J.L. Austin was born in Lancaster on 26 March 1911 and died in Oxford on 8 February 1960. From a middle-class family, he was educated at Shrewsbury School and gained a scholarship to read classics at Balliol College, Oxford. In 1933 he secured a fellowship at All Souls College, a research position; in 1935 he moved to a fellowship at Magdalen College, a teaching position. Austin from early on admired Bertrand Russell for his genius and clarity, but of his direct philosophical acquaintances it was H.A. Prichard who received his greatest respect – for his meticulous rigour. With schol arship and linguistic accuracy to the fore, Austin initially lectured on Aristotle and other historical figures, but his academic life was interrupted by World War II. He was commissioned in the British Intelligence Corps, where his intelligent eye led Lieutenant Colonel Austin, as he became, to be awarded the OBE, the French Croix de Guerre and the American Officer of the Legion of Merit. Returning to Oxford, he gained the White’s Chair in Moral Philosophy in 1952 and a fellowship at Corpus Christi, positions he held until his premature death. During the 1950s he also lectured in the United States – at Harvard and California. He was a private and upright man of routine, though he was not without humour – not without, indeed, silly jokes – and not without a wife (Jean Austin, another philosopher) and four children. Although very much an academic, he regretted having been neither engineer nor architect; and the work that made him famous, ordinary language philosophy, displays a practical interest in linguistic nuts and bolts.
Austin’s lifetime’s published work was scant and before the war promised little of what was to come, though his early ‘Are There a priori Concepts?’ contains sceptical questioning of philosophers’ meanings, a questioning already orally prominent in his encounter with A.J. Ayer’s logical positivism which set the then philosophical agenda. Returning to Oxford after the war, Austin edited H.W.B. Joseph’s Leibniz lectures and translated Frege’s Grundlagen der Arithmetik, but his own distinctive work, with its general empiricist background, lay with ordinary language distinctions, both as pointers to philosophical solutions and as subject-matter for a science of language in use. By his death, although he had merely seven papers published, he was the dominant Oxford philosopher – somewhat to the chagrin of Gilbert Ryle , the more senior figure, and to Ayer, the more attention-seeking. The tall, bespectacled Austin – ‘an inscrutable crane’ – spoke with an assumed authority that no other at Oxford matched. This ascendancy was manifest in his Saturday Mornings, classes at which he influenced, for example, H.P. Grice , P.F. Strawson and G.J. Warnock . Posthumously his influence spread through his lecture notes, published as Sense and Sensibilia and How to do Things with Words.
Plain clear prose marks Austin’s writing, a prose displaying ingenuity, wit – witness his lectures’ title, ‘Sense and Sensibilia’ – and a concern for accuracy, though not always accuracy itself. He was the heart and head of the 1950s Oxford School of Ordinary Language Philosophy (a label applied by others to his work) and many continue to admire him – but no programme ever bound his admirers, save careful attention to everyday language and specific cases. This attention was no novelty, being found in Plato’s early dialogues (though Austin sought no Socratic definitions), Aristotle and G.E. Moore , the latter two’s plain approach receiving his praise. Austin attends to details, idioms (though he criticizes some colloquialisms as loose) and striking examples as few others; he hounds down the minutiae in order to incite. His results, though, are often coloured by etymological conviction and distinctions he, a classicist, makes; yet his technique manifests skill, patience and an escape from hallowed examples and abstractions. When investigating responsibility, he discusses mistakes, inadvertences and accidents; when considering aesthetics, we encounter dainty and dumpy milk jugs; and when discussing philosophers’ ‘material objects’, he goes beyond traditional ‘moderate sized specimens of dry goods’ to wondering about rainbows, shadows and flames. Thus he served subsequent philosophers a rich smorgasbord upon which to dine, and dine some did, be it through the speech act theory of his pupil, John Searle, or the legal philosophy of his colleague, H.L.A. Hart . Although proving a valuable stimulus to discernment, Austin’s work led some to trivial logic-chopping with doubtful philosophical significance – a criticism sometimes levelled at Austin himself.
Austin wrote little about his philosophical method – the main source is ‘A Plea for Excuses’ – and one element of that little should be disregarded, when he speaks mysteriously of prising words off the world to view it unblinkered. Austin’s belief is that every distinction appearing in ordinary language (that is, the non-philosophical), having survived the test of time, has some justification and is likely therefore to be philosophically helpful. This leads Austin neither to think that no new distinctions could profitably be drawn nor to eschew technical terminology; but old habits of Gleichschaltung – ‘the deeply ingrained worship of tidy-looking dichotomies’ – need abandoning and Austin delights in playing Old Harry with dichotomies such as fact/value, appearance/reality, truth/falsity. His piecemeal linguistic investigations and recommendation to ‘work the dictionary’ fit well with his promotion of philosophy as a cooperative endeavour, with participants forming a disciplined investigating team, working on common speech’s nuances, exposing errors of simplification and half-studied facts. When rules were discussed, participants were to study different rule books; discussions of aesthetic judgements were once based on analyses of industrial design manuals. Despite the informality of the Saturday Mornings, the teamwork smacked of the military or civil service: Austin preferred all to be seated round a table, reaching agreement on some subtle linguistic distinction.
