely new form. If the truth of the law of inertia de
pended on the fixed stars as these definite physical individuals, it would be logically unintelligible that we could ever think of dropping this connection and going over to another system of refer ence. The principle of inertia would in this case not be so much a universal principle of the phenomena of motion in general, as rather an assertion concerning definite properties and "reactions" of a given empirical system of objects; and how could we expect that the physical properties found in a concrete individual thing could be separated from their real "subject" and transferred to another? In any case, we see in this example that empiricism and empirical
then
method (Empirie)
are
different.
The only meaning that the
principle of inertia could have, according to the empiristic assump tions, is one that corresponds in no way to the meaning and function
it
has actually
is
fulfilled in scientific
is
The
but
logical principle of
mechanics rather abandoned.
mechanics from the beginning. not grasped and explained here,
Streintz s concept of the "fundamental
principle can be
made
The same objection in body." to every attempt to give the law of inertia a
45 "A free body, affected by a momentary pair of forces, moves in such a fashion that its central ellipsoid, if its center is fixed, rotates without sliding on the tangential plane parallel to the plane of the pair of forces. This is a motion in consequence of inertia. In this, the body makes the strangest turns
with reference to the heavenly bodies. Is it believed now that these bodies, without which the conceived motion cannot be described, are without influence on it? Does what one must name explicitly or tacitly, if one is to describe a phenomenon, not belong to the essential conditions, to the causal nexus of the same? The remote heavenly bodies in our example have no influence on acceleration, but have influence on velocity." (Mach, Erhaltung der Arbeit,
p. 49.)
178
fixed basis
by pointing out
its
system of reference as somehow present
in the reality of things.
system of reference any as it satisfies the double condition of not revolving and not being subjected to the influence of outer forces. The absence of rotation can be definitely indicated by certain instruments of measurement,
designated by Streintz by the name of "gyroscopic compass;" every "absolute" rotation of a body brings about some physical effects, that can be directly perceived and measured. A similarly immediate and positive decision is never possible with regard to the second
aspect, the absence of external forces: here we must simply be content with the fact that, as often as a variation from motion in a straight
Streintz has attempted to define as this arbitrary, empirically given body in so far
from uniform velocity has been observed in the movement of a point in a constant direction relative to a body, we have hitherto always succeeded in indicating some external bodies, which appear as
line or
by virtue of their relative position to the or to the assumed system of reference. If, now, moving point we characterize the body defined by the two conditions indicated by
the causes of this variation
itself
:
the absence of rotation and by perfect independence of all surround ing masses, as the fundamental body (FK), then we have in such a
body a
tial
These equations, which in the manner that they are ordinarily formulated are logically The principle of indefinite, thus gain a fixed and definite meaning. inertia, in particular, can now be expressed in the form, that every point left to itself moves in a straight line and with constant velocity with reference to this fundamental body. 48 This attempted deduc tion, it can be easily shown, rests however on a conversion of the real If Streintz s explanation were true, logical and historical relation. the mechanical principles would be merely inductions, which we have verified in particular bodies with definite physical properties, and which we have then taken as probably true for all bodies of the same
sort.
suitable system with reference to which the equations at the basis of physics are satisfied.
dynamic
differen
The claim
to strict universality
made by
these principles
would then be entirely unintelligible. It could not be understood by what right we opposed them to the observed facts as postulates
predetermining the direction of our explanation, instead of trans forming the principles (which have, indeed, only been gained by
Cf. the more particular exposition by Streintz, Die physikalischen Grundlagen der Mechanik, Leipzig 1883, p. 13 ff., 22 ff
.
48

definite observations), as
179
experiences.
But even
if
soon as they failed to agree with the new we abstract from this, the consideration
decisive, that the fundamental body and the fundamental system of coordinates could never be discovered as empirical facts, if the meaning of both had not been previously established in ideal construction. The seemingly pure inductions, which Streintz makes the basis of his explanation, are already guided and dominated by the fundamental conceptions of analytic mechanics. It is only on the
would be
assumption of these conceptions that the meaning of the two aspects determining the fundamental body can be understood: the absence of rotation and the independence of external forces are the empirical criteria by which we recognize whether a definite given body satisfies the presuppositions of the theory, which we have previously devel oped independently. The property (Merkmal), by which we estab lish whether an individual case can be subsumed under a definite law, is logically strictly separated from the conditions, on which the The idea of inertia did not arise from validity of the law itself rests.
observations of definite bodies, from which we were able, as it were, to read off sensuously the property of their being under no external influence; but it can only be explained conversely, on the basis of
the idea that
we seek for bodies of this sort and give them a special in the structure of our empirical reality. Thus the attempt of place in so far as it is meant to be a true founding of mechanics, Streintz,
involves a circle; for in the experiments and empirical propositions, which form the basis of it, there is already a tacit recognition of the principles which are to be deduced. Analytic mechanics, as history
shows, has come into existence without these experiments, while conversely the mere conception of these experiments could only arise
on the The
basis of this mechanics. 47
theory of C. Neumann: the body alpha. Thus, if we still demand the connection of the law of inertia with some material system of reference, and if we would also explain the rational structure of
mechanics, there remains finally only one escape, viz., of assuming an unknown body not given in experience, and of explaining the funda mental dynamic equations with reference to it. The working out of
this
thought was
first
attempted by C.
Neumann
in his
work on the
principles of the Galileo-Newtonian theory, in which, along with the

47 C/. especially the criticism of Streintz attempt by Hofler, Studien zur gegenw. Philosophic der Mechanik, Leipzig 1900, p. 136 f.
180
explanation of the fundamental physical problem, the methodolog ical question also was clearly expressed. According to Neumann,
the principle of Galileo can only be grasped in its conceptual meaning through the assumption of a definite existential background. Only
an unknown point of space an form and dimensions to all absolutely rigid body, unchanging "The are the propositions of our mechanics intelligible. time, of Galileo, that a material point left to itself moves in a straight saying
in a world in
which there
exists at
in its
line,
appears as a proposition without content, as a proposition hanging in the air, which needs a definite background in order to be
Some special body in the cosmos must be given as a basis of judgment, as that object with reference to which all move ments are to be estimated only then are we in a position to connect
understood.
;
a definite content with those words. What body is it, to which we can give this special position? Unfortunately neither Galileo nor Newton have given us a definite answer to this question. But when we attentively examine the theoretical structure founded by them and expanded continuously up to the present time, its foundations can no longer be concealed from us. We easily recognize that all
the motions present in the universe or conceivable are to be related to one and the same body. Where this body is found, what reasons
are to be given for granting a single body so prominent, as it were sov to this, at all events, we gain no answer." 48 ereign, a position,
We
should not expect to find within physics the proof establishing the existence of this unique body, which is called by Neumann the "body
For the proof is, in fact, of purely ontological nature; the alpha." postulate of a unitary logical point of reference is made into an asser tion of an empirically unknowable existence. And to this existence,
to be of material nature, are ascribed all those predicates usually employed by the ontological argument: it is unchanging^ eternal and indestructible. While, on the one hand, a being with
although
it is
absolute properties is deduced here from mere thought, on the other hand, the converse feature appears, viz., that the conceivability of our
ideal conceptions
If
is
made dependent on
definite properties of being.
force of nature, the propositions of mechanics would necessarily cease not only to be The concept of the strict conapplicable, but even to be intelligible.
48
we conceive the body alpha annihilated by any
Carl Neumann,
f.
liber die
Prinzipien der Galilei-N ewtonschen Theorie,
Leipzig 1870, p. 14
181
stancy of direction, the concept of uniform motion of definite velocity, such as mathematical theory offers us, would at one stroke be deprived of all meaning. Thus not only definite physical consequences
but also the most noteworthy logical consequences would be con nected with a process in the outer world thus it would depend on the being or not-being of an actual spatial thing whether our fundamental mathematical hypotheses had any meaning in them. But how could
;
we ever make a
meaning
established?
of these universal
judgment regarding a physical reality, if the mathematical predicates were not already To all these questions, only one answer could ultimately
rational
be given by the advocates of the body alpha. It is not the existence of the body alpha, they could answer, but the assumption of this exist ence, on which the validity of our mechanical concepts depends. This assumption can never be taken from us it is a pure postulate of our scientific thought, which herein obeys only its own norms and
;
rules.
Such an answer, however, places the problem on an entirely If it is in our power to deal with ideal assumptions, then we do not understand why this procedure should be limited to
new
basis.
the assumption of physical things. Instead of the body alpha, we could then assume (in logically the only unobjectionable and intelli
gible
relations.
manner) pure space itself, and grant it definite properties and Here also we have moved in a circle; thought by its inner necessity has led us back to that very starting-point at which the first doubt and suspicion arose regarding the formulation of
mechanical principles. Space and time as mathematical ideals. We first escape the dilemma when we resolve to place our intellectual postulates in full clarity at the beginning, instead of somehow introducing them in a concealed form in the course of the deduction. The absolute space and the ab
solute time of mechanics involve the as does the pure
geometry.
problem of existence just as little of arithmetic or the pure straight line of Absolute space and time arise in the sure and continuous
number
of these concepts; Galileo emphasized most sharply that the general theory of motion signified not a branch of applied but
development
of pure
The phoronomical concepts of uniform and of accelerated motion contain nothing originally of the sensu uniformly ous properties of material bodies, but merely define a certain relation
mathematics.
between the spatial and temporal magnitudes that are generated and related to each other according to an ideal principle of construction.
182
Thus, in the expression of the principle of inertia, we can rely at first merely on a conceptual system of reference, to which we ascribe all the determinations required. By means of conceptual definitions, we create a spatial "inertial system" and an "inertial time-scale," and take both as a basis for all further consideration of the phenomena
Thus there is no hypomotion and their reciprocal relations. statization of absolute space and absolute time into transcendent things but at the same time both remain as pure functions, by means
of
;
49
of
which an exact knowledge

that
is
fixity,
we must
of empirical reality is possible. The ascribe to the original and unitary system of
50
reference,
have established
not a sensuous but a logical property; it means that we it as a concept, in order to regard it as identical and
unchanging through all the transformations of calculation. The ideal system of axes, to which we look, satisfies the fundamental postulate, which requires independence of all outer forces for the
"fundamental
system of
coordinates:"
for
how
could forces affect

lines in
lines,
pure geometrical forms?
While considering these
our
intellectual abstraction as absolutely constant,
we evolve from them

Only
a general schema for possible spatial changes in general.
experience can ultimately decide whether this schema is applicable to the reality of physical things and processes. Here also it is never possible to isolate the fundamental hypotheses and to point
them out
as valid individually in concrete perceptions; but we can always only justify them indirectly in the total system of connection that they effect among phenomena. We de (Cf. above p. 146 ff.) velop the determination of the "inertial system" and the mathemati
cal
consequences connected with
it
purely in theory.
In so far as
any empirically given body seems to conform to these determinations, we ascribe "absolute" rest and absolute fixity to it also; i.e., we affirm that a material point left to itself must move uniformly in a But at the same time, straight line with reference to that body.
we know that
this postulate is never exactly fulfilled in experience, but always only with a certain approximation. But just as there is
49
Cf.
"inertial
more particularly regarding the mathematical construction of the system," Ludwig Lange, Die geschichtliche Entwicklung des Bewe-
gungsbegriffs,
50
Wundt s Philos. Studien III, (1886), p. 390 ff., 677 ff. See Erkenntnisproblem II, 344, 356 f., 559; cf. now especially the excellent treatment of Edm. Konig, Kant und die Natunvissenschaft, Braunschweig 1907,
p. 129
ff.

no
real straight line, that realizes all the properties of the pure
183
geomet
rical concept, just so there is
no
real
body, that conforms to the
mechanical definition of the inertial system in all respects. Thus the possibility always remains open of establishing, by the choice of a new point of reference, a closer and more exact agreement between
the system of observations and the system of deductions. This relativity is indeed unavoidable for it lies in the very concept of the
;
It is the expression of the necessary difference, object of experience. that remains between the exact conceptual laws we formulate and
their empirical realization.

rest (for example, the
That any system of given bodies is at system of the fixed stars) does not signify a
fact that can be directly established
by perception or measurement, but means that a paradigm is found here in the world of bodies for certain principles of pure mechanics, in which they can be, as it were, visibly demonstrated and represented. The fixed stars stand in relations with the moving bodies of the real world, which relations
are entirely according to the system of these propositions and find The individual material point of their complete expression in it. attachment, to which we connect our equations of motion, may
change; yet the fundamental relation to a definite system of laws of mechanics and physics remains constant. Analogously, we substitute a more exact measure of time for the not wholly exact measure, which the stellar day affords, by taking our stand on the law of the con servation of energy and on the law of gravitation. That unit of time is taken as "absolutely" exact, whose application enables us, on the one hand, to avoid contradiction of the theoretical demands of the principle of energy, and on the other hand to avoid conflict between
moon as actually observed, and as calculated according to the Newtonian law. 51 Thus there is still a relation to the physical concepts of absolute space and absolute
the secular acceleration of the
concepts does not consist in their but in their removing the necessarily every relation, assumed point of reference from the material into the ideal. The system, in which we seek our intellectual orientation, is no individual perceptible body, but a system of theoretical and empirical rules on which the concrete totality of phenomena is conceived to be dependent.
time.
of these
The meaning
stripping off
Hertz s system of mechanics. This meaning of the concepts of absolute space and absolute time was established in its general fea51
Poincar4,
La mesure du temps
(cf.
above, p. 145
f.).
184
tures
by
Leibniz.
For him, both concepts were only another expres
sion of the thorough-going spatial and temporal determinateness, which we demand for all being and process. This determinateness must be demanded, even if there is in the cosmos no strictly uniform
course of any actual process of nature, or any fixed or
immovable
body.
Theoretically it can always be gained; for we can always relate the non-uniform movements, whose law we know, to certain conceived uniform movements, and on the basis of this procedure
calculate in advance the consequence of the connection of different movements. 52 The relation between theory and experience assumed here has found its clearest expression in modern times in the system Hertz in his exposition takes space of mechanics of Heinrich Hertz. and time at first merely in the sense in which they are offered to
"inner
intuition."
The
assertions,
that are
made
of
them, are
"judgments
a priori in the sense of
Kant;"
any appeal to the experi
ence of sensuously perceptible bodies is foreign to them. It is only in the second book, where the transition is made from geometry and kinematics to the mechanics of material systems, that times,
spaces and masses are conceived as signs of outer empirical objects; the properties of these empirical objects, however, cannot contra dict the properties, which we have previously ascribed to these
magnitudes as forms of our inner intuition or by

assertions
definition.
"Our
concerning the relations of times, spaces and masses should thus no longer satisfy merely the claims of our intellect, but at the same time they should correspond to all possible, especially to
These assertions thus rest not only on the laws future, experiences. of our intuition and our thought, but on previous experience besides." By taking fixed units of measurement as a basis, especially within this
and comparing the empirical spaces, times and masses with each other in accordance to them, we gain a general principle of coordination by means of which we allow certain mathematical
field,
symbols definitely to correspond to the concrete sensations and perceptions, and thereby translate the given impressions into the symbolic language of our inner intellectual images. The indefiniteness, that necessarily belongs to these ultimate units of
is
measurement,
not indefiniteness of our images, and not of our laws of transforma tion and correlation, but it is the indefiniteness of the very outer "We mean to say that experience to be copied. by our senses we can
"
Leibniz, Nouveaux Essais, Bk. II, Ch. 14.
185
determine no time more exactly than can be measured with the help
of the best chronometer, no position more exactly than can be referred to the coordinate system of the remote fixed stars, and no
done by the best scales." 53 Thus, while a perfect definiteness of all elements can be gained with the struc tures we generate from the laws of intuition and thought, in the field of empirical phenomena this is merely postulated. We measure the of our abstract dynamic of our experiences by the "reality" concepts and principles. The order of the world which we construe on the presupposition of the motionlessness of the fixed stars, is
mass more exactly than
is
"truth"
for us the true order of things, in so far as all actually observable motions with reference to this system have always hitherto corre
sponded with the greatest approximation to the axioms, by which mechanics characterizes the concept of "absolute movement." If this condition should no longer be fulfilled (and we must consider this in our calculations and assumptions as a possible case), still these axioms, the ideal according to which the construction has taken place, would be wholly unaffected in meaning; its empirical realization would only be shifted to another place.
Thus absolute space, if we mean by it not the abstract space of mechanics but the definite order of the world of bodies, is at all events never finally given, but is always only sought. But in this, there is no lessening of its objective meaning for our knowledge; for, as a sharper analysis shows, relative space also signifies no given fact in the sense of a dogmatic "positivism." Also when we con sider any corporeal masses in their mutual positions and their relative distances, we have already gone beyond the limit of sensuous impres
sions.
When we
speak of
"distance,"
we mean by
this,
strictly
speaking, no relation between sensuous bodies, since these bodies might have very different distances from each other according as we
took one point or another of their volume as a starting-point for measurement. In order to gain an exact geometrical meaning, we must substitute a relation between points for a relation between
bodies, by considering the total mass of the body reduced to the center of gravity. are thus obliged to have the direct empirical intuition transformed by means of the pure geometrical limiting
We
concepts, in order to
of |relative position
63
make a completely
exact assertion regarding the
two material systems.
The
positivistic scruples
ff.,
Heinrich Hertz, Die Prinzipien der Mechanik, p. 53
p. 157
ff.
186
against the
time of mechanics thus space and the prove nothing because they would prove too much logically thought out, they would also have to forbid every representation of physically given bodies in a geometrical system in which there are fixed positions and distances. The physical space of bodies is no isolated essence,
"pure"
"pure"
but
only possible by virtue of the geometrical space of lines and This relation was also expressed by Leibniz in an espe distances. It is indeed true, as he explains, cially striking and pregnant saying.
is
that more is posited in the concept of body than in the concept of mere space but it does not follow from this that the extension, which we perceive in bodies, differs from the ideal extension of geometry in any properties. Number also is something different from the totality of counted things, yet plurality as such has one and the same mean
;
ing,
whether we define
it
in purely conceptual terms or represent

"In
it
in
some concrete example. the same sense, we can also say that we do not have to conceive two sorts of extension, the abstract of space and the concrete of bodies, for the concrete receives its properties
54 We inscribe the data of experience in only through the abstract." our constructive schema, and thus gain a picture of physical reality; but this picture always remains a plan, not a copy, and is thus always
capable of change, although
its
main
features remain constant in the
concepts of geometry and phoronomy.

Construction and convention.
arbitrariness
is
It seems, indeed, that

scientific reflection,
an element of
admitted into our
when we thus
ground our assertion concerning reality in previous constructions. This conclusion is actually drawn, when the concepts of the "inertial
and the "inertial time-scale" are described as mere con which we introduce in order to survey the facts more easily, but which have no immediate objective correlate in empirical fact. 65 Poincare, in an investigation of the conditions of time measurement, has deduced the general consequences very decisively. When we
system"
ventions,
take any natural phenomenon as absolutely uniform, and measure all others by it, we are never absolutely determined in our choice from without; no measure of time is truer than any other, but the only proof we can give for any is merely that it is more convenient.
The question
64
raised here
is
open to no
final
answer so far as our
See Leibniz, Nouveaux Essais, Bk. II, Ch. 4. Cf. Ludwig Lange, op. cit.; also Das Inertialsystem vor dem Naturforschung, Philos. Studien, Vol. II.
55
Forum
der
187
previous investigation is concerned; for it reaches over from the a methodologically alien field. Science has no criterion than truth; and can have no other than unity and higher completeness in the systematic construction of experience. Every
field of science into
scend"
other conception of the object lies outside its field; it must "tran itself even to be able to conceive the problem of another The distinction between an "absolute" truth of sort of objectivity.
being and a "relative" truth of scientific knowledge, the separation between what is necessary from the standpoint of our concepts and what is necessary in itself from the nature of the facts, signifies a metaphysical assumption, that must be tested in its right and validity The characterization of the before it can be used as a standard. ideal conceptual creations as "conventions" has thus at first only one
meaning; it involves the recognition that thought does not proceed merely receptively and imitatively in them, but develops a characteristic and original spontaneity (Selbsttdtigkeif) Yet this
intelligible
.
.,
spontaneity is not unlimited and unrestrained; it is connected, although not with the individual perception, with the system of
perceptions in their order and connection.
It is true this order is
never to be established in a single system of concepts, which excludes any choice, but it always leaves room for different possibilities of
exposition; in so far as our intellectual construction is extended and takes up new elements into itself, it appears that it does not proceed
law
according to caprice, but follows a certain law of progress. This is the ultimate criterion of "objectivity;" for it shows us that
the world-system of physics more and more excludes all the accidents of judgment, such as seem unavoidable from the standpoint of the
individual observer, and discovers in their place that necessity that 56 is universally the kernel of the concept of the object.
VII
The concept of energy. Necessary as are space and time in the construction of empirical reality, they are after all only the universal forms in which it is represented. They are the fundamental orders,
in
which the
real
is
of the real itself.
A new principle is
VI and VII.
arranged, but they do not determine the concept needed to fill these empty forms
with concrete content.

56
This principle has been conceived in different
Cf. later Chs.
188
ways, from Democritus concept of matter opposed to empty space as the Tra/iTrXTjpts 6v, until its final logical definition in the modern conception of energy. Here, for the first time, we seem to have the
ground of reality under our

all
Here we have a being that fulfills feet. the conditions of true and independent existence, since it is
indestructible
physical reasons, energism claims an epistemological advantage. matter, which real type of objective reality for the older natural constitute the science, are reduced to mere abstractions through the closer analysis
and
eternal.
Along with
all
The atom and
of the data
and conditions
of our knowledge.
They
are conceptual
impressions, but they can never be In in real meaning with the immediate sensation itself. compared we grasp the real because it is the effective. Here no mere energy, symbol comes between us and the physical thing; here we are no longer in the realm of mere thought, but in the realm of being. And in order to grasp this ultimate being, we need no circuitous route through complicated mathematical hypotheses, since it is directly revealed unsought in perception itself. What we sense is not the doubtful and in itself entirely indefinite matter, that we assume as the "bearer" of sensuous properties; but it is the concrete effect, which is worked on us by outer things. "What we see is nothing but radiat ing energy, which effects chemical changes in the retina of our eye, that are felt as light. When we touch a solid body, we feel the mechanical work that is involved in the compression of our finger Smell and tips and also in the compression of the body touched. taste rest on chemical activities in the organs of the nose and mouth. Everywhere energies and activities are what inform us as to how the outer world is arranged and what properties it has; and, from this standpoint, the totality of nature appears to us as a division of spatially and temporally changeable energies in space and time, of which we gain knowledge to the extent that these energies go over to our body, in particular to the sense organs constituted for receiving
limits, to which
we attach our
The "thing" as a passive and indifferent [substratum of properties is now set aside. The object is what it lappears to be: a sum of actual and possible ways of acting. In this doctrine, a determination of purely philosophical reflection is
definite
j
57
energies."
admitted into the foundations of

of
reflection
67
is
scientific thought; but the function Henceforth all thereby limited and exhausted.
f.
Ostwald, Vorles. uber Naturphilosophie, p. 159

merely speculative points of view can be
189
strictly excluded, and consid eration limited to the reproduction of the empirical facts. In the of this task, the primal reality itself without abstract accomplishment
felt
must be with regard to this conception. Whatever physical advantages the concept of energy may have over that of matter and the atom,
logically at least, both stand same sphere of consideration.
or conceptual husk becomes increasingly clear. Energy and the sense qualities. There is one doubt that
on the same plane and belong to the

This
is
shown negatively
in the similar
from the sensuously given. The notion that "ener can be seen or heard is obviously no less nai ve than the notion gies" that the "matter" of theoretical physics can be directly touched and
difference of both
grasped with the hands.

of sensation: of
What warm and cold,
is
light
given us are qualitative differences and dark, sweet and bitter, but
not numerical differences of quantities of work. When we refer sensations to such magnitudes and their mutual relations, we have
brought about just such a translation into a foreign language as energism criticizes in the mechanical view of the world. To measure a perception means to change it into another form of being and to
with definite theoretical assumptions of judgment. The advantage, which energism might claim over mechanism, could never consist in that it did without these presup positions, but rather in that it understood them more clearly in their It could not be a matter of entirely excluding logical character. "hypotheses," but only of not making them into absolute properties
approach
it
(Cf. p. 141
ff.)
is done by dogmatic materialism. Energy and the concept of number. If we conceive the problem in this way, it is seen that energism contains a motive from the beginning, which protects it more than any other physical view from the danger of an immediate hypostatization of abstract principles. Its fundamen tal thought, from an epistemological point of view, does not go back primarily to the concept of space, but to that of number. It is to
of things as
numerical values and relations that the theoretical and experimental inquiry are alike directed, and in them consists the real kernel of the fundamental law. Number, however, can not be misunderstood
as substance, unless
we
are to revert to the mysticism of Pythago-
reanism, but it signifies merely a general point of view, by which we make the sensuous manifold unitary and uniform in conception.
The evolution
of the conception of energy offers a concrete physical
190

for this general process of
knowledge. We saw that the mathematical objectification of the given was to conceive it under certain serial concepts. The given only becomes an object of knowledge when it is "established" in this sense, by as cribing to it a definite place in a manifold, ordered and graded accord But the real task of knowledge of nature ing to some point of view. is not thereby exhausted; in fact, in principle it is not yet begun. The insertion of the sensuous manifold into series of purely mathe
example
first
step in the
matical structure remains inadequate, as long as these series are separated from each other. As long as this is the case, the "thing"
of popular experience is not yet wholly grasped in its logical meaning. It is not enough for us to express the individual physical and chemical
i
properties in a pure numerical value,
and to represent the object as
k totality of such "parameters." For the object means more than a mere sum of properties; it means the unity of the properties, and
thus their reciprocal dependency.
If this
postulate
is
to find its
adequate expression in science, we must seek a principle, which enables us to connect the different series, in which we have first
arranged the content of the given, among themselves by a unitary law. Heat, motion, electricity, chemical attraction mean at first only
tions.
certain abstract types, to which we relate the whole of our percep In order to advance from them to a representation of the
real process, a
thorough mediation
is
needed, so that
all
these differ
ent fields again becomes
members
of
an inclusive system.
The concept of
meaning
of
the measure of work. From this point, the general the conception of energism can be surveyed. The
structure of mathematical physics is in principle complete when we have arranged the members of the individual series according to an
exact numerical scale, and when we discover a constant numerical relation governing the transition from one series to the others. Only
when
is done is the way determined from one member to any and prescribed by fixed rules of deduction, no matter what the other, series. Only then it becomes clear how all the threads of the mathe matical system of phenomena are connected on all sides, so that no element remains without connection. This relation was first estab lished empirically in the case of the equivalence of motion and heat; but, once discovered, it was extended beyond this starting-point. The conception was soon extended as a universal postulate to the totality of possible physical manifolds in general. The law of energy
this

directs us to coordinate every
191
member
of a manifold with one
and
only one
any other manifold, in so far as to any quantum of motion there corresponds one quantum of heat, to any quantum of
of
electricity,
all
member
one quantum of chemical attraction,
etc.
In the concept of
these determinations of magnitude are related to a common work, denominator. If such a connection is once established, then every
numerical difference that we find within one series can be completely expressed and reproduced in the appropriate values of any other series. The unit of comparison, which we take as a basis, can arbi
trarily
vary without the result being affected.

field are
any them in any
of
equal when the same amount
series of physical qualities,
If two elements work corresponds to then this equality must be
of
maintained, even when we go over to any other series for the purpose of their numerical comparison. In this postulate, the essential
content of the principle of conservation
is
already exhausted; for
any quantity of work, which arose "from nothing," would violate the principle of the mutual one to one coordination of all series. If
we wish to represent the system schematically, we have here a number of series A, B, C, of which the members ai a 2 a 3 a n bi b 2 b 3 bn
.
stand in a definite physical relation of exchangeability, such that any member of A can be replaced by a definite member of B or C without the capacity of work of the physical system in which
Ci 02 Ca ... c n ,
this
substitution
is
represent this relation of possible substitution not

ing each individual
assumed being thereby changed. We briefly by always coordinat
member with the multitude of corresponding but by ascribing to it once for all a certain value of equivalents, energy, which draws all these coordinations into a single pregnant ex We do not compare the different systems with each other pression. directly, but create for this purpose a common series, to which they
are
all
that
we
is
equally related. It is chiefly owing to technical circumstances traditionally choose mechanical work as this common series,
"types
since the transformation of the different
of
energy"
into this
form
relatively easy
and exactly measurable.
In
itself,
however,
totality of possible relations.
be taken as the basis for expressing the In any case, it appears that energy in this form of deduction is never a new thing, but is a unitary system ,0^ All that can be said of it reference on which we base measurement. on scientific grounds is exhausted in the quantitative relations of equivalence, that prevail between the different fields of physics.
any arbitrary
series could
192
Energy does not appear as a new objective somewhat, alongside of the already known physical contents, such as light and heat, electri city and magnetism; but it signifies merely an objective correlation according to law, in which all these contents stand. Its real meaning and function consists in the equation it permits us to establish between the diverse groups of processes. If we clothed the principle itself, which demands the definite quantitative correlation of the totality of phenomena, in the form of a particular thing, even in the form of thing, the all-inclusive substance, we should create the same dogmatic confusion, that energism charges against materialism. Science at least knows nothing of such a transformation into sub The identity science seeks, and to stance, and cannot understand it. which it connects the chaotic individual phenomena, has always the form of a supreme mathematical law; not, however, of an all-inclusive and thus ultimately propertyless and indeterminate object. Con ceived as a particular thing, energy would be a somewhat, which was at once motion and heat, magnetism and electricity, and yet also nothing of all these. As a principle, it signifies nothing but an intel lectual point of view, from which all these phenomena can be meas ured, and thus brought into one system in spite of all sensuous
"the"
diversity.
The formal presuppositions of energism. At this point, in the midst of the controversial questions of contemporary philosophy of Paradoxical nature, we are led to make a general logical remark. as it may appear, here where consideration seems wholly given over to the facts, the effect of general logical theories is evident.
Whether we conceive energy
as a substance, or as the expression of a causal relation, depends finally on our general idea of the nature of the scientific construction of concepts in general. However frankly the physical investigator intends to stand face to face with nature
can nevertheless be shown that in the construction of energism motives have been at work, which have their real origin in
itself,
it
definite
"formal"
convictions.
At
how
deeply the problems of
"form"
and how
lasting is their influence. to each other in the problem of the concept. The one, that has remained dominant in traditional logic, bases the concept on the procedure of abstraction, i.e., on the separating out of an identical or
this point, we recognize anew penetrate into those of "matter," Two different views were opposed
similar part
from a plurality
of similar perceptions.
The content
193
thus gained is, strictly speaking, of the same property and nature as the object from which it is abstracted; it represents a property that in general is not isolated, yet which can always be pointed out
as a constitutive part in these objects, and therefore possesses a concrete existence. The concept is accordingly the "presentation of the common;"
it is
the unification of those individual features,
that belong uniformly to definite classes of objects. Opposed to this conception, however, is another that is founded above all on the
In this concept, we do not analysis of the mathematical concepts. to separate the given by comparison into classes whose indi attempt
vidual examples all agree in certain properties, but to construct the given from a postulate of unity by a process according to law. Here it is not so much the particular parts of the given, that are isolated,
its organization these are investigated in their characteristic relational structure. rests; The meaning of this opposition (Cf: above esp. p. 14 ff. and 81 ff.)
but rather the connections and relations on which
in views of the concept now appears from a new angle; it is this opposition, that is noticeable in modern discussion as to the formula
tion of the principle of energy. Rankine, who first created the name and concept of a general "energetics," proceeds from purely methodological considerations in his treatise devoted to the first
establishment of the
conception. Physics, as he explains, is from the purely abstract sciences, such characteristically separated as geometry, by the fact that the definitions fundamental in the
new
an abstract science do not necessarily correspond to and the theorems deduced from them need not necessarily be laws of real processes and phenomena, while the true scientific concept is to be nothing but the designation of certain There is, in general, a properties common to a class of real objects.
development
of
any
existing things,
two-fold
way
to separate out such properties.
We
can,
by a pure
method, separate from a group of given things or phenomena that group of determinations which is common to all members of the class, and which belongs to them directly in their sensuous appearance; or we can go behind the phenomena to certain
"abstractive"
Only the
scientific
hypotheses for the explanation of the field of physical facts in question. first procedure strictly corresponds to the demands of
and philosophic criticism. For only here are we sure that we do not falsify the observations by arbitrary interpretation; only here do we remain purely in the field of the facts themselves, for
194
while
we
divide the facts into definite classes,

It
is
feature to them.
of energetics,
we add no foreign an advantage in principle of the new science that from the beginning it uses merely this abstractive
does not refer the phenomenon of heat to molecular of magnetism to any hypothetical in the simple form in which they are
procedure.
It
movements, the phenomenon fluid, but it grasps both merely
"Instead of compounding the different classes offered to perception. of physical processes in some obscure way out of motions and forces,
we would merely separate out those properties, that these classes have in common, and in this way define more inclusive classes, which we represent by appropriate terms. In this way, we finally reach a group of principles applicable to all physical phenomena in general,
since they are merely inductively derived from the facts themselves, are free from that uncertainty, which always attaches even to such mechanical hypotheses whose consequences seem com
and which,
pletely confirmed
by
experience."
deduction of energetics. The first result gained from this mode of investigation is the general concept of energy. It signifies nothing else than the capacity to bring forth changes; and this
Rankine
capacity
is
the most universal determination that
we can
it
distinguish
in the bodies of our world of perception,
and without
they would
cease to be physical phenomena for us. If we discover definite universal laws concerning this property, then, allowing for particular circumstances, they must be applicable to every branch of physics in general, and must represent a system of rules, which every natural
58 The way in which Rankine establishes process as such obeys. and proves these rules concerns merely the historical development of 59 but the logical form he chooses for his thoughts is of the physics;
most general philosophical interest. The laws of energy, we see, owe their universality to the circumstance that the property of things, which we have called energy, is spread throughout the physi cal universe, and somehow attaches to every body as such. No part
of reality can escape these laws, because each part is This form only by this distinguishing property.
68
known
of
as real
deduction
Rankine, Oullines of the Science of Energetics, Proceedings of the Philo sophical Society of Glasgow. Vol. Ill, London and Glasgow, 1855, p. 381 ff. 69 Cf. especially on Rankine, Helm, Die Energetik, p. 110 ff., also A. Rey, La Theorie de la Physique chez les Physiciens contemporains. Paris 1907, p. 49 ff.
195
already determines the general intellectual category, under which energy is conceived here. It stands in principle on the same plane with the perceptible things, whose essential being it constitutes.
Energy is, as it were, concrete substantiality ible and eternal being.
itself,
the one indestruct
Criticism of the method of physical abstraction." From the stand of epistemology, the gap to be pointed out is not so much in point Rankine s physics as in his theory of method. According to Rankine,
the most general property by which the objects of physical reality are distinguished, is the capacity to produce and to receive effects.
Things
first
ceived as
members
gain their real objective character, when they are con of actual or possible causal relations. The
unprejudiced, "abstractive" analysis, which Rankine regards as the ideal of true science, shows with certainty that causality is no
property that can be pointed out as an immediate part of perceptions.
Both
and empiristic criticism agree at least in one that there are no direct impressions corresponding to the conclusion, concepts cause and effect. Thus, if abstraction, as here understood,
rationalistic
is
clear that precisely that aspect
only a division and grouping of the material of perception, it is must escape it, on which the concept
of energy is founded. And even if we grant the "power of producing effects" to be a quality inhering in bodies just as does any other sensuous property, such as their color or odor, even so the real
problem would not yet be solved. In the construction of energetics, are not concerned as to whether this power of producing effects can in general be shown to exist, but with the fact that it can be exactly measured. But as soon as we enquire as to the methods by which this numerical determination is made possible, we are referred to a system of intellectual conceptions and conditions, which has no sufficient basis in the purely abstractive procedure, as has been shown on all hands. The mathematical foundation of energetics already involves all those methods of "construction of series," which can never be adequately grounded from the ordinary standpoint of
we
abstraction.
The problem of
rate,
abstraction in
modern
logic.
Modern
logic, at
any
has substituted for the old principle of abstraction a new one, which may be introduced here. In this new principle of abstrac
tion, the
procedure
relations
but from
is not from things and their common properties, between concepts. If we define a symmetrical and
>
196
transitive relation
R for a number of members a,b,c,

the relations bRa,
(so that
from
60
the relations
aRb and bRc,
cRb and aRc
follow),
then the connection produced in this way can also always be expressed by introducing a new identity x, which stands in a definite relation R to every member of our original series. Instead of comparing the
members
directly with each other in their possible relations, the pos
sible relations
between the
serial
establishing the relation of each to this

tions
members can now be represented by x, and thus forming the rela
aR x, bR x, cR x.
The
relation
R R
is
here an asymmetrical,
many-one relation, so that the members a, b, c can stand in the relation mentioned to no other term than x, while x, on the contrary, with several members. 61 can stand in the corresponding relation
We
have an example of this procedure in the relation between series, which we call their "similarity." Two series s and s are said to be similar to each other in the ordinal sense, when there subsists between them a definite, reciprocal relation of such a sort that to every member of s one member of s corresponds (and conversely), and when,
f
if
in the series in s (x
(n)
precedes a member y, the correlate of the correlate of y(y ). Here we have a symmet precedes
s,
member x
rical
s"
and
transitive
etc.
relation,
by which a
plurality of series
s,
... s
can now which we

all
the basis of this relation, we produce, by the principle of abstraction, a new concept call the common "type of order" of all these series. To
can be connected.
On
the series bound together in this manner, we ascribe one and the same conceptual property. We replace the system of coordinations by the assumption of an identical property, which belongs to all
however, that we do not claim to thin^- oui claim is only that a common ideal point of reference is thereby produced, with reference to which we can make our assertions regarding the relations of the
the series uniformly.
It
is clear,
have discovered a new
self -existent
given series more pregnant, and bring
them
to a single, concentrated
judgment.
of concepts,
If
we now apply
this result to the physical construction

is
an essential feature of the modern concept of energy
Here also we start from the establishment of clearly revealed. certain dependencies between empirically physical series. dis-
We
80
For the concept of the transitive and symmetrical relation see above,
II.
( /.
Ch.
61
more particularly Russell, Principles
of Mathematics, p. 166, 219
ff.
etc.
197
isolated
cover that the manifolds, which at first seem to stand side by side and independent, are connected with each other by a rela
"equivalence,"
tion of
by
in one series one
and
virtue of which there corresponds to a value extend this only one value of the other.
We
connection by taking more and more fields of physical process into account, until finally on the ground of observation and general deductive reasons we draw the conclusion, that whenever any
arbitrary groups of physical
phenomena
are given, definite relations of
equivalence must
Here is given a thorough prevail between them. transitive and symmetrical relation between physical con going 62 and it is the validity of this universally applicable relation, tents; which leads us to introduce a new being by coordinating a certain
work-value, a certain quantity of energy, to every individual
of the
series.
member
This being, however, would obviously lose compared all meaning, if we wished to separate it from the whole system of judgments, in which it has arisen. The being, posited in it, is not
ivthat of
an
the
isolated
"being"
sensuous propertv to be perceived for

of certain laws of connection-. M
"^^
^^*~"
"
itself,
% but
j
we anew what deep, actual oppositions may be concealed recognize" behind the differences regarding the logical schema. If we follow the traditional doctrine of abstraction, then we are almost neces
it is
i
M^^Êiffi^^^^^"Tn^^~^^
""
_.
,_
in
At
this point,
*
sarily forced to a substantial interpretation of energy, as the
example
of
its
Rankine shows; while the functional theory
of the concept finds
natural correlate in a functional determination of the supreme physical "reality." In the one case, consideration ends in the
assumption of a property common to all bodies, and, in the other case in the creation of a highest common standard of measurement for
changes in general. Energy as a relational concept. Some of the representatives of energism have already made the latter logical interpretation. Here we must remember above all Robert Mayer, who also determined the general theoretical position of the new concept, which he introduced. The transformation of force into motion, of motion into heat, meant for him, as he emphasizes, nothing but the establishment of the fact that certain quantitative relations are found between two different groups of phenomena. "How heat arises from disappearing motion
all
62 If we denote the relation of equivalence by A, then from aAb, bAa evi dently follows, as on the other hand the validity of aAb and bAc also involves that of aAc; thus the condition of symmetry and transitivity is fulfilled.
198
or, in
my manner How
to
demand an
of speaking, how answer to this is to
motion
is
transformed into heat

of the
demand too much
human
mind.
give water, why a material of disappearing O and does not result, would trouble no chemist. But other properties whether he does not come closer to the laws, to which his objects,
when he understands that the resulting amount of water can be precisely determined from the disappearing amounts of oxygen and hydrogen, than when he is conscious of no
the materials, are subjected,
the sense of its cannot be questioned." 63 Helm justly remarks here, "energetics is a pure system founder," of relations/ and is not to establish a new absolute in the world. // changes occur, then this definite mathematical relation subsists
such connection,
this
"In
between them, that is the formula of energetics; and certainly it is often also the only formula of all true knowledge of nature." as the spirit of investigation has rested in the bed of sloth of an It may be a comfortable dream absolute, it has been undone. that our questions can find their answer in the atoms, but yet it is a dream. And it would be no less a dream, if we saw an absolute in energy, and not rather the temporarily most adequate expression of
"As
64 quantitative relations among the phenomena of nature." energy, like the atom, is more and more divested of all sensuous
Thus
mean
This (Cf. above p. 217 ff.) ing with the advance of knowledge. development appears most clearly in the concept of potential energy,
which even
in its general
name
points to a peculiar logical problem.
As Heinrich Hertz has emphasized, there
is a peculiar difficulty in the assumption, that the alleged substantial energy should exist in such diverse forms of existence as the kinetic and the potential
form.
Potential energy, as
it
is
ordinarily conceived, contradicts
every definition that ascribes to it the properties of a substance; for the quantity of a substance must necessarily be a positive magnitude, while the totality of potential energy in a system is under some
circumstances to be expressed by a negative value. 85
can, in fact, according to
43 64
Such a relation
Gauss theory
of negatives, only be explained
Mayer
to Griesinger (Kleinere Schriften
und Briefe,
p. 187).
Helm, Die Energetik, p. 20, 562. The same definition of energy as a mere "causal standard of measurement" (Kausalmass) is given by H. Driesch, Naturbegriffe und Natururteile. Berlin, 1904 (Abh. zur Didaktik u. Philos. der Naturwiss., Heft 2). 85 See H. Hertz, Die Prinzipien der Mechanik, p. 26.
199
where that which is counted has an opposite, i.e., "where not sub stances (objects conceivable in themselves) but relations between two objects are counted." (Cf. above p. 56.)
Even where energy is at first intro Energetics and mechanics. duced as a unitary and indestructible object, as in the case of Robert Mayer, this very category of object gradually assumes a new mean ing; this new meaning is for the purpose of doing justice to the new "The eleva content, which appears in a double form of existence. tion of a kilogram to a height of 5 meters," says Robert Mayer, "and the movement of such a weight with a velocity of 10 meters per second, are one and the same object; such a motion can be trans formed again into the elevation of a weight, but then naturally ceases to be motion, just as the elevation of a weight is no longer elevation of a weight, ,when it is transformed into motion." 66 If mere elevation above a certain level (thus a mere state) is here assumed to be identical with the fall over a certain distance (thus with a temporal process), then it is clearly evident that no immediate substantial standard is applied to both, and that they are not compared with each other according to any similarity of factual property, but not merely as abstract measuring values. The two are the because they share any objective property, but because they can occur as members of the same causal equation, and thus can be substituted for each other from the standpoint of pure magnitude. We begin with the discovery of an exact numerical relation, and posit that new we call energy as an expression of this relation. Here a "object" The real radically new turn is taken, in opposition to atomism.
"same"
advantage of energism over the "mechanical" hypotheses, as ordinar understood by its adherents, is that it keeps closer to the given facts of perception, in so far as it permits us to relate two qualitatively different fields of natural phenomena, without previously having reduced them to processes of movement, and thus having divested them of their specific character. The processes remain unaffected
ily
our assertions are merely directed to But, on the other hand, precisely this exclusive reference to the numerical rule of relation involves a new
in their specific character, as all
their
causal connection.
The atom, even while its purely conceptual becomes more pronounced, always appears as the meaning gradually analogue, as it were, the reduced model of the empirical, sensuous
intellectual
moment.
66
Mayer, Kleiner e Schriflen und Briefe,
p. 178.
200
body, while energy belongs to another field in its origin. Energy is able to institute an order among the totality of phenomena, because
it itself is
all
on the same plane with no one of them; because, lacking concrete existence, energy only expresses a pure relation of mutual
dependency. From the epistemological standpoint, the claim, which energism makes to understand the different groups of physical processes in
their specific character (instead of transforming them into mechani cal processes and thereby extinguishing their individual features) is
one that now seems limited although justified within a certain sphere. In fact, the general logical possibility appears here of our shaping nature into a system, without our being obliged to require represen tation of this system in a unitary, intuitive picture, such as is offered
by mechanism.
tative"
world.
But it is an error to see in this tendency to a "quali a reversion to the general Aristotelian view of the physics are forced," says a prominent, modern advocate of "We
"to
take up into our physics other features than the purely quantitative elements, of which the geometrician treats, and thus to agree that matter has qualities; we must face the danger that we shall
energism,
be accused of a reversion to the occult faculties of scholasticism, of recognizing the quality, by which a body is warm or bright, electrical
or magnetic, as an original and not further reducible property in it; in other words, we must abandon all the attempts that have con
tinually been made since the time of Descartes, and connect our theories again with the essential concepts of the peripatetic physics."
But the
of a deeper connection. entirely different
further working out of the thought destroys the appearance The qualities of Aristotle are something
from the qualities of modern physics; for while the former signify only hypostasized sensuous properties, the latter have already passed through the whole conceptual system of mathematics, and have thereby received a new logical form and character. What
energism abandons is only the "explanation" of the particular quali ties out of certain mechanical motions; what it retains, on the other hand, and what is the condition of its existence, is the expression of quality in a definite number which fully represents and replaces it in our consideration. The question whether heat is motion can remain
in the
colder are replaced at the

degree of temperature
background, as long as the indefinite sensations of warmer and same time, by the concept of an exact
and
objectified in
it.
What
is
here retained of

quality
its
201
is not its sensuous property, but merely the peculiarity of mathematical serial form. Duhem, whose judgment regarding the connection between the energistic and the peripatetic physics was just cited, 67 says we can develop a theory of heat and can define the expression "quantity of heat" without borrowing anything from the specific perceptions of cold and warm. 68 In the schema of theoretical physics, the definite empirical system under investiga tion is replaced by a system of numerical values, which express its various quantitative elements. Energism (Cf. above p. 150 f.) shows that this form of numerical order is not necessarily connected without analyzing the things and processes into their ultimate The intuitive parts, and recompounding them from the latter. general problem of mathematical determination can be worked out without any necessity for this sort of concrete composition of a whole out of its parts. Physics as a science of qualities. In this conception, however, physics only completes and applies a thought already recognized in the general doctrine of the principles of mathematics. There is, and can be, a "physics of qualities," because and in so far as there is a mathematics of qualities. The gradual development of this latter can already be traced in its general features. There is a continuous development beginning with Leibniz, who first saw the essence of mathematics in a doctrine of the possible forms of deductive con
nection in general, and who therefore demanded the completion of ordinary algebra (as the science of quantity) by a general science of
(Scientia generalis de qualitate), up to modern projective geometry and the theory of groups. In this whole development, it
quality
clearly appears that there are wide and fruitful fields perfectly accessible to mathematical determination without their objects
being extensive magnitudes which have arisen by the repeated additive The projective theory of dis positing of one and the same unity.
tances shows
how it is possible to place the elements of a spatial manifold in exact correlation to fixed numerical values, and to imprint a definite order upon them by virtue of this correlation, without
applying the ordinary metrical concept of distance. (Cf. above p. 84 ff.) This thought is carried over by universal energetics to the
67
Duhem,
Levolution de la Mecanique, p. 197

ff.
.
f .,
as also H. Driesch, Naturbe-
griffe
68
und Natururteile, p. 51 Duhem, op. cit., p. 233 f
202
It is sufficient for the numerical totality of physical manifolds. formulation of processes if a definite scale of comparison is provided
for the individual qualities,
and
if,
further, the values within these
different scales are reciprocally coordinated with each other
This connection, however, can be established objective law. retained independently of any mechanical interpretation of
particular groups of
by an and
the
phenomena.
The
objection frequently raised
against energism, that it destroys the homogeneity of process since in it nature falls into separate classes of phenomena, is not conclusive. If we take the mathematical general concept as the starting-point and
which standard of judgment, not only such contents are share some intuitive property that can be given for itself, but all which can be deduced from each other by a structures are
"like,"
"like,"
p. 81 ff.) Here, the connection of concepts pro duced by the equivalence between the different series gives a no less definite logical connection than reduction to a common mechanical
fixed
conceptual
rule.
(Cf.
above especially
however, this criterion
is satisfied;
model.
The
intellectual postulate of
homogeneity
is
just as effective
in the energistic as in the mechanistic conception of the natural
realization
processes; the difference only consists in that, in the one case, its is based purely on the concept of number, while, in the
other case, it also requires the concept of space. The conflict between these two conceptions can ultimately only be decided by the history of physics itself; for only history can show which of the two views
can
be most adequate to the concrete tasks and problems. Abstracting from this, however, energism is in any case of preemi nent epistemological interest in so far as the attempt is made to establish the minimum of conditions, under which we can still speak of a "measurability" of phenomena in general. 69 Only those princifinally
thai
has been objected occasionally to the logical possibility of the goal general energetics sets itself, that every measurement of things and processes involves the presupposition that they are compounded out of homo
tt
9l
unit.
genous parts, and can thus be represented by repeated addition of the same Every measure would thus necessarily be a determination of extension;
reference to a unit of measurement would thus already contain the transforma tion of all qualitative differences into extensive distances, and thus reduction to a spatial and mechanical scheme. (Cf. Rey, chez les Physiciens contemporains, p. 264, 286, etc.) of
"measure"
is
"measurement"
of a
obviously taken too narrowly. manifold only its mathematical determination in general,
Theorie de la Physique Here, however, the concept If we understand under the
La

pies
203
and rules are truly general, on which rests the numerical deter mination of any particular process whatever and its numerical comparison with any other process. The comparison itself, however, does not presuppose that we have already discovered any unity of for example, between heat and motion; but, on the con "essence," mathematical physics begins by establishing an exact numerical trary,
relation, on the basis of which it also maintains the homogeneity of such processes as can in no way be sensuously reduced to each other. That the different forms of energy themselves" are of kinetic nature is a proposition that the theory of knowledge cannot defend, for the latter is merely directed on the fundamental aspects of know The demands of the theory of ing, not on those of absolute being.
"in
knowledge are rather
satisfied
when a way
is
shown
for relating every
physical process to values of mechanical work, and thus for producing a complex of coordinations, in which each individual process has its
definite place. A representation of the processes of nature absolutely without hypotheses cannot, indeed, be gained in this way; for trans lation into the language of the abstract numerical concepts, no less than translation into the language of spatial concepts, involves a
theoretical transformation of the empirical material of perception. But it is of logical value to separate the general presuppositions strictly from the particular assumptions; and to separate the "meta
physical,"
from those
particular
because mathematical, principles of knowledge of nature special hypotheses, which serve only in the treatment of a
field.
VIII
The problem of the construction of concepts in chemistry. The exposi tion of the conceptual construction of exact natural science is incom plete on the logical side as long as it does not take into consideration
the fundamental concepts of chemistry. The epistemological inter est of these fundamental concepts rests above all on the intermediate
position which they occupy.
purely empirical description of the particular substances
Chemistry seems to begin with the and their

advances, the more
it
composition; but the further

i.e., the
it
also tends
correlation of its elements with the particular members of the series of numbers, then mathematics itself shows that such a correlation is also possible where the objects of the group in question are not compounded out
of spatial parts.
204
In physical chemistry, this goal in constructive concepts. reached; a leading representative of this discipline is able to give as the connecting feature of physics and chemistry, that both create the systems, which they investigate, on the basis of the empiri
toward
is
fact
cal data. 70
In so far as chemistry has reached this modern form,
Its funda it stands on no other ground logically than physics itself. mental laws, as for example the law of phases (Phasenregel) of Gibbs or the law of chemical masses (Massenwirkung), belong to the same purely mathematical type as the propositions of theoretical physics. It is, however, of special interest to trace the way in which the ideal, which was realized in theoretical physics from its first beginnings in Galileo and Newton, has only been gradually reached in chemistry. The limits of purely empirical and of rational knowledge stand out very clearly in the constant shifting, which they undergo in the progress of chemical knowledge. The intermediate terms and con
ditions of exact understanding are brought into sharp relief. power of scientific moulding is especially striking with the
The more
stubborn material, with which chemistry works. Ultimately physics is only apparently concerned with thing-concepts; for its goal and real field is that of pure law concepts. Chemistry places the problem
of the individual thing decisively in the foreground. Here the particular materials of empirical reality and their particular proper ties are the object of the enquiry. But the "concept," in the specific
has in mathematics and physics, is not available for this For it is only the symbol of a certain form of connec tion, that has more and more lost all material content; it only signifies a type of possible arrangement, not the "what" of the elements ar
meaning
it
new problem.
ranged. Are we here concerned with a gap that is to be filled out by new determinations belonging to the same logical direction of thought, or must we recognize and introduce at this point a form of knowledge
different in principle?
proportions.
The chemistry of sensuous qualities and Richter This question can only be answered
s
if
law of definite we follow the
concrete historical development of chemical doctrines themselves, not grasping the vast wealth of their content in detail, but laying bare
the great logical lines of direction of their advance.
In
fact,
a few
general characteristics soon appear of themselves, according to which the development can be divided and surveyed in all its diversity.
70
Nernst, Die Ziele der physikalischen Chemie. Gottingen 1896.
205
The
older form of the chemical doctrine of elements, which pre dominated up to the time of Lavoisier, and which found its last
characteristic expression in the theory of phlogiston, conceives the element as a generic property belonging to all the members of a definite group and determining their perceptible type. Here the elements are only the hypostatizations of the especially striking sensuous qualities. Thus sulphur by its presence confers on any
body the property of combustibility, salt the property of solubility, while mercury is the bearer of the metallic properties, which are found empirically in any material. 71 This conception is only tran scended in principle when along with the task of dividing bodies into classes according to their generic properties is added the other task of gaining exact quantitative proportions concerning their mutual relation. The demand for strict numerical determination here also forms the decisive turning point. The law of the definite proportions,
in
which the
different elements are connected with each other, con
stitutes the starting-point of
modern chemical theory.
It
is
interest
ing that this law is at first conceived entirely independently of any conception of the constitution of matter, and in particular inde
pendently of the atomic hypothesis.

form, in
In the original,
J.
still
incomplete
D. Richter, it signifies by but the validity of certain harmonic relations, that primarily nothing If we consider a series of prevail between different series of bodies. acids A! A 2 A 3 ... and a series of bases BI B 2 B 3 there prevails between the two a certain relation, which is expressed by saying, that
which
it
was
first
asserted
we coordinate with each member of the first series a definite number while we let correspond to the members of the second mim 2m3 series other constant numerical values n^ns, to be gained by observation. The manner, in which an element of the first series combines with an element of the second, is definitely determined by
.
. .
these numbers; the
two weights, according to which an acid A p combines with any base B q are related according to the correspond Richter seeks to prove in detail ing numerical values p and n q
,
that the series of weights of the bases forms_an arithmetical series, and that of the acids a geometrical series; and that a law is thus
found here, which

71
is
assTimecTtb be analogous to that of the distances

ff; cf.
See Ostwald, Leitlinien der Chemie, p. 4

ff.
also Meyerson, Identite
et
Realite, p. 213
206
of the planets

from the sun. 72
This conception has not proved to
empirically; but it is nevertheless characteristic and significant in its general tendency. It is, as we see, the general Pythagorean doctrine of the "harmony" of the cosmos, which here
be satisfactory
of
modern chemistry, as it is also at the cradle In this connection, Richter is to be compared with Kepler, that is, if we consider not his whole achievements but merely his intellectual tendency, for with Kepler he shares the conception of the thorough-going numerical arrangement of the
is
present at the cradle of
modern
physics.
universe, which
is
continued in
all
particular fields of
phenomena.
founder adds a
Dalton
law of multiple proportions.
The
interpretation of the law
of constant
combination-numbers by
its real scientific
new
first
concrete feature to this general view. Here it is actually at only asserted, that there is a characteristic equivalence-number
for each element,
and
that,
when two
or
more elements enter
into a
combination, their masses are related as whole multiples of these
numbers. But this rule of "multiple proportion" is combined by Dalton with a certain interpretation, and enters only in this form into the system of chemical doctrines. The concept of combinationweight is transformed into that of atomic weight. The law of multi
ple proportions
different in their masses, while within the
atom
is
means, that the atoms of different simple bodies are same chemical genus the one and the same constant mass, and always unchangeably
thus the mass suffices to characterize a given simple material in its In place of the empirically gained proportionspecific character. numbers of the individual bodies, assertions appear regarding an
essential
property of their ultimate constitutive parts. Since, our knowledge only concerns the relations, according to however, which the elements enter into combinations, no definite determination
all
is possible of the absolute values of the atomic weights. If we take the atomic weight of hydrogen as a unit of comparison, then we can, without contradicting the known facts of composition, determine that
of
oxygen by the value

72
O=
8 instead of
O =
16,
whereby we would
cf.,
For the following data from the history of chemistry,
besides the well-
known general historical works, especially Wurtz, La Theorie atomique, Paris 1879; Duhem, Le Mixte et la combinaison chimique, Paris 1902; Lothar Meyer,
Die modernen Theorien der Chemie, 5th
iiber die
On
ed., Breslau 384; Ladenburg, Vortrdge Entmcklungsgeschichte der Chemie, 3rd. ed., Braunschweig 1902.
Richter, see espec.
Duhem,
p. 69
ff.,
Ladenburg,
p. 53
ff .

have to double the number
of
207
atoms
of
oxygen
8, 16,
in all our formulae;
we
... as the atomic weight of sulphur, in so far as we formed the chemical formulae in agreement with one of these assumptions, and thus, for example, characterize sulphide of hydrogen according to our choice by the
could successively take the values S
32
expression HS 2 or HS or possible determinations is
H 2S. The decision between all these made on the basis of several criteria,
which are only gradually worked out in the history of chemistry. One of the most important criteria is the rule of Avogadro, according to which similar quantities of molecules of different combinations occupy the same volume as perfect gases under the same conditions of pressure and temperature. Along with the determination of atomic weights from the density of vapors, which is hereby made possible, there is their determination from heat capacity, which rests on the law of Dulong-Petit; and also the determination on the basis of isomorphism, resting on the law of Mitscherlich, that the same crystal form having different combinations, indicates an equal number of atoms connected in the same way. It is only the totality of all these different points of view, mutually confirming and correcting
each other, that finally after many experiments gives a unitary table of atomic weights, and thus lays the basis of a definite system
of chemical formulae. 73
The development here completed concept. a general logical problem, if we abstract from all details. If we asked merely the individual investigators who cooperated, it would seem to possess only one perfectly definite and clear meaning
The atom as a relational
offers
for
is
them
all.
The
presupposed;
it is
objective existence of the different types of atoms only necessary to discover their properties, and
to define
more
them more exactly. The further we advance and the diverse groups of phenomena we consider, so much the more The substantial definitely the wealth of these properties appears. "inwardness" of the atom is revealed and takes on fixed and tangible
for us.
form
We trace,
how
especially in the involved chemical constitu
the atoms are situated relatively to each other, and tion-formula, how they are mutually connected in the unified structure of the molecule. We see how in their combination they generate, by their
number and
73
relative position, a certain structural outline, such as

esp. Lothar Meyer, Bk.
I,
is
Cf.
more particularly
Pt. II-IV; Ostwald.
Grundriss der allgemeinen Chemie, 4th ed., Lpz. 1909.
208
expressed,

e.g.,
in the
forms of crystals.
If
we look
into the closer
empirical grounding of these assertions, however, the general picture soon changes. It becomes clear that the atom is never the given
starting-point, but always only the goal of our scientific statements. The wealth of content it gains in the progress of scientific investiga
tion never belongs to it fundamentally, but is related to another kind of empirical "subject." As we apparently investigate the
in its manifold determinations, we at the same time these different groups of circumstances in a new relation to place each other. speak of the number of atoms contained in a
atom
itself
We
of a gaseous substance, and thereby express a relation, according to the law of Gay Lussac, subsists between the numerical value of the density of the gas and the value of its com bination-weight. We ascribe to the atom of all simple bodies the
definite
volume
that,
same heat-capacity and thereby express the
fact that,
if
we arrange
a"
. .
combination-weights of the chemical elements in a series a a and the values of their specific heats in another series b b
an
b"
bn
then there
is
a definite correlation between these two

,
a"
series,
in
so far as the products ab, a b etc. possess the same constant value. The characteristic logical function of the concept of the atom appears clearly in these examples we may abstract from all
b"
;
metaphysical assertions regarding the existence of atoms. The atom functions here as the conceived unitary center of a system of coordi
nates, in which we conceive all assertions concerning the various groups of chemical properties arranged. The diverse and originally heterogeneous manifolds of determinations gain a fixed connection
when we
is
relate
them to
this
common
center.
The
particular
property only apparently connected with the atom as its absolute in order that the system of relations can be perfected. In "bearer,"
we are concerned not so much with relating the diverse series to the atom, as rather with relating them reciprocally to each other through the mediation of the concept of the atom. Here again appears the same intellectual process, that we previously met; the complicated
truth,
relations between certain systems are not expressed by our comparing each system individually with all the others, but by putting them all in relation to one and the same identical term. (Cf. above p. 196 ff .) The attempt to determine exactly the atomic weight of the individual elements compels us to appeal constantly to new fields of chemico-physiml phenomena as criteria. To the extent that this
209
determination advances, the circle of empirical relations is extended. If we conceive this progress completed, then in the "absolute"
atomic weights
particular
positive
all
series
could
possible relations would be expressed, that the enter into among themselves. The real
is in the systematic confused factual material originally is organized; it is no longer unrelated, but is arranged around a fixed central point. When we ascribe to one and the same subject the observations on vapor density, on heat capacity, on isomorphism
outcome
of chemical
knowledge here
analysis of these relations.
The
etc.,
is
they thereby enter into true conceptual relation. not the only logical value of this "subject," that
But, indeed,
it
it
describes
and
unifies
productive;
previous observations. This unification is rather directly it produces a general schema for future observations,
and indicates a definite direction for these. The progress of science would be slow, its exposition unwieldy and tiresome, if, on approach ing a new field of facts, it had every time explicitly to repeat the wealth of previously gained material and to present it in all details. While the concept of the atom concentrates all these features, it retains their essential content; on the other hand, it leaves all
the forces of thought free to grasp the
totality of
a single
tion, in
new empirical content. The what is empirically known is condensed, as it were, to point, and from this point issue the different lines of direc
accordance with which our knowledge advances into the Those manifolds already discovered and defined accord ing to law function as a fixed logical unity in opposition to those manifolds newly to be discovered; and it is this unity of the funda
unknown.
mental point of connection, which renders possible our assumption of an ultimate identical subject for the totality of possible properties. The "regulative" use of the concept of the atom. The meaning,
belonging to the general concept of substance within the actual processes of experience, is clearly evident in this example. Empirical knowledge cannot avoid the concept of substance, although genuine
philosophical progress in such knowledge is in understanding and evaluating it as a concept. True, the direct, living work of investi
gation has another standpoint from the beginning, and grasps the problem as if from another side than that of epistemological reflec
Investigation finds its interest in the new fields of facts to be mastered, while it can take known facts as a given condition needing no further analysis. The totality of the "factual" in this sense is
tion.
210
fixed; it is the
permanent substratum, which gives the fundamental What has already been reached at any point, what has been won, must be taken by the investiga tion as something assured and given; for only thus is it possible to move the field of the problematical to another point, and to push it
mould
for all further observations.
further on, so that new questions are always coming in for considera Thus the passive fixity, established by science at certain tion.
points,
is
an element
in its
own
activity.
In
fact, it is justified
and
unavoidable, that science should condense a wealth of empirical relations into a single expression, into the assumption of a particular
thing-like
"bearer."
The
critical self-characterization of
its
thought,
however, must analyze this product once more into

factors,
although it purposes of knowledge. This is done because critical thought is not directed forwards on the gaining of new objective experiences, but
particular conceives this product as necessary for certain
backwards on the origin and foundations of knowledge.
The two
tendencies of thought here referred to can never be directly united; the conditions of scientific production are different from those of
critical reflection.
We
cannot use functions in the construction of
empirical reality and at the same time consider and describe them. Nevertheless, the two standpoints, and therewith the constant alteration of standpoints, are desirable, in order to judge knowledge
as a whole in the motives of
its
advance, and in the permanent logical
peculiar character of knowledge on the tension and opposition remaining between these stand In the light of this, it can be understood that the chemical points.
its
conditions of
existence.
The
rests
concept of the atom also shows a different form, according to the way we approach it. To the first naive consideration, the atom appears as a fixed substantial kernel, from which different properties
can be successively distinguished and separated out, while, con from the standpoint of the critique of knowledge, precisely those "properties" and their mutual relations form the real empirical / data, for which the concept of the atom is created. The given factual
versely,
material
is
anticipated,
united in a sipglefocusVith that which is conceptually and by virj/ue of a natural illusion this focus appears as-
unitary object, (instead of a mere "virtual" point. Thus the of chemistry is an. Idea," in the strict meaning Kant gave / this term, in so far as it possesses most admirable and indispen
real,
atom
"a
sably necessary regulative use, in directing the understanding to a

/
211
iX
converge and which, though it is an idea only (focus imaginarius) that is, a point from which, as lying completely outside the limits, of possible experi
certain aim, towards
which
all
the lines of
its rules
,
ence, the concepts of the understanding do not in reality proceed, serves nevertheless to impart to them the greatest unity and the 74 This function remains as a permanent charac greatest extension."
teristic
concept of the atom, although its content may completely change; thus e.g., the atom of matter becomes the atom of Precisely this sort of change shows that electricity, the electron.
of
the
what
is
essential in the concept does not consist in
any material
properties, but that it is a formal concept, that can be filled with manifold concrete content according to the state of our experience.
The concept
of valency
and
the theory of types.
The second impor
tant step in the construction of chemical concepts, after the concep tion of the atom, and after the value of the atomic weights of the
individual elements has in general been established, is to connect according to conceptual standpoints the various, originally separate determinations thus gained, and to collect them into classes of
definite
character.
The
empirical facts
distinctions
and combinations within the
the relations of chemical substitution.
leading to such relative system are given in If we trace how the atoms of
total
various simple materials replace each other in combinations, and how they can be reciprocally substituted for each other, certain fun
damental rules result governing this form of relation. The form of substitution can be determined once for all and expressed by certain numerical values, which we attach to each element along with the numerical value of its combination weight. If we take the atom of hydrogen as unity, it appears, for example, that an atom of chlorine can in certain combinations replace an atom of hydrogen, while an atom of oxygen always replaces two, an atom of nitrogen three, an atom of carbon four atoms of hydrogen. Thus a new point of view is achieved for the correlation of the individual elements, and a new
characteristic constant for each simple material. The "valency" of the elements is the expression of a definite property in them, that if we belongs to them independently of their chemical affinity.
Now
arrange the chemical combinations according to this

74
new
principle,
Kritik der reinen Vernunft, 2nd. ed., p. 672.
Muller
s trans, p. 518.
212
they are divided into various types, in which the members belong ing to the same type are characterized by the fact that they can all be produced from a certain form by progressive substitutions,
which take place according to the rules

ular atoms.
of the valency of the partic
The concept of the "type" Logical aspects of the concept of type. here comes in for consideration not in its significance for the special
problems of chemistry, but as a paradigm of certain logical relations. In fact, it displays most distinctly a characteristic feature already
established in general
by the
of type
analysis of the exact scientific concept.
islajsa .n,ot formed on the pattern of the The different concept, but on thatXff the serial concept. generic combinations, belonging under a type, are not so conceived because
The chemical concept
of the external similarity of their sensuous properties, or because of the direct agreement in their chemical functions. They belong together in so far as they can be changed into each other by means of
the relation, which subsists between the valency of the individual atoms, while the remote members of the series need no further analogy
In the history of is established by this law of derivation itself. chemistry, the concept of type was only gradually separated from the concept of chemical analogy. 75 The first step in this separation is found in the relation of substitution itself, since here elements, which seem entirely different from each other in their nature and
than
The conception of substi properties, can replace each other. tution, as formulated by Dumas, was at first rejected by Berzelius as paradoxical and inconsistent from this point of view; chlorine
cannot take the place of hydrogen in any combination, as the former
(according to the theory of electro-chemical dualism, which Berzelius advocated,) is negatively electric, while hydrogen possesses positive The more, however, the theory of substitution made electricity.
way, the more the converse view gained acceptance, viz., that entirely dissimilar materials can replace each other in certain com The binations, without altering the nature of the combination. of this view appear even more sharply, when not only consequences the elements, which can be substituted for each other, are contrasted individually, but the whole group of materials, that can issue from
its
repeated substitutions,
75
is
considered.
Here
also the
ff.,
demand
for
cit.,
Cf.
ff.
more particularly
esp.
Duhem, Le
Mixte, p. 97
Wurtz, op.
p. 189
213
analogy was at first upheld, until further investigations showed that the series, which arise in this way, can contain members completely different from each other in all their perceptible properties and in
their essential chemical determinations.
of
To
the
"chemical
type"
which he demanded similar fundamental properties in all members, there is now opposed the "molecular type" of Regnault, which includes materials of very different properties, and considers these materials as issuing from each other by substitution. The conditions, on which the unity of the type rests, thus correspond throughout to those that we found realized in the field of the con Then there were geometrical struction of mathematical concepts. systems and groups of systems, whose elements were not connected by any intuitive feature common to them, but merely by the definite rule of relation, which prevailed from member to member; and the
Dumas,
for
same
lishes
is
true here. The among them such
"valency"
of the particular elements estab
a relation as
produces in
its
continued
The application definite systems of characteristic serial types. variation according to law of this "parameter" generates and founds the form of the concept, which accordingly does not rest on a simi
larity in the
content of
what
is
connected, but on the type
of-
connection.
The chemical concept as a relational concept. The chemical concept is indeed distinguished from the mathematical, in that the relation by which we proceed from one member to another is established by mathematics purely constructively, while the relation of equivalence, on the other hand, is discovered as an empirical relation between the various elements. However, if we abstract from this difference of
origin,
is
recognize that, once the decisive property for comparison gained, the further conceptual construction on both sides takes exactly the same direction. Here, once a general principle of co
we
ordination has been defined, our concern is with carrying this principle through the whole manifold of materials given by observation, and thus with shaping the latter aggregate into a system, within which
we grasp the reciprocal action and interdependence of the particular members according to fixed rules. In this regard, the theory of types
first approach to chemical deduction, since it shows us how, from certain starting-points, to construct the manifold of bodies by adherence to a few general principles, and how to group them around fixed central points. The sensuously heterogeneous now
constitutes the
214
relations.
becomes homogeneous, when we organize it into certain numerical It is the numerical and relational aspect, which is here
;
again decisive for
it
constitutes the genuinely characteristic property
of the scientific interpretation of the chemical concepts.

ency,"
The
"val
ascribed to the particular atoms, must at first appear as a real qualitas occulta, if it is conceived as a substantial quality in them.
We do not know the peculiar property of the atom of chlorine, by which it can only be combined with one atom of hydrogen; we do not know by what force the atom of oxygen combines with two atoms of hydrogen, and the atom of carbon with four of hydrogen. And
this riddle is not solved
when,
in explanation of the different relative
valencies, reference is made to the states of motion of the individual atoms, which states are supposed to agree or to be opposed to each other in such a way that they can only combine with each other in an
For here something absolutely unknown and empirically undemonstrable is substituted for the relations of What distinguished the con substitution, which are alone known. of valency from all scholastic qualities is the intellectual renun cept It does not seek to penetrate into the ciation, which it implies. substantial nature of the connection of atom and atom, but merely to
represent the facts of this connection according to universal quan titative principles of order. The chemical constitution-formula at
76 entirely definite relation.
a direct intuitive picture of the serial order and atoms among themselves; but what it finally achieves is not such a knowledge of the ultimate, absolute elements of reality, as rather a general analysis of the bodies and materials of experi
first
seems to
offer
position of the
of a definite compound does not teach us to merely in its composition, but inserts it into various typical series, and thus refers to the totality of such structures, that can arise by substitution out of a given combination. The individual member becomes the representative of the whole group to which it belongs, and it can issue from the group by variation according to law of certain fundamental parts. Since the constitution-formula
ence.
The formula
know
it
represents this connection, this formula is indeed the real scientific expression of the empirical reality of the body; for it means nothing else than the thorough-going objective connection, in which an
individual
"thing"
and possible experiences.

