J. L. Austin was one of the more influential British philosophers of his time, due to his rigorous thought, extraordinary personality, and innovative philosophical method. According to John Searle, he was both passionately loved and hated by his contemporaries. Like Socrates, he seemed to destroy all philosophical orthodoxy without presenting an alternative, equally comforting, orthodoxy.
Austin is best known for two major contributions to contemporary philosophy: first, his ‘linguistic phenomenology’, a peculiar method of philosophical analysis of the concepts and ways of expression of everyday language; and second, speech act theory, the idea that every use of language carries a performative dimension (in the well-known slogan, “to say something is to do something”). Speech act theory has had consequences and import in research fields as diverse as philosophy of language, ethics, political philosophy, philosophy of law, linguistics, artificial intelligence and feminist philosophy.
This article describes Austin’s linguistic method and his speech act theory, and it describes the original contributions he made to epistemology and philosophy of action. It closes by focusing on two main developments of speech act theory─the dispute between conventionalism and intentionalism, and the debate on free speech, pornography, and censorship.
John Langshaw Austin, born on March 26th 1911 in Lancaster, England. He was trained as a classicist at Balliol College Oxford. He first came to philosophy by studying Aristotle, who deeply influenced his own philosophical method. He also worked on the philosophy of Leibniz and translated Frege’s Grundlagen. Austin spent his whole academic life in Oxford, where he was White’s Professor of Moral Philosophy from 1952 until his death in 1960. During the Second World War Austin was commissioned in the Intelligence Corps, and played a leading role in the organization of D-Day, leaving the British Army in 1945 with the rank of Lt. Colonel. Austin published only seven articles. According to Searle, Austin’s reluctance to publish was partly characteristic of his own attitude, but also it was part of the culture of Oxford at the time: “Oxford had a long tradition of not publishing during one’s lifetime, indeed it was regarded as slightly vulgar to publish” (Searle 2001, 227). Most of Austin’s work was thus published posthumously, and includes a collection of papers (Austin 1961), and two series of lectures reconstructed by the editors on the basis of Austin’s lecture notes: the lectures on perception, edited by Geoffrey Warnock (Austin 1962a), and the William James Lectures held at Harvard in 1955, devoted to speech acts, and edited by James O. Urmson (Austin 1962b) for the first edition, and by Urmson and Marina Sbisa for the second (Austin 1975).
Austin had a profound dissatisfaction not only with the traditional way of philosophizing, but also with Logical Positivism (whose leading figure in Oxford was Alfred J. Ayer). In particular, his dissatisfaction was directed towards a way of practicing philosophy which, in Austin’s view, was responsible for the production of tidy dichotomies, and instead of clarifying the problems at issue, seemed to lead to oversimplifications and dogmatic and preconceived schemes of thought. Austin thus developed a new philosophical methodology and style, which became paradigmatic of Ordinary Language Philosophy. Austin does not claim that this method is the only correct method to adopt. Rather, it represents a valuable preliminary approach to at least some of the most stubborn problems in the tradition of Western philosophy, such as those of freedom, responsibility, and perception. According to Austin, the starting point in philosophy should be the analysis of the concepts and ways of expression of everyday language, and the reconnaissance of our ordinary language. This would help to dismantle the ‘philosophical mistakes’ induced by the way philosophers use certain ordinary words, on the one hand, and, on the other, to gain access to actual features in the world picked out by the expressions we use to describe it.
Ordinary language is not the last word: in principle it can everywhere be supplemented and improved upon and superseded. Only remember, it is the first word. [Austin 1956a/1961, 185]
According to Austin, in ordinary language are deposited all the distinctions and connections established by human beings, as if our words in their daily uses “had stood up to the long test of the survival of the fittest” (Austin 1956a/1961, 182). It is necessary, first of all, to carefully consider the terminology available to us, by making a list of the expressions relevant to the domain at issue: a sort of ‘linguistic phenomenology’ carried out with the help of the dictionary and of the imagination, by searching for combinations of expressions and synonyms, devising linguistic thought-experiments and unusual contexts and scenarios, and speculating about our linguistic reactions to them. The examination of ordinary language enables us to pay attention to the richness of linguistic facts and to tackle philosophical problems from a fresh and unprejudiced perspective.
To be sure, this is not a new methodology in the history of philosophy. Still, this strategy is now carried out with distinctive meticulousness and on a large scale on the one hand, and is undertaken and evaluated collectively, so as to gain a reasonable consensus, on the other. For Austin, philosophy is not an endeavor to be pursued privately, but a collective labor. This was in fact the gist of Austin’s “Saturday mornings,” weekly meetings held during term at Oxford and attended by philosophers of language, moral philosophers and jurists. Austin’s method was better meant for philosophical discussion and research along these lines than for publication: we can nonetheless fully appreciate it in How to Do Things with Words (1962b), and in papers like “A Plea for Excuses” (1956a), and “Ifs and Cans” (1956b).
