This excerpt from

The Nature of Truth.

Michael P. Lynch, editor.
© 2001 The MIT Press.

is provided in screen-viewable form for personal use only by members

of MIT CogNet.

Unauthorized use or dissemination of this information is expressly

forbidden.

If you have any questions about this material, please contact

cognetadmin@cognet.mit.edu.
2
Truth
J. L. Austin

`What is truth?' said jesting Pilate, and would not stay for an answer.
Pilate was in advance of his time. For `truth' itself is an abstract noun,
a camel, that is, of a logical construction, which cannot get past the
eye even of a grammarian. We approach it cap and categories in hand:
we ask ourselves whether Truth is a substance (the Truth, the Body of
Knowledge), or a quality (something like the color red, inhering in
truths), or a relation (`correspondence').1 But philosophers should take
something more nearly their own size to strain at. What needs discussing
rather is the use, or certain uses, of the word `true.' In vino, possibly,
`veritas,' but in a sober symposium `verum.'

What is it that we say is true or is false? Or, how does the phrase `is true'
occur in English sentences? The answers appear at ®rst multifarious. We
say (or are said to say) that beliefs are true, that descriptions or accounts
are true, that propositions or assertions or statements are true, and that
words or sentences are true: and this is to mention only a selection of the
more obvious candidates. Again, we say (or are said to say) `It is true that
the cat is on the mat,' or `It is true to say that the cat is on the mat,' or
` ``The cat is on the mat'' is true.' We also remark on occasion, when
someone else has said something, `Very true' or `That's true' or `True
enough.'
26 J. L. Austin

Most (though not all) of these expressions, and others besides, cer-
tainly do occur naturally enough. But it seems reasonable to ask whether
there is not some use of `is true' that is primary, or some generic name for
that which at bottom we are always saying `is true.' Which, if any, of
these expressions is to be taken au pied de la lettre? To answer this will
not take us long, nor, perhaps, far: but in philosophy the foot of the letter
is the foot of the ladder.
I suggest that the following are the primary forms of expression:
It is true (to say) that the cat is on the mat.
That statement (of his, etc.) is true.
The statement that the cat is on the mat is true.
But ®rst for the rival candidates.
a. Some say that `truth is primarily a property of beliefs.' But it may be
doubted whether the expression `a true belief' is at all common outside
philosophy and theology: and it seems clear that a man is said to hold a
true belief when and in the sense that he believes (in) something which is
true, or believes that something which is true is true. Moreover if, as
some also say, a belief is `of the nature of a picture,' then it is of the
nature of what cannot be true, though it may be, for example, faithful.2
b. True descriptions and true accounts are simply varieties of true state-
ments or of collections of true statements, as are true answers and the
like. The same applies to propositions too, in so far as they are genuinely
said to be true (and not, as more commonly, sound, tenable and so on).3
A proposition in law or in geometry is something portentous, usually a
generalization, that we are invited to accept and that has to be recom-
mended by argument: it cannot be a direct report on current observation
Ðif you look and inform me that the cat is on the mat, that is not a
proposition though it is a statement. In philosophy, indeed, `proposition'
is sometimes used in a special way for `the meaning or sense of a sentence
or family of sentences': but whether we think a lot or little of this usage, a
proposition in this sense cannot, at any rate, be what we say is true or
false. For we never say `The meaning (or sense) of this sentence (or of
these words) is true': what we do say is what the judge or jury says,
namely that `The words taken in this sense, or if we assign to them such
and such a meaning, or so interpreted or understood, are true.'
c. Words and sentences are indeed said to be true, the former often, the
latter rarely. But only in certain senses. Words as discussed by philolo-
gists, or by lexicographers, grammarians, linguists, phoneticians, printers,
 Truth 27