76
or particular event stands with the totality of real (Cf. here esp. Ch. VI.)
p. 175.
Cf.
Wurtz, La Theorie atomique,

The concept of
the
"radical"
215
and the theories of "composite radicals." becomes especially significant, when it is applied not merely to the individual atom but to whole groups of The theory of the "composite radical" now arises and atoms. becomes the real foundation of organic chemistry. Here, according
The conception
of substitution
to the definition of Liebig, the radical is considered as a constant part in a series of compounds, in so far as it can be replaced in the latter by other simple bodies, or in so far as in its combination with a
simple body this body can be excluded and replaced
by equivalents
of other simple bodies. Regarding the manner in which the radicals "exist" in compounds, there is disagreement at first. In the "kernel
theory"
of Laurent, the relation

realistic sense.
is
at
first
conceived and described in
kernels are as such present in a plurality of bodies, which arise from them by their combination with other atoms; they preexist with respect to the more complex struc
a thoroughly
The
In the further development of the theory, this view is more and more superseded. When Gerhardt, in particular, shows that it is possible to assume two radicals in a compound, then the conception
tures.
Since the of the real existence of isolated groups is destroyed. formulae of chemistry are only meant to express by equations certain
relations of structure
and
reaction, since they are not intended to
represent what bodies are in and for themselves, but only what they were or could become, there is nothing to prevent us, it is now urged,
from erecting several rational formulae

according as we wish to express
its
for one and the same body, connection with one or another
group of compounds. The conflict concerning the nature and absolute properties of the radical is thus resolved for the radicals now appear as the result of certain ideal analyses, which we make, and
;
which give taken as a

but
is
different results, according to the standpoint of

basis.
comparison
The
radical
now
possesses no independent reality,

"the
meant to express what Gerhardt calls relations, by which or groups of atoms replace each other." 77 We thus stand elements
at the beginning of a conception, which in general abandons the questions as to whether and how the elements continue to exist in
and
the compounds into which they enter, in order instead to discover represent, according to general rules, merely the measurable
relations, that subsist
77
between the
initial
and
final conditions of
Gerhardt,
Traite de Chimie organique;

ff.)
cited
by Ladenburg,
op. cit.
p. 235.
(Cf. p. 194
216
process of chemical transformation. As soon as this phase is reached, 78 however, chemistry takes its place in the general plan of energetics, and thus passes from the circle of empirical descriptive sciences into that of mathematical science.
The reconstruction of the systematic form of chemistry. However, before this subordination of chemistry to a more general scientific
problem occurs, certain standpoints and tendencies appear within it, which point to this reconstruction in systematic form. The first phase of the determination of the material manifold is marked by the fact that every element is characterized by its atomic weight.
Thus every simple body

"characteristic
gains, in the expression of Leibniz, a definite
number;"
and
its
this
number
is
taken implicitly as
what brings the
wealth of
complete expression. in a manifold of numbers already indicates a new problem. As the real methodological advantage of the field of number is that each
empirical properties conceptually to This representation of the material manifold
member
in
it is
deduced and constructively developed from an
initial
according to unified rules, this demand henceforth is extended to all physical and chemical determinations, such as are known to be dependent on certain numerical values. They must no
structure
longer be conceived as a lawless aggregate; but represent them

in their sequence
it must be possible to and gradual transformation by an
exact law.
The periodic system of

first
the elements.
This general demand has
its
fulfillment in the establishment of the periodic
system
of the
elements.
The various
properties of simple bodies, their hardness
and
malleability, their fusibility and volatility, their conductibility for heat and electricity, etc., now appear as periodic functions of their
atomic weights.
series,
If
we
conceive
all
the elements arranged in a
advancing in the series, the properties of the various elements change from member to member, but that after
find that, in
we
going over a certain period, the same properties recur. The place of an element in this fundamental systematic series determines in detail its physico-chemical "nature." One of the founders of the periodic
On the "energetic" conception and treatment of chemistry, see esp. Ostwald, Elemente und Verbindungen, Faraday-Vorlesung, Leipzig, 1904; also Duhem, Le Mixte, Chs. IX and X.
78
217
system, Lothar Meyer, has clearly characterized the new principle involved in it. "Matter" is here removed from the field of scientific
constants into the
field of variables.
"Up
till
now
there have been
introduced into the calculations of physics as variable magnitudes, on which phenomena depend, place and time in particular, and,
under certain circumstances, heat, temperature, electricity, and a few other magnitudes; matter appeared, as expressed in magnitude and number, only as mass in the equations; its quality was only considered in that the constants for each kind of matter had a differ
ent value in the differential equations. To treat these magnitudes, which are dependent on the material nature of the substances, as
was hitherto not customary; but this advance has now been made. The influence of the nature of matter had previously been considered in physical phenomena by determining the physical
variables
constants for the most various substances.
But
this material nature
always remained something qualitative; the possibility was lacking of introducing this fundamental variable expressed in number and measure into the calculations. An approach to such an introduction,
although a very primitive one, has now been made by proving that the numerical value of the atomic weight is the variable, by which the substantial nature and the properties dependent on it are deter
The qualitative nature of the particular material is made mathematically conceivable by discovering a point of view from
mined."
79
which the material can be arranged in

progression. in that now
series
with a definite law of
The significance of this point of view appears especially members of the manifold, that were hitherto unknown empirically, can be demanded and predicted on the basis of the general systematic principle, and that advancing experience confirms
this
demand.
Chemistry and mathematics. The deductive element, that enters into chemistry, can be understood in its peculiar quality most
clearly,
with the ideal of deduction developed, on the one hand, in the speculative and metaphysical view of nature, and on the other hand, in mathematical physics. If we abstract from the part played by the problem of matter in the philosophy of nature,, the problem of matter has repeatedly had an important
it
when we compare
epistemologthe history of philosophy. Thus Locke, for instance, develops his whole view of the problems and limits of scientific
ical role in
79
Lothar Meyer, op.
cit.,
p. 176.
218
investigation in connection with the example of chemical knowledge For him, genuine of the fundamental elements and their properties.
is only attainable where it is possible to gain universal can speak of genuine knowl insights into necessary connections. all the properties of the object edge, in the strict sense, only where
knowledge
We
are certain and perfectly intelligible from
its original nature, thus with an object to conclude where it is possible from acquaintance This postulate, directly and to determine a priori all its properties. which is fulfilled in all our "intuitive" judgments regarding however, mathematical relations, is not satisfied in our scientific knowledge
Here, where we are merely concerned with the collec and description of various facts of perception, it remains forever impossible to establish that dependence of the individual members on each other, by which alone they could become a rationally connected
of nature.
tion
whole.
No
matter
how many
properties of a substance
we may
dis
cover by observation and investigation, the question as to their inner connection is not advanced a step. If we collect ever so many
properties of gold, its malleability, its hardness, its non-combusti bility, etc., still we cannot discover from them one single new deter
mination, and
we can never understand the form

Such an
of connection,
by
which
definite properties of
one kind always correspond to other

insight, as
definite properties of another kind.
would make
our knowledge of nature a genuine science like mathematics, would only be possible if, instead of merely collecting observations concern
ing the empirical coexistence or empirical incompatibility of proper the other end;" if we could start ties, we could grasp the problem
"at
from some sort of determination of the essence of gold, to deduce from

the totality of secondary properties. 80 Modern science has in part fulfilled the ideal, which Locke here abandons; but it had first to
it
give this ideal a new meaning. Modern science agrees with Locke, that the deduction of the particular properties of a material from
its "substantial
essence"
transcends the problems of exact and empiri
cal knowledge; but it does not thereby disclaim all conceptual con nection of the empirical data themselves. Science today collects the
plurality of elements into a fundamental series, whose members succeed each other according to a definite principle, and then de termine the individual properties of bodies as functions of their
position in this series.

80
How from the assumed fundamental property,

Understanding, Bk. IV, Ch.
6.
Locke, Essay on
Human
219
the further properties follow, how, from a definite atomic weight, there results a definite malleability and hardness, fusibility and
volatility,
remains indeed unanswered; nevertheless, the fact of this dependence itself is used in the attempt to calculate and predict
The certain special properties on the basis of certain special data. thus established contains, indeed, less than functional connection metaphysical insight into ultimate essences, but at the same time it
more than a mere empirical collection of disconnected particu The order of elements, that now arises, offers at least an lars. analogue of mathematics, and thereby an analogue of exact and "intui We penetrate no deeper into the absolute being tive" knowledge. of bodies by this means; but we grasp the rules of their systematic Yet at the same connection more definitely. (Cf. above p. 207 ff.) There arises the problem this solution leads to a new problem. time,
offers
of allowing the atomic weights, that were first introduced as discrete
and
values, to proceed from each other by continuous transformation, of determining the law according to which, in such a variation,
If we regard this problem entered a new form of con logically ceptual construction; instead of a number of rules concerning the concomitant appearance of properties, we would have henceforth a
the dependent properties must change.

as solved, then
we would have
unitary, mathematically representable law of causal dependence be tween the variations of different magnitudes. The atomic weights, by which we express the special character of the elements, would no longer stand next to each other as rigid, given values, but could be traced in The chemical concept would have their origin out of each other.
become the physical concept. The latest phase of natural science, which has resulted from consideration of the phenomena of radio
activity,
seems to attest such a change directly; for here science assumes a continuous transformation of the elements into each other, and for it the particular material with its sensuous definiteness is
only a transition-point in a dynamic process. When the chemical atom is resolved into a system of electrons, it loses the absolute fixity and immutability previously ascribed to it, and appears as a mere
as a cross-section, which thought makes in the relative resting-point continuous flow of process. However we may judge of the positive
truth of such assumptions, at any rate, they show very distinctly the way in which the scientific concept advances. Chemical research
begins with a plurality of actual observations, which at
first
stand
220
unconnected side by side; and these it defines in fixed determinations These numerical values gained through of number and measure. soon arranged into series; the series proceed according observation are to a rule, and the later members can be determined from the precedAs, however, empirical manifolds are transformed into rational ing. this way, there arises the problem of reducing the laws of struc in tural relations to a deeper-lying causal law of process, and of com In this progressive pletely grounding the former in the latter.
mastery of the empirical material, the peculiarity of the logical proc ess is revealed; through it, the concept, while obeying the facts, at the same time gains intellectual dominance over the facts.
IX
The concept
of natural science
and
"reality.
The
real
methodologi
cal interest of the construction of the concepts of chemistry lies in the fact that in it the relation of the universal to the particular The consideration of the physical concepts and is set in a new light.
methods allows only one
side of this
fundamental relation to appear
goal of theoretical physics is and remains the universal clearly. laws of process. The particular cases, in so far as they are taken into
The
Account, serve only as paradigms, in which these laws are represented and illustrated. The further this scientific problem is followed, the sharper the separation becomes between the system of our concepts and the system of the real. For all "reality" is offered to us in individual shape and form, and thus in a vast manifold of particular
features, while all conception, according to its function, turns aside from this concrete totality of particular features. Here is again re
vealed the antinomy that found
its first
striking expression in the
system of Aristotle.
knowledge seeks to be knowledge of the universal, and is only fulfilled in this goal; while true and original being does not belong to the universal, but to the individual sub
All
stances in the dynamic succession of their realization. The historical struggles, that took place regarding the Aristotelian system during the middle ages and far down into modern times, are for the most
part to be explained from this point of view. The conflict of "nomi nalism" and "realism" represents only a further development of the
problem, already latent in the first beginnings of the Aristotelian metaphysics and theory of knowledge.
221
In con Rickert s theory of the scientific construction of concepts. the opposition here referred to finds its temporary philosophy,
sharpest formulation in Rickert s theory of the scientific construction The direction of thought upon the "concept," and its of concepts.
direction
upon the
facts
real,
mutually exclude each other.

fulfills
For to the
field
extent that the concept progressively

perceptible
recedes.
its
task, the
of
The
simplification,
which conception
undertakes with regard to the intensive and extensive manifold of things, means a continuous impoverishment of its significance for
reality.
The
final goal of
is
the material sciences and of

intuition
all
other
from the content of Science does not bridge the gap between "thoughts" their concepts. but it is science, which first creates this gap and and
natural sciences
to
remove empirical
"facts,"
constantly increases it. "Whatever the content of the concept may be, it stands in the most decided opposition to the empirical world of The individual in the strict sense dis the intuitive
appears even in the most primitive construction of concepts, and natural science finally comes to the view, that all reality is always
and everywhere the same, and thus contains absolutely nothing But this is universally not the case; and as individual soon as we only consider that every bit of reality in its intuitive form is different from every other, and further that the particular, the intuitive and the individual, is the only reality that we know, then
the significance of the fact, that all conceptual construction annuls the individuality of reality, must come to mind. If nothing individ ual and intuitive enters into the content of scientific concepts, it
follows that nothing real enters into them.
The gap between the
concepts and the individuals, which is produced by natural science, 81 is thus a gap between concepts and reality in general."
then
If this logical consequence is justified, Criticism of[Rickert s theory. scientific investigation has hitherto been entangled in a strange For all the great exact and empiri self-deception regarding its goal.
cal investigators believed, and still believe, that the task of their science is to permeate the real more and more with knowledge, and
to raise it to more definite intuition. In place of an accidental and fragmentary consideration of things, which is different for each individual observer, a more perfect survey of them should be gained
;
81 Rickert, Die Grenzen der naturwissenschafllichen Begriffsbildung, Tubin gen und Leipzig 1902, p. 235 ff
.
222
and in place of the narrow naive picture of the world, a more compre hensive insight should be achieved, such as to reveal to us the finer structural relations of the real, while permitting us to trace them in
detail.
But how can
this
demand be
satisfied,
is
ment
it?
of investigation, the scientific concept,
if the logical instru in direct conflict with
We must now recognize, that what ought to sharpen our appre hension of the details of empirical intuition, dulls it that what seems to confirm and extend our knowledge of facts, rather separates us The con further and further from the real kernel of the "factual."
;
ceptual understanding of reality amounts to the annihilation of its Peculiar as this result may seem, it follows characteristic import.
from the premises of Rickert s theory. If the concept is, what the dominant logical doctrine holds it to be, nothing but a "presentation of what is common," then it is incapable of grasping
necessarily
the particular as particular. Its function then is not essentially different from that of the word, with which it is placed on a level by
Rickert,
in this follows Sigwart. As Sigwart explains, all that is presented either exists individually or in abstraction from the con
who
ditions of its individual existence, and in this case it is called univer sal, in so far as what is presented, as inwardly present, can be
thought of as existing in any group of individual things or cases.

expression for this inwardly present
of
The
meaning presented is For instance, there is no entirely definite intui tive content corresponding to the word but rather only a certain vague outline of form along with a vague presentation of wing movement, so that a child may call a flying beetle or butterfly a bird; the same is originally true of all our universal presentations. They are only possible because we have, along with the concrete and complete sense perceptions, also less perfect and definite contents of
is
what
the word as such.
"bird,"
consciousness. The indefiniteness of the memory-images of our actual sensations involves that, along with the vivid and immediately present, sensuous intuitions in the real process of consciousness,
pale residua of them are always found, which retain only one or another feature of them; and it is these latter, which contain the real psychological material for the construction of the universal presenta
tion. From this indefiniteness results the capacity of the presenta tion to be applied to what is different, not merely in space and time but in content: "the more indefinite, the easier the application."
The apparent
variety of the conceptual function,
its ability
to intro-
223
duce constantly new and more remote elements for comparison, thus rests rather on the poverty of the psychological substratum. The scientific concept arises in the same way and under the same
It differs from the naive concepts of language and the popular view of the world merely in that the procedure, that is there unconsciously effective, is here practised with critical awareness.
conditions.
The methods of natural abstraction, when left to themselves, are very involved, and never attain a complete, definite result; the achieve ment of science is to remove this ambiguity by establishing by uni versal definition certain rules for the selection of the perceptual
material.
The
various abstract forms thus gain an exact delimita
tion with respect to each other, since each of them comprehends a The essence of the concept is found in single group of properties.
this
content, signified by a definite word. living intuition of the particular facts
constancy and all-round differentiation of the presentational 82 But remoteness from the is even greater than before.
word-meanings, the concrete presentation of still stands in the background of consciousness, although it need not be explicitly clear; while in the case of the scientific concept, the purer it is, the more it frees itself from this final residuum of intuition. The scientific concept thus becomes a whole that can be completely surveyed and mastered by
in the case of the
For
the content they are to signify
it must fail to grasp and reproduce always present only in individual form. What first strikes us in \Yord-nieanings and mathematical concepts. this deduction is that it separates the scientific concept from the connection, in which it logically arises, and from which it continues to draw its real force. The exact scientific concepts only continue an intellectual process already effective in pure mathematical knowledge. Criticism of popular word-meanings does not affect these concepts, for from the beginning they stand on other ground and are rooted
thought, but, on the other hand,

reality,
which
is
in entirely different presuppositions. The theoretical concepts of natural science are in no sense merely purified and idealized wordmeanings; all of them have an element totally foreign to the word as
such.
As we have
seen, they always contain reference to
an exact
serial principle,
that enables us to connect the manifold of intuition

it
in a definite
82
way, and to run through

Logik," 1,
according to a prescribed law.

Cf. Rickert, op.
cit.,
See Sigwart,
45
ff
.,
I,
325, etc.
p. 32
ff.,
47
ff .
(See also Ch. I.)
224

in this sense, the
For the
"concept"
antinomy on which Rickert
founds his argument does not arise. Here no insuperable gap can arise between the "universal" and the "particular," for the univer
has no other meaning and purpose than to represent and to render possible the connection and order of the particular itself. If we regard the particular as a serial member and the universal as a
sal itself
\
serial principle, it is at
once clear that the two moments, without
going over into each other and in any way being confused, still It is not evident refer throughout in their function to each other.
that any concrete content must lose its particularity and intuitive character as soon as it is placed with other similar contents in various
and is in so far "conceptually" shaped. Rather the case; the further this shaping proceeds, and the more systems of relations the particular enters into, the more clearly its peculiar character is revealed. Every new standpoint (and the
serial connections,
the opposite
is
nothing but such a standpoint) permits a new aspect, a become manifest. Here logic again unites with the view of concrete science. In fact, every true scientific
concept
is
new
specific property, to
concept proves
its fruitfulness
"facts."
just
by pointing the way
to hitherto
unknown
In turning aside from the particular material of intuition, the scientific concept does not lose sight of it completely, but always shows us a direction which, if followed further, teaches us new peculiarities in the manifold of intuition. Thus if the chemical "concept" of a certain body is given by its constitutionfields of
formula, in which
istic structure,
it
it is is
chemical
"types,"
grasped as a particular material in its character same time brought under the various and is thus set in a definite relation to the totality
at the
The ordinary chemical formula only gives us the composition in general, but not the type of construction of the individual elements; here it is enriched by a wealth of new relations.
of remaining bodies.
The
general rule now in our possession enables us to trace how and according to what law the given material is transformed into another;
the rule involves not only the form of the existence of the material at a definite moment, but the totality of its possible spatial and
temporal phases.
proceeds, the
The further the chemical construction of concepts more sharply the particulars can be distinguished. Materials, which were called similar, because "isomeric," from the standpoint of the undeveloped concept, are clearly separated and distinctly defined in character, from the standpoint of the developed

concept.
225
of
Thus we never meet here the vague
"universality"
popular word-meanings. The particular serial member, whose place in the system is to be determined, can be retained throughout never theless, its relation with other members of a group possesses a sharply
;
denned meaning, by which

relation.
it is
The
individual in
its
peculiarity
distinguished from other forms of is threatened only by the
universality of the blurred generic image, while the universality of a definite law of relation confirms this peculiarity and makes it known
on
all sides.
Rickert s
confusion
of
"meanings"
and
"presentations .
Thus
Rick~ert7s criticism ultimately concerns only a
that he himself recognizes as unsatisfactory. tual construction comes under the "theory
form of trie concept This form of concep

of
83
subsumption,"
which
nevertheless rejected as far as the foundation of the exact is concerned. If we recognize, with Rickert, that all the thing-concepts of natural science have a tendency to be transformed
is
concepts
more and more
into relation-concepts,
it is
thereby implicitly admitted
that the real logical value of the concept is not connected with the form of abstract "universality." science can only be valuable
"A
with regard to knowledge of the whole of the corporeal world," Rickert says, it has before it in the first beginnings of the con
"if
struction of its concepts the final goal of all natural science, insight into the necessity of things according to natural law. If a science
has this goal before it, it will always seek to abandon the purely classificatory construction of concepts as soon as possible; i.e., it will
never be satisfied with concepts, that are mere complexes of properties, but any collection of such elements into a concept occurs only on the assumption, that the elements either stand directly in a necessary
connection according to natural law (i.e., unconditionally universal connection), or at least represent the preliminary stages of such
concepts,
law
is
which a necessary connection according to natural expressed. The relation of the world of meanings to the world
in
of perceptions assuredly constitutes our knowledge, as least in so far as we are concerned with knowledge in the sense of the natural
sciences; but for that very reason the meanings cannot be presenta tions, but must be judgments in their logical value, and must either
Cf. the pertinent critical remarks of M. Frischeisen-Kohler, Die Grenzen der naturwissenshaftlichenBegriffsbildung, Arch. f. system. Philosophic, XII, (1906), p. 225 ff.
83
226
84 In connection with contain laws or prepare the way for them." this clear explanation, the critical point in Rickert s theory can soon be pointed out the center of the problem is falsely shifted from the
;
necessity of conceptual
presentations.
"meanings"
to the universality of the generic
Only of "presentations" can it be said, that the more general they become the more they lose their intuitive sharpness and clarity, until they are finally reduced to mere schemas without Judgments, on the contrary, determine the significance for reality. individual the more exactly the wider the sphere of comparison and correlation to which they relate it. Increase of extension is here parallel with determination of the content. (See above Ch. I.)
The
judgment does not signify the quantity of the judgment, but the quality of the judgmental connection, so that judgments concerning individuals can be completely universal.
universality of a
S is P, in this case, does not signify that the property contained in a plurality of subjects, but that it belongs uniformly to this particular subject unconditionally and with objective necessity.
The
proposition
P is
When we
tific
conceive the given of sensation as
made
necessary by scien
content, but merely
laws, we thereby change nothing in its material represent it from a new standpoint. A whole
"individual" thing does not pass over into a whole "universal" thing; but a relatively loose aggregate of empirical determinations is united into a system of
objectively valid connections.
peculiar kind of object

is
is
not
given a new categori cal form. The transition to "universality" is thus a secondary aspect, which does not concern the real tendency of the construction of concepts. In so far as it enters, the transition to "universality"
produced, but the very same empirical reality
is
only a
is
which
posited and
symptom and expression of that transition to necessity, demanded by the problem of scientific knowledge
p. 71, 73.
itself. 85
84 85
Op.
tit.,
an indirect confirmation of this view in the latest exposition of the theory of Rickert, which is contained in the writing of Sergius Hessen, "hidividuelle Kausalitat" (Ergdnzungshefte der Kantstudien, Nr. 15, Berlin 1909). In order to reveal clearly the opposition between the scientific and the histori cal construction of concepts, Hessen distinguishes two different forms of causality. The causality, which natural science affirms and makes the basis of its explanations, can be reduced to the idea of universal lawfulness. Ac cording to this view, to conceive an event causally means to subsume it under general laws what is known in this way is thus never grasped in its absolutely
I find
;
227
True, in one
The
concept as the ex^ressionôf individual -relations.
respect the separation between the scientific concept and the "reality" given us in sensuous impressions remains. None of the fundamental
concepts of natural science can be pointed out as parts of sensuous perceptions, and thus verified by an immediately corresponding
impression.
scientific
has become increasingly evident that, the more thought extends its dominion, the more it is forced to
It
intellectual
conceptions
that possess no analogue in the field
of
unique and unrepeatable character, but only as an example of a general con cept. The content of the idea of causality in general is, however, not ex hausted by this one-sided scientific schema. For causality ultimately signifies nothing else than the "necessity in the temporal sequence of the parts of reality;" we must postulate such a necessity also where we are concerned with the succession of purely individual events, which can thus never recur in precisely the same way. The specifically "historical causality" is founded on the application of this point of view; its concept arises, as soon as we insert the idea of necessity and determinateness into a unique, temporally deter mined process, without attempting to conceive it as a special case of universal
Here it appears that there (Cf. Hessenop. cil. esp.p. 32ff.,p. 73 ff., etc.) an inclusive unity for the scientific and the historical "concept," from which both are deduced; and this unity is constituted by the idea of necessity. Hessen himself ascribed this necessity at first to the "objective reality," which as such is to be conceived as free in principle from every form of concep tual interpretation, whether in the direction of the scientific or the historical
laws.
is
concepts.
more exact epistemological analysis shows, however, that
this
reality is not takei* in the sense of an absolute metaphysical existence, but as a regulative idea, which guides our diverse, methodically separate concep
common goal. (Cf. esp. p. 88 ff.) In other words, it thus appears that the methodological distinction of the "universal" concepts of natural science from the "individual" concepts of history does not exclude a connection between the two, but rather requires it; what is logically distinct from the standpoint of "universality" tends to coincide, when we exchange this standpoint with that of necessity. If we hold to the latter thought as the truly original and decisive idea, it is further clear that the distinction in degree of "universality" can never become an unconditional opposition. In so far as we apply the idea of necessity to a particular temporal occurrence, thus in so far as we assert that this individual
tions toward a
necessarily demands and draws after it this individual B, we implicitly assume an element of universality even in this establishment of a, unique state of affairs. For in this judgment, the case is excluded of the total complex A ever recurring in precisely the same character; but at the same time it is asserted that, if A were repeated in this fashion, then B and only B would be demanded as real. Whoever sees more in history than a mere positivistic "descrip tion" of the sequence of various events, whoever grants it a particular form of
228
concrete sensations. Not only the hypothetical concepts, like the atom and the ether, but purely empirical concepts, like matter and motion, show that scientific investigation along with the
"given"
elements of perception cannot do without the purely ideal limiting scientific concepts not given in direct experience; along with the (See above p. 120 ff.) investigation must have the "not-real."
"real,"
Nevertheless
it
owing to
this
would be a mistake to assume that exact science, characteristic feature of its concepts, withdraws
has already
causal judgment,
"universal."
recognized in history this form of the universality does not belong to the categorical but to the hypothetical part of the assertion; the form of connection of A and B is ideally projected into universality, although the particular elements possess only a
The
concept, which seeks to grasp this reality, form of necessity, just as, on the other hand, the exact concept of natural science, which is meant in the first place to be an expression of a universal connection, seeks verification and application in the
unique reality.
The
historical
indirectly refers to the universal
temporally limited particular case.

reference of the
tion of the
"particular"
to the
"universal"
Only, in the two, the direction of the is different, while the correla
two moments is seen to be necessary in both cases. Thus we are not concerned here with an opposition between the "con and absolute "reality," but with a distinction wholly within the cept" system of concepts. Hessen himself emphasizes this fact, and therewith the conceptual character also belonging to history. "The opposite opinion, which makes history a perceptual science and connects it with reality, is guilty of an historical concept-realism as dangerous as that of natural science." The historical concepts are general products of a more or less intense abstrac tion," and are thus as little perceptual as the concepts of the natural sciences. an individualizing science of civilization, history implies a removal from
"in
"As
reality;
it
stands in principle as close to reality as the natural sciences;
it
also
works with concepts, and indeed with individual concepts. This must be especially emphasized against historical concept-realism" (p. 27 ff.). Here it appears from another angle, that the separation of the concepts of natural science from those of history, presupposes a certain connection between the two. The conceptual function as such must be understood and derived in its unitary form before the differentiation into various types of concepts can begin. This fundamental form, however, is not found in the generic con cept, but in the serial concept, which is unavoidable for any sort of
of the perceptually given. An essential task of the historical con the insertion of the individual into an inclusive systematic connection, such as has constantly established itself more distinctly as the real goal of the
"shaping"
cepts
is
scientific construction of concepts. This insertion" can occur under differ ent points of view and according to different motives; nevertheless it has com mon logical features, which can be defined and isolated as the essence of
"the"
concept.

more and more from the tasks
existence.
229
Precisely in this
is
of things, science
by concrete empirical apparent turning away from the reality For those directed upon them in a new way.
offered
very concepts, that have no direct intuitive content, have a neces sary function in the shaping and construction of intuitive reality. The determinations expressed by the scientific concepts are not
perceptible properties of the empirical objects, like their color or taste; but, on the other hand, they are relations of these empirical
objects.
The judgments thus formed, although they cannot be
resolved in their content into mere aggregates of sense impressions, are nevertheless related in their use to the totality of these impres
which they seek to give systematic form. The methodologi becomes metaphysical; for thought only separates itself from intuition in order to return to it with new inde
sions, to
cal opposition thus never
pendent instruments, thereby to enrich it in itself. Every relation, which theory has discovered and given mathematical form, indicates a new way from the given to the not yet given, from real to "possible"
It is thus true indeed that the relational concepts of experience. natural science have no immediate copy in the individual things;
but what they lack
is
not so
rather the thing-like character.
much the element of individuality as They render insight into relations
possible, and guarantee it, although they themselves can never be perceived after the fashion of isolated objects. Thus energy, for example, does not signify a homogeneous thing, in which all inner
differences of the different types of energy are cancelled, but it is a unitary principle of connection, that as such can only be verified in the qualitatively different. Identity of serial form is concealed
behind every assumption of identical objects in natural science; and such identity is only found in the manifold of serial members, which must be retained as such. There is thus no contradiction between
the universal validity of principles and the particular existence of They things, because there is ultimately no rivalry between the two. belong to different logical dimensions thus neither can seek directly
;
to take the place of the other. _ The problem oftiie constants of natural
science^.
sharper conception" matics. 86 Reference

88
whn
we
15ring it
back to the
The problem gains field of mathe
hafe justly
been made,
of
in opposition to Rickert s
(cf. above and emphasized from the
The
I.)
"concrete
universality"
the mathematical concepts
Ch.
has also incidentally been recognized
230
facts of
theory, to the significant rdle due to the establishment of certain magnitude, to the establishment of certain numerical con
87 stants in the structure of natural science.
It is only
when the values
of these constants are inserted in the formulae of these general laws,
that the manifold of experiences gains that fixed and definite struc The scientific construction of reality ture, that makes it "nature."
is
only completed
when
there are found, along with the general
causal equations, definite, empirically established, quantitative values for particular groups of processes: as, for example, when the general principle of the conservation of energy is supplemented by
giving the fixed equivalence-numbers, in accordance with which the exchange of energy takes place between two different fields. As
Robert Mayer has said, these numbers are the desired foundation of an exact investigation of nature. 88 The definite number breaks through the traditional logical schema, which recognizes the concept
only as a generic concept and as comprehending a plurality of or the does not exist as a examples under it. The
"two"
"four"
standpoint of Rickert. "The gap for conceptual knowledge between the uni versal and the particular," says Lask in his work, Fichtes Idealismus und die Geschichte, "and the consequent irrationality is bridged in the mathemati cal view through the possibility of construction. The individual cases real izing the mathematical concept can be generated by the concept itself. From the concept of the circle, we can attain by construction the mathematical individuality of the particular circle, and thus go from the universal to the individual in its individuality In mathematics, also, the intui tive object is an individual, concrete and given object; but it is given a priori, not a posteriori like the material of sensation it is a logical unique! something
We
individual, but at the same time capable of being construed a priori" (p. 40 f.). see here also that Rickert s criticism would have taken another form
if he had conceived the concepts of natural science decisively and from the beginning as products of constructive mathematical procedure, rather than as results of "abstractive" procedure. The insight once gained for mathe matics would have had to be transferred to physics for precisely here lies the real problem that mathematics is no "logical unique," but that it progres sively provides the "special" natural sciences with its own characteristic form of concept. The form of mathematical "deduction" is already contained in the form of physical "induction," by which we grasp the empirically real, and thus the same methodic mastery of the particular by the universal is achieved.
;
(Cf. esp.
87 I, 6, 88
Ch.
5.)
Cf. Riehl, Logik und Erkenntnistheorie (Die Kultur der Gegenwart, p. 101 f .) ; cf. esp. Frischeisen- Kohler, op. cit. p. 255.
(1851) (Die
R. Mayer, Bemerkungen iiber das mechanische Aquivalent der W&rme Mechanik der Wdrme, p. 237).

genus, that
is
231
twos or fours of objects, but it is and occurs only once, although there can be no doubt that it possesses no sensuous, but a purely conceptual "being." (See Ch. II.) On the one hand, it appears that the scientific concept is in no way denied the establishment of the individual; although, on the other hand, it never grasps the individual as isolated, but only as a particular element in an ordered
realized in all concrete
a fixed
member
in the series of unities
Instead of ascending to abstract and empty genera of being and process, investigation seeks to connect the empirical
manifold.
constants, which it discovers and which are represented by definite 89 number-individualities, into series according to necessary laws. The essential object of scientific consideration is the "structural
relations,"
along with the laws of causal dependence.. These struc tural relations are finally reduced to definite numbers, as the example
these
knowledge shows, and the attempt is made to understand numbers as an ordered sequence. Theory considers and defines the possible forms of serial connection in general, while experience shows the definite place, taken by an empirical being or an empirically real process in this connection. In the
of chemical
"real"
conception of the world, the two elements are The universality oL^he" functional rule is only united. inseparably in the particular numericalTconstants, and the particu represented
developed
scientific
larity of the
numerical constants
is
only represented in the universal
This reciprocal relation is ity of a law mutually connecting them. also repeated and confirmed within the special sciences. No natural
science renounces the establishment of particular facts; nor can it establish them without the decisive concurrence of the idea of law.
Even
those
who
start
from the
opjxr!^tT6n,.oj:j[&e^h^
_the individual and the scientific^ generip/concept have to agree ex pressly, that this intellectual division corresponds to no real separa
tion in the sciences
themselves.
Everywhere the two motives
only by the dominance of one or the other that the position of a particular science in the general system of knowledge can be ascertained. But if this is the case, then it is
it is
interpenetrate; and
questionable by what right we can characterize a type of problem and the treatment of it by the name of one science, when the problem
89 Cf. here especially A. Gorland, Aristoteles und Kant beziiglich der Idee der theoretischen Erkenntnis untersucht. Giessen, 1909. (Philos. Arbeiten hrg. von Cohen u. Natorp II, 2), p. 433 ff.
232
is
V
^\
If we collect under the shared by the most various disciplines. all those scientific procedures, generic concept of the "historical" even so it is by which are directed on the gaining of pure
"facts,"
no means shown that the concept thus produced represents a true methodological unity. For the establishment of facts occurs in the different special sciences under very different conditions. JThe
unieral theory of the special discipline is always- necessarily pre supposed, and also gives the judgment of fact its definitafo^jg^ Thus
every astronomical
"factum"
involves in
its
formulation the whole
conceptual apparatus of celestial mechanics, and further the funda mental doctrines of optics, in fact aJLihe essential parts of theoretical
(See above p. 142 ff.) |The "historical" part of each physics. science is methodologically connected with its "theoretical" part by
a genuine inner dependence, while between the descriptive parts of two different disciplines, on the other hand, there subsists only a
The unity here is not one of principle, but is merely classificatory. The procedure, by which astronomy gains its facts, is conceptually connected with the procedure, by which it
loose connection.
constructs
its general theoretical conceptions; but it is sharply and distinguished from the way in which, for example, biology definitely determines and selects its empirical material. Here also it proves to
be impossible to divide our knowledge in such a way that, on the one side, there should be purely the universal, and on the other side,
purely the particular. Only the relations of the two moments, only the function fulfilled by the universal in connection with the particular,
gives a true ground of division.
Magnitudes and other forms of relations. That this function is completed in none of its activities, that beyond every solution a new problem arises, is indeed indubitable. Here, in fact, the "individual" But it reality shows its fundamental character of inexhaustibility. is the characteristic merit of the true scientific relational concepts,
that they attempt this task in spite of the impossibility in principle
of its completion.
with the preceding, forms a
Each new construction, since it is connected new step in the determination of being
and process. The individual, as an infinitely distant point, deter mines the direction of knowledge. This last and highest goal points indeed beyond the circle of scientific concepts and methods. The
"individual"
of natural science includes
and exhausts neither the
individual of aesthetic consideration nor the ethical personalities,

which are the subjects of history.
233
For all particularity in natural science reduces to the discovery of definite magnitudes and relations of magnitudes, while the peculiarity and value of the object of
artistic consideration
and
of ethical
judgment
lies
outside
its field of
vision.
But
this delimitation of the different
methods
produces no dualistic opposition between them. The natural science does not deny the object of ethics and aesthetics, although it cannot construct this object with the means at its disposal;
it
of judging concept of
does not falsify intuition, although
it
from one dominant standpoint, and

mination.
selects
consciously regards intuition one particular form of deter
other types of consideration, which go beyond natural science, are not in contradiction with it, but in a relation of intellect ual supplementation. They also do not approach the individual as
The
a separated and isolated element, but they produce new and signifi cant points of connection. It is a new teleological order of the real, that is added to the mere quantitative order, and in which the individ?ual
is
first gains its full meaning. Logically speaking, the individual taken up and shaped by different forms of relation. The con flict of the "universal" and the "particular" resolves into a system of complementary conditions, such as only in their totality and co operation can grasp the problem of the real.
PART
II
CHAPTER V
ON THE PROBLEM
OF INDUCTION
The metaphysical tendency in induction and deduction. The real knowledge is to deprive the opposition of the universal and the particular of its metaphysical character. The law and the fact appear no longer as two eternally sundered poles of knowledge; but they stand in living, functional connection, related to each other as means and end. There is no empirical law, which is not concerned with the connection of the given and with inferring not-given groups of facts; as, on the
result of the methodological analysis of scientific
other hand, each

cal law,
"fact"
is
and receives
its definite
established with reference to a hypotheti character through this reference.
Empirical natural science, since it first entered "the sure course of a science," has itself taken no considerable part in the struggle of the
philosophical
"deduction."
parties
concerning the rights of
"induction"
and
empirical science examines its own procedure, it has to recognize that there is in this struggle a false and technical separation of ways of knowing that are alike indispensable to its
When
very existence.
is
peculiar to all metaphysics of knowledge appears and acts in the process of knowledge as an inseparable unity of conditions is hypostatized on the meta physical view into a conflict of things. Permanence and change,
The motive
here revealed.
What
being and becoming, unity and plurality,

partial aspects of certain
all of
fundamental ways
of
which signify only knowing, appear in
unconditional opposition. Thus in the philosophy of nature, there is a metaphysics of the particular along with the metaphysics of the universal. While in this latter, concepts expressing the necessary
connection of experiences are raised to independent realities, in the former, the simple sensation in its individual character is made the
bearer and content of true reality.
The
real content of existence,
which
sions
resists
every analysis,
is
sought only in the isolated impres
Qualitative properties. Advancing intellectual serves only to bring out more clearly this fundamental exist insight ence, and to resolve all assertions regarding being into it more
their
and
completely.
237
238
The empirical theory of judgment. If this demand is to be satisfied, the motive of particularization must be sharply and clearly carried through. All our judgments can mean nothing else than the estab lishment of a state of fact, given here and now, and conceived merely An assertion reaching in its spatial and temporal particularity. The validity fall into the field of mere fiction. beyond this would
claimed by any true judgment must be strictly limited to the moment of making the judgment; for as the perception, as a real process, does not go beyond this instant of time, so the concept must recognize
Present these natural limits, if it is not to lose its definite character. and past sensations constitute the kernel of all our judgments, both
the rational judgments and judgments of fact. Indeed the element of past sensations already threatens to break through the general for consciousness in the same schema; for the past does not
"exist"
sense as that in which the concept of reality
is
here taken.
When we
compare a temporally present impression with others, that have occupied consciousness at an earlier point of time, we have already This into the "not-given." taken the first step from the be taken without danger, in so far as it is assumed, that the step may remembered perception is like the actual one in all essential parts.
"given"
The past
is
conceived as present, in spite of
its
temporal remoteness,
and with all the definiteness of the immediate impression. Judg ment rests solely on the comparison of actual and reproduced con
tents of perception.
Mach
s "thought-experiment."
Consistent
"empiricism"
is
obliged
to extend this consequence to all fields of knowledge. Mathematics and physics, physics and biology are, from this point of view, equiva lent; for it is not the analysis of the object, but the psychological
analysis of the act of judgment, which has led to this explanation. The form of judgment must be the same everywhere, because the
presentational material, on which its form exclusively rests, is always the same for the different disciplines of knowledge. The method of
observation and of investigation is independent of whether we experiment with things themselves, or with our presentations and
memories
of things. may cite an instance from Mach. If, for example, the geometrical problem is proposed of inscribing a square in a right-angled triangle, having the two sides a and b, and the
We
hypotenuse
right angle,
c,
while one corner of the square
is
to coincide with the

lie
and the three other corners are to
on the
sides a, b
ON THE PBOBLEM OF INDUCTION

and
c;
239
then thought must submit the given conditions to an investiga
tion, in order to find a solution of the problem.
Now
if
we conceive
an arbitrary distance to be measured off from the vertex of the right angle on one of the two legs, and the corresponding square to be
constructed, then the corner of this square will not in general
fall
on the hypotenuse, but to the right or left of it, outside or inside the surface of the triangle. Between these two possible cases, as is
further shown, there is a continuous transition, since by continually increasing the originally chosen distance, we can move its end-point
from within the triangle to a position outside the triangle. Experi ence directly shows that this movement cannot take place except by crossing the hypotenuse once, as the line which divides the two parts of the plane of the triangle, and thus marking a point on it, which
"Such a tentative represents the point demanded by the problem. division of the field of presentations, in which we (tatonnierende)
have to seek the solution of the problem, naturally precedes its Popular thought may be satisfied with a practically perfect solution. Science demands the most general, sufficient, approximate solution. shortest and most comprehensible solution. We obtain this, when we remember that all inscribed squares have the line which bisects the angle at the intersection of a and 6 in common as a diagonal. Consequently if we draw this bisecting line from this known point, we can complete the required square from its point of intersection with c.
Simple as this purposely chosen example is, it brings out clearly what is essential in all solution of problems, experimen 1 tation with thoughts, with memories Criticism of Mack s theory. However, even this example reveals a
"
....
latent presupposition at the basis of the whole argument. "Memory" in the strict psychological sense can produce no new content; it can
only repeat what has been offered by sensuous presentation. It can thus recall into consciousness those cases, that have been intuitively
presented to us, but
it is
incomprehensible
how
it
can venture any
assertion regarding a whole group of forms, without having run
through
the
particular
examples
possibility is excluded
by the nature
Yet this latter individually. of the problem in the case
mentioned; the number of possible squares is infinite, and is thus The absolutely inexhaustible for concrete sensuous imagination. judgment of memory as such can never survey an infinite group of
1
Mach, Erkenntnis und Irrtum,
p. 39 f .
240
many
However possible cases, but only a limited number of actual cases. points of the line bisecting the angle we may have investigated,
decide whether the next point that
we can never
we
select will
show the same characteristics as those previously observed, if we commit ourselves solely to the described method of experimentation with presentations and memories. From this point of view, there is nothing to hinder the assumption that in further advance there might be found points of the line bisecting the angle, which do not satisfy the assumed condition; or that conversely, there are points that fulfill
the condition without belonging to this line. The solution first gains the character of necessity and determinateness, when we go back from the particular example to the process of construction,
erties.
which the bisecting line arises and it gains its mathematical prop In becoming aware of this unitary rule of construction, we thereby grasp the totality of determinations of the complete struc
in
ture; for these determinations arise only by virtue of the generating law and can be deduced from it in all strictness. do not proceed
We
here from a plurality of particular cases to the connecting law, but from the unity of the geometrical procedure to the particularities of application. Only in this way is a relation posited, that affirms not
only the present presentation, as

ideal connection.
found in consciousness, but a permanent presupposition is established which is meant to hold not of this or that individual triangle with its It is no particular properties, but of "the triangle" absolutely. matter whether this claim can be finally justified; certainly as a mere psychological phenomenon, it breaks through the scheme of
it is
knowledge of the consistent sensationalistic view.

empiricism"
Thus those thinkers, who proclaim the postulate of "radical most decidedly within psychology, are obliged even
this standpoint to recognize the logical
from
and methodological
"pure
difference here.
The unprejudiced
verdict of
experience"
is
opposed on
dogmatic deductions of sensationalism. Unprejudiced analysis of the facts of knowledge shows clearly that to reduce mathematical and logical relations to assertions regarding
the frequent empirical coexistence of particular presentations is a vain endeavor. Mathematical and logical relations do not report
this point to the
whether and how often certain empirical contents have been found in existence in space and time, but rather establish a necessary con nection between ideal structures, the validity of which is to remain
ON THE PROBLEM OF INDUCTION

unaffected
241
all changes in the world of existing sensuous objects. a logical or mathematical proposition as a mere repro interpret duction of particular, actual "impressions" and their empirical
by
To
meaning of the proposition in the an interpretation ascribes to the a sense, which, by the nature of the subject to which it proposition No metaphysical construction refers, it neither does nor can possess. can set aside the psychological and logical phenomenon of this differ ence; "relations between ideas" are separated in principle from purely factual determinations of the coexistence and sequence of particular
attempt to discover its origin; such
empirical properties.
2
relations, is to falsify the real
The more sharply this theory of empirical judgment. is carried through, the more strongly, on the other hand, separation the character of the empirical judgment is revealed. The character
Locke
s
of this latter seems to consist in nothing else than a conscious limita tion of the validity of the judgmental connection to the temporal
moment
of
two kinds
of truths
making the judgment. In this sense, the was already grasped by Locke.
relation of the
According to
him, the validity of mathematical knowledge rests on the principle of the immutability of the same relations between the same intel
What is proved of one triangle can be carried over without further mediation to all triangles; for a particular intuitive presentation of the triangle does not stand for itself in the proof, but is only meant to be an accidentally selected sensuous image for a
lectual objects.
universal and permanent relation. This insight is denied in all the judgments, which go beyond the field of our intellectual presenta
tions to the existence of things. External things manifest themselves to consciousness in no other way than in the sensuous impressions
they arouse in us; their certainty can thus be of no other kind than that of these impressions themselves. The existence of sensation reaches no further than its immediate presence. Once this is removed
we
all
is
lose
our only criterion for the existence of things and the basis of
assertions regarding the
more exact properties
of this existence
taken from us. Consequently, judgments concerning the exist ence of things have only relative and limited truth; for, however
Genuine psychological "empiricism" maintains this separation through is brought out with especial clarity by James in his polemic against Spencer and Mill on this point. (The Principles of Psychology, London 1901,
2
out; this
esp. Vol II, 645, 654, 661, etc.)
242
convincing and evident they may seem as long as the direct sensation is given, we have no assurance that the momentary witness of
sensation will ever be repeated in exactly the same way. Accordingly there is necessary knowledge only of such objects as, like the objects of pure mathematics, renounce all concrete reality; while at the very
moment
is
this reality is
taken into account, the character of knowledge
completely changed.
in all empirical judgment. Although this from the abstract standpoint of theory of illuminating knowledge, it offers a difficult problem when we consider it in con nection with the concrete procedure of natural science. Locke s account, which has been frequently repeated since with slight varia
"element
The
of
eternity"
distinction
is
might appear as a correct account of what the purely empirical and inductive propositions of natural science ought to be; but it No judgment of certainly does not touch what they are in reality. natural science is limited to establishing what sensuous impressions are found in the consciousness of an individual observer at a definite,
tions,
strictly limited point of time.
this, If there are judgments that speak of they are the narrative judgments of psychology, and not the
and descriptive judgments of general natural science. As the mathematician, who treats of the relations between geometri
theoretical
cal forms or between pure numbers, permits nothing in his state ments regarding the properties of the particular presentations, in which he sensuously represents these relations, so also the investiga tor, who gives out the results of an experimental research, constantly
What
goes beyond a simple report of his particular perceptual experiences. he establishes is not the sequence and play of certain sense
impressions, that have occurred in him and again disappeared into nothing, but the constant "properties" of constant things and proc esses. In this advance from the mere process of sensation to definite
assertions, the metaphysical concept of "transcendence" indeed wholly absent. The change, that makes possible the judg ment of natural science, only transforms the data of sense into a new mode of being in so far as it imprints upon sense-data a new form This element of knowledge can be separated out and of knowledge.
"objective"
is
all further metaphysical assertions that be added. It is, first of all, a new sort of temporal validity, may which is now ascribed to the judgment. Even the simplest judgment concerning any empirical matter of fact ascribes to the latter an
retained independently of

existence
243
and a permanence, which the fleeting sense-experience as such cannot establish. The proposition, that sulphur melts at a definite temperature, that water freezes at a definite temperature,
means even
in its simplest terms, and abstracting from the various theoretical assumptions contained in the mere concept of "tempera ture," something that is to be limited to no isolated temporal moment.
embraced by the subject are realized, the consequences expressed in the predicate concept
It asserts that, as often as the conditions
concept will be always and necessarily connected with them. For thought, the moment of immediate perception is extended to the whole course of time, which is surveyed in its totality at one glance.
which gives each experiment its peculiar as proof. Each scientific conclusion based on an experi significance ment rests on the latent assumption, that what is found to be valid
It
is
this logical function,
here and now, remains valid for all places and all times, in so far as the conditions of investigation are unchanged. Only through this principle are the "subjective" facts of sense-perception changed into the seen
for
"objective"
facts of scientific judgment.
From a new
side, it is
how
it is
far, as Goethe said, all that is factual is already theory; only the thought of the necessary determinateness of phe
nomena, that leads us to arrest a single transitory observation, and establish it as a fact. The postulate of necessary determinateness. Even investigators, who think they stand exclusively on the ground of empirical and who repudiate all independence of the intellect with respect to the data of immediate perception, have expressly attested to this In spite of all supposed scepticism this one intellectual function. conviction is occasionally expressed. "The relations between differ ent phenomena," says Ostwald, "which have once been known, remain as indestructible parts of all future science. It can, and indeed often does, happen that the form in which such relations were first expressed is seen to be imperfect; it is seen that the rela tions are not to be maintained as wholly universal, but are subject to other influences, which alter them, of which no thought could be taken at their discovery and first formulation, since they were then unknown. But however science may be transformed, a definite indestructible residuum of that first knowledge remains, and a truth once gained by science has an eternal life in this sense, i.e., it exists as
"facts"
long as
3
human
science will
exist."
This element of
15.
"eternity"
is
Ostwald, Grundriss der Naturphilosophie, p.
244
also
No characteristic of empirical judgments concerning facts. connection between observations, once it has been established objec tively, can be absolutely annulled in the further course of investiga
facts, which we discover, do not displace the earlier in every sense, but only add to them a definite concep experiences And this change really affects not so much the tual determination.
tion.
The new
judgmental connection, as rather the subject to which it refers. For example, if we conceive a certain material as determined by the scheme of its physical and chemical properties, then no matter what
contrary instances may arise in the course of advancing observation, or what change may appear in its action, the previously affirmed connection in the determinations of the material is in no way set If the empirical judgment were bound to a single moment of aside.
time, then a simple relation of annihilation
prevail here; the later
moment would
with
it
all
the
"truths,"
and re-creation would earlier, and which were indeed only established and
set aside the
asserted of the earlier.
Just as, in the real course of phenomena,
the later
replaces the earlier, so also it would involve an But in truth, each inner change in the empirical laws of the thing. body has an identical structure and character, which we ascribe to
it once for all. We never express divergent results by assuming that one and the same body has changed in its fundamental proper ties, but by calling the identity of the observed object into question. What we now see standing before us is not the same empirical object,
moment
as
was previously offered; but we consider it as modified by certain Thus the conditions, that are to be discovered and established. truth of the earlier judgment is P" is not invalidated by its opposite
"S "S
is
not
P,"
but, while maintaining the
first
proposition,
we
trace the
transformation, which the judgment must undergo, when S passes into S The advance of observation thus involves a continuous
.
advance in analysis; it distinguishes with increasing exactness cases which on first vague consideration appear wholly similar and reveals the characteristic differences of each individual case. If we conceive this work of analysis as complete and a wholly determinate subject therewith gained, then the determinateness of this subject will also involve in it the determinateness and necessity of the judgmental
connection.
The element
of uncertainty, contained
by the empirical
judgment sumption
in contrast to the rational, thus concerns only the subof the given under an ideal case. The question is not

whether to a
strictly defined content
245
a the predicate 6 belongs or

.
not, but whether a given content satisfies all the conditions of the The concept a, or is to be determined by a different concept a
problem is not whether a is truly b, but whether x, which is offered by mere perception, is truly a. Here lies the real superiority of the mathematical construction of concepts; for the objects, of mathematical construction are only
what ou^Jde^xonstmctioBk makes them,
while every empirical content conceals unknown determinations in itself, and therefore we can never decide with full certainty under
which of our previously developed hypothetical concepts

brought.
it is
to be
Judgments of perception and judgments of experience. Locke s analysis of the empirical judgment is thus seen to be intrinsically inadequate; for it conceals that element of necessary connection,
which
gives
real
is
them
also characteristic of assertions concerning facts, and which their real meaning. For Kant, this necessity was the
fundamental problem; yet in first introducing his critical ques he still appears dependent in one point upon Locke. The dis tinction between judgments of perception and judgments of experi ence, to which Kant appeals, has not so much a direct, systematic, as a didactic meaning; it joins on to the sensualistic conception of judgment, so as to gain from it a new meaning and a deeper interpre tation. Empirical judgments, in so far as they have objective valid ity, are to be called judgments of experience; while those, that are only subjectively valid, are to be called judgments of perception.
tion,
The latter concept thus includes everything that dogmatic empiricism regards as the genuine mark and character of experience. The "judgment of perception" at least is nothing else, and is intended to
be nothing
experience;
else,
it
than a report concerning a momentary and individual does not connect subject and predicate according to
any standpoint of intellectual dependence and coherence, but only takes them as they are accidentally found together in an individual consciousness, according to the "subjective" rules of association.
In the judgment of perception, we only establish the coexistence of two contents, without setting them in any relation of mutual depend ence. The further the Kantian distinction proceeds, however, the more it appears that the judgment of perception is only meant to
be a methodologically constructed limiting-case, to throw light upon the newly gained concept of scientific objectivity by force of con-.
246
but that the judgments into two a certain measure sphere, no matter

distinction carries with
it
^.trast;
heterogeneous classes. of objectivity within
no real separation of Every judgment claims

self-chosen narrower
its
limited its subject-concept. It is never satisfied with establishing a mere coexistence of presentations, but it erects a functional coordination between them, so that whenever
how
the one content
is given, the other is taken as required. The of the copula is the expression of this connection, and thus enters as a necessary factor into every assertion regarding an individual, empiri
"is"
The proposition, that body is heavy, does not mean cal object. that, as often as I have hitherto lifted a body, certain touch and
pressure sensations have been felt, but it is meant to establish a connection, based in the object and independent of the condition of
this or that sensing individual.
"a
Even the
1
individual
"a
posteriori"
judgment always contains an priori" element in the necessity of the connection, which it affirms. In the final conception of the system of experience, the instrumental concept of the mere judg ment of perception is transcended and excluded. Indeed the indi
vidual as individual can be the object of a scientific statement; hence a state of being, which is given here and now, constitutes the con
tent of judgment.
field of
But in this case also, we do not step outside the objective necessity into that of mere "contingency;" on the contrary, we seek to conceive the particular itself as necessary, by
giving
it
The sphere
a fixed place within the causal process ruled by exact laws. of the necessary is narrowed until it is adequate for
"contingent."
closer determination of the apparently
For example,
we determine the astronomical
position of the heavenly bodies for a of time in this sense, by taking as a basis the universal given point relations offered by the principles of mechanics, as well as by the law
of gravitation.
The
real goal of
"induction"
is
not the absolutely
isolated, temporal fact (Setzung} as such, but the subordination of this fact to the whole process of nature.
Experience as aggregate and as system. often referred to, does not begin where
The "secret of induction," we draw a conclusion from
several observations regarding all the cases, but is already fully contained in the establishment of any individual case. The solution
of the
its
4
import.
problem of induction can only be sought in this broadening of In fact, it cannot be understood how the mere repetition
2nd
ed., p. 141 f
.
Cf. Kritik der reinen Vernunft,

and arrangement
lar a
247
new
logical value.
of particular observations should lend the particu The mere accumulation of elements c_annot
entirely change their conceptual
meaning
it
distinct the determinations already contained in the elements.
can only render more There
must be a factor concealed in the individual case, that raises it out of its limitation and isolation. The function, that constitutes the real kernel of inductive procedure, is that by means of which we trace an empirical content beyond its given temporal limits and
retain
it
in its determinate character for all points of the time series.
The
ible
relation,
which only manifests

until in
itself
at
first for
single, indivis
moment, grows
its content is transferred to the totality of times, and is continued through this totality as if by constant, identical re-creation. The permanent empirical object along with
future points of tune. element of infinity, in so far as
some way it determines the totality of Thus each individual judgment involves an
its constant empirical properties is, in mathematical language, always the integral of the momentary properties, of which the indi vidual enquiry gives evidence. But the logical process of integration would not be possible, if there were not a reference to the whole already in the element, i.e., if the varying content of experience, no matter how scattered and detached it may appear, did not always
involve reference to
its
permanent laws
(gesetzliche
Form).
It is
through this reference that our limited, spatio-temporal experience, which is all we have, becomes the test and image of the system of reality in general. It is only by conceiving all phenomena
_ajL connectedJpynecessary relations,
circle of
that
we can use any
individual
j)hase as a representation and symbol of the total process and of its universal rules. It is this symbolic meaning, which every inductive
inference claims for itself; the particular determination offered by the sensuous impression becomes a norm, that has to be retained as a permanent feature in the intellectual structure of empirical reality.
Each
particular experience, that has been established according to the objective methods and criteria of science, claims to be absolute; what methodically tested experiment has once shown, can never be
entirely logically annulled.
The task of induction is to unify these various assertions, which frequently seem to cross and contradict
each other, by indicating a definite sphere of validity for each of them. What, to the ordinary sensuous view, is an identical group of conditions connected now with one, now with another circum-

stance,
is
here separated into definitely distinguished,
particular
cases; these cases vary in some theoretically-discoverable circum stance, and their variation is thus conceived as necessary.
Discrete
and continuous
"wholes."
In the relation of the inductive
individual ealseTftythe totality of scientific experience, a characteris tic recurs/ which can always be established where the problem is to
define a ("wholey/^hat is not merely the sum of its parts, but is a systematic totality arising from their relations. Logic traditionally
distinguishes between
first,
"discrete"
and
"continuous"
wholes.
In the
the parts precede the whole, and, independently of the con nection into which they subsequently enter, are possible and dis-_ The "element" of the continuum, tinguishable as independent pieces.
on the contrary, is opposed to any such separation; it gains its content only from relations to the totality of the system, to which it belongs, and apart from it loses all meaning. Thus a line can be defined as an infinite manifold of points; but this definition is only possible, because the itself has been previously conceived as the expres sion of a pure relation of position, and thus the relation or spatial
"point"
"coexistence"
with other similar elements
is
contained in
it .A
In the
~"*^*i"^
same sense, it can be said that the law of experience only results" from the particular cases because it is already tacitly assumed in them. The individual empirical judgment contains within it, as an
undeveloped demand, the thought of the thorough-going determinateness of natural processes as a final result of the completed
system of experience.
Every assertion regarding a mere coexistence
of empirical determinations points to the thought that these deter minations are somehow grounded in each other, although the form of
is not directly known, but is only to be gained pro Just as the relational character of position and distance gressively. inheres in the individual point, so the character of a universal law
this
dependence
inheres in the individual experience.
The
individual
cannot be
experienced save in connection with other spatial and temporal, near or remote elements; and this kind of connection presupposes a system of spatial and temporal positions, as well as a unitary whole of causal coordinations. The fact a is only accessible to us in a
functional form as
different
f (a),
<J>
(/3),
\f/
(7), in
which
f,
<,
</-,
represent the
forms of spatial-temporal and causal connection. The logical act of "integration," which enters into in every truly induc tive judgment, thus contains no paradox and no inner difficulty; the
249
advance from the individual to the whole, involved here, is possible because the reference to the whole is from the first not excluded but retained, and only needs to be brought separately into conceptual
The tendency to something the theory of invariants. to something permanent in the coming and going of unchanging, sensuous phenomena, is thus characteristic of inductive thought no
possible to trace in the development of geometrical methods how the manifold tendencies of modern geometrical thought can be unified, in so far
is
prominence. Induction and
than of mathematical thought. The two do not It goal, but in their means for reaching this goal.
less
differ in their
as they are brought under the general standpoint of the "theory of invariants" and characterized by it. Each special form of geometry
then coordinated with a definite group of transformations as its appropriate theory of invariants, and these can be strictly defined
is
and
set over against
each other.
The conceptions
of constancy
and
change are seen to be mutually conditioned by each other; the permanent connections affirmed by geometry can only be formulated
This with reference to possible changes. (See above p. 88 ff.) fundamental logical relation now appears in a new light. Each assertion concerning an empirical connection of elements is meant to be valid independently of any absolute point of space or time.
Maxwell incidentally gives the general "causal law" a turn that The proposition, that like causes always expresses this demand. bring forth like effects, he explains, has no sharply defined meaning, until we establish what is to be understood by like causes and like effects. As each occurrence only takes place once, and is thus fully individualized by the time of its occurrence and distinguished from all others, the likeness here in question cannot be meant in the sense of
absolute identity, but only relatively to a definite point of view, such as needs to be expressly isolated and formulated. The real kernel of the causal principle lies in the assertion that, when the
causes are merely distinguished from each other with reference to absolute space and absolute time, the same is true of the effects:
"the difference between two occurrences does not depend on the pure differences of the times or places, in and at which they take place, but only on differences in the nature, configuration or move ment of the bodies involved." 5 Here it becomes clear, that the
Maxwell, Matter and Motion, Art. XIX.
250
content, on which the physical judgment is directed, is at first sub jected in thought to a certain change, and the judgment separates out those moments unaffected by this change, that can be constantly
affirmed.
certain form
of
its
its
we characterize as the geometrical properties of a those properties, that belong to it independently absolute position in space and of the absolute magnitude of
Just as
all
parts, so here
an analogous mode
of consideration
f
,
with regard to time.

