Professional Documents
Culture Documents
This article aims to disrupt received views about the significanc e of J. L. Austin’s
contributio n to philosophy of language. Its focus is Austin’s 1955 lectures How To
Do Things With Words. Commentators on the lectures in both philosophica l and
literary-theoretica l circles, despite conspicuou s differences , tend to agree in
attributing to Austin an assumption about the relation between literal meaning and
truth, which is in fact his central critical target. The goal of the article is to correct
this misunderstandin g and to show that Austin is deeply critical of a picture of
correspondenc e between language and the world which nearly half a century after he
delivered his lectures continue s to structure philosophica l discussions of language.
And forget, for once and for a while, that ... curious question
‘Is it true?’ May we?1
I. Introduction
Over the last couple of decades, J. L. Austin’s work has slid into a state of
respectable semi-obscurity. While some bits of philosophical terminology he
introduced remain in circulation (e.g. his talk of ‘constative’ and
‘performative’ utterances and his system of classifying utterances according
to their ‘illocutionary forces’), even these bits tend to be regarded, not as
expressing groundbreaking perspectives, but rather as accommodating
themselves comfortably within dominant research programs. My aim in this
article is to raise a question about this image of Austin as a philosophically
relatively innocuous – if also somewhat eccentric and engaging – gure. I will
be concerned above all with his 1955 lectures How To Do Things With Words.
What I hope to show is (i) that one of Austin’s larger goals here is to criticize
an ideal of the ‘statement’ as always in some way reporting on the world or
imparting information about the facts either truly or falsely3 and also (ii) that,
in criticizing this ideal, Austin is developing a powerful attack on a picture of
correspondence between language and the world that continues, nearly half a
century after he delivered his lectures, to structure philosophical discussions
of language.
The picture of correspondence between language and the facts that Austin
or impart straightforward information about the facts’ and also for the fact
that some philosophers have become concerned with asking ‘whether many
apparent pseudo-statements really set out to be statements at all’ (p. 2). In this
connection, Austin touches on the efforts of philosophers who take an interest
in utterances which are not evidently nonsensical, but which do not state facts
or describe states of affairs. He mentions the work of those emotivists who
maintain that ‘ethical propositions’ are perhaps intended’, not to report on
features of the world, but rather ‘to evince emotion or to prescribe conduct or
to in uence it in special ways’ (pp. 2–3).
Austin moves from surveying ways in which philosophica l research into
language has been structured by an ideal of the ‘descriptive statement’ to
outlining a new system for classifying utterances – a system which, like those
proposed within earlier research, leaves the ideal in place. He draws attention
to utterances which look like ‘statements’ but which do not meet the
speci cations of the traditional ideal. He proposes the term ‘performative’ for
utterances which are not properly characterized as describing a state of affairs
and in which the uttering of a sentence (e.g. ‘I promise’) is the doing of an
action (e.g. making a promise) which would not normally be described as
saying something (p. 5). And he proposes the term ‘constative’ for utterances
which, in so far as they describe a state of affairs or state a fact (e.g. ‘That dog
is dangerous’), appear to attain to the traditional ideal of the ‘statement’.
Although the classical ‘statement’ – or, as he now prefers to put it, the
constative utterance – can be judged true or false depending on whether it
accurately or inaccurately describes a state of affairs or states a fact, the
performative utterance is not appropriately assessed in these terms.
It merits emphasis that this initial classi cation of utterances into
constatives and performatives takes for granted the ideal of the ‘statement’
that Austin highlights in the opening moments of his lectures. This is
important because, later in his lectures, Austin will declare that he nds this
ideal deeply problematic. Indeed, even at this early stage, he says that he
wants to question ‘an age-old assumption in philosophy – the assumption that
to say something, at least in all cases worth considering, i.e. all cases
considered, is always and simply to state something’ (p. 12).10 Now, it may
appear that Austin, in isolating a class of utterances that do not meet the
speci cations of the traditional ‘statement’, has already washed his hands of
the assumption that ‘to say something is always to state something’.
Similarly, it may appear that the emotivist, in describing a set of utterances
which are best understood as involving the expression of (non-cognitive)
feeling, has broken with this assumption. Austin, however, believes that
genuine liberation from the assumption requires more than the creation of
new categories of utterances to coexist alongside the traditional ‘statement’.