critics (stylistic or textual) and so on, are not true or false: they are
wrongly formed, or ambiguous or defective or untranslatable or unpro-
nouncable or misspelled or archaistic or corrupt or what not.4 Sentences
in similar contexts are elliptic or involved or alliterative or ungrammati-
cal. We may, however, genuinely say `His closing words were very true'
or `The third sentence on page 5 of his speech is quite false': but here
`words' and `sentence' refer, as is shown by the demonstratives (posses-
sive pronouns, temporal verbs, de®nite descriptions, etc.), which in this
usage consistently accompany them, to the words or sentence as used by
a certain person on a certain occasion. That is, they refer (as does `Many
a true word spoken in jest') to statements.
A statement is made and its making is an historic event, the utterance
by a certain speaker or writer of certain words (a sentence) to an audi-
ence with reference to an historic situation, event or what not.5
A sentence is made up of words, a statement is made in words. A sen-
tence is not English or not good English, a statement is not in English or
not in good English. Statements are made, words or sentences are used.
We talk of my statement, but of the English sentence (if a sentence is
mine, I coined it, but I do not coin statements). The same sentence is used
in making different statements (I say `It is mine,' you say `It is mine'): it
may also be used on two occasions or by two persons in making the same
statement, but for this the utterance must be made with reference to the
same situation or event.6 We speak of `the statement that S,' but of `the
sentence ``S,'' ' not of `the sentence that S.'7
When I say that a statement is what is true, I have no wish to become
wedded to one word. `Assertion,' for example, will in most contexts do
just as well, though perhaps it is slightly wider. Both words share the
weakness of being rather solemn (much more so than the more general
`what you said' or `your words')Ðthough perhaps we are generally being
a little solemn when we discuss the truth of anything. Both have the merit
of clearly referring to the historic use of a sentence by an utterer, and of
being therefore precisely not equivalent to `sentence.' For it is a fashion-
able mistake to take as primary `(The sentence) ``S'' is true (in the English
language).' Here the addition of the words `in the English language'
serves to emphasize that `sentence' is not being used as equivalent to
`statement,' so that it precisely is not what can be true or false (and more-
over, `true in the English language' is a solecism, mismodeled presum-
ably, and with deplorable effect, on expressions like `true in geometry').
28 J. L. Austin

When is a statement true? The temptation is to answer (at least if we

con®ne ourselves to `straightforward' statements): `When it corresponds
to the facts.' And as a piece of standard English this can hardly be wrong.
Indeed, I must confess I do not really think it is wrong at all: the theory
of truth is a series of truisms. Still, it can at least be misleading.
If there is to be communication of the sort that we achieve by language
at all, there must be a stock of symbols of some kind which a communi-
cator (`the speaker') can produce `at will' and which a communicatee
(`the audience') can observe: these may be called the `words,' though, of
course, they need not be anything very like what we should normally call
wordsÐthey might be signal ¯ags, etc. There must also be something
other than the words, which the words are to be used to communicate
about: this may be called the `world.' There is no reason why the world
should not include the words, in every sense except the sense of the actual
statement itself which on any particular occasion is being made about the
world. Further, the world must exhibit (we must observe) similarities
and dissimilarities (there could not be the one without the other): if
everything were either absolutely indistinguishable from anything else
or completely unlike anything else, there would be nothing to say. And
®nally (for present purposesÐof course there are other conditions to be
satis®ed too) there must be two sets of conventions:
Descriptive conventions correlating the words (ˆ sentences) with the
types of situation, thing, event, etc., to be found in the world.
Demonstrative conventions correlating the words (ˆ statements) with
the historic situations, etc., to be found in the world.8
A statement is said to be true when the historic state of affairs to which
it is correlated by the demonstrative conventions (the one to which it
`refers') is of a type9 with which the sentence used in making it is corre-
lated by the descriptive conventions.10