.
A
.
functional relation
and removed from dependency on any individual moment of observation. So far as the remaining conditions are unchanged, any arbitrary moment
,
directly established for one point of tune t discrete times ti, t 2 t 3 ., is freed from this limitation
is applied that is only (a, b), or for a plurality of
of time
may
the present
moment
be taken as equivalent to that first given to us, so that involves a decision for the past and the future.
"invariant"
Thus
tions,
#11
and
experience is directed on gaining certain first in these reaches its real conclusion.
rela
The conception
of the empirical natural object originates and is grounded in this procedure; for it belongs to the concept of this object, that it remains
"identical
with
itself"
in the flow of tune.
We
must indeed con
ceive each natural object as subject in principle to certain physical changes, called forth by external forces; but the reaction to these
forces could not be represented in the form of law, if we were not able to recognize the object as logically permanent and provided with
the same properties.

tions,
we produce
fixed connections
In the midst of the temporal chaos of sensa and coordinations by abstracting
from time; and it is these fixed connections, which constitute the fundamental frame-work of empirical factuality. Induction and analogy. Thus it is always a function of judgment,
that assures us of the permanence of empirical being. This fact finds its expression not only in the inductions of mathematical
is clearly shown within descriptive natural science. Deeper analysis reveals here also how far the apparently purely receptive classification of particulars is dominated by ideal pre Claude suppositions, referring to the structure of the whole. Bernard especially has thrown light on this reciprocal relation of idea and observation for physiology and the whole field of The order of the and factual "experimental medicine." cannot be discovered without an ideal standpoint of comparison,
physics, but also
"real"
without a conceptual anticipation of a possible order.
Although
it

is
251
to experience that we owe the definitive establishment of this order, beyond doubt, yet though there again must work out the scheme of
experience beforehand. The induction of the descriptive sciences is thus always a "provisional deduction." "We can, if we will, call
the tentative thought of the experimental investigator inductive, we characterize the apodictic assertions of the mathematician as deduction; but this is only a difference affecting the certainty or
while
manner
course
uncertainty of the starting-point of our inferences, and not the in which these inferences themselves proceed." The prin
ciple of inference
remains the same in the two cases, although its run in two directions: "there is for the mind only one way of concluding, as there is for the body only one way of going." 6 This unity of principle is especially clear in those limiting fields, where mathematical thought comes in contact, as it were, with experimental investigation. We saw how the progress of geometrical thought tends more and more to allow the subordination
may
The real object of geometrical interest is seen to be only the relational connection between the elements as such, and not the individual properties of
of the particular intuitive figures in the proof.
Manifolds, which are absolutely dissimilar for can be brought to unity in so far as they offer examples and expressions of the same rules of connection. The conceptual construction of exact physics proves to be dominated by a corre sponding logical procedure. Hitherto the analogical inference has
these
elements.
intuition,
been considered an essential part of the method of physics, and especially of the inductive procedure; no less a thinker than Kepler commends it as his truest guide and preceptor, from which no secret
of nature
is
concealed. 7
But the
scientific
value of analogy remains
incomprehensible, as long as one merely bases it on a sensuous simi Indeed the precise task of larity between the individual cases.
theoretical physics, by which it is distinguished from the naive view, Is to separate cases, that appear in direct perception as similar
and analogous, by penetrating analytically further into the conditions

6
Claude Bernard, Introduction a V etude de

ff.
la
medicine experimentale, Paris,
1865, esp. p. 83
7
Kepler, Paralipomena in Vitellionem, Ch. IV, 4 (Op. II, 187). concept of analogy, cf. Mach, Erkenntnis und Irrtum, p. 218 ff.
On
the
252
of their origin. 8
The genuine and
truly frnjf,ful..ajmlngy:
is
on a sensuous agreement
in relational structure.
of properties,
but on a conceptual agreement
When we
sort, in
consider electricity and light as
phenomena
of the
same
the electro-magnetic theory of light,
based on an agreement capable of being grasped but on the form of the equations, which we establish in by perception, both as a quantitative expression of the phenomena, as well as on the relations between the numerical constants characteristic of the two fields. (See above p. 163). The comparison thus does not rest on a mere indefinite similarity, but upon a true identity in the mathe
this assertion is not
matical system of conditions; as in pure mathematics, this identity Thus is isolated as a logical "invariant" and considered for itself.
lar,
which at first seems concerned with the sensuous particu goes over more and more into mathematical "harmony;" the doctrine of Kepler offers the classical example of this. "Analogy"
"analogy,"
passes into a view of the unitary quantitative structural laws which, according to the assumptions of exact physics, rule the whole of being, and thus unify what is most diverse.
ii
The
Induction and analysis, "compositive" and "resolutive" methods. first result claimed by the inductive "concept," in the strict
:
8 Here is an example from Duhem, La Theorie Physique, p. 32 f "La Phy sique expeYimentale nous fournit les lois toutes ensemble et, pour ainsi dire, sur un meme plan. Bien souvent, ce sont des causes tout accidentelles, des analogies toutes superficielles qui ont conduit des observateurs a rappro.
. .
loi. Newton a fixe dans un de lumiere qui traverse un prisme et lea lois des teintes dont se pare une bulle de savon simplement parce que des couleurs 6clatantes signalent aux yeux ces deux sortes de ph6nomenes. La thôrie, au contraire, en deVeloppant les ramifications nombreuses du raisonnement d^ductif qui relie les principes aux lois experimentales, 6tablit parmi celles-ci un ordre et une classification; il en est qu elle r6unit, 6troitement serrees, dans un meme groupe; il en est qu elle s6pare les unes des autres et qu elle place en deux groupes extremement 61oign6s. Ainsi, pres des
cher, dans leurs recherches,
une
loi
d une autre
meme ouvrage les lois de
la dispersion
lois
qui regissent le spectre fourni par un prisme, elle range les lois auxquelles ob6issentles couleurs de l arc-en-ciel; mais les lois selon lesquelles se succedent
les teintes
lois
Newton vont, en une autre region, rejoindre les de couvertes par Young et par Fresnel. Les lois de tous ces phenomenes que leurs eclatantes couleurs confondaient les uns avec les autres aux yeux du simple observateur, sont, par les soins du theoricien,
des anneaux de des
f ranges
. .
classees et
ordonnees."
253
sense of this word, is that it transforms the manifold of observations, which at first appears as a mere unrelated coexistence of particular can realize the meaning elements, into some fixed serial form.
We
of this task
by referring to certain elementary problems of arithmetic, which form an exact example and analogue of the logical relation here
is given, which are connected unknown, then in order to discover this rule, the given sequence must be resolved into a complex of series, that obey relatively simpler laws. If we have before us, for example,
involved.
If
any sequence
of
numbers
according to a definite rule at
first
the sequence of fourth powers 1, 16, 81, 256, 625, then we can estab lish the relation connecting the particular serial members, by finding first the differences between them, and further the differences between
. . .
these differences
series
a simple arithmetical individual members. Thus we are led back to a completely known serial type, and the way is indicated, by which from this fundamental form we can again
etc.,
until
we
finally reach
with a constant difference between
its
reach the given series through continually more complicated steps. This given series is now plainly revealed to us in the conditions of
its
structure,
by reducing
as
it
in its
in all the particular phases of its construction; to the stages of its construction, the series becomes, were, transparent, and shows the same character of necessity advance from member to member, as was shown by the primi
it
and
tive series.
field of
The same
is
"resolutive"
method, that
is
here used in the
numbers,
first
characteristic of the true scientific

fact,
inductive
inference.
The given
as offered
if
by immediate observation,
It
appears at
to thought as
impenetrable.
can be simply
established, but not determined according to its simple beginnings, But genuine, i.e., not deduced from identical rules of progress.
theoretically guided induction is never satisfied merely with estab It replaces the factual coexistence of sensuous lishing facts as given.
data by another kind of connection, which indeed seems poorer in elements, when considered purely materially, but which can be more
clearly
surveyed according to the principle of

institute,
its
construction.
Every experiment we inferences, works in
and on which we base our inductive
this direction. The real object of scientific investigation is never the raw material of sensuous perception; in place of this, science substitutes a system of conditions constructed
and defined by
itself.
254
Experiment as the means of analysis. Strictly speaking, the experiment never concerns the real case, as it lies before us here and now in all the wealth of its particular determinations, but the experiThe ent rather concerns an ideal case, which we substitute for it.
eal beginnings of scientific induction furnish the classical example Galileo did not discover the law of falling bodies by collect of this.
ing arbitrary observations of sensuously real bodies, but by defining hypothetically the concept of uniform acceleration and taking it as a
9 This concept provides for the conceptual measure of the facts. time-values a series of space-values, such as proceed according given Henceforth we to a fixed rule, that can be grasped once for all.
must attempt to advance to the actual process

sive consideration of the
excluded
as,
by a progres that were originally complex determinations, for example, the variation of acceleration according to
of reality
the distance from the centre of the earth, retardation by the resist ance of the air, etc. As in the arithmetical example, we advance from
the simple series, in which the differences between the members are constant, to series of the second, third and fourth order; so we analyse the real into various orders of relations, connected according to law
The sensuous appear ance of simplicity in the phenomenon yields to a strictly conceptual system of superior and subordinate relations. However, as opposed
and progressively conditioning each other.
to the mathematical concept, there is the characteristic difference that, while in mathematics the construction reaches a fixed end, in
experience
it is
how many
"strata"
in principle incapable of completion. of relations we may superimpose
But no matter on each other,
and however
close
we may come
to all particular circumstances of
the real process, nevertheless there is always the possibility that some cooperative factor in the total result has not been calculated,
and
will
analysis.
only be discovered with the further progress of experimental Each result established has thus only the relative value of
a preliminary determination; and as such only holds what is gained in order to use it as a starting-point for new determinations. The
uncertainty, that seems to remain, does not affect the relations within the particular series, but only appears where the whole theoretical
construction
diction here
9
is compared with the actual observations. A contra would not be resolved by our giving up the principles of
Of. more particularly Erkenntnisproblem I, 294; also Honigswald, Beitrdge zur Erkenntnistheorie u. Methodenlehre, Leipzig, 1906.
255
the earlier investigations, but by our adding to these investigations new factors, such as correct the first result yet enable us to retain it in a new meaning. In so far as we abstract from subjective errors
of observation, the truth of the individual determinations remains in general unaffected. What is always being questioned anew is
only the sufficiency of these determinations for explaining the com But the inductive concept plicated factual relations of reality.
its
proves, precisely by the fact that it leaves this question open, that direction does not lead away from reality, but more closely to it.
The
"universal relations," which are at first brought out, do not indeed contain the special properties in themselves, but neither do they deny them. They rather leave room for them from the begin
Galileo s ning, and anticipate their future possible determination. law of falling bodies needs no correction of its content, in order to
represent the
ance of
air,
phenomenon of falling in the case of a definite resist but merely requires a conceptual extension already
"particular"
admitted and foreseen in principle. The re_îonof "universal" and

"sense,
relations.
In this
tneôncept^
of natural science in
ho way abstracts from the
but brings out the particular all the more sharply. Each universal relation, that it establishes, has a tendency to connect itself with other relations, and by this combination to become more and
particular,
more
useful in the mastery of the individual.
Each
of the
fundamen
from which a more complex law is built, signifies in itself indeed only a certain sphere of conditions. However, it cannot be
tal series,
said, that these conditions are only partially or inexactly fulfilled by the concrete process, which results from their totality; rather they must all be fulfilled perfectly and without limitation, if the given
to be possible. Thus the subsequent addition of a new stand which places the phenomenon under consideration in relation to a new circle of facts, changes nothing in the meaning and value of the earlier determinations. Only one thing is demanded, viz., case
is
point,
that the relations progressively established in this

patible with each other.
way
is
shall
be
com
assured in prin ciple by the fact, that the determination of the particular case takes place on the basis of the determination of the general case, and
this compatibility
And
tacitty
assumes the validity of the latter. If we conceive a particular process as a synthesis of various laws, the question as to how the particular can "participate" in the universal ceases to be a metaphys-
256
ical

problem; for the universal
is no longer something substantially which somehow enters as a factual part into the individual, present, but it is a logical moment, implicitly assumed in a more inclusive
system of
rence A,
relations.
We
it
.
by
inserting
determine a particular natural occur into various functional connections f (A, B,
.),
(A,
B C
,
.)
(A,
all
B",
C"
.) etc.,
and conceive
it
as
subordinated to the rules of

tion"
these connections.
The
"participa
of the individual in the universal thus appears
no more of a
riddle than the logical fact that, in general, various conditions can be united intellectually into a unitary result, in which each is fully
retained.
stance"
The question now
is
not
how
is
the individual in
its
"sub
separated from it, but how it is possible for knowledge to bring the rules of universal connections into relation, and to determine them reciprocally by each other, so
issues
from the universal and
that conceptual insight into the particular relations of physical

(Cf. above p. 226 ff.) and "superposition." The real problem of induction lies here, as is shown in many places in the theory of knowledge of modern natural science. The two tendencies of scientific thought, which Galileo contrasted as the "resolutive" and the "compositive" methods, are sometimes distinguished in modern discussion as the 10 The first goal principles of "isolation" and of "superposition." of experimental enquiry is to gain a pure phenomenon, that is, to free the phenomenon under consideration from all accidental circum stances. While reality presents us a mixture of heterogeneous circum stances seemingly inseparably interwoven and confused, thought de mands the separate consideration of each particular moment and the
reality results.
"Isolation"
exact determination of the part it plays in the structure of the whole. But this goal can be achieved only through a technical separation of
what
is
factually connected, only
by
which enable us to trace the individual factors for themselves

action.
establishing special conditions, in their
It is only when this separation is strictly carried through, that the constructive unification is seen in its clarity and sharpness.
The complete
picture of the total process arises when we connect the This partial system and place them as it were upon each other. of the total process appears not merely as a unitary intuition, picture
but as a differentiated conceptual whole, in which the type of de10
See Volkmann, Erkenntnistheoretische Grundzuge der Naturwissenschaft,
Leipzig, 1896.
257
)endency between the individual elements is exactly defined. If the problem of physical induction is conceived in this way, it appears anew that the mathematical point of view is not so much the opposite
as the necessary correlate of inductive conceptual construction. For that very synthesis of relations, which is required and which forms
an
essential part of experimental enquiry, has its ultimate abstract foundation and the guarantee of its general validity in the system of mathematics. As we have seen, the object of mathematics in the
universality of its concept is not the compounding of magnitudes, but the connection and reciprocal determination of relations. (See above p. 95 ff.) Thus the two scientific tendencies here have a
full
point in
common.
Though experiment
is
necessary to analyse an
into
its
originally undifferentiated perceptual whole constitutive elements, on the other hand, to
particular
mathematical theory belongs the determination of the form by which these elements are combined into a unity of law. The system of "possible" relational syntheses already developed in mathematics affords the fundamental
schema
for the connections,
which thought seeks
in the material of
to which of the possible relational connections are actually realized in experience, experiment, in its result, gives its
the real.
As
answer.
answer can only be given when the question has previously been clearly stated and this process of stating the ques tion goes back to conceptions, which analyse immediate intuition
But
this
according to conceptual standpoints. If the real is represented as the result of the interpenetration of elementary series of dependencies, then in principle it has gained the form of a mathematically deter-
minable structure. 11
]1 As I subsequently discover, this relation has received new confirmation and an unusually clear presentation from a modern physicist. Bouasse, Physique generate, in the collective volume De la Methode dans les sciences,
"La physique ne separe pas 1 etude des formes de 1 etude des fails; deduction prevoit les faits que 1 experience confirme." "Qu est-ce done qu expliquer? C est tout uniquement faire rentrer un fail dans une forme. Le fait est explique lorsqu il apparait identique a un des phenomenes qu engendre La un des ces sorites indefmis que nous appelons theories ou formes. physique n est done pas mathematique, parce qu on y trouve des algorithmes algebriques; toute experience devant, en definitive, entrer dans une forme,
Paris 1909:
la
...
toute forme se developpant naturellement sous les symboles mathematiques, toute physique est mathematique (op. cit. p. 76 f., 91, 100).
258

principle of
"isolation"
Thus when the
and
"superposition"
is
explained by the fact that all reality represents only the sum of the manifestations of particular laws of nature, and is to be conceived as having issued from these laws, the real epistemological meaning of
We cannot be concerned here with the but only with the origin and character of our insight as it is grasped in the sensuous impression, The into things. of various elements, but first appears is not in and for itself a to us as an absolutely simple and unanalysed whole. This original of intuition is only transformed into a manifold by the "simplicity"
the thought
is
concealed. 12
origin of things,
"real,"
"sum"
work of the concept. The concept is here just as source of plurality as it elsewhere appears to be a source of Since we conceive a particular process as successively intro unity. duced into different systems, whose general structure can be mathe
logically analytic
much a
matically deduced,
far as
we thereby
increase its determinateness, in so
define its position in the general plan of our thought with The advance of experiment goes hand in hand increasing exactness.
we
with the advancing universality of the fundamental laws, by which
we explain and construct empirical Laws and rules. The methodic

indicated,
reality.
which subsists between

"laws"
difference has occasionally been the mere "rules" of nature and
the truly universal
of nature.
The inductions
"rules"
of Kepler
on
planetary motion express only generalized the fundamental law, on which they rest, was
of process, while first formulated in
Newton
13 In Newton s theory we find the theory of gravitation. ellipse not merely as a real form of the orbit of Mars, but we survey at a glance the whole of "possible" orbits. The Newtonian concept of a centripetal force, that diminishes according to the square of the
distance, leads to a perfect disjunction of the empirical cases in The transition between these cases is henceforth exactly general.
predetermined the magnitude of the initial velocity of a moving body decides (independently of the direction of this velocity) whether the form of its path is to be an ellipse or a hyperbola or a parabola.
;
Thus the
which
it
"law"
of gravitation contains in itself the field of facts,

strict division to its field; while
rules,
and ascribes
empirical rule of planetary stand in loose conjunction

12
13
the merely motion allows the particular cases to
without sharp delimitation.
In the
SeeVolkmann, op. cit., p. 89. See Volkmann, op. cit., p. 59.
ON THE PKOBLEM OF INDUCTION
259
actual progress of science, however, the two views thus logically distinguishable are never strictly separated, but blend imperceptibly into each other. The already contains a tendency to raise
"rule"
itself
to the form of law; while on the other hand, the conceptual per
which the law achieves, remains a mere provisional assump it always contains a hypothetical element. Here we face the apparent circle, that everywhere confronts us in the relation of law and fact. If we conceive the motion of the planets to be deter mined by gravitational forces, which work in the inverse square of the distance, then it is evident that the form of a conic section is necessary for their path; however, it cannot be established that this determination holds with regard to the kind and magnitude of attraction except through this methodic necessity just referred to, i.e., by the power of this assumption to unify observations and to If experience give them definite meaning. (Cf. above p. 146 ff.)
fection,
tion, since
"really"
forced us to forego this assumption hereafter by adducing new mate rial, then the pure concept as such could not prevent it; but in this
case also, the form of the empirical concept would by no means be submerged along with its particular content. require the same
We
field, revealed, and seek to estab The law, to which the earlier is subordinated. changed material conditions a changed manner of connection, while the general function of this connection, viz., the deduction of the
intellectual limit for the
new
that
is
lish this limit in
new
from a supreme serial principle, remains the same. This very function, and not its temporary and varying concrete expression in particular theorems, is assumed in the concept of
particulars
experience
itself;
and thus belongs to the
real
"conditions
.
of its
possibility."
When any series of observations ai
a2 as.
an
is
given,
a double problem is forthwith offered for consideration. On the one hand, we can seek to enrich the material of this series by inter polation arid extrapolation, by inserting between the given members
hypothetical intermediate members, or by tracing the series beyond its original limits. On the other hand, our concern is to unite the
manifold of members into an ultimate identity, by providing a rule, by which the transition from ai to a 2 from a 2 to a 3 etc. is determined
,
and subjected to a
then
it is
If we may call the first procedure fixed principle. preeminently that of "induction" and the second that of "deduction,"
other.
evident that the two mutually cooperate and condition each The completion of the series by the introduction of new
260
individual
follows the direction of a unitary law of deduction, which hovers as a problem before thought. The selec tion and sifting of the material is under the direction of an active
members always
standard of judgment. We seek to trace a law of known conceptual structure through a given sequence of observations, and we measure the truth of this law by its success in characterizing the places left
vacant by immediate perception, and in anticipating their filling-in In this sense, Kepler successively connected .by future investigation. the facts regarding the positions of Mars, left by Tycho de Brahe, by the most diverse geometrical curves and with these curves as known ideal norms, he compared the facts, until he finally reached the ellipse as that line, which enabled him to deduce the greatest number of
;
observations from the relatively simplest geometrical principle of Due to the nature of the problem, this work never progression. reached an absolute limit; for no matter how many points of the
path of the planet might be given,

to connect
plicated
it is
nevertheless always possible
them by any number of lines of different and more com form. Only the methodic requirement remains permanent,
that in the last analysis the processes of nature be reduced to definite
simple rules, no matter how necessary complex assumptions may prove for the representation of a limited concrete field of facts. This reduction to simple rules is comparable to the manner in which we
gradually reduce arithmetical series of any higher order to the funda mental type of a series with a constant difference between its members. The concept of the "fundamental" relation and the relation of mathe matical necessity. This reduction of the manifold and ceaselessly changing material of perception to ultimate constant relations must be granted without limitation by even the most radical "empiricism." For the assumption of this fundamental relation is all that remains for empiricism of the concept of the "object," and thus of the con cept of nature. body," Mach says, "looks differently in each illumination, gives a different optical image in each position, gives a different tactual image with each temperature, etc. However, all these sensuous elements so hang together, that with the same position, illumination and temperature, the same images recur. It is thus a
"A
permanency
here.
If
of the connection of sensuous elements, that
is
involved
we could measure all the sensuous elements, then we would say that the body subsists in the realization of certain equations, that hold between the sensuous elements. Even where we cannot

measure,
261
the expression may be retained symbolically. These equations or relations are thus the really permanent." The logical development of natural science tends more and more to a recognition
at
that the original, nai ve representations of matter are superfluous; most we grant them the value of pictorial representations, and
recognize the quantitative relations, that prevail between phenomena, as what is truly substantial in them. "To the extent that the
conditions of a
is
phenomenon
reduced.
We
what
are known, the impression of materiality recognize the relations between condition and
field,
conditioned, the equations which rule a larger or smaller

constituting
is
as
permanent and substantial, and as con the discovery of which makes possible a stable pic stituting that, ture of the world." 14 So far modern empiricism is in complete agreement with the critical interpretation of the meaning and pro The Critique of Pure Reason gressive tendency of natural science. that all that we know of matter are mere relations; clearly taught but among these relations there are some that are independent and 15 permanent, by which a definite object is given us. Opposition sets in, however, when this concept of permanence, to which the concept of the object is reduced, is to be determined more closely in its logical meaning and logical origin. Is the permanence a property
really
of the sensuous impressions, immediately inhering in them, or is it the result of intellectual work, by which we gradually transform the
given according to definite logical requirements? In the light of our earlier developments, the answer cannot be doubtful. Permanence
is
never found ready made in the sensuous experience as such, for the sensuous experience gives only a conglomeration of the most diverse impressions, limited to a single moment of time and never
recurring in exactly the same way. Permanence only appears to the extent that we are able to transform the sensuous manifold into a
mathematical manifold, i.e., in so far as we let it issue from certain fundamental elements according to rules held as unchangeable. The kind of certainty, that belongs to these rules, is clearly distin guished from the certainty of a particular sensation. Even from the standpoint of a mere "phenomenology" of the facts of conscious ness, it is something entirely different whether the various contents of consciousness merely follow each other factually, or whether the
14
15
Mach, Die Prinzipien der Warmelehre, Leipzig Kritik der reinen Vernunft, 2nd ed., p. 341.
1896, p. 422
ff.
262
succeeding ones are
known
"from"
their predecessors according to
some pervading logical principle. Leibniz illustrated this difference incidentally by pointing to the example of the theory of numbers, which sharply marks the general relation involved here. If we take the series of squares as given, for example, then by making numerous
we can establish purely empirically the fact that the differences between the individual members can be represented by the progres ... On the basis of sive series of the odd numbers 1, 3, 5, 7 this fact, we may expect that when we proceed from the last given member of the series of squares, and add the corresponding odd number to this member, that a square will result; but nothing justifies
tests,
us in taking this psychological expectation as the same as a logical
No matter how many members have been tested and necessity. found to correspond to the rule, it always remains possible that at a certain point the previously constant type of progress may be broken No accumulation of observations regarding particular numbers, no matter how great, could ever enable us to reach a new form of This certainty, that would give us more assurance in this regard. form of certainty is to be gained when we proceed from the "universal" member of the series, i.e., from its identical law of construction, rather than from the enumeration of its particular members. The formula (n n 2 = 2n 1 shows at one stroke, 1Y and without the necessity of several tests, the constant and necessary relation, that subsists between the progression of the squares and that of the odd numbers. When once it is grasped, this formula holds of any arbitrary n, since in its deduction and proof no account is taken of the particularity of any definite number, and thus the particular value can vary arbitrarily without affecting the meaning of the proof itself. The totality of squares and that of the odd numbers are now taken up into one system, in which the one is known through the other, while hitherto, no matter how far we traced their reciprocal correspondence, the two were merely in conjunction. This same fundamental difference can also be shown in the case of every true physical law. If we consider a law, such as Galileo s law of falling bodies or the law of Mariott, we discover that the correlated values of space and time, of pressure and volume, are not simply registered in conjunction, but are taken as conditioned by each other. So far as the mere facts are concerned, it is true that a
off.
list
of numerical values
might accomplish purely materially
all
that
263
the mathematical functional rule could ever give us, provided that
for each particular pressure-value the corresponding volume-value was noted in the list, and for each time-value the corresponding
space-value. And yet such an accumulation of particular numerical data would lack precisely the characteristic element, in which rests the meaning of law. For the decisive element would be gone; the type of determination by which one magnitude is conceived to issue
left in darkness, even if the result were cor In the quantitative equation, this type of determina tion is clearly evident; for the quantitative equation shows by what purely algebraic operations from universal, established rules, the value of the dependent variable is to be gained from the value of the
from another, would be

rectly stated.
argument. And to this mathematical connection, physical theory adds a corresponding, objective and causal connection. Here the values of the function, along with those of the independent variables, belong to a common system of causes and effects, of conditions and conditioned; and they are thus connected with each other in such a way, that the assumption of one draws the others necessarily after it. Here also we do not place the particular values of the one series simply in conjunction with those of the other; but we seek, at least hypothetically, to grasp the two series in their law of construction and thus in the totality of their possible determinations, and to compare them with each other. Methodically guided induction tends to this
goal; it represents what experience knows only as a complex coexist ence of members, as being a resultant of simpler series of dependency, and these for their part progress according to the strict relation of the mathematical "ground" to the mathematical "consequence." The two fundamental types of knowledge. This application of the
concept of ground and consequence

physical implications.
"description,"
is
obviously free from
all
meta
Here again,
in opposition to the theory of this
it is
necessary to maintain the purely logical character
of the relation of
ground and consequence, without making
character
cal
To "describe" a ontological. (Cf. above p. 137 f.) natural process in quantitative equations, in the sense of mathemati
phenomenalism, means also to
"explain"
it
in the only scientifi
cally admissable sense; for the equation
is
the pattern of a purely
conceptual insight. If of observations by the
we
"superposition"
represent mathematically a given totality of several series, we do not
indeed thereby increase our knowledge of the absolute and transcend-
264
ent causes of the process; but we have raised ourselves to a new type It is true we do not grasp the force in things, by which of knowledge.
a definite effect issues from a definite cause; but we comprehend the progress of each step of the theory to the next as strictly and as
exactly as
we grasp the transformation

is
of
any quantitative
it.
relation
into another, that

"description"
is
an active
sense.
The concept of only justified and admissible when we take it in To describe a group of phenomena, then, means not
logically equivalent to
merely to record receptively the sensuous impressions received from From among it, but it means to transform them intellectually. the theoretically known and developed forms of mathematical connection (for instance, from among the forms of pure geometry), a selection and combination must be made such that the elements
given here and now appear as constructively deduced elements in the system, which arises. The logical moment given here cannot be denied even in the theories of empiricism, or under whatever names
The "adjustment of ideas to reality" pre the very concept of this reality, and thus a system of intel supposes lectual postulates. It is in the principle of the unambiguity (Eindeuit
may
be concealed.
tigkeit) of
natural processes above

"I
all,
that
all
these postulates are
am convinced," says Mach himself, ultimately combined. "that in nature so much happens as can happen, and that this only can happen only in one way." All physical process is thus com
by the momentarily effective circumstances, and can thus take place in only one way. 16 However, if we analyse the grounds of this conviction, we are implicitly led back to all those fundamental conceptions, which the sensationalistic explanation
pletely determined
explicitly denies.
The conception of the unambiguity and

is
"stability"
of being obviously
first
not in the content of perception as given in our immediate experience, but it indicates the goal, which the
intellectual labor of science seeks to
have
this content approximate.
This goal can only be reached,
if
we
are able to establish certain
permanent relations in the flux of sensations, which differ and have their truth limited to a single moment of time; the rules of these permanent relations we can call to mind independently of change of
the
momentary
scientific
material. To the extent that this takes place, the concept of nature develops and is confirmed. The biologif.; cf.
16 Mach, Prinzipien der Warmelehre, p. 392 2nd.ed., p.222ff.
Analyse der Empfindungen

cal
265
theory of knowledge, on the other hand, seeks to preserve the constancy of being through the theory that all knowledge is a pro gressive adaptation to being; yet it is not able to ground the assertion
of this constancy in an appropriate instrument of knowledge. permanence of our environment is spoken of, out of which a corre
sponding permanence of thought is to develop. But it is overlooked that ultimately nothing else is meant by this permanence of the environment than the persistence of fixed functional relations, which are formulable between the elements of experience. Along with the content of these elements, there is recognized a form of their con
nection, which at any rate can in no way be reduced to the material oppositions of sensation, i.e., to the bright and the dark, the sweet
and the
bitter, etc.
But
all conflict is
herewith in principle removed.
From
the beginning, it has been admitted and emphasized that the conception of a constant law is indispensable in the definition of a
It only remains to perceive that this conception is a perfectly independent element of knowledge, and resists all reduc tion to the assumed "simple" sense impressions. Progressive analy sis leads to increasingly exact confirmation of this fundamental
natural object.
logical character of the pure concepts of relation is revealed to the extent that they arrange themselves in a fixed system
difference.
The
and are exhibited

dependencies.
in the
whole wealth of their ramifications and mutual
in
The^rjjMem.jtfv.laws of nature. It has been shown that the two fundamental moments of induction: the gaining of particular "facts"
and the connecting of these motive of thought. In both
facts into laws, go

cases, the
problem can be used as constants of theoreti cal construction. Even the establishment of a particular, tempo rally limited occurrence reveals this fundamental feature: viz., that in the changing process, certain permanent connections can be
flux of experience elements, that
is
back to the same to raise from the
grasped and retained, (p. 243 ff.) The scientific explanation of any involved group of phenomena by means of the "isolation" and here a "superposition" of simpler relations carries the task imposed
step further.
We discover in the ultimate empirical
"laws
of
nature"
what may be called constants of a higher order, such as rise above the mere existence of the individual fact, that is established in a
266
definite
magnitude. Nevertheless, the general procedure, which is here everywhere effective, only reaches apparent completion in this The "fundamental laws" of natural science seem at first to result.
represent the final
"form"
of
all
empirical processes, but regarded
from another point

consideration.
of view, they serve only as the material for further In the further process of knowledge, these "constants
of the second
level"
are resolved into variables.
They
are only
valid relatively to a certain sphere of experience, and must be ready, when this sphere is extended, to change their import. Thus we stand before a ceaseless progress, in which the fixed form of being
and process that we believed we had gained, seems to escape us. All scientific thought is dominated by the demand for unchanging elements, while on the other hand, the empirically given constantly
renders this
what we call science appears not as an approximation to any "abiding and permanent" reality, but only as a continually renewed illusion, as a phantasmagoria, in
lose it again. this standpoint,
demand From
fruitless.
We
grasp permanent being only to
which each new picture displaces all the earlier ones, only itself to disappear and be annihilated by another. Laws and constants. This very comparison, however, points to a necessary limit in radical scepticism. The images in the presenta
life of the individual, to which the particular phases of science are here compared, always have a certain inner form of connection with each other, no matter how variegated and diverse in their
tional
succession; and without this inner connection, they could not be grasped as contents of the same consciousness. They all stand at
least in
earlier
an ordered temporal connection, in a definite relation of and later; and this one feature suffices to give them a funda
mental common character through all diversity of individual form. No matter how much the particular elements may differ from each
other in their material content, they must nevertheless agree in those determinations, on which the serial form rests, in which they all
Even in the loosest, most slack succession of members, participate. the preceding member is not absolutely destroyed by the entry of the succeeding member; but certain fundamental determinations
on which rest the homogeneity and uniformity of the series. In the successive phases of science, this postulate is most purely and Each change in the system of scientific concepts perfectly realized. places in a clear light the permanent structural elements to be
persist,
267
ascribed to this system, as it is only under the assumption of these elements that it can be described. If we take as given the whole of
experience, as
this
whole
is
it is represented in any definite stage of knowledge, never a mere aggregate of perceptual data, but is
divided and unified according to definite theoretical points of view. It has already been shown on all sides that, without such points of
view, no single assertion concerning facts, in particular no single concrete measurement, would be possible. (Cf. above p. 140 ff.)
if we view the totality of empirical knowledge at any point of we can represent it in the form of a function, which reproduces time, the characteristic relation according to which we conceive the indi vidual members arranged in mutual dependency. Generally speak ing, we have some such form as F (A, B, C, D ), in which it is
Thus
to be
remembered that what appears
in this expression as
an element
may prove, otherwise considered, to be a very complex system, so that the member A would be replaced by f (EI, a 2 a n ), the member
,
. . .
by $
(bi
b2
...
bn)
etc.
Thus there
arises a
overlapping
certain
fields of
(ineinandergreifender)
syntheses,
complex whole of which stand in a
and subordination. Two and B are first united each according to a phenomena particular law \f/i (i, a z a 3 ), 2 (j8i, /3 2 j8 3 ); these laws are again
mutual relation
of superordination
\f/
connected
among themselves by a new
relation
<b(\l/i,
^ 2 ),
until
we
finally reach the most general relation, which ascribes to each indi vidual factor its definite place with regard to the others. The funda
F is analysed for thought into a structure of mutually dependent determinations, which would be symbolically represented, for example, by an expression F [$i (î ^ 2 ), 2 (^3 ^4)^3 . . . ]. If it is found that some wholly assured observation does not agree with the determinations, that are to be expected and calculated on the basis of this most general theoretical formula, then this formula needs correction. But this correction cannot involve the removal of any element indiscriminately from the formula, for the correction must proceed according to a definite principle of methodic advance. The
mental form
$>
transformation occurs, as it were, "from within outwards;" at first, while retaining the more inclusive relations F, $1, 2 etc., the special
<J>
relations î, would be transformed; 2 attempt to establish again the unbroken

. . .
and
way, we would agreement of theory and

in this
observation.
tion of
The insertion of intermediate terms, and the institu new experiments occurs as part of the intellectual tendency
268
to verify and to

"save"
the more inclusive laws, by deducing the
diverging result from the laws themselves through addition of a
new
determining factor.
The preservation of a general evident here; it is found even when the and their purely empirical necessary revision goes beyond the of connection to the principles and universal laws themselves.
The general form of experience.
of experience
is
"form"
"facts"
"rules"
Also such principles as, for example, those on which Newton founds his mechanics, do not need to be taken as absolutely unchanging dogmas; they can rather be regarded as the temporarily simplest intellectual We "hypotheses," by which we establish the unity of experience.
do not relinquish the content of these hypotheses, as long as any less sweeping variation, concerning a deduced element, can reestablish the harmony between theory and experience. But if this way has been closed, criticism is directed back to the presuppositions them selves and to the demand for their reshaping. Here it is the "func tional form" itself, that changes into another; but this transition never means that the fundamental form absolutely disappears, and another absolutely new form arises in its place. The new form must contain the answer to questions, proposed within the older form; this one feature establishes a logical connection between them, and points to a common forum of judgment, to which both are sub
jected.
this
The transformation must

it is
unaffected; for
leave a certain body of principles undertaken merely for the sake of preserving
body
of principles,
and these reveal
its real goal.
Since
we
never compare the system of hypotheses in itself with the naked facts in themselves, but always can only oppose one hypothetical system of principles to another more inclusive, more radical system, we need for this progressive comparison an ultimate constant stand ard of measurement of supreme principles of experience in general.
Thought demands the identity of this logical standard of measure ment amid all the change of what is measured. In this sense, the
theory of experience would constitute the universal invariant and thus fulfill a requirement clearly urged by inductive procedure itself. The procedure of the "transcendental
critical
theory of experience,
can be directly compared at this point with that of Just as the geometrician selects for investigation those relations of a definite figure, which remain unchanged by certain transformations, so here the attempt is made to discover those
philosophy"
geometry.
269
universal elements of form, that persist through all change in the The "categories" of particular material content of experience. space and time, of magnitude and the functional dependency of
magnitudes, etc., are established as such elements of form, which cannot be lacking in any empirical judgment or system of judgments.
Also the method followed here shows the same
as
"rational"
structure
was found
in mathematics.
Here as
there,
in order to prove the independence of certain changes, that we should actually carry out these changes, but it was sufficient merely to grasp the direction of the change once
for
all,
was not necessary, a conceptual relation from

it
in
making the
decision. 17
We
prove that the meaning of
certain functions of experience is not affected in principle by a change in their material content. For example, the validity of a space-
time dependency of the elements in natural processes expressed by universal causal law, remains unaffected by any change in the
particular causal principles.
The
reached,
if
we succeeded
in isolating in this
goal of critical analysis would be way the ultimate common
element of all possible forms of scientific experience; i.e., if we suc ceeded in conceptually defining those moments, which persist in the
advance from theory to theory because they are the conditions of any theory. At no given stage of knowledge can this goal be perfectly achieved; nevertheless it remains as a demand, and pre scribes a fixed direction to the continuous unfolding and evolution of
the systems of experience.
The concept of
evident.
the
a priori and
the "invariants of experience."
From
is
this point of view, the strictly limited
meaning
of the
"a
priori"
clearly Only those ultimate logical invariants can be called a priori, which lie at the basis of any determination of a connection according to natural law. A cognition is called a priori
not in any sense as if it were prior to experience, but because and in so far as it is contained as a necessary premise in every valid judgment concerning facts. If we analyse such a judgment, we
find,
along with the immediate contents of sensuous data and ele

differing
from case to case, something permanent; we find, as it of which the assertion involved "arguments," represents an appropriate functional value. In fact, this funda mental relation has never been seriously denied by even the most
ments
were, a system of
convinced
17
"empiricism."
When,
for
instance,
the
evolutionary
Cf.
above
p. 239
ff.
270
theory of experience lays weight on the fact that the sensation of time and the idea of time evolve adjustment to the temporal and spatial environment," this uncontested and incontestable proposition contains the very factors here in question, in the concept of "environ
"in
ment,"
which
it
fixed, objective
presupposes. It is herein assumed, that there order of time, and that events do not succeed each
"is"
other in arbitrary, capricious fashion, but proceed "out of each other" according to a definite rule. The truth of this assumption must
fixed, if the conception of evolution is to have any justifica indeed any meaning. And it is to the truth of this judgmental tion, connection, and not to the existence of any sort of presentations in us,
remain
that the concept of the a priori is alone applicable in its purely We are not concerned with the existence of psychic logical meaning. but only with the validity of certain relations and with contents,
their superordination
theory of knowledge, space
and subordination. and not color
is
According to the critical because only priori,

"a
space forms an invariant in every physical construction. But the more sharply the opposition appears between truth and reality, the more clearly it is shown to contain an unsolved problem. Necessary as their separation is, a mediation between them on the other hand, must be assumed, if knowledge is to be brought into a unitary system. It must now be asked whether there is a path within knowledge
leading from the pure logical and mathematical systems to the problem of reality. And if such a path can be shown, what new meaning does the problem gain thereby, and in what direction is its solution to be pointed out to thought?
itself,
CHAPTER
VI

The separation of
teristic
"subjective"
and
"objective"
reality.
The charac
procedure of metaphysics does not consist in transcending the field of knowledge in general, for beyond this field there would not be
even material for a possible question,- but in separating correla tive standpoints within the field of knowledge itself, and thus trans forming what is logically correlative into an opposition of things. At no point is this feature so significant as (Cf. above, p. 237 f.)
in the old question as to the relation of thought and being, of the subject and object of knowledge. This one opposition conceals
and can progressively develop into them. If become conceptually separated, they fall into two separate spatial spheres, into an inner and an outer world, between which there is no intelligible causal connection.
all
the others within

"things"
it
once
and the
"mind"
the conflict constantly grows sharper. If the objects only exist as a plurality, then for the subject the postulate of unity is essential; if the element of change and motion belongs to the essence of reality,
And
then
it is identity and unchangeableness, that is demanded of the true concept. No dialectical solution can ever fully transcend these separations, which are already present in the original formulation of
problems; the history of metaphysics wavers between opposing tendencies, without being able to deduce the one from the other, or
to reduce
them
to each other.
yet at least the system of empirical knowledge has an original unity, which persists in spite of all these oppositions. The constant progress of science is not diverted from its goal by the varying
fortunes of metaphysics. It must be possible to gain clarity regard ing the direction of this advance, without presupposing the dualism of the metaphysical concepts. In so far as this dualism is applic
able to experience, it must make itself intelligible entirely on the Thus the basis of experience and its characteristic principles.
And
question is no longer what absolute separation underlies the opposi tion of the "inner" and the "outer," the "presentation" and the
"object;"
the question merely
is
from what standpoints and by what

271
272
necessity does knowledge itself reach these divisions. The separation and reconciliation of these concepts has been a problem throughout
the whole history of philosophy; are these concepts merely intel lectual phantoms, or do they have a fixed meaning and function in
the structure of knowledge?
The development of the concepts of objectivity and subjectivity. If consult immediate experience unmixed with reflection, the opposi tion of the "subjective" and the "objective" is shown to be wholly For such experience, there is only one plane of "exist foreign to it. and this includes all contents uniformly and without dis ence," tinction. What is grasped by consciousness here and now and is precisely in the form offered by direct experience. Between
we
"is,"
the experiences, in particular, that refer to the individual s own body, and those that refer to "outer" things, there is still no fixed
The temporal limits of the particular experiences are vague; the past, in so far as it is taken up into memory, is just as The various contents are as the present. given and as
line of division.
"real"
arranged, as it were, in one plane; as yet there is no definite point of view, that would establish any preeminence of one over the other.
If we characterize this level at all by the opposition of the subjective and the objective (and it can only be done in an extravagant and unreal sense), then we must ascribe to it thorough-going objectivity; for the contents still have that passivity, that indubitable givenness, which we customarily connect with the conception of the But the very first beginning of logical reflection destroys this impres sion of perfect unity and completeness. The division, whose growth
"thing."
starts here,
is already present, though concealed, in the first attempts at a scientific view of the world. The fundamental tendency of this is
view
values.
not simply to receive sensuous data, but to distinguish their The fleeting, unique observation is more and more forced to
the back-ground; only the "typical" experiences are to be retained, such as recur in a permanent manner, and under conditions that can
be universally formulated and established. When science under takes to shape the given and to deduce it from definite principles, it must set aside the original relation of coordination of all the data of experience, and substitute a relation of superordination and subor
dination.
Every critical doubt, however, that is directed against the universal validity of any perception, bears within it in germ the division of being into a "subjective" and an "objective" sphere. The
273
analysis of the concept of experience has already led to that opposition, which is destined to resolve the metaphysical division of subject and
by taking up into itself its essential conceptual meaning. Changing and constant elements of experience. The goal of all empirical knowledge lies in gaining ultimate invariants as the neces sary and constitutive factors in each empirical judgment. From
object
this standpoint, however, the various empirical assertions
appear of
very different value.
Along
with the loose, associative connection
of perceptions united only under particular circumstances (as, for example, under definite physiological conditions), there are found
fixed connections,
and belong to
particular
any whole field of objects, of the differences given in the this field independently find place and definite tune of the observation.
which are valid
for
We
connections, which
hold their
ground through
all
further experimental
testing and through apparently contrary instances, and remain steadfast in the flux of experience while others dissolve and vanish. It is the former, that we call "objective" in a pregnant sense, while
designate the latter by the term "subjective." We finally call objective those elements of experience, which persist through all change in the here and now, and on which rests the unchangeable
we
all
character of experience; while we ascribe to the sphere of subjectivity that belongs to this change itself, and that only expresses a determination of the particular, unique here and now. The result
of thus deriving the distinction

is
between the subjective and the has merely relative significance. For there are objective, no absolutely changeable elements of experience at any stage of knowledge we have reached, any more than there are absolutely constant elements. A content can only be known as changeable
that
it
with reference to another, with which

first
it is
compared, and which at
claims permanent existence for itself. At the same time, the possibility always remains that this second content will be corrected
by a third, and thus may no longer hold as a true and perfect expres Thus sion of objectivity, but as a mere partial expression of being. we are not concerned here with a fixed line of division, separating
two eternally sundered which constantly shifts
fields of reality, but with a moving limit, The present in the progress of knowledge. phase of experience appears just as "objective" in contrast to the past, as it appears "subjective" in contrast to the later
phases.
Only
this
mutual act
of correction remains standing, only
274
this function of

comparison
persists, while
the material content of
the two fields

division,
is
in constant flux.
The
spatial expression of the
i.e., the analysis of being into an inner and an outer world, thus insufficient and misleading, because it obscures this funda mental relation; it establishes instead of a living, reciprocal relation
is
realized along with advancing knowledge, a fixed and absolutely closed division of things. The opposition involved here is not of but rather of dynamic nature; it signifies the differing power spatial,
judgments to withstand continuous testing by theory and observation, without thereby being altered in content. In this perpetually renewed process, groups, that were originally taken as are always being separated out; and now that they cannot the test, they lose this character of fixity, which constitutes satisfy the fundamental property of all objectivity. But it now becomes clear that our concern with this transition into the subjective is not with a change in the substance of things, but merely with a change
of empirical
"fixed,"
in the critical evaluation of cognitions.
"Things" are not thereby but a judgment, that previously "presentations," seemed to hold unconditionally, is now limited to a certain sphere of
degraded into mere

conditions.
The
plain,
subjectivity of the sensuous qualities.

if
we
consider the best
known example
This relation can be made of this transformation
of objectivity into subjectivity, viz., the discovery of the "subjectivity of the sensuous qualities." Even for Democritus, who made this discovery, it means fundamentally nothing but the fact that the
colors
and
tones, the smells
and
tastes
have a peculiar character for
knowledge, by which they are excluded from the scientific construc tion of reality. They pass from the yvrjalrj yvu^f] into the aKorlrj yvu/jni; they are separated from the pure mathematical ideas of space,
form and motion, to which alone henceforth physical However, this division does not mean that they are denied all part in being in general; rather, a narrower field is marked off for what previously passed as absolute witness of reality; never theless within this narrower field, the sense qualities retain their full validity. The seen color, the heard tone, is and remains some thing only this reality does not subsist in isolation and for itself, but results from the interaction of the physical stimulus and
of
is
"truth"
ascribed.
"real;"
the appropriate organ of sensation. Thus when the qualities are explained as subjective, they fall outside the world of "pure forms,"
275
that mathematical physics constructs, but not outside of nature as such;. for precisely this relation of physical and physiological condi tions, on which the sense qualities rest, is itself a part of "nature,"
indeed, the concept of nature is first realized in the reciprocal causal dependency of particular elements. The same holds true when we
go beyond the sphere of secondary qualities to the illusions and deceptions of the senses. When the straight stick appears in water as if broken, this is no unreal appearance but a phenomenon "wellgrounded" in the laws of the refraction of light; and as such expresses with perfect adequacy a complex connection of elements of experi ence. Error only begins when we transfer a determination, which holds of a particular member, to the total complex; we thus apply a
judgment, that has been found valid under certain limitations, to

experience as a whole, apart from any limiting condition. That the stick is broken, is a valid empirical judgment, in so far as the phenom
we must add
can be confirmed and deduced as necessary; only to this judgment, as it were, a logical index signifying the special conditions of its validity, from which it cannot be abstracted. The series of degrees of objectivity. If we survey the whole of
enon
it
refers to
these considerations, the series of degrees of objectivity is clearly revealed. As long as we remain at the stage of the metaphysical
and outer, we have an opposition that permits absolutely no mediation. Here there is only a simple "either-or." Just as a thing cannot be simultaneously at two different places in space, so the "inner" cannot in any aspect be at the same time an
distinction of the inner
and conversely. In the critical formulation of the question, on the contrary, this limitation is set aside. The opposition is no longer between two members, but between many members; as we have seen, the same content of experience can be called subjective and objective, according as it is conceived relatively to different Sensuous perception, as opposed to the logical points of reference. hallucination and the dream, signifies the real type of the objective; while measured by the schema of exact physics, sense perception can become a phenomenon that no longer expresses an independent property of "things," but only a subjective condition of the observer.
outer,
In truth, we are always concerned here with a relation, holding between a relatively narrower and a broader sphere of experience, between relatively dependent and independent judgments. Instead of a mere duality of determinations a series of values is given,
276
Each member which progresses according to a definite rule. to a successor, and requires this successor for its own comple points tion. Even in the popular and prescientific view, the first, impor tant phases of this development can be recognized. When we charac terize a sensuous impression, that is given us here and now in a or "green," even this primitive act of definite nuance, as is directed from variables to constants, as is essential judgment
"red"
to all knowledge.
Even
here the content of the sensation

(Erlebnis)
is
separated
from
its
momentary experiencing
and
is
opposed as
independent; the content appears, over against the particular tem poral act, as a permanent moment, that can be retained as an identical
determination.
But this intellectual permanence, which is latent even
in the individual impression, lending it a real existence, is nevertheless Here it far behind that in the thing-concept of ordinary experience.
is
not sufficient simply to combine the sensuous perceptions, but along with this mere unification must go an act of logical com The object of experience is conceived as a continuous being, pletion.
whose persistence
moments
of the continuous sequence of postulated as necessary. Direct perception, on the contrary, always offers us only isolated fragments, only entirely discrete values, which in no combination constitute a continuous
in every point of
time
is
whole.
The truly and "heard" furnishes only disconnected, temporally separated masses of perception; while the concept of the requires the perfect filling of the time series, and thus, "object"
"seen"
strictly speaking, requires the
assumption of an
infinite totality of
Thus at this second level, the general procedure is clearly revealed for transforming and enriching the given, on the basis of the It is on the logical demand for its thorough-going connection.
elements.
continuation of this procedure, that science bases its definition of nature and the natural object. The logical tendencies found in the concept of experience of the ordinary view of the world are now
consciously taken up and carried further with methodic purpose. The "things," that arise henceforth, prove, the more distinctly their real meaning is comprehended, to be metaphorical expres
permanent connections of phenomena according to law, and thus expressions of the constancy and continuity of experience This fixity and continuity is never fully realized in any sen itself.
sions of
suously perceptible object; so in order to reach it, thought is led to a .hypothetical substructure of empirical being, which however has no
277
other function than to represent the permanent order of this being Thus there is an unbroken develop (Cf. above p. 164 ff.) ment from the first stages of objectification to its completed scientific
itself.
process would be completed as soon as we succeeded advancing to the ultimate constants of experience in general, which, as we have seen, constitute both the presupposition and the
form.
in
The
goal of investigation. The system of these unchanging elements in so far as this term constitutes the type of objectivity in general, is purely limited to a meaning wholly comprehensible to knowledge.
The logical gradations of the contents of experience. True, it remains an insoluble problem how the "thing-in-itself" passes into the mere "presentation," how absolute existence changes into absolute knowl edge; but with this question we have nothing to do in the critically clarified conception of the opposition of the subjective and the objec tive. Here we do not measure presentations with respect to absolute objects; but different partial expressions of the same total experi ence serve as standards of measurement reciprocally for each other. Each partial experience is accordingly examined as to what it means for the total system; and this meaning determines its degree of In the last analysis, we are not concerned with what objectivity. a definite experience but with what it worth;" i.e., with what function it has as a particular building-stone in the structure of the whole. Dream experiences also are not distinguished from wak
"is
"is,"
ing experiences by any specific character (Ding charakter} attached as a permanently recognizable property. They also have a sort of
grounded in definite physiological bodily states; but this being does not extend beyond the sphere and time-span within which these con ditions are fulfilled. Insight into the subjective character of the
"being,"
in so far as they are

"objective"
conditions, in
dream means nothing
else than the reestablishment of a logical gradation among the contents of consciousness, which for a time threatened to disappear. Thus in general, the opposition of the subjective and objective serves in its development for the increas
ingly strict organization of experience.
We seek to gain permanent contents in place of changing contents; but at the same time we are conscious, that every attempt in this direction only partially ful fills the fundamental demand, and hence requires completion in a
new
thus gain a sequence of superordinated and subordinated moments, representing, as it were, various compleconstruction.
We
278
mentary phases
phases,
None of these of the solution of the same problem. not even those most remote from the goal, can be entirely dispensed with; but, on the other hand, no one of them represents an
absolutely unconditional solution. Thus it is true we can never compare the experience of things with the things themselves, as they are assumed to be in themselves separate from all the conditions of
we can very well replace a relatively narrower aspect of experience by a broader, so that the given data are thereby ordered under a new, more general point of view. The earlier results are
experience but
;
not thereby rendered valueless, but are rather confirmed within a Each later member of the series is definite sphere of validity.
necessarily connected with the earlier ones, in so far as it answers a face here a perpetually self-renewing question latent in them.
We
process with only relative stopping-points; and it is these stoppingpoints, which define the concept of "objectivity" at any time.
The problem of transcendence. The direction of this progress of experience is also directly opposed to what would be expected on the
ordinary metaphysical presuppositions. these presuppositions, it is the subject, it
From
is
the standpoint of the presentations in us,
which alone are given to us in the beginning, and from them we gain access to the world of objects only with difficulty. The history of
philosophy shows, however, how all attempts of this sort fail. If we have once enclosed ourselves in the circle of "self-consciousness," no labor on the part of thought (which itself belongs wholly to this On the other hand, the criticism of circle) can lead us out again.
knowledge reverses the problem; for it, the problem is not how we go from the "subjective" to the "objective," but how we go from the to the "subjective." It recognizes no other and no "objective" higher objectivity than that, which is given in experience itself and according to its conditions. Thus it does not ask whether the whole of experience is objectively true and valid, for this would
presuppose a standard that could never be given in knowledge, but it only asks whether a special, particular content is a permanent
or transitory part of this whole. are not concerned with estimat ing the absolute value of the system in its totality, but with estab
lishing
We
a difference in value
among
its
particular factors.
The
question as to the objectivity of experience in general rests ultimately on a logical illusion, of which the history of metaphysics offers many
examples.
Such a question
is
on the same plane
in principle
with the
279
question, for instance, as to the absolute place of the world. Just as in this latter we falsely carry over to the universe as a whole a
relation,
which
is
only valid for the particular parts of the universe in
their reciprocal relations, so in the former question a conceptual opposition, proper for distinguishing the particular phases of empirical
knowledge,
is
applied to the conceived
totality of
their sequence.
The
full
measure of
"objectivity"
these phases and belongs to each
particular experience, as long as it is not displaced and corrected by another. To the degree that this continual testing and self-correc tion is carried on, there is an increasing exclusion of material from the final scientific conception of reality, although the material retains
its right within a limited sphere. Elements, that at first seemed necessary and constitutive in the concept of empirical being, such as the specific content of the particular sensations, lose this
dominant
position,
and henceforth have no central
significance but
only a peripheral one. Thus the designation of an element as "sub jective" does not belong to it originally, but presupposes a compli cated use of intellectual and empirical controls, which is only gained
at a relatively high level. This designation first arises in the recipro cal criticism of experiences, where the changing existence is separated
from the permanent.
The "subjective" is not the self-evident, given out of which the world of objects is constructed by a starting-point speculative synthesis; but it is the result of an analysis and pre
supposes the permanence of experience and hence the validity of fixed relations between contents in general. The meaning of judgment. The progress of this analysis is evident
to the relation of the "universal" and "particu which appeared in the definition of the inductive judgment. We saw that each particular judgment originally claims uncondi tional validity for itself. As a judgment, it does not intend to
lar,"
when we turn back
describe only momentary sensations in their individual peculiarity, but to establish a matter of fact held to be valid in itself, independent
of all particular temporal circumstances. The judgment as such, by virtue of its logical function, looks beyond the circle of what is given
any moment, while it affirms a universally valid connection between the subject and predicate. (See above p. 246 ff.) Only special motives are able to lead thought to deviate from this first demand, and to limit its assertions expressly to a narrower circle. This limitation only takes place in so far as there is a conflict between
at
280
different empirical assertions.
Assertions, that would be incompati taken absolutely, are now set in harmony with each other by being related to different subjects; thus at least one of them confines itself to expressing the "nature" of things, not abso
ble in content
if
lutely,
but only from a special point of view and under certain As the particular geometrical form, according to a well-known proposition of the Kantian theory of space, is only
limiting conditions.
all-inclusive space, so the particu gained by limitation of the issues from a limitation of the one system lar judgment of experience
"one"
judgments in general, and presupposes this system. The particular judgment of experience arises when a plurality of spheres of experience, each one conceived to be entirely determined according to law, intersect and mutually determine each other. There is no road to the law from the absolutely isolated "impres of logical relation is extinguished; on the sion," where every thought
of experiential
entirely intelligible how, owing to the general demand thorough ordering of experiences according to law, we are led at first to exclude particular contents, such as cannot apparently be
other hand,
it is
for a
brought into the general plan, in order to deduce them subsequently from a particular complex of conditions. The "transcending of sensuous experience. It is thus a logical differentiation of the contents of experience and their arrangement in an ordered system of dependencies, that constitutes the real kernel
1
of the concept of reality.
This connection
is
confirmed anew
when
we
consider the logical character of scientific investigation more closely, for this is indeed the genuine witness of empirical reality.
The
scientific
present,
momentary
experiment never makes a simple report regarding the facts of perception, but it only gains its value by
bringing the particular data under a definite standpoint of judgment, and thus giving them a meaning not found in the simple sensuous
experience as such. What we observe, for instance, is a definite deflection of the magnetic needle under certain conditions; what we
assert,
on the contrary, as the
result of the experiment is
always an
objective connection of physical propositions, which far transcend the limited field of facts accessible to us at a particular moment.
As Duhem has admirably explained, the physicist, in order to reach a real result in his investigations, must always transform the actual
assumes and requires.
case before his eyes into an expression of the ideal case, which theory Therewith the particular instrument before
281
him changes from a group
ideal intellectual determinations.
of sensuous properties into a whole of It is no longer a definite tool,
a thing of copper or steel, of aluminum or glass, to which he refers in his assertions; but in its place are concepts, such as that of the mag
netic field, the magnetic axis, the intensity of the current,
for their part, are again only the
etc.,
which,
symbol and husk

1
of universal
The characteristic mathematico-physical relations and connections. merit of experiment rests on the fact that in it one stroke establishes a thousand connections. The limited circle of facts, that is sensu
ously accessible, expands before our intellectual vision into a univer sal connection of phenomena according to natural law. The imme diate indications of the moment are transcended on every hand; in
their place appears the conception of a universal order, of such a sort
it has equal validity in the smallest as in the greatest, and can be reconstructed again from any particular point. It is only by means of this enrichment of its immediate import, that the content of
that
perception becomes the content of physics, and thus becomes an "objectively real" content.
Thus we have to do here with a type
of
"transcendence;"
for the
particular given impression does not remain merely what it is, but becomes a symbol of a thorough-going systematic organization,
within which
it
stands and to a certain extent participates.
This
change in the empirical impression does not, however, alter its metaphysical "substance," but merely its logical form. What at first appeared in isolation, now combines and shows a reciprocal reference; what previously passed as simple, now reveals an inner wealth and multiplicity, in so far as we can reach from it other and
other data of experience, by a continuous progress according to In connecting the particular contents with each other, definite rules.
still
as
is
it
were, ever with
new
threads,
we
give
them that
fixity,
which
distinguishing property of empirical objectivity. What imprints the mark of true objectivity is not the sensuous vividness
the
but this wealth of inner relations. What elevates above the things of sense and lends them their "things" kind of "reality," is the wealth of consequences proceeding peculiar out of them. Such things of physics only indicate different ways of advancing from one experience to another, so that the totality of being can be finally surveyed as the totality of the system of experience.
of the impression,
the
of physics
Duhem, La Theorie Physique,
p. 251
f., cf.
above
p. 190
ff.
282

"representation"
The concept of
The concept and term,

it,
repre
sentation, has, in spite of all the attacks against
maintained a central position in the history of knowledge; and this concept here receives a new meaning. In metaphysical doctrines, the "presentation" (Vorstellung) refers to the object, which stands behind it. Thus the here is of an different nature than the signified, and belongs to another entirely realm of being. Precisely in this lies the real riddle of knowledge. If the absolute object were already otherwise known to us, then at least it might be understood how we could indirectly read off its par ticular properties from the sort of presentations, that arise from it. Once we have assured ourselves of the existence of two different series, we might attempt by an analogical inference to carry over the rela tions we find in one series to the other; on the other hand, it is incom prehensible how we should succeed in inferring the existence of one As soon as we series, from data belonging exclusively to the other. have even a general certainty of transcendent things beyond all knowledge, we can seek, in the immediate content of experience for signs of this reality, which are given at least in concept; how this con cept itself arises and what makes it necessary, is, on the other hand, not explained by the theory of signs. This difficulty continually reappears in the development of the concept of representation. In
"sign"
persistently the theory of
ancient atomism, the "images" of things, which inform us of their being, are conceived as material parts, which separate themselves
from things and undergo
many
physical changes on the
way
to our
It is, although in reduced measure, the real substance sense-organs. of the body, which enters us in sense-perception and blends with our
own
But this materialistic account cannot reach the logical being. goal for which it was undertaken; for here again the unity of experi
ence
is
only apparently preserved.
Even
if
things gave off a part
of themselves, as it were, in order to become known, it would still remain as obscure as before how this part could possibly be taken, not as what it was in and for itself, but as an expression of an includ
ing whole.
This reference to the whole would always require a
Thus the peculiar function, that is not deduced but presupposed. Aristotelian and scholastic theory of perception seems to come closer
to the real psychological fact, when, instead of explaining this func The whole content of the tion, it postulates it from the beginning. "immaterial species," by which we grasp the being of things, is

accordingly reduced to the act of representation. We do not of the species itself, yet only through it do we
"
283
any determinations
know know
the relations of outer things; cognoscimus non ipsam speciem impressam, sed per speciem." The "similarity," which is to be assumed
between the sign and the signified, is thus not to be conceived as if both belonged to the same logical category. The species agree in no single actual property with the object to which they refer, for they are characterized only by this operation of reference, and not by any property in which they could be similar to other things.
The
turae
conception, that they are

objectorum,"
is
"formales
similitudines ac veluti pic-
the most
"The
expressly contested and rejected, at least in mature and consistent exposition of the theory by Suarez.
assimilation of consciousness with the object does not
mean for
Suarez that elements are thereby brought into consciousness, and that these elements stand in objective relation to other functions of
consciousness,
and appear to these as the object; but rather
his
meaning
is
that the whole consciousness becomes an instrument of
expressa).
knowledge, and in so far an image (better, an expression, species Consciousness performs an act, assumes a peculiar
property, which of
itself directs
consciousness on the real object.
The
living activity of consciousness, a perceiving, not a perceived, which knows the extra-conscious object by means of the species, is
said to be similar to the
object."
weighty new distinction appears
here, although concealed in scholastic terminology.
The
fact,
that
an element
"refers"
to another
and
indirectly represents
it, is
now no
longer explained by a particular property of this element itself, but is reduced to a characteristic function of knowledge, and in particular to the judgment. It is true this insight could not be maintained
with
on the contrary, the functional relation of to which the analysis leads, always threatens to change expression, into a substantial relation of the participation of things in certain
all strictness; for,
objective properties.
of things although
Thus the
species again -become the
"traces"
they no longer possess complete existential import,
"essence." The conflict of these two conceptions the concept of "representation" of its clear and finally deprives In order to let the operation of expression be seen, definite meaning. the content, which serves as a sign, must be stripped more and more
but only a faded
2 H. Schwarz, Die Umwalzung der Wahrnemungshypothesen mechanische Methode. Leipzig 1895, I, p. 25; cf. p. 12 ff.
durch
die
284
same time, however, the objectify ascribed to the content seems to lose its best support. ing meaning Thus the theory of representation always threatens to lapse into
of its thing-like character; at the
skepticism; for what assurance have we that the symbol of being, which we believe we have in our presentation, genuinely reproduces the content of being, and does not misrepresent its essential features? Transformation of the concept of representation and progress to the "whole of experience." The new meaning, which the criticism of knowledge gives to the concept of representation, removes this
danger. It is now recognized that each particular phase of experi ence has a "representative" character, in so far as it refers to another
and
finally leads
experience.
But
this reference
by progress according to rule to the totality of beyond concerns only the transition
from one particular serial member to the totality, to which it belongs, and to the universal rule governing this totality. The enlargement does not extend into a field that is absolutely beyond, but on the contrary, aims to grasp as a definite whole the same field, of which
is a part. It places the individual in the system. But if we ask further, as to whence the particular empirical content has this capacity of representing the whole, we are involved by the question in a reversal of the problem. The connection of
the particular experience
the facts and their reciprocal relation is what is primary and original, while their isolation represents merely the result of a technical
abstraction. sion of
Hence if we understand
ideal rule,
"representation"
as the expres
which connects the present, given particu lar with the whole, and combines the two in an intellectual synthesis, then we have in "representation" no mere subsequent determination, but a constitutive condition of all experience. Without this apparent representation, there would also be no presentation, no immediately
it is
an
present content; for this latter only exists for knowledge in so far as brought into a system of relations, that give it spatial and tem
poral as well as conceptual determinateness.
for positing a being outside of all relation to
Just as the necessity knowledge can not be
deduced from the mere concept of knowledge, so must the con cept of knowledge necessarily have within it that postulate of connection to which the critical analysis of the problem of reality leads. The content of experience becomes "objective" for us when we understand how each element is woven into the whole. If we attempt to characterize this whole itself as an illusion, it is a mere
285
play of words; for the difference between reality and appearance presupposed here is itself only possible within the system of experi
ence and under
as to the
offers
its
conditions.
(C
above
p.
273
ff.)
The question
"similarity"
of the empirical sign
no further
difficulty.
with what it signifies The particular element, which serves as a
indeed not materially similar to the totality that is signified, for the relations constituting the totality cannot be fully expressed
sign, is
and "copied" by any particular formation, but a thorough going logical community subsists between them, in so far as both belong in principle to the same system of explanation. The actual
similarity is changed into a conceptual correlation; the two levels of being become different but necessarily complementary points of view for considering the system of experience.
istic
Association as a principle of explanation. Indeed the sensationaltheory of knowledge might attempt to bring this fact into its own
it, by reducing it to the psychological The concept of association seems to offer
explanation, without contesting
concept of
all
"association."
reality, for it
the principles for the formulation and solution of the problem of proceeds from the content of the particular impres
is
sions to the fixed connections
explanation
The defect of the when we analyse more closely the shown, however,
between them.
form, of connection assumed here and which, according to the concepts of associationistic psychology, appears alone admissible. The
"connection"
between the particular members of the
series
here
means nothing but

coexistence
their frequent empirical coexistence. of particular presentations produces not so
And this much a
connection between them, as rather an appearance of one. No conceptual principle, such as can be expressed and established in The strict logical identity, unifies the elements of association.
limited.
paths from one element to another are in and for themselves un Which of these paths is followed in real psychological
thought depends merely on the preceding psychical "dispositions," and thus upon a circumstance, which is to be regarded as variable from moment to moment, and from individual to individual. Here
that constancy and exactitude is feature of the conception of reality.
is the distinguishing only through the critical evaluation of concepts that the formation of the "object" comes to If we proceed from the particular content of experience at a light.
lost,
which
But
it is
particular
moment
of time, there are given in
it
not only certain
286
elements but also certain lines of direction, according to which thought can gradually expand the particular phase into the whole system. The advance is not left to individual caprice, but is demanded accord
ing to law.
strictly
As
science grasps the whole of these

it
demands more
progressively gains the concept of the real. It has already been shown on all hands that this development must everywhere transcend the sphere of mere association. Association,
exactly,
and
understood in the most favorable sense, is merely able to formulate the question; the answer lies only in the universal serial principles,
which predetermine and order the possible logical transitions from member to member, in accordance with certain points of view. The specific meaning of these points of view must remain constant, if the advance is not to lose itself in the indefinite. The necessary guiding concepts of association cannot arise from association itself, but 3 belong to another field and logical origin. In general it appears, that the further we advance into the particu
lar conditions of the problem of reality, the more clearly it unites with the problem of truth. If it is once understood how knowledge attains a constancy of certain predicates and establishes judgmental
connections, then the "transcendence" of the object as opposed to the mere presentation no longer offers any difficulty. And the
means used by knowledge

problems.
"copying"
are
shown
to be the
same
in
both
fields of
Just as the real achievement of the concept is not in a given manifold abstractly and schematically, but in
constituting a law of relation and thus producing a new and unique connection of the manifold, just so it is shown to be the form of connection of experiences that transforms changeable "impressions"
into
"thought"
The
constant objects. In fact, the most general expression of is the same as the most general expression of "being." opposition, that metaphysics could not reconcile, is resolved by
in
going back to the logical function from the application of which
both problems arose, and

explanation.
which they must
finally find their
ii
The concept of objectivity and the problem of space. In the history of and speculative thought, the problem of reality has always been inseparably connected with the problem of space. So close is
scientific
3
Cf.
above esp.
p. 14
and
p. 261
ff.

this connection,
all
287
it governed logical interest, connected with the concept of the real are con questions sidered solved, as soon as the question of the reality of the "external world" has been finally decided. Even the Critique of Pure Reason
and so exclusively has
that
could only approach its real theme by starting with a transformation of the theory of space. Thereby, however, the fate of its historical effect was in large part decided; in the minds of contemporaries and
successors, what was intended to be a criticism of the concept of experience was misapprehended as a metaphysics of the concept of In truth, here again the order of problems must be reversed. space.
We
tive"
cannot proceed from a fixed view of the "subjective" or "objec character of space and from it determine the concept of empiri cal reality in general; but it is the supreme and universal principles
of empirical knowledge, that the "nature" of space.

cal view,
must
finally decide the question as to
The theory of projection and its defects. The empirical-physiologi which was founded by Johannes Miiller, conflicts with this, in so far as it starts with an axiom of unconcealed metaphysical character. It is presupposed, that what we perceive are not things themselves in their real shape and real mutual position and distance, but are immediately only certain determinations of our own body. The object of visual sensation is not the external objects, but parts of the retina, which we can grasp in their real spatial magnitude and
The problem of the physiology of vision is to describe the transition, by which we pass from this consciousness of the retinal images to knowledge of the spatial order of objects. It must be
extension.
shown how we come to place the sensations given us" in the outer world, and how we come to comprehend them as a self-subsisting spatial world. However, if we take the problem in this form, it
"in
soon proves to be insoluble.

process of
ences,"
"projection,"
which
All attempts to reduce the peculiar is assumed, to "unconscious infer
move
in
circle;
of the
is
"external,"
yet this
they always assume a general knowledge In fact, there is what is to be deduced.
no phase of experience, in which sensations are given as inner and separated from all "objective" reference. Sensation in this sense is no empirical reality, but only the result of an abstrac We pass from tion, resting upon very complex logical conditions. the seen objects to the assumption of certain nerve excitations and the sensations corresponding to them; we do not proceed in the
states
288
reverse direction from the sensations
known
in themselves to the
4 Thus the general form objects, which may correspond to them. the coexistence and externality of the particular ele of spatiality,
ments
is
no mediated
result,
with the elements themselves. 5

arises in
and
for itself;
it
a fundamental relation posited cannot be asked how this form can merely be enquired how the form is
but
is
It
more
What
closely determined and specialized in empirical knowledge. needs explanation is not the fact that we go from the inner
for the absolutely "inner" is a mere fiction, but to the outer, how we are led to regard certain contents of the original external
world as finally
"in
us,"
i.e.,
how we
are led to determine
them not
only as in general spatial, but as in necessary correlation with our (Cf. bodily organs, with certain parts of our retina or our brain.
above
p.
274
ff.)
It is
not localization in general, but this particular
localization, that is to be explained; and any such explanation obviously take the general relation of spatiality as a basis.
must
Concept and perception distinguished (Helmholtz). To determine the concept of "reality" means also to find a motive of differentiation, enabling us to separate the originally homogeneous totality of
experiences into groups of different value and import. For example, if we conceive the different perceptual images, which we receive from one and the same "object" according to our distance from it and
according to changing illumination, as comprehended in a series of perceptual images, then from the standpoint of immediate psycholog
ical experience,
of these varying
no property can be indicated at first by which any images should have preeminence over any other.
Only the totality of these data of perception constitutes what we call empirical knowledge of the object; and in this totality no single element is absolutely superfluous. No one of the successive per
spective aspects can claim to be the only valid, absolute expression of the "object itself;" rather all the cognitive value of any particular perception belongs to it only in connection with other contents,
it combines into an empirical whole. And yet this thoroughgoing connection does not mean the complete equivalence of the particular factors. We only gain perception of a definite
with which
hypotheses,"
for the psychological refutation of the "projection Stumpf, Uber den psychologischen Ur sprung der Raumvorstellung, p. 184 ff as well as James, Principles of Psychology, II, 31 ff 5 Cf. later Ch. 8.
Cf.
more particularly
.,
289
As Helmspatial form, when we break through this equivalence. holtz explains incidentally, when we ask what is to be understood
by a body extended in three dimensions, we are led psychologically to nothing else than a series of particular visual images, which mutually pass into one another. Closer analysis, however, shows
that the mere succession of these images, no matter how many we assume, would never of itself give us the presentation of a corporeal object, if the thought of a rule were not added, by which a cer
and position is ascribed to each in the total complex. In this sense, the presentation of the stereometric form plays "the role of a concept compounded from a great series of sensetain order
perceptions, which, however, could not necessarily be construed in verbally expressible definitions, such as the geometrician uses,
but only through the living presentation of the law, according to which the perspective images follow each other." This ordering by a concept means, however, that the various elements do not lie along side of each other like the parts of an aggregate, but that we estimate each of them according to its systematic significance. We also dis tinguish the "typical" experiences, such as we assume to occur uni formly, from "accidental" impressions only found in individual cir
cumstances. And it is the "typical" experiences, that we exclusively apply in the construction of the "objective" spatial world, while endeavoring to exclude all contents in conflict with them. Helmholtz account has illumined this process in detail. Here
the general rule is first established "that we always perceive as present in the field of vision such objects as ought to be present to
produce, under ordinary normal conditions of the use of our eyes, the same impression on the nervous apparatus." A stimulation, occurring under unusual conditions, is at first given the meaning
that would belong to it if it were conceived to arise in the ordinary way. "To use an example, let us assume that the eye-ball is mechan
ically stimulated at the outer corner of the eye.
We
then think
In the see a light on the nasal side of the field of vision. ordinary use of our eyes, where they are stimulated by a light coming
that
we
from without, the external

side,
if
light must enter the eye from the nasal a stimulation of the retina in the region of the outer corner of the eye is to occur. It is thus in accordance with the previously
established rule, that
place a luminous object in such a case in this aforementioned position in the field of vision, in spite of the
we
290

that the mechanical stimulus comes neither from in front in
field of vision
fact,
the
nor from the nasal side of the eye, but on the from the external surface of the eye and more from behind." 6 contrary,
The
particular observations are thus adjusted to a certain set of In these conditions we conditions, that we regard as constant.
have a fixed system of reference, to which each particular experience is tacitly referred. And it is only through this peculiar interpreta tion of the material of sensation, that the whole of objective visual and tactual space arises. This whole of space is never a mere dead copy of particular sense perceptions, but it is a construction, following from certain universal rules. To the extent that the unchanging elements of experience are separated from the changing elements in accordance with these rules, there results a division into an objective and a subjective sphere. And here there is no doubt, that the knowl
edge of subjectivity does not represent the original starting-point, but a logically mediated and later insight. Helmholtz expressly
emphasizes that the knowledge of

sensations,
objects
precedes the knowledge of
and that
it
is
far superior in clarity
and
distinctness.
Under ordinary psychological conditions

tion
is
of experience, the sensa
so exclusively directed on the object and enters into it so completely, that it disappears behind it, as it were. The apprehen
sion of sensation as sensation
is
conscious reflection.
We must first always learn to

"and
always the work of subsequent direct our atten
we ordinarily learn this only means of knowledge of the outer world." "While we attain an extraordinary degree of niceness and sureness in objective observation, we not only do not gain this for
tion on our particular sensations, for the sensations, that serve as
subjective observation, but we rather gain to a high degree the capacity to overlook these subjective sensations, and to proceed in the estimation of objects independently of them, even where they
could easily enough be noticed by reason of their
strength."
The division into circles of objectivity. What is described here as a merely negative achievement, as an act of overlooking and forgetting, is really that highly positive function of the concept already seen on every side. It is the retention of the identical relations in the vary ing content of presentation, that characterizes every stage, even the
earliest stages, of objectively valid
knowledge.
What
is
absolutely
Helmholtz, Handbuch der Physiolog. Optik, Helmholtz ov. cit.. D. 432.
26, p. 428.
291
changeable falls away from the momentary content, and only that remains, which can be fixed in permanent conceptions. By the
central tendency of thought, a certain circle of experiences, which satisfy certain logical conditions of constancy, is raised out of the
flux of experience
In this
first
and distinguished as the "fixed kernel" of being. construction and characterization of the the
"real,"
relatively fleeting contents, on the other hand, in which no thorough determinateness of experience is expressed, can be left unnoticed. Deeper reflection, however, teaches that these elements also do not
absolutely outside the circle of experience, but that they also claim a place in it in so far as their variation is itself not arbitrary, but subjected to definite rules. Thus the variable itself, by a new
fall
interest of knowledge, is made the object of consideration. This of the "subjective" means an obj edification of a higher knowledge level, in that it discovers an element of determinability in a material,
that was at
first left absolutely undetermined. The given is now divided into wider and narrower circles of objectivity, which are
distinctly separated
and arranged according to

is
definite points of
henceforth determined not only particular experience the material content of the impressions, but through the charac by teristic function by which some experiences serve as a fixed point of
view.
reference for measuring
Each
and interpreting
others.
In this
way we
produce definite, conceptually distinguished centers, around which the phenomena are ordered and divided. The particular phenomena do not flow by uniformly, but are limited and separated from each other; the initial surface-picture gains foreground and background, as it were. The real origin of the concept of object proves to be, not the "projection" of the inner and outer, but the division of phenomena
Each into partial fields, distinguished by their systematic meaning. receives an index, which signifies its position in the whole, particular
and
in this index its objective value is expressed.
it is
For the naive
given from the beginning, and that is view, "thing," expressed and partially copied in each of our perceptions. Thus the naive view also presupposes a whole, with which we compare each
the
that
is
The demand particular experience arid thereby measure its value. made by the naive view still holds from the standpoint of critical
consideration.
The
defect of the nai ve view
is
not that
its
it
makes
this
it
demand, but that
it
confuses the
demand and
fulfillment;
takes the problem, that knowledge has to work out, as already solved.
292

is
The whole, which we seek and on which the concept must not be conceived as an absolute being outside of
experience
;
directed,
all
possible
it is
nothing
else
than the ordered totality of these possible
experiences themselves.
The modern psychology of space the projection theory by another view, perception has replaced which expresses the facts of knowledge more purely and sharply, since
"Projection"
and
"selection."
it
represents According to
them independently
it,
of all metaphysical assumptions.