At various points in his rst several lectures, he suggests that philosophers’
preoccupation with the category of the ‘statement’ has its roots in a confused
The Happy Truth 63
Austin notes that the performatives he has thus far considered (e.g. ‘I name
this ship ...’, ‘I give and bequeath this watch ...’, ‘I bet you sixpence that ...’,
etc.) all have verbs in the rst person singular present indicative active. He
accordingly asks whether this mood and tense are distinctive of the
performative. What he discovers is that there are performatives that do not
observe this rule. There are verbal formulae in the rst person plural (e.g. ‘We
promise ...’), in the second person (e.g. ‘You are hereby authorized to pay ...’),
and in the third person (e.g. ‘Passengers are warned ...’) which are sometimes
issued with performative force (p. 57). He concludes that ‘mood and tense ...
break down as absolute criteria’ for the isolation of the performative (p. 58).
Austin now considers another suggestion for isolating the performative.
The performative might be distinguished by a special vocabulary. This
suggestion, however, is undermined by the observation that for any candidate
pair of performative and ‘performative words’ it is possible to produce the
performative without the words, or to use the words without producing the
performative (p. 59).
These failures lead Austin to ask whether a more complex criterion – one
‘involving both grammar and vocabulary’ (p. 59) – is required, and to propose
such a criterion for ‘explicit performatives’. The mark of an explicit
performative is that it makes explicit the agent who is performing the action in
question and also that its verb or verb phrase makes explicit what action is
being performed on the occasion of utterance (e.g. ‘I promise to pay you
tomorrow’). Austin considers the suggestion that we can demarcate
performative language by counting as performatives utterances which are
‘reducible, expandable, or analysable’ into this explicit form (pp. 60–61). But
the suggestion encounters intractable dif culties. Some verbal formulae (e.g.
‘I approve’) seem sometimes to be explicit performatives (e.g. ‘I approve’
when it has the performative force of giving approval) and sometimes to be
descriptives (e.g. ‘I approve’ when it means ‘I favor this’). Austin concludes
that ‘it is often not easy to be sure that, even when it is apparently in explicit
form, an utterance is performative or that it is not’ (p. 91), and he abandons his
attempt to nd a precise method of distinguishing constatives and
performatives.
It seems clear that Austin’s narrative about his failure to nd a grammatical
criterion for distinguishing constative and performative utterances is intended
to establish that there is something confused in principle about the idea of
such a criterion. Recall his discussion of how constative utterances can be
infelicitous and can even ‘mis re’. Here Austin is attacking the idea that we
can isolate constative combinations of words in advance of a consideration of
ways in which those combinations of words are used on particular occasions.
He challenges our tendency to regard the contexts in which strings of words
are used as essentially irrelevant – or ‘merely pragmatic’. This relatively
dismissive treatment of context appears justi ed where the ‘statement’ is
66 Alice Crary
concerned with language in this sense, with speech acts. He writes that ‘[t]he
total speech act in the total speech situation is the only actual phenomenon
which, in the last resort, we are engaged in elucidating’ (p. 147). And he
signals that he thinks we can classify bits of language with regard to the kind
of speech act that they are – with regard, as he will put it, to their
‘illocutionary forces’.16
the sentence ‘The car is red.’ If any sentence has a clear ‘literal meaning’, this
sentence seems like a good candidate. It appears to have a constant and
unvarying descriptive meaning – one which its indexical elements (along with
contextual clues) will pin down to the particular contexts in which it is used.
Imagine the following situation:
Two people, A and B are sitting at a table that has on it a checker set and
a small green toy car. A picks up the green car and says: ‘The car is red.’
B is confused. Nonetheless, because she takes herself to understand
(what she thinks of as) the ‘literal meaning’ of A’s sentence, she thinks
she can determine (by appealing to this ‘literal meaning’ along with a
few contextual clues) what factual statement A is making. Suppose that B
now responds: ‘Of course it’s not red. It’s green.’ A laughs and rejoins:
‘That’s not what I meant. This checker set is missing one of its red
pieces. I was suggesting that we use this little car here as one of the red
pieces!’