3a
Troubles arise from the use of the word `facts' for the historic situations,
events, etc., and in general, for the world. For `fact' is regularly used in
 Truth 29

conjunction with `that' in the sentences `The fact is that S' or `It is a fact
that S' and in the expression `the fact that S,' all of which imply that it
would be true to say that S.11
This may lead us to suppose that
i. `fact' is only an alternative expression for `true statement.' We note
that when a detective says `Let's look at the facts' he does not crawl
round the carpet, but proceeds to utter a string of statements: we even
talk of `stating the facts';
ii. for every true statement there exists `one' and its own precisely cor-
responding factÐfor every cap the head it ®ts.
It is (i) which leads to some of the mistakes in `coherence' or formalist
theories; (ii) to some of those in `correspondence' theories. Either we
suppose that there is nothing there but the true statement itself, nothing
to which it corresponds, or else we populate the world with linguistic
DoppelgaÈnger (and grossly overpopulate itÐevery nugget of `positive'
fact overlaid by a massive concentration of `negative' facts, every tiny
detailed fact larded with generous general facts, and so on).
When a statement is true, there is, of course, a state of affairs which
makes it true and which is toto mundo distinct from the true statement
about it: but equally of course, we can only describe that state of affairs
in words (either the same or, with luck, others). I can only describe the
situation in which it is true to say that I am feeling sick by saying that it is
one in which I am feeling sick (or experiencing sensations of nausea):12
yet between stating, however truly, that I am feeling sick and feeling sick
there is a great gulf ®xed.13
`Fact that' is a phrase designed for use in situations where the distinc-
tion between a true statement and the state of affairs about which it is a
truth is neglected; as it often is with advantage in ordinary life, though
seldom in philosophyÐabove all in discussing truth, where it is precisely
our business to prize the words off the world and keep them off it. To ask
`Is the fact that S the true statement that S or that which it is true of?'
may beget absurd answers. To take an analogy: although we may sen-
sibly ask `Do we ride the word ``elephant'' or the animal?' and equally
sensibly `Do we write the word or the animal?' it is nonsense to ask `Do
we de®ne the word or the animal?' For de®ning an elephant (supposing
we ever do this) is a compendious description of an operation involving
30 J. L. Austin

both word and animal (do we focus the image or the battleship?); and so
speaking about `the fact that' is a compendious way of speaking about a
situation involving both words and world.14

3b
`Corresponds' also gives trouble, because it is commonly given too re-
stricted or too colorful a meaning, or one which in this context it cannot
bear. The only essential point is this: that the correlation between the
words (ˆ sentences) and the type of situation, event, etc. which is to be
such that when a statement in those words is made with reference to an
historic situation of that type the statement is then true, is absolutely and
purely conventional. We are absolutely free to appoint any symbol to
describe any type of situation, so far as merely being true goes. In a small
one-spade language tst nuts might be true in exactly the same circum-
stances as the statement in English that the National Liberals are the
people's choice.15 There is no need whatsoever for the words used in
making a true statement to `mirror' in any way, however indirect, any
feature whatsoever of the situation or event; a statement no more needs,
in order to be true, to reproduce the `multiplicity,' say, or the `structure'
or `form' of the reality, than a word needs to be echoic or writing picto-
graphic. To suppose that it does, is to fall once again into the error of
reading back into the world the features of language.
The more rudimentary a language, the more, very often, it will tend to
have a `single' word for a highly `complex' type of situation: this has such
disadavantages as that the language becomes elaborate to learn and is
incapable of dealing with situations which are nonstandard, unforeseen,
for which there may just be no word. When we go abroad equipped only
with a phrase-book, we may spend long hours learning by heartÐ
A 1 -moest-fai nd-e tschaÃr woume n,
Mai hwõÃl-iz-waur pt (beÁnt),
and so on and so on, yet faced with the situation where we have the pen
of our aunt, ®nd ourselves quite unable to say so. The characteristics of
a more developed language (articulation, morphology, syntax, abstrac-
tions, etc.), do not make statements in it any more capable of being true
 Truth 31

