1992 Lockeondefinition

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/244345372
The Deﬁnition of Deﬁnition in John Locke’s Essay Concerning Human

Understanding
Chapter · January 1992
CITATIONS READS
2 985
1 author:
Rüdiger Schreyer
RWTH Aachen University
28 PUBLICATIONS 61 CITATIONS
SEE PROFILE
All content following this page was uploaded by Rüdiger Schreyer on 16 May 2014.
The user has requested enhancement of the downloaded file.

Rüdiger Schreyer 1992 "The definition of Definition in LOCKE'S Essay concerning human
understanding". Lexicographica 8/1992 ed. F.M. Dolezal et.al. Tübingen: Niemeyer. 26-
51. (slightly revised and reformatted version)
THE DEFINITION OF DEFINITION IN LOCKE'S

ESSAY CONCERNING HUMAN UNDERSTANDING
By Rüdiger Schreyer
RWTH Aachen
1. Introduction.................................................................................... 2
2. The wider context............................................................................ 3
3. A KWIC view of Locke's use of the term definition ............................... 5
4. Words and Ideas ............................................................................. 6
5. The private and public uses of language ............................................. 8
5.1 Civil use .................................................................................... 9
5.2 The philosopher's dilemma............................................................ 10
6. Definition ..................................................................................... 11
6.1 What definition is good for ......................................................... 11
6.2 What definition is ...................................................................... 12
6.3 Two types of definition .............................................................. 12
6.4 Why simple ideas cannot be defined ............................................ 13
6.5 Why the definition of complex ideas is difficult.............................. 15
6.5.1 Defining mixed modes ......................................................... 16
6.5.2 Defining substances............................................................. 17
6.6 Definitions and trifling propositions ............................................. 19
7. Against a philosophical language ..................................................... 20
Appendix A: Selections from a KWIC Concordance ................................ 23
1. Definition of .... .......................................................................... 23
2. Signification of ... ....................................................................... 23
Literature ......................................................................................... 24
understanding". Lexicographica 8/1992 ed. F.M. Dolezal et.al. Tübingen: Niemeyer. 26-51.
2
(slightly revised and reformatted version)
But I desire it may be considered how difficult it is to lead another by words into the thoughts of
things, stripped of those special differences we give them; which things, if I name not, I say
nothing, and if I do name them, I thereby rank them into some sort or other and suggest to the
mind the usual abstract idea of that species, and so cross my purpose. (Locke III, vi, 43)
Abstract
In this paper I investigate the place of definition in LOCKE'S theory of language. John
LOCKE'S ESSAY CONCERNING HUMAN UNDERSTANDING (1699) had an enormous influence on
l8th century linguistic theory. There is reason to assume that his theory, and his
observations on definition in particular, had an Impact on Enlightenment lexicography.
Since LOCKE'S linguistic thought mainly revolves around the meaning of words I
reconstruct LOCKE'S link between words, ideas and things. Both words and ideas are
signs, and I show that LOCKE'S considers the relationship between sign and signified as
arbitrary and therefore fragile. This fragility is held responsible for the unreliability of
words and for much misunderstanding in philosophical and scientific discourse. LOCKE
believes that such communication problems can, to a certain extent, he remedied or
prevented by definition. LOCKE'S analysis of the nature, purpose, value and limits of
definition demonstrates that no definition is forever: The object of any definition is an
idea, and since ideas are necessarily changeable, definitions must change with them.
LOCKE'S theory of definition contains arguments sufficient to shatter the belief in the
possibility of a philosophical language and may have contributed to the abandonment of
universal language projects in the l8th century.
1. Introduction
The 18th century was the first great period of English lexicography. Its greatest
achievement was DR. JOHNSON'S DICTIONARY OF THE ENGLISH LANGUAGE (1755), compiled
with the express goal of fixing and ascertaining the English language. The motives
for his endeavour have been sufficiently discussed in the literature (SLEDD/KOLB
1955; STARNES/NOYES 1991). Let me just say that confidence and pride in the
efficiency of English as a tool for communication went hand in hand with the fear
that it had reached its peak and was on the verge of decline. The cry for an English
Academy, for a body to watch over and cultivate, purify and prune the language, by
such well-known writers as ROSCOMMON, DEFOE and SWIFT bears testimony to this
(FLASDIECK 1929). The academy was never established and Johnson's dictionary was
hailed as a substitute authority and compared - favourably, of course - with the
dictionaries produced by the Italian and French Academies. It was recognized then,
and is still recognized, that Johnson's main achievement lies in the definition of his
terms. BOSWELL (1964 I: 292) praises the definitions as "proofs of acuteness of
intellect and precision of language". Nevertheless, Johnson recognised that his very
definitions were open to criticism, not only because of his shortcomings as a
lexicographer, but also because of the ambiguity and vagueness of words on the one
hand and of the limits of definitions purely verbal on the other. In his preface to the
dictionary he discusses these difficulties, sometimes in detail.
That part of my work on which I expect malignity most frequently to fasten, is the
Explanation; in which I cannot hope to satisfy those, who are perhaps not inclined to be
pleased, since I have not always been able to satisfy myself. To interpret a language by
itself is very difficult; many words cannot be explained by synonimes, because the idea
signified has not more than one appellation; nor by paraphrase, because simple ideas
cannot be described. When the nature of things is unknown, or the notion unsettled and
3
indefinite, and various in various minds, the words by which such notions are conveyed,
or such things denoted, will be ambiguous and perplexed. And such is the fate of
hapless lexicography, that not only darkness, but light, impedes and distresses it;
things may be not only too little, but too much known to be happily illustrated. To
explain requires the use of terms less abstruse than that which is to be explained, and
such terms cannot always be found: for as nothing can be proved but by supposing
something intuitively known, and evident without proof, so nothing can be defined but
by the use of words too plain to admit a definition. (JOHNSON 1755: PREFACE)
These are valuable insights, but they hardly amount to a theory of definition.
Eighteenth century lexicological theory generally lags behind lexicographical practice.
Prefaces of Enlightenment dictionaries rarely allow more than a glimpse of the
linguistic views underlying the definition of dictionary entries. It is not that
lexicographers had no such views. It can be shown that Dr. Johnson, for instance,
had a rather coherent theory of language, which may be pieced together from the
scattered remarks in his works - a theory never made explicit, but consistent with his
practice of definition. This theory is, however, unoriginal and in the main derived
from JOHN LOCKE'S ESSAY CONCERNING HUMAN UNDERSTANDING (see SCHREYER 1971)1,
which is according to HOWELL (1971:277) "the most popular, most widely read, most
frequently reprinted, and most influential of all English books of the 18th century".
All the Enlightenment philosophers and linguists had read LOCKE and sometimes
discussed or criticised aspects of his Essay in their works. Samuel Johnson knew his
LOCKE very well and his discussion of definition, as far as it goes, is in close parallel
with LOCKE'S thinking: There is not a single idea in the passage quoted above that
cannot also be found in LOCKE'S Essay.
One might be tempted to speculate that LOCKE'S epoch-making work relieved

Johnson and other lexicographers of the need to ponder the theoretical foundations
of their practice. Be that as it may, a scrutiny of LOCKE'S views on the methods, uses
and limits of definition will not only throw light on this aspect of LOCKE'S linguistic
theory but may also contribute to a better understanding of the theoretical
foundations of 18th century lexicographical practice. The following inquiry is an
attempt to establish the place of definition within LOCKE'S theory of language.
2. The wider context
LOCKE'S Essay is an inquiry into "the original, certainty and extent of human
knowledge" (I, i, 2). At first blush, there is no apparent connection with language
and if we are to believe LOCKE himself the relevance of language to his subject
dawned on him only slowly. In his third book (III, ix, 21) he points out, as a belated
insight of his, that very little can be said about knowledge without considering "the
1
I suspect the same to be true of other 18th century dictionary makers.
4
force and manner of signification of words". As a result of this insight he devotes his
third book to language, or, more precisely, to words considered as signs of ideas.2
In LOCKE'S division of the sciences, the study of words, logic, is part of the study
of signs. He attaches so much importance to this study that at the very end of his
Essay (IV, xxi) - much to the delight of modern semioticians - he names semeiotikè,
the doctrine of signs, as one of the three principal branches of science. Semeiotikè
studies either the nature of ideas, i.e. the signs the mind uses for understanding the
nature of things, or it studies words, the signs used for conveying knowledge to
others. The study of signs is not, however, a goal in itself. Beside pure semiotics,
there is also an applied branch, the teaching of the "right use of signs in order to
knowledge".
LOCKE'S Essay can be read as an essay in semiotics, pure and applied. The
reason is simple. Human knowledge can only be about ideas, and ideas are signs;
they are the only objects directly accessible to the mind. Knowledge is "the
perception of the connexion and agreement, or disagreement and repugnancy, of
any of our ideas" (IV,i,2), mental truth the joining or separating of ideas (i.e. a
mental proposition) as they agree or disagree, verbal truth the joining or separating
of words, as their ideas agree or disagree (IV, v). It follows that words are not
necessary for the thought processes which produce knowledge, but they are
essential for its transmission: True mental propositions are made known by verbal
propositions. This seems rather straightforward. But there are difficulties:
1. Words "seem scarce separable from our general knowledge" (III, ix, 21).
2. Words "interpose themselves so much between our understandings and the truth
which it would contemplate and apprehend that, like the medium through which
visible objects pass, their obscurity and disorder does not seldom cast a mist
before our eyes and impose upon our understandings" (2:87-88).
This all too intimate connection between the medium and the message often
results in communication failures, arguments or worse. The causes of these
infelicities must be analysed and means be devised to overcome them.
This is the wider context in which LOCKE'S concern with definition must be seen.
Definition is the most important means to avoid or remedy communication failures
and thus to improve the communication of knowledge. There are at least three ways
of approaching LOCKE'S concept of definition: look at his statements about definition,
look at his use of the word definition or look at his definitions. Here I shall confine
myself to the first two options, beginning with LOCKE'S use of definition and related
terms. Then I take a closer look at LOCKE'S concept of idea and at his analysis of
communication and communication failures, followed by a scrutiny of LOCKE'S views
2
LOCKE'S main interest is in words; his third book, which contains most of his linguistic thought,
bears its title (Of Words) rightfully. The only larger unit deserving of LOCKE'S attention is the proposition.
This is the linguistic vehicle of knowledge. The term sentence occurs only five times and syntax is not
even mentioned.
5
of the purpose, structure, achievements, difficulties and limits of definition. Finally I