The correct lesson to draw from this example is not that the sentence ‘The car
is red’ as A uses it does not have the particular ‘literal meaning’ that B takes it
to have. The example leaves open the question of whether the sentence as A
uses it is in the end best understood as having what B takes to be its ‘literal
meaning’. Indeed it would be in keeping with the point of the example to
append to it a scenario in which B, after she has understood what A is using
the sentence ‘The car is red’ to say, tries to gure out whether or not the
sentence as A uses it has (what B thinks of as) its ‘literal meaning’. Which
decision B ultimately reaches is of no importance. What the example teaches
is that it is only once B has understood how A’s sentence is to be taken that she
is in a position to make such a decision at all. Since it is only once B has
understood how A’s sentence is to be taken that she is in a position to say what
the sentence as A uses it means, it is only at this point that she is in a position
to consider what similarities and differences obtain between the meaning it
has in this context and (what she takes to be) its ‘literal meaning’. So it is not
in virtue of her grasp of (what she thinks of as) the ‘literal meaning’ of A’s
sentence that B understands what A is saying. This means that B’s appeal to
the ‘literal meaning’ of A’s sentence does not here play the role that it is often
called on to play in philosophy.
Austin’s account of locutionary and illocutionary acts, taken as a whole,
brings into question the idea that we might develop a theory that could be
used to identify the locutionary acts performed whenever particular sentences
are used. But his strategy for combating this idea involves allowing himself at
rst to entertain it. After initially elaborating his (suspect) notion of the
locutionary act, he criticizes it by rejecting as awed an idea that it
presupposes, viz., that it is possible to isolate the locutionary act that is
performed when a particular sentence is employed in the absence of a grasp of
70 Alice Crary
the illocutionary force with which it is being used. Austin’s misgivings about
the viability of his own notion of the locutionary act are best understood as
directed towards his initial description of it and should not be interpreted as a
sign that he lacks con dence in his overall treatment.
After contrasting the illocutionary act with the locutionary act, Austin goes
on to contrast it with the perlocutionary act. He distinguishes the sorts of
conventional effects that belong to the illocutionary act (e.g. my saying
something which has the illocutionary force of a promise has the effect of my
being committed by my promise [p. 102]) from the sorts of consequential
effects that uttering a string of words may have (e.g. your saying ‘Look what
you’re doing!’ may have the effect of my stopping my work and looking to
see whether I have perhaps spilled my coffee or put my elbow in the jam). He
calls the production of the latter kind of effects a perlocutionary act (p. 101)
and says he wants to keep us from confusing illocutionary acts with
perlocutionary ones (p. 103). In urging us not to understand his remarks about
illocutionary force as remarks about perlocutions, Austin is, once again,
guarding against the picture of meaning at play in a traditional philosophical
ideal of the ‘statement’. What interests him is the fact that it is possible to
identify the perlocutionary (or consequential) effects of an utterance without
understanding what someone is saying in uttering the relevant string of words.
It follows that a system of classifying utterances according to their
perlocutionary forces would be one that did not presuppose an understanding
of the utterances in question. Our employment of such a system would be
consistent with the view that it is possible to grasp the meaning of a sentence
without considering the force with which it is issued in a particular context;
we could study perlocutions without distancing ourselves from the notion of
literal sentence-meaning. The result is that if we read Austin as concerned not
with illocutionary effects but with perlocutionary ones, we will be inclined
not to recognize that one of his central concerns is criticizing this notion.
One further comment is in order here. After contrasting illocutionary acts
with locutionary and perlocutionary acts, Austin goes on to classify utterances
according to their illocutionary forces. At this juncture, he starts to refer to
those utterances that have the illocutionary forces of promises, bets, warnings,
etc., as ‘promises’, ‘bets’, ‘warnings’, and he starts to use the term ‘statement’
(without single quotes) – not, as he did previously, to refer to all historical
utterances of sentences, but rather – to refer exclusively to those utterances
that have the illocutionary force of statements.20
calls the ‘true/false fetish’. He realizes that, to the extent that it seems possible
to identify sentences as such as constative, it will seem reasonable to think
that we can characterize the relation of correspondence between language and
the facts without investigating how what we say bears on the facts in
particular circumstances. Now it will seem as though we are justi ed in
conceiving ‘true’ and ‘false’ as terms of assessment which (because, by our
lights, they pertain to a relationship between sentences and world which is not