Austin’s method has been regarded by some as pedantic, as a mere insistence that all we must do is use our words carefully, with no genuine interest in the phenomena which arouse our philosophical worries. There are indeed limitations to his methodology: on the one hand, many philosophical questions still remain untouched even after a meticulous reformulation; on the other hand, our everyday language does not embody all the distinctions which could be relevant for a philosophical inquiry (compare Searle 2001, 228-229). But Austin is far from being concerned merely by language: his focus is always also on the phenomena talked about – as he says, “It takes two to make a truth” (Austin 1950/1961, 124fn; compare Martin 2007).
With the help of his innovative methodology, Austin takes a new stance towards our everyday language. As is well known, philosophers and logicians like Gottlob Frege, Bertrand Russell, the earlier Ludwig Wittgenstein, Alfred Tarski and Willard Quine want to build a perfect language for philosophical and scientific communication, that is, an artificial language devoid of all the ambiguities and imperfections that characterize natural languages. Conversely, ordinary language philosophers (besides Austin, the later Wittgenstein, Friedrich Waismann, Paul Grice, Peter Strawson) view natural language as an autonomous object of analysis – and its apparent imperfections as signs of richness and expressive power.
In a formal language, semantic conventions associate with each term and each sentence a fixed meaning, once and for all. By contrast, the expressions of a natural language seem essentially incomplete; as a result, it seems impossible to fully verify our everyday sentences. The meanings of our terms are only partially constrained, depending on the beliefs, desires, goals, activities, and institutions of our linguistic community. The boundaries, even when temporarily fixed, are unstable and open to new uses and new conventions in unusual situations. In “The Meaning of a Word,” Austin takes into consideration different contexts of utterance of sentences containing familiar terms, in order to create unusual occasions of use: extraordinary or strange cases are meant to contrast with our intuitions and reveal moments of tension hidden in natural language. What are we to say about “It’s a goldfinch” uttered about a goldfinch that “does something outrageous (explodes, quotes Mrs. Woolf, or what not)”? (Austin 1940/1961, 88). Austin’s main point is that it is in principle impossible to foresee all the possible circumstances which could lead us to modify or retract a sentence. Our everyday terms are extremely flexible, and can still be used in odd cases. Concerning “That man is neither at home nor not at home,” Austin writes: “Somehow we cannot see what this ‘could mean’ – there are no semantic conventions, explicit or implicit, to cover this case: yet it is not prohibited in any way – there are no limiting rules about what we might or might not say in extraordinary cases” (Austin 1940/1961, 68). Of a dead man, lying on his bed, what would we say? That he is at home? That he is not at home?
Similarly, should the utterance “France is hexagonal” be taken as true or false? According to Austin we must take into consideration the speaker’s goals and intentions, the circumstances of utterance, and the obligations we undertake in asserting something. Assertions are not simply true or false, but more or less objective, adequate, exaggerated, and rough: “‘true’ and ‘false’ […] do not stand for anything simple at all; but only 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” (Austin 1975, 145).
Austin’s most celebrated contribution to contemporary philosophy is his theory of speech acts, presented in How to Do Things with Words (Austin 1975). While for philosophers interested mainly in formal languages the main function of language is describing reality, representing states of affairs and making assertions about the world, for Austin our utterances have a variety of different uses. A similar point is made in Philosophical Investigations by Wittgenstein, who underlines the “countless” uses we may put our sentences to (Wittgenstein 1953: § 23). Austin contrasts the “desperate” Wittgensteinian image of the countless uses of language with his accurate catalogue of the various speech acts we may perform – a taxonomy similar to the one employed by an entomologist trying to classify the many (but not countless) species of beetles.
Not all utterances, then, are assertions concerning states of affairs. Take Austin’s examples
(1) I name this ship the ‘Queen Elisabeth’
as uttered in the course of the launching of a ship, or
(2) I bet you sixpence it will rain tomorrow.
The utterer of (1) or (2) is not describing the launching ceremony or a bet, but doing it. By uttering these sentences we bring about new facts, “as distinguished from producing consequences in the sense of bringing about states of affairs in the ‘normal’ way, that is, changes in the natural course of events” (Austin 1975: 117): by uttering (1) or (2) we modify the social reality, institute new conventions, and undertake obligations. In the first lessons of How to Do Things with Words, Austin traces a tentative distinction between constatives and performatives, to be abandoned in the subsequent lessons. Constatives, on the one hand, are sentences like
(3) The cat is on the mat:
they aim to describe states of affairs and are assessable as true or false. Performatives like (1) and (2), on the other hand, do rather than report something: they perform acts governed by norms and institutions (such as the act of marrying or baptizing) or social conventions (such as the act of betting or promising) and do not seem assessible as true or false. This last controversial claim is not argued for (“I assert this as obvious and do not argue it,” Austin 1975, 6): the claim is only provisional and subject to revision in the light of later sections (Austin 1975, 4n).