will show that in LOCKE'S theory of language the creation of a universal philosophical
language by strict definition of a standard vocabulary is theoretically impossible.
3. A KWIC view of Locke's use of the term definition
LOCKE'S ESSAY CONCERNING HUMAN UNDERSTANDING is available on magnetic disk

and formatted for use with the Oxford Concordance Program (OCP). A lexicographical
approach to our question is therefore not quite as time-consuming as it would have
been in Dr. Johnson's day. One aspect of LOCKE'S idiolect which lends itself to
computer-aided research is the vocabulary. An unsophisticated frequency list reveals
that the most frequent noun lexemes in LOCKE'S Essay are idea, mind, man, thing,
word, knowledge in that order. The lexeme definition occurs only 67 times in the
Essay, but, strikingly, 57 of these occurrences cluster in Book III. If we are
interested in definition, this is the book to study. However, for a first impression of
LOCKE'S use of the term definition a KWIC (keyword in context) concordance will
furnish sufficient information. Concordances may be sorted according to the context
on the right or the left of the keyword. A right-sorted concordance reveals that
definition/s of W is a frequent syntactic structure. Here some typical examples3:
(1) definition of the name man
(2) definition of the word man
(3) definition of that name
(4) definition of that word
(5) definition of a man
W frequently stands for a determiner plus name, word, term sometimes
followed by the definienda, of which man and gold enjoy the greatest popularity.
Rarely is name, word, term omitted. In this context the three terms seem to be in
free variation. The same is true of their occurrence with some form of the verb
define. For instance, LOCKE uses the phrases "the word/term/name defined"
indiscriminately.4 What is being defined, then, is the word or name or term. I shall
therefore merely look at LOCKE'S use of name. Name is a relational term: People have
or want names for ideas. The phrase names of ideas occurs with high frequency and
is expanded in various ways. The maximum construction can be represented by the
following formula:
(ADJ) (ADJ) (N)

general simple colours
proper names of complex ideas of substances
relative compound modes
abstract general species
3
For more examples see Appendix A.
4
On this interpretation example 5 is elliptical with the specifier word understood.
6
The relation between names and ideas marked in the above table by the
preposition of is specified by the verbal expressions stand for, signify, be
marks/signs of, belong to, be annexed to. In other words, names represent ideas
because men give, assign, apply or annex names to them. Men signify, mark, or
denote their ideas by them. They sort, rank, unite, or comprehend ideas under a
name which allows them to call or express them by that name. The idea is the
signification of that name.
In a right-sorted concordance the term signification occurs in the same contexts

as definition. However, modifiers on the left of our terms show them not to be
synonyms.5 The signification of words may be determined, determinate, distinct,
clear or confused, uncertain, doubtful; it can be received, usual, proper or new,
abstruse. Signification unlike a definition is never qualified as good or bad. Definition
can be paraphrased as the explanation of the signification of a word. It consists of
parts and/or terms jointly standing for (or representing) the idea meant by a
speaker.
This may suffice as a preliminary account of the use of definition, its cognates
and related terms. Ad hoc KWIC concordances and collocation lists indicate that
LOCKE is rather consistent in his use of key terms. Computer-aided semantic analyses
may well provide new insights into the meaning of such terms as meaning,
signification, resemble, represent, agree or conform, terms which have given a
headache to philosophers. Other problems arise which had better be handled by
reference to the full text. What, for example, is the purpose of definition? What is a
good or a bad definition? Which words cannot be defined and why? Here the
computer may still be of some help; it can retrieve information and point the student
to pertinent passages. Their interpretation will have to be done manually, as they
say in computerese. Clearly, the relation between the definition and the signification
of a word must be analysed further. The following section investigates the relation
between word and idea.
4. Words and Ideas
LOCKE distinguishes two types of words, proper names and general terms, and
he makes it quite clear that his interest is in the latter. If we examine the nature of
general names, says he (III, i, 6), "we shall the better come to find the right use of
words, the natural advantages and defects of language, and the remedies that ought
to be used, to avoid the inconveniences of obscurity or uncertainty in the
signification of words." Why are general terms so important? And why are proper
names comparatively unimportant?
5
Only once does LOCKE speak of "the received signification or definition of" a word.
7
Proper names are signs of a collection of ideas found to coexist at a given time
in a given place. Such ideas are ideas of a particular thing (individual). A language of
proper names is impossible, useless and uninformative. It is impossible, because the
number of individuals being infinite, the number of words would be infinite and could
not be memorised. It is useless, since different people being acquainted with
different individuals will use different words: Communication would fail. A language
of proper names is uninformative because knowledge is always knowledge of general
ideas and their relation. For this reason proper names are of little interest to LOCKE.
General terms/names are signs of general ideas.

Words are general ... when used for signs of general ideas, and so are applicable
indifferently to many particular things; and ideas are general when they are set up as
the representatives of many particular things: but universality belongs not to things
themselves, which are all of them particular in their existence, even those words and
ideas which in their signification are general. (LOCKE III, iii, 11)
The existence of general names results from the need to speak about things
"with dispatch". Men "do thereby enable themselves to consider things and discourse
of them, as it were in bundles, for the easier and readier improvement and
communication of their knowledge, which would advance but slowly were their words
and thoughts confined only to particulars" (III, iii, 20). This aspect of LOCKE'S
linguistic thought was emphasized by none more pointedly than JOHN HORNE TOOKE
(1798 I: 29-43), who considers "the whole of Mr. LOCKE'S Essay as a philosophical
account of the first sort of abbreviations in Language", the first sort being
abbreviations in terms. According to Tooke not enough writers had realized how
important speed (dispatch) was for communication and thus for the structure of
language. LOCKE had mistakenly explained abbreviation as the result of the
composition of ideas, when "it was merely a contrivance of language; and that the
only composition was in the terms..." (1798 I: 37), an interesting thought,
considering that artificial intelligence as well as linguistic semantics explain signs as
signs of signs ... of semantically primitive signs.
In LOCKE'S view general ideas make for brevity in thinking, general terms for
brevity in speaking. That "general thought ... would probably never have arisen
without the need for general words" (AYERS 1981:16) is a misinterpretation of
LOCKE'S views. For him the need to think and to speak about a certain field of
experience are both socially conditioned; languages "being suited only to the
convenience of communication, are proportioned to the notions men have, and the
commerce of thoughts familiar amongst them, and not to the reality or extent of
things ..." (II, xxviii, 2). Therefore "it is no wonder men should have framed no
names for those things they found no occasion to discourse of" (II, xxviii, 2).
Nonetheless, general terms make up the largest part of every language and there is
good reason for this.6
6
This argument was elaborated sixty years later by James Harris (1968:327-49).
8
Need is the mother of invention both of general ideas and their names. General
ideas are signs because they are general: Only signs, "the inventions and creatures
of the understanding, can be general or universal". It follows that general words are
signs (names) of signs (ideas). However, while words bear no resemblance to the
ideas they stand for, general ideas do. They are created by the human mind by
leaving out one or more ideas in which similar individuals differ: "Ideas become
general by separating from them the circumstances of time and place and any other
ideas that may determine them to this or that particular existence" (LOCKE III, iii, 6).
This process of separation is called abstraction. A general or abstract idea is
therefore a subset of ideas common to a number of individuals. Each individual
conforming to this abstract idea is of that class and has a right to its name.7 I can
rank an idea under a general name if it conforms to the general, abstract idea it
marks, i.e. if it contains all the ideas the abstract idea is made up of.
By successive abstraction the mind can create a hierarchy of ideas, each more
general than the previous one. Abstraction being a process of elimination, the
generality of an idea, and thus the number of individuals it can represent, is in
inverse proportion to the number of its constituent ideas. "The more general term is
always the name of a less complex idea" (III, vi, 32). By definition, each idea at one
level of a hierarchy of ideas represents all ideas at the lower levels. Inversely, the
constituent ideas of a complex idea at a higher level form a subset of the
constituents of the ideas of the lower levels (III, vi, 32). LOCKE'S concept of general
ideas at different levels of abstraction corresponds to a taxonomy of genera and
species.
... this whole mystery of genera and species, which make such a noise in the schools
and are with justice so little regarded out of them, is nothing but abstract ideas, more
or less comprehensive, with names annexed to them. In all which, this is constant and
unvariable: that every more general term stands for such an idea as is but a part of any
of those contained under it. (III, iii, 9)
General ideas enable men to think at any level of generality, to think with
dispatch, as it were; general terms enable men to speak about these thoughts, again
with dispatch. If we could only speak about individuals our knowledge would not
extend further than individuals, if we had only proper names our communication
would not extend further either.
5. The private and public uses of language
Words are articulate sounds (III, i, 1).8 Their primary use "is to be sensible
marks of ideas and the ideas they stand for are their proper and immediate
signification" (III, ii, 1). More specifically, "Words, in their primary or immediate
signification, stand for nothing but the ideas in the mind of him that uses them"
7
. For a more detailed discussion of general terms see Schreyer (1990:399-414).
8
ASHWORTH (1985:30) identifies this with the scholastic term vox.
9
(LOCKE III, ii, 2; III, ii, 7). The connection between words and ideas is arbitrary,
created by "a voluntary imposition" (III, ii, 1) of each individual. In this sense words
are an extremely private possession. They can represent nothing outside the mind of
the individual, neither things nor ideas in other minds. When confined to the
individual, language serves the recording of our thoughts. When used for this
purpose it is sufficient that the individual has ideas and that he uses the same words
as marks for the same ideas. This is the private use of language, a use logically prior
to the public and chief use of language.
Language is "the great instrument and tie of society" (III, i, 1) and its chief end
is communication. It "is to be the easiest and shortest way of communicating our
notions" to others (III, vi, 33).9 The chief use of words is "to make our ideas within
us known to others" (III, vi, 45)10. To be useful for communication, words have to
become a public property: A speaker speaks intelligibly if his "words excite the same
ideas in the hearer which he makes them stand for in speaking" (LOCKE III, ii, 8; ix,
4). In the process of understanding, a hearer "represents to himself other men's
ideas by some of his own" (LOCKE III, ii, 2). If these ideas differ, communication will
fail. Therefore the first principle of communication is that speaker and hearer should
use the same word for the same idea, "for if the idea be not agreed on ... for which
the word stands, the argument is not about things11 but names." (III, xi, 6) As a
corollary people should not use words "without clear and distinct ideas" (III, ix, 2)
and one word should denote only one idea. Synonyms are superfluous, ambiguous
words open to misunderstanding.
5.1 Civil use
A word can thus be seen as an aide de mémoire or as a vehicle of