caught up with our conventional procedures for performing speech acts) are
essentially different from the terms of assessment (e.g. ‘bizarre’, ‘appro-
priate’, ‘rough’, ‘misleading’, etc.) employed in assessing speech acts. Austin
accordingly believes that if we abandon the idea that we can grasp the
meaning of sentences independently of a grasp of the (illocutionary) force
with which they are used in particular contexts, we will want to say that the
terms we use for appraising statements ‘overlap with those we use in the
appraisal of performatives’ (pp. 141–2).21 We will take ‘true’ and ‘false’ to
stand, not for a ‘simple’ relation (i.e. one ‘uncontaminated’ by conventions of
illocutionary force), but rather for ‘a general dimension of being a right or
proper thing to say as opposed to a wrong thing, in these circumstances, to this
audience, for these purposes and with these intentions’ (p. 144). Austin
writes:
Is the constative, then, always true or false? When a constative is confronted with
the facts, we in fact appraise it in ways involving the employment of a vast array of
terms which overlap with those that we use in the appraisal of performatives. In real
life, as opposed to the simple situations envisaged in logical theory, one cannot
always answer in a simple manner whether it is true or false (pp. 141–2).22
By arguing along these lines, Austin wishes to get us to jettison the view
that ‘true’ and ‘false’ are essentially different from evaluative terms we use in
the assessment of speech acts and to leave room for the possibility that some
bits of language that report on the facts are most accurately characterized, not
either as true or as false, but rather as, say, rough or deceptive. But it would be
misleading to portray him as simply claiming that sometimes statements are
neither true nor false – that sometimes they are ‘approximately true’ or ‘not
strictly false’. This claim is consistent with the view of literal sentence-
meaning Austin rejects. Someone who endorsed the rejected view might
maintain that there are cases in which we know the literal meaning of a
sentence that is being used to make a ‘statement’ – and in which, moreover,
we also know how the indexical elements of the sentence pin it to the
particular context of use – but in which we nevertheless cannot determine
whether the relevant ‘statement’ is true or false. The idea might be that our
inability to make judgments of truth or falsity in these cases is a function of
(what might be thought of as) the indeterminacy of literal meaning. However,
although a conception of the indeterminacy of literal meaning might thus
72 Alice Crary
follow from Austin’s discovery that statements are speech acts that some
speech acts are capable of being true or false.26
Searle is not unaware that Austin’s talk of statements as speech acts is
intended to draw attention to the role the circumstances in which a statement
is made play in making up its content. Searle repeatedly observes that one of
Austin’s larger philosophical ambitions is to show that features of the
circumstances in which a sentence is uttered play a role in constituting its
meaning. And he tries to accommodate this moment in Austin’s thought by
claiming that features of the way in which a sentence is used (such as its
‘stress and intonation contour’) affect (what Searle calls) its ‘intended
meaning’.27 Searle’s thought seems to be that, once he has made this initial
claim about the signi cance of the circumstances in which linguistic
expressions are used, he can then go on – without betraying the spirit of
Austin’s thought – to advance a further claim about how sentences have (in
addition to their ‘intended meanings’) ‘literal meanings’ they carry with them
into every context of use. Searle thus believes he is licensed to present himself
as faithful to Austin’s thought in arguing that the way in which a sentence
corresponds to the world is a function of elements of its ‘literal meaning’.28
And he believes he is licensed to claim that Austin would have supported an
ideal of the ‘statement’ that presupposes some idea of literal sentence-
meaning if not for (what Searle regards as) an error in Austin’s analysis of
statements.
Searle does not argue that it is possible to defend the attribution to Austin of
an idea of literal sentence-meaning by mustering internal, textual evidence.
What he suggests instead is that it is possible to defend this attribution by
appealing to evidence of an external kind. In relatively recent writings, Searle
defends an assumption about how it is necessary to retain the idea of literal
sentence-meaning in order to preserve the notion of objective truth.29 And he
signals that he thinks this assumption speaks in favor of searching for an
interpretation of Austin’s doctrine of speech acts which is consistent with the
idea of such meaning.
It is worth noting in this connection that, although Derrida is conspicuously
critical of Searle’s interpretation of Austin’s lectures,30 his interpretative
strategy nevertheless resembles Searle’s in an important respect. Derrida
(together with members of the group of commentators on Austin he has
in uenced) agrees with Searle in approaching Austin’s lectures equipped with
the assumption that the notion of objective truth stands or falls with the idea
of literal sentence-meaning.