According to Austin it is possible and fruitful to shed light on standard cases of successful communication, and to specify the conditions for the smooth functioning of a performative, by focusing on non-standard cases and communicative failures. As we have said, performatives cannot be assessed as true or false, but they are subject to different types of invalidity or failure, called “infelicities.” In some cases the attempt to perform an act fails or “misfires.” The act is “null and void” on the basis of the violation of two kinds of rules:
A.1: there must exist an accepted conventional procedure having a certain conventional effect, that procedure to include the uttering of certain words by certain persons in certain circumstances;
A.2: that procedure must be invoked in adequate circumstances and by appropriate persons.
Further infelicities concern the execution of the procedure, for it must be executed by all participants both
B.1: correctly, and
Finally, there are cases in which the performance of an act is achieved, but there is an abuse of the procedure, due to the violation of two kinds of rules:
C.1: the procedure must be executed by the speaker with appropriate thoughts, feelings or intentions;
C.2: the participants must subsequently conduct themselves in accordance with the procedure performed.
As we said, in How to Do Things with Words Austin draws the distinction between constatives and performatives merely as a preliminary to the presentation of his main thesis, namely that there is a performative dimension in any use of language. The putative class of performatives seems to admit only specific verbs (like to promise, to bet, to apologize, to order), all in the first person singular present. Any attempt to characterize the class with grammatical or lexical criteria, however, is bound to fail. We may in fact perform the act of, say, ordering by using an explicit performative, as in
(4) I order you to close the door
but also with
(5) Close the door!
Similarly, there are performative verbs also for acts of stating, asserting, or concluding, as in
(6) I assert that the Earth is flat.
The very distinction between utterances assessable along the dimension of truth and falsehood (constatives) and utterances assessable along the dimension of felicity or infelicity (performatives) is a mere illusion. To show this, Austin presents two arguments:
a) on the one hand, constatives may be assessed as happy or unhappy: like performatives, assertions require appropriate conditions for their felicitous performance (to give an example, it does not seem appropriate to make an assertion one does not believe);
b) on the other hand, performatives may be assessed in terms of truth and falsehood, or in terms of some conformity to the facts: of a verdict we say that it is fair or unfair, of a piece of advice that it is good or bad, of praise that it is deserved or not.
By a) and b) Austin is led to the conclusion that the distinction between constatives and performatives is inadequate: all sentences are tools we use in order to do something – to say something is always to do something. Therefore it is necessary to develop a general theory of the uses of language and of the acts we perform by uttering a sentence: a general theory of what Austin calls illocutionary force.
Within the same total speech act Austin distinguishes three different acts: locutionary, illocutionary and perlocutionary.
In the last lesson of How to Do Things with Words Austin tentatively singles out five classes of illocutionary acts, using as a starting point a list of explicit performative verbs: Verdictives, Exercitives, Commissives, Behabitives, Expositives.
Austin characterizes the illocutionary act as the conventional aspect of language (to be contrasted with the perlocutionay act). As we said before, for any speech act there must exist an accepted conventional procedure having a certain conventional effect (condition A.1): if the conventional procedure is executed according to further conditions, the act is successfully performed. This claim seems plausible as far as institutional or social acts (like naming a ship, or betting) are concerned: the conventional dimension is here manifest because it is our society (and sometimes our laws) that validates those acts. The claim seems less plausible as far as speech acts in general are concerned: nothing conventional, or semantic, makes of (5) an order, or a challenge, or an entreaty – the illocutionary force of the utterance is fixed by the context of utterance. More generally, according to Austin the speaker’s intentions play only a minor role in the performance of a speech act (violation of condition C.1 leads to an abuse of the procedure, but not to a failure of the speech act). Drawing on Gricean ideas, Peter Strawson argues that what makes of (5) an illocutionary act of ordering instead of entreating are the speaker’s intentions – intentions the speaker may (but need not) make available to the audience using linguistic conventions:
I do not want to deny that there may be conventional postures or procedures for entreating… But I do want to deny that an act of entreaty can be performed only as conforming to some such conventions. What makes X’s words to Y an entreaty not to go is something complex enough, no doubt relating to X’s situation, attitude to Y, manner, and current intention. [Strawson 1964, 444; compare Warnock 1973 and Bach & Harnish 1979]
Marina Sbisa disagrees with Strawson’s reading of the conventionality of illocutionary acts and identifies two distinct claims in Austin’s conventionalism: (a) illocutionary acts are performed via conventional devices (like linguistic conventions); and (b) illocutionary acts produce conventional effects. Austin specifies three kinds of conventional effects: the performance of an illocutionary act involves the securing of uptake, that is, bringing about the understanding of the meaning and force of the locution; the illocutionary act takes effect in conventional ways, as distinguished from producing consequences in the sense of bringing about changes in the natural course of events; and many illocutionary acts invite by convention a response or sequel (Austin 1975, 116-117). According to Sbisa, Austin deals too briefly with the conventionality of the effects produced by illocutionary acts (b) as opposed to the conventionality of the devices by which we perform illocutionary acts (a), leaving room for Strawson’s distinction between two groups of illocutionary acts: those depending on some convention the speaker follows, and those depending on a particular kind of intention of the speaker (Sbisa 2013, 26). On this point see below, § 5.a.