communication. In the first case it is a sensible mark a person voluntarily and
arbitrarily annexes to ideas of the same sort. In the second case it is a sign used by
a speaker to signify (make known) his ideas to a hearer. In the first case a person
may choose any signs he pleases, in the second case the signs must be the common
9
As a vehicle of communication language is either "by words as may serve for the upholding common
conversation and commerce about the ordinary affairs and conveniences of civil life" (civil use), or by
words "as may serve to convey the precise notions of things, and to express in general propositions
certain and undoubted truths which the mind may rest upon and be satisfied with in its search after true
knowledge" (philosophical use). Philosophical use requires more exactness than civil use (III, ix, 3) and
therefore clear and unambiguous terms.
10
The term idea(s) is the most frequent noun in the Essay (occurrence 3680 times) and its
interpretation perhaps more controversial than that of any other term in LOCKE scholarship. LOCKE
obviously and consciously uses it with greater latency than suits his interpreters. In his introduction (I, i,
8) he writes: " ... I must here in the entrance beg pardon of my reader for the frequent use of the word
idea, which he will find in the following treatise. It being that term which, I think, serves best to stand for
whatsoever is the object of the understanding when a man thinks, I have used it to express whatever is
meant by phantasm, notion, species, or whatever it is which the mind can be employed about in
thinking...". ROBINSON (1954:74-77) argues that LOCKE himself does not adhere to his stipulated
meaning because he was unconsciously influenced by the traditional meaning of idea as "something
essentially part of the thinker or perceiver and not of the objects he surveys".
11
A rather loose (civil?) use of things for ideas.
10
property of speaker and hearer. Ideal communication demands first that all users
have the same ideas and second, that they annex the same ideas to the same
sounds. In the real world neither of these strict conditions are fully met, nor is it
always necessary that they should be. Many words of civil (i.e. everyday) use, for
instance, are of a doubtful and uncertain signification (LOCKE III, ix, 4). In existing
languages a standard of propriety is often missing. If use is the standard, asks
LOCKE, what if there is no standard use? For civil communication this may not be so
necessary: "Vulgar notions suit vulgar discourses". In philosophical (i.e. scientific)
use, however, the vagueness and obscurity of many words are intolerable. If human
knowledge is the knowledge of the relations between ideas, there can be neither
knowledge nor information where the ideas are not clear and determinate.
5.2 The philosopher's dilemma
Needless to say, it is the philosophical use of language that LOCKE is concerned

with. As we have seen, words are intricately and perhaps inextricably connected with
both the acquisition and the transmission of knowledge. LOCKE'S inquiry into
language is meant to help reduce obscurity and obscurantism in science by
identifying, explaining, and finding remedies for misinformation and
misunderstandings arising from the misuse of language. LOCKE does this by analysing
the relation between words, ideas and things and creating an awareness for
problems that may arise from their peculiar relationship. To improve the quality of
philosophical argumentation it is necessary to "understand better the use and force
of language" (III, i, 6)
The complaint that words, albeit necessary, can be an impediment to science is

not new. BACON in his ADVANCEMENT OF LEARNING (1605) (HÜLLEN 1989:36-42) analyses
"the false appearances that are imposed upon us by words"12. Fifty years later the
unreliability of words was debated in the Royal Society, which decided to remedy this
imperfection of language by adopting their frequently invoked but rarely attained
plain style (HÜLLEN 1989:98-113; AARSLEFF 1982:225-238). Improving language was
also one motive underlying the project of a universal, philosophical language, so
frequently proposed in the 17th century, which led to the attempt at rigorous logical
and philosophical definitions in JOHN WILKINS' ESSAY CONCERNING A REAL CHARACTER
12
"... certain it is that words, as a Tartar's bow, do shoot back upon the understanding of the wisest,
and mightily entangle and pervert the judgement. So as is almost necessary, in all controversies and
disputations, to imitate the wisdom of the mathematicians, in setting down in the very beginning the
definitions of our words and terms, that others may know how we accept and understand them, and
whether they concur with us or no. For it cometh to pass, for want of this, that we are sure to end there
where we ought to have begun, which is, in questions and differences about words." ( BACON 1873:146;
XIV, 9). HOBBES (Leviathan 1968: 105) uses a different metaphor. A scientist needs "to remember what
every name he uses stands for, and to place it accordingly; or else he will finde himselfe entangled in
words, as a bird in lime-twigges; the more he struggles, the more belimed". Hobbes therefore
recommends the method of geometers who "begin at settling the significations of their words; which
settling of significations they call Definitions..."
11
(1668). Some twenty years after the publication of Wilkins' Essay, LOCKE, like Wilkins
a fellow of the Royal Society, still complains about the adverse effects of the
imperfection of language. In a passage reminiscent of Bacon he writes:
And here I desire it may be considered and carefully examined whether the greatest
part of the disputes in the world are not merely verbal and about the signification of
words ... (LOCKE III, xi, 7)
The problem is that the mind has no immediate access to the world of things.
We can grasp reality only through the mediation of ideas. And we communicate ideas
only by the mediation of the words13 of our language. A corollary of this is what
ASHWORTH (1985:65) calls the double conformity of ideas; "on the one hand our
ideas must conform to the ideas of other language users; and on the other hand our
ideas must conform to the reality of things". In the real world comprehensibility and
exactness are often incompatible. The philosopher is therefore caught on the horns
of a dilemma. Philosophy requires him to have determinate ideas conformable to
reality, communication requires him to apply his words "as near as may be to such
ideas as common use has annexed them to" (III, xi, 11). To be understood he must
use words signifying the vague, skewed ideas signified by the words of his language,
ideas created centuries ago by his rude and unlearned forbears. It seems that a good
philosopher must remain silent or misunderstood - or he can tamper with the
language, at least with his own.
6. Definition
6.1 What definition is good for
Bacon, Hobbes and LOCKE all recommend the definition or better- re-definition -
of terms as a remedy for this philosophical plight. Definitions help remove the
linguistic obstructions to the progress of science. They are necessary wherever the
meaning of terms is unclear or different in different minds.
In communication with others, it is necessary that we conform the ideas we make the
vulgar words of any language stand for to their known proper significations ...or else to
make known the new signification, we apply them to. (III, vi, 51)
Many communication failures could be avoided if speakers used their words
"constantly in the same sense" and - if not- made their sense at least known. In
philosophical, i.e. scientific, discourse the use of definition is to settle and/or reduce
the significations of the obscure, unsteady or equivocal terms of everyday language.
Definition is a way of making sure that scientists speak about the same thing. It will
often be necessary, says LOCKE "to make known the new signification" we apply
words to (III, xi, 14); he asks his readers to consider
whether, if the terms … were defined and reduced in their signification (as they must be
where they signify anything) to determined collections of the simple ideas they do or
13
For a discussion of the scholastic sources of this and other aspects of LOCKE'S linguistics see
ASHWORTH (1985:299-357).
12
should stand for, those disputes would not end of themselves and immediately vanish.
(LOCKE III, xi, 7)
They would, of course, and so definition would promote peace on earth, or at
least in the scientific community. Definition is necessary because in common and
philosophical use words do not - or not always - stand for one and the same
determined collection of ideas. Whenever necessary philosophers should "declare
their meaning where ... common use has left it uncertain and loose" (III, xi, 11). But
is definition a miracle cure against ambiguity, vagueness and vacuousness, against
ideas being various in various minds? Bacon had his doubts: Definitions use words to
explain words, which need to be defined by words and so ad infinitum. LOCKE'S
theory of signification attempts to stop the infinite regress.
6.2 What definition is
Now we know what definition is good for, but what is it?