or capable of being any more true, they make it more adaptable, more
learnable, more comprehensive, more precise, and so on; and these aims
may no doubt be furthered by making the language (allowance made for
the nature of the medium) `mirror' in conventional ways features descried
in the world.
Yet even when a language does `mirror' such features very closely (and
does it ever?) the truth of statements remains still a matter, as it was with
the most rudimentary languages, of the words used being the ones con-
ventionally appointed for situations of the type to which that referred to
belongs. A picture, a copy, a replica, a photographÐthese are never true
in so far as they are reproductions, produced by natural or mechanical
means: a reproduction can be accurate or lifelike (true to the original), as
a gramophone recording or a transcription may be, but not true (of ) as a
record of proceedings can be. In the same way a (natural) sign of some-
thing can be infallible or unreliable but only an (arti®cial) sign for some-
thing can be right or wrong.16
There are many intermediate cases between a true account and a
faithful picture, as here somewhat forcibly contrasted, and it is from the
study of these (a lengthy matter) that we can get the clearest insight into
the contrast. For example, maps: these may be called pictures, yet they
are highly conventionalized pictures. If a map can be clear or accurate or
misleading, like a statement, why can it not be true or exaggerated? How
do the `symbols' used in mapmaking differ from those used in state-
mentmaking? On the other hand, if an air-mosaic is not a map, why is it
not? And when does a map become a diagram? These are the really illu-
minating questions.

Some have said thatÐ

To say that an assertion is true is not to make any further assertion at
all.
In all sentences of the form `p is true' the phrase `is true' is logically
super¯uous.
To say that a proposition is true is just to assert it, and to say that it is
false is just to assert its contradictory.
32 J. L. Austin

But wrongly. TstS (except in parodoxical cases of forced and dubious

manufacture) refers to the world or any part of it exclusive of tstS, i.e.,
of itself.17 TstST refers to the world or any part of it inclusive of tstS,
though once again exclusive of itself, i.e., of tstST. That is, tstST refers to
something to which tstS cannot refer. TstST does not, certainly, include
any statement referring to the world exclusive of tstS which is not in-
cluded already in tstSÐmore, it seems doubtful whether it does include that
statement about the world exclusive of tstS which is made when we state
that S. (If I state that tstS is true, should we really agree that I have stated
that S? Only `by implication.')18 But all this does not go any way to show
that tstST is not a statement different from tstS. If Mr. Q writes on a
notice-board `Mr. W is a burglar,' then a trial is held to decide whether
Mr. Q's published statement that Mr. W is a burglar is a libel: ®nding
`Mr. Q's statement was true (in substance and in fact).' Thereupon a
second trial is held, to decide whether Mr. W is a burglar, in which
Mr. Q's statement is no longer under consideration: verdict `Mr. W is a
burglar.' It is an arduous business to hold a second trial: why is it done if
the verdict is the same as the previous ®nding?19
What is felt is that the evidence considered in arriving at the one ver-
dict is the same as that considered in arriving at the other. This is not
strictly correct. It is more nearly correct that whenever tstS is true then
tstST is also true and conversely, and that whenever tstS is false tstST is
also false and conversely.20 And it is argued that the words `is true' are
logically super¯uous because it is believed that generally if any two
statements are always true together and always false together then they
must mean the same. Now whether this is in general a sound view may
be doubted: but even if it is, why should it not break down in the case
of so obviously `peculiar' a phrase as `is true'? Mistakes in philosophy
notoriously arise through thinking that what holds of `ordinary' words
like `red' or `growls' must also hold of extraordinary words like `real'
or ``exists.' But that `true' is just such another extraordinary word is
obvious.21
There is something peculiar about the `fact' which is described by
tstST, something which may make us hesitate to call it a `fact' at all;
namely, that the relation between tstS and the world which tstST asserts
to obtain is a purely conventional relation (one which `thinking makes
 Truth 33

so'). For we are aware that this relation is one which we could alter at
will, whereas we like to restrict the word `fact' to hard facts, facts which
are natural and unalterable, or anyhow not alterable at will. Thus, to
take an analogous case, we may not like calling it a fact that the word
elephant means what it does, though we can be induced to call it a (soft)
factÐand though, of course, we have no hesitation in calling it a fact
that contemporary English speakers use the word as they do.
An important point about this view is that it confuses falsity with
negation: for according to it, it is the same thing to say `He is not at
home' as to say `It is false that he is at home.' (But what if no one has
said that he is at home? What if he is lying upstairs dead?) Too many
philosophers maintain, when anxious to explain away negation, that a
negation is just a second order af®rmation (to the effect that a certain ®rst
order af®rmation is false), yet, when anxious to explain away falsity,
maintain that to assert that a statement is false is just to assert its nega-
tion (contradictory). It is impossible to deal with so fundamental a matter
here.22 Let me assert the following merely. Af®rmation and negation are
exactly on a level, in this sense, that no language can exist which does not
contain conventions for both and that both refer to the world equally
directly, not to statements about the world: whereas a language can quite
well exist without any device to do the work of `true' and `false.' Any
satisfactory theory of truth must be able to cope equally with falsity:23
but `is false' can only be maintained to be logically super¯uous by making
this fundamental confusion.