The point here is not that the assumption operates uniformly across the
different interpretations. Commentators who follow in Derrida’s footsteps
differ from Searle in taking the assumption to support an interpretation of
Austin on which he appears to be – not receptive to the idea of literal
sentence-meaning, but rather – uneasy about the very notion of objective
74 Alice Crary
truth. What accounts for this difference from Searle, given the shared
interpretative framework, is the fact that, unlike Searle, the members of this
group of commentators start with the recognition that a central aim of
Austin’s lectures is to discredit the idea of literal sentence-meaning.31
One characteristic expression of this form of recognition, within the
writings of Derrida and other literary-theoretical commentators, concerns
their understanding of Austin’s view of truth. It is typical of Derrida – and, for
instance, Stanley Fish – to claim that Austin rejects a familiar conception of
truth and falsity that takes for granted that we can understand what someone is
saying independently of a grasp of how her words satisfy the demands of a
conventional procedure for saying something on a particular occasion. Thus
Derrida writes that Austin hopes to free us ‘from the true/false opposition ... at
least in its classical form’.32 And Fish tells us that Austin wishes to wean us
from the view that answers to ‘the question “Is it true or false?”... [are]
absolute judgments made independently of any particular set of circum-
stances’.33
In so far as the account of Austin’s view of truth that Derrida and Fish
champion portrays Austin as repudiating a traditional philosophica l regime of
truth (and, in particular, one that presupposes the idea of literal sentence-
meaning), it resembles the account presented in this article. But this
resemblance coexists with a philosophically fundamental difference. Within
the Derridean account – in sharp contrast to the account laid out above – an
interpretatively compelling story about how Austin abandons a familiar
regime of truth serves as a preface to an interpretatively quite dubious story
about how he is skeptical about the very notion of objective truth. Austin is
read as maintaining that there is no such thing as a statement that possesses
the kind of universal authority that is distinctive of objective truth. He is taken
to be denying the existence of statements that are objectively true in the sense
that anyone who is thinking about the pertinent matter correctly ought to
recognize their correctness. The larger suggestion of the Derridean account of
Austin’s view of truth is that Austin should be read as proposing to exchange
an understanding of truth and falsity as terms of assessment with universal
authority for an understanding of them as terms of assessment whose
authority is merely relative to the linguistic conventions that govern particular
discursive practices. The idea, in Derrida’s parlance, is that Austin wants to
‘substitute [the authority of the truth value] for ... the value of force’.34 Fish in
turn claims that Austin jettisons a conception of truth as concerned with ‘a
special relationship [language] bears to the world’, in favor of a conception of
it as concerned with ‘a special relationship it bears to its users’.35
It is at this juncture that we see the in uence, in the writings of Derrida and
Fish, of the very basic exegetical assumption shared with Searle. Like Searle,
Derrida and Fish assume that we cannot rid ourselves of the idea of literal
sentence-meaning without placing the very notion of objective truth at risk.
The Happy Truth 75
Where they differ most dramatically from Searle is simply in maintaining that
Austin is willing to run this risk.
This brings me to my main criticism of prevailing interpretations of
Austin’s lectures. An assumption about a direct conceptual dependence of
objective truth on literal sentence-meaning is foreign to Austin’s thought.
Moreover, it is just as foreign when made by philosophers like Searle as it is
when made by literary-theoretical commentators like Derrida. The dif culty
isn’t merely that Austin himself never explicitly makes such an assumption –
that he launches an assault on the idea of literal sentence-meaning without
any indication that he takes his assault to threaten his right to talk about
objective truth.36 The more serious dif culty is that the assumption depends
for its apparent plausibility on an internally inconsistent understanding of the
implications of abandoning the idea of literal sentence-meaning.
This more serious dif culty might be expressed as follows. To give up the
idea of literal sentence-meaning is to give up an understanding of ourselves as
capable of identifying and assessing bits of language from outside
‘conventions of illocutionary usage’. If we present ourselves as breaking
cleanly with this idea, we commit ourselves to the view that there is no such
thing as a linguistically formulable conception of the world that is accessible
independently of our dealings with (irredeemably conventional) illocutionary
forces. Now we are in effect claiming to have repudiated any thought of an
extra-conventional standpoint from which we might nd a priori metaphy-
sical grounds for representing our preferred conception of the world as merely
relative to a set of illocutionary conventions; or, in other words, from which
we might nd a priori metaphysical grounds for representing our preferred
conception as falling short of objective authority. Indeed, given the
constraints of our envisioned philosophical outlook,34 there cannot even be
a meaningful question about whether our preferred conception of the world
lacks (universal, objective) authority in relation to some hypothetical, extra-
conventionally available alternative. For it is part of this outlook that the idea
of such an alternative is intrinsically confused. The end-result is that we
cannot consistently treat our rejection of the idea of literal sentence-meaning
as undermining the notion of objective truth. If we try, we wind up presenting
ourselves both as repudiating any thought of an extra-conventional standpoint
on language and, at the same time, also as endorsing a view of language which
(in so far as it represents us as hopelessly cut off from objective truth as a
result of our failure to adopt such a standpoint) remains structured by such a
thought. In order to distance ourselves from the idea of literal sentence-
meaning in a rigorously consistent manner, we therefore need to resist the
temptation to represent any distance we achieve as simultaneously a departure
from the notion of objective truth.
