Logic and Logical Philosophy

Volume 15 (2006), 217–237

Maria K. Timofeeva

INTROSPECTIVE AND TRADITIONAL

VIEWS OF LANGUAGE

Abstract. The present-day traditional view of language has an essentially

pedagogical background inherited from antiquity. Certain features of this
heritage have passed through the centuries, reached our days and continue
to be a sort of implicit postulates penetrating almost into every scientific
conception of language. This causes specific divergences between the scien-
tific view of language and its introspective view, i.e., the way it is actually
conceived by an individual during ordinary communication. Studying those
divergences is important for a better grasping the nature of natural language
and for a more adequate approaching to certain problems.

Keywords: introspection, ordinary communication, proposition, truth value,

natural language.

1. Introductive remarks

What do we really conceive when we conceive the content of some text during
our ordinary communication? Do we generally conceive the senses of sepa-
rate words? Do we in fact comprehend a proposition as some well-defined
part of content separated from its pragmatic constituents? Do we really dis-
tinguish truth and falsity as some external characteristics of a proposition?
We consider these questions in sections 3, 4; this allows us to see the
core of the divergences between introspective and traditional (pedagogically-
based) views of language and to depict the essential principles of the intro-
spective framework. The outline of certain introspective theories and con-
siderations is given in Section 5. The whole reasoning about this approach
is summarized in Section 6.

Received September 22, 2006; Revised Januar 17, 2007

218 Maria K. Timofeeva

2. Traditional view of language

It should be noted that a person learning a foreign language for communica-

tive purposes generally gets to know only the introductory part of scientific
linguistic knowledge. That is why the corresponding pedagogical activity
should be called propaedeutic. It has also been propaedeutic in historical
perspective: it was this very activity that marked the beginning of the sci-
ence about language, linguistics.
The present-day scientific view of language originated in antiquity when
the sciences concerning human languages (primarily logic and linguistics)
arose as normative ones. In ancient Greece linguistic investigations were
driven by the aim of teaching and included two lines of pedagogic activity.
First, it was necessary to teach foreigners Greek and Latin thereby assur-
ing mutual understanding essential for successful trade relations. The first
prominent ancient center of such practices was in Greek Alexandria. Led
by the above pedagogical and substantially propaedeutic purpose, ancient
linguists—grammarians, as they were called at that time—elaborated the
normative rules governing the processes of building grammatically correct
texts. By what sort of rules were ancient grammarians constrained? They
sought to represent a language as a whole, for the pupils were to acquire
a complete mechanism of producing and understanding correct texts. The
number of intelligible sentences which can be potentially produced by peo-
ple is infinite. But the pedagogically-aimed representation of those sentences
should be finite. That is why those representations inevitably took the shape
of a certain calculus-like (albeit informal) system including (1) classes of ba-
sic entities, features, relations (words, parts of speech, syntactic functions,
etc.) and (2) (syntactic, semantic, pragmatic) rules governing the processes
of constructing complex texts from those basic objects. Moreover, for the
same reason (i.e. finiteness of representation) ancient scientists were com-
pelled to posit the shared intelligibility of senses associated with the above
basic objects and with complex texts. It is this presumed shared intelligibil-
ity that can allow different people to understand each other during ordinary
communication.
Second, at about the same time Greek philosophers paid their attention
to the contents of the texts and tried to understand the laws of valid reason-
ing (as logicians did) or the means of deceptive reasoning (as logicians and
especially sophists did) in order to teach people the art of (valid or mislead-
ing) argumentation. The ancient logicians worked out the systems of rules
governing the processes of arranging senses.
 Introspective and Traditional Views of Language 219

On the whole, ancient logicians and linguists elaborated the principles

of speaking and reasoning which were suitable for teaching people to speak
and to reason. In the subsequent centuries the theories of language were
developed in many important aspects but at the same time their pedagog-
ical bent was largely retained. This was strongly stimulated by the fact
that during those centuries the pedagogically-based framework was in good
accordance with the predominant areas of its application—teaching contem-
porary languages; conserving the intelligibility of ancient texts for posterity;
historical, comparative, logical and structural studying of languages. As
a matter of fact those applications were based on viewing language data
through a conceptual scheme based on the pedagogical view. This way of
treating languages is still taken as traditional.
Thus, the nowadays traditional (pedagogically-based) approach consists
in representing a language as a whole, as a certain system or mechanism
which somehow underlies or governs a process of language activity.
Another guiding line of traditional framework stems from the fact that
natural language is seen as some non-individual object with shared accessi-
bility. For viewing a language as an impersonal shared thing the following
concepts are substantial:
1. Propositions, their structure (logical form), their constituents (words/
phrases and their senses);
2. Judgments about truth (falsity) of a proposition;
3. Presuppositions;
4. Associated ideas.
The basic features of the notion of proposition (see, e.g., [1]) have been in-
vestigated by G. Frege, K. Ajdukiewicz, B. Russell and other authors. The
proposition is traditionally considered as a thought explicitly expressed by a
sentence. It is a proposition, rather than a sentence that can be true or false.
The proposition is an abstraction, an element of scientific framework, espe-
cially of logic and linguistics (this matter is discussed in detail by J. Peregrin
[8]). The proposition is not considered as a subjective (and hence unique)
content conceived by an individual mind; instead, it is posited as a shared
entity accessible for different minds.
The proposition can have a structure, or a “logical form”. The structured
proposition can have constituents—senses of words and/or phrases.
The truth value is traditionally thought of as a certain external feature
of a proposition, it adds to a proposition the information about its relation
to a certain state of affairs.
220 Maria K. Timofeeva