There are two main examples of Austin’s philosophical method applied to epistemological issues: the paper “Other Minds” (1946), and the series of lectures Sense and Sensibilia, delivered at Oxford and Berkeley during the decade from 1947 to 1958, and published in 1962. “Other Minds” is a paper given in the homonymous symposium at joint sessions of the Mind Association and the Aristotelian Society (John Wisdom, Alfred J. Ayer, and Austin were the main participants), and its topic was a much debated one in the middle decades of the twentieth century. In Sense and Sensibilia Austin applies his linguistic analysis to the sense-data theory and the more general foundational theory of knowledge, within which sense-data played the role of the basis of the very structure of empirical knowledge, in order to gain a clarification of the concept of perception.
These lectures represent a very detailed criticism of the claims put forward by A. J. Ayer in The Foundations of Empirical Knowledge (1940), and, to a lesser extent, of those contained in H. H. Price’s Perception (1932) and G. J. Warnock’s Berkeley (1953). Austin challenges the sense-data theory, according to which we never directly perceive material objects. On the contrary, it is claimed by such theory, we perceive nothing but sense-data.
The notion of sense-data is introduced to identify the object of perception in abnormal, exceptional cases, for example, refraction, mirages, mirror-images, hallucinations, and so forth. In such cases perceptions can either be ‘qualitatively delusive’ or ‘existentially delusive,’ depending on whether sense-data endow material things with qualities that they do not really possess, or the material things presented do not exist at all. In all such cases, the sense-data theorist maintains, we directly perceive sense-data. The subsequent step in this argument, named the argument from illusion, is to claim that in ordinary cases too we directly perceive merely sense-data.
Austin’s goal is not to answer the question “What are the objects of perception?” Austin aims to get rid of “such illusions as ‘the argument from illusion’” on the one hand, and to offer on the other a “technique for dissolving philosophical worries” by clarifying the meaning of words such as ‘real,’ ‘look,’ ‘appear’ and ‘seem’ (Austin 1962a, 4-5). The argument from illusion amounts to a misconception inasmuch as it introduces a bogus dichotomy: that between sense-data and material objects. Austin challenges this dichotomy, and the subsequent claim that abnormal, illusory perceptions do not differ from normal, veridical ones in terms of quality (in both cases sense-data are perceived, though in different degrees), by presenting different cases of perceptions in order to show that “there is no one kind of thing that we ‘perceive’ but many different kinds, the number being reducible if at all by scientific investigation and not by philosophy” (Austin 1962a, 4).
Besides chairs, tables, pens and cigarettes, indicated by the sense-data theorist as examples of material objects, Austin draws attention to rainbows, shadows, flames, vapors and gases as cases of things we ordinarily say that we perceive, even though we would not classify them as material things. Likewise, Austin argues, there is no single way in which we are ‘deceived by senses’ (that is, to perceive something unreal or not material), but “things may go wrong […] in lots of different ways – which don’t have to be, and must not be assumed to be, classifiable in any general fashion” (Austin 1962a, 13). Moreover, Austin asks whether we would be prone to speak of ‘illusions’ with reference to dreams, phenomena of perspective, photos, mirror-images or pictures on the screen at the cinema. By recalling the familiarity of the circumstances in which we encounter these phenomena and the ways in which we ordinarily consider them, Austin intends to show how the dichotomies between sense-data and material objects, and between illusory perceptions and veridical ones, are in fact spurious alternatives.
The facts of perceptions are “diverse and complicated” and the analysis of words in their contexts of use enables us to make the subtle distinctions obliterated by the “obsession” some philosophers have for certain words (for example, ‘real’ and ‘reality’), and by the lack of attention to the (not even remotely interchangeable) uses of verbs like ‘look,’ ‘appear,’ and ‘seem.’ The way the sense-data theorists use the words ‘real’ and ‘directly’ in the argument from illusion is not the ordinary use of these words, but a new use, which nevertheless fails to be explained. Austin does not want to rule out the possibility of tracing, for theoretical purposes, new distinctions, and thus of emending our linguistic practices by introducing technical terms, but he rather proposes always to pay attention to the ordinary uses of our words, in order to avoid oversimplifications and distortions.
As an example, Austin examines the word ‘real’ and contrasts the ordinary, firmly established meanings of that word as fixed by the everyday ways we use it to the ways it is used by sense-data theorists in their arguments. What Austin recommends is a careful consideration of the ordinary, multifarious meanings of that word in order not to posit, for example, a non-natural quality designed by that word, common to all the things to which that word is attributed (‘real ducks,’ ‘real cream,’ ‘real progress,’ ‘real color,’ ‘real shape,’ and so forth).