I think it is agreed that a definition is nothing else but the showing the meaning of one
word by several other not synonymous terms. The meaning of a word being only the
ideas they are made to stand for by him that uses them, the meaning of any term is
then shown or the word is defined when, by other words, the idea it is made the sign of
and annexed to, in the mind of the speaker, is as it were represented or set before the
view of another, and thus its signification ascertained: this is the only use and end of
definitions, and therefore the only measure of what is or is not a good definition.(III, iv,
6)
In other words, the definition of a word W is a sequence of different words
which ideally stand for the same determined collection of simple ideas which W
stands for in the mind of the speaker.
6.3 Two types of definition
The traditional method of definition was by genus and differentia. LOCKE

proposes a better and more explicit type of definition related to, but distinct from,
the old method of the schools:
... definition being nothing but making another understand by words what idea the term
defined stands for, a definition is best made by enumerating those simple ideas that are
combined in the signification of the term defined. (III, iii, 10)
The greatest merit of a definition by genus and differentia is not clarity but
brevity.14 It arose "not out of necessity, but only to save the labour of enumerating
the several simple ideas which the next general word or genus stands for, or,
perhaps, sometimes the shame of not being able to do it" (LOCKE III, iii, 10).
Definition by genus and differentia is but a convenient shortcut and can always be
replaced by a complete enumeration of all the simple ideas that make up a complex
one. The reverse is, however, not the case: Consistent definition by genus and
differentia presupposes a consistently hierarchical system of meanings without gaps,
hardly a realistic assumption, for "languages are not always so made according to
14
"...and if, instead of such an enumeration, men have accustomed themselves to use the next
general term, it has not been out of necessity or for greater clearness, but for quickness and dispatch
sake." (III, iii, 10)
13
the rules of logic that every term can have its signification exactly and clearly
expressed by two others." As we have seen above, the hierarchical structure of the
vocabulary is not a necessity, but a convenience, which enables speakers to
communicate their ideas with brevity. It allows them to use one single term for a
complex idea, rather than having to enumerate all the names of the simple ideas
separately.
The distinction between simple and complex ideas is crucial to an

understanding of LOCKE'S concept of definition. Complex ideas are composed of
simple ideas. Complex ideas fall into two major classes, ideas of mixed modes and of
substances. "The names of simple ideas are not capable of any definitions; the
names of all complex ideas are" (III, iv, 4). Complex ideas can - at least in principle
- be defined by enumerating all simple ideas they are made up of.
Notice that for LOCKE definition is purely verbal: In his use of the term a non-
verbal definition is a contradictio in adiecto15. Nevertheless, non-verbal (e.g.
ostensive) methods of making one's meaning known to a hearer may be useful - and
sometimes even necessary: Some terms are beyond definition. The definability of a
word depends on the type of idea it stands for. There are limits to definition, which
LOCKE believes he is the first to observe.
As the ideas men's words stand for are of different sorts, so the way of making known
the ideas they stand for, when there is occasion, is also different. For though defining
be thought the proper way to make known the proper significations of words, yet there
be some words that will not be defined, as there be others whose precise meaning
cannot be made known but by definition, and, perhaps, a third which partake somewhat
of both the other, as we shall see in the names of simple ideas, modes and substances.
(III, xi, 13)
6.4 Why simple ideas cannot be defined
LOCKE'S concept of simple ideas is open to interpretation. The discussion in the

literature is controversial and inconclusive. O'CONNOR (1952:47-51) recognizes four
different senses of the phrase simple idea whose mutual relation is not altogether
clear. One explanation given by LOCKE runs as follows: Simple ideas are gained by
"internal or external sensation" (II, xi, 17). The latter are the result of the
impressions of external objects on one or more of our senses. A person lacking a
sense cannot therefore form a simple idea entering through this channel; a blind
man cannot form an idea of redness (III, iv, 13). Simple ideas of internal sensation
are the operations performed by the mind on ideas gained by introspection (e.g.
remembrance, discerning, reasoning). Here simple idea refers to the smallest unit of
experience. Another interpretation of simple idea is that it is the final unit of
15
Cf. LOCKE (III, iii, 10): "...definitions, as has been said, being only the explaining of one word by
several others so that the meaning or idea it stands for may be certainly known..."; LOCKE (III, iv, 7):
"...and therefore a definition which is properly nothing but the showing of the meaning of one word by
several others, not signifying each the same thing..."
14
abstraction, the ultimate indivisible result of comparing and abstracting from

complex ideas of particulars.
Whatever the philosophical problems inherent in LOCKE'S concept of simple

ideas, for his theory of definition it is important that simple ideas are primitives and
that they are related to the material world. Obviously simple ideas cannot be broken
down into simpler ones16: The names of simple ideas are beyond definition.
Ignorance of this truism has led to many trifling definitions of simple ideas in the
schools, some incomprehensible, some tautologous, some circular. The following
methods LOCKE dismisses as useless:
(1) A scientific description of the cause of a simple idea cannot produce this idea in
another mind: "For the cause of any sensation and the sensation itself, in all the simple
ideas of one sense, are two ideas, and two ideas so different and distant one from
another that no two can be more so". (III, iv, 10)
(2) An explanation by similarity cannot produce the idea either. Similarities within one
sense are but approximations, similarities across the senses are absurd, as they
suppose "that sounds should be visible or colours audible" (III, iv, 11). Words, being
sounds, can only produce simple ideas of sounds.
(3) An explanation of the name of a simple idea by synonyms presupposes the
existence of that idea, it does not create it. "This is to translate and not to define, when
we change two words of the same signification one for another; which, when one is
better understood than the other, may serve to discover what idea the unknown stands
for, but is very far from a definition, unless we will say every English word in the
dictionary is the definition of the Latin word it answers... (III, iv, 9)
Explaining the name of a simple idea to a person is the same as creating the
idea in his mind. Since simple ideas are related to the real world they can only be
created by exposing a person to that world, mediately or immediately. To do this you
must name "the subject wherein that simple idea is found" or (if he is unacquainted
with the subject or its name) present him with the subject that will excite it (III, xi,
14). Thus in LOCKE'S theory verbal definitions can always be reduced to what we, but
not he, would call ostensive definitions.17
Simple ideas ... can only be got by experience from those objects which are proper to
produce in us those perceptions. When, by this means, we have our minds stored with
them and know the names for them, then we are in a condition to define and by
definition to understand the names of complex ideas that are made up of them. (III, iv,
14)18
16
GONSALVES' (1991:257-8) accuses LOCKE of inconsistency, since he should have noted "that just as in
many cases definitions have more than two components, in many others they have, being simple and not
complex, only one". This totally misses the point: Definitions, for LOCKE, are verbal explanations, and the
only explanation of the name of a simple idea would be by that name; in other words, it would be a
tautology. What cannot be broken down into component ideas with names attached, cannot be defined.
A "definition of redness that says that it is a physical quality that is a colour that is redness", where the
use of redness alone" is justified by virtue of its nature as a primitive", Locke would argue, cannot give
anybody an idea of redness, and is thus useless and superfluous.
17
ROBINSON (1954:43) argues against the claim that things that cannot be analysed cannot be defined:
The fact that we cannot analyse something, does not prove it to be unanalysable. True as this may be,
the fact that a speaker cannot analyse a thing at a given moment means that he cannot define it at that
moment. Furthermore, however refined the analysis, it will, in real life, always end with constituents that
are not analysed and thus treated as primitives. Ars longa, vita brevis.
18
LOCKE uses a two tier method of explaining meaning: Definition (= explanation by words) for
complex ideas is ultimately based on ostension (explanation by showing) for simple ideas. LOCKE does
15
6.5 Why the definition of complex ideas is difficult
Whatever they are, simple ideas are the result of the inescapable impression of
the real world of things on our senses. Any ideas composed of simple ideas are,
therefore, also related to the real world, albeit only mediately. All complex ideas are
ultimately reducible to simple ideas since they are but simple ideas combined. LOCKE
thus advocates an atomistic theory of signification with simple ideas as semantic
primitives. The names of complex ideas can be defined by enumerating all the names
of the simple ideas they are made up of (II, xxii, 9).
The suggestion that complex ideas may be decompounded into a series of

simple ideas is not new. Take for instance this quotation from VINDICIAE ACADEMIARUM
(1654):
But it did presently occurre to me, that by the helpe of Logick and Mathematiticks
[sic!] this might soon receive a mighty advantage, for all Discourses being resolved
in sentences, those into words, words signifying either simple notions or being
resolvible into simple notions, it is manifest, that if all the sorts of simple notions be
found out, and Symboles attached to them, those will be extreamly few in respect
of the other, (which are indeed Characters of words, such as Tullius Tiro's) the
reason of their composition easily known, and the most compounded ones at once
will be comprehended, and yet will represent to the very eye all the elements of
their composition, & so deliver the natures of things: and exact discourses may be
made demonstratively without any other paines then is used in the operations of
specious Analyties (WARD in DEBUS 1970:215).
LOCKE'S Essay appeared 36 years after these observations by SETH WARD and
JOHN WILKINS. And they were not new even then. The idea of the technical word
occurs, for instance, in a letter from KINNER to HARTLIB of 1647 (cf. HÜLLEN 1989: 160
ff.).
In Lockian semantics definition will be successful if the hearer holds these