"objective"
the presentation of
space
is
not a work of
but of "selection;" it rests on an intellectual selection "projection" made among our sense perceptions, especially in the field of visual and tactual impressions. In the homogeneous mass of these impres
sions,
we only
retain those contents, that correspond to
"normal"
physiological conditions; while we more and more suppress others, that arise under extraordinary conditions and lack the same repeat Because our apperception separates a certain ability as the first.
group of experiences in this way from the flux of the others, this group tends to be emphasized. These pass as reality absolutely, while all the other contents only have value in so far as they refer
as
"signs"
to this reality.
It
is
thus no absolute difference in being,
but a difference of emphasis, that separates the objective from the The construction of spatial reality has in it a process of subjective. logical elimination, and without this its result could not be understood.
The mass
ture. 8
of spatial "perceptions" is gradually organized according to a definite plan, and in this organization gains fixed form and struc
From the standpoint of logic, it is of especial interest to trace the function of the concept in this gradual process of construc
tion. Helmholtz touches on this question when he affirms, that even the presentation of a connection of contents in temporal sequence according to law would not be possible without a conceptual rule. "We can obviously learn by experience what other sensations of vision, or some other sense, an object before us would give us, if we should move our eyes or our bodies and view the object from different sides, touch it, etc. The totality of all these possible sensa tions comprehended in a total presentation is our presentation of
the body; this

sensations,
8
we
call
perception
when
it
it is
is
and memory-image when

of
not.
supported by present In a certain sense,

James, Principles of
Cf.
more particularly on the theory

II,
"selection,"
Psychology
237
ff.
293
although contrary to ordinary usage, such a presentation of an individual object is already a concept because it comprehends the whole possible aggregate of particular sensations, that this object can
,
arouse in us
that
cal
when viewed from

9
investigated."
is
Here Helmholtz
different sides, touched or otherwise is led back to a view of the concept
even to him.
foreign to traditional logic and that at first appears paradoxi But in truth the concept appears here in no mere
extravagant and derivative sense, but in its true and original meaning. It was the "serial concept," in distinction from the "generic concept,"
that was decisively revealed in the foundations of the exact sciences, and that, as is now seen, has further applications, proving itself to
be an instrument of objective knowledge.
in
The psycho repetition. logical analysis of the idea of space thus confirms the concept of
The function of judgment; permanence and
objectivity derived from the logical analysis of knowledge. The mysterious transition between two different, essentially separated
spheres of being now disappears; in its place appears the simple problem of the connection of the particular, partial experiences into
an ordered
totality.
The
of
particular content, in order to be called

its
truly objective,
must expand beyond
temporal narrowness and
the total experience. Henceforth the content stands not only for itself but for the laws of this experience, which it brings to representation. The particular moment is the starting-point of an intellectual construction, that determines and
includes in
its
become the expression
narrower and broader consequences the whole of
The procedure of this logical "integration" experiential reality. is discoverable in outline in the simplest judgment about "facts."
Everywhere, even where only a concrete, particular property is ascribed to an individual thing, a connection once established logically This persistence is posited with the very form of judgment, persists. and is found even where the content, on which the judgment is In the simplest schematic characterization directed, is changeable.
of this relation, it is enough to point to the fact, that in any asser tion a is b an element of permanence is included, since a dependency
is
therewith established, and this dependency

3
is
intended to be valid
Handbuch der physiolog. Optik, 2nd ed.,
p. 947
f.
294
not only for a particular instant, but is regarded as identically The transferable to the whole sequence of moments of time. b belongs to the "thing" a not only at a definite point of "property"
time to, in which it is apprehended by the act of perception, but it is maintained for the whole series ti t 2 1 3 ... (Cf. above p. 243 ff., 268.) Thus at first one and the same determination is posited as absolutely
act,
capable of repetition and as fixed in judgment. To this original however, another may be added, in which the change of the
ment has
so
it
As judg particular elements is conceived as logically determined. ascribed to the subject a, at the time ti, the predicate b,
can ascribe to
b",
it
at the time
the predicate
in so far as this
at the time ta ti the predicate 6 change of properties does not take

,
place without rule, but is conditioned and demanded according to law by appropriate changes in another series. Only thus does the general schema for the concept of the empirical "object" result; for the scientific concept of a definite object comprises in its ideal completion not only the totality of its properties given here and now, but also the totality of necessary consequences, that could develop from it under definite circumstances. We connect a series of different, circumstances by means of a unitary complex temporally separate,
of causal rules;
and
this connection gives the individual, in Platonic
language, the seal of being.

differential gains objective
The content
of the particular time-
meaning, can be reconstructed from it according to a definite method. The problem of the ll transsubjective." In contrast to this continuous
process,
in so far as the total experience
by which the
originally
fragmentary and disconnected
experiences are gradually shaped into a system of empirical knowl edge, the metaphysical view at some point necessarily reaches a
chasm, which thought cannot bridge even though it can leap across it. This defect appears most clearly and sensibly exactly where the most strenuous effort is made to break through the limits of the mere
"world
of
presentation,"
in order to reach the
world of real
"things."
When
undertakes to show the inferences, that lead from the field of the subjective (at first the only accessible field), to the realm of the "transsubjective," a barrier between
"transcendental realism"
thought and being is already raised by this way of stating the ques It is taken as an assump tion, that no logical efforts can remove. tion needing no further scrutiny, that all consciousness is first directed on the subjective states of our ego and that nothing except these

states
295
is immediately given. There is a sphere of "immanence," which never goes beyond these first, original data; there is a type of self-consciousness, which limits itself merely to taking in passively the content of particular, present impressions, without adding any new element or logically judging the content according to a definite
conceptual point of view. Only this is asserted, that this first stage, while sufficient for the consciousness of the ego, in no way suffices
to ground the consciousness of the object. The object of natural science especially cannot be treated by these primitive means. The objects of science, "mass" and "energy," "force" and "acceleration"
are strictly
and unmistakably distinguished from all contents of immediate perception. He who grants science the right to speak of
objects and of the causal relations of objects, has thereby already left the circle of immanent being and gone over into the realm of
"transcendence."
The
the
it is
correlation of the consciousness of the ego
and
the consciousness of
object.
One can
fully accept these
conclusions so far;
but
a strange error to believe that they affect not only psychological, presentational idealism but also critical idealism in its ground and
root.
Critical idealism
is
distinguished from the
"realism"
here
advocated, not by denying the intellectual postulate at the basis of these deductions of the concept of objective being, but conversely
by the fact that it grasps this intellectual postulate more sharply and demands it for every phase of knowledge, even the most primi tive. Without logical principles, which go beyond the content of
given impressions, there is as little a consciousness of the ego as there is a consciousness of the object. From the standpoint of critical
the concept of "transcendence" as rather the concept of "immanence," that is here presupposed. The thought of the ego is in no way more original
idealism,
is is
what
to be contested
not so
much
and logically immediate than the thought of the object, since both arise together and can only develop in constant reciprocal relation.
No
known and experienced as "subjective," without contrasted with another content, which appears as objective. being Thus the conditions and presuppositions of (Cf. above p. 273.) as a supplement, after the "objective" experience cannot be added
content can be
subjective world of presentations has been completed, but they are already implied in its construction. The "subjective" signifies only
an abstract moment of a conceptual
distinction,
and as such has no
296

its
independent existence, because

logical correlate
whole meaning
is
rooted in
its
and opposite. The separation of thought and

clear,
experience.
Although
this relation
seems very
it is
when once the metaphysical
difference of subject
and object is changed into a methodological distinction, nevertheless worth while to pause over it. For the kernel of all the miscon
ceptions among the various epistemological tendencies is found here. The deeper reason, that the outer and inner world are opposed as two
heterogeneous
realities,
lies
in
an analogous opposition between
experience and thought. The certainty of pure experience is taken to be entirely different from that of thought. And as both are different in their origin, they are accordingly related each to a
separate sphere of objects, within which they possess unique and It is pure experience, free from any admixture exclusive validity. of the concept, which assures us of the states of our ego; while all
knowledge
of thought.
of the outer object
is really guaranteed by the necessity Inner perception, by which the ego comprehends itself,
has a peculiar and, of its sort, unsurpassable "evidence" (Evidenz); but this "evidence" is purchased at the price of conceiving the content as absolutely individual, and as merely given in its unique properties here and now. When from our presentations, intellectual
"all
removed, and they exist only in a blending of the similar and the like:" then and only then can they be taken as self-evident. Thus the beginning of every theory of knowledge must involve, that we divest ourselves of all connection with the realms of intellect and nature, of all commerce with the common values of civilization, so that we merely retain our individ ual consciousness its whole emptiness and nakedness." Only in this way do we reach a type of certainty, in which thought in no
necessity, all logical order is
"in
way
participates,
certainty, that
postulates,
only to find, that we cannot remain with such a must be broadened by logical assumptions and by means of which we posit an object of knowledge. 10
it
This apparently presuppositionless beginning, however, contains a premise, that is unjustified both from the standpoint of logic and from that of psychology. The division here between perception
and thought negates the concept

objective concept of experience.
10
of consciousness
All consciousness
no less than the demands some

Miinchen 1906;
Cf. Volkelt, Die Quellen der menschlichen Geivissheit. esp. p. 15 ff.; cf. Erfahrung undDenken, Leipzig 1878, Ch. I.

sort of connection
297
and every form
of connection presupposes a rela
tion of the individual to an inclusive whole, presupposes the insertion of the individual content into some systematic totality. However
primitive and undeveloped this system may be conceived to be, it can never wholly disappear without destroying the individual content An absolutely lawless and unordered something of percep itself. tions is a thought, that cannot even be realized as a methodological
fiction; for
the mere possibility of consciousness includes at least the
conceptual anticipation of a possible order, even though the details may not be made out. Therefore if we characterize every element, that goes beyond the mere givenness of the individual sensation, as "transsubjective," then the paradoxical proposition holds, that not
only the certainty of the object, but also the certainty of the subject conceals a "transsubjective" element. For even the mere "judgment
of
perception"
gains
its
meaning only through reference to the system
of empirical judgments, and must therefore acknowledge the intel lectual conditions of this system. (Cf. above p. 245 f.)
The concept of the object in critical idealism. If we determine the object not as an absolute substance beyond all knowledge, but as the object shaped in progressing experience, we find that there is no
"epistemological
gap"
to be laboriously spanned
by some authorita
tive decree of thought, by a "transsubjective command." 11 For this object may be called "transcendent" from the standpoint of a.
psychological individual from the standpoint of logic and its supreme principles, nevertheless it is to be characterized as purely "imma nent." It remains strictly within the sphere, which these principles
;
determine and
cal
limit, especially
and
scientific
knowledge.
the universal principles of mathemati This simple thought alone constitutes
the kernel of critical
"idealism."
When
Volkelt repeatedly urges in
his criticism, that the object is not given in mere sensation, but that 12 it is first gained on the basis of intellectual necessity, he defends the
most
characteristic thesis of this idealism.
The
ideality,
which
is
here alone asserted, has nothing in common with the subjective "presentation;" it concerns merely the objective validity of certain
axioms and norms of

this alone
11
scientific
is
affirmed
knowledge. The truth of the object depends on the truth of these axioms, and has
f., etc.;
Cf. Volkelt, Quellen der menschlichen Gewissheit, p. 46

p. 186
ff.
Erfahrung
und Denken,
12
Quellen der menschlichen Gewissheit, p. 32
ff.,
etc.
298
no other and no firmer basis. It is true that there results, strictly But this relativity speaking, no absolute, but only a relative being. obviously does not mean physical dependency on particular thinking subjects, but logical dependency on the content of certain universal
uct"
The proposition, that being is a "prod principles of all knowledge. of thought, thus contains no reference to any physical or meta
relation,
physical causal relation, but signifies merely a purely functional a relation of superordination and subordination in the
If we analyse the definition of the validity of certain judgments. if we bring to clear consciousness what is assumed in this "object,"
concept, we are led to certain logical necessities, which appear as the inevitable constitutive "factors" of this concept. Experience and its object are conceived as dependent variables, which are succes
sively reduced to a sequence of logical "arguments;" and it pure dependence in content of functions on their arguments,
is
is
this
which
characterized in the language of idealism as the dependence of the
"object"
on
"thought."
(Cf.
above
esp. p.
267
ff.)
pure mathematics. This sort of dependence is so unmistakable, that it is also attested and emphasized by the opposing side. It is finally admitted, that the actual necessity by
objectivity within
The
which we go beyond the circle of particular disconnected sensations and rise to the conception of continuous objects connected by strict
is in the name of causal rules, is ultimately a logical necessity. that the certainty of actual necessity dominates me and com reason, pels me to make transsubjective assumptions. Everything that we
"It
call
judgment,
reflection,
if
thought, understanding, reason, science

to this
certainty."
would seem uprooted
we acted contrary
is
By
validity of being, nothing
subjective significance, which virtue of intellectual necessity." 13

sal rules of reason, that
to be understood except "the transwe give to the contents of judgment by
Thus it is according to the univer we construct the concept of being and deter
every sort This limita
mine
of
it
in detail.
The
justification, as well as the limits of
thereby exactly characterized. tion of transcendence is most clearly seen when we

"transcendence," is
compare the
object of experience with the object of pure mathematics.
The
object of pure mathematics also can in no way be reduced to a complex of sensations it is also characteristic for it to transcend the
;
given in an intellectual construction, that has no direct correlate in

13
Volkelt, Die Quellen der menschlichen Gewissheit, p. 33 and 37.

any particular content
of presentation.
299
yet the objects of
And
mathematical knowledge, the numbers and the pure forms of geom etry, do not constitute a field of self-subsisting, absolute existences, but are only the expression of certain universal and necessary ideal connections. If this insight is once gained, it can be carried over to the objects of physics, which, as we have seen, are nothing but the result of a logical work, in which we progressively transform experi ence according to the demands of the mathematical concept. The "transcendence," which we ascribe to the physical objects, in dis tinction from the flowing, changing content of perception, is of the same sort and rests on the same principles as the distinction between the mathematical idea of the triangle or of the circle and the particular intuitive image, by which it is represented here and now in a real In both cases, the momentary sensuous image is presentation. raised to a new logical significance and permanence; but no entirely
is grasped by us by virtue of this division in either but a new character of conceptual necessity is imprinted on cer case, tain contents. The same conditions, that were fundamental in the transition from the empirical data of touch and vision to the pure forms of geometry, are necessary and sufficient for the trans formation of the content of mere perception into the world of empiri cal-physical masses and movements. Here, as there, a constant standard of measurement is introduced, to which the changeable is henceforth referred; and on this fundamental function rests the
foreign being
positing of
any sort of objectivity. Thus is justified when it urges, that what makes a judg ment a judgment, and what makes knowledge knowledge, is not
"realism"
something given, but something that is added to the given. "We mean anything, if we were limited merely to the given; for all attempts to remain purely within the given in the case of
could never
meaning, of judgment, become tautologies and lead to meaningless propositions. Judgment and knowledge go beyond the given in their meaning; what is meant in them is transcendent to the given and thus
to
them
in so far as
psychical contents. Each thought ent to itself, in so far as it can never

tions are entirely satisfactory; but
14
they are considered only as given, as present is thus transcend mean itself." 14 These proposi
....
only a slight change in their
W.
Freytag, Der Realismus und das Transzendenzproblem, Halle 1902,
p. 123.
300
formulation
is needed in order to draw an entirely different conse quence from them than the one here drawn. If all thought is really "transcendent to itself," if it belongs to its original function not to remain with the present sensations but to go beyond them, then the converse inference holds. The "transcendence," which can be grounded and proved by thought, is no other than that which is 15 posited and guaranteed in the fundamental function of judgment.
The
"object"
is
just as
much and
just as little transcendent as
judgment
Thereby, however, the correlation of knowledge and the object is admitted in the critical sense; for, much as judgment transcends the mere content of present, sensuous perception, we can not affirm that it stands beyond the logical principles of knowledge. It was dependence on these principles, and not on any concrete psychic contents or acts, that methodic idealism alone represented
is.
and demanded. "Immanence" in the sense of psychologism must indeed be overcome in order to reach the concept of the physical object; but this object itself, while it transcends the sphere of sensa tion, gains its existence in conceptual relations, from which its essence and its definition are inseparable. To the psychological immanence
is opposed, not the metaphysical transcendence of but rather the logical universality of the supreme principles things, of knowledge. The fact must be granted unconditionally, that the
of impressions
"presentation" reaches beyond itself, and that all that is means something that is not directly found in itself 16 but it has given already been shown that there is no element in this "representation," which leads beyond experience as a total system. Each particular
particular
member
of experience possesses a symbolic character, in so far as the law of the whole, which includes the totality of members, is posited and intended in it. The particular appears as a differential, that is not fully determined and intelligible without reference to its integral.
15
"in
this general conviction of the objective Cf. Freytag, op. cit., p. 126: nature of truth, however, the transcendence of judgment is involved as a neces sary presupposition. For if judgment were not transcendent, had it no mean ing going beyond what is given in it, did all its meaning consist in what it is as a psychic process, then the truth of the judgment would be directly estab lished; whether I judge a is b or a is not b, each judgment would be correct in itself, because in the first precisely the a was meant, which is actually judged to be b, and thus would also have been given; in the second the a is meant, that is actually judged as not being 6, and therefore would be also thus given."
16
Cf. e.g.
Uphues, Kant und seine Vorganger, Berlin
1906, p. 336.

Metaphysical
"realism"
301
misunderstands this
here argued,
it is,
shift of logical
meaning
else
when
is
it
conceives
it
as a sort of transubstantiation.
it is
"must
"Whatever
to
mean anything/
it is;
17
mean something
than
for
what
this
it is,
that
and therefore does not need to
mean."
But
"other"
heterogeneous; rather
relation brings
it
we
need not at all be something actually are here concerned with a relation between
different empirical contents,
about, that
which belong to a common order. This we can pass from a given starting-point
through the whole of experience in a progress according to rule, but not that we can pass beyond experience. The constant reaching out beyond any given, particular content is itself a fundamental function of knowledge, which is satisfied in the field of the objects of knowledge. Among the philosophical physicists, it is Fechner, who--, has clearly grasped the problem involved here. "The fact, that the ^V world of phenomena is such that one phenomenon can exist onlyîn and through another, can easily lead one (and has led some) to deny real existence to phenomena in general; and can lead them to assume, as a basis for the varying multiplicity of phenomena, independently existing, fixed things in themselves behind the phenomena; these things in themselves can never appear as they are among phenomena, but they rather produce the whole dependent appearance of phenom ena either by external interaction among themselves or by their inner working. For it is said: if one phenomenon is always only to appeal to another with reference to the ground of its existence, then a ground for all existence is ultimately lacking; if A says, I can only
B exists/ and B in return, I can only exist in so far as But then both of them are founded on nothing instead of seeking for the ground of the existence of A and B (which neither can find in the other) in something behind them, that gives a ground and kernel to their appearance, it is to be sought in the
exist in so far as
A exists/
....
totality, of
which they are both members; the whole

all
is
the basis and
In the whole, not to besought in something particular, in something other behind the individual, concerning whose ground we would have to ask anew; but we can investigate by what rules the individual is united to the whole, and what are the ultimate
that
is
kernel of the whole and of
in
it
the basis of the individual
is
elements
What we can
thing never consists in an

17
find objective in a material obscure thing behind and independent of
E. v. Hartmann, Das Grundproblem der Erkenntnistheorie, p. 49.
302
perceptions or phenomena, but consists of a unified connection according to law of the same, of which connection each phenomenon
but which connection goes beyond the particular 18 perceptions or the particular phenomena, which the thing affords." Although these propositions draw the line between metaphysics and
realizes a part,
physics very distinctly, Fechner still betrays an inner obscurity in the definition of the object of physics. The unity of the physical world. In order to avoid the conception
of matter as a completely unknown, indeterminate something, at the basis" of the sensuously perceivable properties, which Fechner defines matter by these very sensible properties the matter of the physicist is entire agreement with the most common usage;" Matter is thus it is nothing else than what is perceptible by touch. equivalent to the "tangible." What may lie behind the data of touch and feeling need not concern the physicist; to him, tangibility alone is able to be pointed out, to be grasped by experience and to be traced further; and this suffices to give the concept the fixed founda tions necessary for his purposes. 19 Here the attempt to exclude the metaphysical elements of the concept of matter has again led to
"lies
;
"in
setting aside the logical element, that

critical
is
characteristic of
it.
The
view stands between the two views.
It defines the object
by relation to the "whole of experience;" but aware that this whole can never be represented and grounded as a mere sum of particular sense data. The whole gains its form and system only by the assumption of original relations, of which no one can be pointed out as "tangible" like a given sensuous content; and it is one of the manifold expressions of these relations, that is
of natural science
it is
also
presented in the concept of matter, as also in that of force or energy. (Cf. above p. 169 f.)
IV
The
historical transformation of the
"thing."
The reduction
of the
concept of thing to a supreme ordering concept of experience disposes of a dangerous barrier to the progress of knowledge. For the first na ive view of reality, it is true, the concept of thing offers no riddle or
difficulty.
18
Thought does not need to reach things gradually and by

Uber die physikalische und philosophische Atomenlehre. 2nd
ff.
Fechner,
ed., Leipzig 1864, p. Ill

19
Op.
cit.,
p. 106 f
303
complicated inferences; but possesses them immediately, and can grasp them just as our bodily organs of touch grasp the corporeal But this naive confidence is soon shattered. The impres object.
sion of the object and the object itself are separated from each other; instead of identity, the relation of representation appears. No
matter how complete our knowledge may be in itself, it never offers us the objects themselves, but only signs of them and their reciprocal relations. An increasing number of determinations, that were formerly taken as belonging to being itself, are now changed into mere expressions of being. Following the well-known path of the history of metaphysics, the thing is to be conceived as free not only from all the properties belonging to our immediate sensations (for the thing is in itself neither luminous nor odorous, neither colored nor sounding), but also as having all spatial-temporal properties, such as plurality and number, change and causality, stripped from
is known, all that is knowable, is set over in opposition the absolute being of the object. The same ground, that against assures us of the existence of things, shows them to be incomprehen
it.
All that
sible.
Skepticism and mysticism are united in this one point.
No
matter how
ence
many new
relations of
"phenomena"
scientific experi
teach us, the real objects seem not so much revealed as rather the more deeply concealed by these relations.
may
All this doubt and suspicion disappear, however, when we reflect that what appears here as the uncomprehended residuum of knowl edge really enters into all knowledge as an indispensable factor and
necessary condition. To know a content means to make it an object by raising it out of the mere status of givenness and granting it a
certain logical constancy and necessity. Thus we do not know as if they were already independently determined and "objects"
given as
objects,
but we know
objectively,
by producing
certain
fixating certain permanent elements and con nections within the uniform flow of experience. The concept of the
limitations
and by
object in this sense constitutes no ultimate limit of knowledge, but is rather the fundamental instrument, by which all that has become
permanent possession is expressed and established. The object marks the logical possession of knowledge, and not a dark beyond is thus no longer forever removed from knowledge. The
its
"thing"
something unknown, lying before us as a bare material, but is an expression of the form and manner of conceiving. What metaphysics
304
ascribes as a property to things in themselves now proves to be a necessary element in the process of objectification. While in meta physics the permanence and continuous existence of objects is
spoken of as distinguishing them from the changeableness and dis continuity of sense perceptions, here identity and continuity appear as postulates, which serve as general lines of direction for the progres sive unification of laws. They signify not so much the known proper ties of things, but rather the logical instrument, by which we know. In this connection the peculiar changeableness is explained, that is
manifest in the content of scientific concepts of objects. According as the function of objectivity, which is unitary in its purpose and
nature,
is
realized in different empirical material, there arise different
concepts of physical reality; yet these latter only represent different stages in the fulfillment of the same fundamental demand. Merely
this
Helmholtz
demand is unchanging, not the means by which it is satisfied. Thus natural science, even where theory of signs.
it
retains the concept of the absolute object, can find no other expres sion for its import than the purely formal relations at the basis of
experience.
This fact appears in significant form in Helmholtz
theory of signs, which is a typical formulation of the general theory of knowledge held by natural science. Our sensations and presenta
tions are signs, not copies of objects. From a copy we demand some sort of similarity with the object copied, but we can never be sure of
our presentations. The sign, on the contrary, does not require any actual similarity in the elements, but only a functional correspondence of the two structures. What is retained
this in the case of
in
it is
relations, in
t
not the special character of the signified thing, but the objective which it stands to others like it. The manifold of sen-
is correlated with the manifold of real objects in such a way, that each connection, which can be established in one group, indicates a connection in the other. Thus by our presentations we do not know, indeed, the real absolutely in its isolated, self-existent proper
sations
but we rather know the rules under which this real stands and in it changes. What we discover clearly and as a fact without any hypothetical addition is the law in the phenomenon; and this law is a condition of our comprehending the phenomenon, and is the only property, which we can carry over directly to things themselves. 20 But in this view also no entirely new content is
ties,
accordance with which
20
Helmholtz, Handbuch der physiolog. Optik, 2nd
ed., p. 586
ff.,
947
f.
305
fact.
posited, but only a double expression is produced for one and the same The lawfulness of the real means ultimately nothing more and
nothing else than the reality of the laws; and these exist in their unchanging validity for all experience, in abstraction from all particu
When we declare connections, that at first could appear as mere regularities of sensation, to be laws of things, we have merely produced a new designation for the universal meaning
lar limiting conditions.
The known facts are not changed in their nature our choosing this form of expression, but are merely strengthened by in their objective truth. Thinghood is always only such a formula
ascribed to them.
of confirmation and,
of the empirical has no significance. The accredited, objects of physics are thus, in their connection according to law, not so much "signs of something objective" as rather objective signs, that
when separated from the whole

it
is
connections by which
satisfy certain conceptual conditions
and demands.
It follows of
The
itself
logical
and
the ontological conceptions of relativity.
that
we never know
mutual permanence and change.

in their
things as they are for themselves, but only relations: that we can only establish their relations of
But this proposition involves none of the that are connected with it in realistic meta skeptical consequences If we proceed from the existence of absolute elements, then physics. it must indeed appear as a defect of thought, that this existence can
pure and separate form. According to this themselves; but they are only known to us in view,^hmgs_ea;isf their interactions, and their interactions influence and obscure the
never be mastered in
its
.for-
nature
of
each.
Helmholtz
formulates
his
view
thus:
its
"Each
property or -quality of a thing is in reality nothing but produce certain effects on other things
effect
capacity to
call
We
such an
a property when we hold the reagent, on which it works, in mind as self-evident without naming it. Thus we speak of the
solubility of a substance, that is its reaction to water; we speak of its weight, that is its attraction for the earth; and with the same justifi cation we call it blue, since we thereby presuppose as self-evident, that
we
If,
are merely concerned with defining its effect on a normal eye, however, what we call a property always involves relation between two things, then such an action naturally never depends on the nature of one of the agents alone, but always exists only in relation to and in dependence upon the nature of a second being, on which
306
the effect
is

exerted.
^1
These propositions, with their striking
formulation of the general principle of relativity, have been appealed to as a ground for demanding the exclusion in principle of all ontolog-
Yet in truth, they too contain an unmistakable ontological element. A closer interpretation of the principle of the relativity of knowledge does not find this principle to be a mere consequence of the universal interaction of things, but
ical
elements from natural science. 22
recognizes in
itself.
it
This
Its
is
a preliminary condition for the concept of thing the most general and most radical meaning of relativy
is
not that we can only grasp the relations between elements of being by conceiving these elements still as obscure, Vself-existent kernels, but rather that we can only reach the category
itv.
meaning
of thing
through the category of relation.
We
do not grasp the
relations of absolute things from their interaction, but we concen trate our knowledge of empirical connections into judgments, to which
we
ascribe objective validity.
Therefore the
"relative"
properties
do not signify in a negative sense that residuum of things, that we are able to grasp, but they are the first and positive ground of the concept of reality. There is a circle in attempting to explain the
relativity of
for this interaction
knowledge from the thorough-going interaction of things, is rather only one of those ideas of relation, that
knowledge posits in the sensuous manifold in order to unify it. The unity of the scientific view of the world. It is of especial interest to follow the manner in which this view gains methodic clarity and
sharpness within modern physics. A characteristic example of it is lately offered by a prominent physicist in his exposition of the
progress and general goals of the physical method.
In a short sketch
in his treatment of the unity of the physical world, Planck has defined the general points of view, that account for the continual
transformation of physical theories. definitions is marked by the fact that
If
the
we
of sensation directly in the concept, all sists in removing this dependence more
first stage of our physical try to reproduce the content further logical progress con
and more. Sensation as such contains an anthropomorphic element, in so far as it neces21
und
22
Helmholtz, Die neueren Fortschritte in der Theorie des Sehens. (Vortrage JKeden, 4th ed., Braunschweig 1896, p. 321), cf. Physiologische Optik*,
p. 589.
Cf. Stallo, The Concepts and Theories of Modern Physics, Leipzig 1901, p. 131, 186 ff.
German
ed.
307
sarily involves a relation to a definite sense-organ, thus to a specific
human organism. How this anthro constantly forced into the background, so that of this, the history it entirely disappears in the ideal plan of physics of natural science furnishes a single continuous example. 23 It
physiological structure of the
pomorphic element
is
must now be asked, what compensation

scheme
tage soon
is
is
offered
by the
scientific
advan and necessity? It is made clear that the required compensation cannot be based on anything material, but purely on a formal moment. While science renounces the wealth and varied multiplicity of immediate
of the
world for these
lost contents?
its
What
positive
there at the basis of
significance
sensation,
it
regains
its
apparent
loss in content
through increased
inner heterogeneity of the impres unity and completeness. sions has disappeared along with their individual particularity, so that fields, that are absolutely incomparable from the standpoint of
sensation, can
The
and
in reciprocal connection.
lies
now be conceived as members of the same total plan The peculiar value of the scientific
in
construction
this alone: that in the scientific construction
there appears as connected by continuous conceptual intermediates, what in the first naive view appears in mere conjunction and unre
this tendency is carried through, the more we look more closely, the investigation fulfills its task. perfectly we see that the old system of physics was not like a single picture, lated.
The more purely
"If
but rather a collection of pictures; there was a special picture for each class of natural phenomena. And these various pictures did not hang together; one of them could be removed without affecting the others. That will not be possible in the future physical picture of the world. No single feature will be able to be left aside as un each is rather an indispensable element of the whole and essential; as such has a definite meaning for observable nature; and, conversely, each observable physical phenomenon will and must find its cor
responding place in the picture." We see that the characteristics of the true physical theory, as developed here, entirely coincide with
See Planck, Die Einheit des physikalischen Weltbildes, Leipzig 1909. of the development of the conceptual constructions of natural science in the fourth chapter (cf. esp. p. 164 ff .) was already completed when Planck s lecture appeared. It is so much the more welcome to me be cause our own result is confirmed in all essential points in the philosophical part of Planck s treatment and is thereby illumined from another point of view.
23
The exposition
308
the criteria of empirical reality, as shown by epistemological analysis. Planck requires as the condition of any physical theory: "unity in respect of all features of the picture, unity in respect of all places and
times, unity with regard to all investigators, all nations, all civiliza tions." The realization of all these demands is what constitutes the
real
meaning
of the concept of the object.
In opposition to the
phenomenalistic view, which remains with the data of mere sensation, Planck can justly characterize his view as "realistic." This "real ism," however, is not the opposite but the correlate of a rightly understood logical idealism. For the independence of the physical object of all particularities of sensation shows clearly its connection with the universal logical principles; and it is only with reference to these principles of the unity and continuity of knowledge, that the
content of the concept of the object
is
established.
CHAPTER
SUBJECTIVITY
VII
OF
AND
OBJECTIVITY
THE
RELATIONAL
CONCEPTS
The problem of
the subjectivity and objectivity of relational concepts. analysis of knowledge ends in certain fundamental relations, which rests the content of all experience. Thought cannot go
The
on
further than to these universal relations; for only within
them are
thought and
its
seem as
if
we
object possible. And yet with this answer it may have moved in a circle. The end of the enquiry leads
us back to the same point at which we began. The problem seems shifted, but not solved; for the opposition of the subjective ancTthe
objective
fall
still
,.
persists as sharply as ever.
The pure
relations also
under the same question, that was previously directed upon and presentations. Are pure relatiqji8"an element of Is the nature of being, or are they mere constructions of thought? things revealed in them, or are they only the universal forms of expression of our consciousness, and thus valid only for consciousness and the sphere of its contents? Or does there exist a mysterious pre-established harmony between mind and reality, by virtue of which both must necessarily agree ultimately in fundamental
sensations
character?
The universal functions of rational and empirical knowledge. But to state the problem in this way shows that it has been over come in principle as the result of the preceding enquiry. At any rate, "common" ground for resolving the opposition of
thought and being exists. Yet this common ground is not to be sought in an absolute ground of all things in general, but merely in the universal functions of rational and empirical knowledge. These functions themselves form a fixed system of conditions; and only relative to this system do assertions concerning the object, as well as concerning the ego or subject, gain an intelligible meaning. There is
no objectivity outside of the frame of number and magnitude, per manence and change, causality and interaction: all these determina tions are only the ultimate invariants of experience itself, and thus The same point of all reality, that can be established in it and by it.
309
310
Without a temporal of view directly includes consciousness itself. sequence and order of contents, without the possibility of combining them into certain unities and of separating them again into certain
pluralities, finally
without the possibility of distinguishing relatively
constant conditions from relatively changing ones, the conception of the ego has no meaning or application. Analysis teaches us very
definitely that all these relational forms enter into the t5nce]pt of as into that of "thought;" but it never shows i(s how .-they "being"
In fact, every are combined, nor whence they have their origin. as to this origin, every reduction of the fundamental forms question to an action of things or to an activity of mind would involve an obvious petitio principii; for the "whence" is itself only a certain
form
of logical relation.
tion, all
With
If causality is once understood as a rela as to the causality of relations in general disappears. question regard to relations in general, we can only ask what they are
meaning, and not as to the manner in which and the from which they have arisen.__^fter J;hes^j^]^tiojQS._ai:e.beginnings established in their meaning, we can seek out the origin of particular objects and processes with their help and with the guidance of experi It is hopeless, on the contrary, to seek to trace them to more ence. as is done with ultimate beginnings, to psychical or physical
in their logical
"forces,"
a changing empirical existence.
The reciprocal
relation of the
"form"
and
"matter"
Thus the
possibility disappears of separating the

its
"form"
"matter"
of knowledge. of knowl
edge from
by
referring
them each
to a different origin in
absolute being; as when, for example, we seek the origin of one factor in "things," and of the other in the unity of consciousness. 1 For all
the
determinateness, that
we can
ascribe
to
the
"matter"
of
knowledge, belongs to it only relatively to some possible order and thus to a formal serial concept. The particular, qualitatively specific sensation gains its character only by distinction from other contents of consciousness; it exists only as a serial member and only as such
result not
can be truly conceived. Abstraction from this condition would merely in a greater or less "indefiniteness" of its content,
1 Cf. Riehl, Der philosophische Kriticismus (esp. II, 1, p. 285 ff.) and also the treatment by Honigswald, Beilrage zur Erkenntnistheorie u. Melhodenlehre, Lpz. 1906. On the following, cf. my criticism of this writing, Kant Studien XIV, p. 91-98; on the problem of constant magnitudes, cf. above p. 230 ff.
SUBJECTIVITY AND OBJECTIVITY OF CONCEPTS

but would absolutely annihilate
lation withstands every
it.
2
311
This indissoluble logical corre
attempt to explain the relation here by two causal factors, which are assumed to exist and to act in separate themselves. Matter is only with reference to form, while form, on
the other hand,
is valid
only in relation to matter.
If
we
abstract
from their correlation, neither matter nor form has any "existence," concerning whose ground and origin we could enquire. The material particularity of the empirical content can thus never be adduced as proof of the dependence of all objective knowledge on an absolutely transcendent" ground for this determinateness, which as such undeni ably exists, is nothing but a characteristic of knowledge itself, by which the concept of knowledge is first completed. If we give it its most purely scientific expression, it means ultimately nothing but the proposition, that universal rules of connection and universal equations of natural phenomena do not suffice for the construction of experience and the constitution of its object; but the knowledge of
;
particular constants is also needed, such as can only be gained by experimental observation. Still it is not evident how these con
stants attest to
experience, foundations.
more than the empirical
reality of the objects of

its
or
how they
reveal anything regarding

of a
absolute
law presupposes this law and is only intelligible with reference to it; hence the particular, fixed value always remains in the sphere of that concept of being, that is defined and limited by the universal principles of mathematics.
This limitation, however, constitutes the true "ideality" of a value. Such ideality cannot be established by correlation with the presenta tions and acts of thought of psychological individuals, but only by
correlation with the universal principles truth. (See above p. 296 f.)
For the particularization
and conditions
of scientific
The existence of the êternal truths" But while the question as to the metaphysical origin of these conditions proves to be a miscon ception, if the problem as to whether they are to be deduced from the
mind
or from things or from an interaction between the two comes to nothing, the old opposition of the "subjective" and the "objec tive" has not yet been removed in every sense. On the contrary,
seems to press anew for solution when we enquire as to what are the instruments of knowledge, what are the forms of judgment and
it
2
Cf.
now
especially:
G. F. Lipps, Mythenbildung und Erkenntnis, Lpz.
1907, p. 154
ff.
312

which we can present to ourselves temporally,
of relating thought, in
pure timeless validity of the ideal to reject this question entirely also, principles. tempted for the sake of the strictness and purity of the logical foundation.
in actual empirical experience, the
We may be
Leibniz declares (in close conjunction with Plato), that the

truths"
"eternal
are valid in complete independence of every fact of reality, whatever its properties may be. These eternal truths merely repre
sent hypothetical systems of consequences; they connect the validity of certain conclusions with the validity of certain premises, without
considering whether examples of these abstract connections can be in the world of empirical things, indeed, without asking whether there are individuals in whose thought the transition from the
found
premises to the consequences is ever actually made. The truths of pure arithmetic would remain what they are, even if there was noth 3 ing that could be counted nor anyone who knew how to count.
In the real classics of idealism, the negation of any merely psychologi cal foundation is expressed with the greatest clarity. These classi
cal writers all incline to the same thought, that has found paradoxical expression in Bolzano s conception of a realm of "propositions and truths in themselves." The "subsistence" of truths is logically
independent of the fact of their being thought.
For example, the
meaning
of the propositions of pure geometry, as they issue from each other in a strict, necessary sequence, thus forming an ideal
whole, can be perfectly deduced and expressed without going back to the psychological acts, in which we bring the content of these propo sitions to intuitive or conceptual presentation. No matter whether
these acts are different in different individuals or whether they are uniform and have constant properties, in no case is it these proper
ties,
that
lines,
surfaces
is
we mean, when we speak of the objects of geometry, of and angles. The being, that we ascribe to these
no sort of temporal
reality,
is
such as belongs to concrete merely their reciprocal being determined; it affirms an objective dependency in the realm of what is thought of, but no actual causal connection in the realm of thinking. The concept of truth of modern mathematics. It is especially the modern extension of mathematics, that has fully revealed this fact
objects
physical or psychical contents, but
See Leibniz, Juris et aequi elementa (Mollat, Mitteilungen aus Leibnizens ungedruckten Schriften. Lpz. 1893, p. 21 f; cf. my ed. of Leibniz Hauptschriften zur Crundlegung der Philosophic. Lpz. 1904 ff., II, p. 504 f.)
313
and has thereby prepared the ground anew for the logical theory that rests upon it. These structures, with which the general theory of the manifold is concerned, are the true and complete objects of mathematics, and their conceptions first represents mathematics in its full extension. These structures can be fully developed, however, without going into the complex and difficult psychological problems as to in what intellectual processes we present to ourselves the mean ing of the infinite totalities here under consideration. Further,
since all the properties of these groups are fixed by their original concept and belong to them in a necessary, unalterable way, there is no room left for any arbitrary activity of thought; on the contrary,
thought is resolved entirely into its object and is determined and as a modern representative of mathematical guided by it. it as you will, there is a world that is peopled with logic says, "phrase
"For,"
ideas, ensembles, propositions, relations,
and implications,
in endless
variety and
multiplicity, in structure ranging from the very simple to the endlessly intricate and complicate. That world is not the
product but the object, not the creature but the quarry of thought, the entities composing it propositions, for example, being no more identical with thinking them than wine is identical with the
drinking of it. The constitution of that extra-personal world, its intimate ontological make-up, is logic in its essential character and
substance as an independent and extra-personal form of being. Just as the astronomer, the physicist, the geologist, or
. .
other student of objective science looks abroad in the world of sense, so, not metaphorically speaking but literally, the mind of the mathe
matician goes forth into the universe of logic in quest of the things that are there; exploring the heights and depths for facts ideas,
These propositions define the problem very sharply, on the positive as well as on the negative side. The necessity of universal mathematical connections must remain undisputed; and this necessity constitutes a peculiar entity, an objective content, which is set over against the psychological activity But does this content of of thought as an absolutely binding norm. truth stand on the same plane as sensuous reality, of which we can of the mathema have merely empirical knowledge? Are the tician nothing else, and have they no other meaning than those, for
classes, relationships, implications."
"facts"
C.
J.
Keyser, Mathematics
1907, p. 25
f.
a lecture delivered at Columbia University,
New York,
314
example, which the comparative anatomist and zoologist establish The in the description and comparison of various bodily structures? very logic of mathematics and of mathematical physics definitely
forbids
any such
identification of the exact
and the
descriptive
methods.
Necessities cannot be simply described, cannot as such be absolutely found holds only for the moment for "found;" for what is merely
which
it is
The question
hend these
I
established, and thus signifies an empirically unique fact. as to the intellectual operations, in which we compre
necessities,
thus
presses
anew
for
solution.
These
operations can indeed never blend indistinguishably into what is known by them the laws of the known are not the same as those of
;
.^knowing.
Nevertheless, the two types of law are related to each other, in so far as they represent two different aspects of a general problem. Thus there exists a deeper and more intimate mutual
between the object and the operation of thought than between the wine and the drinking of the wine. The wine and the drinking are not exactly correlated but every pure act of cognition is directed
relation
;
on an objective truth, which it opposes to itself, while on the other hand, the truth can only be brought to consciousness by these acts of cognition and through their mediation.
ii
The relational concepts and the activity of the ego. Hence it is necessary, on the basis of the concept of "objectivity" gained from the analysis of scientific principles, to define the concept of "sub
jectivity"
in a
new
sense.
The
object, resulting
from such
analysis, has implicit in
general characterization of the it the answer to
which we reach
out decisively.
the question as to what are the intellectual means and methods by this knowledge. One aspect above all stands
As long as the object was simply the "thing" in the customary sense of naive dogmatism, the particular "impres sion" or mere sum of such impressions might grasp it and copy it.
But
this sort of appropriation
is
denied,
when the
validity of certain
logical relations is established as the necessary condition and real kernel of the concept of the object. For pure relations can never
be represented in mere sensuous impressions; the similarity or dis similarity, the identity or difference of the seen or touched, is never
itself
6
something that
is
seen or touched. 5
Everywhere we must go
Cf.
more particularly Ch. VIII.
315
back from the passive sensation to the activity of judgment, in which alone the concept of logical connection and of logical truth is
adequately expressed. We may conceive the idea of the thing as a complex of sensuous properties to have arisen through these properties having been perceived for themselves, and having com bined of themselves by virtue of an automatic mechanism of "associa
tion."
But necessary connection,
in order to be defined psychologi
an independent activity of consciousness. The progress of the judgment according to law is the correlate of the unification of relations according to law in the object of knowledge. Constancy and change in knowledge. The content of truth, and with it the content of seems once more to be in flux; for according to this whole view, we can only realize what a certain
cally, requires reference to
"being,"
truth
by intellectually regenerating it, i.e., by allowing it to develop before us out of its particular conditions. But this "genetic" view of knowledge is not opposed to the demand for permanence.
"is"
For the activity of thought, to which we here recur, is not arbitrary, but a strictly regulated and constrained activity. The functional activity of thought finds its support in the id&^structure of what is thought, a structure that belongs to it once for all independently of any particular, temporally limited act of thought. The two elements of structure and function in their interpenetration determine
the complete concept of knowledge. All of our intellectual opera tions are directed on the idea of a "fixed and permanent" realm of
It is evident, therefore, that all objectively necessary relations. knowledge contains a static and a dynamic motive, and only in their
unification
is
its
in a succession of logical acts, in

successively-,
r
concept realized. Knowledge realizes itself only a series that must be run through
-so^
that
we may become aware
But
if
this series^ is to
of the rule of itsjjrogress. be grasped as a unity, as an expression of an
Identical reality/ which is defined the more exactly the further we advance, then we must conceive the series as converging toward an
ideal limit.
This limit
it is
although for us
and exists in definite determinateness, not attainable save by means of the particular
"is"
members
view
of the series
and
their
change according to law.
A different
results, according as we choose our standpoint at the conceived limit or within the series and its progress yet each of the two aspects
;
requires the other for its completion. The change is directed toward constancy, while the constancy, on the other hand, can only come to consciousness in the change. There is no act of knowledge, which
L
316
is
not directed on some fixed content of relations as its real object; while, on the other hand, this content can only be verified and com
prehended in acts of knowledge.
/The independence
o/L2aaicaL,laj.ths
qf the
thinkinciafrsubrectr*
At
dominating present opistemologiOn the one hand, attempt is cal discussion divide most distinctly. made to maintain the pure objectivity of the logical and the mathe
matical,
this point, the general tendencies
by giving up
If
all
"thinking mind."
we analyse the
reference in principle to thought and to the ideal structure of mathematics
and
if
we
clearly represent the

it is
whole of
is
its definitions,
axioms and
con
theorems,
ject to
stants"
urged,
this
we do not
find the concept of the thinking sub
which
whole system
given,
among
the
"logical
that ultimately remain. Therefore, this concept of the subject does not belong to the field of pure logic and mathematics; it is rather to be accounted one of those "entirely meaningless"
by the help of philoso lacking between the ideal truths of mathematics and logic and the activity of thought; it is rather em phasized that the mind, in grasping these truths, receives them only
phy.
6
conceptions, that have only gotten into science
Thus
all closer
relation
is
The mind is as completely passive receptively as given material. in cognizing a definite system of inferences as sense is according to
the ordinary view
all
in the perception of
sensuous objects. 7
of being
"In
short,
mere delusion; arithmetic must be discovered in just the same sense in which Co lumbus discovered the West Indies, and we no more create numbers than he created the Indians. The number 2 is not purely mental, but is an entity which may be thought of. Whatever can be thought of has being, and its being is a precondition, not a result, of its being 8 The "objectivity" of pure concepts and truths is thought accordingly put on a plane with that of physical things. Neverthe less, the difference between them is sharply revealed again, when we recall that in the sphere of logic and mathematics we cannot
knowledge must be recognition, on pain
of."
reach absolute
but always only relative objects.v ,_Not "objects," but rather the numbers constitute a true "entity." Here number,
The Principles of Mathematics, I, p. 4: Philosophy asks of Mathe it mean? Mathematics in the past was unable to answer, and Philosophy answered by introducing the totally irrelevant notion of mind. But now Mathematics is able to answer, so far at least as to reduce the whole of its propositions to certain fundamental notions of logic.
Cf. Russell,
matics:
What
does
7 8
Russell, op.
cit.,
37, p. 33.
Russell, op.
cit.,
427, p. 451.

..
.the individual gains its
meaning and content only from the whole; but this whole cannot be presented like a quiescent object of per ception, but, in order to be truly aipvftyed, must be grasped and
in the
_ ^determined
the
number
series as
nature, there is is considered sufficient for the perception of a particular thing of

sense), but always a manifold of such acts,
law of it^construction,. In order to comprehend a series and thus to penetrate into its systematic needed not merely a single apperceptive act (such as
which reciprocally con
Thus a movement of thought is always demanded, yet a movement which is no mere change of presentation, but in which what is first gained is retained and made the starting-point of new developments. Thus from the activity itself flows the recog
dition each other.
body of truths. Within the act of production, there arises for thought a permanent logical product, in so far as thought is aware that this act does not proceed arbitrarily but ac
nition of a fixed
cording to constant rules, which tainty and definiteness.
it
cannot avoid
if it is
to gain cer
The problem of pragmatism. The "spontaneity" of thought is thus not the opposite but the necessary correlate of "objectivity," which can only be reached by means of it. If this fundamental
between spontaneity and objectivity is not fully grasped, a single element of it is stressed one-sidedly, a reaction will ensue such as to endanger the constancy of logic itself. From this general motive, we can perhaps most easily understand the attack, that has been made by "pragmatism" on "pure logic." In so far indeed
relation
if
as pragmatism consists of nothing but the identification of the con cepts "truth" and "utility," we can confidently abandon it to the
general fate of philosophical catch-words. What has been advanced in defence of this view is clad almost entirely in a rhetorical and
when we attempt to translate it The very concept of utility defies all attempt at exact definition: now it is the particular individual with his special wishes and inclinations, now it is some common,
polemical style, and melts away into the sober language of logic.
generic structure of
man, with reference to which

first
utility is defined
and measured.
The
case leaves unsolved precisely the decisive
problem, the possibility of exact scientific knowledge; for a science of nature is no more constructed from individual feelings and tendencies than from individual sensations, and is rather directed on the exclu
sion of all purely
"anthropomorphic"
elements from
its
system of the
318
world.
(Cf.

above
p.
306
is
ff.)
In the second case, a
constant
psycho-physical subject
tion,
assumed having a permanent organiza
which develops under conditions themselves possessing objec
tive laws; thus the whole concept of being, which should be deduced, There is "utility" only in a world, in which is in fact presupposed.
everything does not issue arbitrarily from everything, but where certain consequences are connected with certain presuppositions.
The standpoint
being
of utility
is
and
a definite order of process.

the
"practical."
only intelligible and applicable inside of (Cf. above p. 284 f .)
However, such considerations do not more subtle interpretation that pragmatism has affect the finer and received especially from Dewey and his school. Here the problem is freed at least from those obscurities and ambiguities that attached Here it clearly appears, to it in popular philosophical discussion. that pragmatism is concerned with the relation to be assumed between the objectively valid propositions of science and the activity
Truth and
of thought.
For, as closer examination shows, thought has here

"doing."
Our inferences become the pure and complete expression of and conclusions, our investigations and experiments are "practical" not because they are necessarily directed on gaining an outer purpose, but merely in the sense that it is the unity of all thought, which stands constantly before us as an ultimate goal and directs our
The truth of any particular proposition can only be cognition. measured by what it contributes to the solution of this fundamental problem of knowledge, by what it contributes to the progressive unifi cation of the manifold. We can never set a judgment directly over against the particular outer objects and compare it with these,
fulfils
as things given in themselves; we can only ask as to the function it in the structure and interpretation of the totality of experi
ence.
fixed reality
a proposition not because it agrees with a all thought and all possibility of thought, but because it is verified in the process of thought and leads to new and fruitful consequences. Its real justification is the effect, which it produces in the tendency toward progressive unification. Each hypothesis of knowledge has its justification merely with reference to this fundamental task; it is valid to the degree that it succeeds in
call
"true,"
We
beyond
intellectually organizing isolated sensuous data. 9
and harmoniously shaping the
originally
9 Cf. Studies in Logical Theory, edited by Dewey (The Decennial Publica tions of the University of Chicago, First Series, Vol. Ill, Chicago 1903).
319
The critical concept of truth. The critical view of knowledge and of its relation to the object is not affected, however, by these develop ments; for in them a thought is merely spun out, which the critical
view has recognized and taken as a basis from the beginning. For the critical view, too, concepts do not gain their truth by being copies of realities presented in themselves, but by expressing ideal orders by which the connection of experiences is established and guaranteed. The ".realities," which physics affirms, have no meaning beyond that of being ordering concepts. They are not grounded by pointing out a particular sensuous being, that "corresponds" to them, but by being recognized as the instruments of strict connection and thus of thorough-going relative determinateness of the "given." (Cf. above The recognition of this fact, however, involves none of p. 164 ff.) the consequences, that pragmatism is accustomed to attach to it.
No
matter how-far we
may
grant and urge the
"instrumental"
meaning
cal
we are obviously concerned here of scientific hvpotnyaesL. is to be sought with purely theoreti with a purely theoretical goal, that
The will, that is here to be satisfied, is nothing but the The direction of the progress of knowledge is not will to logic. determined by any sort of individual needs, such as vary from one subject to another, but by the universal intellectual postulate of
means.
unity and continuity. In fact, this consequence has occasionally come to light very clearly, in spite of all the ambiguity in the con cept of the "practical." James himself emphasizes the fact that our knowledge is subject to a double compulsion; just as we are bound
to the properties of our sensuous impressions in our knowledge of facts, so our thought is determined by an "ideal compulsion" in the field of pure logic and mathematics. Thus, for example, the
*/\
hundredth decimal of the number T is ideally predetermined, even "Our ideas must agree with if no one has actually calculated it. be such realities concrete or abstract, be they facts or be realities, they principles, under penalty of endless inconsistency and frus 10 tration." It is clear that the assumption of such "coercions of the ideal order" does not differ from the assumption of an objective,
logical criterion of truth;
both are only different expressions of the accomplished is no refutation of "pure logic," but at any rate a further development of the thought on which it It is not a new solution that is offered, but a new problem, rests.
same
fact.
What
is
10
James, Pragmatism,
New York
1907, p. 209
ff.
320
such as could be neglected at

knowledge.
first in the general attempt to ground universal truths of logic and mathematics are not only deprived of any empiristic grounding, but they also seem to lack any relation to the world of empirical objects. Their apriority is
The
based on their
"freedom
.from
existence,"
and only holds to the

turns
extent that this condition
is satisfied.
The moment thought
to the empirical existence of the objects, it seems to separate itself from the real foundation of its certainty. True insight into the necessity of a connection can only be reached, where we renounce
all
that enter into the relation. 11
attempts to assert anything regarding the reality of the elements, Nevertheless we cannot hold to this
unconditional separation, however unavoidable it may appear at from a methodological standpoint for the mere possibility of a mathematical science of nature contradicts it. For in mathematical
first
;
two types of knowledge, that are here opposed, more immediately related to each other. It is empirical being itself that we seek to grasp in the form of rational mathematical orde^f That this demand is never ultimately and definitely satis For the material, fied, results from the nature of the task itself.
science of nature the
are
oripe
is never found but is only shaped in the process, assumes new forms. It is no constant, but a variable datum and is to be comprehended in its variability, in the possible transformation that it can undergo through new
that
is
here subjected to mathematical treatment,

"facts,"
finished as a complete store of and in the process constantly
observations and investigations. But this variability is of the essence of the empirical and involves no element of "subjective" arbitrariness.
The change
is itself determined and necessary, since we do not go from one stage to another but according to a definite law. arbitrarily It is customary to appeal especially to the purely relative validity of our astronomical system of the world, in order to prove the relativ
ity of the concept of empirical truth.
It is argued, that since the absolute motions of the heavenly bodies are not given to us in experi ence and can never be given to us, and since therefore we can never
compare our astronomical constructions with these motions them and thus test them, there is no meaning in granting to any one system (for instance, the Copernican) the merit of above all others. All systems are equally true and equally real,
selves
"truth"
11
See above p. 319
im System
f.; cf. Meinong, Uber die Stellung der Gegenstandstheorie der Wissenschaften, Lpz. 1907, section 5 ff.

since they are all equally remote
321
from the absolute
reality of things
and signify only subjective conceptions of phenomena, which can and must differ according to the choice of the intellectual and spatial standpoint. The defect of this inference, however, is plain; for the abolition of an absolute standard in no way involves the abolition The latter of differences in value between the various theories.
all strictness in so far as the general presupposition is retained that the changing phases of the concept of experience do not lie absolutely outside of each other, but are connected by logical
remain in
The system and convergence of the series takes the place an external standard of reality. Both system and convergence can be mediated and established, analogously to arithmetic, entirely by comparison of the serial members and by the general rule, which they follow in their progress. This rule, on the one hand, is given by
relations.
of
the fact that the form of experience persists; thus the particular configurations in space, which we take as a basis for our construction
system of the world, vary, while space and time, number and magnitude are retained as instruments of any construction. (Cf. above p. 266 ff .) Further there are also certain material features of the system, that are unaffected by the transition from one stage to
of the
the following: the variation does not absolutely cancel the earlier All the stage, but allows it to remain in a new interpretation.
observations of
Tycho de Brahe enter
into the system of Kepler,
although they are connected and conceived in a nejsr way. The claim of any such connection, however, is not measured with respect to things themselves, but by certain supreme principles of the knowl edge of nature, that are retained as logical norms. The spatial order is called "objective," that corresponds to these principles, that the presupposi e.g., has been constructed by us in accordance with tion and the requirements of the law of inertia. (Cf. above p. 183 ff .) Going back to such supreme guiding principles insures an inner
homogeneity of empirical knowledge, by virtue of which all its various phases are combined in the expression of one object. The is thus exactly as true and as necessary as the logical unity "object" but also no truer or more necessary. Yet of empirical knowledge;
little
as this unity ever appears as complete, much as it remains rather a "projected unity," its concept is none the less definitely determined. The demand itself is what is fixed and permanent, while every form of
its
realization points
beyond
itself.
The one
reality can only be
322
indicated and defined as the ideal limit of the
many changing theories; yet the assumption of this limit is not arbitrary, but inevitable, since only by it is the continuity of experience established. No single
as the Ptolemaic, cosmic order, but only the whole of these systems as they unfold continuously according to a definite connection. Thus the instrumental character of scien
little
"true"
astronomical system, the Copernican as can be taken as the expression of the
tific
concepts and judgments
is
not here contested.
These concepts
are valid, not in that they copy a fixed, given being, but in so far as they contain a plan for possible constructions of unity, which must be
progressively verified in practice, in application to the empirical material. But the instrument, that leads to the unity and thus to
the truth of thought, must be in itself fixed and secure. If it did not possess a certain stability, no sure and permanent use of it would be possible it would break at the first attempt and be resolved
;
into nothing. need, not the objectivity of absolute things, but rather the objective determinateness of the method of experience.
We
The reconciliation of permanence and change. The real content of the object of thought, to which knowledge penetrates, corresponds therefore to the active form of thought in general. In the realm of
rational knowledge, as in that of empirical, the
same problem
is set.
In the process of knowledge, the conception is established of a body of ideal relations, that remain self-identical and unaffected by the
accidental, temporally varying conditions of psychological appre hension. The affirmation of such a constancy is essential to every act
of thought; only in the way in which this affirmation is proved is founded the difference of the different levels of knowledge. As long as we remain in the field of purely logical and mathematical proposi tions, we have a fixed whole of truths, that remain unchanged. Each proposition is always what it once is; it can be supplemented by others being added to it, but it cannot be transformed in its own import. Pure empirical truth, however, seems in principle to be deprived of this determinateness; it is different today from what it was yesterday, and merely signifies a temporary hold, that we
gain amid the flux of presentations only to relinquish again. And still both motives, in spite of their opposition, are ultimately united
t
type of knowledge. It is only in abstraction that we can separate the absolutely permanent element from the changing and oppose them to each other. The genuine, concrete problem of
in a unitary
323
knowledge
is
to
make
body
of eternal truths
the permanent fruitful for the changing. The becomes a means for gaining a foothold in
The changing is considered as if it were when we attempt to understand it as the result of univer permanent, sal theoretical laws. The difference between the two factors can never be made to disappear entirely; still in the constant assimila tion between them consists the whole movement of knowledge. The
the realm of change.
changeableness of the empirical material proves in no way to be a hindrance, but rather a positive furthering of knowledge. The
opposition between mathematical theory and the totality of observa tions known at any time would be irreducible, if there were fixed,
unchanging data involved on both sides. The possibility of remov ing the conflict only opens up, when we become aware of the con ditional character of our empirical cognitions and thus of the plastic The ity of the empirical material with which knowledge works.
of the given and the demanded is established by our investi the given in the light of the theoretical demands, and thus gating broadening and deepening its concept. The constancy of the ideal
harmony
forms has no longer a purely static, but also and especially a dynamic meaning; it is not so much constancy in being, as rather constancy in
logical use.
The
ideal connections spoken of
by
logic
and mathe
matics are the permanent lines of direction, by which experience is orientated in its scientific shaping. This function of these con
nections
as
is
their
identical
permanent and indestructible value, and is verified through all change in the accidental material of
experience. The double form of the concept.
Identity and diversity, constancy and change appear from this side also as correlative logical moments. To establish an absolute, real opposition between them would be to cancel not only the concept of being, but that of thought. For
thought is not limited to separating out the analytically common element from a plurality of elements, as has been shown but reveals its true meaning in its necessary progress from one element to another. Difference and change are not points of view in principle "alien to x2 but belong in their fundamental meaning to the character thought,
"
istic
function of the intellect and represent this function in
its full
scope.
If this correlative
12
double form of the concept