Austin highlights the complexities proper to the uses of ‘real’ by observing that it is (i) a substantive-hungry word that often plays the role of (ii) adjuster-word, a word by means of which “other words are adjusted to meet the innumerable and unforeseeable demands of world upon language” (Austin 1962a, 73). Like ‘good,’ it is (iii) a dimension-word, that is, “the most general and comprehensive term in a whole group of terms of the same kind, terms that fulfil the same function” (Austin 1962a, 71): that is, ‘true,’ ‘proper,’ ‘genuine,’ ‘live,’ ‘natural,’ ‘authentic,’ as opposed to terms such as ‘false,’ ‘artificial,’ ‘fake,’ ‘bogus,’ ‘synthetic,’ ‘toy,’ but also to nouns like ‘dream,’ ‘illusion,’ ‘mirage,’ ‘hallucination.’ ‘Real,’ is also (iv) a word whose negative use “wears the trousers” (a trouser-word) (Austin 1962a, 70).
In order to determine the meaning of ‘real’ we have to consider, case by case, the ways and contexts in which it is used. Only by doing so, according to Austin, can we avoid introducing false dichotomies (for a criticism of Austin’s attack on sense-data see Ayer 1967 and Smith 2002).
In this paper Austin tackles the philosophical problems of the possibility of knowing someone else’s mental states (for example, that another man is angry) and of the reliability of the reasons to which we appeal when we justify our assertions about particular empirical facts (for example, that the bird moving around my garden is a goldfinch). The target of Austin’s analysis is the skeptical outcome of the challenge of such a possibility on the part of certain philosophers (in this case Austin is addressing some of John Wisdom’s contentions). As to the knowledge of particular empirical facts, given that “human intellect and senses are, indeed, inherently fallible and delusive” (Austin 1946/1961, 98), the skeptic claims that we should never, or almost never, say that we know something, except for what I can perceive with my senses now, for example, “Here is something that looks red to me now.” On the other hand, the possibility of knowing someone else’s mental states is challenged by means of the idea of a privileged access to our own sensations and mental states, such that only about them can we not ‘be wrong’ “in the most favoured sense” (Austin 1946/1961, 90).
Austin engages in an examination of the kinds of answers we would provide, in ordinary, concrete and specific circumstances, to challenges to our claims of knowledge. For instance, in responding to someone’s question, “How do you know?” in the face of my claim “That is a goldfinch,” my answer could appeal to my past experience, by virtue of which I have learned something about goldfinches, and hence to the criteria for determining that something is a goldfinch, or to the circumstances of the current case, which enable me to determine that the bird moving around my garden now is a goldfinch. The ways in which, in ordinary circumstances, our claims can be challenged, or be wrong, are specific (ways that the context helps us to determine), and there are recognized procedures appropriate to the particular type of case to which we can appeal to justify or verify such claims.
The precautions to take, in ordinary cases, in order to claim to know something “cannot be more than reasonable, relative to current intents and purposes” (Austin 1946/1961, 88), inasmuch as in order to suppose that one is mistaken there must be some concrete reason relative to the specific case. On the contrary, the “wile of the metaphysician,” Austin claims, amounts to phrasing her doubts and questions in a very general way, and “not specifying or limiting what may be wrong,” “so that I feel at a loss ‘how to prove’” what she has challenged (Austin 1946/1961, 87).
By drawing a parallel with the performative formula ‘I promise,’ Austin claims that in uttering ‘I know’ the speaker does not describe her mental state (this would be, in Austin’s terms, a descriptive fallacy). Rather, in the appropriate circumstances, she does something: she gives others her word, that is, her authority for saying that ‘S is P.’
Austin’s analysis of the epistemological terms in their ordinary, specific uses is meant to determine the conditions under which our claims are felicitous, successful speech acts. These conditions are the ones we typically appeal to in order to justify our claims of knowledge should they be challenged.
Whether Austin’s strategy proves to be successful against the skeptical challenge, which rests on a metaphysical and logical possibility, is a further issue to be resolved.
Austin objects to the idea (as claimed for example, by Wisdom) that we know (if ever we do) someone else’s feelings only from the physical symptoms of these feelings: we never know someone else’s feelings in themselves, as we know our own. According to Austin, claims about someone else’s mental states can be tackled like those about particular empirical facts, even though the former are more complex, owing to “the very special nature (grammar, logic) of feelings” (Austin 1946/1961, 105). It is on this special nature that the analysis of Austin in this paper is meant to shed light.
To affirm of someone, ‘I know he is angry,’ requires, on the one hand, a certain familiarity with the person to whom we are attributing the feeling; in particular, familiarity in situations of the same type as the current one. On the other hand, it seems to be necessary to have had a first-person experience of the relevant feeling/emotion.