simple ideas in his mind and knows their names as used in the definition. If he does
not know the terms of the definition it may be saved by the use of synonyms, which
may be known to him (III, iv, 14). This sounds easy enough. In practice, however,
the definition of the terms for mixed modes and substances proves difficult. The
ideas they stand for are often unsettled or vague even in philosophical minds. It is a
mistake to believe that the "names of complex ideas having so settled determined
significations that they are constantly used for the same precise ideas" (III, xi, 25).
The real problem is the determination of the idea itself. These ideas are hard to
analyze because they are hard to learn. And they are hard to learn because they are
hard to analyse. Every child must learn anew the conventional connection between
name and idea. Especially in the case of complex ideas, learners often encounter
names before they have the corresponding idea. They must, therefore, re-create in
their mind the idea conventionally annexed to a name. This may be difficult for
several reasons. The ideas the names of mixed modes "stand for are very complex,
not say how we isolate a simple idea from the multitude of impressions assailing our senses and how we
16
and made up of a great number of ideas put together" and furthermore they "have
no certain connexion in nature, and so no settled standard anywhere in nature
existing, to rectify and adjust them by". In the case of substances "the signification
of the word is referred to a standard, which standard is not easy to be known" or
"the signification of the word and the real essence of the thing are not exactly the
same". (III, ix, 5)
6.5.1 Defining mixed modes
Complex ideas of mixed modes (murder, adultery) have no archetype in nature.

They are created by the understanding voluntarily and arbitrarily out of "scattered
independent ideas". (III, iv, 10; cf. II, xxii, 3-8 ). The name attached to them "is as
it were the knot that ties them fast together". Since these ideas "have not so much
anywhere the appearance of a constant and lasting existence as in their names",
these names are "very apt to be taken for the ideas themselves" (II, xxii, 8).
The more complex the idea to be conveyed by the name of a mixed mode the
greater linguistic economy, and the more difficult its analysis. "The multiplicity of
simple ideas put together in mixed modes" explains why (cf. II, xxii, 9) people using
a term may connect no, or only an imperfect, idea with it. Thus the signification of
these words has some latitude . "The measure of propriety itself being nowhere
established" (III, ix, 8) the proper use of a word is often a matter of dispute. As a
result many significations of terms for mixed modes in common use are difficult to
analyse and thus difficult to define.
Since mixed modes have no standard in nature there is no limit to their number
or complexity. In fact, their complexity is a natural consequence of the reason for
their invention: They serve the chief end of language "to signify with ease and
dispatch general conceptions, wherein not only abundance of particulars may be
contained, but also a great variety of independent ideas collected into one complex
one. In the making, therefore of the species of mixed modes, men have had regard
only to such combinations as they had occasion to mention to one another" (III, iv,
7; cf. II, xxii, 5); they depend on the "customs, fashions and manners of one
nation". Mixed modes being culture-dependent and changeable combinations of
simple ideas they are often untranslatable, albeit not indefinable (II, xxii, 6).
The problem with the names of mixed modes is that in common use their
signification is vague and variable. Appeal to common use is therefore "not sufficient
to adjust them to philosophical discourses". The obscurity of the terms for mixed
modes is however, avoidable. Having no archetype in nature they must correspond
with, they "may be perfectly and exactly defined" (III, xi, 15). In fact, they need to
be defined, as definition is "the only way whereby the signification of the most of
them can be known with certainty" (III, xi, 18). There is no standard of propriety in
identify it to a hearer in order to make him have that same idea.

17
the language to regulate the use of these terms, nor is there a standard in nature to
be appealed to as in the case of the names of substances.
6.5.2 Defining substances
The names of substances are arbitrary, the ideas of substances are not. Ideas
of particular natural substances (such as gold) are "collections of such qualities as
have been observed to co-exist in an unknown substratum, which we call
substances" (IV, iv, 7). Thus, the complex idea of a particular substance is a
collection of simple ideas plus "the confused idea of something to which they belong,
and in which they subsist" (II, xxiii, 3, 6). However, this substratum is unknowable;
our "faculties carry us no further towards the knowledge and distinction of
substances than a collection of those sensible ideas which we observe in them" (III,
vi, 9).
This, then, in short, is the case: nature makes many particular things, which do agree
one with another in many sensible qualities, and probably too in their internal frame
and constitution; but it is not this real essence that distinguishes them into species: it is
men who, taking occasion from the qualities they find united in them and wherein they
observe often several individuals to agree, range them into sorts, in order to their
naming, for the convenience of comprehensive signs; under which individuals, according
to their conformity to this or that abstract idea, come to be ranked as under ensigns: so
that this is of the blue, that the red regiment; this is a man, that a drill; and in this, I
think, consists the whole business of genus and species. (III, vi, 36)
General ideas of substances are creatures of the understanding, "signs of such
complex ideas wherein several particular substances do or might agree, by virtue of
which they are capable of being comprehended in one common conception and be
signified by one name" (III, vi, 1; 32). They function as "patterns or forms" (III, iii,
13) which serve the distribution of things into their classes.
The boundaries of all species are man-made; the only constraint to randomness
is the condition of co-existence. Thus the main problem in forming ideas of
substances seems to be to determine how many simple ideas are "constantly and
inseparably joined in nature" (III, vi, 30). LOCKE'S answer is that we do not know and
that there is no way of knowing: "How shall the mind, when it perceives nothing but
its own ideas, know that they agree with things themselves?" (iv, vi, 3). Thus there
can be no natural constraint on the number of co-existent qualities joined in the
complex idea of a substance; this number "depends on the various care, industry, or
fancy" of the maker of an idea: "Men generally content themselves with some few
sensible obvious qualities, and often, if not always, leave out others as material and
as firmly united as those they take" (III, VI, 29). This explains why there is no
agreement on "the precise number of simple ideas or qualities belonging to any sort
of things, signified by its name" (III, vi, 30).
The difficulty in defining names of substances is that to different speakers these

do not signify the same, or the same number of, co-existent simple ideas:
For we, having need of general names for present use, stay not for a perfect discovery
of all those qualities which would show us their most material differences and
agreements; but we ourselves divide them by certain obvious appearances into species,
18
that we may the easier, under general names, communicate our thoughts about them.
(III, vi, 60)
Existing languages are historical products, created and developed by ignorant
people. Their ideas were neither precise nor scientific and their words therefore have
a variable meaning. Our ideas of natural substances must be considered defective, if
one demands, as LOCKE does, that they "must also be conformable to things as they
exist" (III, xi, 10): "For the attaining of knowledge and certainty, it is requisite that
we have determined ideas, and to make our knowledge real, it is requisite that the
ideas answer their archetypes." (IV, iv, 8;3)
And therefore in substances, we are not always to rest in the ordinary complex idea
commonly received as the signification of that word, but must go a little further and
inquire into the nature and properties of the things themselves, and thereby perfect, as
much as we can, our ideas of their distinct species, or else learn them from such as are
used to that sort of things and are experienced in them. (III, xi, 24)
Philosophers in particular need a determinate idea of each substance, but,
ideally, also a complete idea, i.e. a complex idea composed of all "those simple ideas
wherein they observe the individuals of each sort constantly to agree" (III, xi, 25).
From this point of view, definitions of our ideas of substances leave much to be
desired, precisely because our ideas of substances do. In other words, "to define
their names right, natural history is to be inquired into"19.
In philosophical use a definition of our ideas is insufficient where ideas

represent real things. When we claim to speak of real things, we really claim that we
speak of ideas that co-exist. Which co-existent ideas go into our definition of a
substance, and how many, we are at liberty to decide. Thus our definitions of
substances must not combine ideas that do not co-exist in nature and they should
combine those that co-exist frequently.
LOCKE does not, however, hold that there is one correct definition of a name of
a substance. Since the real essence of any substance (from which all its properties
are supposed to flow) is unknowable, none of the simple ideas we assume to
correspond to the properties of things, stand in a necessary relation. Our liberty to
pick and choose from "the vast numbers of similarities and regularities" (Ashworth
1985:69) which we observe in nature follows from our having no reason for
preferring one choice to another, other than practical interest.
We are certainly, in LOCKE'S view at least, in no position to claim that our definitions
capture real distinctions among things in such a way that we can make definitive
classifications, nor that our definitions will capture real properties in such a way that
they can be used to expand our knowledge by making deductions and necessary
predictions based on this set of real properties. (ASHWORTH 1985:69)
19
This is precisely what Johnson attempted and gave up for reasons foreseen by LOCKE 60 years
previously. Says he: "But a dictionary of this sort, containing, as it were, a natural history, requires too
many hands as well as too much time, cost, and sagacity ever to be hoped for; and till that be done, we
must content ourselves with such definitions of the names of substances as explain the sense men use
them in". (III, xi, 25).
19
The names of substances are best explained by both "showing and defining"
(III, xi, 19). The reason for this combined approach is that not all component ideas
are of equal importance: In ideas of substances there are "ordinarily ... some leading
qualities, to which we suppose the other ideas which make up our complex idea of
that species annexed".... These leading or characteristical (....) ideas, in the sorts of
animals and vegetables are (...) mostly figure, and in inanimate bodies colour, and in
some both together". Since these leading sensible qualities make up the most
"observable and invariable part in the definitions of our specific names" (III, xi, 19-
20) they "are best made known by showing and can hardly be made known
otherwise".20 Definition comes into its own where, as is often the case, simple ideas
of substances are powers21 rather than sensible qualities: Here verbal enumeration
must be resorted to.
6.6 Definitions and trifling propositions
Definitions are very similar to what LOCKE calls trifling propositions. He defines
a proposition as the "joining and separating of signs" and truth as the "the joining or
separating of signs, as the things signified by them do agree or disagree with one
another" (IV, v, 2; 5). Since both ideas and words are signs, LOCKE distinguishes
mental and verbal propositions, and consequently mental and verbal truth. Truth is
verbal, when the words in a proposition stand for ideas which do not agree with the
reality of things. It is real when such an agreement exists. This is important for
understanding what LOCKE means by a trifling proposition.
Whenever the whole or a part of a complex idea is affirmed of itself, we have a