The ordinary language investigation is intended as a prelude to – an under-labouring for – solving particular philosophical problems, particularly in epistemology and action theory. It is neither the be-all nor the end-all, but, as Austin quips, the begin-all. It runs the danger of a paralysing effect – in never risking hunches and hypotheses until all niceties are examined. Further, the relationship between linguistic evidence and philosophical tasks remains unexplained; and Austin himself, as he acknowledges, makes little philosophical progress beyond the preparation of the evidence. His lecture series, How to do Things with Words, a series polished over the years, ends with the comment that the fun of applying what has been said to philosophical problems is yet to come. Despite lack of immediate philosophical pay-off, Austin certainly held that philosophical progress could be made; he was traditional in understanding philosophy to be in the same dimension as the sciences. Not for Austin Wittgenstein’s view of philosophical problems as grounded in linguistic bewitchments. Not for Austin the rejection, in philosophy, of all theories. For Austin, there was no fear of drowning within hidden depths of perplexities. Indeed, he held Wittgenstein somewhat in disdain and, in his published works, Wittgenstein is mentioned only once – though, in discussion, parts of his Philosophical Investigations received minute examination. Unlike Wittgenstein, Austin exhibited none of the anguished soul-searching and almost religious bafflement with philosophy. Unlike Wittgenstein, Austin writes in straightforward continuous prose.
Austin’s name is probably most associated with speech acts in general and performatives in particular. Although others, such as Wittgenstein, certainly drew attention to the importance of different uses of language, including performing uses, such as ‘I mourn …’, it is Austin who aspires to a theory and who led others to develop speech act theory within linguistics, literary theory and philosophy. The paradigm that Austin first focuses upon is one that troubled Prichard: how, in saying ‘I promise’, am I doing the promising and not saying that I am doing it? Austin’s questioning of what the doing is when saying something had its first published outing in his 1946 ‘Other Minds’, the detailed systematic treatment appearing later in How to do Things with Words, where his delight in assembling Linnaean-type classifications is on full display. Without the taxing taxonomy yet to come, ‘Other Minds’ suggests that as saying ‘I promise’ can perform an act of promising, so saying ‘I know’ can also perform: I give others my word, my authority for saying whatever is in question. Just as with promising, one needs the right setting for the performance to come off, and people are then entitled to act on what has been said, so too with knowing: one needs the right circumstances to be justified in making the performance, and people are then entitled to act on that performance. This assimilation of knowing and promising to a performing conceals significant differences. You might promise, yet not do what you promise, but you cannot know, yet what is known not be true. Another can report that you promised to do something but are unlikely to pull it off; someone cannot report that you both knew but that what you knew was false. Austin, underplaying these differences, finds philosophers guilty of a descriptive fallacy in ignoring the performative heart of ‘I know’.
Thus it was that ‘performative’ became the watchword for many philosophers when any descriptive analyses struck philosophical brick walls. In view of this approach, Austin was expected to have stressed a performative understanding of ‘is true’; and others, such as Strawson, doubting the value of ‘correspondence’ and ‘facts’, promoted such an understanding. Curiously, although Austin recognized ‘true’s performative features, he remained wedded to a correspondence theory, wedding facts to true statements. Truth bearers must be statements, yet, using the statement that France is hexagonal, he reminds us of rough descriptions and of the more or less true. He looks to common uses of ‘true’, adding the less than common thought that in vino, possibly, ‘veritas’, but in sober symposium ‘verum’.
Austin’s claim, that ‘I know’ is a performative parading as descriptive, accords with the backdrop that indicative sentences are sometimes nonsense parading as sense or emotive whoops as factual assertions: Austin has in mind logical positivist views of metaphysical and ethical assertions. In How to do Things with Words Austin extends his search for such passing-off; he investigates common or garden indicative sentences which grammatically look like statements, yet which lack truth value, stating nothing at all. In using such sentences one is doing something more than just saying. It is here that Austin explicitly and formally christens such uses ‘performatives’, thus performing his own performative. He uses humdrum verbs in the first person singular present indicative active — ‘christen’, ‘advise’, ‘salute’ – to display a doing that is not solely a saying – bringing out similarities with typical imperatives. Saying, for example, ‘I christen this ship “Poppy”’ would seem to be neither true nor false, though it could be happy or unhappy. Appropriate circumstances are required, otherwise misfirings – infelicities other than falsehoods – occur whereby nothing, or only something inappropriate, gets done. Appropriate intentions and beliefs are required, otherwise abuses occur: a promise is made all right, but ought not to have been. In some striking cases, as with ‘I promise …’, the utterance explicitly reveals what is being performed.