A feeling (say anger), Austin claims, presents a close connection both with its natural expressions/manifestations, and with the natural occasions of those manifestations, so that “it seems fair to say that ‘being angry’ is in many respects like ‘having mumps’. It is a description of a whole pattern of events, including occasion, symptoms, feeling and manifestation, and possibly other factors besides” (Austin 1946/1961, 109).
Against Wisdom’s claim that we never get at someone else’s anger, but only at the symptoms/signs of his anger, Austin draws attention to the ways we talk about others’ feelings, and highlights the general pattern of events “peculiar to the case of ‘feelings’ (emotions)” on which our attributions are based (Austin 1946/1961, 110). Moreover, emphasis is placed on the fact that in order for feelings and emotions to be attributed, and also self-attributed, a problem of recognition, and of familiarity with the complexities of such pattern, seems to be in place, due to the very way in which the uses of the relevant terms have been learnt.
Emotion terms, for their part, are vague, for, on the one hand, they are often applied to a wide range of situations, while on the other, the patterns they cover are rather complex, such that in “unorthodox” cases there may be hesitation in attribution. Apart from this kind of intrinsic vagueness, doubts may arise as to the correctness of a feeling attribution, or its authenticity, due to cases of misunderstanding or deception. But these cases are special ones, and there are, as in the goldfinch case, “established procedures” for dealing with them.
Unlike the goldfinch case, in which “sensa are dumb” (Austin 1946/1961, 97), in the case of feeling attribution a special place is occupied, within its complex pattern of events, by “the man’s own statement as to what his feelings are” (Austin 1946/1961, 113). According to Austin, “believing in other persons, in authority and testimony, is an essential part of the act of communicating, an act which we all constantly perform. It is as much an irreducible part of our experience as, say, giving promises, or playing competitive games, or even sensing coloured patches” (Austin 1946/1961, 115). Austin thus aims at blocking the skeptical argument by claiming that the possibility of knowing others’ states of minds and feelings is a constitutive feature of our ordinary practices as such, for which there is no “justification” (ibid.). Again, it may still be open to contention whether this is sufficient to refute skepticism.
Austin’s contribution to the philosophy of action is traceable mainly to two papers: “A Plea for Excuses” (1956a) and “Three Ways of Spilling Ink” (1966), where the notions of ‘doing an action,’ and ‘doing something’ are clarified by means of the linguistic analysis of excuses, that is, by considering “the different ways, and different words, in which on occasion we may try to get out of things, to show that we didn’t act ‘freely’ or were not ‘responsible’” (Austin 1966/1961, 273). According to the method dear to Austin, through the analysis of abnormal cases, or failures, it is possible to throw light on the normal and standard cases. An examination of excuses should enable us to gain an understanding of the notion of action, by means of the preliminary elucidation of the notions of responsibility and freedom.
As for the case of ‘knowing,’ Austin’s contribution is one of clarification of use, which sheds light on the notion of ‘doing an action.’
According to Austin, from the analysis of the modifying expressions occurring in excuses (for example, ‘unwittingly,’ ‘impulsively’) and in accusations (‘deliberately,’ ‘purposely,’ ‘on purpose’), it is possible to classify the different breakdowns affecting actions, and thus to dismantle the complex internal details of the machinery of action. Far from being reducible to merely making some bodily movements, doing an action is organized into different stages: the intelligence, the appreciation of the situation, the planning, the decision, and the execution. Moreover, apart from the stages, “we can generally split up what might be named as one action in several distinct ways, into different stretches or phases” (Austin 1956a/1961, 201). In particular, by using a certain term to describe what someone did, we can cover either a smaller or larger stretch of events, distinguishing the act from its consequences, results, or effects. Austin thus points out that it is crucial to determine ‘what modifies what,’ that is, what is being excused, because of the different possible ways in which it is possible to refer to ‘what she did.’ Depending on the way we delimit the act, we may hold a person responsible or not, hence it is extremely important to determine with precision what is being justified.
In order to ascertain someone’s responsibility for a certain action, for example, a child that has spilled ink in class, it is necessary to consider as distinct cases whether he did it ‘intentionally,’ ‘deliberately,’ or ‘on purpose.’ Austin reveals the differences among the three terms in two steps. First, he considers imaginary actions the description of which presents two of the terms as expressly dissociated: for example, to do something intentionally, but not deliberately (for example, to impulsively stretch out one’s hand to make things up at the end of a quarrel; see Austin 1966/1961, 276-277); intentionally, but not on purpose (for example, to act wantonly); deliberately, but not intentionally (for example, unintended, though foreseeable, consequences/results of some actions of mine – to ruin one’s debtor by insisting on payment of due debts; see Austin 1966/1961, 278); and on purpose, but not intentionally (this last case seems to verge on impossibility, and has a paradoxical flavor). The logical limits of these combinations enable Austin to single out the differences among the concepts under investigation. Subsequently, as a second step, such differences are made apparent by an analysis of the grammar and philology of the terms (adjectival terminations, negative forms, the prepositions used to form adverbial expressions, and so forth). From this examination each term turns out to mark different ways in which it is possible to do an action, or different ‘aspects’ of action.