("
is
misunderstood, then
Jonas
For the concept of alienation from thought
Denkfremdheit"), see
ff.
Cohn, Voraussetzungen
u. Ziele des Erkennens, Lpz. 1908, esp. 107
324
a gap, that cannot be bridged, must arise between knowledge and phenomenal reality. We reach once more the fundamental view of
Eleatic metaphysics, which has had in fact an interesting and signifi cant revival in modern criticism of knowledge. Inorder to understand reality by our mathematico-physical concepts, it is concluded that we
destroy it in its real nature, i.e., in its multiplicity and changeableness. Thought does not tolerate any inner heterogeneity and
must
from which it constructs its form of being. physical qualities of things are resolved into the one concept of the other, which is itself nothing else than the hyposchange
in the elements,
The manifold
The living intuition tatization of empty space devoid of qualities. of the temporal flow of events is transfixed by thought into the per manence of ultimate constant magnitudes. To explain nature is
thus to cancel
it
as nature, as a manifold
and changing whole.

of
Parmenides"
The
con
eternally homogeneous,
It is only
motionless
"sphere
stitutes the ultimate goal to
approaches.
efforts
which all natural science unconsciously owing to the fact that reality withstands the of thought and sets up certain limits, that it cannot transcend,
;
that reality maintains

it is
itself against the logical leveling of its content only by such opposition from reality, that being itself does not 13 Paradoxical as this disappear in the perfection of knowledge.
consequence may seem, it is nevertheless exactly and consistently deduced from the assumed explanation of the intellect and its But this explanation requires a limitation. The peculiar function.
identity, toward which thought progressively tends, is not the iden tity of ultimate substantial things, but the identity of functional
orders and correlations.
These functional correlations, however, do not exclude the element of diversity and change, but are deter mined only in it and by it. It is not the manifold as such that is cancelled, but only one of another dimension; the mathematical
manifold
manifold.
substituted in scientific explanation for the sensuous is not the extinction of plurality in general, but its mastery by the mathematical con and change In establishing this continuity, tinuity of serial laws and serial forms.
is
What thought demands
thought requires the standpoint of diversity no less than it requires the standpoint of identity. The former also is not absolutely forced upon it from without, but is grounded in the character and task of scientific "reason." When analysis resolves the given, sensuous
13
See E. Meyerson, Identite
el
R6alite, esp. p. 229
ff.
325
qualities into a wealth of elementary motions, when the reality of the becomes the reality of the "vibration," then it is "impression"
seen that the path of investigation does not merely consist in going over from plurality to unity, from motion to rest, but that it also
leads in the reverse direction,
and that the
setting aside of the appar
ent constancy and simplicity of perceptual things is no less neces sary. Only by thus setting aside the simplicity of the perceptual things can we reach the new meaning of identity and permanence,
which
The complete concept of lies at the basis of scientific laws. thought thus reestablishes the harmony of being. The inexhaustibleness of the problem of science is no sign of its fundamental insolu bility, but contains the condition and stimulus for its progressively
complete solution.
CHAPTER
VIII
Logical relations and the problem of self-consciousness. The prob lem of knowledge, instead of leading us to a metaphysical dualism of the subjective and the objective worlds, has led us to a totality
of relations, that contains the presupposition of the intellectual opposition of the "subject" and the "object." Confronted with
this totality, the
This totality
knowledge judgment, while, on the other hand, it can only be comprehended in judgment and thus in the activity of thought. From this, it is evident that the characterization of this totality is subject to a double method and can be attempted in a twofold way. What these relations are in their purely logical meaning, can only be learned from
the meaning they gain in the total system of science. Each particu lar proposition is bound and connected with another within this
customary separation is seen to be impracticable. objective, in so far as all the constancy of empirical rests upon it as well as the whole possibility of objective
is
system; and the position, that the proposition thus gains within the whole of possible knowledge, indicates the measure of its certainty. The question, as to how this whole is realized in the knowing individ uals, must be subordinated, as long as we are concerned with under standing the pure system of foundations and deducing it in its truth.
The development
the
of science itself urges us to leave this question in background. Science advances from one objectively valid proposition to another for which it claims the same form of validity,
without being diverted from this path by psychological considera And yet precisely this independent tions and psychological doubt. creates a new problem for psychology also. progress ultimately It is evident, that in so far as psychological analysis proceeds from
mere sensuous experience and seeks to remain in this type of experi ence, it can in no way do justice to the problems, that are continually
raised from the side of science.
The
object, that science presents to
us as distinct and certain, requires new psychological means for its Thus the general demand for a psychology of relations description.
326
327
leads to a transformation of psychological methods in general. This msformation in the principles of psychology constitutes a weighty
lem of epistemology.
it is
It is clear that here, as in other fields,
the type of conceptual construction, that undergoes a charac
teristic
change. Plato s psychology of relations. For a long time modern psychology seemed completely to have lost sight of the peculiarity of the pure relational concepts. Only in relatively recent times and by remark
round-about ways, has it begun to approach them again. From a historical point of view, there is a strange anomaly in this; for what the modern psychologist may easily consider the end of his
able,
The conception science, really constitutes its historical beginning. With of scientific psychology goes back historically to Plato.
Plato, the concept of the soul emerges for the first time from the general concept of nature, and receives peculiar and independent features.
The
no longer the mere breath of life, which has in it the and self-movement, but the soul passes over from this general meaning to that of self-consciousness. This transition, however, is only possible owing to the fact that Plato
soul
is
principle of its preservation
sure of his necessary intermediate terms, in pure pure geometry and arithmetic. There is no path that from mere perception as such to the new concept of the For perception seems a mere part of the processes is to be established.
has already
made
logic as well as in
"self,"
of nature; as represented of nature, perception
by Empedocles and all the older philosophy seems nothing else than the adjustment, which f takes place between our body and the material things of its environ ment. In order to know the corporeal things in perception, the In the soul must be of the same nature and constitution as they. of Protagoras in development, which Plato has given the proposition the Theatetus, this view is still plainly re-echoed. "Subject" and are related as two forms of motion directed on each other; "object" these forms of njoiion we can never separate in their purity and independence, but can only grasp in their reciprocal determination by each other. We always grasp only the result, without being able To this view, Plato conforms to analyse it into its real components.
for a short space
as long as he
it
sensation,
but leaves
is concerned with the analysis of behind when he turns to the analysis of the
reac pure concepts. The image and analogy of physical action and and diversity, equality and tion now has to be abandoned. Unity
328
real forces.
inequality are not corporeal objects, pressing in upon us with corpo The manner, in which they are presented to the ego, is
fundamentally new and characteristic. The eye may distinguish the bright and the dark, and touch may distinguish the light and the heavy, the warm and the cold, but the whole of knowledge is never exhausted in the totality of such sensuous differences. It is knowl edge when we say of color or tone, that each of them is, that one of them is different from the other, or that both of them united are two.
But although being and non-being, similarity and dissimilarity, unity and plurality, identity and opposition are objectively necessary elements of every assertion, they cannot be represented by any content of perception. For just this is their function, that they go
beyond the particularity of sensuous contents in order to establish a connection between them; and while both the sensuous contents participate in this connection in the same sense, the latter can be pointed out directly in neither of the two particular elements as such. The relation between the heterogeneous fields of sense-perception could not be attained, if there were no structures which remained outside of their special characters and thus outside of their qualita tive oppositions. These universal elements neither require nor are
from
connected with any special organ; rather the soul gains them purely itself in free construction. The concept of the unity of con
sciousness here
first
we
rest in the content of the particular sensation,
If gains a firm foothold and sure foundation. nothing is offered
us but a chaos of particular experiences. The perceptions are packed 1 together in us like the heroes in the wooden horse, but nothing is
found that refers them to each other and combines them into an identical self. The true concept of the self is connected with the concepts of the one and the many, the like and the unlike, being and non-being, and finds its true realTzalibfr only in these. When we comprehend the perceptions under these concepts, we combine them into one idea, whether or not we designate this unity as The is thus, as it were, conceived and postulated as the unitary expression of the content and as the systematic arrangement of pure relational concepts. The fundamental problem of psychology is defined with reference to the fundamental problems of purejogic and mathematics; and it is this connection, which definitely frees the Platonic concept of the soul from the Orphic speculation and that
"soul."
"soul"
1
Cf. esp. Theatetus, 184 C.
ff.
ON THE PSYCHOLOGY OP RELATIONS

of the philosophy of nature with
329
which
it
at
first
seems closely
view is un but already
connected.
Aristotle s
doubtedly
its
Doctrine of the KOLVOV. The Platonic influential in Aristotle s doctrine of the KOLVOV
Aristotle starts
In his division of sense-perceptions, center of gravity has shifted. from the fact that there is a particular content belong
ing to each sense, peculiar to it alone and distinguishing it from all the other senses. Thus color belongs to vision, tone as such, an tdiov belongs to hearing, while touch includes indeed a plurality
of qualities, but is related to each of sense is related to its specific content.
them But
in the
same way
as
any
this sort of relation does
not
suffice,
when we
are determining the psychological correlate of

rest,
concepts
like
motion and
magnitude and number.
These
concepts represent something truly "common," that reaches beyond all particular differences. As Aristotle further concludes, a univer
organ must correspond to the universality of the object. When, for example, we correlate the white with the sweet or oppose them to each other, it is necessarily s,ense itself that
sality of the receptive
performs this act of comparison. For with what other faculty could we grasp contents of purely sensuous nature? But here sense no
longer functions in any special capacity, as mere vision or mere taste, but as a "common sense" in an inclusive meaning. To this common
sense, all the individual data of perception are referred, to be collected
and related to each other. 2 Thus what was conceived by Plato as a spontaneous and free function of "consciousness itself," here appears as a power at once abstract and sensuous, wherein is combined This all in which the different types and fields of perception agêe. psychological view of Aristotle s corresponds to his fundamental logical view; it rests on the view of. the "concept" as nothing but a sum of sensuQUS properties, which are found uniformly in a plurality
in it
of objects.
"Thoughts
of
relation"
in modern psychology.
At
first,
modern
psychology makes only isolated attempts to reach a new interpreta tion of the problem. Leibniz harks back directly to Plato when he urges, that the contents, which the traditional doctrine ascribes to the "common sense" (in particular, extension, form and motion), are ideas of the pure understanding:- and althcugh they are-formed on
the occasion of sense-impressions, can never be completely grounded
2
Cf. especially Aristotle,
Trep^xns
II, 6,
41Sa; III,
2,
426b.
330
in them.
who
In modern German psychology, it is Tetens especially, takes up this suggestion and develops it into a theory of the pure "thoughts of relation." But, on the whole, Locke s schema is
dominant throughout; and according to it a concept can only be taken as truly understood and deduced, when we are able to show the simple sense-contents out of which it is compounded. The ideas of
"reflection"
also,
although they seem at
first
to occupy a special
They possess truly positive content only in so far as they can be expressed im
position,
finally
are
measured by
this
standard.
mediately in individual, perceptually given presentations. This reliance on a perceptual criterion appears most clearly in the concept of the infinite, which is criticized for no other reason than because it
clear that what it means is never realized in actual presentations, but subsists merely with reference to a possible, unlimited intellec tual progress. Even if, from the standpoint of logic and mathe
is
matics, the general rule of this progress constitutes the existence and truth of the infinite, nevertheless, from the psychological view, this
rule carries the stamp of the merely negative. For within this psychological view no adequate expression has been discovered for the validity and peculiarity of relations. Nevertheless the thought of
these relations constantly returns, no matter how much it may be Like a ghost of uncertain nature and origin, it mixes repressed. the clear, certain impressions of perception and memory. among
No
ridicule, with Berkeley, the infinitesi mathematics as the "ghosts of departed quantities," these ghosts cannot be laid to rest. Analysis here strikes an ulti mate remainder, which it can neither comprehend nor set aside. The concepts, which prove effective and fruitful in actual scientific
matter how
much one may

of
mal magnitudes
use, are never contained in those elements that the psychological view recognizes as the unique bearers of "objectivity." The signifi
scientific concepts rests on the fact, that they persis transcend the type of reality that is here taken as a pattern. tently
cance of such
The concept
however,
of substance.
The deeper ground
of this conflict,
has by no means freed itself from the presuppositions it struggles against. The concept, which Locke attacks most sharply and thoroughly,
rests in the fact that psychological criticism
is
moned
the concept of substance. All the weapons of ridicule are sum against the assumption of an independent, separate "some thing" without properties, which is assumed to be the "bearer" of

sensuous qualities.
is
331
He shows repeatedly that the real validity of overthrown by the assumption of substance, since what knowledge is best known to us and most certain from the standpoint of experi ence is explained by means of something entirely without content and
unknown.
A mysterious
"I
know not
qualities
what"
is
made
the conceptual
and properties. In this po ground lemic against the concept of substance, Locke believes he has struck the real kernel of all metaphysics and of all scholastic explanation of And after Hume has transferred its result from outer to reality. inner experience, the work of criticism seems ended. The substance of the ego seems now explained away like the substance of the thing; mere associative connections of presentations take the place of both. In spite of all this, the view of physical and psychical reality, that is constructed on these foundations, has in it the general category of
of all truly
knowable
Only the applications of this substantiality in its decisive meaning. category have been changed, while it itself retains its old rank and
is only ap The substantiality of the parently done away with; for it lives on in the substantiality of the sensuous "impression." After, as before, the conviction prevails, and the ground of the real, which stands that only that is truly
position unnoticed.
"soul"
"real"
for itself alone
and is intelligible purely of itself as an isolated exist Here what is unchanging and essential in known reality lies ence. before us, while all connections, which are established subsequently between the particular contents, form a mere addition of the mind. They thus only express an arbitrary tendency of the imagination, but not an objective connection of things themselves. This result
the were, the negative proof of the stability that When of everything. substantialistic view still possesses in spite the attempt is made to conceive the pure concepts of connection not as impressions and (in particular the concepts of cause and effect)
constitutes, as
it
copies of objects,
"impression"
is by be any inner value through the fact that it appears to and established with a kind of based in the "nature" of the mind universality and regularity under definite conditions. The doctrine of the "form-quality" in modern psychology. Modern
their logical import disappears. that very fact a mere fiction.
What
is
not
this
Nor does
fiction gain
psychology tried for a long time to avoid the skeptical consequences it of Hume s doctrine, without submitting the premises on which
rests to a
thorough-going change.
In
its
own
concepts, there de-
332

which was at
first
veloped a new form of
"reality,"
naively and con
All the peculiarities of psychological analysis fidently accepted. were directly taken as properties of the psychical object. Thus, there resulted that self-deception discovered and described by James as
the
"psychologist s fallacy."
The means used
to represent a definite
psychical fact and to make it communicable in a simple way were taken as real moments of this fact itself. Imperceptibly, the stand
point of analysis and reflective observation took the place of the

3 A typical example of this whole standpoint of real experience. view is the doctrine of the "simple" elements, out of which each The ultimate parts state of consciousness is to be compounded.
which we can conceptually discriminate become the absolute atoms out But this being of which the being of the psychical is constituted. remains ambiguous in spite of everything. Properties and charac teristics constantly appear in it, that cannot be explained and de duced from the mere summation of the particular parts. The further introspection advances and the further it follows its own way without prejudice, the more do new problems come to light. At first they are formulated only from a limited viewpoint and a special interest. The questions, which are combined in psychology under the concept of the form-quality, furnish the first stimulus to a renewed revision of the general concepts. Here it is shown in specially striking examples that not every spatial or temporal whole, that experience offers us, can be represented as a simple aggregate of its
particular parts.
When
our consciousness follows and apprehends a
simple melody, at first all the content present to it seems to consist in the perception of the particular tones. Closer observation shows,
however, that such a description does not touch the real fact. We can, by transition into another kind of tone, allow all the individual
tones originally in the melody to disappear, and we can replace them by others without ceasing to retain and to recognize the melody itself as a unity. Its specific character and properties thus cannot depend
for us
on the particularity of the elements, since the melody remains change of this particularity. Two entirely different complexes of tone-sensations can give us the same melody while, on the other hand, two complexes of elements having the same content
in spite of all
can lead to entirely different melodies, in so far as these elements are different from each other in their relative sequence. The same
3
Cf. James, Principles of Psychology,
I,
196
ff.,
278
ff.,
etc.
333
thought can be carried over from the unity of the "tone-form" to the unity of the "spatial form." In space, we call certain figures "simi to each other and conceive them as identical according to their
lar"
mere geometrical concept even though they are
built out of qualita This consciousness of tively wholly different spatial sensations. the identity of wholes that differ from each other in all their particu
not a special explanation, at least a special psychological designation. The concept of form-quality contains this designation, though at first it only defines the problem without giving it a definite solution. What connects the manifold presenta
lar
elements requires,
if
tional contents into one psychological form, is not to be found in these contents themselves nor in their mere coexistence as an aggre
gate.
But a new function
is
found here embodied in an independent
structure of definite properties. The existence of such a structure and of the peculiar increase in content given with it is to be recog nized as an empirical datum, no matter from what theoretical pre
suppositions one
may
interpret
it.
is already found, when one to explain the unity of the complex psychical structures, attempts not from the mere coexistence of their parts, but from the reciprocal It is objected and is not to be action they exert on each other.
Such a theoretical interpretation
disputed, that the melody appears as an independent content as opposed to the particular tones which enter into it. But in the explanation of this fact it is by no means necessary to introduce
entirely new elements along with the ordinary elements of sensa To form a whole tion and presentation and as added to them.
means psychologically nothing

4
else
than to
act as
a whole.
f.
Not only
Cf. esp. Ehrenfels,
Uber Gestaltqualitdten, Vierteljahrschr.
wiss. Philos.,
XIV
u. (1890), p. 249 ff., also Meinong, Zur Psychologic der Komplexionen Relaiionen, Zeitschr, fur Psychologie u. Physiologic der Sinnesorgane II (1891), The psychological explanation of the problem of the "form-quali p. 245 ff.
ties"
would doubtless have been of more general significance, if it had gone into the corresponding logical problems more closely, that immediately present themselves here. As the psychological examples adduced always show, the
concern here is with a general process of liberation and separate consideration of the relational content, which is particularly characteristic and of fundamen tal importance for wide fields of mathematics. (See especially Ch. Ill, p. 93 ff.) The possibility of retaining a relation as invariant in its meaning, while the members of the relation undergo the most various transformations,
is
only illumined and established from a new side in purely psychological
considerations.
334
the parts as such, but also their whole complex always produces And it is by definite effects upon our feeling and presentations.
virtue of these effects, which proceed from the complex and are thus dependent on the order of the elements within it, that we judge
of wholes.
concerning the similarity or dissimilarity, the equality or inequality This explanation seems in its general application to
render unnecessary any assumption of special relational presenta It is again the simple perception, tions and relational concepts.
which decides not only regarding sensuous properties such as color and tone, smell and taste, but which also informs us of unity and plurality, of permanence and change, and of the temporal sequence and the temporal persistence of contents. For all these determina tions are distinguished from the simple sense impressions only by the not of single stimuli, but of complexes of fact that they are
"effects"
Just as a certain ether vibration produces in us the impres stimuli. of a definite color, so a certain composition and connection of sion
stimuli
affecting our consciousness produce in it an impression of Thus, e.g., when similarity or difference, of change or permanence. different tone sensations are aroused in us at definite temporal dis
tances from each other, we can measure the length of the intervening pauses by each other and accordingly speak now of a more rapid,
now
for the
special intellectual act is needed of temporal intervals, but the simple assump "comparison" tion suffices that a certain effect issues from the complex of rapidly
of a slower succession.
No
succeeding tones, that is different from the effect produced with a 5 However, if we follow this greater distance between the tones.
attempted explanation to the end, we discover at once the epistemoThe totality of pure relational logical circle that it involves. is reduced to an actual effect resulting from certain mani thoughts folds, while however the mere application of the standpoint of cause and effect involves a special relational thought. It is not the case that we can be led from the knowledge of certain causal connections
to
the understanding of relational concepts in general; rather, conversely, what these concepts mean must be already presupposed When in order to be able to speak of causal connections of reality.
5
Schumann, Zur Psychologic der Zeitanschauung. Ztschr. f. Psychol. XVII ff. For a criticism of Schumann, cf. esp. Meinong s treatment, Uber Gegenstande hoherer Ordnung u. der en Verlidltnis zur inner en Wahrnehmung. Ztschr. f. Psychol. XXI. (1899), p. 236 ff.
(1899) p. 106
335
the psychological explanation proceeds from the factual elements and the action which they exert in the whole psychic process, it has
already taken for granted everything the logical justification of which stands in question. It places a world of things, in which
various objective relations prevail, at the starting point of its con while it had appeared at first as if this whole kind of sideration, reality could and should be deduced out of simple sensations as the
unique data of pure experience, without the assumption of any other element. This reversal is indeed not strange; for it actually
only reestablishes the order of problems that was perverted in the first approach to the problem. What is truly known and given in the field of consciousness, is not the particular elements, empirically
which then compound themselves into various observable effects, but it is rather always a manifold variously divided and ordered by relations of all sorts, such a manifold as can be separated into particular elements merely by abstraction. The question here can never be how we go from the parts to the whole, but how we go from the whole to the parts. The elements never "subsist" outside of every form of connection, so that the attempt to deduce the pos sible ways of connection from them moves in a circle. Only the total result itself is in the sense of experience and of psychological process, while its individual components have only the value of hypothetical assumptions. Their value and justification acc is to be measured by whether they are able to represent and recon struct in their combination the totality of the phenomena.
"real"
Ebbinghaus
physiological account of relations.

itself
Thus within psy

to
chological investigation
the speculative attempts
deny
the peculiar character of pure relations and to replace them by mere complexes of sensations gradually die out. The ideal of conceptual that was here taken as a standard, is in general is recognized that our actual experience and our Just real empirical psychological knowledge prevent its realization. as we must assume certain classes of simple sensations as ultimate
"explanation,"
retained; but
it
facts so, it is
now explained, we must recognize also as fundamental and irreducible data of consciousness certain specific relations, such as unity and plurality, similarity and difference, spatial coexistence and temporal duration. "Naturally this does not mean to explain it means to "but things," a representative of this view remarks, One way prefer honest poverty to the appearance of
wealth.""
Ebbinghaus, Grundzuge der Psychologic, 2nd
ed.,
Lpz. 1905,
p. 462.
336
seems still to remain, that would permit us to resolve the conceptual manifold of relations into the unity of a single causal origin. What
denied to the purely psychological type of consideration here seems to receive a physiological explanation. The general relational determinations, which are uniformly found with all sensations regard less of their qualitative difference, thus prove to be a common con dition of sensation, for which a corresponding community is to be demanded in the appropriate physiological processes. This agree ment in the physical foundations of each perception, no matter to what particular field it may belong, can easily be pointed out. At first the sense organs and their appropriate nervous centers appear to be very differently built and differently equipped apparatuses; nevertheless, they form a unity in so far as they are all built up out of the same material, out of nervous elements according to certain uniform principles. external stimuli affect them, then the called forth in them must naturally be different, in so far processes
is
"If
as the physico-chemical properties of the stimulus and the function At the same of the receiving apparatus adapted to it are different.
time, those processes must be like or similar, in which the connection of the stimuli into a whole in the outer world
mode
of
and the of the nervous matter within the sense organ and the properties It general principles of its construction are the same
owing to the special character of the stimulations involved in seeing, hearing and tasting, that we sense their mental effects as something entirely disparate, as bright, loud, bitter; it is due to the agreeing features of the same stimulations, that we are aware of these impres
is
sions as
permanent or intermittent, as changing
etc.,
according to
circumstances."
"intuitions"
The nervous
processes, that
lie
at the basis of the
and change,
and time, of unity and plurality, of constancy thus found entirely in the same processes that are correlated with sensations; but they are not found in all features, but only in the common characteristics of these processes, which as
of space
"are
yet cannot be indicated more

first
closely."
Criticism of the physiological explanation of relational concepts. At glance, this explanation seems to operate wholly with the means
of
modern
true
7
scientific investigation; nevertheless in principle it leads

"common
sense."
us back to the Aristotelian doctrine of the

is
It
we have not a
cit.,
special organ at our disposal for

p. 442
f.
apprehending
Ebbinghaus, op.

relations, as
337
we have
for the particular sense qualities;
exists a sort of
common organ by which we
but yet there take up into our con

I
sciousness the real relations of outer objects. However, if this causal explanation is meant also to be a logical deduction of the validity of the relational concepts, then it would also involve that
For in vffrepov Tporepov in itself, which we have previously met. order to explain the presentation of equality or difference, of identity or similarity, such an account must obviously go back to similarity
or difference in things,
organs of perception.
already contains in
more specifically in the peripheral and The concept of being, assumed at the
central
outset,
those categorical determinations that are drawn out of it in the course of the psycho-physiological subsequently deduction. The truth of these determinations has to be presupposed,
it all
although the manner in which they reach the consciousness of the individual subject may be open to an indeed only hypothetical explanation. Further, every purely physiological representation of the matter must leave in obscurity the point that is above all in ques
An identity or community in the outer stimuli never suffices to explain the correlative expression of these relations in conscious ness. The physically like must be recognized and judged as like,
tion.
the actually different must be conceived as different for there to be that separation of the ^general content from the particular content of
Thus the pure conscious func sensation, which is here assumed. tions of unity and difference can never be dispensed with or be re
demand
placed by reference to the objective physiological causes. The is that they be represented by a type of explanation that remains within the field of the psychical phenomena themselves,
without going back to their hypothetical grounds. Thus from all sides we are directed with increasing clarity to a second great field of psychological investigation, a field that was at first neglected and subordinated. In contrast to the psychology of sensation, appears the psychology of thought. Such a psychology is dominated from the first by an entirely different formulation of the problem,
and by a new order

tive"
of value
between the
"absolute"
and the
"rela
elements of consciousness.
ii
Meinong
theory of
"founded
contents."
The
involved in the theory of the
"form-qualities"
general problems receive sharper expres-
338

which
this doctrine
sion in the transformation
has undergone in the

Inhalte).
theory of
clearly
Here it is shown that the questions thereby introduced into psychology do not tend to a mere extension of its field, but to an inner trans
"founded
contents"
(fundierten
formation of
definitely
Two forms of psychic "objects" are now its concept. opposed to each other. Over the simple sensations and
qualities of the different senses, there are built "objects of higher orders." These latter objects are borne by the elementary contents
and require them as supports, but cannot be reduced to them.
We
cannot,
indeed,
speak of equality
or
difference,
of
or plurality, without
thinking of equality or difference,
unity unity or
But this something may vary arbitrarily; plurality of something. it can appear as color or tone, as odor or taste, as concept or judg
ment, of all of which, difference or unity can be predicated, without the real meaning being thereby affected. Thus the depend ence, which seems to belong to pure relations in their actual occur rence and, as it were, in their psychic existence, does not exclude a
complete independence in their characteristic meaning. The uni versally valid relations do not exist as temporally or spatially bounded
parts of psychical or physical reality; but they "subsist" absolutely by virtue of the necessity which we recognize in certain assertions.
Whoever
represents to himself four real objects does not represent fourness as a particular piece of reality along with them, although he claims a certain objective truth and validity for his judgment regard
ing the numerical relation.
relations
Thus
in general,
in contrast to the
relations:
between existences, there appear pure ideal
and
corresponding to this distinction, there is further a characteristic opposition in the value of the cognitions relating to these objects. Wherever the judgment refers to an object of actual reality and is
meant to designate a
particular determination of it, the judgment is necessarily limited to the here and now, thus to an assertion of merely empirical validity. To this case, in which we merely ascribe to an
individual thing an individual property known by experience, there is opposed the other case, in which the type of dependency between the two elements a and b is determined and exactly prescribed by the
"nature" of the members. Concerning ideal relations of this sort, judgments are possible that do not need to be tested by different successive cases in order to be grasped in their truth, but which are
recognized once for
all
by
insight into the necessity of the connection.
339
Along with the empirical judgments concerning objects of experience, there are thus priori" judgments concerning "founded objects." While the psychic "phenomenon," like color or tone, can simply be established in its occurrence and properties as ja,,aet, there are
"a
judgments connected with the "metaphenomenal" objects, like equality and similarity, that are made with consciousness of timeless and necessary validity. In place of the mere establishment of a fact, there appears the systematic whole of a rational connection with elements that reciprocally demand and condition each other. 8
"Objects
of a higher
order."
However
energetically this theory
may
strive to extend the field of psychological
problems beyond the usual it is still under the influence of the traditional theory of the limits, concept at one point. It starts with the simple sensational contents as recognized data in order to proceed from them to the more complex structures. The "objects of a higher order" cannot be separated from whatever perceptual elements they may be founded in, without
thereby losing
all
meaning.
On
the other hand, the converse of
this proposition does not hold; while the "superius" is necessarily referred to the "inferiora," the latter for their part are characterized
by the
in themselves.
fact that they exist for themselves alone and rest entirely The relations, which are built up over the perceptual
elements, appear as a subsequent result; their being or non-being adds nothing to the existence of the elements, and can neither ground
nor imperil it. Sharper analysis, however, destroys this last appear ance of the independence of the simple. In place of a succession,^ of a superordination and subordination of contents, analysis fixes a
relation of strict correlativity.
Just as the relation requires reference the^~eTements, so the elements no less require reference to a form of relation, in which alone they gain fixed and constant meaning. Each conceptual assertion regarding an "inferius" considers this
from the standpoint of. some- relation, which we correlate with the content in question. The "foundations" are always determinable and determined only as foundations of possible relations. What deceives us at first is the fact, that all the relational determina
inferius
tions into
it
which an individual can enter are somehow contained in but in no way actually realized from the beginning. In order to
8 and "objects Cf. more particularly on the theory of "founded contents" of higher order" Meinong, Z. f. Psychologic XXI, 182 ff.; cf. also Hofler, Zur
gegenw. Naturphilosophie, p. 75
ff
340
transform the
there
is
"potential" logical import into "actual" import, needed a series of involved intellectual operations and con But the possibility of separating stantly renewed conceptual work. a content from this or that conceptual determination, and of consider ing it before this determination, as it were, must not lead us to divest it of all forms of determination in general. Consciousness as con sciousness would be extinguished, not only if we conceived the sensu ous phenomena, such as the colors and tones, the smells and tastes to be removed, but also if we conceived the "metaphenomenal" objects, such as plurality and number, identity and difference to be removed. The existence of consciousness is rooted merely in the mutual correlativity of the two elements, and neither is to be pre
ferred to the other as
"first"
and
original.
From this stand conflict between empiricism and nativism. new light falls on the old psychological controversy between point, This controversy we see is also rooted "empiricism" and "nativism."
The
an unclarified statement of the problem. It is asked whether common determinations, which appear with the sensations are immediate properties of the sensations, whether their unity and
in
the
plurality, their spatial arrangement, their longer or shorter
temporal
duration are properties just as immediate as the differences of the sensations themselves, and are apprehended at the same time as
or do they form a later product of mental comparison, imprints a definite form on the unordered perceptual material? In other words, is it a peculiar spiritual activity that leads to these determinations, or are they directly given in the first sensations,
which
first
act of perception implicitly as elements?

different standpoints,
There are, however, two which unnoticed are identified with each
The logical separation of the moments of by a temporal separation in the occurrence of certain psychic contents; and an attempt is made to solve both of these completely heterogeneous problems at the same time and by means of each other. When it is shown from the standpoint of
other in these questions.
knowledge
is
replaced
assumed or conceived, exhibits some form
nativism, that even the earliest state of consciousness, which can be of spatial-temporal or
conceptual connection, it is believed that the logical value of the connections has therewith been reduced to that of mere sensation.
It
is
concluded that there
is
in the
same way that there
an immediate consciousness of relations is an immediate consciousness of colors

or tones.
341
We apprehend in inner perception, in mere "feeling," the meaning of "and," and "therefore" just as truly as we or gain information by perception concerning the content "cold." The actus purus of the understanding thus proves to be
"but,"
"if"
"blue"
unnecessary, for everything that it was to produce is in fact already contained in the first data of perception. 9 If we would be critically
just to this conception,
we must
in
from the special way meant to be emphasized above

ruling
is it
which
distinguish the general tendency it is worked out. What is
all is this:
that the element of order

relation of before
this dis
related to the element of content in

after, of earlier or later.
no temporal
and
Only analysis can discover
tinction in the originally unitary material of the "given." In this sense, it is correct that even the most elementary psychic state in
cludes the general elements of form. The inference, however, that these elements therefore belong to the mere passivity of perception, is not thereby justified. On the contrary, the converse inference
holds.
The
fact that there
is
no content
of consciousness,
which
is
not shaped and arranged in some manner according to certain rela tions proves that the process of perception is not to be separated from that of judgment. It is by elementary acts of judgment, that the
particular content is grasped as a member of a certain order and thereby first fixed in itself. Where this is denied, judgment itself
is is
understood only in the external sense of a comparing activity, which subsequently adds a new predicate to a "subject" already fixed and Such an activity appears indeed as accidental and arbitrary, given.
Whether or in contrast to the material with which it is connected. not such an activity takes place, this material remains what it is and retains the properties, which belong to it prior to all logical activities. In its real form, on the contrary, judgment does not
signify such
an arbitrary
act,
but
is
rather the form of objective
determination in general, by which a particular content is distin guished as such and at the same time subordinated systematically
From this form, we cannot abstract without thereby the qualitative differences in content. Thus it may be that when we consider the pure temporal relation, relations will be
to a manifold.
losing all
"found"
at the
same time as the sensational contents; nevertheless

very
"finding"
it
is
true, that this
involves within
If
itself
the elementary
forms of intellectual activity.

9
we
I,
consider these forms removed,

244
f.
;
James, The Principles of Psychology
cf.
esp. II, 148.
342
all possibility would thereby disappear of further application of the concept of consciousness. Whatever the content might be or mean in and for itself,- for us, for the unity of the self, it
would not be present. For the self comprehends and constitutes It is always with certain types itself only in some form of activity.
of the "apperception of unity," that the apprehensionoi definite relations between objects is necessarily connected psychologically. 10
Thus the inseparable correlation of sensations with thkjmre relations, when its consequences are consistently traced, implies thôpposite of what was initially deduced from it; it shows not the passivity of the
ego in apprehending these thoughts, but conversely the element of activity, that belongs to every process of perception in so far as it
but belongs to the whole of conscious can, in fact, attempt to deduce the rela tions from sensation; but then we have already placed in sensation determinations, that go beyond the isolated impression. It is no longer the abstract "simple" sensation; but it rather signifies merely the initial, still unordered content of consciousness, to which, how
does not stand for
itself alone,
ness and experience.
We
and connections are always essential, leading from it to other elements. The psychology of the idea of space. This is a^.1 the clearer when we consider the special problem, that has always been the center of discussion between, empiricism and nativism. The fate of the
ever, definite relations
is decided by the question as to the psychological and the psychological meaning of the idea of space. If it is possible to deduce space from absolutely non-spatial sensations differing only in their quality and intensity, then there is nothing in principle to prevent us from carrying through the same explanation
different theories
origin
for all the different types of relation in general.
However,
it is
at
once evident that the empiristic theory, when it undertakes to deduce the origin of spatial order from the mere material of perceptions and
the simple forces of associative connection, is obliged to be untrue to its own methodic ideal. For there can be no doubt that such an
origin,
if
indeed
it is
to be assumed, cannot be pointed out in our
actual experience. Each experience, whatever its properties may be, reveals some primitive form of the "coexistence" of particular
elements and thus the specific element in which every spatial con10 Cf. above all the treatment by Th. Lipps, Einheiten und Relationen, Eine Skizze zur Psychologie der Apperception, Lpz. 1902.

struction,
343
no matter how complex, is originally rooted. If we attempt to go behind this psychological fact, if we attempt to show
how order itself arises and evolves out of the absolutely unordered, we surrender ourselves to an hypothesis, which goes beyond the limits of experience in two directions. Empirically we know just as little of a simple, absolutely unlocalized perception, as we know of a
particular function of the soul, which transforms the previously formless into a form, on the basis of unconscious "inferences." How ever we may judge of the methodic value of such concepts, it would be
dangerous and misleading to misunderstand them as an expression

of concrete facts.
Here the
criticism of the
"empiristic"
theories of
space in modern psychology (especially by Stumpf and James) is in the right, in so far as it urges that mere "association" as such can
produce no new psychic content.
No mere
repetition
if
and arrange
ment
of contents could give
them
spatiality,
spatiality were not
already
11
somehow placed
in
them. 11
But here again the temporal
James, op. cit., II, 270, 279, etc. The effects of the general schema psychology are also shown in its critics, however, in the fact that they can only affirm the originality of the spatial order by condensing it into a peculiar and original content of perception. Stumpf in particular explains, that the spatial order would not be intelligible without there being a positive, absolute content at its basis. This content first gives the spatial
Cf., e.g.,
of associationistic
its peculiar character, by which it is distinguished from other orders. distinguish the various orders from each other, we must recognize every where a special absolute content with reference to which the order is found.
order
"To
And thus
space also
is
not a mere order, but
is
precisely
what distinguishes the

(Uber den psy-
spatial order, adjacency (Nebeneinander),
from other
orders."
chologischen Ursprung der Raumvorstellung, p. 15, cf. p. 275.) Two different points of view, that are not strictly differentiated, are combined in this argu.
It may be accepted thai-every relation is the relation of something, and therefore presupposes some "foundation," on which it is built, although it is also to be remembered that the dependence is throughout reciprocal, so that the "foundation" requires the relation just as much as the latter requires the
ment
"foundation." But this does not involve that what constitutes the peculiar character of a definite principle of order must be able to be pointed out as a property of the ordered elements. For if we assumed this, we would finally have to ascribe to the content as many special "qualities" as there are ways of
connecting the content with, and referring it to, others. A specific perceptual quality would be demanded not only for spatial order, but for temporal order and further for all sorts of quantitative or qualitative comparison. But in general it is not obvious how a mere difference in the content of the compared elements should serve to define and separate the various possible types of
relation.
If
two orders are distinguished as
orders,
some means
of conscious-
344
way proves their logical equiva knowledge distinguishes the spatial and temporal form from the content of sensation and treats it as an inde pendent problem, it does not require the conception of a real separateness of the two in some mythical stage of consciousness. What it affirms and defends is merely the simple thought, that the judgments based and constructed on these forms of relation have a characteristic validity of their own, which is denied to mere assertions regarding the existence of the sensation given here and now. The original
lence.
connection of the two factors in no
When the
criticism of
unitary content
is differentiated, when we recognize it as the start ing-point of two different systems of judgments, which are separated according to their dignity. According to whether we emphasize the
moment of a particular sensation (the blue and red, the rough and smooth, etc.), or merely consider the universal relations subsisting between these particular elements, different propositions arise,
specific
belonging to thoroughly different types of grounding. Psychology, indeed, within the limits of its task, cannot trace and survey this conceptual division in its totality, since psychology merely de
and analyses thought as a temporal process, but not the content of what is thought. The tendency of the process only becomes clear in its final result; only the fully developed system
scribes
of geometry,
constructed according to unitary rational principles,
ness must be given

such,
by which the type of connection can be grasped purely as and distinguished from all others. If we ascribe to consciousness the capacity of distinguishing the simple data of perception from each other, it is not obvious how we can deny it the same capacity in the case of the various original functions of order. The deeper ground of the difficulty here seems to
lie
not so
much
in
psychology
itself,
as in the ordinary conception of logic.
Logic, in its traditional form, is founded on the thought of identity, and seeks to reduce all types of connection and inference ultimately to identity. If,
>V
however, identity is taken as an expression of the relational form in general, then the diversity of relations (which in any case must be explained) can be founded merely in the content of the elements, that are related to each other. The modern form of logic, however, has destroyed the basis of this view; it has shown with increasing distinctness that it is impossible to reduce the di verse forms of judgment to the single type of identity. (For further particulars, cf. e.g., Jonas Cohn, Voraussetzungen und Ziele des Erkennens, p. 85 ff.) ill Just as we are here forced to recognize an original plurality of diverse relational etc.), that are not mutually reducible to each other, so if-Hsyntheses (R, R
,
R",
psychological consideration must ultimately recognize differences belonging to the manner and way of "apperceptive connection" itself, without finding
their expression in
any particular quality
of sensation.
345
contains the definite characterization of the elements of space. But although psychology cannot establish this characterization, it need not contradict it at any point. Its own treatment of the problem of
it with inner necessity to a point where a new type and direction of consideration begins. The separation of the element of relation from the element of content, to which psychology is forced,
relations leads
remains, so to speak, proleptical in tion in a broader field.
it,
and only receives
full
illumina-
The psychology of thought. Even the purely empirical-experi mental view of mental phenomena significantly reveals such a tendency of the problem. The effort more and more is to apply the method of experiment not only to the facts of sensuous perception, but with its help to discover the complex processes of conceptual 12 But in applying the understanding in their fundamental features. method of experiment here, it becomes constantly clearer that it is
not the intuitive factual presentations, not the direct perceptual images, that bear and support this process of the understanding.
The understanding
and grammatical structure
of the simplest proposition in its definite logical requires, if it is to be apprehended as a
proposition, elements, that are absolutely
removed from
intuitive
representation. pictorial presentations of the concrete objects, of which the assertion holds, can vary greatly or even wholly disap
pear, without the apprehension of the unitary meaning of the proposi tion being endangered. The conceptual connections, in which this
The
meaning
is
rooted,
must
therefore be represented for consciousness in
peculiar categorical acts which are to be recognized as independent, not further reducible factors of every intellectual apprehension.
The manner,
is
in
which psychological investigation has reached
this
indeed noteworthy enough; and it further signifies the insight, historical dependence of the methods and problems of psychology. observed here in its independent "Thought" is not conceived and
activity; but the attempt is made to establish its peculiar character in the reception of a finished content from without. Accordingly,
the
factor thus gained appears rather as a paradoxical, incom understood remainder left by analysis, than as a positive and pletely The criticism of knowledge reverses this characteristic function.
new
relation; for
12
it,
that problematic
"remainder"
is
what
is
really first
A neat and synoptical treatment of this tendency in

is
gation
psychological investi given by Messer, Empfindung und Denken, Lpz. 1908.
346

"intelligible"
and is the point of departure. It does not study where thought merely receptively receives and reproduces thought the meaning of an already finished judgmental connection, but where it creates and constructs a meaningful system of propositions. When
psychology pursues this
line of
and
enquiry and considers thought equally
problems, seek their solution in logic and in their application to science.
in the concrete totality of its productive functions, the initial opposi tion of methods is more and more resolved into a pure correlation. Psychology in this sense gives the approach to which must
SUPPLEMENT
EINSTEIN S THEORY OF RELATIVITY CON SIDERED FROM THE EPISTEMOLOGICAL
STANDPOINT
AUTHOR S PREFACE
The following essay does not claim to give a complete account of the philosophical problems raised by the theory of relativity. I am aware that the new problems presented to the general criticism of knowledge by this theory can only be solved by the gradual work
both physicists and philosophers; here I am merely concerned with beginning this work, with stimulating discussion, and, where pos sible, guiding it into definite methodic paths, in contrast to the un certainty of judgment which still reigns. The purpose of this writing will be attained if it succeeds in preparing for a mutual understand ing between the philosopher and the physicist on questions, con cerning which they are still widely separated. That I was concerned, even in purely epistemological matters, to keep myself in close contact with scientific physics and that the writings of the leading physicists of the past and present have everywhere essentially helped to deter mine the intellectual orientation of the following investigation, will be gathered from the exposition. The bibliography, which follows, how ever makes no claim to actual completeness; in it only such works are adduced as have been repeatedly referred to and intensively con
of
sidered in the course of the exposition. Albert Einstein read the above essay in manuscript
the benefit of his critical comments; I cannot expressing my hearty thanks to him.
University of
let it
and gave it go out without
ERNST CASSIRER.
Hamburg.
CHAPTER
CONCEPTS OF MEASURE AND CONCEPTS OF THINGS

"The
use,
which we can make
in philosophy, of mathematics,"
Kant wrote
in the year 1763 in the Preface of his Attempt to Introduce the Concept of Negative Magnitudes into Philosophy, "consists either
methods or in the real application of its proposi tions to the objects of philosophy. It is not evident that the first has to date been of much use, however much advantage was originally
in the imitation of its
promised from it. The second use, on the contrary, has been so much the more advantageous for the parts of philosophy concerned, which, by the fact that they applied the doctrines of mathematics for their
purposes, have raised themselves to a height to which otherwise they could make no claim. These, however, are only doctrines belonging to the theory of nature As far as metaphysics is
concerned, this science, instead of utilizing a few of the concepts or doctrines of mathematics, has rather often armed itself against them and, where it might perhaps have borrowed a sure founda
tion for its considerations, we see it concerned with making out of the concepts of the mathematician nothing but fine imaginings,
which beyond his field have little truth in them. One can easily decide where the advantage will fall in the conflict of two sciences, of which the one surpasses all others in certainty and clarity, the other of which, however, is only striving to attain certainty and clarity. Metaphysics seeks, e.g., to discover the nature of space and the supreme ground from which its possibility can be understood. Now nothing can be more helpful for this than if one can borrow
proved data to take as a basis for one s several data, which concern the most general properties of space, e.g., that space does not consist of simple parts; but these are passed by and one sets his trust merely on the ambiguous consciousness of the concept, which is conceived in a wholly abstract fashion The mathematical con
from somewhere
consideration.
sufficiently
Geometry
offers
with knowledge of space furnishes many data to guide the metaphysical speculations of the times in the track of truth. The celebrated Euler, among others, has given some opportunity for this, but it seems more comfortable to remain with
sideration of
motion
in connection
351
352
EINSTEIN S THEOKY OF KELATIVITY

test,
obscure abstractions, which are hard to
than to enter into con
nection with a science which possesses only intelligible and obvious

insights."
The essay
is
of Euler, to
s
which Kant here
refers the metaphysician,
Reflexions sur I espace et le temps, which appeared in the year 1748 among the productions of the Berlin Academy of Science. This essay sets up in fact not only a program for the
the former
construction of mechanics but a general program for the epistemology It seeks to define the concept of truth of of the natural sciences. mathematical physics and contrasts it with the concept of truth of
the
metaphysician. Materially, however, the considerations of Euler rest entirely on the foundations on which Newton had erected the classical system of mechanics. Newton s concepts of absolute space and absolute time are here to be revealed not only as the
necessary fundamental concepts of mathematico-physical knowl edge of nature, but as true physical realities. To deny these realities
on philosophical, on general epistemological grounds, means, as Euler explains, to deprive the fundamental laws of dynamics above In such an all the law of inertia of any real physical significance. the outcome cannot be questioned the philoso alternative, however, pher must withdraw his suspicions concerning the "possibility" of an absolute space and an absolute time as soon as the reality of both can be shown to be an immediate consequence of the validity of the fundamental laws of motion. What these laws demand, also and it is, it exists in the highest sense and highest degree of objectivity which is attainable for our knowledge. For before the reality of nature as it is represented in motion and its empirical laws all logical doubt must be silent; it is the business of thought to accept the exist ence of motion and its fundamental rules instead of attempting to prescribe to nature itself from abstract considerations concerning what can or cannot be conceived. This demand, however, illuminating as it appears and fruitful as the methodic stimulus of Euler proved in the development of the Kantian problem, 1 becomes problematical when considered from the standpoint of modern physics and epistemology. Kant believed that he possessed in Newton s fundamental work, in the Philosophiae
:
"is"
Naturalis Principia Mathematica, a fixed code of physical

1
"truth"
For more detail concerning Euler and Kant

(7), II,
relation to him,
cf.,
Erkennt-
nisproblem
472
ff.,
698, 703
f.
CONCEPTS OF MEASURE AND OF THINGS
353
and believed that he could

;
definitively
ground philosophical knowl
edge on the "factum" of mathematical natural science as he here found it but the relation between philosophy and exact science has
Ever more clearly, ever more comtoday that the Archimedean point on which Kant supported himself and from which he undertook to raise the whole system of knowledge, as if by a lever, no longer offers an uncondition
since changed fundamentally.
pellingly
do we
realize
ally fixed basis.
The factum of geometry has
lost its
unambiguous
find ourselves
definiteness; instead of the one geometry of Euclid,
we
facing a plurality of equally justified geometrical systems, which all claim for themselves the same intellectual necessity, and which, as the example of the general theory of relativity seems to show, can
rival the
system of
classical
fruitfulness for physics.
And
geometry in their applications, in their the system of classical mechanics
has undergone an even greater transformation, since in modern physics the "mechanical" view of the world has been more and more
superseded and replaced by the electro-dynamic view. The laws, which Newton and Euler regarded as the wholly assured and im pregnable possession of physical knowledge, those laws in which they believed to be defined the concept of the corporeal world, of matter and motion, in short, of nature itself, appear to us today to be only abstractions by which, at most, we can master a certain region, a definitely limited part of being, and describe it theoretically in a
first approximation. And if we turn to contemporary physics with the old philosophical question as to the "essence" of space and time, we receive from it precisely the opposite answer to that which Euler
gave the question a hundred and

cepts of absolute space
fifty
years ago.
Newton
count
con
and absolute time may
still
many
among the "philosophers," but they seem definitively removed from the methodic and empirical foundations of physics. The general theory of relativity seems herein to be only the ultimate consequence of an intellectual movement, which receives its decisive s motives equally from epistemological and physical considerations. s The working together of the two points of view has always come
adherents
to light with special distinctness at the decisive turning points in the evolution of theoretical physics. A glance at the history of physics
shows that precisely
its
most weighty and fundamental achievements
stand in closest connection with considerations of a general episte mological nature. Galileo s Dialogues on the Two Systems of the
354

filled
World are
with such considerations and his Aristotelian oppo
nents could urge against Galilei that he had devoted more years to the study of philosophy than months to the study of physics. Kepler lays the foundation for his work on the motion of
of the world in which he gives a complete methodo Apology for Tycho, logical account of hypotheses and their various fundamental forms; an account by which he really created the modern concept of physical Newton also, in the theory and gave it a definite concrete content. midst of his considerations on the structure of the world, comes back to the most general norms of physical knowledge, to the regulae In more recent times, Helmholtz introduces his philosophandi. Uber der Erhaltung der Kraft (1847), with a consideration of work,
Mars and
for
his
chief
work on the harmony
his
in
the causal principle as the universal presupposition of all "comprehensibility of nature," and Heinrich Hertz expressly asserts in the
preface of his Prinzipien der Mechanik (1894), that what is new in the work and what alone he values is "the order and arrangement of the whole, thus the logical, or, if one will, the philosophical side of
the
subject."
But
all
these great historical examples of the real
epistemological problems and physical are almost outdone by the way in which this connection problems has been verified in the foundation of the theory of relativity. Ein
inner connection between
especially in the transition from the special to the of relativity appeals primarily to an epistemological general theory
stein himself
motive, to which he grants, along with the purely empirical and 3 And even the special physical grounds, a decisive significance.
theory of relativity is such that its advantage over other explana such as Lorentz s hypothesis of contraction, is based not so much on its empirical material as on its pure logical form, not so
tions,
much on
4 In this physical as on its general systematic value. connection the comparison holds which Planck has drawn between
its
the theory of relativity and the Copernican cosmological reform. 5 The Copernican view could point, when it appeared, to no single new by which it was absolutely demanded to the exclusion of all
"fact"
earlier astronomical explanations,

2 3
but
its
value and real cogency
Helmholtz (29, p. 4); Cf. Einstein (17, p. 8).

Cf.
H. Hertz
(31, p.
XXVII).
See below, Sect.,
II.
f.
>Cf.
Planck
(68), p. 117
355
lay in the fundamental and systematic clarity, which it spread over the whole of the knowledge of nature. In the same way, the theory
of relativity, taking its start in a criticism of the concept of time, extends into the field of epistemological problems, not merely in its applications and consequences, but from its very beginning. That the sciences, in particular, mathematics and the exact natural sciences furnish the criticism of knowledge with its essential material is
scarcely questioned after Kant; but here this material is offered to philosophy in a form, which, even of itself, involves a certain episte-
mologiâl interpretation and treatment. Thus, the theory of relativity, as opposed to the classical system of mechanics, offers a new scientific problem by which the critical
philosophy must be tested anew. works on Kant urged repeatedly
If
Kant
as
Hermann Cohen
all
and proved from
angles
intended to be the philosophical systematizer of the Newtonian natural science, is not his doctrine necessarily entangled in the fate
of the
Newtonian physics, and must not all changes in the latter react directly on the form of the fundamental doctrines of the critical philosophy? Or do the doctrines of the Transcendental Aesthetic
a foundation, which is broad enough and strong enough to bear, along with the structure of the Newtonian mechanics, also that of modern physics? The future development of the criticism of
offer
If it is will depend on the answer to these questions. shown that the modern physical views of space and time lead in the end as far beyond Kant as they do beyond Newton, then the time would have come when, on the basis of Kant s presuppositions, we would have to advance beyond Kant. For the purpose of the Critique of Pure Reason was not to ground philosophical knowledge once for all in a fixed dogmatic system of concepts, but to open up
knowledge
for
it
the
"continuous
development of a
science"
in
which there can
be only relative, not absolute, stopping points.
Epistemology, however, closely as its own fate is connected with the progress of exact science, must face the problems which are presented to it by the latter, with complete methodic independence. It stands to physics in precisely the relation, in which, according to the Kantian account, the "understanding" stands to experience and order to be taught by it: but nature: it must approach nature
"in
not in the character of a pupil, who agrees to everything the master to answer likes, but as an appointed judge, who compels the witnesses
356
the questions which he himself proposes." Each answer, which physics imparts concerning the character and the peculiar nature of
fundamental concepts, assumes inevitably for epistemology the form of a question. When, for example, Einstein gives as the essen tial result of his theory that by it "the last remainder of physical objectivity" is taken from space and time (17, p. 13), this answer of
its
the physicist contains for the epistemologist the precise formulation of his real problem. What are we to understand by the physical
objectivity,
which
is
To
the physicist physical objectivity
here denied to the concepts of space and time? may appear as a fixed and sure
and as an entirely definite standard of comparison; must ask that its meaning, that what is to be expressed epistemology by it, be exactly defined. For epistemological reflection leads us everywhere to the insight that what the various sciences call the
starting-point
is nothing given in itself, fixed once for all, but that it is determined by some standpoint of knowledge. According to the changes of this ideal standpoint, there arise for thought various classes and various systems of objects. It is thus always necessary to recognize, in what the individual sciences offer us as their objects and "things," the specific logical conditions on the ground of which they were established. Each science has its object only by the fact that it selects it from the uniform mass of the given by certain formal concepts, which are peculiar to it. The object of mathematics is different from that of mechanics, the object of abstract mechanics different from that of physics, etc., because there are contained in all
"object"
first
these sciences different questions of knowledge, different ways of referring the manifold to the unity of a concept and ordering and mastering the manifold by it. Thus the content of each particular
knowledge is determined by the characteristic form of judg ment and question from which knowledge proceeds. In the form of judgment and question the particular special axioms, by which the sciences are distinguished from each other, are first defined, if we
field of
attempt to gain a objectivity" from

feature.
definite explanation of the concept of

this standpoint,
"physical
we
are
first
coincide with
Whatever this objectivity may mean, what the naive view of the world
led to a negative in no case can it

is
accustomed to
regard as the reality of things, as the reality of objects of sensuous perception. For the objects, of which scientific physics treats and
for
which
it
establishes its laws, are distinguished
from
this reality
357
by their general fundamental form. That concepts, such as those of mass and force, the atom or the ether, the magnetic or electrical
even concepts, like those of pressure or of temperature, are no simple thing-concepts, no copies of particular contents given in perception: this scarcely needs any further explanation, after all That the epistemology of physics itself has established concerning the meaning and origin of these concepts. What we possess in them are obviously not reproductions of simple things or sensations, but the oretical assumptions and constructions, which are intended to trans form the merely sensible into something measurable, and thus into an Planck s "object of physics," that is, into an object for physics.
potential,
neat formulation of the physical criterion of objectivity, that every thing that can be measured exists, may appear completely sufficient
from the standpoint of physics from the standpoint of epistemology, it involves the problem of discovering the fundamental conditions of this measurability and of developing them in systematic complete\ness. For any, even the simplest, measurement must rest on certain theoretical presuppositions on certain "principles," "hypotheses," or from the world of sense, but which "axioms," which it does not take
;
,
In this sense, tEe brings to this world as postulates of thought. reality of the physicist stands over against the reality oflmmediat
it
perception as something through and through mediaj;e4i &s a system. not of existing things or properties, but of abstract intellectual symv
L.
~T5ols7wnicn serve to express certain relations of magnitude and measure, certain functional coordinations and dependencies of phenoma&a.. If we start from this general insight, which within physics itself has
been made very
clear, especially
by Duhem
analysis of the physical
construction of concepts, the problem gains its full logical definiteness. That physical objectivity is denied to space and time by this theory must, as is now seen, mean some thing else arid something deeper than the knowledge that the two are
of the theory of relativity
not things in the sense of "na ive realism." For things of this sort, we must have left behind us at the threshold of exact scientific
physics, in the formulation of its first judgments and propositions. The property of not being thing-concepts, but pure \ concepts of
measurement, space and time share with all other genuine physical have a concepts; if, in contrast to these, space and time are also to it must be shown that they are removed in special logical position,
the same direction as these, a step further from the ordinary thing-
358
EINSTEIN
THEORY OF RELATIVITY
concepts, and that they thus represent, to a certain extent, concepts and forms of measurement of an order higher than the first order.
The fact appears even in the first considerations, from which the theory of relativity starts, that the physicist does not have only to hold in mind the measured object itself, but also always the particular
conditions of measurement. The theory distinguishes between physical determinations and judgments, which result from measure
ment from resting and moving systems
of reference,
and
it
emphasizes
the fact that before determinations, which have been gained from diverse systems of reference, can be compared with each other, a
must be
universal methodic principle of transformation and permutation To each objective measurement, there must be given.
added a certain subjective index, which makes known its particular conditions and only when this has taken place can it be used along
with others in the scientific construction of the total picture of reality, in the determination of the laws of nature, and be combined with
these others into a unitary result. What is gained by this reflection on the conditions of physical measurement in a pure epistemological
regard appears as soon as one remembers the conflicts, which have (tilted from the lack of this reflection in the course of the history
It seems almost the unavoidable^" approach to the world that each new and fruit ful concept of measurement, which it gains and establishes, should be transformed at once into a thing-concept. Ever does it believe that the truth and the meaning of the physical concepts of magnitude
of philosophy
and
of exact science.
fate of the scientific
are assured only spond to them.

lates
when it permits certain absolute realities to corre Each creative epoch of physics discovers and formu
characteristic measures for the totality of being and natural but each stands in danger of taking these preliminary and process, relative measures, these temporarily ultimate intellectual instruments of measurement, as definitive expressions of the ontologically real. The history of the concept of matter, of the atom, of the concepts of
new
the ether and of energy offer the typical proof and examples of this. All materialism and there is a materialism not only of "matter"
but also of force, of energy, of the ether, etc., goes back from the standpoint of epistemology, to this one motive. The ultimate constants of physical calculation are not only taken as real, but they are ultimately raised to the rank of that which is alone real. The development of idealistic philosophy itself is not able to escape this

tendency.
359
Descartes as an idealistic mathematician was at the of the "mechanical view of the world." Since only extension offers us exact and distinct concepts and since all clearly comprehended truth is also the truth of the existing, it follows, in his view, that mathematics and nature, the system of measure-
same time the founder
vments and the totality of material existence, must be identified. The manner, in which the same step from the logico-mathematical
to the ontological concept has been repeated in the development of modern energetics, is known. Here, after energy has been discovered
is not limited to the but spans equally all physical fields, it was made into an all-inclusive substance, which rivalled "matter" and But on the whole, we are finally took it up into itself completely. here concerned with a metaphysical by-way, which has not seduced science itself from its sure methodic course. For the concept of
as a fundamental measure, as a measure which

of motion,
phenomena
energy belongs in conception to that general direction of physical thought, which has been called the "physics of principles" in con A "prin trast to the physics of pictures and mechanical models. of ciple," however, is never directly related to things and relations things, but is meant to establish a general rule for complex functional dependencies and their reciprocal connection. This rule proves to be the really permanent and substantial: the epistemological, as well as the physical, value of energetics is not founded on a new
be substituted for the old concepts of but on the gaining of equivalence-numbers, such as were expressly demanded and discovered by Robert Mayer as the "foundation of exact investigation of nature." (Cf. 52, p. 145. 237 ff.) Even in these two examples we can learn that through the whole history of physics there is a certain intellectual movement, which
pictorial representation to
"matter"
and
"force"
throughout runs parallel to the movement in epistemology that mediates and passes to and from between the "subject" and the of knowledge. "object" Physical thought is always concerned at
with establishing a characteristic standard of measurement in Then in a certain natural constant. with understanding more it is concerned, in the further development, and more clearly the constructive element that is contained in any such original constant, and with becoming conscious of its own
first
an objective physical concept,
conditionality. For, whatever particular properties they may have, no constants are immediately given, but all must be conceived and
360
One of the most sought before they can be found in experience. pregnant examples of this is found in the history of the concept of
the atom.
The atoms were postulated by Democritus as ultimate constants of nature long before thought possessed any means of concretely realizing this postulate. Fundamentally, such a realiza
such a strictly quantitative meaning of the concept of the atom, in the beginnings of modern chemistry in the law
tion,
was only reached
To the extent, however, that we add, to this of multiple proportion. particular realization of the concept of the atom in the law of mul
and still others and the concept of the atom comes to characterize and to organize intellectually the most diverse fields, its character as a pure principle, which was originally fused with its thing-character, comes to light. The content of the idea of the atom changes and shifts from place to place in the course of the development of physics and chemistry, but the function of the atom as the temporarily ultimate unit of measurement remains. When we pass from the consideration of "ponderable" matter to the consideration of the ether, when we seek a unity, which compre hends not only the mechanical but also the optical and electrical phenomena, the atom of matter becomes the atom of electricity, the electron. In recent physics, there appears further, with Planck s Quanta Theory the thought of an atomistic structure not only of matter but of energy. It would be in vain were one to attempt to combine all these various applications of the concept of the atom in chemistry, in the kinetic theory of gases, and in the doctrine of light and heat radiation, etc., into a unitary picture. But the of its meaning requires no such pictorial unity; it is satisfied, unity indeed verified in a far stricter logical sense, when it is shown that
tiple proportion, others
finally
here a
common
relation, a peculiar
"form"
of connection, prevails,
in the
which as such can be

contents.
verified
and represented
most diverse
The atom shows
itself
thereby to be, not an absolute
minimum
of
minimum of measure. It was one modern philosophy, Nicholas Cusanus, who, with true speculative profundity, anticipated and announced this as the function of the concept of the atom, which was to be actually
of being, but a relative
the founders of
realized only in the history of natural science. Cusanus fundamental doctrine of the infinite and of the unity of opposites in the infinite
rested entirely on this insight into the relativity in principle of all determinations of magnitude, on the coincidence of the "greatest"
361
and the "smallest." knowledge brings the
(Cf. 7, I,
riddle,
40 ff, 265 ff.) Modern criticism of with which Cusanus doctrine of the
minimum struggles, to enters when we attempt

different forms,
a simple expression. Contradiction only to unify after the fashion of a thing all the
which the thought of the "smallest" assumes, in the but it disappears as soon as we reflect that the true unity is never to be sought in things as such, but in intellectual constructions, which we choose according to the peculiar ity of the field to be measured, and which are thus in principle pos sessed of an unlimited variability. It follows from this that, as what is to be measured is unlimited in variety, so what measures can be represented in infinitely many and infinitely diverse ways. In other words, the unity that we have to seek lies neither in the one nor the other member, but merely in the form of their reciprocal connection, i.e., in the logical conditions of the operation of measurement itself. This receives new confirmation when we pass from the concept of matter, of energy and of the atom to the real concept of objectivity
different fields of thought;
of modern physics, that of motion. modern theory of motion in Galileo

cal question,
The
historical beginnings of the
which has
refer directly to the epistemologireceived its definitive formulation in the
general theory of relativity. What Galileo gained with his idea of relativity was the cancelling of the absolute reality of place, and this first step involved for him the most weighty logical consequences,
viz.,
the
new concept
of the lawfulness of nature
and the new
inter
pretation of the particular laws of dynamics.
Galileo s doctrine of
motion
a
of
is
rooted in nothing less and nothing more than in the choice of

in the universe.
new standpoint from which to estimate and measure the phenomena

motion
By
this choice, there
was given him ac
once theJw-4ineiiaj3,nd in it the real foundation of the new view ot nature. The ancient view saw in place a certain physical property and "there," that produced definite physical effects. The the "above" and "below," were for it no mere relations; but the particular point 6f space was taken as an independent real, which consequently wast provided with particular forces. In the striving
"here"
of bodies to their\
"natural places," in the pressure of air and fire upwards and in th^ sinking of heavy masses downwards, these forces seemed given as iminediate empirical realities. Only when one takes
account of these fundamental features, not only of ancient astron omy and cosmology^ but also of ancient physics, does one under-
362
EINSTEIN
resulting
stand the boldness of the new intellectual orientation,
from the Copernican system of the world. One of the most fixed and certain realities on which Grecian thought had constructed its picture of the world now became a mere illusion, a purely "subjective" fea ture. Even the first adherents of the new doctrine drew the decisive
conclusion with reference to the doctrine of place. What Gilbert, e.g., urges against the Aristotelian physics and cosmology is above
all
this
and the
our
epistemological feature, i.e., that it confuses the ideal real. Differences belonging merely to our thought, to; subjective reflection, are throughout made into objective
But in truth, no place in itself is opposed to any oppositions. but there are in nature only differences in the mutual posi other,
tions of bodies
and
of material masses.
"It
is
not place which,
works and produces, which determines the rest and motion of bodies. For it is in itself neither a being nor an effective cause; rather bodies determine their mutual place and posi tion by virtue of the forces which are immanent in them. The is a nothing; it does not exist and exerts no force, but all natural place power is contained and grounded in bodies themselves." (7, I, 360 f.) It is implied in this that what we call the "true place" is never given to us as an immediate sensuous property, but must be discovered on the basis of calculation and of the "arithmetic of
in the nature of things,
forces"
in the universe. All determination of place as Kepler sharply and clearly expresses this insight which for him resulted equally from astronomical convictions, physiological optics and
is a work of the mind^: From this opus. (37, II, 55, cf. 7, I, 339.) point the way is open to Galileo s foundation of dynami cs :for since place has ceased to be something real, the question as to the ground of the place of a body and the ground of its persistence in one and
analysis of the general problem of perception
omnis
locatio
mentis
est
the same place disappears. Objective physical reality passes from place to change of place, to motion and the factors by which it is
in
determined as a magnitude. If such a determination is to be possible a definite way, the identity and permanence, which were hitherto
ascribed to mere place, must go over to motion; motion must possess that is, from the standpoint of the physicist, numerical "being,"
This demand for the numerical constancy of motion itself and its realization in the law of inertia. We recognize here again how closely, in Galileo, the mathematical motive
constancy.
finds its expression
363
of his thought was connected with an ontological motive, how his conception of being interacted with his conception of measure.*
The new measure, which is found in ineî^ and in the concept of uniform acceleration, involves also anew determination of reality. In contrast with mere place, which is infinitely ambiguous and differs
;-.
according to the choice of the system of reference, the inertial movement appears to be a truly intrinsic property of bodies, which belongs to them themselves" and without reference to a definite system of mrison and measurement The velocity of a material system is more than a mere factor for calculation; it not only really belongs to
"in
the system but defines its reality since it determines its vis viva, In its measure of i.e., the measure of its dynamic effectiveness. in the differential quotient of the space by the time, Galileo s motion,
physics claims to have reached the kernel of have defined the intensive reality of motion.
all
physical being, to By this reality, the
dynamic consideration
is
The concept
of the
"state
distinguished from the merely phoronomic. of motion," not as a mere comparative
magnitude, but as an essential element belonging to the moving system intrinsically, now becomes the real mark and characteristic of physical reality. Leibniz, too, in his foundation of dynamics, stands throughout at this standpoint, which becomes for him a Motion conceived as starting-point for a new metaphysics of forces. a mere change of place in the purely phoronomical sense, he explains, remains always something purely relative it only becomes an expres sion of a true physical and metaphysical reality when we add to it an
;
inner dynamic principle, a force conceived as an

principle of
"originally
implanted
permanence and
(42,
per sever antiae.