The proposition cannot be meaningful (either true or false) unless its

presuppositions are true. The presupposition is also a certain thought, i.e.,
it can be represented by another proposition.
There is one particular condition a statement must fulfill to be meaning-
ful: it is necessary to conceive a question which the statement answers, to
see what the utterance has been made for; this allows one to assess whether
the utterance suits the purpose and the circumstances of its usage. As
R.G. Collingwood [2] puts it, a statement is not meaningful unless it is clear
what question it is due to answer. Let us call this question the hidden ques-
tion for the sentence. It seems natural to consider the hidden question as a
sort of presupposition.
The proposition and its constituents can have some associated ideas (also
representable by propositions). Presuppositions can be thought of as special
cases of associations but we would treat them as a separate class. Among
the ideas associated with a certain proposition there can also be some con-
sequences inferable from this proposition.
Thus the concepts mentioned in (1)–(4) are considered by science as cer-
tain abstractions. Their denotata are traditionally thought of as commonly
accessible “pieces of sense” equally intelligible for different people. However
everyone knows from his own experience that each language is available only
for an individual mind and is thereby subjective (and only on this basis it can
exist as a social institution). This contradiction was explicitly formulated
in the 19th c. by Wilhelm von Humboldt as the antinomy of objectivity
and subjectivity in language. Really we do not conceive those “pieces of
sense” during our ordinary communication; no one can find in his mind any
direct evidence for their accessibility: an individual mind has only subjective
contents, available to this very individual; and some of those contents are
postulated (explicitly or implicitly) as representatives of certain commonly
accessible or objective entities.
Each of us associates with a text a certain but a unique content. A com-
monly accessible content or identity of individual contents (or their parts) for
different people can only be hypothetical; nobody can fix this by introspec-
tion, nobody can be absolutely sure about such hypotheses (one can easily
learn this from misunderstandings occurring in ordinary communication).

3. Introspective view of language

Let us consider a text U uttered (or heard) by a person N at a moment t

during an ordinary communication; and let C be a content of U conceived
 Introspective and Traditional Views of Language 221

by N at a moment τ . Let us label the above abstract “pieces of sense” (1)–

(4)—i.e., the proposition expressed by U , judgment about its truth value, its
presuppositions, its associative ideas—as P , V , PP, A, respectively. Thus,
C, as well as P , V , PP, A, are understood as temporal (not timeless) entities.
Here C is thought of as radically different from P : the former is indi-
vidual, while the latter is commonly accessible; C includes all the contents
connected with U by N at τ : thoughts, assessments, perceptions, feelings,
volitions, images—everything that is conceived by N as constituting the un-
derstanding of U . All the so-called material objects mentioned within U are
first of all the objects of N ’s perception, i.e., the constituents of N ’s mind,
since the real world is given to us only through our perception. Thus, within
the introspective stance (i.e., phenomenologically), the so-called “external”
(or “material”) and “internal” worlds are—for each individual—first of all
the contents of his mind; and the utterance U , since it is perceived by the
individual, is also first of all the content of his mind.
Besides, within introspective perspective we do not speak about mean-
ings and senses, we do not even make any distinctions between those two
kinds of things. In ordinary communication we conceive the content of an
utterance as a single whole; we do not conceive separately any senses for
connecting them with meanings. This wholeness also forces us not to rely
on the principle of compositionality: we do not conceive contents as some-
how constructed from simpler senses. Introspectively, it is not clear, how the
contents of words/phrases participate in the creation of contents ascribed
to the sentences containing those words/phrases. We have no introspective
reasons to insist that a content of some word (e.g. “tree”) coincides with a
certain part of the content ascribed to a sentence containing this word (e.g.
“It is a tree”).
Within the traditional view the proposition P may be considered as
structured (e.g. constructed from the senses of the words) but this doesn’t
mean that C is also structured, or (if it is) that its structure is somehow sim-
ilar (e.g. isomorphic) to the structure of P . It may happen that P is a struc-
tured proposition, most adequately representable in first-order logic by a
certain predicate formula, while C is a non-structured integral content, most
adequately representable by a term. As one can find out by his own intro-
spection and by studying other people we do not conceive (during ordinary
communication) the senses of single words nor do we consciously combine
them in more complex senses. We do not even conceive any clear boundaries
between the mentioned “pieces of sense”. Only as logicians or linguists we can
judge to which of them it would be better to ascribe one or another part of C.
222 Maria K. Timofeeva