The line of reasoning traced out in the last paragraph doesn’t amount to an
attack on the thinker who both protests that talk of objective truth presupposes
76 Alice Crary
NOTES
‘bi-polar’, claims that Austin’s hostility to this conception is a function of an error in his
analysis of statements and that, when allowances are made for this error, the doctrine of
speech acts becomes fully compatible with bi-polarity. See Searle (1968, pp. 422–4) and
(1969b, chs 1–3).
7 Searle (1978).
8 Derrida (1988c).
9 Stanley Fish’s writings contain one of the clearest and most in uential readings of Austin
along these lines. See especiall y Fish (1980). Fish (1989) points out that his interpretatio n
of Austin resembles Derrida’s in important respects. I touch brie y on Derrida’s and Fish’s
work on Austin below in section V.
10 This passage continues: ‘[t]his assumption is no doubt unconscious , no doubt is wrong, but
it is wholly natural in philosoph y apparently’.
11 A number of commentators have argued that Austin overestimates the dif culty of
preserving a sharp distinctio n between constative s and performatives . According to Max
Black, Austin would be in a position to retain his original distinctio n if he simply stipulate d
that no utterance in which a speaker makes a truth claim counts as a performativ e – not
even if the speaker makes the claim in ‘a special way indicated by the utterance ’ (e.g. ‘I
state that it is time to do it’). Black claims that if we impose this constrain t on what counts
as a performativ e ‘many if not all of the dif culties that Austin encountere d [in attempting
to distinguis h constatives and performatives ] will be overcome’ (1969, p. 406). Here Black
is considering passages in which Austin suggests that his initial distinction between
constatives and performative s is threatene d by the discover y that some performatives , while
not true or false per se, have some bearing on the facts. Black also consider s passages in
which Austin suggests that statements, like performatives , are the performance s of
conventiona l procedures . Black says he agrees with Austin that there is a sense in which
even statements are conventional . He adds that if we are to safeguard the constative /
performativ e dichotomy we need to distinguis h between the sense in which the action
performed when a speaker utters a performativ e formula is conventiona l and the sense in
which any act of speech is conventional , and he suggests that this task will be a quite
straightforwar d one (1969, pp. 406–8). But Austin gives us reason to think that the task of
making a general distinctio n between the conventiona l characte r of performative s and that
of constatives will be more dif cult than Black imagines. Indeed Austin suggests that the
task is a hopeless one. When he points out that constative s are in general subject to the same
range of infelicitie s to which performative s are subject, he is effectivel y pointing out that
(at least some) constativ e utterances are the performances of conventiona l procedure s in
just the same way in which performative s are the performance s of conventiona l procedures .
12 Austin writes that ‘in order to explain what can go wrong with statements we cannot just
concentrat e on the proposition involved (whatever that is) as has been done traditionally ’
(p. 52).
13 See above, note 3.
14 Austin (1979c, esp. pp. 119–20) explicitly discusse s this distinctio n between sentences and
statements. He sometimes restricts the applicatio n of the term ‘statement’ to speci c sorts
of utterance s (1979, p. 131). I touch on this further restriction below, at the end of section
III.
15 At the opening of his lectures, Austin claims that philosopher s endorse a traditiona l ideal of
the ‘statement’ in part because they fail to distinguish sentences from statements. They tend
to obscure the fact that ‘not all “sentences ” are (used in making) statements’ through ‘some
loose use of “sentence ” for “statement”’ (p. 1).
16 Indeed Austin believes that we can classify bits of language as either constative s or
performatives – as long as we are clear that what we are doing is classifying acts of speech
accordin g to whether or not, in performing them, the relevant speaker constate s something
and not classifyin g sentences accordin g to whether or not they are suited for constatin g
things. The fact that Austin is willing to use a vocabular y of ‘constatives ’ and
‘performatives ’ later in his lectures should, for this reason, not be taken as a sign that he
equivocate s about whether or not it is possible to sort sentence s as such into constatives and
performatives.