There is another way of coming to see that the phrase `is true' is not
logically super¯uous, and to appreciate what sort of a statement it is to
say that a certain statement is true. There are numerous other adjectives
which are in the same class as `true' and `false,' which are concerned, that
is, with the relations between the words (as uttered with reference to an
historic situation) and the world, and which nevertheless no one would
dismiss as logically super¯uous. We say, for example, that a certain state-
ment is exaggerated or vague or bald, a description somewhat rough or
misleading or not very good, an account rather general or too concise.
34 J. L. Austin

In cases like these it is pointless to insist on deciding in simple terms

whether the statement is `true or false.' Is it true or false that Belfast is
north of London? That the galaxy is the shape of a fried egg? That
Beethoven was a drunkard? That Wellington won the battle of Water-
loo? There are various degrees and dimensions of success in making
statements: the statements ®t the facts always more or less loosely, in
different ways on different occasions for different intents and purposes.
What may score full marks in a general knowledge test may in other
circumstances get a gamma. And even the most adroit of languages may
fail to `work' in an abnormal situation or to cope, or cope reasonably
simply, with novel discoveries: is it true or false that the dog goes round
the cow?24 What, moreover, of the large class of cases where a statement
is not so much false (or true) as out of place, inept (`All the signs of
bread' said when the bread is before us)?
We become obsessed with `truth' when discussing statements, just as
we become obsessed with `freedom' when discussing conduct. So long
as we think that what has always and alone to be decided is whether a
certain action was done freely or was not, we get nowhere: but so soon
as we turn instead to the numerous other adverbs used in the same con-
nection (`accidentally,' `unwillingly,' `inadvertently,' etc.), things become
easier, and we come to see that no concluding inference of the form
`Ergo, it was done freely (or not freely)' is required. Like freedom, truth is
a bare minimum or an illusory ideal (the truth, the whole truth and noth-
ing but the truth about, say, the battle of Waterloo or the Primavera).25

Not merely is it jejune to suppose that all a statement aims to be is `true,'

but it may further be questioned whether every `statement' does aim to
be true at all. The principle of Logic, that `Every proposition must be true
or false,' has too long operated as the simplest, most persuasive and most
pervasive form of the descriptive fallacy. Philosophers under its in¯uence
have forcibly interpreted all `propositions' on the model of the statement
that a certain thing is red, as made when the thing concerned is currently
under observation.
 Truth 35

Recently, it has come to be realized that many utterances which have

been taken to be statements (merely because they are not, on grounds of
grammatical form, to be classed as commands, questions, etc.) are not in
fact descriptive, nor susceptible of being true or false. When is a state-
ment not a statement? When it is a formula in a calculus: when it is a
performatory utterance: when it is a value-judgment: when it is a de®ni-
tion: when it is part of a work of ®ctionÐthere are many such suggested
answers. It is simply not the business of such utterances to `correspond to
the facts' (and even genuine statements have other businesses besides that
of so corresponding).
It is a matter for decision how far we should continue to call such
masqueraders `statements' at all, and how widely we should be prepared
to extend the uses of `true' and `false' in `different senses.' My own feel-
ing is that it is better, when once a masquerader has been unmasked, not
to call it a statement and not to say it is true or false. In ordinary life we
should not call most of them statements at all, though philosophers and
grammarians may have come to do so (or rather, have lumped them
all together under the term of art `proposition'). We make a difference
between `You said you promised' and `You stated that you promised':
the former can mean that you said `I promise,' whereas the latter must
mean that you said `I promised': the latter, which we say you `stated,' is
something which is true or false, whereas for the former, which is not
true or false, we use the wider verb to `say.' Similarly, there is a difference
between `You say this is (call this) a good picture' and `You state that this
is a good picture.' Moreover, it was only so long as the real nature of
arithmetical formulas, say, or of geometrical axioms remained unrecog-
nized, and they were thought to record information about the world, that
it was reasonable to call them `true' (and perhaps even `statements'Ð
though were they ever so called?): but, once their nature has been recog-
nized, we no longer feel tempted to call them `true' or to dispute about
their truth or falsity.
In the cases so far considered the model `This is red' breaks down
because the `statements' assimilated to it are not of a nature to corre-
spond to facts at allÐthe words are not descriptive words, and so on.
But there is also another type of case where the words are descriptive
36 J. L. Austin