It would be too much to insist that a given speaker does perceive the
content C in the same way as we do, say, colors, sounds or other qualita-
tive (phenomenal) experiences, so-called qualia. Although some philosophers
(e.g. G. Strawson) believe that to grasp the content C means to grasp certain
qualia, i.e., that the nature of C is qualitative, a large number of other schol-
ars (e.g. M. Tye) do not think so: they suppose that this content can include
some phenomenal components but they are constituted only by the present
and/or past perceptions, feelings, emotions and, in sum, never amount to
the whole content as such [13]. Although there may be other arguments in
favor of both positions, there is no indubitable fact that could give us an
unambiguous evidence to make us prefer one of them. Besides, it is generally
accepted in psychology that no one can directly observe his own thoughts
(and therefore the content C) at the very moment of their presence in his
mind. But in any case, it is doubtless that we do somehow conceive the
contents of texts. Otherwise what do we point at by asking “Do you really
think so?” (or “Have you really meant this?”, or “That is what you have
told me an hour before!”, or “But in fact you have not told me that!”) if
not to the content of some previously uttered text? If there were no such
contents those sentences would be senseless. The previously uttered text is
a sort of a (temporal) name, indicating certain content and thereby allowing
one to point at this content by different means, e.g. by the above sentences.
It is possible to state the introspective principles of dividing C into parts
which can be considered as introspective counterparts or prototypes of P , V ,
PP, A. Let us label those parts C P ,C V , C P P , and C A and call them proto-
typical proposition, prototypical judgment about its truth value, prototypical
presuppositions, and prototypical associated ideas, respectively. (Those parts
are not considered as self-dependent, they exist only within C.)
Prototypes of a proposition and its presuppositions should be thought
of as obligatory parts of C: their presence cannot be disregarded by those
who have understood the utterance. We cannot make sense of a prototypical
proposition unless its prototypical presuppositions are true: every content
is understandable only due to some presupposed data. That is, if N con-
ceives C P , he inevitably conceives also C P P too; C P P penetrates into C,
so to speak, “automatically”. However, disregarding other prototypes may
be compatible with a successful understanding of U . For example, some
thinkable ideas, which might be associated with P , can pass unnoticed and
not conceived by N at the moment τ ; thereby C A would not include the
corresponding prototypes.
 Introspective and Traditional Views of Language 223

We do not really conceive P , V , PP, A and can only claim to conceive

their prototypes. Studying those prototypes constitutes the introspective
view, to which we shall now switch our attention.
Within the introspective framework we are not bound by the traditions
intrinsic to the pedagogically-based stance and are to divide C in a way
plausible from the point view of our introspection. It is reasonable to think
that the division comes along the lines separating

I. the ideas expressed by U explicitly from those which are not expressed
explicitly but can enter into C only implicitly: the prototypical propo-
sition is expressed explicitly, the other parts of C are only implicit;
II. the necessary and accidental parts of C: the prototypes of the propo-
sition and the corresponding presuppositions are necessary, while the
prototypes of associations are accidental;
III. the ideas, making it possible to assess U as (in)adequate for the cir-
cumstances of its usage, from the other constituents of C: it seems
natural to qualify the former as the prototypical judgment about the
truth value.

For example let the text U be the sentence:

It has stopped raining now. (1)

In the subsequent examples we shall use English sentences for referring to

the parts of C. In those cases a sentence will be enclosed in “corners” (‘h’,
‘i’) in order to distinguish it from an ordinary usage of the sentence during
communication. It should be once more stressed that in such a case we do
not consider the sense of the sentence in “corners” to be constructed from
simpler senses (e.g. from the senses of words), we treat it as a single whole.
Nor is the sense of such a sentence considered to be self-contained, it is an
inseparable part of C; it is not even thought (within C) to be well articulated
and may be conceived only vaguely.
What does N precisely conceive when he conceives the content C of the
utterance (1)?
At this moment N can conceive prototypical ideas like hIt has been rain-
ing recentlyi (presupposition), hThe speaker reports about the observable
world and is not a subject to hypnosis nowi (the reason for assessment),
hIt isn’t necessary to put on the raincoati (consequence), hThis summer is
rather rainyi (another associated idea). All those ideas are not expressed by
224 Maria K. Timofeeva

the utterance in question, but they can be conceived at the time of compre-
hending U as parts of the content C. While conceiving C, the person N does
not conceive those groups of ideas as separate ones, i.e., he does not really
conceive in C the boundaries between the above-mentioned prototypes, he
conceives C as a single whole. But he can—due to his introspection—divide
the content C along the lines (I)–(III).
We shall use (I)–(III) for determining C P , C V , C P P , and C A . Those
principles of dividing C seem to be in accordance with the most abstract
features of a proposition, judgment about its truth value, presuppositions,
and associated ideas.
Now it will be useful to compare the traditional approach with the in-
trospective one.
In fact switching from the pedagogically-based view to the introspective
one means changing the way of conceptualizing and choosing the language
data. Therefore it will be helpful to compare the way used within the tradi-
tional approach with that suitable for the introspective one. We shall state
the difference by appealing to the framework of first-order logic. If we are to
use first-order logic for describing those two approaches, the crucial point is:
how to determine the intended universes and standard interpretations. The
two ways of conceptualization differ in their answer to this question. Within
the traditional framework we use the universe which is constituted by the
entities thought of as commonly accessible for many people (propositions,
their truth values, etc.); within the introspective framework—by individu-
alized entities thought of as unique for each human (contents of a certain
individual mind). Thus, the basic differences between the traditional view
and the introspective one can be summarized as follows:

1. the former’s universe is thought to be a set of hypothetical commonly

accessible (usually timeless) senses, while the latter’s is thought to be a
set of directly observable but fundamentally individual temporal contents;
2. the former’s universe includes the whole arsenal of those potentially pos-
sible (for a language in question) senses which are basic for representing
the language in its entirety as a certain autonomous system (of entities,
rules, relations, etc.); the latter’s includes only those contents which cor-
respond to separate acts of understanding/producing the utterances and
are not supposed to constitute a whole system;
3. in other words, the former interprets terms as abstract constituents of a
certain total system, the latter—as contents of a certain individual mind;
 Introspective and Traditional Views of Language 225

4. in the former a predicate formula can be interpreted as rendering a sense

of a sentence; in the latter (i.e., within an introspectively plausible de-
scription of a language) only a term can be interpreted as a content of
a sentence (a predicate formula can be interpreted here only as denoting
a (meta)relation between a given content and some other contents of an
individual mind, e.g. some decisions or contents of some other sentences).
Those are the most essential features of the introspective view. For more
detailed reasoning about this approach see [12].