Austin’s analysis of the notions of action and responsibility casts light on that of freedom, which is clarified by the examination of “all the ways in which each action may not be ‘free’” (Austin 1956a/1961, 180). Like the word ‘real,’ also in the case of ‘free’ it is the negative use that wears the trousers: the notion of freedom derives its meaning from the concepts excluded, case by case, by each use of the term ‘free’ (for example, consider the following utterances: “I had to escape, since he threatened me with a knife;” “The glass slipped out of my hand because of my astonishment;” “I can’t help checking the email every five minutes”). By considering the specific, ordinary situations in which it is possible or not to ascertain someone’s responsibility for an act (where an issue of responsibility arises), the notions of freedom and responsibility emerge as closely intertwined. In Austin’s words: “As ‘truth’ is not a name for a characteristic of assertions, so ‘freedom’ is not a name for a characteristic of actions, but a name of a dimension in which actions are assessed” (Austin 1956a/1961, 180). Austin does not provide a positive account of the notion of freedom: it is rather elucidated through attention to the different ways in which our actions may be free.
Since the end of the 1960s Speech Act Theory (SAT) has been developed in many directions. Here we will concentrate on two main strands: the dispute between conventionalism and intentionalism on the one hand, and the debate on pornography, free speech, and censorship on the other.
Peter Strawson’s contribution to Austin’s SAT, with ideas drawn from Paul Grice’s analysis of the notion of non-natural meaning, marks the beginning of the dispute over the role played by convention and intention respectively in the performance of speech acts. Although almost all the developments of SAT contain, in different degrees, both a conventionalist and an intentionalist element, it may be useful to distinguish two main traditions, depending on the preponderance of one element over the other: a conventionalist, Austinian tradition, and an intentionalist, neo-Gricean one. There are two versions of conventionalism: John Searle’s conventionalism of the means, and Marina Sbisa’s conventionalism of the effects (the distinction is marked by Sbisa).
Central to Searle’s systematization of SAT is the hypothesis that speaking a language amounts to being engaged in a rule-governed activity, and that human languages may be regarded as the conventional realizations of a set of underlying constitutive rules, that is, rules that create, rather than just regulate, an activity, to the effect that the existence of such activity logically depends on that of the rules constituting it. Searle’s analysis aims at specifying a set of necessary and sufficient conditions for the successful performance of an illocutionary act (Searle concentrates on the act of promising, but the analysis is further extended to the other types of illocutionary acts, of which several classifications have been developed within SAT). Starting from the felicitous conditions for the performance of an act, a set of semantic rules for the use of ‘the illocutionary force indicating device for that act is extracted. Searle’s position thus qualifies as a conventionalism of the means because the felicitous performance of an illocutionary act obtains thanks to the conformity of the meaning of the utterance used to perform it to the linguistic conventions proper to the language within which the act is performed.
Sbisa developed Austin’s thesis of the conventionality of the illocutionary force by linking it to the second of the three kinds of conventional effects characteristic of the illocutionary act: the production of states of affairs in a way different from natural causation, namely, the production of normative states of affairs, such as obligations, commitments, entitlements, rights, and so forth. According to Sbisa, each type of illocutionary act produces specific conventional effects, which correspond to “assignments to or cancellations from each one of the participants of modal predicates” (Sbisa 2001, 1797), which are conventional insofar as their production depends on intersubjective agreement. The hallmark of such effects is, unlike physical actions, their being liable to annulment, their defeasibility.
The main exemplification of the intentionalist tradition within SAT is the analysis developed by Kent Bach and Robert Harnish. In the opposite direction relative to Searle’s views, Bach and Harnish claim that the link between the linguistic structure of an utterance and the illocutionary act it serves to perform is never semantic, or conventional, but rather inferential.
In order for the illocutionary act to be determined by the hearer, in addition to ‘what is said’ (the semantic content of the utterance), the following are also necessary: the speaker’s illocutionary communicative intention (this intention is reflexive: its fulfillment consists in its being recognized, on the part of the hearer, as intended to be recognized); the contextual beliefs shared by the interlocutors; general beliefs about the communicative situation which play the role of presumptions, that is, expectations about the interlocutor’s behavior which are so basic that they constitute the very conditions of possibility of the communicative exchange (for example, the belief of belonging to the same linguistic community, and that in force to the effect that whenever a member of the community utters an expression, she is doing so with some recognizable illocutionary intent); and, finally, conversational assumptions, drawn from Gricean conversational maxims.