it is
VI, 100
cf.
change," principium mutationis et 290 ff.) In all these examples, 5, p.
hand the physical thought of modern times has grasped the thought of the relativity of place and
evident
sharply, on the one
of motion, and, on the other hand, how it has shrunk back from following it to its ultimate consequences. If not only place but the
how
velocity of a material system is to signify a magnitude that entirely depends on the choice of the reference body and is thus infinitely
variable
and
infinitely
ambiguous, there seems no possibility of an
exact determination of magnitude and thus no possibility of.aa, exact objective determination of the state of physical reality. Pure
mathematics may be constructed as the ideal doctrine of the compari son and connection of magnitudes, as a system of mere relations and
364
functions and
may come to recognize itself as such ever more clearly, but physics seems necessarily to reach an ultimate limit, a non plus ultra, if it is not wholly to lose any basis in reality. The difficulty, which remains in the structure of classical mechanics
,,
^\
\
is expressed in an epistemological circle, from which there seems no escape. To understand the meaning of the law of inertia, we need the concept of "equal times" but a practicable physical measure of equal times can,
in the formulation of the principle of inertia,
as
is
discovered, only be gained
by presupposing,
fact, since
in its content
and
validity, the
law of
inertia.
In
Carl
Neumann
work,
Uber die Prinzipien der Galilei-Newton schen Theorie (57), which set motion the modern discussion on the law of inertia, it is customary in mechanics to define equal times as times within which a body left
in
to itself traverses equal distances. Maxwell too, in his exposition of the Newtonian mechanics, conceived the law of inertia as a pure
The first law of Newton, as he explains and pregnantly, tells under what conditions no external force is present Thus in the progress of mechanics the princi (5 1 p 3 1 ) of inertia is recognized with increasing distinctness as what it ple meant fundamentally to Galileo. It is no longer taken as a direct
definition of measure.
clearly
empirical description of given processes of nature, but as the "axiom of the field," the fundamental hypothesis by which the new science
dynamics prescribes to itself a certain form of measurement. Inertia appears, not as an absolute and inherent property of things and of bodies, but as the free establishment of a certain standard and
of
symbol
of
measurement, by virtue of which we can hope to reach a

In this alone
is
systematic conception of the laws of motion.

its reality, i.e., its
rooted
objective and physical significance. Thus, within the historical development of physics itself what measures is separated with increasing distinctness from what is measured, with which it at
seems to coincide the observable data of experience are separated with increasing distinctness from what must be presupposed and used as a condition of observation and of measurement.
first
;
And what
field is
here seen in a particular example and within a narrow repeated, on closer examination, in all the special fields of
is
Everywhere physical thought must determine for itself of measurement before it proceeds to observation. There must be established a certain standpoint for the comparison and correlation of magnitudes; certain constants must be established
physics.
its
own standards

at least hypothetically
365
\ contains a
with our
it
measurement can take
and in preliminary fashion before a concrete In this sense, each measurement place.
not so much with the sensuous we measure natural processes as
is
purely ideal element; it instruments of measurement that
thoughts. The instruments of measurement are, as the visible embodiments of these thoughts, for each of were, only them involves its own theory and offers correct and useful results
own
ff.)
only in so far as this theory is assumed to be valid. (Cf. 8, p. 189 Itîs not clocks and physical measuring-rods but principles ari^l
postulates that are the real instruments of measurement For in the multiplicity and mutability of natural phenomena, thought In the possesses a relatively fixed standpoint only by taking it.
choice of this standpoint, however, it is not absolutely determined by the phenomena, but the choice remains its own deed for which ulti mately it alone is responsible. The decision is made with reference
i.e., to the connection of observations according to law, not prescribed in a definite way by the mere sum of observa tions. For these in themselves can always be expressed by a number of intellectual approaches between which a choice is possible only with reference to logical "simplicity," more exactly, to systematic
to experience,
but
it is
unity and completeness, of scientific exposition.

in accordance with its claims
"simple"
When
thought,
and demands, changes the form of the fundamental measuring relations, we stand before a new
The previously of the world with regard to content also. gained relations of experience do not indeed lose their validity, but, since they are expressed in a new conceptual language, they enter
"picture"
Archimedean point of the former view of the world moves; the previous TTOV OTW of thought appears transcended. But it is soon seen that thought, by virtue of its peculiar function, can only transcend an earlier construction by replacing it by a more general and more inclusive one; that it only shifts, among phenomena, the constancy and identity, which it cannot cease to demand, to another and deeper place. That every
into a
new system
of meanings.
The
fixed
which the demand of thought for ultimate constants can find within the empirical world is always only conditioned and rela tive, is guaranteed by the unconditionality and radicalism of precisely The critical theory of knowledge would not only show this demand.
realization,
but for it the concrete movement of the continual oscillation between experience and concept, thought,
this connection in abstracto,
366

in the history of physics,
forms a In the midst of the change perpetually of particular theoretical instruments of measurement, the critical theory holds fast to the thought of the unity of measurement, which indeed signifies for it no realistic dogma but an ideal goal and a never-
between facts and hypotheses
new
source of instruction.
Each new physical hypothesis erects, as it were, a new logical system of coordinates, to which we refer phenom ena, while nevertheless the doctrine is retained as a regulative
to-be-concluded task.
idea for investigation that
definite limiting value.
all these systems converge on a certain In the confusion and continuous flux of
phenomena, the understanding seems at first almost arbitrarily to fix and separate out certain points in order to learn through them a definite law of change, but everything which it regards as determined and valid in this sense proves, in the course of further progress, to be a mere approximation. The first construction must be both limited and more exactly defined logically by the second, this again by the
third, etc.
Thus, ever anew does the temporarily chosen theoretical
center of thought shift; but in this process, the sphere of being, the
thought.
facts
sphere of objective knowledge, is more and As often as it seems that thought
is
more penetrated by overturned by new
and observations, which are outside
its
previously formulated
laws, it is seen that, in fact, thought has found in them a new point of leverage, around which moves henceforth the totality of empiri
cally provable
"facts."
The
tion of each
ideal center
new
physical theory
epistemological exposition and evalua must always seek to indicate the
and turning-point around which it causes the totality phenomena, the real and possible observations, to revolve, whether this point is clearly marked or whether the theory only refers to it indirectly by the intellectual tendency of all its propositions and
of
deductions.
CHAPTER
11
THE EMPIRICAL AND CONCEPTUAL FOUNDATIONS OF THE THEORY OF

RELATIVITY
there can be no doubt, according to the opening words of the Critique of Pure Reason, that all our knowledge begins with experi
If
ence, then this holds especially when we are concerned with the The question here can never run as to origin of a physical theory.
whether the theory has issued from experience but merely as to how it is based on experience, and what is the relation of the diverse ele
ments which characterize and make up the concept of experience as such. There is accordingly needed no special epistemological analysis to make clear the relation of the special and general theories of relativity to experience, to the whole of observation and of physical experiment; such an analysis will only have to decide whether the theory in its origin and development is to be taken as an example and witness of the critical or of the sensualistic concept of experience. Does "experience," as it is used here, mean merely the bare sum of particular observations experimentorum multorum coacervatio, as a sensualistic thinker once described it or is there involved in it an
independent intellectual form? Is the construction of the theory to merely a matter of joining perception to perception, in this connection of particulars, have there been effective all or, along certain universal and critical norms, certain methodic presup No "empiricism" however extreme can ever seek to deny positions?
"fact"
"fact,"
the role of thought in establishing and grounding physical theories, and just as little is there, on the other hand, a logical idealism, which would attempt to free "pure thought" from reference to the world of the
"factual" and from being bound to it. The question dividing the two views can only be as to whether thought consists in a simple registration of facts, or whether, even in the establishment, in gaining an interpretation of "particular facts," thought reveals its charac
teristic
Is its work completed in arranging immediately taken from sense perception, like or does it face them with its own original measures pearls on a thread
power and function.

data,
particular
as independent criteria of
judgment?
367
368
The problem raised here received its first sharp and clear systematic formulation in the Platonic doctrine of ideas. For Platonic idealism,
too, the proposition holds that
it is
basis
of
some perception:
of the
ov
But the function

the
"logic
not possible to think save on the dwarov tvvotv tf e/c TWOS in consists indeed not in finding
at<r$7J<reoos.
us"
sum
the
"idea
of the particular perceptions, not in deriving and deducing of the equal" from the "equal pieces of wood and stone,"
us"
in is revealed in discriminating and judging what "logic given in perception. This discriminating constitutes the real fundamental character of thought as diavoic*, as discursus. Not
is
but the
perceptions and observations stimulate equally the critical and discriminating activity of thought. There are some which do not summon the understanding to reflection, since satisfaction is done
all
them by mere
forth
thought,
solid.
nothing
sensation, but there are others which in all ways call as in their case perception by itself could gain "Not stimulating, namely, is that which does not pass
into an opposite perception; stimulating objects I call those which give opposite perceptions, because here perception gives no more vivid
of its opposite. Much in percep a paraclete of thought (irapanXriTiKa. TT?S diavoiai), while other perceptions are not such an awakener of thought, namely, is ihich comes into sense at the same time as its opposite; but what does not, that also does not arouse thought." (Republic
idea of
tion
is
any particular object than
pj!(>rythi.-n.(i
523-524.)
In this Platonic characterization of the relation of thought
and sensation, of reason and sensibility, we have, as Cohen has urged, "one of the most fundamental thoughts in the evolution of the cri tique of cognition." (12, p. 16 ff .) Just as for Plato thought becomes what it is in assertion and contradiction, in dialectic, so only a perception to which this feature corresponds can become the awakener and paraclete of thought. The dialectic of percep tion summons that of thought to judgment and decision. Where the perceptions, as it were, rest peacefully side by side, where there is no inner tension between them, thought rests also; only where they contradict each other, where they threaten to cancel each other does thought s fundamental postulate, its unconditional demand for unity stand forth and demand a transformation, a reshaping of
experience itself. The evolution of the theory of relativity has furnished a new It \vas in fact a fundamental typical proof of this general relation.
contradiction between physical experiments from which the theory
_^.^
EMPIRICAL AND CONCEPTUAL FOUNDATIONS
369
of relativity took its start. On the one side stood the investigation of Fizeau, on the other, that of Michelson, and the two seemed in their results absolutely irreconcilable. Both sought to gain an answer to the question as to how the velocity of light in a moving medium was related to its velocity in a resting medium; and they answered
ways. The investigation of Fizeau showed that the velocity of light in flowing water was greater than in water at rest; that, however, not the whole velocity of the
this question in completely opposite
flowing water, but only a fraction of it was added to the velocity of medium at rest. If we call the velocity of light in the and the velocity of light in a medium at rest w moving medium
light in a
and the velocity
of the flowing
v,
it
results not simply that
w+
n
v,
r>
but rather that

signifies
W = w -f v
),
in
which the magnitude

This
the exponent of the refraction of the liquid.
result, as interpreted by the theory of Lorentz, spoke directly for the assumption of a motionless ether not carried along by the body in its movement. But the attempt of Michelson, to discover the consequences of the movement of the earth with reference to this motionless ether, failed. In no way could any influence be shown of the motion of the earth on the velocity of the propagation of light; it was rather shown with increasing evidence that all optical phenom ena take place as if there were no translation of the earth against the ether. 1 And behind this conflict of there stood, as one
"facts"
was forced to recognize more and more, a conflict in general principles, to which the theories of mechanical and of optical and electromagnetic phenomena seemed to lead necessarily. Experiments in the latter could finally be summarized in a single proposition, the principle of the constancy of the velocity of light in a vacuum. The validity of the fundamental equations of electrodynamics of Maxwell and Hertz involved the assumption that light in an empty space is
always propagated with a definite velocity V independently of the motion of the body emitting it. From whatever system one made the observation and from whatever source the light issued there would always be found the same determinate value for its
state of
velocity of propagation.
But
this
assumption of the velocity of
1 For more detail concerning the investigations of Fizeau and Michelson as well as concerning the negative outcome of other investigations on the influ ence of the movement of the earth on optical and electrical phenomena, cf.
Laue
(40), p. 10
ff.
370
EINSTEIN
same for all systems, necessarily principles of electrodynamics, now comes into opposition with the principle of relativity of the Galileo-Newtonian
light as a universal constant the
demanded by the
mechanics.
reference
"left
body
This principle demands that, when any definite Galileian is given i.e., a body relatively to which a body
persists in its state of rest or of
to
itself"
uniform motion in a
straight line
body K K which
,
the laws, which are valid relatively to this reference remain valid when one passes to the system of reference
all
is,
In the transition from

formation"
with reference to K, in uniform translatory motion. to the equations of the "Galileo-trans with hold, (where v signifies a constant "velocity of
x
reference to
vt,
y, z
K parallel to the x and x axes), to which there is to be added the identical transformation from the time t = t, which is
not especially noticed in classical mechanics. If we seek, however, to apply the principle of relativity of mechanics to electrodynamics,
i.e.,
to recalculate its equations according to the formulae of the
it is seen that this cannot be done: the electrodynamic equations, in contrast to the Newtonian equations of motion, alter their form when we insert the coordinates x y z t into them in place of the coordinates x, y, z, t according to the rules
Galileo-transformation,
of the Galileo-transformation.
The
effort to unite
mechanics and
electro-dynamics by carrying over the principle of relativity of the first into the second thus has to be given up the Hertzian theory,
:
which represented such an attempt, came into irreconcilable
conflict
with assured experimental results. Physical investigation stood before the dilemma of giving up a principle which had been verified
without exception in all the phenomena of motion and which formed a corner-stone in the structure of classical mechanics or of retain ing it within its field but denying its applicability to optical and electromagnetic phenomena. In both cases, the unity of the explana tion of nature, the unity of the very concept of nature, seemed de
stroyed. Here in fact the condition set up by Plato of the intellec tual fruitfulness of experience was fulfilled: here experience stood at a point at which assured observation seemed to pass directly into
The conflict between the principle of the constancy opposite. of the propagation of light and the principle of relativity of mechanics became the "paraclete of thought" the real awakener of the theory
its
of relativity.
371
since
But how did physical thought go about overcoming this it was bound to the outcome of observation as such,
conflict,
since
it
could neither set aside the facts expressed in the principle of the constant velocity of light in a vacuum, nor those expressed in the If we look back on the histori principle of relativity of mechanics?
cal
development of the theory
of relativity,
we
recognize that the
which was once given by Goethe. art in theoretical and practical Hfq," wrote Goethe to_ "The greatest Zelter, "consists in changing the problem into a postulate; that way__ one succeeds." In fact, this was the course which Einstein followed in his fundamental essay. Zur Ekktrodynamik bewegter Systeme
latter has followed here a counsel
of"
the year 1905.
principle of the constancy of the velocity of light was given first place as a postulate, but, supported by the negative result of all attempts to establish an "absolute" motion with reference to a chosen system of reference, i.e., the "motionless
ether,"
The
the supposition was made that there correspond to the concept of absolute rest no properties of phenomena in either mechanics or electrodynamics, but rather that the same electrodynamic and optical laws hold for all systems of coordinates of
which the mechanical equations hold. And this "supposition" does not continue such, but is expressly "made a presupposition," i.e., a shaping of theory is demanded which will simultaneously satisfy the conditions of the principle of relativity and those of the principle of
The two assump the constant propagation of light. (Cf. 16, p. 26). to the means and habits of tions are indeed not compatible according
thought at the disposal of the kinematics generally accepted before the establishment of the theory of relativity, but they ought no longer to be incompatible. The demand made of physical theory was that it remove this incompatibility by subjecting precisely these
means and habits
of thought themselves to a critical examination. an analysis of the physical concepts of space and time, it now By
appears that in fact the incompatibility of the principle of relativity with the law of the propagation of light is not to be found; that rather
is only needed a transformation of these concepts in order to reach a logically unobjectionable theory. The decisive step is taken when it is seen that the measurements, to be gained within a system by definite physical methods of measurement, by the application of
there
fixed measuring-rods
once for
all,
and clocks, have no "absolute" meaning fixed but that they are dependent on the state of motion of
372
EINSTEIN
latter.
V"
the system and must necessarily result differently according to the There now arises the purely mathematical problem of dis-
covering the laws of permutation, according to which the spacetime values of an event are changed in going from one reference
body to another, which

to the
first.
is
in
This problem
uniform translatory motion with regard is solved, as is known, by the funda

x
mental equations of the
"Lorentz-transformation:"
vt
y = y
*-**
On the
basis of these equations,
we
see that the law of the propaga
tion of light in a
K and K
equally fulfilled for all justified systems on the other hand, it is seen that Maxwell s fundamental
is
vacuum
equations of electrodynamics do not change their form when the formulae of the Lorentz-transformation rather than those of the
There is thus a univer which comprehends the totality of physical phenomena; the laws, according to which the states of physical systems change, are independent of whether they are referred to one or the other of two systems of coordinates in uniform translatory motion relative to each other. (Cf. 16, p. 29). The principle of relativity of classical mechanics is so little contradicted by this
Galileo-transformation are applied to them.
sal principle of relativity,
general principle that it is rather contained in it as a special case; the equations of the Galileo-transformation directly issue from those
of the Lorentz-transformation
ties
that the values
v as are very small v v2

cz c 2
in
considers only such veloci with the velocity of light so comparison

left
when one
can practically be
out of account.
It follows
from
this that the principle of relativity of electrodynamics, carried
over to mechanics, can come into conflict with no empirical result, while the converse carrying-over of the principle of relativity of mechanics to electrodynamics proves to be impossible, as the col
lapse of Hertz s theory showed.
More
closely considered, however,
373
in the special theory of relativity, the electrodynamic processes are
not used as a key to the mechanical, but a truly universal principle,
maxim of investigation in general, is established, which claims to contain a criterion of the validity and permissibility of all
a heuristic
particular physical fields
initial contradiction,
and
theories.
Thus
it
is
seen that the
appearing between the principles of mechanics and those of electrodynamics, has shown the way to a far more perfect and deeper unity between them than previously existed. And this result was not reached entirely by heaping up experiments by newly instituted investigations, but it rests on a critical trans formation of the system of fundamental physical concepts.
On the purely epistemological side, there thus appears with special distinctness in this intellectual process in which the theory of relativ ity originates, that peculiar "Copernican revolution." that variation
in the conceptual foundations of the theory of nature, which we have previously traced in the example of classical mechanics and the older
physics.
fact,
An
it
essential part of its
achievement seems based on the
has shifted the previous logical constants of physical knowledge, that it has set them at another place than before. For
classical
that
mechanics, the fixed and immovable point was the assump

in the various systems.
tion of the identity of the spatial
measurement
and temporal values gained by This identity was taken to
be the unquestionable and sure foundation of the concept of objectiv ity in general: as that which first really constituted the object of nature" as a geometrical and mechanical object and distinguished
it
from the changeable and

O.VTO
relative data of sensation.
TO
p.ev
<rxM
eon, TO de J\VKV KO.L oXcos TO aiadrjTov irpos dXXo /ecu runs the proposition, which Democritus brought into the foundations of atomism, and which in modern times was taken
icat?
kv aXXots
thus
up by
mary"
Galileo to support the fundamental distinction between "pri and "secondary" qualities^ and thus the whole "mechanical"
view of the world.
Although the principle here established proved
to be very fruitful and has been frequently confirmed in mathematical physics, the modern evolution of physics shows, with increasing evidence, that it was conce-ived too narrowly in a philosophical
and methodological sense. ./The true goal of science is not mecha nism but unity as Henri Poincare once formulated the guiding maxim of modern physics. But concerning this unity the physicist does not need to ask whether it is, but merely how it is; i.e., what is
374
the
minimum
to provide
an exact exposition
of presuppositions that are necessary and sufficient of the totality of experience and its
In order to maintain this (72 p. 172 ff.) which seemed endangered by the conflict of the principle of the constancy of the velocity of light and the principle of relativity of mechanics, and to ground it more deeply and securely, the theory of relativity renounces the unity of the values of spatial and temporal magnitudes in different systems. It surrenders the assumption that the temporal interval between events is a magnitude fixed once for all independently of the state of motion of the reference body and that in the same way the spatial distance between two points of a rigid body is independent of the state of motion of the reference body. By going back to the method of measuring time and to the fundamental role that the velocity of light plays in all our physical time measurements, it discovers the relativity of the simultaneity of two processes and further leads to the insight that the magnitude of the length of a body of its volume, its form, its energy and tempera ture, etc., are, as results from the formulae of the Lorentz-transformation, to be assumed as different according to the choice of the system of reference in which measurement takes place. But these "relativisystematic connection.
unity,
zations"
and unity
are not in contradiction with the doctrine of the constancy of nature they are rather demanded and worked out in the
;
very unity. The variation of the measurements of space and time constitutes the necessary condition through which the new invariants of the theory are discovered and grounded. Such invariants are found in the equal magnitude of the velocity of light for all systems and further in a series of other magnitudes, such as the entropy of a body, its electrical charge or the mechanical equiva lent of heat, which are unchanged by the Lorentz-transformation and which thus possess the same value in all justified systems of reference. But above all it is the general form of natural law which we have to recognize as the real invariant and thus as the real logical frame
of this
name
work of nature in general. While the special theory of relativity limits itself to regarding all reference bodies which are moving in a straight line relatively to a definite justified reference uniformly
system K, as equivalent for the formulation of natural laws, the

general theory extends this proposition to the assertion that all reference bodies whatever their state of motion may be, are
KK
to be taken as equivalent for the description of nature.
(17 p. 9;

18, p. 42.)
375
But the path by which alone
the concept of nature and of natural law,
this true universality of i.e., a definite and objec
tively valid description of phenomena independent of the choice of the system of reference, is to be reached, leads, as the theory shows,
necessarily through the "relativization" of the spatial and temporal magnitudes, that hold within the individual system; to take these as
changeable, as transformable, means to press through to the true invariance of the genuine universal constants of nature and universal
laws of nature.
light
The
and the postulate
postulate of the constancy of the velocity of of relativity show themselves thus as the two
as the fixed intellectual poles around which phenomena revolve; and in this it is seen that the previous logical constants of the theory of nature, i.e., the whole system of conceptual and numerical values, hitherto taken as absolutely deter minate and fixed, must be set in flux in order to satisfy the new and
fixed points of the theory,
more strict demand for unity made by physical thought. Thus reference to experience, regard for phenomena and
their
unified exposition, proves to be everywhere the fundamental feature, but at the same time it is seen that, in the words of Goethe, experi ence is always only half experience; for it is not the mere observa
tional material as such, but the ideal
form and the
intellectual inter
pretation, which it is given, that is theory of relativity and of its advantage over other types of explana As is known, the investigation of Michelson and Morley, tion. which gave the impetus and starting-point for the development of the theory of relativity, was explained as early as the year 1904 by
the basis of the real value of the
purely physical demands. each body moving with reference to The Lorentzian hypothesis, that the motionless ether with a velocity v undergoes a certain shortening in the dimension parallel to the motion, and indeed in the ratio of
Lorentz in a manner which
fulfilled all
vl
J,
was
sufficient to give
a complete explanation of
decision
possible;
all
known
observations.
An
experimental
Einstein s theories
was thus not
it
between Lorentz s and was seen that between
them there could fundamentally be no experimentum crucis* The advocates of the new doctrine accordingly had to appeal an unusual
spectacle in the history of physics
3
to general philosophical grounds,

p. 16
ff.
For more detail
cf. e. g.
Ehrenfest (15a),
376
EINSTEIN
to the advantages over the assumption of Lorentz which the new doctrine possessed in a systematic and epistemological respect. really experimental decision between the theory of Lorentz and
"A
the theory of relativity," Laue, e.g., explains in his exposition of tho principle of relativity in the year 1911, indeed not to be
"is
gained,
and that the former, in spite background, is chiefly due to the fact,
of this, has receded into the
that, close as
it
comes to the
theory of relativity, it still lacks the great simple universal principle, an the possession of which lends the theory of relativity
....
3 Lorentz s assumption appeared above all imposing appearance." to be epistemologically unsatisfactory because it ascribes to a physi cal object, the ether, definite effects, while at the same time it results from these effects that ether can never be an object of possible obser Minkowski too explains in his lecture on space and time vation. that the Lorentzian hypothesis sounds extremely fantastical; for
is not to be conceived as a physical consequence of the resistance of the ether but rather purely as gift from above," What as an accompaniment of the state of motion. (47 p. 60 f.)
the contraction
"a
thus, in the last analysis, decided against this assumption was not an empirical but a methodological defect. It conflicted most
sharply with a general principle, to which Leibniz has appealed in his struggle against the Newtonian concepts of absolute space and
time,
and which he formulated as the
"principle
of
observability"
When Clarke, as the representative of (principe de Vobservabilite.} referred to the possibility that the universe in its motion Newton,
relatively to absolute space might undergo retardation or accelera tion which would not be discoverable by our means of measurement,
Leibniz answered that nothing fundamentally outside the sphere of

3 cf. 41, p. 106. 40, p. 19 f Cf. also the characteristic remark of Lorentz himself in his Haarlem lecture: "The estimation (of the fundamental con cepts of Einstein s theory of relativity) belongs to a very large extent (grosstenteils) to the theory of knowledge, and one can leave the judgment to the
.
latter in confidence that it will consider the questions mentioned with the necessary thoroughness. But it is certain that it will depend for a great part on the type of thought to which one is accustomed, whether one feels drawn more to the one or the other conception. As far as concerns the lecturer himself, he finds a certain satisfaction in the older conceptions, that ether pos
sesses at least some substantiality, that space and time can be sharply sepa (46a, rated, that one can talk of simultaneity without further specification."
p. 23.)
377
observation possessed "being" in the physical sense: quand il n y a point de changement observable, il n y a point de changement du tout, It is precisely this principle of "observabil (5, p. 247 ff.).
which Einstein applied at an important and decisive place in his theory, at the transition from the special to the general theory of relativity, and which he has attempted to give a necessary connec
tion with the general principle of causality.
Any
physical explana
tion of a
phenomenon, he
urges,
only when there enter into it no law of causality is an assertion concerning the world of experience 2). (17, only when observable facts occur as causes and effects. Here we stand before one of the fundamental intellectual motives of
the theory of relativity a motive which not only gives it the advan tage over the empirically equivalent hypothesis of Lorentz, but which also produces the advance from the more limited interpretation of
epistemologically satisfactory non-observable elements; for the
is
the postulate of relativity in the special theory to the completely ûniversal formulation.
The way
to
in
which
this
advance has taken place
is
especially suited
and empirical presuppositions of the theory and their reciprocal connection. The special theory of relativity rests, as has been shown, on two different assumptions,
make
clear the conceptual
on the postulate of the uniformity of the propagation of light in a vacuum and on the pre supposition that all reference systems in rectilinear, uniform and
which stand equally
justified, side
by
side:
one stand in inseparable con considers these presuppositions, which nection in the empirical structure of the special theory of relativity, from a purely methodological standpoint, it is seen that in this stands respect they belong to different strata. On the one side, a general fact, a constant of nature, which results the assertion of from the experimental findings of optics and electrodynamics; on the other side stands a demand, which we make of the form of natural In the first case, it is empirically established that there is a laws. retains this peculiar velocity with a definite finite value, which value in any system independently of the state of motion of the latter.
If
non-rotary motion relatively to a definite justified system equally permissible for the formulation of the laws of nature.
are
In the second, a general

nature, which
is
established for the investigation of to serve as a "heuristic aid in the search for the
maxim
is
general laws of
nature."
In the formal limitation, which
is
placed
378

this
on natural laws by
the characteristic
ity.
maxim,
lies
as Einstein himself has urged
"penetration"
(Spurkrafi) of the principle of relativ

principles,
(18,
pp.
28,
67.)
But the two
the
"material"
are not distinguished from each other in the shaping of the special theory of relativity. The fact that this dis tinction is made and that the general and "formal" principle is
"formal"
and the
placed above the particular and "material" principle constitutes, from the purely epistemological standpoint, the essential step taken by the general theory of relativity. And this step seems to lead to a
strange and paradoxical consequence; for the particular result is not taken up into the general, but rather is cancelled by it. From the standpoint of the general theory of relativity, the law of the constancy of the velocity of light in a vacuum no longer possesses unlimited
validity. According to the general theory of relativity the velocity of light is dependent on the gravitation potential and must thus in
"to
general vary with places. The velocity of light must always depend on the coordinates when a field of gravitation is present; it is only be regarded as constant when we have in mind regions with a
-constant gravitation potential. This consequence of the general of relativity has often been regarded as a refutation of the theory
"presupposition
start
from which the special theory of and on which it based all its deductions.
relativity
took
its
But with
justice
Einstein rejects any such conclusion. The special theory of relativ ity, he explains, is not rendered valueless by the fact that one comes
to see that
to the
propositions refer to a definitely limited field, namely, phenomena in an approximately constant field of gravitation.
its
"Before
the establishment of electrodynamics, the laws of electro
were regarded as the laws of electricity in general. Today we know that electrostatics can only describe electrical fields cor rectly in the case, that is never exactly realized, in which the electric masses are exactly at rest relatively to each other and to the system of coordinates. Is electrostatics overthrown by Maxwell s electrostatics
Electrostatics is contained dynamical equations? Not in the least as a limiting case in electrodynamics; the laws of the latter lead directly to those of the first for the case that the fields are temporarily
!
unchangeable. The most beautiful fate of a physical theory is to point the way to the establishment of a more inclusive theory, in which it lives on as a limiting case." (18, p. 52.) In fact, in the
advance from the special to the general theory of
relativity,
we
\
)
J
379
have only a verification of the same principle of the construction of concepts of natural science that is found in the advance from classical mechanics to the special theory of relativity. The constants of measurement and of the theory of nature in general are shifted and magnitudes, which were regarded as absolute from the earlier stand
point, are again, with the gaining of a
new theoretical unit of measure
into merely relative determinations valid only under While classical mechanics, like the special definite conditions. theory of relativity, distinguishes between certain reference bodies
ment,
made
relatively to
which the laws of nature were valid and certain relatively which they were not, this distinction is now cancelled. The expression of the universal physical laws is freed from any connection
to
with a particular system of coordinates or with a certain group of such systems. To be expressed the laws of nature always require
some
definite system of reference; but their meaning and value is independent of the individuality of this system and remains selfidentical whatever change the latter may undergo. Only with this result do we reach the real center of the general
Now we know where lie its truly ultimate constants, its cardinal points, around which it causes phenomena to revolve. These constants are not to be sought in particular given
theory of relativity.
which are selected as chosen systems of reference from all others, such systems as the sun was to Copernicus and as the fixed No sort of things are truly in stars were for Galilei and Newton. variant, but always only certain fundamental relations and functional dependencies retained in the symbolic language of our mathematics
things,
This result of the general theory a paradox from the standpoint of_ is the criticism of knowledge, that it can rather be regarded as the natural logical conclusion of an intellectual tendency characteristic 4 of all the philosophical and scientific thought of the modern age.
in certain equations.
and physics,
of relativity, however,
so
little
To the popular view and

"things"
its
habits of thought the radical resolution of
for this
\
mere relations remains as ever suspicious and alienating, view believes that it would lose with the thing-concept the
into
one sure foundation of all objectivity, of all scientific truth. And of thus, from this side not so much the positive as the negative aspect
for its Here, indeed, I can only make this assertion in a general way; I must refer to the more specific explanation in my work Substance proof and Function. (8, pp. 148-310.)
4
380
the theory of relativity has been emphasized; what it destroys, not what it constructs has been comprehended. But it is remarkable
to find this interpretation not only in popular expositions of the theory of relativity but in investigations of its general "philo sophical" significance; and to meet in the latter also the view that it
brings an element of subjective arbitrariness into the formulation of the laws of nature and that, along with the unity of space and time, the unity of the concept of nature is destroyed. In truth, as closed consideration shows, the theory of relativity is characterized through
out by the opposite tendency. It teaches that to attain an objective and exact expression of natural process, we cannot take without further consideration the space and time values, gained by measure
system of reference as the only and universal in scientifically judging these measure ments, take account of the state of motion of the system from which trie measurement is made. Only w hen this is done can we compare measurements which have been made from different systems. Only those relations and particular magnitudes can be called truly objec tive which endure this critical testing, that is, which maintain them That not only selves not only for one system but for all systems. are there such relations and values, but that there must be such, in so far as a science of nature is to be possible, is precisely the doctrine the
ment within a
definite
values, but that
we must,
theory of relativity sets up as a postulate. If we start, as practically we must do at first, from a definite system of measurement, we must bear in mind that the empirical values, which we gain here, do not
signify the final natural values but that, to undergo an intellectual correction. What
become
such, they
must
nature only arises when we first made from the standpoint of a particular reference body, with those made from other reference bodies, and in principle with those
the system of combine the measurements, which are

call
we
made from
all "possible" reference bodies, and bring them ideally into a single result. there can be found in this assertion any limitation of the "objectivity" of physical knowledge is not evident;
How
obviously
it is
meant to be nothing but a

it is
definition of this very
we have only to do clear," says Kant, with the manifold of our presentations and that X, which corresponds to them (the object), since it is to be something distinct from all our
objectivity.
"But
"that
is for us nothing; the unity, which makes the object can be nothing else than the formal unity of consciousness necessary,
presentations,

in the synthesis of the
381
we know the
manifold of presentations. Thus we say: object \vhen in the manifold of intuition we have pro
duced synthetic unity." The object is thus not gained and known by our going from empirical determinations to what is no longer empirical to the absolute and transcendent, but by our unifying the totality
and measurements given in experience into a single complete whole. The theory of relativity shows the whole complexity of this task; but it retains the postulate of the possibility of such a system all the more strenuously and points out a new way to realize
of observations
It Classical mechanics believed itself at the goal too soon. clung to certain reference bodies and believed that it possessed, in connection with them, measures in some way definitive and universal, and thus absolutely "objective." For the new theory, on the con
it.
trary, true objectivity only in the manner and
never
lies in
empirical determinations, but
The space and

relative;
way, in the function, of determination itself. time measurements in each particular system are
but the truth and universality, which can be gained never
by physical knowledge, consist in the fact that all these measurements correspond mutually and are coordinated with
theless
each other according to definite rules. More than this indeed knowledge cannot achieve, but it cannot ask for more, if it under stands itself. To wish to know the laws of natural processes inde
pendently of
all relation
to
any system
of reference,
is
an impossible
is
and self-contradictory
system
desire; all
that can be
demanded
that the
content of these laws not be dependent on the individuality of the It is precisely this independence of the acciden of reference.
tal standpoint of the observer that we mean when we speak of the in them "natural" object and the "laws of nature" as determinate
selves.
Measurements
of
"justified"
in
one
number
ties,
systems would
system, or even in an unlimited in the end give only particulari

of the object.
but not the true
"synthetic unity"
The theory
may go a totality of invariant determinations. The of the world, anthropomorphism of the natural sensuous picture 6 the overcoming of which is the real task of physical knowledge,
general theory,
definite
of relativity teaches, first in the equations of the Lorentz-transformation and then in the more far-reaching substitution formulae of the from each of these particularities to a how we
whole,
to
Cf.
Planck
(66) p. 6
ff.
and
(67) p. 74.
382
is
here again forced a step further back. The mechanical view of the world thought to have conquered it, when it resolved all being and natural process into motion and thus put everywhere pure magnitudes
in place of qualitative
elements of sensation.
But now
it is
seen that
precisely the determination of these values, the measurements, which it applies to motions, are still bound to certain limiting presupposi
Reflection on the manner in which we make empirical meas urements of space and time shows how anthropomorphism reaches into this field that was thought withdrawn from it in principle. It is, as it were, this earthly remainder still belonging to classi cal mechanics with its assumption of finite fixed reference bodies and motionless inertial systems, from which the theory of relativity seeks to free itself. The conceived unit of connection determined by a system of mathematical equations here takes the place of any sensuously given, and also sensuously conditioned, unit of measure ment. As is seen, there is involved here not a cancellation but a critical correction of the empirical concept of objectivity, by which a correction of our empirical spatial and temporal measures and their
tions.
transformation into the one system of natural laws are gained. We are brought to the same outcome by consideration of the
historical
problems out of which the theory of relativity has grown.
give the propositions of abstract mechanics, especially the prin ciple of inertia a definite physical meaning had been attempted repeatedly by trying to point out some empirical systems for which
strict validity. But these attempts were all in particular, by the discovery of the motion of the solar thwarted, system and of the fixed stars; to find a fixed and clear empirical
To
they would possess
for the equations of the Galileo-Newtonian mechanics, remained save to postulate, with Carl Neumann, an abso .nothing
meaning
But lutely motionless body at some unknown place in space. such a postulate of the existence of a particular physical object, a body which can never be discovered by observation, remains the
strangest
238
ff.)
The
anomaly, from the epistemological standpoint. (8, p. absolutely motionless ether too, which seemed for a
offer the lacking physical reference system of the GalileoNewtonian mechanics, showed itself unsuited to this purpose; since the negative outcome of Michelson s investigation the question seemed to be decided here also. At this point, as has been seen,
time to
the theory of relativity begins.
It
makes a
virtue out of the difficulty
383
into which philosophical thought had fallen in its attempt to find a particular privileged system of coordinates. Experience had shown
that there
is
no such system, and the theory, in
its
most general
interpretation, makes it a postulate that there cannot and must not be such. That, for the physical description of the processes of nature, no particular reference body is to be privileged above any
other
is
now made a
principle.
"In
classical mechanics, as well as
the special theory of relativity," drawn between reference bodies
says
relatively to
nature are valid and reference bodies
a distinction is which the laws of relatively to which they are

Einstein,"
no consistently thinking man possible that certain reference bodies (and their states of motion) are privileged over other reference bodies (and their states of motion)? In vain, I seek in classical
not valid.
With
this state of affairs
can be
satisfied.
He
asks:
how
is it
in the behavior of the

."
mechanics for something real to which I might trace the difference body with reference to the systems of refer ence K and K In this argument from the principle of (18, p. 49.) the physicist seems to move on slippery ground. insufficient reason, One is inevitably reminded of the argument of Euler, who thought that he proved the principle of inertia of classical mechanics by explaining that, if a body changed its state of motion without the influence of external forces, there would be no reason why it should choose any particular change of magnitude and direction of its
state The circle involved here, namely, that is assumed to be a determinate magnitude, body
"the
velocity. (23.) of motion" of a
only defined as such by the law of inertia itself, is easily In Einstein s appeal to the "principle of reason," there is doubtless involved a more general and deeper epistemological motive.
while
it is
seen.
If we assume that the final objective determinations, which our physical knowledge can reach, i.e., the laws of nature, are provable and valid only for certain chosen systems of reference, but not for
others, then, since experience offers no certain criterion that we have before us such a privileged reference system, we can never reach a truly universal and determinate description of natural processes.
This is only possible if some determinations can be pointed out, which are indifferent to every change in the system of reference taken as a basis. Only those relations can we call laws of nature, i.e., ascribe
to
them
objective universality, whose form
is
independent of the
particularity of our empirical
measurements
of the special choice
384
of the four variables
Xi, x 2 x 3 x 4 which express the space and time In this sense, one could conceive the principle of the parameters. universal theory of relativity, that the universal laws of nature are not changed in form by arbitrary changes of the space-time variables,
, ,
as an analytic assertion; as an explanation of what is meant by a law of nature. But the demand,, that there must in "universal" be such ultimate invariants, is synthetic. general
In fact, it can be shown that the general doctrine of the invariabil ity and determinateness of certain values, which is given first place by the theory of relativity, must recur in some form in any theory of
nature, because it belongs to the logical and epistemological nature of such a theory. To start from the picture of the world of general energetics Leibniz, in establishing the law of the "conservation of vis viva" as a universal law of nature, referred to this logical element in it. He first defines the vis viva of a physical system as a quantity of work; he determines that forces are to be called equal, when they
are able to perform equal mechanical work, no matter what their properties may be in detail; thus if they produce an equal degree of tension in an equal number of elastic springs, raise an equal weight
to the same height, communicate to an equal number of bodies the same amount of velocity, etc., they are equal. In this definition it is assumed that measurement of the vis viva by different systems of measurement will give results equivalent to each other, and thus that force which, when measured by a certain effect, prove to
be equal or in a definite relation of greater or smaller,

this
will
retain
If this
same
relation
if
we measure them by any other
effect.
were not the case, and did there result a different relation of forces
according to the different effect one uses as a measure, nature would be without laws; the whole science of dynamics would be superfluous;
and it would not be possible to measure forces, for forces would have become something indeterminate and contradictory, quiddam vagum etabsonum. (42, III, 208 ff.; VI, 209 f.; cf. 5, p. 305 ff.) The same process of thought has been repeated on broader physical lines in the discovery and grounding of the modern principle of energy. Here, too, the energy of a material system in a certain state was defined e.g., by W. Thompson first as the amount of all the effects,
expressed in mechanical units of work, called forth outside the system when the system passes in any way from its state into a definite but
arbitrarily defined state of nullity.
This explanation at
first
leaves

it
385
what
entirely undecided as to whether there exists a determinate value of is here called "energy," i.e., whether the results of the measure
of the
ment
amount
of
work
of a
system turn out the same or
differ
ently according to the method of bringing the system from the given state into a definite state of nullity. But that this determinateness in fact exists, that there always results the same amount of energy no matter what effect we use as the measure of work and
what type of transition we choose, is precisely what the principle of the conservation of energy affirms. This affirms nothing else
and has no other physically comprehensible meaning than that the amount of all the effects, measured in units of mechanical work, which a material system calls forth in its external environment, when it passes from a definite state in any arbitrary manner to an arbitrarily defined state of nullity, has a determinate value, and is
thus independent of the type of transition. If this independence did not exist and that it exists only experience can teach us it would follow that what we called "energy" is not an exact physical
determination energy would not be a universal constant of measure ment. We would have to seek for other empirical values to satisfy the fundamental postulate of determinateness. But it holds, con
;
versely, that
ment,
it
if energy is once established as a constant of measure thus becomes a constant of nature also, a "concept of a
Now from a physical standpoint a "substantial" object." conception of energy can be carried through without arousing suspi cion; energy can be regarded as a sort of "reserve supply" of the 6 physical system, the quantity of which is completely determined by the totality of the magnitudes of the states, which belong to the
definite
system involved.
From
the epistemological standpoint,

is
it
must be
remembered that such an interpretation
nothing more than a con
venient expression of the relations of measurement, that alone are
known, an expression which adds to them nothing essential. The unity and determinateness of measurement can be immediately understood and expressed as the unity and determinateness of the
but object, precisely because the empirical object means nothing It follows from this analogy .totality of relations according to law.
a
from a new angle that the advance

I
in
"relativization"
which takes
place in the
6
theory of relativity, represents no contrast to the general

ff.
In more detail in Planck (63) p. 92
386
task of objectification, but rather signifies one step in it, since, by the nature of physical thought, all its knowledge of objects can consist
knowledge of objective relations. "Whatever we here, too, we can cite the Critique of Pure Reason, nothing but relations, some of which are independent and (34, p. 341; permanent and by which a certain object is given The general theory of relativity has cf. Muller s Trans, p. 232.) shifted these "independent and permanent relations" to another place by breaking up both the concept of matter of classical mechanics and the concept of the ether of electrodynamics; but it has not con tested them as such, but has rather most explicitly affirmed them in its own invariants, which are independent of every change in the
in nothing save
may know
of
matter,"
"is
us."
system of reference.
The
criticism
made by
the theory of relativity
of the physical concepts of objects springs thus from the same method of scientific thought, which led to the establishment of these con
cepts, and only carries this method a step further by freeing more from the presuppositions of the naively sensuous and
stantialistic"
it still
"sub-
view of the world. To grasp this state of affairs in its full import we must go back to the general epistemological questions offered to us by the theory of relativity; we must go back to the transformation of the physical concept of truth involved in it by which it comes into direct contact with the fundamental problem of logic.
CHAPTER
III
THE PHILOSOPHICAL CONCEPT OF TRUTH AND THE THEORY

OF RELATIVITY
principle of the relativity of knowledge received its complete systematic working out in the history of ancient skepticism. Here it possessed, according to the fundamental charac
first
The general
ter of skepticism,
an exclusively negative meaning;
it
signified the
limit in principle which is set to all knowledge and by which it is separated once for all from the definitive apprehension of the truth
as
"absolute."
Among
the skeptical
"tropes"
intended to show the
uncertainty of sensuous and conceptual knowledge, the "trope" of Trpos n stands in the first place. To know the object, our knowl edge would, above all, have to be in a position to grasp it in its pure
from all the determinations, which only and other things. But this separation is impossible, not only actually, but in principle. For what is actually given to us only under certain definite conditions can never be made out logically as what it is in itself and under abstraction from In what we call the perception of a precisely these conditions. thing, we can never separate what belongs to the objective thing froia what belongs to the subjective perception and contrast* th^ Two
"in
itself"
and to separate
it
belong to
it
relatively to us
independent factors. TheJfcMrm qf subjective organization enters as a necessary element into all our so-called objective knowledge of things and properties. The "thing" appears, accordingly, not only differ
ently to the various senses but it is limitlessly variable for the same organ according to the time and varying conditions of perception.
whole character depends on the relations under which it is presented to us. No content is given us in experience unmixed with others in a purely self-identical character, but what is given us is always only a general combination of impressions. It is not one or
its
For
the other,
"this"
or
"that"
definite quality,
but only the reciprocal

is
relation of the one to the other
and the other to the ongjhat
here
nown, indeed that is alone knowable. Modern science has overcome the objections of skepticism to the possibility of knowledge, not by contesting their content, but by
drawing from them a wholly
different,
indeed,
opposite logical
387
388
consequence. Modern science also assumes the reduction of what is taken in the naive view of the world, as fixed and absolute "proper ties" of "With regard to the things to a system of mere relations. of the objects of the outer world," we read in, e.g., Helmproperties
holtz s
Handbuch
the properties
we can
der physiologischen Optik, is easy to see that all ascribe to them, signify only the effects they
"it
produce either on our senses or on other natural objects. Color, sound, taste, smell, temperature, smoothness, solidity belong to the first class; they signify effects on our sense organs. The chemical are likewise related to reactions, i.e., effects, which the properties
natural body in question exerts on others. It is thus with the other physical properties of bodies, the optical, the electrical, the magnetic.
Everywhere we are concerned with the mutual relations of bodies to each other, with effects which depend on the forces different bodies exert on each other From this it follows that in fact,
the properties of the objects of nature do not signify, in spite of their name, anything proper to the particular objects in and for themselves,
The type
effect
is
but always a relation to a second object (including our sense organs). of effect must naturally always depend on the peculiarities
body on which the whether cinnabar is question really red as we see it, or whether this is only an illusion of the senses, is therefore meaningless. The sensation of red is the normal reac tion of normally constituted eyes to the light reflected from cinnabar. A color-blind person will see the cinnabar black or dark grey; this
of the effecting
body
as well as
on those
of the
exerted
To
also
. .
is
.
the
.
correct
itself,
reaction
of
his
peculiarly
is
"
constituted
eye.
In
the one sensation
not more correct and not

(30,
more
false
than the other
p.
588
f.)
The
old
skeptical
"trope,"
us again in all things involves no longer renunciation of the objectivity of knowledge. For the truly objective element in modern knowledge of nature is not
so
the argument of the irpos n here stands before distinctness. But renunciation of the absoluteness of
much things as laws. Change in the elements of experience and the fact that no one of them is given in itself, but is always given with reference to something else, constitute no objection to the possi
bility of objectively real knowledge in so far as the laws establish The constancy and absolute precisely these relations themselves. ness of the elements is sacrificed to gain the permanency and neces
sity of laws.
If
we have gained the
latter,
we no longer need the
THE PHILOSOPHICAL CONCEPT OF TRUTH

former.
389
For the objection of skepticism, that we can never know the absolute properties of things, is met by science in that it defines the concept of property in such a way that the latter involves in itself
is overcome by being outdone. can mean absolutely nothing save a rela tion to a seeing eye, that "heavy" means nothing save a relation of
the concept of relation. When it is seen that

reciprocal acceleration
Doubt
"blue"
and that
in general all
"having"
of properties
can be resolved purely and simply into ,a-~~being-related" of the elements of experience, then the longing to grasp the ultimate absolute
qualities of things, secretly at the basis of skepticism, loses its
is
mean
ing. Skepticism refuted, not by showing a way to a possible fulfillment of its demands, but by understanding and thus rendering
îneffective the dogmatic import of these demands themselves. In this transformation of the general ideal of knowledge, modern
science
one
is
and modern logic are both involved; the development of the in closest connection with that of the other. Ancient logic is
on the relation
of
"subject"
entirely founded
and
"predicate,"
on
the relation of the given concept to its also given and final properties. It seeks finally to grasp the absolute and essential properties of
.
Modern logic, on the ccîirary; comes more anymore to/ abandon this ideal and to be made into a pure doctrine of form and relation., The possibility of all determinate character of the content of knowl edge is grounded, for it, in the laws of these forms, which are not reducible to mere relations of subsumption but include equally all the different possible types of relational construction and connection of elements of thought. But here doubt must begin in a new and
absolute self-existent substances.
in the course of its development,
deeper sense. If knowledge of things is understood as knowledge of laws and if an attempt is made to ground the former in the latter and to protect it from the attacks of skepticism, then what guarantees the objectivity, the truth and universality of the knowledge of laws? Do we have, in the strict sense, knowledge of laws or does all that
we can gain
resolve itself in the
most favorable case
into knowledge
of particular cases?
Here as we see, the problem of skepticism is reversed on the basis of the modern conception of law. What; perplexed the ancient skeptic, who sought the substance of things,;
was the
limitless relativity of all
phenomena; it was the fact that would not remain fixed individual data, but were reduced phenomena for knowledge ever again into mere relations and relations of relations. -
390
EINSTEIN
for the
modern skeptic, to whom the objective truth, in so far attainable, means the one all-inclusive and necessary law of all process, the basis of doubt lies in the fact that reality is never
But
as
it is
given us in this universal intellectual form, but
is
always divided
and broken up into mere punctual particularities. We grasp only a here and a now, only a particularity isolated in space and time, and it is not to be seen how we could ever pass from this perception of the individual to a view of the objective form of the whole. No more than the continuum can be built up and generated by the summation of mere unextended points can a truly objective and necessary law be gained and deduced by the simple aggregation of however many This is the form of Hume s skepticism, which particular cases. While the ancient characteristically distinguished from the ancient.
skeptic could not reach the absolute substance because of the relativi
ties in
fails
is"
which the phenomenal world involved him, the modern skeptic to reach laws as universal relations because of the absolute
While in the former it is the certainty particularities of sensation. of things that is questionable, in the latter it is the certainty of causal connections. The connections of processes become an illusion; what
remains is only their particular atoms, the immediate data of sensa of "matter of fact" ultimately tion, in which all knowledge of
"facts,"
consists.
If it is
possible to
overcome
this essentially
more
radical
form of
skepticism also, it can only be by there being shown in it too a concealed dogmatic assumption, which lies implicitly at its basis.
assumption consists in fact in its concept of empirical itself. This givenness of impressions in which abstraction is made in principle from all elements of form and con When this is nection, proves to sharper analysis to be a fiction. understood, doubt is directed, not on the possibility of knowledge, but on the possibility of the logical measuring-rod with which knowledge is measured here. Instead of the criterion of the "impression" making the universal formal relations of knowledge and its axioms questionable, the validity of this criterion must be contested on the basis of these relations. The only refuge from radical doubt lies in its being not set aside but intensified, in our learning to question, as ultimate elements of knowledge known in themselves, not only "things" and but especially sensations. The skepticism of Hume left the "simple" sensation as a completely
this
"givenness"
"bare" "laws"
And
unproblematic certainty, as a simple and unquestionable expression

of
"reality."
391
While antique skepticism rested completely on the assumption of absolute things, ."that of Hume rests on the assumption of absolute sensations. The hypostasization in the one case concerns being, in the other, being, but its general form is the- same. And only by this hypostasization does
tacit
"outer"
"inner"
the doctrine of the relativity of knowledge gain its skeptical character. Doubt does not result directly from the content of this doctrine, but,
on the contrary, it depends on the fact that the doctrine is not truly and consistently thought through. As long as thought contents itself with developing, with reference to phenomena and according to demands of its own form, its logical axioms, and truth as a system of pure relations, it moves within its own circle with complete But when it affirms an absolute, whether of outer or certainty.
inner experience, it is forced skeptically to annihilate itself with reference to this absolute. It strikes this absolute of things or of
sensations again and again as if against the wall of the cell in which it is enclosed. Relativity, which is, fundamentally, its immanent
becomes its immanent limit. It is no longer the principle, which renders possible and governs the positive advance of knowl edge, but is merely a necessary instrument of thought, which by that fact confesses itself not adequate to being the absolute object and
force,
the absolute truth.

is indeed changed when we contrast to both the and the skeptical concept of truth, which are united by a dogmatic common root, the idealistic concept of truth. For the latter does not measure the truth of fundamental cognitions by transcendent
This relation
it grounds conversely the meaning of the concept of the on the meaning of the concept of truth. Only the idealistic object concept of truth overcomes finally the conception which makes knowledge a copying, whether of absolute things or of immediately of knowledge changes from a given "impressions." The mere pictorial to a pure functional expression. In the history of modern philosophy and logic, this change is first represented in
objects, but
"truth"
complete clarity by Leibniz, although in his case, the new thought appears in the setting of a metaphysical system, in the language of the monadological scheme of the world. Each monad is, with all its
contents, a completely enclosed world, which copies or mirrors no outer being but merely includes and governs by its own law the whole of its presentations; but these different individual worlds
express,
nevertheless,
common
universe and a
common
truth.
392
EINSTEIN
This community, however, does not come about by these different pictures of the world being related to each other as copies of a
common
"original"
but by the fact that they correspond functionally
and in the general form of their For one fact, according to Leibniz, expresses another when there exists between what cari be said of the one and of the other a constant and regular relation^ Thus a perspective projection expresses its appropriate geometrical figure, an algebraic equation expresses a definite figure, a drawn model a machine; not as if there existed between them any sort of factual likeness or similarity, but in the sense that the relations of the one structure correspond to those of the other in a definite conceptual fashion. (43, VII, 263 f, This Leibnizian concept of truth was 44, II, 233; cf. 7, II, 167.) taken up and developed by Kant who sought to free it from all the unproved metaphysical assumptions that were contained in it. In this way he gained his ow n interpretation of the critical concept of the object, in which the relativity of knowledge was affirmed in a far more inclusive meaning than in ancient or modern skepticism, but in which also this relativity was given a new positive interpre tation. The theory of relativity of modern physics can be brought without difficulty under this interpretation, for, in a general epistemoto each other in their inner relations
structure.
r
it is characterized by the fact that in it, more clearly and more consciously than ever before, the advance is made from the copy theory of knowledge to the functional theory. As long as physics retained the postulate of absolute space, the question still had a definite meaning as to which of the various paths of a moving body that result when we regard it from different systems of reference, represents the real and motion; thus a higher objective truth had to be claimed for certain spatial and temporal values, obtained from the standpoint of certain selected systems, than
logical regard,
"true"
for others.
The theory
it
of relativity ceases to
make
this exception;
not that
this
it
would abandon the determinateness of natural process,

has at
its
but because
disposal
new
intellectual
means
of satisfying
is
demand.
The
infinite multiplicity of possible
systems
not
identical with the infinite indeterrninateness of the values to be
gained in them in so far as all these systems are to be related and connected with each other by a common rule. In this respect, the
with
principle of relativity of phj^sics has scarcely more in common "relativistic positivism," to which it has been compared, than
the name.
When
there
is
seen in the former a renewal of ancient
393
sophistical doctrines, a confirmation of the Protagorean doctrine that man is the "measure of all things," its essential achievement is
mistaken. 1
The
appears to each person
physical theory of relativity teaches not that what is true to him, but, on the contrary, it warns
against taking appearances, which hold only from a particular system, as the truth in the sense of science, i.e., as an expression of an inclu
and final law of experience. The latter is gained neither by the observations and measurements of a particular system nor by those
sive
of however many such systems, but only by the reciprocal coordina tion of the results of all possible systems. The general theory of to show how we can gain assertions concerning relativity purports
how we can rise above the fragmcntariness of the indi vidual views to a total view of natural processes. It (Cf. above.) abandons the attempt to characterize the "object" of physics by any
all of these,
sort of pictorial properties, such as can be revealed in presentation, and characterizes it exclusively by the unity of the laws of nature. When, for example, it teaches that a body regarded from one system
possesses spherical form and, regarded from another system, in motion relatively to the first, appears as an ellipsoid of rotation, the
question can no longer be raised as to which of the two optical images here given is like the absolute form of the object, but it can and must be demanded that the multiplicity and diversity of the sensuous
ence.
data here appearing can be united into a universal concept of experi Nothing more is demanded by the critical concept of truth and the object. According to the critical view, the object is no absolute
model to which our sensuous presentations more or less correspond as copies, but it is a "concept, with reference to which presentations have synthetic unity." This concept the theory of relativity no longer represents in the form of a picture but as a physical theory, in the form of equations and systems of equations, which are covariant with reference to arbitrary substitutions. The "relativizawhich is thus accomplished, is itself of a purely logical and mathematical sort. By it the object of physics is indeed determined
tion,"
w orld "object in the phenomenal world;" but this phenomenal no longer possesses any subjective arbitrariness and contingency. For the ideality of the forms and conditions of knowledge, on which physics rests as a science, both assures and grounds the empirical and in the name of realhVy of all that is established by it as a
as the
r
"fact"
objective validity.
Cf. Petzoldt (61).
CHAPTER
IV
MATTER, ETHER AND SPACE

In the structure of physics we must, it seems, distinguish two differ ent classes of concepts from each other. One group of concepts concerns only the form of order as such, the other the content that
enters into this form the first determines the fundamental
;
schema whi ch
physics uses, the other concerns the particular properties of the real by which the physical object is characterized. With regard to the pure formal concepts, they appear to persist as relatively fixed unities
in spite of all
changes of physical ideals in
detail.
In
all
and
conflict of the systematic concepts of physics, space
the diversity and time
are distinguished as the ultimate, agreeing unities. They seem, in this sense, also, to constitute the real a priori for any physics and
the presupposition of its possibility as a science. But the first step from these bare possibilities to reality, which is a matter not of the spatio-temporal form, but of the somewhat that is thought to be
somehow
circle of
"given"
the a priori,
in space and in time, seems to force us beyond the Kant indeed, in the Metaphysischen Anfangs-
grunden der Naturwissenschaft, attempted an a priori deduction and construction of the concept of "matter" as a necessary concept of physics; but it is easy to see that this deduction does not stand on the same plane and cannot claim the same force as the Transcendental Aesthetic or the Analytic of the Pure Understanding. He himself believed that he possessed in these deductions a philosophical
grounding of the presuppositions of the science of Newton; today we recognize to an increasing extent that what he so regarded was in fact nothing but a philosophical circumlocution for precisely these
presuppositions. As a fundamental definition of the physical _concepj^pf the object, the classical system of mechanics is only one Heinrich Hertz, structure, by the side of which there are others.
in his
new grounding
three such structures: the
of the mechanical principles, distinguished first is given in the Newtonian system,

<>!
is founded on the concepts of space, of time, of force, and as given presentations; the second leaves the presuppositions mass, of space, time and mass unchanged, but substitutes for the concept of
which
force as the mechanical
"cause
of
acceleration"
the universal concept
394

of energy, which
is
395
energy.
concepts,
Here, too,
divided into two different forms, potential and we have four mutually independent
relations to each other are to constitute the content Hertz s own formulation of mechanics offers a third structure in which the concept of force or of energy as an independent idea is set aside and the construction of mechanics is accomplished by only three mdependent^ fundamental Jdeasx_srjacej_time and mass. The circle of possibilities would thus have seemed completely sur veyed had not the theory of relativity once more given a new inter pretation to the mutual relation between the pure formal concepts
of mechanics.
whose
and the physical concepts of the object and substance, and thus trans formed the problem not only in content but in principle. The concept of "nature," the gaining of which is the real methodic
problem of physics, leaves room, as the history of physical thought shows, for a dualism of presuppositions, which as such seems neces sary and unavoidable. Even in the first logical beginnings of genuine natural science, which are found in Greek thought, this dualism appears in full distinctness and clarity. Antique atomism, which is the first classical example of a conceptual and scientific picture of the world, can only describe and unify the of nature by it up out of two heterogeneous elements. Its view of building nature is founded on the opposition of the and the the lull and the void, prove necessary elements for the The_ constitution of the object of physics. To the being of the atom and matter as the Tra^TrAjjpes ov, there is opposed by Democritus the not-being, the jui) ov of empty space; both this being and this notbeing possessed for him, however, uncontested physical truth and thus indubitable physical reality. The reality of motion was only intelligible by virtue of this dual presupposition; motion would disappear if we did not both distinguish empty space from the mate rial filling of space and conceive the two as in inseparable mutual At the relation, as fundamental elements in all natural processes.
"being"
"full"
"void."
_tw.<v
beginning of modern times, Descartes attempted philosophically to overcome this duality in the foundations of physical thought. Pro ceeding from the thought of the unity of consciousness, he postulated also a new unity of nature. And this seemed to him only attainable of and the by abandoning the opposition of the of the body and the matter and extension. The physical being geometrical being of extension constitute one and the same object:
"full"
"void,"
396
the
"substance"

of a
body is reduced to its spatial and geometrical Thus a new approach to physics, methodologically deeper and more fruitful, was found, the concrete realization of which, however, could not be accomplished by Descartes physics. When Newton fought the hypothetical and speculative premises of the Cartesian physics, he also abandoned this approach. His picture of the world was rooted in the dualistic view, which was even intensified in it and which set its seal on his universal law of nature and the cosmos. On the one side, there stands space as a universal receptacle and vessel; on the other, bodies, inert and heavy masses, which enter into it and determine their reciprocal position in it on the basis of a universal dynamic law. The "quantity of matter,"
determinations.
on the one hand, the purely spatial "distance" of the particular masses from each other, on the other, give the universal physical law of action, according to which the cosmos is constructed. _ Newton for a as a physicist always declined to ask for a further reason for this rule. It was for him the unitary mathematical formula, which included all empirical process under it and thus That perfectly satisfied the task of the exact knowledge of nature. this formula concealed in the expression for the cosmic masses and in the expression for their distance two wholly different ele ments seemed a circumstance that no longer concerned the physicist but only the metaphysician and the speculative philosopher of nature. The proposition "hypotheses non Jingo" cuts off any further investiga tion in this direction. For Newton as for Democritus, matter and space, the full and the void, form for us the ultimate but mutually irreducible elements of the physical world, the fundamental buildingstones of all reality, because as equally justified and equally neces sary factors, they enter into the highest law of motion taught u* In
"why,,"
experience. If we contrast this view with the picture of the world of modern and very recent physics, there results the surprising fact that the latter seems to be again on the road to Descartes, not indeed in
It too strives from various sides content, but certainly in method. toward a view in which the dualism of and "matter" is in which the_two no longer occur as~ditterent classes of cancelled, physical object-concepts. There now appears in the concept of the a new mediating concept between "matter" and "empty and this it is which henceforth appears with increasing space;"
"space"
"field"
..
397
is
definiteness as the genuine expression of the physically real since it the perfect expression of the physical law of action. In this
concept of the field, the typical manner of thought of modern physics has gained, from the epistemological standpoint, its sharpest and
most distinct expression. There now takfis^ .pkce^starting from electrodynamicaâ progressive transformation of the concept of matter. Already with Faraday, who constructed matter out of
"lines of force," there is expressed the view that the field of force cannot depend on matter, but that on the contrary, what we caJL matter is nothing else than specially distinguished places of .this aljgli^ In the progress of electrodynamics, this view is confirmed
and assumes ever more radical expression. The doctrine is carried through more and more of a pure "field-physics," which recognizes neither bare undifferentiated space by itself nor matter by itself
subsequently entering into this finished space, but which takes as a basis the intuition of a spatial manifold determined by a certain law
and qualified and differentiated according to it. Thus, e.g., there was established by Mie a more general form of electrodynamics on the basis of which it seemed possible to construct matter out of
field.
magnetic
the
as
as
its
JTb^j^c^Q.ej)X.o|,^.,syb8.tance existing along with the electro field seemed unnecessary in this approach; according to
new conception,
the
field
is
an
bearer, but matter "outgrowth of the
no longer requires for its existence matter considered and treated, on the contrary,
.It~is
field."
the 4ast-cunseqnHe
_Q
this
type^joLthought that is drawn by the theory of relativit the real difference finally disappears between an "empty" it, too, space andaspace-filling substance, whether one calls this matter or
ether, since it includes
both moments in one and the same act of
methodic .determination..
The
"riddle
of
weight"
is
revealed to us,
according to. the fundamental thought of Einstein s theory of gravita tion, in the consideration and analysis of the inner relations of
measurement
ten functions
of the four dimensional spftra-flpift manifold. For the of the linear g/j.v, which occur in the determination
4
elements of the general theory of relativity ds 2
^Kuj^XfAxv
(/z,
v =
1,2, 3, 4), represent also the ten
components
of the gravitation
potcntiaLpf Einstein s theory. It is thus the same determinations, which, on the onehancl, designate and express the metrical properties
1
On Faraday,
cf.
Buek
(4,
esp. p. 41
ff.); cf.
also
Weyl
(83, p. 142).
398
of the four-dimensional space and, on the other, the physical proper ties of the field of gravitation., The spatio-temporal variability of
the magnitudes
gfj.v
and the occurrence
of such a field prove to be
equivalent assumptions differing only in expression. Thus it is shown most distinctly that the new physical view proceeds neither
I
$
from the assumption of a space in itself," nor of "matter" nor of "force in itself" that it no longer recognizes space, force and matter as physical objects separated from each other, but that for it exists only the unity of certain functional relations, which are differently designated according to the system of reference in which we express them. All dynamics tends more and more to be resolved into pure metrics, a process in which indeed the concept of metrics undergoes, in contrast with classical geometry, an extraordinary broadening and generalization whereby the measurements of Euclidean geometry
appear as only a special case within the total system of possible measurements in general. "The world," as is said by Weyl, in whose account of the general theory of relativity one can trace and
survey this development most
metrical manifold; metrics
all
clearly,"
is
(3
1)
dimensional
expressions of world The dream of^Descartes of a purely geometrical physics seems to be about to be fulfilled in a wonderful way, which could not have been foreseen by him." (83, p. 244; cf. p. 85 ff.,
physical
phenomena are
J70ff).
Just as the dualism of matter and space is superseded here by a unitary physical conception, so the opposition between "matter"
and
"force"
is
to be overcome
by the
principle
and law
of the
new
Since Newton, as a physicist, established this opposition physics. between the "inert masses" and the forces that affect them in the
have not been lacking to overcome

speculative side.
Philosophiae Naturalis Principia Mathematica, attempts, indeed, it from the philosophical and
Leibniz led the way here; but although, in his he wholly resolved substance into force he retained in metaphysics, the construction of his mechanics, the duality of an "active" and a of "passive" force, whereby matter is subsumed under the concept the latter. The essence of matter consists in the dynamic principle immanent in it; but this expresses itself, on the one hand, in activ ity and striving for change, on the other hand, in the resistance which a body opposes, according to its nature, to change coming upon it from

without. 2
399
As for Newton, the opposition in fundamental concepts, which he assumes, threatens finally to destroy the unity of the physi cal structure of his world; he can only retain this unity by introducing at a certain place a metaphysical factor. The principle of the conser
disputed by him because all bodies consist of atoms, and in the rebounding of such atoms, mechanical energy must be lost; the sum total of force is in a con tinuous decrease, so that for its preservation the world needs from time to time a new divine impulse. - (58, p. 322 ff .) Kant attempted in a youthful work, the Monadologia Physica of the year 1756, a reconciliation and mediation between the principles of the Leibnizian philosophy and those of Newtonian mechanics; and in the
vation of
vis viva is
hard"
"absolutely
Metaphysischen Anfangsgriinden der Naturwissenschaft he returns to the attempted purely dynamic deduction and construction of matter.
\
The "essense" of matter i.e., its pure concept for experience, according to which it is nothing else than a totality of external relations, is resolved into a pure interaction of forces acting at a distance; but since these forces themselves occur in a double form, as attracting
\
and repelling forces, the dualism is not fundamentally overcome, but is only shifted back into the concept of force itself.
Modern physics has sought, from essentially different standpoints and motives, to overcome the old opposition between matter and force, which seemed sanctioned and made eternal in the classical system of mechanics. Heinrich Hertz s Prinzipien der Mechanik
takes the opposite course to that of previous philosophical specula by placing the sought unity in the concept of mass, instead of in the concept of force. Along with the fundamental concepts of
tion
space and time, only the concept of mass enters into the systematic construction of mechanics. The carrying out of this view presup poses, indeed, that we do not remain with gross perceptible mass and
gross perceptible motion but supplement the sensuously given ele ments, which by themselves do not constitute a lawful world, by assuming certain "concealed" masses and "concealed" motions.
it is shown to be necessary and calculation of phenomena, and without arous ing suspicion since mass is conceived by Hertz from the beginning
This supplementation takes place when
for the description
merely as a definite factor of calculation.