4. Truth and falsity as internal features of content

We have distinguished in C the nonintersecting parts C P , C P P , C A . Which

of them includes C V ?
First of all we should draw distinctions between “truth” or “falsity” as
logical notions and their prototypes in C. The former concepts do exist
within the framework of symbolic logic and are applied to propositions; the
latter (i.e., perceiving something as true) exist in an individual mind. We
shall speak about the latter. Among the most well-known theories of truth
(see, e.g., [9]) none of them corresponds to this kind of truth/falsity and
thereby suits the introspective theory of ordinary communication.
In an ordinary dialogue we do not necessarily conceive two separate
things:
(a) the prototype of a proposition and (b) its evaluation as true or false.
According to our introspection truth (or falsity) is not something applied
to a proposition. How could we reconcile this vision with the tradition
of formalizing truth and falsity as external characteristics of propositions?
Where is the prototype of truth/falsity located?
Let us consider another variant of the utterance U :

It has really stopped raining now. (2)

Here (2) can be understood as informing about the truthfulness of (1).

What is the difference between (1) and (2)? (Or, in more general form,
between “Q” and “Q is true”?)
Usually, during an ordinary communication, we do not actually express
nor conceive our judging about truth in a form similar to Tarskian T-sentence
[11], i.e., when we utter (1), we don’t evaluate the truthfulness of this sen-
tence according to the rule
226 Maria K. Timofeeva

‘It has stopped raining now’ is true if and only if

it has stopped raining now

and therefore we do not conceive the content of this utterance as a kind of

syllogism

The sentence ‘It has stopped raining now’ is true if and only if
it has stopped raining now.
But it has stopped raining now.
Then the sentence ‘It has stopped raining now’ is true.

In this case the difference between (1) and (2) would be unclear.
The T-sentence constitutes one of the ways of formalizing our ordinary
notion of truth but not the way in which we conceive it in C. It seems that
within the introspective framework the difference between the utterances (1)
and (2) lies in the difference between the location of C V : in the first case
it is located in the prototypical presupposition C P P , in the second case—in
the prototypical proposition C P .
Let us now return to (1). Here C P P includes the information which
is necessary for understanding C P . This presupposed information tells us
(among other things) that the utterance reports a real (observable) fact.
Thus CP P includes the idea which can be indicated by the sentence

hThe utterance reports a real (observable) facti. (3)

Such an idea is included neither in P , nor in C P . The very fact which

CP reports (the absence of rain) is not “visible” from (3), this prototypical
presupposition states only the mode of its existence (as a real, not imagined
fact).
During our everyday communication we are usually interested not in
the truth per se but in that very fact which we suppose a true sentence
reports. And the utterance informs us about this very fact, not about its
own truth. Therefore here truthfulness is tantamount to conceiving some
kind of “presupposed existence”.
Thus, for those who consider whether truth is a substantive or a relational
notion (see, e.g., [10]) the introspective approach gives reasons in favor of
the former: here truth is thought of as a certain idea, as a content of one’s
mind (or, ultimately, a state of one’s mind), i.e., as a certain substance, not
a relation.
 Introspective and Traditional Views of Language 227

Actually, if we remove (3) from the content of the sentence (1) the re-
mainder would not yield the understanding of this sentence because it will
not be clear what kind of facts it is supposed to be about.
Thereby, the prototypical propositions expressed by (1) and (2) are not
equal, and thus nothing gives us any right to think that hit has stopped
raining nowi from (1) is interchangeable with hit has stopped raining nowi
from (2). In fact it is not the invariant common part of the contents ascribed
to (1), (2). In the first case the speaker is interested in the very situation;
strictly speaking, it is the answer to the question like ‘What is being observed
now?’, it is not the question about ‘is’. In the second case the speaker is
interested in the reality of this situation; it is the answer to the question
‘Is this situation really being observed now?’. Thus these two texts have
different contents; their contents have different prototypical propositions and
are thought of as the answers to different questions.
When we utter (1) the assertion about the location (in one’s mind, in
the observable world, in an imagined world, etc.) lies in the area of C P P ,
when we utter (2) it lies in the area of C P . And this appears to be the
only substantial difference between (1) and (2). In ordinary communication
we do not usually discover the truthfulness (by comparing the prototypical
proposition with some state of affairs), but presuppose it.
Now let us suppose that U is false. In what case would N , for example,
conceive the sentence ‘It is raining now’ as false? It happens when the
adequate prototypical presupposition appears to be absent and C contains
a prototypical presupposition about some other—wrong—location (if N is
mistaken or lies). It is impossible to find the rain (or Q) ‘now’ due to its
incompatibility with the presupposed wrong location. There appears to be
no situation which can be indicated by uttering this sentence; this fact makes
it false. And really, in such a case the reaction will often be: ‘You are wrong,
where did you seen the rain?’. It is important that introspectively the non-
existence of a situation compatible with this sentence is seen directly, it is
not yielded by any conscious inference. The failure of designation constitutes
the corresponding truth value, and this failure is seen directly.
Truth/falsity does not characterize the content of an utterance from out-
side (relative to some state of affairs), it is a constituent of this very content,
its internal feature.
To elucidate the introspective perspective let us look at a well-known
problem connected with the notions of truth/falsity, namely, at that of in-
terchangeability. The following reasoning should be considered not as solving
this problem (in the form of its usual definition within the traditional frame-
228 Maria K. Timofeeva