17 In this connection , see Searle (1968).
78 Alice Crary
18 This part of Austin’s lectures sheds helpful light on the bits of his work outside How To Do
Things With Words that have attracted the most critical attention – bits in which he
investigate s ways in which we ordinaril y use a phrase or a sentence and conclude s that it is
‘meaningless’ to use that sentence or phrase in some set of circumstance s in which it is not
ordinaril y used. In the light of Austin’s discussio n of locutionary and illocutionar y acts, it
seems reasonabl e to read Austin as claiming that when expressions are used in
circumstance s which are so different from the circumstance s of their ordinary use that
we have no clue what someone might be saying in using them, it is best to regard those
expression s as mere nonsense . But most commentators take Austin to be making – not a
logical objection to our coming out with combination s of words which, in the particula r
circumstance s of utterance , make no sense, but rather – a pragmatic objection to our
coming out with combinations of words which, while meaningful , are strange or bizarre in
the situations in which they are uttered. Thus, e.g., Searle (1969a) takes Austin to be
concerne d with showing that we don’t ordinaril y advance factual claims unless there is
some reason to doubt their truth. He reads Austin as maintaining that if there is no reason to
doubt the truth of a particula r claim, making it will be ‘bizarre’ or ‘odd’. (For a favorabl e
assessment of Searle’s interpretation , see Grice [1989, pp. 3–21].) Alan White (1969)
suggests a helpful criticism of relevant portions of Searle’s interpretation . White argues that
Searle’s Austin differs from the true Austin in the following respect. Where the former
might subscrib e to the motto ‘if you have nothing worth saying, then keep it to yourself’,
the latter would more likely champion the motto with which Wittgenstei n closes the
Tractatus: ‘Whereof one cannot speak, thereof one must be silent.’ White’s use of this
Tractarian motto draws attention to a striking parallel between Austin’s thought and a
central moment in Wittgenstein’s philosophy , viz., the moment at which Wittgenstei n
employs ‘nonsense ’ as a term of philosophica l assessment . When Wittgenstei n says that a
form of words that we are tempted to utter in philosophy is nonsense , he is making – not a
pragmatic objection to a form of words that attempts to say some impermissible (or bizarre
or strange) thing that cannot properly be formulated , but rather – a logical objection to a
form of words that fails to say anything , that hasn’t yet been given any sense. For
discussio n of the signi cance of this aspect of Wittgenstein’s philosophy , see, e.g., the
essays in Crary and Read (2000).
19 For a straightforwar d version of this sort of view of ‘literal meaning’, see Grice (1989).
20 See above, note 14.
21 One way in which Austin attempts to demonstrat e overlap between the terms we use for
appraisin g statements and the terms we use for appraisin g speech acts (or performatives ) is
by listing a series of sentences and asking us to determine whether they are true or false.
(See the discussio n of ‘France is hexagonal’, ‘Lord Raglan won the Battle of Alma’, and
‘All snow geese migrate to Labrador’ [pp. 142–3].) The point of this exercise is to show
that we can only make judgments of truth or falsity if we imagine a sentence being used, in
some situation, to make a statement.
22 For the signi cance of Austin’s willingness to continue to talk about constativ e and
performativ e utterances even late in his lectures, see above, note 16.
23 This is a central project of Searle (1968).
24 Searle (1968, p. 422).
25 Searle (1968, p. 422). Notice that Searle here uses the term ‘statement’ to refer to what
Austin would call a sentence . If we wanted a statement in Austin’s sense, we might take:
‘The statement the minister made to the family of the deceased to the effect that all men are
mortal is true.’
26 Searle (1968, pp. 421–3).
27 Searle (1968, pp. 415–16). See also (1969b, pp. 44–49).
28 Searle takes the ‘literal meaning’ of a sentence to include elements that have the function of
correspondin g to the world (Searle describe s these elements as making up the
‘propositiona l content’ of the sentence ) and also elements that place constraint s on the
illocutionar y force with which the sentence can be uttered (1968, p. 412). His thought is
that, for any given illocutionar y act, it is possible to come up with a linguistic formula
which (on account of constraints its ‘literal meaning’ places on illocutionar y force) can only
be used to perform that illocutionar y act. (See also 1969b, pp. 20–21.) Thus, e.g., he thinks
The Happy Truth 79
that there are linguistic formulae which (on account of their ‘literal meanings’) can only be
used to make promises and also linguistic formulae that can only be used to make
statements. In presentin g his view of the relation between ‘literal meaning’ and
illocutionar y force, Searle resurrect s in a slightly new guise an idea Austin plainly rejects
– namely that some combination s of words are ideally suited (without regard to the
circumstance s in which they are used) for making statements.