words and the `proposition' does in a way have to correspond to facts,

but precisely not in the way that `This is red' and similar statements set-
ting up to be true have to do.
In the human predicament, for use in which our language is designed,
we may wish to speak about states of affairs which have not been observed
or are not currently under observation (the future, for example). And
although we can state anything `as a fact' (which statement will then be
true or false26) we need not do so: we need only say `The cat may be on
the mat.' This utterance is quite different from tstSÐit is not a statement
at all (it is not true or false; it is compatible with `The cat may not be on
the mat'). In the same way, the situation in which we discuss whether and
state that tstS is true is different from the situation in which we discuss
whether it is probable that S. Tst it is probable that S is out of place,
inept, in the situation where we can make tstST, and, I think, conversely.
It is not our business here to discuss probability: but is worth observing
that the phrases `It is true that' and `It is probable that' are in the same
line of business,27 and in so far incompatibles.

In a recent article in Analysis, Mr. Strawson has propounded a view

of truth which it will be clear I do not accept. He rejects the `semantic'
account of truth on the perfectly correct ground that the phrase `is true' is
not used in talking about sentences, supporting this with an ingenious
hypothesis as to how meaning may have come to be confused with truth:
but this will not suf®ce to show what he wantsÐthat `is true' is not used
in talking about (or that `truth is not a property of') anything. For it is
used in talking about statements (which in his article he does not distin-
guish clearly from sentences). Further, he supports the `logical super¯uity'
view to this extent, that he agrees that to say that ST is not to make any
further assertion at all, beyond the assertion that S: but he disagrees with
it in so far as he thinks that to say that ST is to do something more than
just to assert that SÐit is namely to con®rm or to grant (or something of
that kind) the assertion, made or taken as made already, that S. It will be
clear that and why I do not accept the ®rst part of this: but what of the
 Truth 37

second part? I agree that to say that ST `is' very often, and according to
the all-important linguistic occasion, to con®rm tstS or to grant it or
what not; but this cannot show that to say that ST is not also and at the
same time to make an assertion about tstS. To say that I believe you `is'
on occasion to accept your statement; but it is also to make an assertion,
which is not made by the strictly performatory utterance `I accept your
statement.' It is common for quite ordinary statements to have a per-
formatory `aspect': to say that you are a cuckold may be to insult you,
but it is also and at the same time to make a statement which is true
or false. Mr. Strawson, moreover, seems to con®ne himself to the case
where I say `Your statement is true' or something similarÐbut what of
the case where you state that S and I say nothing but `look and see' that
your statement is true? I do not see how this critical case, to which
nothing analogous occurs with strictly performatory utterances, could be
made to respond to Mr. Strawson's treatment.
One ®nal point: if it is admitted (if ) that the rather boring yet satis-
factory relation between words and world which has here been discussed
does genuinely occur, why should the phrase `is true'' not be our way of
describing it? And if it is not, what else is?