work) but as depicting the situations prototypical for this problem within the
introspective framework. The point is that this problem does exist within
the traditional approach but it does not generally arise within ordinary com-
munication: we usually have no difficulties with interchangeability. How
does the difference between introspective and traditional views explain this
discrepancy?
Suppose that N utters the following sentence:

Cervantes and Shakespeare are contemporaries. (4)

Since we know that Cervantes is the author of “Don Quixote”, we can form
the sentence:

The author of “Don Quixote” and Shakespeare are contemporaries. (5)

Is it possible to conclude that it does not matter which utterance was as-
cribed to N : (4) or (5); that (4) can be equally replaced by (5)? Within
the traditional approach it is usually thought that we cannot do so (if this
sentence is considered de dicto) because N may be fully ignorant of the
authorship of “Don Quixote”. However, both expressions indicate the same
person and thus should be interchangeable. This fact is treated (by G. Frege,
S. Kripke, W.V. Quine and other authors) as puzzling. But let us look at
the situation from the introspective standpoint and compare the two cases:
in the first N utters (4), in second (5).
Generally we cannot speak about substituting Y for X unless we are
sure that this operation is applied within some unaltered context, i.e., the
context of X will not be changed after replacing X by Y. But it does not
seem to be the case. Let us take X to be hCervantesi and Y to be hthe
author of “Don Quixote”i. After replacing, the rest of C should remain the
same. But in (4) X is included in the prototypical proposition while Y only
may be among the prototypical associative ideas (i.e., X is the necessary part
of C, while Y only may be its accidental part); in the second case it is vice
versa: Y is contained in the prototypical proposition, X—(may be) among
the prototypical associated ideas. How could it be possible to substitute the
prototypical proposition for the prototypical associated idea or vice versa?
This operation doesn’t meet the above-mentioned condition and hence it
cannot be though of as warranted.
 Introspective and Traditional Views of Language 229

5. The proponents of the introspective view

The introspective and traditional views are complementary, each of them

serves its own purposes. At the same time the former should not be seen as
opposite counterpart to the latter: the latter is so deep-rooted that it is dif-
ficult to notice its disagreement with introspective observations. Still, there
are proponents of the introspective view both in logic and in philosophy. An-
drzej Grzegorczyk proposes to interpret certain antinomies in a distinctly
introspective spirit. Robin G. Collingwood has developed a rather detailed
introspective theory of language. Even some ideas of Gottlob Frege—who
is often ranked among the founders of antipsychologism in logic—are con-
sonant with the introspective stance. As for Ludwig Wittgenstein, I doubt
whether he could be placed among the adherents of the introspective view
although some parts of his reasoning can seem similar to it.
A. Grzegorczyk [4] claims that paradoxes can be caused by inadequacy
of the conceptual apparatus. He discusses self-referential sentences and il-
lustrates his ideas by two well-known examples: the Liar paradox and the
Grelling-Nelson Heterological paradox.
Let us, for instance, consider a situation of lying. Suppose somebody
says: ‘The sentence I said yesterday is false’. Such refutation of the idea
formulated earlier is quite normal. But in the Liar paradox we have another
situation. Here a speaker rejects a sentence which has not been formed yet.
That is why this utterance is not a judgment; it is only an intention to judge,
a declaration about a certain future action. But an intention is not equal
to an action: only some intentions become actions. Similarly, only some
declarations are realizable. For example, the declarations proposed in the
Crocodile paradox or in the Barber paradox appear to be unrealizable.
Why have logicians been so long uneasy about the Liar paradox? Grze-
gorczyk thinks that there are two reasons for that. First: it is very hard
to analyze mental processes; this task requires a refined skill of introspec-
tion. Second: senses and meanings of linguistic expressions are traditionally
regarded as ideal entities that exist independently of speakers and their feel-
ings. This statement is misleading: those entities, just as truth or falsity,
are only relative; they depend on the circumstances of their usage. A word
can also refer to a certain thing only relatively. Thus, we should say
A person X uttering a word Y refers it to a thing Z
instead of the traditional
A word Y refers to a thing Z.
230 Maria K. Timofeeva