trifling proposition. Trifling propositions are analytic and therefore always true: their
certainty is but a verbal certainty and therefore uninstructive. Identical propositions,
where a term is affirmed of itself (e.g. A soul is a soul), are one type of trifling
proposition.
It is but like a monkey shifting his oyster from one hand to the other; and had he had
but words, might no doubt have said: Oyster in right hand is subject, and oyster in left
hand is predicate; and so might have made a self-evident proposition of oyster, i.e.
Oyster is oyster, and yet, with all this, not have been one whit the wiser or more
knowing... (IV, viii, 3)
A second sort of trifling proposition is "when a part of the complex idea is
predicated of the name of the whole: a part of the definition of the word defined",
e.g. when the genus is predicated of the species (e.g. Lead is a metal). This
proposition is also uninstructive. People who know the signification of lead do not
learn anything new. The use of this second type of proposition is as a definition; its
20
"For the shape of a horse or cassowary will be but rudely and imperfectly imprinted on the mind by
words, the sight of the animals doth it a thousand times better." (III, xi, 21) Therefore LOCKE suggests
that in dictionaries words for things whose leading quality is their shape should be "expressed by little
draughts and prints made of them", as naturalists did then and as some dictionaries do now.
21
Here LOCKE uses the term simple idea expressly in a loose sense to include observable powers "to
change some sensible qualities in other subjects" (II, xxiii, 7), although these powers "are truly complex
ideas".
20
form that of the traditional definition by genus and differentia specifica, with the
specific difference left out:
Indeed, to a man that knows the signification of the word metal, and not of the word
lead, it is a shorter way to explain the signification of the word lead, by saying it is a
metal, which at once expresses several of its simple ideas, than to enumerate them one
by one... (IV, viii, 4)
This type of trifling proposition can be used "where a man goes to explain his
terms to one who is supposed to, or declares himself not to, understand him; and
then it teaches only the signification of that word or the use of that sign" (IV, viii, 7).
In other words, trifling propositions can be used as (partial) definitions, but they do
not add to real knowledge.
This, I think I may lay down as an infallible rule: that wherever the distinct idea any
word stands for is not known and considered, and something not contained in the idea
is not affirmed or denied of it, there our thoughts stick whole in sounds and are able to
attain no real truth or falsehood. (IV, viii, 13)
7. Against a philosophical language
Definitions resemble trifling propositions: They convey merely verbal

knowledge. With a definition the scientist can, at best, make known what complex
idea he has annexed to a word. Definition can thus standardize ideas in different
minds. It cannot create knowledge, but, being the prerequisite for scientific
communication, it can help advance it. Considering the practical importance of
definition for science it may seem curious that LOCKE does not speak up for a
wholesale reform of the language of science. He exhorts us to avoid confused ideas,
to use our words consistently,to use them like the other language users. He points
us to the difficulties inherent in traditional languages and he advises us when and
how we can avoid them by ostension or definition. He preaches clarity but he does
not preach reform.
I am not so vain to think that anyone can pretend to attempt the perfect reforming the
languages of the world, no not so much as of his own country, without rendering
himself ridiculous. To require that men should use their words constantly in the same
sense and for none but determined and uniform ideas, would be to think that all men
should have the same notions and should talk of nothing but what they have clear and
distinct ideas of. Which is not to be expected by anyone who hath not vanity enough to
imagine he can prevail with men to be very knowing or very silent. (III, xi, 2)
AARSLEFF (1982:72) suggests that this remark may well be directed against the
project of a universal and philosophical language, which was intended to reform
language by inventing it afresh, but on philosophical principles. This new, artificial
language was to be a language whose words would, by the selection and
combination of vocal and/or graphical signs, provide a definition of their meanings
and of the reality of things. In the first half of the 17th century this proposal had
been aired and discussed by many authors, such as WARD, LODOWICK, DALGARNO,
MERSENNE and DESCARTES. The discussion culminated in the invention of a universal
language in JOHN WILKINS' Essay of 1668, an enormous work whose largest part
represents an attempt to isolate and arrange meanings (which were mostly the
21
meanings of English words) into one enormous hierarchy according to the old logical
principle of genus and differentia specifica. But words are related to things and by
divine decree things in this world are ordered in one global hierarchy.22 The idea of
such a great chain of being was still very present in the minds of the philosophers of
the Royal Society:
Such is the Dependence amongst all the Orders of creatures; the inanimate, the
sensitive, the rational, the natural, the artificial; that the Apprehension of one of them is
a good step toward the understanding of the rest: And this is the highest Pitch of
human reason; to follow all the links in this chain, till all their Secrets are open to our
minds; and their Works advanc'd, or imitated by our hands. This is truly to command
the World; to rank all the varieties, and Degrees of things, so orderly one upon another,
that standing on top of them, we may perfectly behold all that are below, and make
them all serviceable to the Quiet, and Peace, and Plenty of Man's Life.
These plain words were written by THOMAS SPRAT (1667:110), historian - or
should we say propagandist - of the Royal Society, but there is little doubt that they
were underwritten if not ghosted by the then secretary of the Society - John Wilkins.
It seems as if the scholars and gentlemen of the Society believed that there was a
divine order of things and that they could discover it.
LOCKE did not share these views. He subscribed to MERSENNE'S dictum of 1625
that "no matter what we do, we shall never arrive at the point of making our intellect
equal to the nature of things" (AARSLEFF 1982:13). Even if there were a divine order,
man would not be able to discover it; for human knowledge is not of things, but of
general ideas, which are entirely the product of reason working on sense
impressions. It is the human mind that is continually creating order, hierarchies of
ideas, to escape from the particular, to make its ideas comprehensible, effable and
communicable.23 Consequently not all ideas are the same for all men. LOCKE even
casts doubt on the stability of ideas of things, such as substances, which have an
archetype in the external world,.
Universal language schemes rested on the firm conviction that the ideas of all
men were basically the same everywhere. For universal communication it was
deemed sufficient to create a nomenclature to be used by all men. LOCKE refutes the
naive belief in a universal translatability of terms. It is impossible for all men to have
the same ideas. Not even the ideas of speakers of the same language are all the
22
I think Salmon (1979:197) is too radical when claiming that "Wilkins treats names as direct
representations of objects". However, he never clearly distinguishes between ideas and things, as
demonstrates his frequent formula "notions or things", and furthermore he believes it possible for all
people to have the same ideas, so that their ideas are or can be standardized.
23
As FRED DOLEZAL pointed out in a personal communication, a similar thought is found in WILKINS
(1968:24): "Tho particulars are first in order of Being, yet Generals are first in order of Knowing,
because by these, such things and notions as are less general, are to be distinguished and defined." The
question seems to be whether man creates generals to impose his own order or to recreate in his mind a
pre-existing divine order. I believe that both LOCKE and Wilkins opted for the second alternative, LOCKE,
however, with the proviso, that the limited mind of man will never be able to fully recognize the divine
order of the things of this universe. As for Wilkins and the Royal Society the question merits further
investigation.
22
same. In other words, not all speakers of the same language speak the same
language, and it is unlikely they ever will.
AARSLEFF (1982:27) writes "Owing to the impenetrable subjectivity of ideas to