29 Searle (1992).
30 See the exchang e between Derrida and Searle that includes Derrida’s ‘signature Event
Context’ (Derrida [1988c], rst published in vol. 1 of Glyph in 1977), Searle’s response in
‘Reiterating the Differences : A Reply to Derrida’ (Searle (1977), publishe d in vol. 2 of
Glyph) and Derrida’s counter to Searle in ‘Limited Inc a b c ...’ (Derrida [1988b], which
also rst appeared in vol. 2 of Glyph).
31 Derrida (1988c) tells us that Austin’s efforts to discredit this idea – which Derrida refers to
as Austin’s ‘critique of linguisticism ’ – represen t the moment in Austin’s thought ‘that most
intereste d and convince d [him] in Austin’s undertaking ’ (p. 19). It is, however, important to
acknowledg e that, although Derrida thus presents himself as a fan of Austin’s attack on a
classical ideal of the ‘statement’, one of Derrida’s main concerns in his discussion s of How
To Do Things With Words is criticizin g what he sees as Austin’s tendency to allow this
ideal to continue to shape his own thinking. (See esp. Derrida [1988c, pp. 13–15].) Derrida
is committed to defendin g the somewhat peculiar exegetica l position that, when Austin
draws what by Derrida’s own lights counts as the main conclusion of his lectures (viz., that
there is no such thing as identifying the locutionar y act someone performs in using certain
words apart from an appreciatio n of the illocutionar y force with which it is uttered in
particula r circumstances) , he is assenting to a view that he himself in various ways actively
resists. (See Fish [1989] for a favorabl e assessment of this Derridean position. ) I am
inclined to think that Derrida’s claim that Austin betrays his own central insight rests on a
misinterpretation , but I cannot take up this matter in this article. (For a detailed and very
helpful discussion, see Cavell [1994].) My primary concerns here are: (i) the fact that
Derrida, unlike Searle, takes Austin to be rejecting the idea of literal sentence-meanin g and
also (ii) the further fact that the conclusion s Derrida draws from this Austinian gesture of
rejection reveal that, like Searle, he is drawing on an assumption about a conceptual
dependenc e of objective truth on literal sentence-meanin g – an assumption which (I am
arguing) is not at home in Austin’s thought.
32 Derrida (1988c, p. 13).
33 Fish (1980, p. 198), emphasis in the original.
34 Derrida (1988c, p. 13), emphasis added.
35 Fish (1980, p. 241).
36 See sections III and IV, above.
37 This article grew out of a series of conversation s I had in the mid-1990s with three of my
teachers – Nuel Belnap (with whom I read several of Austin’s papers in the Spring of 1993
while serving as the University of Pittsburgh Philosophy Department’s Alan Ross Anderson
Fellow), Stanley Cavell (with whom I discusse d the exchang e between Austin and Derrida
on several occasion s during the academic year 1993–94), and Annette Baier (with whom I
read How To Do Things With Words in the Fall of 1994). I am grateful to all three for their
guidance , and to Cavell and Baier for their help with an early draft. I owe thanks to Jim
Conant and Nancy Bauer for stimulating discussions about Austin over the years, as well as
for many constructiv e criticisms of this article; to Cora Diamond and Elijah Millgram for
their encouragemen t and for their helpful comments on a relatively recent draft; and to an
anonymous referee for useful suggestions about preparin g the article for publication .
R E FE R E NC E S
Austin, J. L. 1962. How To Do Things With Words. Cambridge, MA: Harvard University Press.
Austin, J. L. 1979a. Philosophica l Papers, 3rd ed., in J. O. Urmson and G . J. Warnock (eds).
Oxford: Oxford University Press.
Austin, J. L. 1979b. ‘A Plea for Excuses’, Philosophica l Papers, 175–204.
80 Alice Crary
Alice Crary, Department of Philosophy, Graduate Faculty of Political and Social Science,
The New School, 65 Fifth Avenue, New York, NY 10003, USA. E-mail: CraryA@
newschool.edu