Notes

1. It is suf®ciently obvious that `truth' is a substantive, `true' an adjective and `of'

in `true of' a preposition.
2. A likeness is true to life, but not true of it. A word picture can be true, just
because it is not a picture.
3. Predicates applicable also to `arguments,' which we likewise do not say are
true, but, for example, valid.
4. Peirce made a beginning by pointing out that there are two (or three) different
senses of the word `word,' and adumbrated a technique (`counting' words) for
deciding what is a `different sense.' But his two senses are not well de®ned,
and there are many moreÐthe `vocable' sense, the philologist's sense in which
`grammar' is the same word as `glamour,' the textual critic's sense in which the
`the' in l. 254 has been written twice, and so on. With all his 66 divisions of signs,
Peirce does not, I believe, distinguish between a sentence and a statement.
5. `Historic' does not, of course, mean that we cannot speak of future or possible
statements. A `certain' speaker need not be any de®nite speaker. `Utterance' need
not be public utteranceÐthe audience may be the speaker himself.
38 J. L. Austin

6. `The same' does not always mean the same. In fact it has no meaning in the
way that an `ordinary' word like `red' or `horse' has a meaning: it is a (the typical)
device for establishing and distinguishing the meanings of ordinary words. Like
`real,' it is part of our apparatus in words for ®xing and adjusting the semantics
of words.
7. Inverted commas show that the words, though uttered (in writing), are not to
be taken as a statement by the utterer. This covers two possible cases, (i) where
what is to be discussed is the sentence, (ii) where what is to be discussed is a
statement made elsewhen in the words `quoted.' Only in case (i) is it correct to say
simply that the token is doing duty for the type (and even here it is quite incorrect
to say that `The cat is on the mat' is the name of an English sentenceÐthough
possibly The Cat is on the Mat might be the title of a novel, or a bull might be
known as Catta est in matta). Only in case (ii) is there something true or false, viz.
(not the quotation but) the statement made in the words quoted.
8. Both sets of conventions may be included together under `semantics.' But they
differ greatly.
9. `Is of a type with which' means `is suf®ciently like those standard states of
affairs with which.' Thus, for a statement to be true one state of affairs must be
like certain others, which is a natural relation, but also suf®ciently like to merit
the same `description,' which is no longer a purely natural relation. To say `This
is red' is not the same as to say `This is like those,' nor even as to say `This is like
those which were called red.' That things are similar, or even `exactly' similar, I
may literally see, but that they are the same I cannot literally seeÐin calling them
the same color a convention is involved additional to the conventional choice of
the name to be given to the color which they are said to be.
10. The trouble is that sentences contain words or verbal devices to serve both
descriptive and demonstrative purposes (not to mention other purposes), often
both at once. In philosophy we mistake the descriptive for the demonstrative
(theory of universals) or the demonstrative for the descriptive (theory of monads).
A sentence as normally distinguished from a mere word or phrase is characterized
by its containing a minimum of verbal demonstrative devices (Aristotle's `refer-
ence to time'); but many demonstrative conventions are nonverbal (pointing,
etc.), and using these we can make a statement in a single word which is not a
`sentence.' Thus, `languages' like that of (traf®c, etc.) signs use quite distinct
media for their descriptive and demonstrative elements (the sign on the post,
the site of the post). And however many verbal demonstrative devices we use as
auxiliaries, there must always be a nonverbal origin for these coordinates, which
is the point of utterance of the statement.
11. I use the following abbreviations:
S for the cat is on the mat.
ST for it is true that the cat is on the mat.
tst for the statement that.
I take tstS as my example throughout and not, say, tst Julius Caesar was bald
or tst all mules are sterile, because these latter are apt in their different ways
 Truth 39