Surely we do not think about this relativity every time, at every moment of
an ordinary dialogue. But we should keep it in mind.
In his summary Grzegorczyk challenges antipsychologism:
Semantic antinomies seem to be consequences of antipsychologistic
paradigm adopted by logicians at the beginning of 20 Century. We
shall loose more radically from troubles caused by antinomies if we dis-
card antipsychologism which is, in effect, a huge simplification of the
description of semantic situations. [4, p. 126]
Introspective view is a sort of psychologism (if we mean by this word all the
appeals to the psychological reality), and, as we have seen earlier, a shift
from antipsychologism towards psychologism can give us a renewed vision
of certain language problems.
G. Frege is traditionally considered to be a proponent of antipsycholo-
gism, but his own thoughts are not so one-sided. Let us consider some of
his ideas without going too deep into the heart of the confrontation between
psychologism and antipsychologism.
In connection with the sign Frege speaks about the reference, the idea,
and the sense. The reference is “an object perceivable by the senses”. (We
can note here that the act of perception can be only individual, that is why
one can have only individualized subjective information about the object of
reference.)
The idea is subjective: my idea is an internal image, arising from my
memories of sense impressions which I have had and acts, both internal
and external, which I have performed. Such an idea is often saturated
with feeling; the clarity of its separate parts varies and oscillates. The
same sense is not always connected, even in the same man, with the
same idea. [5, pp. 25–26)]
Thus, feelings, impressions, all aspects of individual experience may enter
into the content of a text. This part of Frege’s reasoning is compatible with
the introspective view.
Still, one should distinguish the sense of a sign from the associated idea.
When Frege says that a sense “may be common property of many and there-
fore not a part or a mode of the individual mind” [5], that it is “not the
subjective performance of thinking but its objective content” [5, p. 28], he
moves away from the introspective view.
Hence, Frege does not wholly deny the subjective nature of the content of
a text, but he implicitly postulates the existence of the objective constituent
of this content and is interested in investigating this very constituent. Grze-
gorczyk thinks that postulating such objective entities is risky.
 Introspective and Traditional Views of Language 231

Reflections about the psychological aspects of content can be found in

other works by Frege, for example, in his Logic.
As a matter of fact, Frege constructs a metalanguage suitable for the
purpose of his investigations. But, as Grzegorczyk shows, this may give
rise to antinomies caused precisely by the metalanguage, not by the object
language (i.e., not by the ordinary language of human communication).
Hence, two statements can be used for characterizing the content of a
text:
(a) the content of a text is conceived by an individual as subjective,
(b) within a successful communication a text has some objective sense.
It is not easy to co-ordinate those statements with each other. Hence
a researcher should give preference to one of them. The crucial question
is: are we interested in investigating individual understanding or mutual
communication? For exploring the former we should accentuate (a) and
consider the content as subjective, for exploring a latter we should accentuate
(b) and postulate the existence of objective senses. Frege chooses the latter.
At the same time, it is not necessary to do so and there are researchers
who give preference to the former. Among them is R.G. Collingwood. We
can never be completely sure that within a certain communication both of
its participants conceive one and the same content of a text or of a store of
texts. Thus, we are not obliged to postulate the existence of common senses.
As Collingwood puts it, there can
never be any absolute assurance, either for the hearer or for the speaker,
that the one has understood the other. [. . . ] The only assurance we
possess is an empirical and relative assurance, becoming progressively
stronger as conversation proceeds, and based on the fact that neither
party seems to the other to be talking nonsense. [3, p. 251]

Frege’s reasoning is partly based on considerations consonant with the

introspective stance, but he is interested in mutual communication and de-
ducibility, therefore he accentuates (b) at the expense of (a). He explicitly
states it, for example, in his Logic. Of course, this interest and this decision
make his theory different from the introspective perspective. For instance,
he considers truth (falsity) as something added to a proposition. In [6] he
introduces the notations −A and ⊢ A: the content of A and the assertion of
this content respectively. Thus, for instance, when A is “Opposite magnetic
poles attract each other”, “−A will not express this judgment, but should
simply evoke in the reader the idea of the reciprocal attraction of opposite
magnetic poles” [6, p. 112], perhaps for deriving some conclusions from it.
232 Maria K. Timofeeva

The horizontal stroke, which is part of the symbol ⊢, ties the symbols which
follow it into a whole; “and the assertion, which is expressed by means of the
vertical stroke at the left end of the horizontal one, relates to this whole”
[6]. Here Frege also raises an objection against the traditional notions of
subject and predicate. He believes that the subject contains the whole con-
tent, and the predicate serves only to present this as a judgement. There
is only one predicate common for all judgements (namely, “is a fact”) and
there are no subject and predicate in the usual sense. This single predicate
is expressed by the symbol ⊢. Thus, Frege postulates the abstraction “the
common predicate for all judgements” and thereby disjoins the truth value
from the content of a text. This disagrees with the introspective perspective
described above, in sections 3 and 4.
Thus, in contrast to the introspective view, Frege postulates not only
senses as common entities but also the separations of assertion (and thereby
truth and falsity) from the content of a text. Those two decisions refer to
metalanguage which he uses for describing human language. Appealing to
Grzegorczyk we can repeat that such a metalanguage only relatively corre-
sponds to the nature of human language and although we need not think
about this relativity all the time we should keep it in mind.
But how can we articulate or assess this relativity? For this purpose
we should develop the introspective view and investigate the language of
ordinary human communication viewed from the introspective perspective.
Rather a detailed analysis of the introspective perspective can be found
in the works by Robin G. Collingwood, especially in his Autobiography [2]
and in The Principles of Art (Book II, Ch. XI. Language) [3]. Collingwood’s
introspective view of language is well articulated and detailed; he considers
the whole subjective process of acquiring and using a language.
Before adverting to the ideas of Collingwood we should also mention the
name of Benedetto Croce who was rather influential in the first half of the
20th century. Among his followers were Robin G. Collingwood and Karl
Vossler. All of them are adherents of the introspective stance.
Characterizing the linguistic conception of Collingwood it is first of all
necessary to say that he treats the content of a text as a whole which includes
individual impressions, feelings, etc. and which is not divided into senses of
words, phrases or any other entities; truth (falsity) can also be part of this
content, it is not considered as something added to the content (as it was
considered by Frege).
Collingwood treats words, phrases and other well-known linguistic no-
tions as metaphysical fictions; he does not assume that the science about
 Introspective and Traditional Views of Language 233

such entities investigates the nature of human language. I think that the
large quotation below clearly expressing his position is appropriate here:

The grammatical manipulation of language is so familiar to ourselves,

who have learnt it from the Greeks as an essential part of those trans-
mitted and developing customs which make up our civilization, that we
take it for granted and forget to inquire into its motives. We vaguely
suppose it to be a science; we think that the grammarian, when he
takes a discourse and divides it into parts, is finding out the truth
about it, and that when he lays down rules for the relations between
these parts he is telling us how people’s minds work when they speak.
This is very far from being the truth. A grammarian is not a kind
of scientist studying the actual structure of language; he is a kind of
butcher, converting it from organic tissue into marketable and edible
joints. Language as it lives and grows no more consists of verbs, nouns,
and so forth than animals as they live and grow consist of forehands,
gammons, rump-steaks, and other joints. The grammarian’s real func-
tion (I do not call it purpose, because he does not propose it to himself
as a conscious aim) is not to understand language, but to alter it [. . . ].
[3, p. 257]

“And to teach it”, as we would add.

For Collingwood language is a sort of activity but a grammarian does not
analyze this very activity. He analyses its product, and this product is not
real: it is a metaphysical fiction, a “thing”. When this “thing” is “scientifi-
cally studied” it is cut up into parts, first of all into words; then the scientist
elaborates the scheme of relations between those parts. This division and
those relations are not discovered, they are devised. The traditional investi-
gation presupposes that such a grammatical analysis has already been made,
i.e., that language has already been transformed into the “thing” which is
made up of parts joined by relations and attached to senses.
Now what about truth and falsity? Collingwood treats those notions in
the introspective spirit and sees their origins in certain primitive decisions
of one’s consciousness, in the “the corruption of consciousness”. Such cor-
ruption appears at the level of perception where we have two alternatives.
According to them we may accept a certain feeling or deny it, i.e., we may
implicitly choose between the statements like: This is how I feel or (the
opposite) This is not how I feel. To assert one means to deny the other.

A true consciousness is the confession to ourselves of our feelings; a

false consciousness would be disowning them, i.e., thinking about one
of them ‘That feeling is not mine’. [3, p. 216]
234 Maria K. Timofeeva

[. . . ] whenever some element in experience is disowned by conscious-

ness, that other element upon which attention is fixed, and which con-
sciousness claims as its own, becomes a sham. [. . . ] The picture which
consciousness has painted of its own experience is not only a selected
picture (that is, a true one so far as it goes), it is a bowdlerized picture,
or one whose omissions are falsifications. [3, p. 218]

How does Collingwood explain the nature of the content of a text?

He treats our ordinary language primarily as the means of expressing our
emotions and ideas, not as the way of their transmission, and thinks that
“the expression of any given thought is effected through the expression of
the emotion accompanying it” [3, p. 225].
The emotions which are expressed by a human are always temporal and
individual. Thus each text is tied with its content also only temporally and
relatively to the person. We cannot transmit anything by a text because
we can never be sure of reaching its equal understanding; we can only hope
that this is so:
emotions cannot be shared like food or drink, or handed over like
clothes. [. . . ] Understanding what some one says to you is thus at-
tributing to him the idea which his words arouse in yourself; and this
implies treating them as words of your own. [3, p. 249–250]

And this does not presuppose community of language between the speaker
and hearer. Instead Collingwood presupposes the existence of some pri-
mary “natural” (non-verbal) language natively assigned for expressing the
emotions and qualifies the traditional understanding of sign as inversed.
According to this tradition we speak about symbols that conventionally
correspond to some contents or senses. But there should be some basis for
such a convention, some other language out of which the agreement is arrived
at. Collingwood imputes this role to the primary language.
A similar idea about the necessity of a basic primary language can also
be found in the works by L. Wittgenstein and T. Reid (Reid called the latter
the “natural language”, Wittgenstein used the term “natural expression”);
R. Harré and D.N. Robinson [7] draw a comparison between their thoughts
on this subject.
According to the theory of such a primary language one initially has
some feeling and expresses it in a “natural” (non-verbal, e.g. physiological)
way, only then does s(he) learn the other, verbal, way of expressing the
same feeling. Thus, one becomes aware of words for feelings as alternatives
to natural ways of expressing feelings. It is precisely in this way that one
 Introspective and Traditional Views of Language 235