*
It is
intended to express
Cf. (44),
I,
204, 267
ff.,
332
II,
290
ff.,
303.
400
nothing but certain coordination of space and time values: a property by which we particle of mass," as Hertz defines it, coordinate unambiguously a certain point of space to a certain point
"is
"a
of time (and) 3 a certain point of space to every other
time."
(31, p.
29
Another attempt was made by general energetics to ff., 54.) reach a unified foundation for physics and with it for mechanics. Inert mass appears here merely as a definite factor of energy, as the capacity-factor of the energy of motion, which with certain other
capacity-factors shares with the
different types of energy, e.g., the empirical property of quantitative conservation. Energetics refuses to grant this law of conservation a special place and to recognize matter as a particular substance along with energy.
electricity,
(Cf. 60, p,
is
282
ff.)
But
logically unsatisfactory,
precisely in this we see very distinctly what which consists in that the principle of
inner connection
conservation refers to wholly different is not to be seen.
moments between which an
of relativity brings important clarification here too in combines the two principles of conservation, that of the conservation of energy and that of the conservation of mass into a This result it gains by applying its characteristic single principle.
The theory
it
that
manner
of thought;
it is
led to this result
on the conditions of measurement.
by general considerations The demand of the theory of
relativity (at first of the special theory) is that the law of the con servation of energy be valid not only with reference to any system of coordinates but also with reference to any other in uniform
motion relatively to it; it results from this presupposition, however, combined with the fundamental equations of Maxwell s electrodynamics that when a body in motion takes up energy E in the form of radiation its inert mass increases by a definite amount The mass of a body is thus a measure of its content of energy; (f ). if the energy content alters a definite amount then its mass alters 4 Its independent constancy is thus only an appear proportionately. it holds good only in so far as the system takes up and gives off ance; no energy. In the modern electron theory, it follows from the wellknown investigation of Kaufmann that the of an electron is not unchangeable, but that it rapidly increases with the velocity of the electron as soon as the latter approaches the velocity of light.
rectilinear
2
"mass"
3 *
Trans. Einstein (16a) and Planck (64 and 65).

While previously a distinction had been made between a and a "fictitious" mass of electrons, i.e., between an inertia, which came from its ponderable mass, and another, which they possessed solely because of their motion and their electric charge, in so far as this
"real"
opposed a certain resistance to every change of velocity, it now turns out that the alleged ponderable mass of the electrons is to be taken as
strictly
0.
matter thus seems completely replaced by the inertia of energy; the electron and thus the material atom as a system of electrons possesses no material but only "electromagnetic" mass.
inertia of
The
What was
of matter, as its substantial kernel,
previously regarded as the truly fundamental property is resolved into the equations of
the electro-magnetic field. in the same direction; but
The theory
it
of relativity goes further
and character.
gains one of its the equivalence of phenomena of inertia and weight. Here it is at first merely a calculation, a consideration of the same phenomena
its peculiar nuance This comes out especially in the process by which it fundamental propositions: in the establishment of
reveals in this too
from
different systems of reference,

it
which points the way.
We
shows, regard one and the same phenomenon now as a pure inertial movement and now as a movement under the influence The of a field of gravitation according to the standpoint we choose.
can, as
equivalence of judgment, here indicated, grounds for Einstein the

If certain physical identity of phenomena of inertia and weight. accelerated motions occur for an observer within his sphere of obser vation, he can interpret them either by ascribing them to the effects
of a field of gravitation or conceive the system of reference from which he makes his measurements as in a certain acceleration. The
two assumptions accomplish precisely the same in the description and can thus be applied without distinction. We can as Einstein expresses it produce a field of gravitation by a mere change of the system of coordinates. (17, p. 10; cf. 18, p. 45 ff.) Hence, it follows that to attain a universal theory of gravitation we need only assume such a shift of the system of reference and estab
of the facts
lish its
fashion
consequences by calculation. It suffices that in purely ideal we place ourselves at another standpoint to be able to deduce certain physical consequences from this change of standpoints. What was previously done in the Newtonian theory of gravitation by the dynamics of forces is done by pure kinematics in Einstein s
4\U
EINSTEIN
of different systems of reference
theory,
i.e.,
by the consideration
moving relatively to each other. In emphasizing this ideal element in Einstein s theory of gravita tion, the empirical assumption on which it rests must naturally not
be forgotten. That we change in thought, by the mere introduction of a new system of reference, a field of inertia into a field of gravita
tion,
and a
field of
gravitation of special structure into a field of
on the empirical equality of inert and gravitating masses of bodies, as was established with extraordinary exactitude by the Only the fact that investigation of Eotvos to which Einstein refers. to all bodies found at the same place in the field gravitation imparts of gravitation, the same amount of acceleration, and that thus it is for any definite body the same constant, i.e., mass, which determines its inertial effects and its gravitational effects, renders possible that transformation of the one into the other, from which the Einstein 6 But it is especially interesting and important from a theory starts. general methodological standpoint that this fundamental fact is given a completely different interpretation than in the Newtonian
inertia, rests
Einstein urges against the latter is that it regis of the equivalence of gravitating and inert What was established masses, but did not interpret it. (18, p. 44.) In as a fact by Newton is now to be understood from principles.
mechanics.
tered the
What
phenomenon
problem one can trace how gradually the question as to the of matter and of gravitation is superseded by another epistemological formulation of the question, which finds the "essence"
this
"essence"
and
of a physical process expressed wholly in its quantitative relations its numerical constants. Newton never ceased to reject the
question as to essence, which met him ever again, and the phrase that physics has to do merely with the "description of phenomena" was first formulated in his school and is an expression of his method. 6 But so little was he able to escape this question that he expressly
urged that universal attraction was not itself grounded in the essence of body, but that it came to it as something new and alien. Weight is, as he emphasizes, indeed a universal but not an essential property
of matter.
(59,
Vol. Ill, p. 4.)
What
this distinction
between
of the
the universal and the essential

For more
ff
;
means from the standpoint

.
p. 27
8
Freundlich (24), pp. 28 and 60 f and Schlick (79) 45 ff. Keill, Introductio ad veram Physicam (1702), (36); cf. 7, II, 404 ff.
detail, cf.
cf.
Einstein
(18), p.
403
physicist, who has to do merely with the laws of phenomena, and thus with the universality of the rule to which they are subjected, is here left in the dark. Here lies a difficulty, which has been felt again and
again in the tedious controversy of physicists and philosophers on the actuality and possibility of force acting at a distance. Kant,
in his Metaphysischen Anfangsgrunden der Naturwissenschaft, urges against Newton that, without the assumption that all matter merely by virtue of its essential properties exercises the action we call
is
gravitation, the proposition that the universal attraction of bodies proportional to their inert mass, would be a totally contingent and
mysterious fact. (35, IV, p. 421.) In its solution of this problem the general theory of relativity has followed the path prescribed by the peculiarity of the physical method. The numerical proportion
which is universally found between inert and heavy masses becomes the expression of physical equivalence, of the essential likeness of the two. The theory of relativity concludes that it is the same
quality of the body, which is expressed according to circumstances as "inertia" or as "weight." have here in principle the same
We
which, e.g., in the electromagnetic theory of led to insight into the "identity" of light waves and electrical light waves. For this identity too means nothing else and nothing more
us,
procedure before
mysterious than that we can represent and master the phenomena of light and the phenomena of dielectric polarization by the same
equations and that the same numerical value results for the velocity of light and for that of dielectric polarization. This equality of values means to the physicist likeness in essence since for him
essence
is
defined in terms of exact determinations of measure and
magnitude.
In the advance to this insight, there
may
be traced
historically a definite series of steps, a culmination of physical theories. The physics of the eighteenth century was in general rooted in a substantialistic view. In the fundamental investigations of Sadi Carnot on thermodynamics heat was still regarded as a
material,
electricity
and the assumption seemed unavoidable, in understanding and magnetism, of a particular electric and magnetic
"matter." Since the middle of the nineteenth century, however, there appears in place of this "physics of materials," ever more definitely and distinctly the physics that has been called the "physics of principles." Here a start is not made from the hypothetical existence of certain materials and agents, but from certain universal
404
relations,
of particular
lar
which are regarded as the criteria for the interpretation phenomena. The general theory of relativity stands methodologically at the end of this series, since it collects all particu
systematic principles into the unity of a supreme postulate, in the postulate not of the constancy of things, but of the invariance of certain magnitudes and laws with regard to all transformations of
e
system of reference.
evolution, that is characteristic of physical conceptual construction in general, is seen when we go from the concept of matter to the second fundamental concept of modern physics, to the concept
The same
of the ether. 7
magnetic
and and things. A sensuous description of its fundamental properties was sought by comparing it now with a perfectly incompressible But the more one attempted fluid, now with a perfectly elastic body. to work these pictures out in detail, the more distinctly was it seen that they demanded the impossible of our faculty of presentation, that they demanded the unification of absolutely conflicting proper ties. Thus modern physics was more and more forced to abandon But in principle this sort of sensuous description and illustration. the difficulty was unchanged also when one asked, not concerning any concrete properties of the ether, but merely concerning the abstract laws of its motion. The attempt to construct a mechanics of the ether led little by little to the sacrifice of all the fundamental
principles of classical mechanics; it was seen that, really to carry it through, one would have to give up not only the principle of the equality of action and reaction, but the principle of impenetrability
in which, e.g., Euler saw the kernel and inclusive expression of all mechanical laws. Ether was and remained accordingly, in an expres sion of Planck, the "child of sorrow of the mechanical theory;" the assumption of the exact validity of the Maxwell-Hertzian differential
idea of the ether, as the bearer of optical and first conceived in the greatest possible analogy affinity with our presentations of empirically given materials
The
effects
was at
equations for electrodynamic processes in the pure ether excludes the possibility of their mechanical explanation. 8 An escape from
7 Here I do not go into details in the development of the hypothesis of the ether; they have been expounded from the standpoint of epistemology by e. g., Aloys Muller (55, p. 90 ff.) and Erich Becher (2, p. 232 ff.). On the
Substance and Function (8, p. 215 ff.). Lenard (4oa and 6), especially declares for (67), p. 64 ff. the possibility and necessity of a "mechanics of the ether."
following
8
cf.
Cf.
Planck

this
405
antinomy could only be reached by reversing the treatment. Instead of asking about the properties or constitution of the ether as a real thing, the question must be raised as to by what here in
right
general one seeks for a particular substance with particular material properties and a definite mechanical constitution. What if all the difficulties of the answer are based on the question itself, there
being in it no clear and definite physical meaning? That is, in fact, the new position which the theory of relativity takes to the question of the ether. According to the outcome of Michelson s investigation
and the
principle of the constancy of the propagation of light, each observer has the right to regard his system as "motionless in the ether;" one must thus ascribe to the ether simultaneous rest with
reference to wholly different systems of coordinates K, K", which are in uniform translatory motion relatively to each other.
,
That, however, is an obvious contradiction and it forces us to abandon the thought of the ether as a somehow moving or motionless "sub
stance,"
as a thing with a certain
"state
of
motion."
Physics,
instead of imagining
some
sort of hypothetical substratum of
phenom
ena and losing itself in consideration of the nature of this substra tum, is satisfied, as it becomes a pure "physics of fields," with the body of field-equations themselves and their experimentally verifiable "One cannot define," says e.g., Lucien Poincare, "ether validity. material properties without committing a real fallacy, and to by characterize it by other properties than those, the direct and exact knowledge of which is produced for us by experiment, is an entirely The ether useless labor condemned to sterility from the beginning. is defined when we know the two fields, which can exist in it, the
electric
point.
and magnetic fields, in their magnitude and direction at each The two fields can change; by custom we speak of a motion
propagated in the ether; but the phenomenon accessible to experi Here we is the propagation of these changes." (75, p. 251.) face one of those triumphs of the critical and functional concept again over the naive notion of things and substances, such as are found
ment
more and more

the ether
is
in the history of exact science.
The
physical role of
ended as soon as a type of exposition is found for the electrodynamic laws into which it does not enter as a condition. "The theory of relativity," remarks one of its representatives, "rests on an entirely new understanding of the propagation of electromag netic effects in empty space; they are not carried by a medium, but
406
neither do they take place
the electromagnetic
field in
by unmediated action at a distance. But empty space is a thing possessing self-
existent physical reality independently of all substance. Indeed, one must first accustom himself to this idea; but perhaps this habituaeasier by the remark that the physical properties which are given most adequate expression in Maxwell s equations, are much more perfectly and exactly known than the properties of any substance." (Laue, 41, p. 112.) Habituation with to a "thing independent of any substance" can indeed be as regard
tion will be
made
of this field,
little
attributed to
common human
understanding as to the episte-
mologically trained understanding; for precisely to the latter does substance mean the category on the application of which rests all
But it is obvious that we have here possibility of positing "things." an inexactitude of expression and that the "independent physical only
reality" of the electromagnetic field can mean nothing but the reality of the relations holding within it which are expressed in the equations of Maxwell and Hertz. Since they are for us the ultimate attainable of physical knowledge, they are set up as the ultimate attain object
substance
The idea of the ether as an inexperienceable excluded by the theory of relativity in order to give conceptual expression merely to the pure properties of empirical
able reality for us.
is
knowledge.
For
we do not need
this purpose, however, according to the theory of relativity, the fixed and rigid reference body, to which classical
mechanics was ultimately referred. The general theory of relativity no longer measures with the rigid bodies of Euclidean geometry and classical mechanics, but it proceeds from a new and more inclusive standpoint in its determination of the universal linear element ds. In place of the rigid rod which is assumed to retain the same unchang ing length for all times and places and under all particular conditions of measurement there now appear the curved coordinates of Gauss. If any point P of the space-time continuum is determined by the four parameters Xi, x 2 x 3 x 4 then for it and an infinitely close point P there is a certain "distance" ds, which is expressed by the formula:
,
ds 2
= gndxi 2
g22dx 2
gssdxs*
+
.
g 4 4dx 4
.
2g 12 dxidx 2
2gi 3 dx 1 dx 3 +.
in which the magnitudes gn, g 22 with the place in the continuum.
.
.g44
have values, which vary
In this general expression, the
MATTER, ETHEK AND SPACE
407
formula for the linear element of the Euclidean continuum is con tained as a special case. We need not here go into details of this
determination;
9
its
in general different
from each other
essential result, however, is that measurements result for each place in the spaceis
time continuum.
Each point
system of reference outside of
it,
referred, not to a rigid and fixed but to a certain extent only to itself
Thus all measurements become infi compared with the rigid straight lines of Euclidean geometry, which are freely movable in space without change of form; and yet, on the other hand, all these infinitely various determinations are collected into a truly universal and unitary system. We now apply, instead of given and finite reference bodies, only "reference
to infinitely close points.
nitely fluid as
mollusks"
and
these
that
"mollusks"
as Einstein calls them; but the conceptual system of all satisfies the demand for an exact description of
natural processes. For the universal principle of relativity demands all these systems can be applied as reference bodies with equal
right
and with the same consequences

;
in the formulation of the

is
universal laws of nature the form of the law
to be completely inde
p.
pendent
of the choice of the mollusk.
(18,
67.)
Here
is
ex
pressed again the characteristic procedure of the general theory of relativity; while it destroys the thing-form of the finite and rigid reference body it would thereby only press forward to a higher form
form of nature and its laws. Only and outdoing the difficulties which resulted even for by heightening classical mechanics from the fact of the relativity of all motions, does it hope to find an escape in principle from these difficulties. "The clearer our concepts of space and time become," as was said in the outline of mechanics, which Maxwell has given in his gjjozt more do we see tJ8&*#*effi*$img to which Matter and Motion, our dynamic doctrines refer, beloi^ trTj^s^ngl^y^sêm^ At first we might think that we, as conscious IJemgs, must have as~necessary elements of our knowledge, an absolute knowledge of the place, in But which we find ourselves, and of the direction in which we move
of object, to the true systematic
>
w0rk>
"the
which was undoubtedly that of many sages of antiquity, In space, disappears more and more from the idea of the physicist. like any other there are no milestones; one part of space is precisely We find ourselves in a part, so that we cannot know where we are.
this opinion,
Cf. Einstein (17
and
18,
pp. 59
ff.); cf.
below VI.
408
EINSTEIN
waveless sea without stars, without compass and sun, without wind and tide, and cannot say in what direction we move. We have no
log that
we can
cast out to
make
a calculation;
we can indeed
deter
mine the degree of our motion in comparison with neighboring bodies, but we do not know what the motion in space of these bodies From this mood of "ignorabimus," into which physics (51, p. 92 f.) was sinking more and more, only a theory could free it which grasped
is."
the problem at
tions,
its root; and, instead of modifying the previous solu transformed fundamentally the formulation of the question. The question of absolute space and absolute motion could receive
ual mobile
only the solution which had been given to the problem of the perpet and the squaring of the circle. It had to be made over
from a mere negative expression into a positive expression, to be changed from a limitation of physical knowledge to a principle of such knowledge, if the true philosophic import, which was con cealed in it, was to be revealed.
CHAPTER V
THE CONCEPTS or SPACE AND TIME OF CRITICAL IDEALISM AND THE THEORY OF RELATIVITY
have hitherto sought primarily to understand the special and general theory of relativity on its physical side. In fact, this is the standpoint from which it must be judged and one does it poor service if one seeks precipitately to interpret its results in purely "philo sophical" or indeed in speculative and metaphysical terms. The theory contains not one concept, which is not deducible from the intellectual means of mathematics and physics and perfectly representable in them. It only seeks to gain full consciousness of precisely
these intellectual
We
means by seeking not only to represent the
result of
physical measurement, but to gain fundamental clarity concerning the form of any physical measurement and its conditions.
Thereby
it
of the critical
"possibility
seems indeed to come into the immediate neighborhood and transcendental theory, which is directed on the of experience;" but it is nevertheless different from it in
general tendency. For, in the language of this transcendental criticism, the doctrine of space and time developed by the theory of relativity is a doctrine of empirical space and empirical time, not of pure space and pure time. As far as concerns this point, there is
its
scarcely possible a difference of opinion; and, in fact, all
critics,
Kantian and the Eingtein-Minkowski theories of space and time seem to have reached essentially the same result. 1 From the standpoint of a strict empiricism, one could attempt to dispute the possibility of a doctrine of "pure space" and of "pure time;" but the conclusion cannot be avoided that in so far as such a doctrine is justified, it must be independent of all results of concrete measurement and of the particular conditions, which prevail in the latter. If the concepts of pure space and pure
the
who have compared
time have in general any definite justified meaning, to use a phrase of the theory of relativity, then this meaning must be invariant with regard to all transformations of the doctrine of the empirical measure ment of space and time. The only thing that such transformations
*
Cf, esp.
Natorp
p. 323
(56, p.
ff.)
sen-Kohler
(26,
392 ff.). Honigswald and more recently Sellien

409
(33,
p.
88
ff.).
Frischei-
(81, p. 14 ff.).
410
can and
will accomplish is that they teach us to draw the line more between what belongs to the purely philosophical, "transcen sharply dental," criticism of the concepts of space and time and what belongs
merely to the particular applications of these concepts. Here, in fact, the theory of relativity can perform an important indirect service for the general criticism of knowledge, if we resist the to translate its propositions directly into propositions of temptation
the criticism of knowledge. Kant s doctrine of space and time developed to a large extent on the basis of physical problems, and the conflict carried on in the
natural science of the eighteenth century on the existence of absolute time and absolute space affected him keenly from the beginning. Before he approached the problems of space and time as a critical philosopher, he had himself lived through the various and opposite solutions by which contemporary physics sought to master these
problems. Here, at first, contrary to the dominant scholastic opinion, he took his stand throughout on the basis of the relativistic In his Neuen Lehrbegriff der Bewegung und der Ruhe of the view.
year 1758, the thirty-four year old Kant set up the principle of the relativity of all motion with all decisiveness and from it attacked the traditional formulation of the principle of inertia. "Now I begin to see," he says after he has illustrated the difficulties of the
concept of
"absolute
motion"
with well-known examples,
"that
lack something in the expression of motion
and
rest.
should
a body rests without adding with regard to what thing it and never say that it moves without at the same time naming the objects with regard to which it changes its relation. If I wish to imagine also a mathematical space free from all creatures as a receptable of bodies, this would still not help me. For by what should I distinguish the parts of the same and the different places, which are occupied by nothing corporeal?" (35, II, 19.) But Kant, in his further development did not at first remain true to the norm, which he here set up so decisively and of which a modern
never say
rests,
physicist has said that it deserves to be set up in iron letters over each physical lecture hall. 2 He ventured to abandon the concept of inertial force, of vis inertiae; he refused to pour his thoughts on the
principles of mechanics
2 "into
the mill of the Wolffian or of any
Streintz (82), p. 42.
SPACE AND TIME OF CEITICAL IDEALISM
411
other famous system of doctrine." But while he opposed in this way the authority of the leading philosophers, he could not perman ently withdraw himself from the authority of the great mathematical In his Versuch, den Begriff der negativen physicists of his time.
Grossen in die Weltweisheit einzufiihren of the year 1763, he took his place at the side of Euler to defend with him the validity of the Newtonian concepts of absolute space and absolute time, and six years later, in his essay on the first grounds of the difference of
regions in space (1769), he sought to support the proof, that Euler
had attempted, of the existence of absolute space from the principles of mechanics, by another, purely geometrical consideration, which
"would give practical geometricians a conclusive reason to be able to affirm the reality of their absolute space with the "evidence" which is customary to them." (35, II, 394.) But this is indeed only an
episode in Kant s evolution; for only a year later the decisive critical turn in the question of space and time had taken place in his Inaugural Dissertation of the year 1770. By it the problem receives an entirely
new form;
dental
it is
philosophy"
removed from the field of physics to that of "transcen and must be considered and solved according to
the general principles of the latter. But the transcendental philosophy does not have to do primarily with the reality of space or of time, whether these are taken in a
metaphysical or in a physical sense, but it investigates the objective significance of the two concepts in the total structure of our empirical knowledge. It no longer regards space and time as things, but as
"sources of knowledge." It sees in them no independent objects, which are somehow present and which we can master by experi ment and observation, but "conditions of the possibility of experi ence," conditions of experiment and observation themselves, .which, again for their part are not to be viewed as things. What like time and space makes possible the positing of objects can itself never be given to us as a particular object in distinction of possible experience, the forms of from others. For the
"forms"
^"
intuition as well as the pure concepts of the understanding, are not met again as contents of real experience. Rather the only possible manner in which we can ascribe any sort of "objectivity" to these
forms must consist in that they lead to certain judgments to which we
must ascribe the values of necessity and universality. The meaning is thus indicated, in which one can henceforth inquire as to the objec-
412
tivity

of
space or time.
Whoever demands absolute
thing-like
"being"
correlates for
them
strains after shadows.
For their whole
consists in the
of
meaning and function they possess for the complexes judgments, which we call science, whether geometry or arithmetic,
mathematical or empirical physics. What they can accomplish as presuppositions in this connection can be exactly determined by transcendental criticism; what they are as things in themselves is a
vain and fundamentally unintelligible question. This basic view conies out clearly even in the Inaugural Dissertation. Even here
absolute space and time possessing an existence separate from empiri cal bodies and from empirical events, are rejected as nonentities, as
mere conceptual fictions (inane rationis commentum.) The two, space and time, signify only a fixed law of the mind, a schema of connection by which what is sensuously perceived is set in certain of coexistence and sequence. Thus the two have, in /"relations
spite of their "transcendental
ideality," "empirical reality,"
but this
means always only their validity for all experience, which however must never be confused with their existence as isolated ob
reality
jective contents of this experience itself. "Space is merely the form of external intuition (formal intuition) and not a real object that can
be perceived by external intuition. Space, as prior to all things which determine it (fill or limit it), or rather which give an empirical intuition determined by its form, is, under the name of absolute space, nothing but a mere possibility of external phenomena. If we try to separate one from the other, and to place space
outside
all
phenomena, we arrive at a number
of
empty determi
nations of external intuition, which, however, can never be possible perceptions; for instance, motion or rest of the world in an infinite
space, i.e., a determination of the mutual relation of the two, which can never be perceived, and is therefore nothing but the predi cate of a mere idea." (34, p. 457; Muller trans., p. 347.) when Einstein characterizes as a fundamental feature Accordingly, of the theory of relativity that it takes from space and time "the last remainder of physical objectivity," it is clear that the theory only accomplishes the most definite application and carrying through of
empty
the standpoint of critical idealism within empirical science itself. Space and time in the critical doctrine are indeed distinguished in
their validity as types of order from the contents, which are ordered in them; but these forms possess for Kant a separate existence neither
OF CRITICAL IDEALISM
in the subjective
413
ni*"fci ttaf Objective sense. The conception, that space and time as SUdêctivc forms into which sensations enter ready in the mind" before all experience, not as "physical" but as
"lie
This con "psychical" realities, today scarcely needs refutation. ception indeed seems to be indestructible, although Fichte poured upon it his severe but appropriate scorn; but it disappears of itself for everyone who has made clear to himself even the first conditions
of the transcendental formulation of the question in opposition to the The meaning of the principle of order can in general psychological.
be comprehended only in and with what is ordered; in particular, it is urged in the case of the measurement of time that the determina tion of the temporal positions of particular empirical objects and processes cannot be derived from the relations of the phenomena to absolute time, but that conversely the phenomena must determine and make necessary their positions in time for each other. "This unity in the determination of time is dynamical only, that is, time
not looked upon as that in which experience assigns immediately place to every existence, for this would be impossible because abso lute time is no object of perception by which phenomena could be
is
its
held together; but the rule of the understanding through which alone the existence of phenomena can receive synthetical unity in time determines the place of each of them in time, therefore a priori and
as valid for all
trans., p. 175.)
It is
time."
(34, p.
245 and 262;
cf.
56, p. 332;
cf.
Muller
such a
"rule
of the understanding," in
which
is
expressed the
synthetic unity of phenomena and their reciprocal dynamical rela relations tion, on which rests all empirical spatial order, all objective
of spatial
spatii"
"community"
i.e.,
language
is
The "communio in the corporeal world. that a priori form of coexistence, which in Kant s characterized as "pure intuition" is, as he expressly
of sub urges, only empirically knowable for us by the commercium stances in space, i.e., by a whole of physical effects, that can be
pointed out in experience.
We
read, in a passage of the Critique of
Pure Reason, which appears especially significant and weighty in connection with the development of the modern theory of relativity: "The word communion (Gemeinschaft) may be used in two senses,
,
communio or commercium. We use it here in the the latter sense: as a dynamical communion, without which even known empirically. We can local communio -spatii could never be
meaning
either
414
EINSTEIN S THEORY ef .RELATIVITY
easily perceive in our experience, that c^nn t^t^s -influences only can lead our senses in all parts of from one>e$yjct to another; that the
,sy;a<r
which plays between our eyes and celestial bodies produces a mediate communion between us and them, and proves the coexistence of the
light
we cannot change any place empirically (perceive such a change) unless matter itself renders the perception of our own place possible to us, and that by means of its reciprocal influence only matter can evidence its simultaneous existence, and thus (though
latter; that
mediately only)
(34, p. 260;
c/.,
its coexistence, even to the most distant objects." Miiller trans., p. 173 f.) The spatial order of the
corporeal world, in other words, is never given to us directly and sensuously, but is the result of an intellectual construction, which takes its start from certain empirical laws of phenomena and from that point seeks to advance to increasingly general laws, in which
grounded what we call the unity of experience as a spatiotemporal unity. But is there not found in this last expression the characteristic and
finally is
decisive opposition between the theory of space and time of critical idealism and the theory of relativity? Is not the essential result of
and time measurement time dependent on the state of motion of the system from which it is made there seem to result only infinitely many and infinitely diverse "place-times," time. We which, how ever, never combine into the unity of have already seen, however, that this view is erroneous, that the destruction of the substantialistic unity of space and time does not destroy their functional unity but rather truly grounds and confirms it.
this theory precisely the destruction of the unity of space
demanded by Kant?
If all
of
is
"the"
In fact, this state of affairs is not only above, p. 33 ff. 54 ff.) the representatives of the theory of relativity among the granted by
(C/.
physicists, but is expressly emphasized by them. "The boldness and the high philosophical significance of Einstein s doctrine consists," we that it clears away the traditional read, e.g., in the work of Laue,
"in
prejudice of one time valid for all systems. Great as the change is, which it forces upon our whole thought, there is found in it not the
slightest epistemological difficulty.
r
For
in
Kant
manner
of expres
sion time is, like space, a pure form of our intuition; a schema in which w e must arrange events, so that in opposition to subjective and
may gain objective meaning. This arranging can only take place on the basis of empirical knowlhighly contingent perceptions they
SPACE AND TIME OF CRITICAL IDEALISM
415
edge of natural laws. The place and time of the observed change of a heavenly body can only be established on the basis of optical laws. That two differently moving observers, each one regarding himself at rest, should make this arrangement differently on the basis of the
same laws of nature, contains no logical impossibility. Both arrange ments have, nevertheless, objective meaning since there may be deduced exactly from each of them by the derivative transformation
formulae that arrangement valid for the other moving observer." This one-to-one correlation and not the oneness of the (40, p. 36 f .) values gained in the different systems, is what remains of the notion
"unity of time"; but precisely in it is expressed all the more sharply the fundamental view that this unity is not to be represented in the form of a particular objective content, but exclusively in the
of the
The "dynamic unity of temporal relations. retained as a postulate; but it is seen that we cannot satisfy this postulate if we hold to the laws of the Newtonian mechanics, but that we are necessarily driven to a new and more
determinations" is
form of a system of valid
mination shows
classical
its
universal and concrete form of physics. The "objective" deter itself thus to be essentially more complex than the
mechanics assumed, which believed it could literally grasp hands the objective determination in its privileged systems of reference. That a step is thereby taken beyond Kant is inconfor he shaped his "Analogies of Experience" essentially on testible; the three fundamental Newtonian laws: the law of inertia, the law of the proportionality of force and acceleration, and the law of the equality of action and reaction. But in this very advance the doc trine that it is the "rule of the understanding," that forms the pattern In of all our temporal and spatial determinations, is verified anew.
with
the special theory of relativity, the principle of the constancy of the velocity of light serves as such a rule; in the general theory of relativ ity this principle is replaced by the more inclusive doctrine that all
Gaussian coordinate systems are of equal value for the formulation It is obvious that we are not con cerned here with the expression of an empirically observed fact,
of the universal natural laws.
but with a principle which the understanding uses hypothetically as a norm of investigation in the interpretation of experience, for how could an --infinite -totality be "observed"? And the meaning and its justification of this norm rest precisely on the fact that only by
application could
we hope
to regain the lost unity of the object,
416
namely, the
relations."
EINSTEIN S THEORY or RELATIVITY

unity of
"synthetic
phenomena according
to temporal
neither on the constancy of with which the naive sensuous view of the world rests those objects nor on the constancy of particular spatial and temporal measure
physicist
The
now depends
ments gained from a particular system, but he
affirms, as a condition
of his science, the existence of "universal constants" and universal laws, which retain the same values for all systems of measurement.
Anfangsgrunden der Naturwissenschaft, to the problem of absolute space and time, formu Kant, returning lates a happy terminological distinction, which is suited to character
Metaphysischen
ize
In his
more sharply the
relation of critical idealism to the theory of

itself
Absolute space, he urges here too, is in relativity. and indeed no object; it signifies only a space relative
nothing
to every other
To make it a real I can think outside of any given space. means to confuse the logical universality of any space with thing which I can compare any empirical space as included in it with the physical universality of real extension and to misunderstand the Idea of reason. The true logical universality of the Idea of space thus
which
not only does not include its physical universality, as an all-inclusive container of things, but it is precisely of a sort to exclude it. We
an absolute space, i.e., an ultimate unity of all spatial determinations; but not in order to know the absolute movements of empirical bodies, but to represent in the same "all movements of the material as merely relative to each other, as
should, in fact, conceive
lute space
"Abso alternatively reciprocal, but not as absolute motion or rest." is thus necessary not as a concept of a real object, but as
an Idea, which should serve as a rule for considering all motions in it as merely relative, and all motion and rest must be reduced to the
absolute space,
if
the
phenomena
of the
same are
to be
made
into a
definite concept of experience that unifies phenomena."
(35,
IV,
472 f.) The logical universality of such an idea does not conflict with the theory of relativity; it starts by regarding all motions in space as merely relative because only in this way can it combine em into a definite concept of experience, that unifies all phenomena.
383
f.,
On
the basis of the
demand
for the totality of determinations
it
negates every attempt to make a definite particular system of reference the norm for all the others. The one valid norm is merely
e idea of the unity of nature, of exact determination itself. The mechanical view of the world is overcome from this standpoint.
417
grounded by the general theory of relativity includes under a supreme principle of knowl edge along with the phenomena of gravitation, which form the real classical field of the older mechanics, the electrodynamic phenomena.
"unity
The
in a
of
nature" is
new
sense, since
it
That
many
advance to this "logical universality of the Idea," trusted presentational pictures must be sacrificed need not disturb us; this can affect the "pure intuition" of Kant only in so
in order to
it is
far as
misunderstood as a mere picture and not conceived and estimated as a constructive method. In fact, the point at which the general theory of relativity must implicitly recognize the methodic presupposition, which Kant calls It lies, in fact, in the "pure intuition" can be pointed out exactly.
"coincidence" to which the general theory of relativity reduces the content and the form of all laws of nature. ultimately If we characterize events by their space-time coordinates Xi, x 2 x 3 x 4 x i, x 2 x 3 x 4, etc., then, as it emphasizes, everything that physics
concept of
can teach us of the "essence" of natural processes consists merely in assertions concerning the coincidences or meetings of such points. reach the construction of physical time and of physical space
We
merely in this way; for the whole of the space-time manifold is noth 3 Here is the point at ing else than the whole of such coordinations. which the ways of the physicist and of the philosopher definitely What the part, without their being thereby forced into conflict. and is for him a concrete measurable calls physicist
"space"
"time"
result of coordination, according to law, of the particular points; for the philosopher, on the contrary, space and time signify nothing else than the forms and modi, and
manifold, which he gains as the
thus the presuppositions, of this coordination itself. They do not result for him from the coordination, but they are precisely this coordination and its fundamental directions. It is coordination from the standpoint of coexistence and adjacency or from the
standpoint of succession, which he understands by space and time as "forms of intuition." In this sense, both are expressly
defined in the Kantian Inaugural Dissertation.
.
"Tempus
non
est
sed subjectiva conditio, per objectivum aliquid et reale naturam mentis humanae necessaria, quaelibet sensibilia certa lege sibi
. . .
coordinandi f
3
et
intuitus
purus
Spatium
est
Einstein
(17), p. 13 f.; (18), p. 64.
418
subjectivum
et

ideale et e natura mentis stabili lege profisiscens veluti
schema omnia omnino externe sensa sibi coordinandi. (35; II, 416, Whoever recognizes this law and this schema, this possibility 420.) of relating point to point and connecting them with each otherA has recognized space and time in their "transcendental meaning," can abstract here from any psychological by-meaning of the con^pu of form of intuition. We can thus conceive the "world-points" Xi x 2 Xs x 4 and the world-lines, which result from them, so abstractly that we understand under the values Xi x 2 x 3 X4 nothing but certain
mathematical
of
parameters;
the
"meeting"
of
involves a comprehensible
"possibility
meaning only succession," which we
if
we take
time.
such world-points as a basis that
call
coincidence,
a unification, which is still a separation, since the same point is conceived as belonging to different lines: all this finally demands that synthesis of the
is
which
not to
mean
identity,
manifold, for which the term
"pure
intuition"
was formulated.
The most general meaning of this term, which indeed was not always grasped by Kant with equal sharpness, since more special meanings
.
and applications were substituted involuntarily in his case, is merely that of the serial form of coexistence and of succession. Nothing is
thereby presupposed concerning special relations of measurement in the two, and in so far as these depend in particular on the relations of the physical in space, we must guard against seeking to find an exhaus
tive determination in the of the
"real."
mere
"forms
of
possibility"
of the relations
(Cf.
below VI.)
When
e.g.,
in the
mathematical
foundations of the theory of relativity the formula is deduced for the "distance" of the two infinitely close points Xi x 2 x 3 x 4 and Xi
,
dxi, x 2
-J-
dx 2 x 3
,
dx 3 x 4
,
dx 4 this cannot indeed be conceived

,
as a rigid Euclidean distance in the ordinary sense, since there is involved in it, by the addition of time as a fourth dimension, not a
magnitude of space but rather one of motion; but the fundamental form of coexistence and succession and their reciprocal relation and
unmistakably contained in this expression of the general Not that the theory, as has been occasionally objected, presupposes space and time as something already given, for it must be declared free of this epistemological circle, but in the sense that it cannot lack the form and function of spatiality and temporality
"union"
is
linear element.
in general.
419
What seems to render understanding difficult at this point between the physicist and the philosopher is the fact that a common problem
is
found here, which both approach from entirely different

p
"ar
sides.
The
;
"
process of measurement interests the critic of knowledge only as he seeks to survey in systematic completeness the concepts, are used in this process, and to define them in the utmost
sharpness.
But any such
definition
is
unsatisfying and
fundamen
tally unfruitful to the physicist as long as it is not connected with any definite indication as to how the measurement is to be made in the
concrete particular case. "The concept exists for the physicist," says Einstein in one place neatly and characteristically, "only when the possibility is given of finding out in the concrete case whether the concept applies or not." (18, p. 14.) Thus the concept of simul
taneity, for example, only receives a definite meaning, when a method is given by which the temporal coincidence of two events is deter
mined by certain measurements, by the application of optical signals; and the difference which is found in the results of this measurement seems to have as a consequence the ambiguity of the concept. The
philosopher has to recognize unconditionally this longing of the physicist for concrete determinateness of concepts; but he is ever
again brought to the fact that there are ultimate ideal determina tions without which the concrete cannot be conceived and made intel
ligible.
To make
clear the opposition in formulation of the question
which is here fundamental, one can contrast to Einstein s expression one of Leibniz. "On pent dire," we read in Leibniz Nouveaux Essais, il nefaut point s imaginer deux etendues, I une abstraite, de Vespace, I autre concrete, du corps; le concret n etant tel que par t abstrait." (43, V, 115.) As we see, it is the unity of the abstract and the concrete, of the ideal and the empirical in which the demands of the physicist and the philosopher agree; but while the one goes from experience to the idea, the other goes from the idea to experience. The theory of relativity holds fast to the "pre-established harmony between pure mathematics and physics;" Minkowski, in the well-known concluding words of his lecture, "Space and Time," has expressly taken up again and brought to honor this Leibnizian term. But this harmony is for the physicist the incontestable premise from which he strives to reach the particular consequences and applications, while for the
"qu
critic of
knowledge the
"possibility"
of this
harmony
constitutes the
real problem.
The
basis of this possibility he finds ultimately in the
420
fact that
any physical assertion, even the simplest determination of magnitude established by experiment and concrete measurement, is connected with universal conditions, which gain separate treatment
in pure mathematics,
i.e.,
we can point out the concept of the concept of space, the concept of time, and the concept of number, function as the fundamental elements, which enter as presuppositions
into every question which physics can raise. None of these concepts can be spared or be reduced to another so that, from the standpoint of
logico-mathematical constants. constants into a short formula,
that any physical assertion involves certain If we desire to bring all of these
the critique of cognition, each represents a specific and characteristic motive of thought; but on the other hand, each of them possesses
an actual empirical use only along with the others and in systematic connection with them. The theory of relativity shows with especial distinctness how, in particular, the thought of function is effective as a necessary motive in each spatio-temporal determination. Thus knows its fundamental concepts never as logical "things in physics
but only in their reciprocal combination; it must, be open to epistemology to analyze this product into its however, It thus cannot admit the proposition that the particular factors. of a concept is identical with its concrete application, but it meaning will conversely insist that this meaning must be already established before any application can be made. Accordingly, the thought of space and time in their meaning as connecting forms of order is not
themselves,"
by measurement but is only more closely defined and We must have grasped the concept of the as something spatio-temporal, we must have understood the meaning expressed in it, before we can ask as to the coincidence of events and seek to establish it by special methods of measurement.
first
created
given a definite content.

"event"
In general, physics sees itself placed by its fundamental problem from the beginning between two realms, which it has to recognize and between which it has to mediate without asking further as to their On the one side, stands the manifold of data of sensation, on the other a manifold of pure functions of form and order. Physics, as an empirical science, is equally bound to the "material" content, which sense perception offers it, and to these formal principles in which is expressed the universal conditions of the It has to "invent" or to derive deduc "possibility of experience." tively the one as little as the other, i.e., neither the whole of empirical
"origin."
421
contents nor the whole of characteristic scientific forms of thought, but its task consists in progressively relating the realm of "forms" to the data of empirical observation and, conversely, the latter to the former. In this way, the sensuous manifold increasingly loses
its
"contingent"
anthropomorphic character and assumes the imprint
of thought, the imprint of systematic unity of form. Indeed "form," just because it represents the active and shaping, the genuinely
creative element,
moving.
must not be conceived as rigid, but as living and Thought comprehends more and more that form in its
peculiar character cannot be given to it at one stroke, but that the existence of form is only revealed to it in the becoming of form and
law of this becoming. In this way, the history of physics represents not a history of the discovery of a simple series of "facts," but the discovery of ever new and more special means of thought.
in the
But
in all
change of these means of thought there

"sure
is
nevertheless
science,"
revealed, as surely as physics follows the
course of a
the unity of those methodic principles upon which rests the formula tion of its question. In the system of these principles, space and
time take their fixed place, although they are not to be conceived as fixed things or contents of presentation. The ancient view believed
that
it
possessed and encompassed the spatio-temporal unity of
being directly in presentation. To Parmenides and fundamentally the whole ancient world being was given "like the mass of a wellrounded sphere." With the reform of Copernicus, the security of Modern science knows that this possession was gone once for all.
there
is
a definite spatio-temporal order of phenomena for knowledge
only in so far as knowledge progressively establishes it, and that the only means of establishing it consists in the scientific concept of law. But the problem of such a general orientation remains for
thought and becomes the more urgent the more thought knows it as a problem never to be solved definitively. Precisely because the unity of space and time of empirical knowledge seems to flee eternally before all our empirical measurements, thought comprehends that it must
seek
it must avail itself of new and ever sharper the merit of the theory of relativity not only to have proved this in a new way but also to have established a prin universal laws of ciple, i.e., the principle of the co-variancy of the it
eternally
and that
instruments.
It is
nature with regard to all arbitrary substitutions, by which thought can master, out of itself, the relativity which it calls forth.
422
In the analysis of spatial and temporal measurements,
made by
the theory of relativity this fundamental relation can be traced in detail. This analysis does not begin by accepting the concept of the
"simultaneity"
of
two processes as a
self-evident
known datum, but by demanding an explanation

tion, which, as
of
and immediately it an explana
a physical explanation, cannot consist in a general conceptual definition, but only in the indication of the concrete methods of measurement, by which "simultaneity" can be empirically
pointed out.
practically in
adjacency
is
of such processes as take place of space or in immediate spatial point at first presupposed; we assume, as Einstein explains,
"the
The simultaneity
same"
the determinability of
"simultaneity"
for events,
diately adjacent spatially, or,
more
exactly, for events in
which are imme immediate
spatio-temporal adjacency (coincidence), without defining this concept In fact, recourse here to a mediating physical method of (17, 3.)
measurement seems neither desirable nor possible; for any such method would always presuppose the possibility of making a temporal coordination between diverse events, thus, e.g., of establishing simultaneity" of a definite event with a certain position of the hands
of a clock found at the
"same"
"the
place.
The
real
problem of the theory
are no longer concerned with temporally connecting spatially adjacent series of events with each other, but rather series of events spatially removed from each other.
of relativity begins only
If
when we
we assume
common to A and B. It is seen that "B-time" but no time every attempt to establish such a common time as an empirically measurable time, is bound to a definite empirical presupposition concerning the velocity of light. The assumption of the uniform
the simultaneity of what
velocity of light enters implicitly into all our assertions concerning time common to A is spatially distant.
and and
that there is established for the two points of space A a certain "place-time," then we possess only an "A-time"
and B is gamed when one establishes by definition that the which it which light takes in going from A to B is equal to the takes in going from B to A. Let us assume that a ray of light is sent at A-time tA from a clock found in A to B, and then at B-time, tfi, the ray of light is reflected to A and reaches A again at A-time, t Aj then we establish by definition that the two clocks of A and B are to be called "synchronous" if t B Thus for the * A IBi\ first time an exact determination is made of what we are to under"time,"
"time,"

stand by the
"time"
423
of
of
an event and by the
"simultaneity"
two
processes; "the time" of an event is what is told us by a motionless clock found at the place of the event simultaneously with the event, a clock which runs synchronously with a certain motionless clock and indeed synchronously with the latter at all times." (16, p. 28 f.)
That the "forms" of space and time as definite forms of the co ordination of different contents already enter into the concrete determinations, which are here made for the procedure of the physical measurement of time, scarcely needs special explanation. The two are immediately assumed in the concept of the "place-time;" for the
possibility
"now"
is
involved in
it
of grasping a definitely distinguished

"here."
"here"
This and does not signify indeed the whole of space and time, to say nothing of all the concrete relations within the two to be established by measure ment; but it represents the first foundation, the unavoidable basis
"now"
in a definitely distinguished
of the two. The first primitive difference, which is expressed in the mere positing of a and a remains thus, for the theory of relativity, too, an indefinable on which it grounds its complex physical definitions of space and time values. And while for these
"here"
"now"
definitions it appeals to a definite assumption concerning the law of the propagation of light, this, too, involves the presupposition that a certain condition that we call occurs in succession at different
"light"
and according to a definite rule, in which what space and time mean as mere schemata of coordination, is obviously contained. The epistemological problem seems indeed to be heightened when we reflect on the reciprocal relation of space and time values in the fundamental equations of physics. What is given in these equations is the four-dimensional the continuum of events in general, the temporal determinations in this continuum not being separated from the spatial. The intuitive difference between a spatial dis tance and a temporal duration, which we believe ourselves to grasp immediately, plays no role in this purely mathematical determina
places
"world,"
tion.
According
:
to
the
temporal
equation
of
the
Lorentz-
transf ormation
424
EINSTEIN
the time differential At
between two events with reference to
does not disappear when the time differential At of the same disap pears with reference to K; the purely spatial distance of two events with reference to has as a consequence in general the temporal
sequence of the
same with reference
to
This leveling of space
and time values is developed even further in the general theory of Here it is seen to be impossible to construct a reference relativity. out of fixed bodies and clocks of such a sort that place and system
time are directly indicated by a fixed arrangement of measuring rods and clocks relatively to each other; but each point of the continuous series of events is correlated with four numbers, xi, x 2 x 3 x 4 which possesses no direct physical meaning, but only serve to enumerate the points of the continuum in a definite but arbitrary way. This
, , ,
correlation need not have such properties that a certain group of values Xi x 2 x 3 must be understood as the spatial coordinates and The (18, p. 38, 64.) opposed to the "temporal" coordinate x 4
.
demand
of
Minkowski that
"space
for itself
and time
"a
for itself
be
sort of union of completely degraded to shadows" and that only the two shall retain independence" seems thus now to be strictly
realized.
Now
at
any
rate,
this
demand
contains
nothing
ceased to conceive space and time as things in themselves or as given empirical objects. For the realm of ideas is for him a "realm of shadows," as Schiller called it,
terrible for the critical idealist,
who has
since
no pure idea corresponds directly to a concrete real object, but rather the ideas can always only be pointed out in their syste
matic community, as fundamental moments of concrete objective knowledge. If it thus appears that physical space and time measure ments can be assumed only as taking place in common, the difference in the fundamental character of space and time, or order in coexist ence and succession is not thereby destroyed. Even if it is true that, as Minkowski urges, no one has perceived a place save at a time and a time save at a place, there remains a difference between what is to be understood by spatial and by temporal discrimination. The
factual interpenetration of space and time in all empirical physical measurements does not prevent the two from being different in prin
ciple,
not as
objects,
Although two observers
but as types of objective discrimination. can assume in different systems and
the arrangement of the series of events in the orders of space and time to be different, it is still always a series of events and thus a
SPACE AND TIME OP CRITICAL IDEALISM
425
continuum both spatial and temporal, which they construct in their measurements. Each observer distinguishes from his standpoint of measurement a continuum, which he calls from another, which he calls but he can, as the theory of relativity shows, not assume without further consideration that the arrangement of phenomena in these two schemata must be similar from each system of reference. There may thus, according to Minkowski s "world postulate," be given only the four-dimensional world in space and
"space,"
"time;"
time,
"with
and
"the
projection into space and

freedom;"
a certain
time" may be possible this only affects the different spatio-
temporal interpretations of phenomena, while the difference of the form of space from that of time is unaffected. For the rest, here too the transformation-equation reestablishes
objectivity
results
and unity, since it permits us to translate again the found in one system into those of the other. Also, if one
seeks to clarify the proposition of Minkowski that only the insepa rable union of space and time possesses independence, by saying that this union itself, according to the results of the general theory of
becomes a shadow and an abstraction, and that only the unity of space, time and things possesses independent reality, 4 then this classification only leads us back again to our first epistemoFor that neither "pure space" nor "pure time" nor logical insight.
relativity,
the reciprocal connection of the two, but only their realization in
some empirical material gives what w e call i.e., the physical of things and of events, belongs to the fundamental doctrines being of critical idealism. Kant himself did not weary of referring re
r
"reality,"
peatedly to this indissoluble connection, this reciprocal correlation of the spatio-temporal form and the empirical content in the exist ence and structure of the world of experience. "To give an object," we read, this is not meant again as mediate only, but if it means
"if
to represent something immediately in intuition, is nothing else but to refer the representation of the object to experience Even space and time, however pure these concepts may be of all
and however certain it is that they are represented would lack nevertheless all objective validity, all sense and meaning, if we could not show the necessity of their use with reference to all objects of experience. Nay, their
that
is
empirical,
in the
mind
entirely a priori,
See Schlick
(79), p. 51; cf. p. 22.
426
representation
EINSTEIN S THEORYÔF RELATIVITY

is
tive imagination,
a pure ^drema, always referring to that reproduc which calls up the objects of experience, without
"
which objects would-be meaningless. (34, p. 195; cf. Muller 127 f.) The that space and time possess p. meaning,
"iaeal"
trans.,
"in
the
thus does not involve any sort of particular existence, which would possess prior to things and independently of them, but it they rather expressly denies it the ideal separation of pure space and
mind"
pure time from things (more exactly, from empirical phenomena), not only permits but demands precisely their empirical "union." This union the general theory of relativity has verified and proved in a new way, since it recognizes more deeply than all preceding physical
theories the dependency belonging to all empirical measurement, to all determination of concrete spatio-temporal relations. 5 The rela tion of experience and thought that is established in the critical doctrine does not contradict this result in any way, but rather it
it and brings it to its sharpest expression. It is indeed at glance strange and paradoxical that the most diverse epistemological standpoints, that radical empiricism and positivism as well
confirms
first
have all appealed to the theory of relativity in fundamental views. But this is satisfactorily explained by the facts that empiricism and idealism meet in certain
as critical idealism
of
support
their
presuppositions with regard to the doctrine of empirical space and of empirical time, and that the theory of relativity sets up just such a doctrine. Both here grant to experience the decisive role, and both
laws. b
teach that every exact measurement presupposes universal empirical But the question becomes all the more urgent as to how we
reach these laws, on which rests the possibility of all empirical measurement, and what sort of validity, of logical "dignity" we grant to them. Strict positivism has only one answer to this ques
tion: for it all knowledge of laws, like all knowledge of objects, is grounded in the simple elements of sensation and can never go beyond their realm. The knowledge of laws possesses accordingly in prin the same purely passive character that belongs to our knowledge ciple
of
any particular sensuous qualities. Laws are treated like things whose properties one can read off by immediate perception. Mach
attempts, quite consistently with his standpoint, to extend this
6
On
the
"relativization"
of the difference of space

ff.
and time,
cf.
also below,
VII.
6
(8), p.
191
ff.; cf.
Sellien (81), p. 14
427
of consideration to pure mathematics also and the deduction relations. The way in which we gain the differen tial quotient of a certain function, as he explains, is not distinguished of its
manner
fundamental
in principle from the way in which we establish any sort of properties or changes of physical things. As in the one case we subject the thing, so in the other case we subject the function to certain
opera
and simply observe how it m of the function y = x to the operation

tions
"reacts"
to them.
The
reaction
of differentiation out of which
dy = the equation mx m -
results
"is
a distinguishing
mark
of x
as the blue-green color in the solution of copper in Here we find clearly before us the (49, p. 75.) sharp line of distinction between critical idealism and positivism of
just as
much
sulphuric
acid."
Mach
type.
fields are to
be regarded as what
That the equations governing larger or smaller is truly permanent and substantial,
make possible the gaining of a stable picture of the world, 7 that they thus constitute the kernel of physical objectivity: this is the fundamental view in which the two theories combine. The
since they
question concerns only the manner of establishing, only the exact grounding, of these equations. Idealism urges against the stand
"pure experience" as the standpoint of mere sensation, that equations are results of measurement; all measurement, however, presupposes certain theoretical principles and in the latter certain universal functions of connection, of shaping and coordination. We never measure mere sensations, and we never measure with mere
point of
all
sensations, but in general to gain
any
sort of relations of
measure
>
of perception and replace it by "given" a conceptual symbol, which possesses no copy in what is immediately sensed. If there is anything that can serve as a typical example of
this state of affairs,
it is
ment we must transcend the
the development of
modern physics
in the
theory of relativity. It is verified again that every physical theory, to gain conceptual expression and understanding of the facts of
experience,
facts are
is
must
free itself
from the form
in
which at
first
these
of
8 immediately given to perception.
That the theory

is,
relativity
7 8
founded on experience and observation

!
of course,
See
C/.
"Les fails d experience, pris dans leur brulalite ne sauraient servir au raisonnement mathematique; pour alimenter ce native, raisonnement, Us doivent etre transformes et mis sous forme symbolique."
Mach (49), p. 429. Duhem (15, p. 322):
428

But, on the other hand, its essential achievement interpretation that it gives to the observed
beyond question.
consists in the
facts, in
new
led to subject
classical
It
the conceptual interpretation by which it is progressively the most important intellectual instruments of
older physics
to a critical
revision.
mechanics and the
has been pointed out with justice that it has been precisely the oldest empirical fact of mechanics, the equality of inert and heavy
masses, which, in the new interpretation it has received from Ein stein, has become the fulcrum of the general theory of relativity.
(24a.)
The way
this fact
in
the foundations of the
which the principle of equivalence and with it new theory of gravitation have been deduced
from
pure
can serve as a logical example of the meaning of the We conceive ourselves in "thought-experiment" in physics. the position of an observer, who, experimenting in a closed box,
establishes the fact that all bodies left to themselves
move, always with constant acceleration, toward the floor of the box. This fact can be represented conceptually by the observer in a double manner: in the first place, by the assumption that he is in a temporarily con
stant field of gravity in which the box is hung up motionless, or, in the second place, by the assumption that the box moves upward with a constant acceleration whereby the fall of bodies in it would repre sent a movement of inertia. The two the inertial movement and the
:
effect of gravitation, are
thus in truth a single phenomenon seen and judged from different sides. It follows that the fundamental law that we establish for the movement of bodies must be such that it
includes equally the
As
is
seen,
phenomena of inertia and those of gravitation. we have here no empirical proposition abstracted from
particular observations, but a rule for our construction of physical
concepts: a demand that we make, not directly of experience, but rather of our manner of intellectually representing it. "Thoughtexperiments"
of such force
and
fruitfulness cannot be explained
and
justified
It is
empiristic theory of physical knowledge. not in contradiction with this that Einstein refers gratefully
by the purely
to the decisive stimulus, which he received from Mach (20); for a sharp distinction must be made between what Mach has accomp
Newton s fundamental con and the general philosophical consequences he has drawn from cepts, this achievement. Mach himself has, as is known, granted wide
lished as a physicist in his criticism of
scope to the pure
"thought-experiment"
in his
own
logic of physics;
SPACE AXD TIME OF CRITICAL IDEALISM

but, of a
429
closely considered, he has thereby already left the ground purely sensatory founding of the fundamental concepts of 9 That there is no necessary connection between the theory physics. of relativity and Mach s philosophy may be concluded from the
more
fact,
among
other things, that
it is
precisely one of the

all
first
advocates
of this theory,
Max
Planck,
who among
if
modern
most
sharply criticized
(09.)
and fought against the presuppositions
physicists has of
this philosophy.
Even
one take the theory of relativity
as an achievement and outcome of purely empirical thought, it is thereby a proof and confirmation of the constructive force imma nent in this thought by which the system of physical knowledge is
distinguished from a
See
mere
"rhapsody
of
perceptions."
Mach
(50, p.
180
ff.); cf. (8), p.
316
ff.
and
(39), p.
86
f.
CHAPTER
VI
EUCLIDEAN AND NON-EUCLIDEAN GEOMETRY

In the preceding considerations, however, we have taken up only an achievement of the general theory of relativity,
incidentally
above all others, seems to involve a "revolution of In the working out of the theory, it is seen that the previous Euclidean measurements are not sufficient; the develop ment of the theory can only take place by our going from the Euclidean continuum, which was still taken as a basis by the special theory of relativity, to a non-Euclidean four-dimensional space-time continuum and seeking to express all relations of phenomena in it. Thus a question seems answered physically which had concerned the epistemology of the last decades most vitally and which had been answered most diversely within it. Physics now proves not only the possibility, but the reality of non-Euclidean geometry; it shows that we can only understand and represent theoretically the rela tions, which hold in space, by reproducing them in the lan guage of a four-dimensional non-Euclidean manifold. The solution of this problem from the side of physics was, on the one hand, for a long time hoped for as keenly, as, on the other hand, its possibility was vigorously denied. Even the first founders and representatives of the doctrine of non-Euclidean geometry sought to adduce experiment and concrete measurement in confirmation of their view. If we can establish, they inferred, by exact terres
which,
thought."
"real"
or astronomical measurements, that in triangles with sides of very great length the sum of the angles differs from two right angles, then empirical proof would be gained that in empirical space
trial
"our"
the propositions not of Euclidean geometry, but of one of the others were valid. Thus, e.g., Lobatschefsky, as is known, used a triangle
EI
S,
of the earth
whose base EI E 2 was formed by the diameter of the orbit and whose apex S was formed by Sirius and believed that
he could, in this \vay, prove empirically a possible constant curvature of our space. The fallacy in method of any such attempt (48.) must be obvious, however, to any sharper epistemological analysis
of the problem and it has been pointed out from the side of the mathematicians with special emphasis by H. Poincare. No measure430
431
ment, as Poincare objects with justice, is concerned with space itself but always only with the empirically given and physical objects in No experiment therefore can teach us anything about the space. ideal structures, about the straight line and the circle, that pure
basis; what it gives us is always only knowledge of the relations of material things and processes. The propositions of geometry are therefore neither to be confirmed nor refuted
geometry takes as a
by
ever come into conflict with the postulates of Euclid; but, on the other hand, no experiment will ever contradict the postulates of Lobatschefsky. For granted,
experience.
No
experiment
will
that some experiment could show us a variation in the sums of the angles of certain very great triangles, then the conceptual represen tation of this fact would never need to consist in, and methodologi
cally could not consist in,
changing the axioms of geometry, but
rather in changing certain hypotheses concerning physical things. What we would have experienced, in fact, would not be another
structure of space, but a new law of optics, which would teach us that the propagation of light does not take place in strictly recti linear fashion. "However, we turn and twist," Poincare therefore
concludes,
"it
is
in geometry."
(72, p.
impossible to attach a rational meaning to empiricism 92 ff.) If this decision holds and if it can be
proved, on the other hand, that among all possible self-consistent geometries the Euclidean possesses a certain advantage of "sim which plicity" since it defines the minimum of those conditions under
experience
it
is
possible in general, there
would then be established
for
an exceptional position from the standpoint of the critique of knowledge. It would be seen that the different geometries, which are equivalent to each other from a purely, formal standpoint, as regards
their logical conceivability, are yet distinguished in their fruitfulness in the founding of empirical science. "The geometries are dis
tinguished from each other in principle," one can conclude, "only by reference to their epistemological relation to the concept of experience; for this relation is positive only in the case of the Euclid
ean geometry." 1 In connection, however, with the new development of physics in the general theory of relativity, this epistemological answer seems to become definitely untenable. Again and again the fact has been
1
Cf.
Honigswald
(32);
on the following
c/.
Bauch
(1), p.
126
ff.
432
appealed to in the controversy concerning the epistemological justi fication of the different geometries that what determines value must not be sought in formal but in transcendental logic; that the com patibility of a geometry with experience is not involved but rather its
"positive fruitfulness," i.e.,
the
"founding
of
experience,"
that
it
can
was thought to be found in Euclidean give. The latter appeared as the real and unique "foundation geometry. of possibility of knowledge of reality," the others, on the contrary, always as only the foundations of the possible. But with regard to the extraordinary role that the concepts and propositions of Riemannian geometry played in the grounding and construction of Einstein s
this latter
And
theory of gravitation, this judgment cannot be supported. Supported by the same logical criterion of value, one now seems forced rather
to the opposite conclusion: non-Euclidean space is alone while Euclidean space represents a mere abstract possibility. In any event, the logic of the exact sciences now finds itself placed before
"real,"
a new problem.
fact of the fruitfulness of non-Euclidean geom can no longer be contested, since it has been verified, not only in particular applications, but in the structure of a complete new system of physics; what is in question is the explana tion to be given to this fact. And here we are first forced to a nega tive decision, which is demanded by the first principles of the theory
The
etry for physics
of relativity.
Whatever meaning we may ascribe to the idea
of
non-Euclidean geometry for physics, for purely empirical thought, the assertion has lost all meaning for us that any space, whether Euclidean or non-Euclidean, is the space. Precisely this was
"real"
the result of the general principle of relativity, that by it "the last remainder of physical objectivity" was to be taken from space.
fold,
of measurement within the physical mani within that inseparable correlation of space, time, and the physically real object, which the theory of relativity takes as ulti
Only the various relations
mate, are pointed out; and
it
is
affirmed that these relations of
measurement
find their simplest exact mathematical expression in the language of non-Euclidean geometry. This language, however, is and remains purely ideal and symbolic, precisely as, rightly under
"
geometry could alone be. The which alone it can express is not that of things, but that of reality laws and relations. And now we can ask, epistemologically, only one question: whether there can be established an exact relation
stood, the language of Euclidean
433
and coordination between the symbols of non-Euclidean geometry and the empirical manifold of spatio-temporal If physics
"events."
answers
this
question
affirmatively,
it
then
answering negatively. space that it affirms as the condition of every physical theory involves, as has been seen, no assertion concerning any definite particular structure of space in itself, but is concerned only with that function
ground
for
For
epistemology has no the of priori"

"a
of
in general, that is expressed even in the general con "spatiality" cept of the linear element ds as such, quite without regard to its character in detail.
If it is seen thus, that the determination of this element as is done in Euclidean geometry, does not suffice for the mastery of certain problems of knowledge of nature then nothing can prevent us, from a
methodological standpoint, from replacing it by another measure, in so far as the latter proves to be necessary and fruitful physically.
in either case one must guard against taking the "preestablished harmony between pure mathematics and physics," that is revealed to
But
us in increasing fullness and depth in the progress of scientific knowl edge, as a naive copy theory. The structures of geometry, whether Euclidean or non-Euclidean, possess no immediate correlate in the world of existence. They exist as little physically in things as they
do psychically
their
in our
"presentations"
but
all
their
"being,"
i.e.,
validity
and
truth,
consists
in
their
ideal
meaning.
The-
virtue existence, that belongs to them by virtue of their definition, by in principle, not to be inter of a pure logical act of assumption is, changed with any sort of empirical "reality." Thus also the appli-
which we grant to any propositions of pure geometry, can never rest on any direct coinciding between the elements of the ideal In place geometrical manifold and those of the empirical manifold. of such a sensuous congruence we must substitute a more complex and more thoroughly mediate relational system. There can be no copy or correlate in the world of sensation and presentation for what the
cability,
points, the straight lines and the planes of pure geometry signify. of similarity, Indeed, we cannot in strictness speak of any degree from the ideal, for the of or less difference of the
greater
"empirical"
two belong to fundamentally different species. The theoretical between the two, relation, which science [nevertheless establishes fast to consists merely in the fait, that it, while granting and holding a more the difference in contend of the two series, seeks to establish
434