to make us overlook the distinction between sentence and statement: we have,

apparently, in the one case a sentence capable of being used to refer to only one
historic situation, in the other a statement without reference to at least (or to any
particular) one.
If space permitted other types of statement (existential, general, hypothetical,
etc.) should be dealt with: these raise problems rather of meaning than of truth,
though I feel uneasiness about hypotheticals.
12. If this is what was meant by ` ``It is raining'' is true if and only if it is raining,'
so far so good.
13. It takes two to make a truth. Hence (obviously) there can be no criterion of
truth in the sense of some feature detectable in the statement itself which will reveal
whether it is true or false. Hence, too, a statement cannot without absurdity refer
to itself.
14. `It is true that S' and `It is a fact that S' are applicable in the same circum-
stances; the cap ®ts when there is a head it ®ts. Other words can ®ll the same role
as `fact': we say, e.g., `The situation is that S.'
15. We could use `nuts' even now as a codeword: but a code, as a transformation
of a language, is distinguished from a language, and a codeword dispatched is not
(called) `true.'
16. Berkeley confuses these two. There will not be books in the running brooks
until the dawn of hydrosemantics.
17. A statement may refer to `itself' in the sense, for example, of the sentence
used or the utterance uttered in making it (`statement' is not exempt from all
ambiguity). But paradox does result if a statement purports to refer to itself in a
more full-blooded sense, purports, that is, to state that it itself is true, or to state
what it itself refers to (`This statement is about Cato').
18. And `by implication' tstST asserts something about the making of a state-
ment which tstS certainly does not assert.
19. This is not quite fair: there are many legal and personal reasons for holding
two trialsÐwhich, however, do not affect the point that the issue being tried is
not the same.
20. Not quite correct, because tstST is only in place at all when tstS is envisaged
as made and has been veri®ed.
21. Unum, verum, bonumÐthe old favorites deserve their celebrity. There is some-
thing odd about each of them. Theoretical theology is a form of onomatolatry.
22. The following two sets of logical axioms are, as Aristotle (though not his
successors) makes them, quite distinct:
(a) No statement can be both true and false.
No statement can be neither true nor false.
(b) Of two contradictory statementsÐ
Both cannot be true.
Both cannot be false.
40 J. L. Austin

The second set demands a de®nition of contradictories, and is usually joined

with an unconscious postulate that for every statement there is one and only one
other statement such that the pair are contradictories. It is doubtful how far any
language does or must contain contradictories, however de®ned, such as to
satisfy both this postulate and the set of axioms (b).
Those of the so-called `logical paradoxes' (hardly a genuine class) which con-
cern `true' and `false' are not to be reduced to cases of self-contradiction, any
more than `S but I do not believe it' is. A statement to the effect that it is itself true
is every bit as absurd as one to the effect that it is itself false. There are other types
of sentences which offend against the fundamental conditions of all communica-
tion in ways distinct from the way in which `This is red and is not red' offendsÐ
e.g., `This does (I do) not exist,' or equally absurd `This exists (I exist).' There are
more deadly sins than one; nor does the way to salvation lie through any hierarchy.
23. To be false is (not, of course, to correspond to a nonfact, but) to miscor-
respond with a fact. Some have not seen how, then, since the statement which
is false does not describe the fact with which it miscorresponds (but misdescribes
it), we know which fact to compare it with: this was because they thought of all
linguistic conventions as descriptiveÐbut it is the demonstrative conventions
which ®x which situation it is to which the statement refers. No statement can
state what it itself refers to.
24. Here there is much sense in `coherence' (and pragmatist) theories of truth,
despite their failure to appreciate the trite but central point that truth is a matter
of the relation between words and world, and despite their wrongheaded Gleich-
schaltung of all varieties of statemental failure under the lone head of `partly
true' (thereafter wrongly equated with `part of the truth'). `Correspondence'
theorists too often talk as one would who held that every map is either accu-
rate or inaccurate; that accuracy is a singly and the sole virtue of a map; that
every country can have but one accurate map; that a map on a larger scale or
showing different features must be a map of a different country; and so on.
25. Austin pursues this line of thought further in How to Do Things with Words
(Oxford: The Clarendon Press, 1962), pp. 139ff.ÐEd.
26. Though it is not yet in place to call it either. For the same reason, one cannot
lie or tell the truth about the future.
27. Compare the odd behaviors of `was' and `will be' when attached to `true' and
to `probable.'
This excerpt from

The Nature of Truth.

Michael P. Lynch, editor.
© 2001 The MIT Press.

is provided in screen-viewable form for personal use only by members

of MIT CogNet.

Unauthorized use or dissemination of this information is expressly

forbidden.

If you have any questions about this material, please contact

cognetadmin@cognet.mit.edu.