which words are tied, each individual has a radically private language that
precludes all hope of perfect communication…". As a statement of LOCKE'S linguistic
relativism this seems a little too radical: LOCKE neither believes the subjectivity of
ideas to be impenetrable nor the language of an individual to be radically private.
Why else would he devote so much space to definition and other methods of
standardizing ideas, especially in science. I would argue that (for LOCKE) scientific
communication cannot be perfect because science never ends, and that everyday
communication will never be perfect because there is no need for perfection.
By demonstrating that general ideas are man-made, largely arbitrary and thus
subject to change or revision LOCKE proved something his fellows in the Royal Society
may have suspected, viz. that a universal, philosophical language was a pipe-dream,
not merely for practical reasons, but because of its theoretical impossibility.
Philosophical (i.e. scientific) terminology is bound to change because our ideas are
bound to change. Therefore definition can merely explainand standardise our ideas,
it cannot fix them.24 Any attempt at reforming and fixing the language of science
once and for all is doomed.
LOCKE'S Essay may or may not have been responsible for the as yet unexplained
discontinuation of Wilkins' project by the Royal Society. It certainly confirmed the
worst fears of all those charged with the continuation of Wilkins' great work (SALMON
1979:191-206, AARSLEFF 1982:239-277).
In the Eighteenth century Wilkins' project was judged a noble failure; his Essay
was sometimes praised for being "a very good English vocabulary" (MONBODDO 1967:
II :456). A more pragmatic Dr. Johnson would write an English dictionary, more
ambitious than Wilkins' Essay in its comprehensiveness and its semantic information
on each word, but considerably less ambitious in its universality and definitional
rigour. Wilkins was, by his goal and his method, forced to adopt a hierarchical order
and to use a more controlled definitional vocabulary. Johnson was 'only' expected to
ascertain and fix a living language, not to design an artificial one. He used an
alphabetical order, the most practical, but semantically the worst possible
arrangement of words. Thus, in Johnson's definitions there is no recognizable global
structure. When Johnson started work on his dictionary his ideas showed some
similarity with those of Wilkins. He planned an encyclopaedic work where each entry
was to be strictly and logically defined
24
Wilkins saw this too (HÜLLEN 1989:162), at least for certain areas of the vocabulary. He was an
ardent promoter of the new science, and like LOCKE, he rejected essentialism, but unlike LOCKE he did not
recognise the implications of this rejection for science and the language of science: "His Essay is
therefore an anomaly, and the scientists were unimpressed." (Aarsleff 1982: 13)
23
I resolved likewise to shew my attention to things; to pierce deep into every science, to
enquire the nature of every substance of which I inserted the name, to limit every idea
by a definition strictly logical, and exhibit every production of art or nature in an
accurate description, that my book might be in place of all other dictionaries whether
appellative or technical. But these were the dreams of a poet doomed at last to wake a
lexicographer. (JOHNSON 1755: Preface)
When the poet turned lexicographer he realized that his plan was not only
impracticable but futile. In the end he was forced to recognize that linguistic change
was inevitable. "Those who have much leisure to think, will always be enlarging the
stock of ideas; and every increase of knowledge, whether real or fancied, will
produce new words, and new combinations of words." (JOHNSON 1755: Preface) The
words are Johnson's, the argument is LOCKE'S. If language must change so must
definitions.
Appendix A: Selections from a KWIC Concordance
LOCKE, John Essay Concerning Human Understanding (1690)
1. Definition of ....
III X 17 500 is not a good definition of a Man; 'tis plain,
III VI 22 450 ose of them to whom the definition of the Word Man, or t
III XI 20 519 should not be as good a Definition of the Name Man, stan
III VI 3 440 now is contained in our Definition of that Species, be i
III IV 10 423 is another Peripatetic definition of a simple Idea; whi
III IV 9 423 us, prove a much better definition of Motion, when well
III XI 25 523 could have from a long Definition of the Names of eithe
III VI 50 470 not being a part of the definition of the Word Gold, is
IV VIII 5 612 e any other part of the Definition of the Term defined,
IV VIII 4 612 he whole; a part of the Definition of the Word defined.
III VI 30 457 ords we have yet setled Definitions of, may, with Reason
III XI 26 523 ing of their Words, and Definitions of their Terms are n
2. Signification of ...
III XI 27 524 ten as a Man varies the signification of any Term; yet t
IV VIII 10 615 ine, i.e. determine the signification of his Names of Su
III XI 25 522 o an agreement in the signification of common Words, w
III IV 5 422 e of our Ideas, and the Signification of our Words shew,
III XI 22 520 pear; therefore, in the signification of our Names of Su
II XXI 5 236 . The Perception of the signification of Signs. . The Pe
IV VIII 12 616 r, are barely about the signification of Sounds. For sin
II XXI 26 247 eive where the confused signification of terms, or where
III IX 16 485 r Dispute was about the signification of that Term; and
III XI 24 520 ommonly received as the signification of that Word, but
III XI 9 513 he have so examined the signification of that Name, and
III IV 11 425 never come to know the signification of that Word, by a
III II 8 408 which so far limits the signification of that Sound, tha
III IV 12 425 ns, or the teaching the signification of one word, by se
III III 15 417 is the proper original signification of the Word, as is
III XI 15 516 . For since the precise signification of the names of mi
III IX 15 484 down, there the precise signification of the names of Su
II XXI 4 235 hich is the more proper signification of the word Power
III IX 13 483 s establish'd the right signification of the Word Gold?
III V 9 434 ed. And if the doubtful signification of the word Specie
III VI 43 465 same time, the ordinary signification of the Name Man, w
III IX 17 486 hority to determine the signification of the Word Gold,
24
III XI 19 518 For the explaining the signification of the Names of Su

III VI 40 464 reof; and so settle the signification of the Names, wher
Literature
AARON, RICHARD I. 1971 John Locke. Oxford: Oxford University Press.

AARSLEFF, HANS 1982 From LOCKE to Saussure: Essays on the Study of Language and
Intellectual History. Minneapolis: University of Minnesota Press.
ARMSTRONG, ROBERT L. 1965 "John LOCKE'S 'Doctrine of Signs': A New Metaphysics". Journal
of the History of Ideas 26, 369-382.
ARNDT, HANS WERNER 1979 "John LOCKE: Die Funktion der Sprache". Grundprobleme der
großen Philosophen. Philosophie der Neuzeit 1. Ed. by Josef Speck. Göttingen: Vandenhoek
& Ruprecht.
ASHWORTH, E. JEAN 1985 "LOCKE on Language". Studies in Post-Medieval Semantics, 327 -
355. London Variorum Reprints.
ASHWORTH, E. JEAN 1985 "'Do Words Signify Ideas or Things?' The Scholastic Sources of
John LOCKE'S Theory of Language". Studies in Post-Medieval Semantics, 299-326. London:
Variorum Reprints.
AUSTIN, JOHN L. 1979 (1939). "Are There 'A Priori` Concepts?". Philosophical Papers, 32-54..
Ed. by J. O. Urmson and G. J. Warnock. Oxford etc.: Oxford University Press.
AYERS, MICHAEL 1981 "LOCKE'S Doctrine of Abstraction: Some Aspects of its Historical and
Philosophical Significance". John LOCKE. Symposium Wolfenbüttel 1979. Ed. by Reinhard
Brandt. 1981, 5-24. Berlin: de Gruyter.
BACON, FRANCIS (1561-1626) 1857-74 The Works of Francis Bacon, Lord Verulam, Viscount
St. Alban, and Lord High Chancellor of England. Collected and edited by James Spedding,
Robert Leslie Ellis and Douglas Denon Heath. 6 vols. Vol 4: Of the Advancement of
Learning 1872 (1605). London: Longmans et. al.
BENNETT, JONATHAN 1971 LOCKE, Berkeley, Hume. Central Themes. Oxford: Oxford
University Press.
BOSWELL, JAMES 1964 The Life of Samuel Johnson, ed. by George Birkbeck Hill, enlarged by L.
F. Powell, 6 vols, Oxford: Clarendon University Press
BRANDT, REINHARD (ed.) 1981 John LOCKE. Symposium Wolfenbüttel 1979. Berlin: de
Gruyter.
CHRISTENSEN, FRANCIS 1946 "John Wilkins and the Royal Society's Reform of Prose Style". 2
pts. Modern Language Quarterly 7: 179-187; 279-290.
CRANSTON, MAURICE 1985 (1957). John LOCKE. A Biography. Oxford: Oxford University
Press.
DANTO, ARTHUR 1969 "Semantic Vehicles, Understanding, and Innate Ideas". Language and
Philosophy. Ed. by S. Hook. 1969, 122-137. New York: New York University Press.
DEBUS, ALLEN G. 1970 Science and Education in the Seventeenth Century. The Webster -
Ward Debate. London: Macdonald & New York: American Elsevier Inc.
DOLEZAL, FREDRIC 1985 Forgotten But Important Lexicographers: John Wilkins and William
Lloyd. A modern Approach to Lexicography before Johnson. Tübingen: Niemeyer.
DOLEZAL, FREDRIC 1989 "The Bibliography of Words and Notions", International Journal of
Lexicography 2, 167-175.
DONEY, WILLIS 1956 "LOCKE'S Abstract Ideas". Philosophy and Phenomenological Research
16, 407-408.
DUNN, JOHN 1984 LOCKE. Oxford: O.U.P..
25
FLASDIECK, HERMANN M. 1928 Der Gedanke einer englischen Sprachakademie in