All verification,
exact and perfect correlation between them.
which
the propositions of geometry can find in physics, is possible only in this way. The particular geometrical truths or particular axioms, such as the principle of parallels, can never be compared with particu
1
but we can always only compare with the whole of physical experience the whole of a definite system of axioms. What Kant says of the concepts of the understanding in general, that they make letters out of phenomena so that we can read only serve
lar experiences,
"to
1
them as experiences" holds in particular of the concepts of space. They are only the letters, which we must make into words and proposi tions, if we would use them as expressions of the laws of experience.
If the goal of harmony is not reached in this indirect way, if it appears that the physical laws to which observation and measurement lead us cannot be represented and expressed with sufficient exactitude and
simplicity by a given system of axioms, then we are free to determine which of the two factors we shall subject to a transformation to
harmony between them. Before thought ad vances to a change of one of its "simple" geometrical laws it will first make the complex physical conditions that enter into the measurement responsible for the lack of agreement; it will change the If this does not lead "physical" factors before the "geometrical."
reestablish the lost
to the goal
and if it is seen on the other hand, that surprising unity and systematic completeness can be reached in the formulation of the "laws of nature" by accepting an altered conception of geometrical
methods, then in principle there
is
nothing to prevent such a change.
we conceive the geometrical axioms, not as copies of a given reality, but as purely ideal and constructive structures, then they are subjected to no other law than is given them by the system of
For
if
thought and knowledge. If the latter proves to be realizable in a purer and more perfect form by our advancing from a relatively simpler geometrical system to a relatively more complex, then the criticism of knowledge can raise no objection from its standpoint. It will be obliged to affirm only this: that here too "no intelligible meaning can be gained" for empiricism in geometry. For here, too,
experience does not ground the geometrical axioms, but it only selects from among them, as various logically possible systems, of which
is derived strictly rationally, certain ones with regard to their concrete use in the interpretation of phenomena. 2 Here, too,
each one
On this relation of the problem of metageometry to the problem of "experi

cf.
ence,"
esp. Albert
Gorland
(28, p.
324
ff.)
435
Platonically speaking, phenomena are measured by Ideas, by the foundations of geometry, and these latter are not directly read out of the sensuous phenomena.
But when one grants to non-Euclidean geometry
in this sense
meaning and fruitf ulness for physical experience, the general methodic difference can and must be urged, that still remains between it and
Euclidean geometry.
This difference can no longer be taken from
their relation to experience, but it must be recognized as based on certain "inner" moments, i.e., on general considerations of the theory
of relations.
special
and exceptional
its
logical position, a
in
fundamen
Euclidean
tal simplicity of ideal structure,
can be recognized
previous sovereignty within And here it is precisely the fundamental doctrine of the physics. general theory of relativity, that, translated back into the language of
logic
geometry even
if it
must abandon
and general methodology, can

is
this special position.
of relativity, which
constant,
^nrvfli.iirq
establish and render intelligible Euclidean geometry rests on a definite axiom peculiar to it. As the_geometrv of space of a
it.
J
is
cWnêrized by
the thorough-going
relativity of all places and ma^pi Its formal determinations tnrlf^. are in principle independent of any absolute determinations of
magnitude.
While,
e.g.,
in the
geometry
of Lobatschefsky, the
sum
of the angles of a rectilinear triangle is different from 180 and indeed the more so, the more the surface area of the triangle increases, the of Euclidean geometry.
absolute magnitude of the lines enters into none of the propositions Here for every given figure a "similar"
can be constructed; the particular structures are grasped in their pure "quality," without any definite "quantum," any absolute value of number and magnitude, coming into consideration in their definition. This indifference of Euclidean structures to all absolute determinations of magnitude and the freedom resulting here of the
erties,
particular points in Euclidean space from all determinations and prop form a logically positive characteristic of the latter. For the
proposition,
omnis determinatio
est
negatio,
holds here too.
The
assumption of the indeterminate serves as the foundation for the more complex assumptions and determinations which can be joined to it. In this sense, Euclidean geometry is and remains the "simplest," not in any practical, but in a strictly logical meaning; Euclidean space is, as Poincar6 expresses it, "simpler not merely in consequence of our mental habits or in consequence of any direct intuition, which
436

but
it is
we
possess of
it,
in itself simpler, just as a polynomial of the
simpler than a polynomial of the second degree." This logical simplicity belonging to Euclidean space in (72, p. 67.) the system of our intellectual meanings wholly independently of its relations to experience, is shown, e.g., in the fact that we can make
first
degree
is
any
"given"
space, that possesses
any
definite curvature, into Euclid
ean by regarding sufficiently small fields of it from which the differ ence conditioned by the curvature disappears. Euclidean geometry shows itself herein as the real geometry of infinitely small areas, and thus as the expression of certain elementary relations, which we take as a basis in thought, although we advance from them in certain
cases to
more complex forms. The development of the general theory
of relativity leaves this
methodic advantage of Euclidean geometry unaffected. For Euclid ean measurements do not indeed hold in it absolutely but they hold for certain "elementary" areas, which are distinguished by a certain The Euclidean expression of the simplicity of physical conditions. linear element shows itself to be unsatisfactory for the working out of the fundamental thought of the general theory of relativity, since it does not fulfill the fundamental demand of retaining its form
in
every
arbitrary
alteration
of
the
system
of
reference.
4
It
),
must be replaced by the

which
satisfies
2 general linear element (ds
= S
i
g
t*v
dx dx
M
v
this
demand.
If,
small four-dimensional
fields, it is
however, we consider infinitely expressly demanded that the pre
suppositions of the special theory of relativity, and thus
measurements
its Euclidean remain adequate for them. The form of the universal linear element here passes over into the Euclidean element of the special theory when the ten magnitudes g, which occur in this as functions of the coordinates of particular points assume definite constant values. The physical explanation of this relation, however consists in that the magnitudes are recognized as those which g^ v describe the gravitational field with reference to the chosen system of reference. The condition, under which we can pass from the pre
shall
suppositions of the general theory of relativity to the special theory, can accordingly be expressed in the form that we only consider regions
within which abstraction can be

gravitation.
This
is
made from the effects of fields of always possible for an infinitely small field and

it
437
holds further for finite fields in which, with appropriate choice of the system of reference, the body considered undergoes no noticeable acceleration. As we see, the variability of the magnitudes g
v,
which expresses the variation from the homogeneous Euclidean form of space, is recognized as based on a definite physical circum If we consider fields in which this circumstance is absent or stance.
it in thought, we again stand within the Euclidean Thus the assertion of Poincare that all physical theory and physical measurement can prove absolutely nothing about the if
we
cancel
world.
Euclidean or non-Euclidean character of space, since it is never concerned with the latter but only with the properties of physical in force. The abstraction reality in space remains entirely better expressed, the pure function) of homogeneous Euclidean (or,
space is not destroyed by the theory of relativity, but as such through it more sharply than before.
is
only
known
fact, the pure meaning of geometrical concepts is not limited what this theory teaches us about the conditions of measurement. by These concepts are indeed, as is seen now anew, neither an empirical datum nor an empirical dabile, but their ideal certainty and meaning It is shown that in fields where is not in the least affected thereby. we have to reckon with gravitational effects of a definite magnitude, the preconditions of the ordinary methods of measurement fall short, that here we can no longer use "rigid bodies" as measures of But this change as measures of time. length, nor ordinary of relations of measurement does not affect the calculation of space,
"clocks"
In
but the calculation of the physical relation between the measuring rods and rays of light determined by the field of gravitation. (Cf. The truths of Euclidean geometry would only be also 83, p. 85 ff.) affected if one supposed that these propositions themselves are noth
ing but generalizations of empirical observation, which we have established in connection with fixed bodies. Such a supposition, however, epistemologically regarded, would amount to a petitio
principii.
Even Helmholtz, who greatly emphasizes the empirical origin of the geometrical axioms occasionally refers to another view, which might save their purely ideal and "transcendental" character. The Euclidean concept of the straight line might be conceived not as a from certain physical observations, but as a purely
generalization
ideal concept, to be confirmed or refuted
by no experience,
since
438

to decide
we would have
by
it
whether any bodies
of nature
were to
be regarded as fixed bodies. But, as he objects, the geometrical axioms would then cease to be synthetical propositions in Kant s
they would only affirm something that would follow analyti from the concepts of the fixed geometrical structures neces It is, however, overlooked (30a, II, 30.) sary to measurement. by this objection that there are fundamentally synthetic forms of unity besides the form of analytic identity, which Helmholtz has here in mind and which he contrasts with the empirical concept as if the form of analytic identity were unique, and that the axioms of geome
sense, as
cally
try belong precisely to the former. Assumptions of this sort refer to the object in so far as in their totality they "constitute" the object and render possible knowledge of it; but none of them, taken for itself,
things.
can be understood as an assertion concerning things or relations of Whether they fulfill their task as moments of empirical knowledge can be decided always only in the indicated indirect way
:
by using them as building-stones in a theoretical and constructive system, and then comparing the consequences, which follow from the latter, with the results of observation and measurement. That the elements, to which we must ascribe, methodologically, a certain "simplicity," must be adequate for the interpretation of the laws of But even so, thought does nature, can not be demanded a priori. not simply give itself over passively to the mere material of experi ence, but it develops out of itself new and more complex forms to satisfy the demands of the empirical manifold.
retain this general view, then one of the strangest and, at appearance, most objectionable results of the general theory of It is a necessary consequence of this relativity receives a new light. that in it one can no longer speak of an immutably given theory
If
first
we
geometry
world.
of
measurement, which holds once for
all for
the whole
by
Since the relations of measurement of space are determined the gravitational potential and since this is to be regarded as in
general changeable from place to place, we cannot avoid the conclu sion that there is in general no unitary "geometry" for the totality of
space and reality, but that, according to the specific properties of the field of gravitation at different places, there must be found different
forms of geometrical structure. This seems, in fact, the greatest conceivable departure from the idealistic and Platonic conception of geometry, according to which it is the "science of the eternally exist-

ent,"
439
knowledge
uvavTus
of
what always
Relativism
"is
in the
same
x
\
state"
(del /card
sharpest expression of the fact that the problem of space has lost all ontological meaning in the theory of relativity. The purely methodological
seems here to pass over directly exov). into the field of logic; the relativity of places involves that of geometri cal truth. And yet this view is, on the other hand, only the
ravTo.
question has been substituted for the question of being. We_axe no longer concerned with wna* space T 1s^"aM with whether any definite character, whether Euclidean, Lobatschefskian or Iliemannian, is to be ascribed to it, but rather with what use is to be made of
the different systems of geometrical presuppositions in the inter pretation of the phenomena of nature and their dependencies accord
ing to law.
indeed
rests
If we call any such system a particular we can no longer attempt to grasp all of these spaces
"space,"
then
as intui-
Jtiye_parts to be united into
an intuitive whole. But this impossibility on the fact that we have here to do wjth a fundamentally problem, which as such stands outside the limits of intuitive repre
sentation in general. The space of pure intuition is always only ideal, being only the space constructed according to the laws of this intui tion, while here we are not concerned with such ideal syntheses and
their unity, but with the relations of measurement of the empirical and the physical. These relations of measurement can only be!
by proceeding from the^ each other, aad by permit dynamic dependency phenomena upon the spaceting phenomena to determine their positions reciprocally time manifold by virtue of this dependency. Kant too decisively
gained on the basis of natural laws,
of
i.e.,
belong to. urged that this form of dynamic determination didintuition as such, but that it is the "rules of the understanding"^ which alone give the existence of phenomena synthetic unity and
"riot
enable
(C/.
them
above
p. 79.)
to be collected into a definite concept of experience. The step beyond him, that we have now to make
results of the general theory of relativity, consists
I
\
on the basis of the
in the insight that geometrical axioms and laws of other than Eucli dean form can enter into this determination of the understanding,
which the empirical and physical world arises for us, and that the admission of such axioms not only does not destroy the unity of the
in
world,
the unity of our experiential concept of a total order of in this phenomena, but first truly grounds it from a new angle, since which we have to calculate way the particular laws of nature, with
i.e.,
440
EINSTEIN
brought to the unity of a that of the universal postulate of relativity. supreme principle, The renunciation of intuitive simplicity in the picture of the world
in space-time determination, are ultimately
thus contains the guarantee of its greater intellectual and systematic completeness. This advance, however, can not surprise us from the
epistemological point of view; for
scientific
it
expresses only a general law of
in particular of physical thought. Instead of speaking of the being or indeed of the coexistence of diversely ontologically constituted "spaces," which results in a tangible contradiction, the
and
theory of relativity speaks purely methodologically of the possibility

or necessity of applying different measurements, i.e., different geo metrical conceptual languages in the interpretation of certain physical manifolds. This possible application tells us nothing concerning the
"existenceof space, but merely indicates that by an appropriate choice of geometrical presuppositions certain physical relations, such as the field of gravitation or the electromagnetic field, can be described.
The connection between the purely conceptual thought, involved

working out of the general doctrine of the manifold and order, and physical empiricism (Empiric) here receives a surprising con firmation. A doctrine, which originally grew up merely in the immanent progress of pure mathematical speculation, in the ideal
in the
transformation of the hypotheses that lie at the basis of geometry, now serves directly as the form into which the laws of nature are
poured. The same functions, that were previously established as expressing the metrical properties of non-Euclidean space, give the equations of the field of gravitation. These equations thus do not
need for their establishment the introduction of new unknown forces

acting at a distance, but are derived from the determination and specialization of the general presuppositions of measurement.
Instead of a
in setting
the consideration of a
new complex of things, the theory is satisfied here by new general complex of conditions. Riemann,
his theory, referred to its future physical meaning in of which one is often reminded in the discussion of
up
prophetic words
the general theory of relativity. In the "question as to the inner ground of the relations of measurement of space," he urges, "the
remark can be applied that in a discrete manifold the principle of measurement is already contained in the concept of this manifold, but in the case of a continuous manifold it must come from else-
441
where. Either the real lying at the basis of space must be a discrete manifold or the basis of measurement must be sought outside it in
binding forces working upon it. The answer to this question can only be found by proceeding from the conception of phenomena,
founded by Newton and hitherto verified by experience and gradually reshaping this by facts that cannot be explained from it; investiga tions, which, like the one made here, proceed from universal concepts, can only serve to the effect that these works are not hindered by limitations of concepts and the progress in knowledge of the con
nection of things not hindered by traditional prejudices." (77). What is here demanded is thus full freedom for the construction of
geometrical concepts and hypotheses because only thereby can physi cal thought also attain full effectiveness, and face all future problems
resulting from experience with an assured and systematically per fected instrument. But this connection is expressed, in the case of Riemann, in the language of Herbartian realism. At the basis of
the pure form of geometrical space a real is to be found in which is to be sought the ultimate cause for the inner relations of measurement
If we carry out, however, with reference to this formu lation of the problem, the critical, "Copernican," revolution and thus conceive the question so that a real does not appear as a ground
of this space.
of space
struction
but so that space appears as an ideal ground in the con and progress of knowledge of reality, there results for us at
once a characteristic transformation. Instead of regarding "space" as a self-existent real, which must be explained and deduced from rather whether the "binding forces" like other realities, we ask now
a priori function, the universal ideal relation, that we call "space" involves possible formulations and among them such as are proper
to offer an exact
of certain
"fields
and exhaustive account

of
force."
of certain physical relations,

of the general theory
The development
answered this question in the affirmative; it has shown what appeared to Riemann as a geometrical hypothesis, as a mere possibility of thought, to be an organ for the knowledge of
of relativity has
here resolved into pure kine matics and this kinematics ultimately into geometry. The content of the latter must indeed by broadened and the "simple" Euclidean
reality.
The Newtonian dynamics
is
type of geometrical axioms must be replaced by a more complex type; but in compensation we advance a step further into the realm of without leaving the being, i.e., into the realm of empirical knowledge,
442
EINSTEIN S THEORY or RELATIVITY
sphere of geometrical consideration. By abandoning the form of Euclidean space as an undivided whole and breaking it up analytically
and by investigating the place of the particular axioms and their reciprocal dependence or independence, we are led to a system of pure a priori manifolds, whose laws thought lays down constructively, and in this construction we possess also the fundamental means for
representing the relation of the real structures of the empirical manifold.
The realistic view that the relations of measurement of space must be grounded on certain physical determinations, on "binding forces" of matter, expresses this peculiar double relation one-sidedly and
and unsatisfactorily. ground would destroy the methodological unity, which should be brought out. What relativistic physics, which has developed strictly and consistently from a theory of space and time measurement, offers us is in fact only the combination, the reciprocal determination, of the metrical and physi cal elements. In this, howeve^ there is found no one-sided relation and consequent, but rather a purely reciprocal relation, a pf ground correlation of the "ideaP and moments, of "matter" and In so far as we assume form," of the geometrical and the physical. any division at all in this reciprocal relation and take one element as and derivative, this prior" and fundamental, the other as distinction can be meant only in a logical, not in a real sense. In this sense, we must conceive the pure space-time manifold as the logical prius; not as if it existed and were given in some sense outside of and before the empirical and physical, but because it constitutes a principle and a fundamental condition of all knowledge of empirical and physical relations. The physicist as such need not reflect on this state of affairs; for in all the concrete measurements, which he makes, the spatio-temporal and the empirical manifold is given always only in the unitary operation of measurement itself, not in the abstract isolation of its particular conceptual elements and conditions. From these considerations the relation between Euclidean and non-Euclidean geometry appears in a new light. The real superiority of Euclidean geometry seems at first glance to consist in its concrete and intuitive determinateness in the face of which all "pseudo-geomthus,
epistemologically
regarded,
inexactly
For
this metaphysical use of the category of
"real"
"later"
etries"
fade into logical
"possibilities."
"being;"
only for thought, not for
JThese possibilities exist they seem analytic plays with

concepts, which can be
left
443
experience and with

this
"nature,"
knowledge. When we look back over our earlier considerations, view must undergo a peculiar and paradoxical reversal. Pure Euclidean space stands, as is now seen, not closer to the demands of empirical and physical knowledge than the non-Euclidean mani folds but rather more removed. For precisely because it represents the logically simplest form of spatial construction it is not wholly adequate to the complexity of content and the material determinateness of the empirical. Its fundamental property of homogeneity, its
unconsidered when we are concerned with with the synthetic unity of objective
axiom of the equivalence in principle of all points, now marks it as an abstract space; for, in the concrete and empirical manifold, there
,
never
is
V reigns
in
such uniformity, but rather thorough-going differentiation If we would create a conceptual expression for this it.
fact of differentiation in the sphere of geometrical relations
them
selves, then nothing remains but to develop further the geometrical conceptual language with reference to the problem of the hetero We find this development in the construction of metagegeneous." ometry. When the concept of the special three-dimensional mani fold with a curvature is broadened here to the thought of a system of manifolds with different constant or variable curvatures, a new
ideal means is discovered for the mastery of complex manifolds new conceptual symbols are created, not as expressions of things, but of
;
realized within
Whether these relations are possible relations according to law. phenomena at any place only experience can decide.
But it is not experience that grounds the content of the geometrical concepts; rather these concepts foreshadow it as methodological anticipations, just as the form of the ellipse was anticipated as a
conic section long before it attained concrete application and signifi cance in the courses of the planets. When they first appeared, the
systems of non-Euclidean geometry seemed lacking in all empirical meaning, but there was expressed in them the intellectual prepara tion for problems and tasks, to which experience was to lead later. Since the "absolute differential calculus," which was grounded on purely mathematical considerations by Gauss, Riemann and
a surprising application in Einstein s theory of held open gravitation, the possibility of such an application must be for all, even the most remote constructions of pure mathematics and especially of non-Euclidean geometry. For it has always been
Christoffel, gains
444

in the history of
shown
mathematics that its complete freedom the guarantee and condition of its fruitfillness. Thought contains does not advance in the field of the concrete by dealing with the
phenomena like pictures to be united into a single mosaic, but by sharpening and refining its own means of determination while guided by reference to the empirical and by the postulate If a proof were needed for of its determinateness according to law. this logical state of affairs, the development of the theory of relativity would furnish it. It has been said of the special theory of relativity
particular
it "substituted mathematical constructions for the apparently most tangible reality and resolved the latter into the former." (38, The advance to the general theory of relativity has brought p. 13). this constructive feature more distinctly to light; but, at the same time, it has shown how precisely this resolution of the "tangible" realities has verified and established the connection of theory and experience in an entirely new way. The further physical thought advances and the higher universality of conception it reaches the more does it seem to lose sight of the immediate data, to which the naive view of the world clings, so that finally there seems no return to these data. And yet the physicist abandons himself to these last and highest abstractions in the certainty and confidence of finding in them reality, his reality in a new and richer sense. In the progress of knowledge the deep words of Heraclitus hold that the way upward and the way downward are one and the same: 666f avo) KO.TW /UT?. Here, too, ascent and descent necessarily belong together: the direction of thought to the universal principles and grounds of
that
knowledge finally proves not only compatible with its direction to the particularity of phenomena and facts, but the correlate and condition of the latter.
CHAPTER
VII
THE THEORY OF RELATIVITY AND THE PROBLEM OF REALITY
We have attempted to show how the new concept of nature and of the object, which the theory of relativity establishes, is grounded in the form of physical thought and only brings this form to a final
Physical thought strives to determine and merely the natural object, but thereby necessarily expresses itself, its own law and its own principle. Here is revealed again that "anthropomorphism" of all our concepts of
to express in pure objectivity
conclusion and clarity.
nature to which Goethe, in the wisdom of old age, loved to point. "All philosophy of nature is still only anthropomorphism, i.e., man, at unity with himself, imparts to everything that he is not, this unity, draws it into his unity, makes it one with himself
We
we
can observe, measure, calculate, weigh, etc., nature as much as will, it is still only our measure and weight, as man is the measure of all things." Only, after all our preceding considerations, this
/ anthropomorphism"
itself is not to be understood in a limited but in a universal, critical and transcendental psychological way sense. Planck points out, as the characteristic of the evolution of
the system of theoretical physics, a progressive emancipation from anthropomorphic elements, which has as its goal the greatest possible
separation of the i^ystem of physics from the individual personality of the physicist. But into this "objective" system, free (68, p. 7.)
from
all
the accidents of individual standpoint and individual per
sonality, there enter those universal conditions of system, on which depends the peculiarity of the physical way of formulating problems.
The sensuous immediacy and particularity of the particular perceptual qualities are excluded, but this exclusion is possible only through the concepts of space and time, number and magnitude. In them physics determines the most general content of reality, since they
specify the direction of physical thought as such, as it were the form of the original physical apperception. In the formulation of the theory of relativity this reciprocal relation has been confirmed
throughout. The principle of relativity has at once an objective and a subjective, or methodological meaning. The "postulate of the absolute world," which it involves according to an expression of
445
446
Minkowski, is ultimately a postulate of absolute method. The general relativity of all places, times and measuring rods must be the last word of physics, because "relativization," the resolution of the
natural object into pure relations of measurement constitutes the kernel of physical procedure, the fundamental cognitive function of
physics.
If
we understand, however, how,
in this sense, the affirmation of
relativity develops with inner consequence
and necessity out
of the
very form of physics, a certain

*
critical limitation of this affirmation
The postulate of relativity may be the purest, most universal and sharpest expression of the physical concept of objectiv ity, but this concept of the physical object does not coincide, from
also appears.
the standpoint of the general criticism of knowledge, with reality
The progress of epistemological analysis is shown in absolutely. that the assumption of the simplicity and oneness of the concepts of Each of the reality is recognized more and more as an illusion.
original directions of knowledge, each interpretation, which it makes of phenomena to combine them into the unity of a theoretical
connection or into a definite unity of meaning, involves a special understanding and formulation of the concept of reality. There result here not only the characteristic differences of meaning in the objects of science, the distinction of the "mathematical" object from
the
"physical"
object, the
"physical"
from the
"chemical,"
the
"chemical"
from the
"biological,"
but there occur also, over against the
knowledge, other forms and meanings of such as the ethical, the aesthetic "form." laws, It appears as the task of a truly universal criticism of knowledge not to level this manifold, this wealth and variety of forms of knowledge
theoretical scientific
whole of
independent type and
and understanding of the world and compress them into a purely abstract unity, but to leave them standing as such. Only when we resist the temptation to compress the totality of forms, which here result, into an ultimate metaphysical unity, into the unity and simplic ity of an absolute "world ground" and to deduce it from the latter, do we grasp its true concrete import and fullness. No individual form can indeed claim to grasp absolute as such and to give it complete and adequate expression. Rather if the thought of such an ultimate definite reality is conceivable at all, it is so only as an Idea, as the problem of a totality of determination in which each particular function of knowledge and consciousness must
"reality"
RELATIVITY AND THE PROBLEM OF REALITY
447
cooperate according to its character and within its definite limits. If one holds fast to this general view, there results even within the
pure concepts of nature a possible diversity of approaches of which each one can lay claim to a certain right and characteristic validity. The "nature" of Goethe is not the same as that of Newton, because there prevail, in the original shaping of the two, different principles
of form, types of synthesis, of the spiritual and intellectual combina tion of the phenomena. Where there exist such diversities in funda
mental direction of consideration, the results of consideration cannot be directly compared and measured with each other. The naive realism of the ordinary view of the world, like the realism of dogmatic
falls into this error, ever again. It separates out of the of possible concepts of reality a single one and sets it up as a totality norm and pattern for all the others. Thus certain necessary formal points of view, from which we seek to judge and understand the world
metaphysics,
we
Whether things, into absolute beings. characterize this ultimate being as "matter" or nature" or "history," there always results for us in the end confusion in our
of
"life,"
phenomena, are made into
view of the world, because certain spiritual functions, that cooperate in its construction, are excluded and others are over-emphasized. It is the task of systematic philosophy, which extends far beyond the theory of knowledge, to free the idea of the world from this onesidedness.__JUiasJ jo. .grasp the whole system of symbolic fgrms^Jlie^ application of which produces for us the concept of an ordered reality, and by virtue of which subject and object, ego and world are
J
separated and opposed to each other in definite form, and

refer each individual in this totality to its fixed place.
If
it
must
as
we
would be assured, and the problem limits fixed, of each of the particular forms of the concept and of knowledge as well as of the general forms of the theoretical, ethical,
sume
this
solved, then the rights
Each particular aesthetic and religious understanding of the world. form would be "relativized" with. regard to the others, but since this "relativization" is throughout reciprocal and since no single
form but only the systematic totality can serve as the expression of and "reality," the limit that results appears as a thor oughly immanent limit, as one that is removed as soon as we again
"truth"
relate the individual to the
We trace the general problem,

use
it
system of the whole. which opens up here, no further but even the merely to designate the limits, that belong to any,
448
most universal, physical formulation of problems, because these limits are necessarily grounded in the concept and essence of this-
/way
of formulating the question.
All phj^sics considers
phenomena
1
under the standpoint and presupposition of their measurability.
It seeks to resolve the structure of being and process ultimately into a pure structure or order of numbers. The theory of relativity has brought this fundamental tendency of physical thought to its sharpest t
expression.
planation"
According to
it
the procedure of every physical
"ex
of natural process consists in coordinating, to

,
continuum, four numbers, Xi, x 2 possess absolutely no direct physical meaning but only serve to enumerate the points of the continuum a definite, but arbitrary The ideal, with which scientific physics began way." (18, p. 64). with Pythagoras and the Pythagoreans, finds here its conclusion;
of the space-time
,
,
each point x 3 x 4 which
"in
including those of pure space and time, are translated into pure numerical values. The logical postulate contained in the of number, which gives this concept its characteristic form, concept
all qualities,
seems now
classical
fulfilled in
a degree not to be surpassed;
all
sensuous and
intuitive heterogeneity has
mechanics and
passed into pure homogeneity. The physics seeks to reach this immanent goal of
conceptual construction by relating the manifold of the sensuously given to the homogeneous and absolutely uniform time. All differ ence of sensation
hereby reduced to a difference of motions; all a mere variety of spatial and temporal positions. But the ideal of strict homogeneity is not reached here since there are still always two fundamental forms of the
is
possible variety of content is resolved into
homogeneous
pure time.
itself
that are opposed to each other as pure space and

of relativity in its
it
The theory
development advances
beyond
this opposition also;
seeks to resolve not only the differences
but also those between spatial and temporal determina tions into the unity of numerical determinations. The particularity of each "event" is expressed by the four numbers Xi, x 2 xs, x 4
of sensation
, ,
whereby these numbers among themselves have reference to no differences, so that some of them Xi, x 2 x 3 cannot be brought into a special group of coordinates and contrasted with the time coordinate" x 4 Thus all differences belonging to spatial and temporal apprehension in subjective consciousness seem to be con
, ,
"spatial"
sistently set aside in the
same way that nothing
of the subjective

visual sensation enters into the physical concept of light only are all spatial and temporal values
>[ot
449
and
color. 1
exchangeable with each
inner differences of the temporal itself, unavoidable or the subjective consciousness, all differences of direction, which ,ve designate by the words "past" and "future," are cancelled. The
,ther,
all
but
direction into the past
and that into the future are distinguished
Trom each other in this form of the concept of the world by nothing more than are the + and directions in space, which we can deter
mine by arbitrary definition. There remains only the "absolute world" of Minkowski; the world of physics changes from a process
world, in which time
in a three-dimensional world into a being in this four-dimensional is replaced as a variable magnitude by the
"ray
imaginary
of
light"
(Lichtweg) (x 4
V -i
c t)-
This transformation of the time-value into an imaginary numerical value seems to annihilate all the "reality" and qualitative determinateness, which time possesses as the "form of the inner sense," as the form of immediate experience. The "stream of process," which, psychologically, constitutes consciousness and distinguishes it as
has passed into the absolute rigidity of a mathe There remains in this formula nothing of that form of time, which belongs to all our experience as such and
such, stands
still; it
matical cosmic formula.
enters as
an inseparable and necessary
factor into all
its
content. 3
experience,
But, paradoxical as this result seems from the standpoint of this it expresses only the course of mathematical and physical
objectification, for, to estimate it correctly
standpoint,
process,
we must understand
it
not in
its
from the epistemological mere result, but as a
a method.
In the resolution of subjectively experienced
qualities into pure objective numerical determinations, mathematical physics is bound to no fixed limit. It must go its way to the end;
it
can stop before no form of consciousness no matter
how
original
fundamental; for it is precisely its specific cognitive task to translate everything enumerable into pure number, all quality into quantity, all particular forms into a universal order and it only "con
arid
ceives"
them
scientifically
by virtue
of this transformation.
Philoso
phy would seek

1 2
in vain to bid this tendency halt at
any point and to
declare ne plus ultra.
The task of philosophy must rather be limited to

cf.
On thia
Cf.
latter point
Minkowski
(54, p.
(14,
62
Cf., e. g., J.
Cohn
Planck, Das Wesen des Lichts (71). Einstein (18, p. 82 f.). p. 228 ff.).
now
ff.);
450
recognizing fully the logical meaning of the mathematical and physical
concept of objectivity and thereby conceiving this meaning in its logi All particular physical theories including the theory cal limitedness.
and import only through the which stands back of them. The moment that we transcend the field of physics and change not the means but the very goal of knowledge, all particular concepts assume a new aspect and form. Each of these concepts means something different, depending on the general "modality" of consciousness and knowledge with which it is connected and from which it is considered. Myth and scientific knowledge, the logical and the aesthetic con sciousness, are examples of such diverse modalities. Occasionally concepts of the same name, but by no means of the same meaning, meet us in these different fields. The conceptual relation, which
of relativity receive their definite meaning
unitary cognitive will of physics,
generally call "cause" and "effect" is not lacking to mythical thought, but here its meaning is specifically distinct from the meaning that it receives in scientific, and in particular, in mathematical and
we
In a similar way, all the fundamental concepts a characteristic intellectual change of meaning when we undergo trace them through the different fields of intellectual consideration.
physical thought.
Where the copy theory of knowledge seeks a simple identity, the functional theory of knowledge sees complete diversity, but, indeed, at the same time complete correlation of the individual forms. 4
If
we apply
it is
then
these considerations to the concepts of space and time, obvious what the transformation of these concepts in modern
physics means, in its philosophical import, and what it cannot mean. The content of physical deductions cannot, without falling into the logical error of a //era/foc-is" etf aXXo yevof be simply carried over into
the language of fields whose structure rests on a totally different principle. Thus, what space and time are as immediate con tents of experience and as they offer themselves to our psychologi
cal
and phenomenological analysis
is
unaffected
by the use we make
of
them
in the determination of the object, in the course of objective conceptual knowledge. The distance between these two types of
consideration and conception

4
is
only augmented by the theory of

:
1 am aware of the fragmentary character of these suggestions for their supplementation and more exact proof I must refer to some subsequent more exhaustive treatment. Cf. also the essay Goethe und die mathematische Physik (11).

relativity
451
is
and thus only made known more
distinctly,
but
not
first
produced by it. Rather it is clear that even to attain the first ele ments of mathematical and physical knowledge and of the mathe matical and physical object we assume that characteristic transfor mation of "subjective" phenomenal and of
space
"subjective"
customary transformation, between the "physiological" space of our sensation and presentation and the purely "metrical" space, which we make the basis of geometry. The latter rests on the assumption of the equivalence of all places and directions, while for the former the distinction of places and directions and the marking out of the one above the others is essential. The space of touch, like that of is
this opposition
vision,
leads, in its ultimate consequences, to the results of the general theory of From the standpoint of relativity. strict sensualism too, it is to admit this
phenomenal time, which
and inhomogeneous, while metrical Euclidean space is distinguished by the postulate of isotropism and homogeneity. Compared with "metrical" time, physiological time shows the same characteristic variations and differences of meaning; one must, as
anistropic
Mach
sensation of duration and the measuring tion of warmth and temperature. 5

6
himself urges, as clearly distinguish between the immediate number as between the sensa
(50, p. 331 ff., 415 ff.). If, with Schlick (79, p. 51ff.), one would the psychological space of sensation and presentation the space of intui tion, and contrast with it physical space as a conceptual construction, no objec tion could be made against this as a purely terminological determination; but
Mach
call
one must guard against confusing this use of the word "intuition" with the Kantian, which rests on entirely different presuppositions. When Schlick sees in the insight that objective physical time has just as little to do with the intuitive experience of duration as the three-dimensional order of objective space with optical or "haptical" extension, "the kernel of truth in the Kantian doctrine of the subjectivity of time and space," and when he on the other hand combats, on the basis of this distinction, the Kantian concept of "pure intuition," this rests on a psychological misunderstanding of the meaning of the Kantian concepts. The space and time of pure intuition are for Kantnever sensed or perceived space or time, but the "mathematical" space and time of Newton; they are themselves constructively generated, just as they form the presupposition and foundation of all further mathematical and physi cal construction. In Kant s thought, "pure intuition" plays the r61e of a definite fundamental method of obj edification; it coincides in no way with
"subjective,"
i.e.,
psychologically
experienceable
Kant speaks
of the subjectivity of space
time and space. When and time, we must never understand
452
This contrast between subjective, "phenomenal" space and time, on the one hand, and objective and mathematical space and time, on the other, comes to light with special distinctness, when one considers a property which seems at first glance to be common to them. Of both we are accustomed to predicate the property of continuity, but we understand thereby, more closely considered, something wholly different in the two cases. The continuity, which we ascribe to time and processes in it on the basis of the form of our experience, and that which we define in mathematical concepts by certain constructive methods of analysis, not only do not coin cide but differ in their essential moments and conditions. The
experiential continuity affirms that each temporal content is given to us only in the way of certain characteristic tyholes," which can not
be resolved into ultimate simple "ekments;" analytic continuity demands reduction to such elements. The first takes time and duration as "organic" unities in which according to the Aristotelian definition, "the whole precedes the parts;" the second sees in them only an infinite totality of parts, of particular sharply differentiated In the one case, the continuity of becoming signifies temporal points. that living flux, that is given to our consciousness only as a flux, as a transition, but not as separated and broken up into discrete parts; in the other, it is demanded that we continue our analysis beyond all limits of empirical apprehension; it is demanded that we do not allow the division of elements to cease where sensuous perception, which is bound to definite but accidental limits in its capacity for discrimination, allows it to end, but that we follow it purely intellectually adinfinitum. What the mathematician calls the "continuum" is thus never the purely experiential quality of "continuity," of which there is no longer possible any further "objective" definition, but it is a purely conceptual construction, which he puts in the place of the latter. Here too he must follow his universal method he must reduce the quality of continuity to mere number, i.e., precisely to the fundamental form of all intellectual discreteness. (Cf. 6, p. 21). The only continuum he knows and the one to which he reduces all others, is always the continuum of real numbers which modern analysis and
;
experiential subjectivity but their "transcendental" subjectivity as condi tions of the possibility of "objective," i.e., of objectifying empirical knowl
edge.
19, 39).
(Cf. also
the significant remarks of Selliens against Schlick; 81, p.
453
theory of groups seek, as is known, to construct strictly conceptually with renunciation in principle of any appeal to the "intuition" of
The continuum thus considered, as Henri Poincare especially has urged with all emphasises nothing but a totality of individuals,, which are conceived in a definite order and are
given
space and time.
opposed to the others as something separate and external. We are here no longer concerned with the ordinary view, according to which there exists between the elements a sort of "inner bond" by which they are connected into a whole, so that, e.g., the point does not precede the the famous formula, that the line, but the line the point.
is
"Of
indeed in infinite number, of which each one
continuum
remains
disappeared. The analysts are nevertheless right when they define continuity as they do, for in all their inferences they are concerned, in so far as they claim
the unity of the only the manifold,

is
manifold,"
concludes Poincare,
"there
the unity has
But this circum rigor, only with this concept of the continuous. stance suffices to make us attentive to the fact that the true mathe matical continuum is something totally different from that of the
and the metaphysician." (72, p. 30.) In so far as physics an objectifying science working with the conceptual instruments of mathematics, the physical continuum is conceived by it as related to and exactly correlated with the mathematical continuum of pure numbers. But the "metaphysical" continuum of the pure and original "subjective" form of experience can never be comprehended in this way, for the very direction of mathematical consideration is
physicist
is
such that, instead of leading to this form, it continually leads away from it. The critical theory of knowledge, which does not have to select from among the different sorts of knowledge, but merely to establish what each of them and means, can make no normative
"is"
decision as to the opposite aspects under which the continuum here appears, but its task consists in defining the two with reference to
each other in utmost distinctness and clarity. Only by such a delimitation can be reached, on the one hand, the goal of phenomenological analysis of the temporal and spatial consciousness, and, on
the other hand, the goal of the exact foundation of mathematical "With regard to theanalysis and its concepts of space and time.
objection, "says a modern mathematical author in concluding his inves tigation of the continuum, "that nothing is contained in the intuition of the continuum of the logical principles that we must adduce in the
454
exact definition of the concept of the real number, we have taken account of the fact that what can be found in the intuitive continuum
and
mathematical world of concepts are so alien to each other, that the two coincide must be rejected as absurd. In spite of this, those abstract schemata, which mathematics offers us, are helpful in rendering possible an exact science of fields of The exact temporal or spatial objects in which continua play a role.
in the
that the
demand
point does not lie in the given ("phenomenal) duration or extension as an ultimate indivisible element, but only reason reaching through this can grasp these ideas and they crystallize into full determinateness only in connection with the purely formal arithmetical and 6 analytical concept of the real number."
If
we bear in mind
this state of affairs, the deductions of the theory
of relativity in its determination of the four-dimensional space
and
time continuum lose the appearance of paradox, for it is seen that they are only the final consequence and working out of the fundamen tal methodic idea on which rests mathematical analysis in general. But the question as to which of the two forms of space and time, the psychological or the physical, the space and time of immediate
experience or of mediate conception and knowledge, expresses the true reality has lost fundamentally for us all definite meaning. In
the complex that
ego and of moments.

exclude
it
call our "world," that we call the being of our the two enter as equally unavoidable and necessary things, We can cancel neither of them in favor of the other and
we
from
this complex,
If
but we can refer each to
its definite
the physicist, whose problem consists in affirms the superiority of "objective" space and time objectification, over "subjective" space and time; if the psychologist and the meta
place in the whole.
are directed upon the totality and immediacy of draw the opposite conclusion; then the two judgments experience express only a false "absolutization" of the norm of knowledge by which each of them determines and measures In which direction this "absolutization" takes place and whether it is directed on the or the is a matter of indifference from the standpoint of pure epistemology. For Newton it was certain that the absolute and mathematical time, which by its nature flowed time of which all empirically given temuniformly, was the
physician,
"reality."
"outer"
"inner"
"true"
who
Weyl,
84, p. 83, 71.
455
poral determination can offer us only a more or less imperfect copy; time of Newton is a conceptual fiction and abstraction, a barrier, which intervenes between our apprehension and the original meaning and import of reality. But it is forgotten that what is here called absolute reality, duree reelle, is itself no abso
for Bergson, this
"true"
lute but only signifies a standpoint of consciousness opposed to that of mathematics and physics. In the one case, we seek to gain a unitary and exact measure for all objective process, in the other we
are concerned in retaining this process itself in its pure qualitative character, in its concrete fullness and subjective inwardness and The two standpoints can be understood in their "contentuality." meaning and necessity; neither suffices to include the actual whole
The symbols that "being for us." the mathematician and physicist take as a basis in their view of the outer and the psychologist in his view of the inner, must both be understood as symbols. Until this has come about the true philo
ence
sophical view, the view of the whole, is not reached, but a partial experi is hypostasized into the whole. From the standpoint of mathe
of being in the idealistic sense of
matical physics, the total content of the immediate qualities, not only the differences of sensation, but those of spatial and temporal consciousness, is threatened with complete annihilation; for the
metaphysical psychologist, conversely, all reality is reduced to this immediacy, while every mediate conceptual cognition is given only the value of an arbitrary convention produced for the purposes of our action. But both views prove, in their absoluteness, rather perver sions of the full import of being, i.e., of the full import of the forms
of knowledge of the self and the world. While the mathematician and the mathematical physicist stand in danger of permitting the real world to be identified with the world of their measures, the meta physical view, in seeking to narrow mathematics to practical goals, It violently loses the sense of its purest and deepest ideal import.
meaning and the
closes the door against what, according to Plato, constitutes the real real value of mathematics; that, namely, "by each
of these cognitions
an organ of the soul is purified and strengthened, which under other occupations is lost and blinded; for its preserva tion is more important than that of a thousand eyes; for by this alone is the truth seen." And between the two poles of consideration, which we find here, there stand the manifold concepts of truth of the
concrete sciences
different
and therewith
their concepts of space
456
EINSTEIN
and time. History, to set up its temporal measure, cannot do with out the methods of the objectifying sciences: chronology is founded on astronomy and through this on mathematics. But the time of the
historian
is
nevertheless not identical with that of the mathematician
and
form.
but possesses in contrast to it a peculiar concrete In the concept of time of history, the "objective" content of knowledge and the "subjective" experiential content enter into a
physicist,
new
characteristic
reciprocal
relation.
An
analogous relation
is
presented, the forms of space and time. Painting presupposes the objective laws of perspective, architecture the laws of statics, but the two
when we survey the
aesthetic
meaning and shaping
of
serve here only as material out of which develops the unity of the picture and of the architectural spatial form, on the basis of the
For music, too, the Pythagoreans original artistic laws of form. sought a connection with pure mathematics, with pure number; but the unity and rythmical division of a melody rests on wholly differ ent structural principles than those on which we construct time in the sense of the unity of objective physical processes of nature.
What
space and time truly are in the philosophical sense would be determined if we succeeded in surveying completely this wealth of nuances of intellectual meaning and in assuring ourselves of the under The lying formal law under which they stand and which they obey. theory of relativity cannot claim to bring this philosophical problem to its solution for, by its development and scientific tendency from the
;
beginning,
of space
it is
and time.
limited to a definite particular motive of the concepts As a physical theory it merely develops the
in the
sense, final
meaning that space and time possess and physical measurements. In this
system of our empirical judgment on it be
In the course of its history, physics longs exclusively to physics. will have to decide whether the world-picture of the theory of relativ
ity is securely
experimental verification. anticipate; but even now

nles of physics.
founded theoretically and whether it finds complete Its decision on this, epistemology cannot it can thankfully receive the new impe
tus which this theory has given the general doctrine of the princi-
BIBLIOGRAPHY
457
(1)
BAUCH,
BRUNO: Studien zur Philosophic der exakten Wissenschaften. Heidelberg, 1911.

Leipzig, 1906.
(2)
BECKER, ERICH: Philosophische Voraussetzungen der exakten Naturwissenshaften.
OTTO: Das Relativitdtsprinzip der Elektrodynamik. (Abt. der Friesschen Schule, N. F., Gottingen, 1912.) (3a) BLOCK, W.: Einfuhrung in die Relativitdtstheorie, Lpz. und Berl., 1918. (A us Natur. u. Geisteswelt, Bd. 618.) (4) BUEK, OTTO: Mich. Faradays System der Natur und seine begriffiichen Grundlagen. Philos. Abhandlungen zu H. Cohens 70. Geburtstag.
(3)
BERG,
Berlin, 1912.
(5)
(6)
CASSIRER, ERNST: Leibniz System in seinen mssenschaftlichen Grund lagen. Marburg, 1902. CASSIRER, ERNST: Kant und die moderne Mathematik. Kant-Studien
XII
(7)
(1901).
CASSIRER, ERNST: Das Erkenntnisproblem in der Philosophic und Wissenschaft der neueren Zeit. Bd. I und II, 2. Aufl., Berlin, 1911; Bd. Ill,
Berlin, 1920.
(8)
(9)
CASSIRER, ERNST: Substance and Function. CASSIRER, ERNST: Erkenntnistheorie nebst den Grenzfragen der Logik. In Jahrb. fur Philosophic, hg. von Max Frischeisen-Kohler, Berlin;
1913.
(10) (11)
CASSIRER, ERNST: Kants Leben und Lehre, Berlin, 1918. CASSIRER, ERNST: Goethe und die mathematische Physik
Gestalt,
(In: Idee
und
Funf Aufsdtze,
COHEN, HERMANN: COHEN, HERMANN: Logik der reinen Erkenntnis, Berlin, 1902. (13a) COHN, EMIL: Physikalisches uber Raum und Zeit, 4- Aufl., Berlin, (14) COHN, JONAS: Relativitdt und Idealismus, Kant-Studien (XXI),
(12) (13)
Berlin, 1920). Kants Theorie der Erfahrung, 3. Aufl., Berlin, 1918.

1920.
1910,
p. 222
(15)
ff.
DUHEM, PIERRE: La
1906.
Theorie Physique, son objet
et
sa structure.
Paris,
(15a)
(16)
EHRENFEST,
P.:
Zur Krise
der Lichtaether-Hypothese, Berlin, 1913.
EINSTEIN, ALBERT: Zur Elektrodynamik bewegter Systeme. Physik, 4- F., XVII, p. 891 (1905), (Cited from No. 47)
.
Annalen der
(16a) EINSTEIN,
(17)
die Trdgheit eines Korpers von seinem EnerAnnal. der Physik (17), 1905. EINSTEIN, ALBERT: Die Grundlagen der allgemeinen Relativitatstheorie,
ALBERT:
1st
giegehalt abhdngig?
Lpz., 1916.
(18)
EINSTEIN,
theorie
(19)
(20)
ALBERT: Uber die spezielle und die allgemeine Relativitats (Sammlung Vieweg, Heft 38) 2. Aufl., Braunschweig, 1917. Relativi EINSTEIN, ALBERT: Die formalen Grundlagen der allgemeinen tatstheorie. Sitzungsber. der Berliner A kad. d. Wiss., XLI, 1916. EINSTEIN, ALBERT: Ernst Mach, Physikalische Zeitschrift, XVII (1916),
p. 101
ff.
458
(21)

ERDMANN, BENNO: Die Axiome der Geometric. Eine philosophische Untersuchung der Riemann-H elmholtzschen Raurntheorie (1877). EULER, LEONHARD: Reflexions sur I espace et le temps. Hist^. de I Acad.
des Sciences
et
(22)
Belles Lettres, Berlin, 1748.

: .
(23)
(24)
EULER, LEONHARD Briefe an eine deutsche Prinzessin (1768) FREUNDLICH, ERWIN: Die Grundlagen der Einsteinschen Gravitationstheorie.
Berlin, 1916.
(24a)
FREUNDLICH, ERWIN: Die Entwicklung des physikalischen Weltbildes

bis zur allgemeinen Relativitdtstheorie,
(25)
FRISCHEISEN-KOHLER, MAX: Wissenschaft

Berlin, 1912.
Weisse Blatter, 1920, p. 174 ff u. Wirklichkeit, Leipzig and

.
(26)
FRISCHEISEN-KOHLER, MAX: Das Zeitproblem.

phic
[s.
In: Jahrb. f.
Philoso
Berlin, 1913. (27) GALILEI: Opere, ed. Alberi, Firenze, 1843 ff. (28) GORLAD, ALBERT: Aristoteles und Kant beziiglich der Idee der theoretischen Erkenntnis untersucht. [Philos. Arbeiten, hg. von H. Cohen u.
9]
.
NT.
P. Natorp, Band II.] Giessen, 1909. HELMHOLTZ, HERMANN: Uber die Erhaltung der Kraft (1847). Ostwalds Klassiker der exakten Wissenschaft, H. 1, 1889). (30) HELMHOLTZ, HERMANN: Handbuch der physiologischen Optik, 2. Aufi., Hamburg and Leipzig, 1896. (30a) HELMHOLTZ, HERMANN Uber den Ur sprung und die Bedeutung der geometrischen Axiome (1870) in: Vortrage und Reden, 4- Aufl., Braun
(29)
:
(31)
(32)
schweig, 1896, Bd. II. HERTZ, HEINRICH: Die Prinzipien der Mechanik, Lpz., 1904. HONIGSWALD, RICHARD: Uber den Unterschied und die Beziehungen der logischen und der erkenntnistheoretischen Elemente in der kritischen
Philosophic der Geometric (Verh. des III. internal. Kongr. fur Philos.
Heidelb., 1908.
(33)
HONIGSWALD, RICHARD:
Zum
Streit uber die
Grundlagen der Mathematik.

2,
Heidelb., 1912. (34) KANT: Kritik der reinen Vernunft (cited from Ed.
(35)
1787).
herausgegeb. von Ernst Cassirer, Bd. I-X, Berlin, 1911 ff. (36) KEILL: Introductio ad veram Physicam, Oxford, 1702. (37) KEPLER: Opera omnia, ed. Frisch, Vol. I-VIII, Frankf. and Erlangen,
KANT: Werke,
1858
(38)
ff
(39)
(40)
KNESER, ADOLF: Mathematik und Natur. Rede. Breslau, 1911. KONIG, EDMUND: Kant und die Natururissenschaft, Braunschweig, 1907. LAUE, MAX: Das Relativitdtsprinzip (Die Wissenschaft, H. 38), Braun
schweig, 1911.
(41)
LAUE, MAX: Das Relativitdtsprinzip.

Berlin, 1913.
In: Jahrb. f. Philosophic

J.
(s.
NT.
9),
(42)
LEIBNIZ: Mathematische Schriften, Bd. I-VII, hg. von C.

Berlin, 1849
ff.
Gerhardt,
Gerhardt.
(43)
LEIBNIZ: Philosophische Schriften, Bd. I-VII, hg. von C.

Berlin, 1875
ff.
J.
BIBLIOGRAPHY
(44)
459
iibersetzt
LEIBNIZ: Hauptschriften zur Grundlegung der Philosophic, A. Buchenau, hg. von E. Cassirer, Bd. I, II,
Lpz.,
Bibliothek).
von
1904/06.
(Philos.
LEIBNIZ: Neue Versuche uber den menschlichen Verstand, ubers. u. herausgeg, von E. Cassirer, Lpz., 1915 (Philos. Bibliothek). (45a) LENARD, PHILIPP: Uber Ather und Materie, 2. Aufl., Heidelberg, 1911. Uber Relativitatsprinzip, Ather, Gravitation, Lpz., 1918.
(45)
(46)
LORENTZ, H. A.: Electromagnetic phenomena in a system moving with any

velocity smaller than that of light.
Proceed. Acad. Sc. Amsterd. 6 (1904) (Cited from No. 47). (46a) LORENTZ, H. A.: Dos Relativitatsprinzip, drei Vorlesungen, gehalten in Teylers Stiftung zu Haarlem. Lpz. and Berl., 1914.
(47)
LORENTZ-EINSTEIN-MINKOWSKI Das Relativitatsprinzip. Eine Sammlung von Abhandlungen. Mil Anmerk. von A. Sommerfeld und Vorwort
:
von 0. Blumenthal. ( Fortschritte der mathemat. Leipzig and Berlin, 1913.)

(48)
Wissensch., Heft
v.
2,
LOB ATSCHEFSKY Zwei

:
geometrische Abhandlungen, ubers.
Fr.
Engel,
Lpz., 1898.
(49)
(50)
MACH, ERNST: Die Prinzipien der Warmelehre, Lpz. ,1896. MACH, ERNST: Erkenntnis und Irrtum, Skizzen zur Psychologic
der
Forschung, Lpz., 1905. (51) MAXWELL, J. C.: Substanz und Bewegung (Matter and Motion), dtsch. von E. v. Fleischel, Braunschw., 1881.
(52)
MAYER, ROBERT: Die Mechanik

von Weyrauch,
8. Aufl.,
der
Warme
in gesammelten Schriften, hg.
Stuttgart, 1893.
(53)
(54)
(55)
(56)
(57)
MEYERSON, EMILE: Identite et realite, Paris, 1908. MINKOWSKI, HERMANN Raum und Zeit, Vortrag, Coin, 1908 (Cited from No. 47). MtiLLER, ALOYS: Das Problem des absoluten Raumes und seine Beziehung zum allgemeinen Raumproblem. Braunschweig, 1911. NATORP, PAUL: Die logischen Grundlagen der exakten Wissenschaften. Lpz. and Berl., 1910. NEUMANN, CARL: Uber die Prinzipien der Galilei-N ewtonschen Theorie,
:
Lpz., 1870.
(58)
(59)
NEWTON, ISAAC: Optice, lat reddid. S. Clarke, Lausanne and Geneva, 1740. NEWTON, ISAAC: Philosophiae naturalis principia mathematica (1686).
Ausg. von Le Seur und Jacquier, 4. Bd., Genf, 1739 ff. OSTWALD, WILHELM: Vorlesungen uber Natur philosophic, Leipzig, 1902. PETZOLDT, Jos.: Die Relativitatstheorie im erkenntnistheoretischen Zu-
(60)
(61)
sammenhang
Gesellschaft
(62)
des relativistischen Positivismus.

1912.
Verh. der Dtsch. Physik.
XIV., Braunschweig,
:
PFLUGER, A.
Das Einsteinsche
Bonn, 1920.
Relativitatsprinzip, gemeinverstandlich
dargestellt, 3. Aufl.,
(63) (64)
PLANCK, MAX: Das Prinzip der Erhaltung der Energie, Leipzig, 1887. PLANCK, MAX: Zur Dynamik bewegter Systeme. Annalen der Physik,
4. F.,
XXVI,
1 ff.
(1908).
460
(65)

PLANCK, MAX: Bemerkungen zum Prinzip der Aktion und Reaktion in derallg. Dynamik. Physikal. Zeitschrift IX (1908). PLANCK, MAX: Die Einheit des physikalischen Weltbildes, Leipzig, 1909. PLANCK, MAX: Die Stellung der neuen Physik zur mechanischen Welt
anschauung.
1910,
(66) (67)
(Verh. der Ges. dtsch. Naturf. u. Arzte in Konigsberg, Leipzig, 1911, S. 58-75.)
(68)
(69)
PLANCK, MAX: Acht Vorlesungen uber theoretische Physik, Leipzig, 1910. PLANCK, MAX: Zur Machschen Theorie der physikalischen Erkenntnis.
Vierteljahresschr.fur wissensch. Philos., 1910.
(70)
PLANCK, MAX: Neue Bahnen der physikalischen Erkenntnis (Rektoratsrede), Berlin, 1913.
(71)
(72)
(73)
(74) (75) (76)
PLANCK, MAX: Das Wesen der Lichts, Vortrag, Berlin, 1920. POINCARE, HENRI: La Science et I hypothese, Paris. POINCARE, HENRI: Der Wert der Wissenschaft, dtsch. v. E.Weber, mil. Anmerk und Zusdtzen von H. Weber, Berlin and Leipzig, 1906. POINCARE, HENRI: Science et methode, Paris, 1908. POINCARE, LUCIEN: Die moderne Physik, dtsch. von Brahn, Leipzig, 1908. RIEHL, AL: Der philosophische Kritizismus und seine Bedeutung fur die
(77)
RIEMANN, B: Uber
(1854).
positive Wissenschaft I/II, Leipzig, 1876-97. die Hypothesen, welche der Geometrie
zugrunde liegen
(78)
(79)
RIGHI: Die moderne Theorie der physikalischen Erscheinungen, dtsch. von Dessau, Leipzig, 1908. SCHLICK, MORTIZ Raum und Zeit in der gegenwdrtigen Physik, Berlin,
:
1917.
(80) (81)
SCHLICK, MORITZ: Allgemeine Erkenntnislehre, Berlin, 1918. SELLIEN, EWALD: Die erkenntnistheoretische Bedeutung der
tdtstheorie, Kieler Inaug.-Diss., Berlin, 1919.
Relativi-
(82)
STREINTZ,
1883.
H..:
Die physikalischen Grundlagen der Mechani k, Leipzig,

Zeit.
(83)
WEYL, HERMANN: Raum.
Materie.
Vorlesungen uber allgemeine
Relativitdtstheorie, 3. Auft., Berlin, 1920.
Kritische Untersuchungen uber die Grundlagen der Analysis. Leipzig, 1918. The work of Dr. Hans Reichenbach on the meaning of the theory of rela tivity for the concept of physical knowledge (Die Bedeutung der Relativitdts theorie fur den physikalischen Erkenntnisbegriff) became accessible to me in manuscript while this essay was being printed. I can here only refer to this thorough and penetrating work, which has much in common with the present essay in its way of stating the problem; I cannot, however, completely agree with its results, especially with regard to the relation of the theory of relativity to the Kantian critique of cognition.
(84)
WEYL, HERMANN: Das Kontinuum.
INDEX
ABSTRACTION, 6,
ioff.,
Activity of ego, 314. Acts, logical, 25.
Analogy and induction,

Analytic geometry, Anaxagoras, 152f.
72f.
250ff.
Chemistry, types in, 212; construc tion of concepts in, 203-220; and mathematics, 217. Class and number, 44ff.
Anaximander,
151f.
priori, 246, 269f. Apollonius, 70.
Cohen, H., 99, 355, 368. Cohn, 323. Compositive method, 252. Concept, 3ff., 9, 19, 53, 68,
351ff.
113, 221,
Aristotle, 134, 220, 329.
Conceptual foundations of relativity,

367.
Aristotelian logic,
3ff.
Arithmetic;
27ff.
empiristic theory
of, 28; and psychology, 32. Association, 285. Atom, 159ff., 188; as a relational con
Consciousness of ego, 295. Conservation of energy, 157. Constant relations, reduction

Constants, 230, 265. Continuity, 79, 158, 452f.
to, 260.
cept, 207, regulative use of, 209f.
Atomic weight, criteria of, 206. Atomism and number, 155ff.

cient, 395.
Continuum,
;
248.
an
Critical idealism, 295, 297.
Cusanus, 360.
Ausdehnungslehre, 96 ff.
DALTON, 2oef.
BACON,
147, 154.
Dedekind,
Democritus,
Descartes,
35f., 38, 58, 60.
Being, a product of thought? 298.
Definite proportions, law of, 204f.

167ff., 274, 373, 395.
Beltrami, 110. Bergson, 455.

Berkeley, 9ff. Bernard, C., 250f.
Berzelius, 219. Body Alpha, 179-181.
70f., 76, 95, 395-398.
Description and explanation, 263.
Dewey,
318f.
Differential equations, 159ff.
Drobisch, 21.
Boltzrnann, 159f. Bolzano, 312.
du Bois-Reymond,
122ff., 129.
Duhem,
138,
143,
175,
200f.,
216,
Boscovich, 159. Bouasse, 257.
252, 280. Dumas, 212.
CANTOR,
MS.,
65.
46ff.
EBBINGHAUS,
335.
Cardinal number, 41, Carnot, 403.

Categories, 269.
Einstein, 354, 356, 371, 375, 377ff.,

383, 397, 401f., 419, 422, 428.
407,
412,
414,
Causality, 226n, 249.
Eleatics, 167.
Cayley, 86.
Characteristic, 92.
461
Empirical foundations of relativity,

367ff.
462
Empiricism, 128f., 340. Energism, 192f., 199-^03. Energy, 170, 188-192, 197f., 359. Eotvos, 402. Ephodion, 69. Equivalence, 197.
INDEX
HAMILTON,
w., 40.
301.
Hartmann, Helm, 198.
Helmholtz, 41, 288, 304f., 388, 438. Hertz, H., 170, 184f., 198, 394f., 399f. Hessen, 226.
Hilbert, 93.
Erdmann,
23.
Eternity, in empirical judgment, 242.
Hume,
331.
Ether, 162f., 397ff., 404ff. Ethics and aesthetics, 232.
Husserl, 24.
Hypothesis,
concept
of,
135-138,
Euclidean
and
430ff.
Non-Euclidean
140f., 146ff.
Geometry,
Existence,
Euler, 352ff., 383, 404, 411.

123ff.
IDEALISM,
129f.
123, 125.
Ideal and reality,
relation
of,
Experiment, 254.
Experience, 246, 268, 270, 273, 278,
284.
Identity and diversity, 324.

Illusions, 275. Imaginary in geometry, 82f.
JTACT and
Fichte, 413.
theory, 143f.
397.
Induction, 237ff.
Inertia,
176f., 364; principle of, Galileo s conception of, 168f.
;
Faraday, Fechner, 159, 301.

Field, concept of, 396f., 405.
and body alpha,

183.
177ff
and gravi
tation, 401ff., 428.
Fixed stars,
Fizeau, 369.
Inexhaustibility of individual, 232. Infinitesimal geometry, 73.

310.
Form (Aristotelian), 7. Form and matter of knowledge,

Form-quality, 311. Frege, 28f., 43, 45f., 52. Freytag, 299. Functions, categorical, 16. Functions and magnitude,
Infinity in judgment, 247. Intellectual experiment, 176f.
Intension and extension, 5. Intuition and thought in geometry of

position, 76.
Invariants,
75.
and induction, 249; of experience, 269; consonant with

relativity, 374, 379.
GALILEO,
139,
156, 168,
175, 254,
Isolation and superposition, 256.
36 If., 373.
Gauss, 55, 406. Gerhardt, 215. Generic image,
KANT,
15.
;
40,
ioe, 210, 245, 351, 355,
380, 386, 392, 394, 399, 403, 415ff.,
Geometry,
68ff.
ancient method
in,
425, 434, 439, 451n, on relativity of motion, 410; on space and
69; and theory of groups, 88; constancy and change in, 90;
time, 410ff., 416.
Kaufmann,
400.
135f., 251, 258.
as doctrine of relations, 95; Eucli dean, 398^07; Non-Euclidean, f ruitf ulness of, 430ff not a copy
.
Kepler, 116, 118, Keyser, 313.

Kirchhoff, 119. Klein, 86, 90.
of real world, 433.
Goethe, 243, 371, 375, 445.
Knowledge, problem
thought,
131ff.
of,
in
ancient
Grassmann,
96.
INDEX
Kronecker, 41.
Michelson, 369, 375. Mie, 397.
Mill, J. S., lOff., 28, 34.
463
Kummer,
117.
LAMBERT,
19.
Lask, 230. Laue, 376, 406, 414. Laurent, 215.
Minkowski, 376, 419, Motion, 361f.
424f.
Muller, J., 287ff. Multiple proportion, law
of, 206.
Law, and
rules, 258;
meaning
of, 263.
Laws
of nature, 265, 383. Leibniz, 43, 77, 91f., 95, 184, 186, 201, 262, 312, 329, 363, 376, 384, 391f.,
398, 419.
NATIVISM,
Neumann,
Newton,
340.
Nature, 264, 275, 395. Natural science and reality, 220.

C., 179f., 364, 382.
136ff., 170f., 258, 268, 352ff.,
Liebig, 215.
Light, uniform velocity
of, 369, 422.
396, 398f., 402.
Limiting concept,
Lipps, 342.
122f., 125.
Non-being,
Null, 50.
167.
Lobatschefsky, 430. Locke, 217, 214f., 330.

Logic,
Number,
of, 3;
27ff.; logical
foundations
;
of,
35ff.; ordinal, 39f.
cardinal, 41,
new development
Aristo
46ff.;
telian, 4ff., 7; function of, 21; of relations, 36.
extensions of, 54ff.; imagi nary, 55; negative, 55; transfinite
62ff.;
second principle of genera
Lorentz, 375ff.
tion of, 66.
Lorentz-transformation, 372, 423. Lotze, 21.
QBJECTIVITY,
275;
290f.; in
272;
degrees
of,
and space,
286; circles of,
JVJACH,
174,
176f., 238f., 260, 264,
427f., 451.
Mass, 399f. Mathematics, symbols in, 43; con cepts of, 112, 116; and teleology,
133ff.,
mathematics, 298; func tion of, 304; physical, 356ff., 361. Object, 293; of first and second
orders, 23, 338; physical 164ff.; in critical idealism, 297; as pro
Plato
interpretation
223.
of,
jected unity, 321.

Observability, principle
of, 376f.
134;
and word meanings,

169f., 188, 397ff.
Matter,
Ostwald, 130,
311.
loif.
141, 205, 216, 243.
Matter and form of knowledge, Maxwell, 249, 364, 407. Mayer, R., 139, 197ff. Measurement, presuppositions
141f.; units of, 144f.
PASCH,
of,
Pearson, I20f. Perception, 245, 282, 342. Periodic system, 216.
Mechanism and motion,

Mechanics,
119ff.
;
117ff.
postulate of rela
Permanence, of mathematical tions, 80; and change, 322.
rela
tivity of, 370.
Phenomenology,
Physics,
logical
24.
Medicine, experimental, 250.
ideal
of,
114ff.
Meinong, 320, 337ff. Metageometry, 100, 443. Meyer, Lothar, 217. Meyerson, 324.
Aristotelian, 154; as science of qualities, 201; and philosophy,
419;
constants
in,
420;
posi-
tivistic, 427ff.
464
Plato,
131ff.,
INDEX
134f.,
312,
327,
368,
370, 455.
Serial order, 15, 26, 190, 148f., 259. Signs, 304.
Planck, 306f., 357, 360, 404, 445. Poincare, H., 108, 373, 431, 435, 453. Poincare L., 405.
,
Sigwart, 222.
Simultaneity, 422f.
Skepticism, 387ff.
Socrates, 133.
Poncelet, 78.
Power
(Machtigkeit), 62.
317ff.
Pragmatism,
Progression, 38.
Projection, theory of, 287, 292.
Soul (Platonic), 377. Source, logic of, 99. Space, 68ff.; and number, 70; and
order,
87;
mathematical
104;
and
108,
Projective metric, 86.

Propositions, 345.
sensuous,
Euclidean,
Psychologist
s fallacy, 332.
443; psychology of, 342; without ontological meaning, 439.
Psychology of relations,
326ff.
Space and time, 118, 170f., 397ff.; Kant s theory of, 106, 409ff., 414, 416, 423; as mathematical ideals,
181ff.;
{RADICAL,
Rankine,
chemical, 215. Radio-activity, 219.

193f., 195.
Newton
theory
of, 171f.,
352f.,
for physicist
and philos
Reality, 271ff.
opher, 417; transformation of values of 449f psychological and

.
Relations, 255, 310, 336, 398.
physical, 454f.
Stallo, 306.
Relativity of knowledge, 306. Relativity, theory of, presupposi

tions
of
Staudt, 85f.
Steiner, 77f.
Streintz, 178f.
relation
special theory, 377; of special to general

378f.;
theory,
and
functional
Structure
and function
in
knowl
theory of knowledge, 392; not Protagorean relativism, 393; and reality, 445ff resolves being into pure order of numbers, 448. Representation, 282, 284. Resolutive method, 252.
.
edge, 315. Stumpf, 288.
Suarez, 283.
Subjectivity and objectivity of rela tional concepts, 309ff.
Substance,
8, 151ff., 164, 330.
Richter, 205.
System
of reference, 172f.
Rickert, 221ff.
Riehl, 310.
,
concept
384.
of, 17,
149,
204,
Riemann,
440f.
302, 357f.
Russell, B., 46, 50, 88, 95, 316.
Thompson, W.,
gCHWARZ,
and
of
H., 283.
Science, natural, concepts of, 113ff.; as description, 114f., 119f.;

history, 231; transformation
Time, 145 (See Space). Thought, experiment, 238; and ex perience, 296; psychology of, 345. Transcendence, 278, 300. Transcendental philosophy, 268. Transcendental realism, 294.
Transubjective, 294, 297.
system of, 266. Sensuous qualities, subjectivity

274.
of,
Truth, 311, 319,
387ff., 391f.
Tycho de Brahe,
260.
INDEX
465
UNAMBIGUITY,
Unity. 50.
principle of, 204.
Voikeit, 2%ff.
IT-
Unity of nature, 373f ., 417. Universal and particular, 232. Universal laws, 416.
Volkmann, 2o6f
"YYÊYL,
398.
VALENCY,
211.
Wholes, discrete and continuous, 248.
Veronese, 101.
1433 34
BD 221 C373
cop. 2
Cassirer, Ernst Substance and function
PLEASE
DO NOT REMOVE
FROM
THIS
CARDS OR
SLIPS
POCKET
UNIVERSITY
OF TORONTO
LIBRARY

Cassirer Substance and Function Einsteins Relativity

Uploaded by

Copyright:

Available Formats

Cassirer Substance and Function Einsteins Relativity

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Cassirer Substance and Function Einsteins Relativity

Uploaded by

Copyright:

Available Formats

SUBSTANCE AND FUNCTION

SUBSTANCE AND FUNCTION

EINSTEIN S THEORY OF RELATIVITY

WILLIAM CURTIS SWABEY,

MARIE COLLINS SWABEY,

THE OPEN COURT PUBLISHING COMPANY

COPYRIGHT 1923 BY THE OPEN COURT PUBLISHING COMPANY CHICAGO

in United States of America

COMPOSED AND PRINTED AT THE

regarded in different logical perspectives.

through the most important

new meaning when we advance from purely

which are generally considered under

directly significant in judging the questions of fact &quot;Criticism of Knowledge,&quot;

The transformation which

ever wider and more concrete

fields into its circle.

in the English language

these requirements best of

will find here a construc

Knowledge and that perhaps these

guide-posts might help the student in finding his difficult material.

WILLIAM CURTIS SWABEY, MARIE COLLINS SWABEY.

THE CONCEPT OF THING AND THE CONCEPT OF RELATION

ON THE THEORY OF THE FORMATION OF CONCEPTS

THE CONCEPT OF NUMBER

sensationalistic deduction of number.

IV. Extension of the concept of number.

and imaginary numbers.

THE CONCEPT OP SPACE AND GEOMETRY

ontological &quot;hypotheses.&quot; IV. Robert Mayer s methodology of natural science.

Construction and con

Chemistry and mathematics IX. The concept of natural science and

THE SYSTEM OF RELATIONAL CONCEPTS AND THE PROBLEM OF REALITY

The metaphysical tendency

induction and deduction.

fundamental types of knowledge

Laws and constants.

ciple of explanation II. The concept of objectivity

division into circles of objectivity.

The concept of modern psy

THEORY OF RELATIVITY CONSIDERED FROM THE

Concepts of measure and concepts of things

The empirical and conceptual foundations

409 430 445 457

OF THE FORMATION OF CONCEPTS

The new view that

follow the evolution of science within the last decades

The appearance of unconditional now begins to be applied to those very

SUBSTANCE AND FUNCTION

criticism of the general doctrine of the construction of concepts

this doctrine are

some common property

THEORY OF THE FORMATION OF CONCEPTS

upon the same plane as the concepts

way that we ascend from

its content, this

magnitude increases as we descend from

Cf. e.g., Drobisch,

Uberweg, System der Logik, Bonn,

directly significant in judging the questions of fact "Criticism of Knowledge,"

ontological "hypotheses." IV. Robert Mayer s methodology of natural science.