replaces, say, groaning by speaking. A groan or a smile is but a physical

corollary of the feeling until the time when it is realized and obtains the
capability to replace the feeling. After this the physical corollary begins
to signify something. “The verbal expression of pain replaces crying and
doesn’t describe it” [14, § 244]. The subsequent reasoning of Wittgenstein
differs from that of Collingwood because the latter considers the expressive
function of language to be the basic one while the former primarily discusses
the denotational and regulative aspects of language, the learnable rules of
its usage.
As Collingwood reasons, at first, the primary acts of non-verbal expres-
sion are completely uncontrollable, but at the level of awareness they are
controllable and due to this fact become part of our conscious experience.
In this way the primary language develops from physical level to verbal. By
combining and accumulating such processes lead to the possibility of express-
ing more and more complicated emotions. Nevertheless, in the traditional
theory of language those relations are reversed,
with disastrous results. Language as such is identified with symbolism;
and if its expressive function is not altogether overlooked an attempt
is made to explain it as a secondary function somehow arrived at by
modifying the symbolic function. [. . . ] To-day it is almost an ortho-
doxy that language as such is symbolic in the above sense of the word
[3, p. 226–27]
i.e., symbols are considered as conventionally corresponding to some contents
or senses.
As opposed to Collingwood who has designed the introspective prospects
for studying human language, Ludwig Wittgenstein should not be considered
as an adherent of such views. Beginning with the question “What do the
words of this language signify?” [16, §10], he speaks about the process of
learning the language and arrives at the conclusion that for “a large class
of cases—though not for all—in which we employ the word ’meaning’ it
can be defined thus: the meaning of a word is its use in the language” (§
43). In his reasoning he tests the possibility of explicating the meaning of
a sign (particularly that of a “private language”) with the aid of some rules
or language games. Thus, his vision can also be considered as reversed (in
Collingwood’s terms): he is interested in the lasting meanings (corresponding
to a series of utterances) rather than in expressed contents (which are always
individual and momentary).
Wittgenstein does not consider language as a belonging of a person; he
discusses and tries to find some general rules substantial for communication.
236 Maria K. Timofeeva

He remarks that logic does not treat of language—or of thought—in the sense
in which a natural science treats of a natural phenomenon, and the most that
can be said is that we construct ideal languages (§ 81); he specially stipulates
that we are analyzing not a phenomenon (e.g. thought) but a concept (e.g.
that of thinking) (§ 383). Whereas the introspective view can just be equated
with that of natural sciences; and it considers a language as a phenomenon.
At the same time, some of Wittgenstein’s ideas express the introspective
view. Indeed, the confusions which occupy us arise when language, like
an engine, is idling, not when it is doing work (§ 132) and philosophical
problems arise when language goes on holiday (§ 38). Indeed, we understand
the meaning of a word when we hear or say it; we grasp it in a flash, and
what we grasp in this way is surely something different from the ‘use’ which
is extended in time! (§ 138) Nevertheless, his ideas of this kind do not form
any explicit introspective conception.

6. Concluding remarks

Scientific theories deal only with abstractions but those abstractions can
be of different kinds. Prototypes of a proposition, judgment about its truth
value, presupposition, and associated idea are abstractions, but—in contrast
to the traditional view—they do not posit the unobservable commonly acces-
sible entities, instead, they posit certain introspective (i.e., observable for a
certain individual) principles. Making difference between introspective and
traditional views is important because it helps avoid ascribing objectivity
to the traditional abstractions (e.g. to propositions): we do not see those
entities by introspection; we use them only because of their convenience and
usefulness in a certain type of our reasoning. But there are still other types
of reasoning, in which it is more adequate to posit different ideas—those
which correspond to the introspective view.

References
[1] Ajdukiewicz, K., “Proposition as the Connotation of Sentence”, Studia Logica
V.XX (1967).

[2] Collingwood, R.G., An Autobiography London, 1944.

[3] Collingwood, R.G. The Principles of Art, US: Oxford University Press, 1958.
 Introspective and Traditional Views of Language 237

[4] Grzegorczyk, A., “Psychologistic Semantics and Avoidance of Antinomies”

(Russian. English summary), A.S. Karpenko et al. (ed.), Logical Investigations,
no. 6. Papers from the 1st International Conference “Smirnov’s Readings”,
Moscow, Russia, March 1997. Moscow: ROSSPEHN. 1999.
[5] Frege, G., “On sense and reference”, in: A.W. Moore (ed.) Meaning and Ref-
erence, Oxford: Oxford University Press. 1993.
[6] Frege, G., “Concept script, a formal language of pure thought modelled upon
that of arithmetic”, in: Gottlob Frege. Conceptual Notation and Related Arti-
cles, Oxford University Press 1972.
[7] Harré, R., and D.N. Robinson, “What Makes Language Possible?, Ethological
Foundationalism in Reid and Wittgenstein”, The Review of Metaphysics 50
(1997).
[8] Peregrin, J., “The ‘Natural’ and the ‘Formal’ ”, Journal of Philosophical Logic
29 (2000).
[9] Priest, G., “Truth and Contradiction”, The Philosophical Quarterly 50 (2000).
[10] Sher, G., “In Search of Substantive Theory of Truth”, The Journal of Philos-
ophy, vol. CI, no. 1 (2004).
[11] Tarski, A., “The Semantic Conception of Truth and the Foundations of Se-
mantics”, Philosophy and Phenomenological Research vol. 4, no. 3 (1944).
[12] Timofeeva, M.K., Natural and Formal Languages: Logico-Philosiphical Analy-
sis (in Russian), Novosibirsk, 2003.
[13] Tye, M., “Qualia”, The Stanford Encyclopedia of Philosophy (Summer
2003 Edition), Edward N. Zalta (ed.). http://plato.stanford.edu/archives/
sum2003/entries/qualia/.
[14] Wittgenstein, L., Philosophical Investigations, Blackwell Publishing, 2003.

Maria K. Timofeeva
Laboratory of Mathematical Logic
Institute of Mathematics
Koptug prospect, 4
630090, Novosibirsk, Russia
[email protected]