Vergangenheit und Gegenwart. ( Jenaer Germanistische Forschungen 11). Jena:
Frommsche Buchhandlung.
GIVNER, DAVID A. 1962 "Scientific Preconceptions in LOCKE'S Philosophy of Language".
Journal of the History of Ideas 23, 340-354.
GONSALVES, RENNIE 1991 "LOCKE'S Definitional Semantics, Condillac's Critique, and the
Consequences". The Seventh LACUS Forum 1990, 254-263. Ed. by Angela Della Volpe.
Lake Bluff: Linguistic Association of Canada and the United States.
HOBBES, THOMAS (1588-1679) 1968 [1651] Leviathan. Ed. with an introduction by C.B.
Macpherson. Harmondsworth: Penguin.
HARRIS, JAMES ( 1709 - 1780) 1967 (1751) Hermes: or a Philosophical Inquiry Concerning
Language and Universal Grammar. Menston: Scolar Press.
HORNE TOOKE, JOHN (1736-1812) 1968 [1786/1798]. EPEA PTEROENTA or, The Diversions of
Purley. 2 vols. Menston: Scolar Press.
HOWELL, WILBUR SAMUEL 1971 Eighteenth Century British Logic and Rhetoric. Princeton:
Princeton University Press.
HÜBNER, WALTER 1920 "John LOCKE als Sprachphilosoph". Die Neueren Sprachen 27, 385-
412.
HÜLLEN, WERNER 1989 `Their Manner of Discourse'. Nachdenken über Sprache im Umkreis der
Royal Society. Tübingen: Gunter Narr.
JOHNSON, Samuel 1967 [1755] A Dictionary of the English Language. 2 vols. New York: AMS
Press.
JONES,RICHARD FOSTER 1932 "Science and Language in England of the Mid-Seventeenth
Century". The Journal of English and Germanic Philology 31: 315 - 331.
KRETZMANN, NORMAN 1976 "The Main Thesis of LOCKE'S Semantic Theory". History of
Linguistic Thought and Contemporary Linguistics. Ed. by Herman Parret. Berlin: de
Gruyter.
LAND, STEPHEN K. 1986 The Philosophy of Language in Britain from Hobbes to Reid. New
York: AMS Press.
LEYDEN, W. VON 1969 "What is a Nominal Essence the Essence of?". John LOCKE: Problems
and Perspectives. A collection of new Essays ed. by John W. Yolton. 1969, 224-253.
Cambridge: Cambridge University Press.
LINNELL, JOHN 1956 "LOCKE'S Abstract Ideas". Philosophy and Phenomenological Research
16, 400-405.
LOCKE, JOHN (1632-1704) 1975 [1690] An Essay Concerning Human Understanding. Ed. by
P.H. Nidditch. Oxford: O.U.P.
LOCKE, JOHN 1975 [1690] An Essay Concerning Human Understanding. Ed. by P.H. Nidditch.
[Oxford Electronic Text Library]. Oxford: O.U.P.
LOCKE, JOHN 1961 (1690). An Essay Concerning Human Understanding. 2 vols. With an
introduction by John W. Yolton. London: Dent & Dutton.
MACKIE, J. L. 1976 Problems from LOCKE. Oxford: Clarendon Press.
MCCOLLEY, GRANT 1938 "John Wilkins: A Precursor of LOCKE". Philosophical Review 47, 642-
643.
MONBODDO, JAMES BURNETT, LORD (1714-1799) 1967 [1773 -1792] Of the Origin and Progress
of Language. 6 vols. Menston: Scolar Press.
O'CONNOR, D. J. 1952 John LOCKE. Harmondsworth: Penguin.
PARRET, HERMAN 1975 Idéologie et sémiologie chez LOCKE et Condillac. La question de
l'autonomie du langage devant la pensée. Amsterdam: Benjamins.
26
REININGER, ROBERT 1922 LOCKE, Berkeley, Hume. München: Ernst Reinhardt.

ROBINSON, RICHARD 1954 Definition. Oxford: Clarendon Press.
RYLE, GILBERT 1968 "John LOCKE on the Human Understanding". LOCKE and Berkeley. A
Collection of Critical Essays. Ed. by C.B. Martin and D.M. Armstrong. 1968 , 14-39. London,
Notre Dame: University of Notre Dame Press.
SALMON, VIVIAN 1979 The Study of Language in 17th-Century England. Amsterdam:
Benjamins.
SCHOULS, PETER A 1980 The Imposition of Method: A Study of Descartes and LOCKE.
Oxford: Oxford University Press.
SCHREYER, RÜDIGER 1971 Untersuchungen zur Sprachauffassung Dr. Johnsons. Diss.
Saarbrücken.
SCHREYER, RÜDIGER 1990 "Of General Terms. On Language and Mind in 18th-Century
Linguistic Theory". History and Historiography of Linguistics. Proceedings of the Fourth
International Conference on the History of the Language Sciences (ICHoLS), Trier, 24 - 27
August 1987, 399-415. Amsterdam: Benjamins.
SLEDD, JAMES H. and KOLB, GWIN J. 1955 Dr. Johnson's Dictionary. Essays in the Biography of
a Book. Chicago: The University of Chicago Press.
SPECK, JOSEF (ed.) 1979 Grundprobleme der großen Philosophen. Philosophie der Neuzeit 1.
G”ttingen: Vandenhoek & Ruprecht.
SPRAT, THOMAS (1635 - 1713) 1667 The History Of the Royal-Society of London, For the
Improving of Natural Knowledge... London: J. Martyn and J. Allestry.
STARNES, DE WITT T. and NOYES, GETRUDE E. 1991 [1946] The English Dictionary from
Cawdrey to Johnson 1604 - 1755. Reprint of the 1946 edition, updated with an
introduction by Gabriele Stein. Amsterdam: Benjamins.
WALES, KATHLEEN 1985 "Johnson's Use of Synonyms in Dictionary and Prose Style: The
Influence of John LOCKE.". Prose Studies No.8, 25-34. London: Frank Cass.
WARD, SETH and WILKINS, JOHN 1654 Vindiciae Academiarum. Debus 1970:193-259.
WILKINS, JOHN (1614 - 1672) 1968 [1668] Essay Concerning a Real Character and a
Philosophical Language. Menston: Scolar Press.
YOLTON, JOHN W. 1956 John LOCKE and the Way of Ideas. Oxford: Oxford University Press.
YOLTON, JOHN W. 1968 "LOCKE'S Concept of Experience". LOCKE and Berkeley. A Collection
of Critical Essays. Ed. by C.B. Martin and D.M. Armstrong. 1968, 41-52. London, Notre
Dame: University of Notre Dame Press.
YOLTON, JOHN W. (ed.) 1969 John LOCKE: Problems and Perspectives. Cambridge:
Cambridge University Press.
YOLTON, JOHN W 1985 John LOCKE. An Introduction. Oxford: Blackwell.
ZOBEL, ARTHUR 1928 "Darstellung und kritische Würdigung der Sprachtheorie John LOCKEs".
Anglia 52, 289-324.
View publication stats

1992 Lockeondefinition

Uploaded by

Copyright:

Available Formats

1992 Lockeondefinition

Uploaded by

Document Information

Original Title

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

1992 Lockeondefinition

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

The Deﬁnition of Deﬁnition in John Locke’s Essay Concerning Human

Chapter · January 1992

The user has requested enhancement of the downloaded file.

THE DEFINITION OF DEFINITION IN LOCKE'S

2. The wider context............................................................................ 3

3. A KWIC view of Locke's use of the term definition ............................... 5

4. Words and Ideas ............................................................................. 6

5. The private and public uses of language ............................................. 8

5.1 Civil use .................................................................................... 9

5.2 The philosopher's dilemma............................................................ 10

6.1 What definition is good for ......................................................... 11

6.2 What definition is ...................................................................... 12

6.3 Two types of definition .............................................................. 12

6.4 Why simple ideas cannot be defined ............................................ 13

6.5 Why the definition of complex ideas is difficult.............................. 15

6.5.1 Defining mixed modes ......................................................... 16

6.5.2 Defining substances............................................................. 17

6.6 Definitions and trifling propositions ............................................. 19

7. Against a philosophical language ..................................................... 20

Appendix A: Selections from a KWIC Concordance ................................ 23

1. Definition of .... .......................................................................... 23

2. Signification of ... ....................................................................... 23

One might be tempted to speculate that LOCKE'S epoch-making work relieved

2. The wider context

of the purpose, structure, achievements, difficulties and limits of definition. Finally I

3. A KWIC view of Locke's use of the term definition

LOCKE'S ESSAY CONCERNING HUMAN UNDERSTANDING is available on magnetic disk

(ADJ) (ADJ) (N)

In a right-sorted concordance the term signification occurs in the same contexts

4. Words and Ideas

General terms/names are signs of general ideas.

5. The private and public uses of language

5.1 Civil use

A word can thus be seen as an aide de mémoire or as a vehicle of

5.2 The philosopher's dilemma

Needless to say, it is the philosophical use of language that LOCKE is concerned

The complaint that words, albeit necessary, can be an impediment to science is

6.1 What definition is good for

6.2 What definition is

Now we know what definition is good for, but what is it?

6.3 Two types of definition

The traditional method of definition was by genus and differentia. LOCKE

The distinction between simple and complex ideas is crucial to an

6.4 Why simple ideas cannot be defined

LOCKE'S concept of simple ideas is open to interpretation. The discussion in the

abstraction, the ultimate indivisible result of comparing and abstracting from

Whatever the philosophical problems inherent in LOCKE'S concept of simple

6.5 Why the definition of complex ideas is difficult

The suggestion that complex ideas may be decompounded into a series of

In Lockian semantics definition will be successful if the hearer holds these

6.5.1 Defining mixed modes

Complex ideas of mixed modes (murder, adultery) have no archetype in nature.

identify it to a hearer in order to make him have that same idea.

6.5.2 Defining substances

The difficulty in defining names of substances is that to different speakers these

In philosophical use a definition of our ideas is insufficient where ideas

6.6 Definitions and trifling propositions

Whenever the whole or a part of a complex idea is affirmed of itself, we have a