All these elements combine to bring about the inference that enables the hearer to move from the level of the utterance act up to the illocutionary level, going through the locutionary one (note that Bach and Harnish, and Searle himself, part company with Austin in distinguishing the components of the speech act). Bach and Harnish put forward a model, the Speech Act Schema (SAS), which represents the pattern of the inference a hearer follows (and that the speaker intends that she follow) in order to understand the illocutionary act as an act of a certain type (as an order, a promise, an assertion, and so forth).
The merit of the intentionalist analysis offered by Bach and Harnish is that it attempts to integrate SAT within a global account of linguistic communication whose aim is to provide a psychologically plausible account. A shortcoming of this approach seems to be concerned with an essential element of SAT itself: the emphasis put on the normative dimension produced by the performance of speech acts. The intentionalist analysis fails to provide an account of such a normative dimension, since it is maintained that the creation of commitments and obligations is a “moral” question not answerable by SAT. On the other hand, the conventionalist tradition in both its variants seems to show an opposite and equally unsatisfactory tendency. It has in fact been argued, specifically by relevance theorists (see Sperber and Wilson 1995), that the illocutionary level, as identified by SAT, does not effectively play a role in the process of linguistic comprehension. That it does play such a role, as the SAT claims, is an unjustified hypothesis, so runs the relevance theorists’ objection. More generally, this objection urges speech act theorists to confront the cognitive turn in the philosophy of language and linguistics.
One more surprising applications of Speech Act Theory concerns the debate on free speech, pornography and censorship (compare Langton 1993, Hornsby 1993, Hornsby & Langton 1998, Saul 2006, Bianchi 2008). Liberal defenders of pornography maintain that pornography – even when violent and degrading – should be protected to defend a fundamental principle: the right to freedom of speech or expression. Conversely, Catharine MacKinnon claims that pornography violates women’s right to free speech: more precisely pornography not only causes the subordination and silencing of women, but it also constitutes women’s subordination (a violation of their civil right to equal civil status) and silencing (a violation of their civil right to freedom of speech; MacKinnon 1987). MacKinnon’s thesis has been widely discussed and criticized. Rae Langton and Jennifer Hornsby offer a defence of her claim in terms of Austin’s speech act theory: works of pornography can be understood as illocutionary acts of subordinating women or of silencing women. On the one hand, pornography (or at least violent and degrading pornography where women are portrayed as willing sexual objects) subordinates women by ranking them as inferior and legitimizing discrimination against them. On the other hand, pornography silences women by creating a communicative environment that deprives them of their illocutionary potential. The claim that pornography silences women can be analysed along the lines drawn by Austin. According to Langton (1993), one may prevent the performance of a locutionary act by preventing the utterance of certain expressions, using physical violence, institutional norms, or intimidation (locutionary silence). Or one may prevent the performance of a perlocutionary act by disregarding the illocutionary act, even if felicitously achieved (perlocutionary frustration). Or else one may prevent the performance of an illocutionary act by creating a communicative environment that precludes the uptake of the force of the speech act, or the recognition of the speaker’s authority, particularly with respect to women’s refusals of unwanted sex (illocutionary disablement). Indeed in Langton and Hornsby’s perspective, pornography might prevent women from performing the illocutionary act of refusing sex – by contributing to situations where a woman’s ‘No’ is not recognized as a refusal, and rape occurs as a result. Against the protection of pornography as a form of expression (a simple form of locution), Langton and Hornsby argue that pornography does more than simply express: it acts to silence the expressions of women (a form of illocution), thereby restricting their freedom of speech. We are thus confronted with two different, and conflicting, rights or, better, with different people in conflict over the same right, the right to free speech.
Austin’s legacy has been taken up in epistemology by individual authors such as Charles Travis (2004), who, with regard to the debate about the nature of perception, claims it to be nonrepresentational, and, along the same lines, by Michael G. F. Martin (2002), who defends a form of nonrepresentational realism.
Austin’s work may be considered as a source of inspiration for contextualism in epistemology, but some scholars tend to resist tracing much affinity by pointing out important differences between Austin’s method and the contextualist approach (see, for example, Baz 2012 and McMyler 2011).
More generally, Mark Kaplan (2000, 2008) has insisted on the necessity of considering our ordinary practices of knowledge attribution as a methodological constraint for epistemology, in order for it to preserve its own intellectual integrity. According to Kaplan, the adoption of an Austinian methodology in epistemology would undermine skeptical arguments about knowledge.
Such an Austinian trend is surely a minority one within the epistemological scenery, but later contributions (Gustafsson and Sørli 2011, Baz 2012) have shown how Austin’s method can play a significant role in dealing with several issues in the contemporary philosophical agenda, in epistemology and philosophy of language in particular.
University of Bologna
Vita-Salute San Raffaele University
Last updated: July 14, 2013 | Originally published: July 14, 2013
Article printed from Internet Encyclopedia of Philosophy: http://www.iep.utm.edu/austin/
Copyright © The Internet Encyclopedia of Philosophy. All rights reserved.