‘I bet …’ can show a betting. The possibility of a ‘hereby’ insertion is a test for spotting such explicit performatives. As Austin progresses through these lectures, performatives pop up everywhere: even in stating that the minister lied, one is doing something – a stating. Indeed, the speaker might more formally have used the form ‘I state that the minister lied’ or even ‘I hereby state that the minister lied’. Constatives (statings) – and here they resemble performatives – can have infelicities which do not amount to logical inconsistencies: stating that p when I lack belief that p is no logical contradiction, yet possesses an unhappiness. Indeed, performatives – and here they resemble constatives – can generate something remarkably like logical contradictions: ‘I promise, but I am not bound to perform’. Austin, despite his scepticism of big dichotomies, remains determined to install some classifications; and so – curiously, halfway through the lecture series (even after many revisions) – he announces that he will start again.
The fresh start leads to a terminological battery. Taking the case of someone speaking literally, the speaker makes noises, that is performs the phonetic act; he speaks in a language (not much discussed by Austin) with a vocabulary and grammar; hence, the noises are words making up a sentence – this he labels a ‘phatic act’. A phatic act might, of course, go wrong: it might be ungrammatical; it offers scope for ambiguity. These issues sorted out, the speaker thus says something – and this saying is the rhetic act, an act that is identified in terms of what he means, covering both reference and sense. In performing this trinity (phonetic, phatic and rhetic acts) – and now Austin switches nomenclature from Greek to Latin based – we perform a locutionary act; but there remains the question of how we are using the locution. My saying what I say typically has effects on the audience, speaker and others, effects not necessarily uniform or conventional. I might frighten one listener, by saying ‘There’s a policeman’, yet reassure another. Producing such effects (intended or not), one has performed a perlocutionary act. Between the locutionary and perlocutionary lies the illocutionary. Performing a locutionary act, I might, for example, warn, advise or christen – and these are examples of illocutionary acts. To warn, advise and christen are not my doings’ effects that depend on how the audience takes them, but are what I am doing. I might say as such, whatever the results on the audience. I might have meant my remark as a warning, and a warning it be, even if listeners fail to take it this way. Its being a warning is grounded in convention and circumstances. This is sometimes explicitly shown, as when the words ‘I warn you …’ are used; and, in using those words, intrinsically I warn you. That there are these different types of function – warning, undertaking, ordering – shows the existence of different illocutionary forces, yet with the possibility of common content. Although Austin’s own attempts to establish simple criteria for drawing these distinctions fail, his outlined distinctions have stimulated much research and theory concerning how utterances give rise to locutionary acts with various illocutionary forces. Austin certainly demonstrated how understanding language use extends well beyond understanding the meaning of words and grammatical rules.
Austin drew closest to handling directly some traditional philosophical problems in his published papers such as ‘A Plea for Excuses’ and ‘Ifs and Cans’. These – at the most general level – raise questions of what it is to do something, an understandable Austinian interest, given his fascination with linguistic performings. Philosophers have been blind to the complexity of doings, according to Austin. To open our eyes, he examines how things can go wrong in doings; how we can do things unwittingly, unintentionally, recklessly, negligently, mistakenly, inadvertently, carelessly, accidentally – and he explains in detail how these differ from each other. In his paper ‘Pretending’ there is a similar concern with how doings might not be quite as they seem. Doing something ‘freely’ might suggest that there must be some feature present to justify the ‘freely’ accolade, but this puts the dialectical boot on the wrong foot. We need some particular reason for suggesting that a person failed to be acting freely: perhaps he acted under duress or by mistake. If there are no specific reasons for thinking that the person acted under duress or by mistake or in some other way suggestive of failing freely to act, it does not follow that the person was acting freely. The motto is ‘no modification without aberration’: in everyday cases, if there are no special reasons to raise questions, then no modifying expression is properly deployable – be the modification of the aggravating or excusing ilk. To yawn is just that – to yawn, neither voluntarily nor involuntarily.
Philosophers traditionally connected questions of responsibility with whether people could have acted otherwise. Moore had tentatively proposed, with variations, that to say that I could have acted otherwise amounts to saying that I could have acted otherwise, if I had chosen; such a proposal secures the compatibility of determinism and being able to act otherwise. Austin digs away at Moore’s analyses, at the ‘could’s, ‘should’s and ‘would’s, at the ‘can’s and ‘if’s, asking whether an ‘if’ is in the offing when we speak of what we could have done. Austin argues that, while some ‘if’s can be introduced, showing a condition on my being able to do something – I could have done that, had I been fully fit – often we are saying what we are able to do simpliciter. Any introduced ‘if I had chosen’ sets no more a condition on my ability than ‘if you want them’ sets a condition on there being biscuits on the sideboard when I tell you that there are – ‘if you want them’.
Moore makes other proposals, namely that the correct analysis of a ‘can’ or ‘could’ sentence, in the relevant contexts, is that one will do something, if one chooses – or if one wants or tries – or would have done something, if one had chosen, wanted or tried. Austin dismisses these suggestions pretty quickly. After all, having missed a very short putt, it might yet be true that I could have holed the ball easily; yet that does not amount to the truth that I should have holed it, if I tried – for I did try.
Austin’s writings in this area – as ever – stimulated others to further work, to careful analyses of the ‘can’ and ‘could’ and related idioms, but that work, while in Austinian spirit, has been critical of his accuracy and has not made the significant progress with the philosophical problem that Austin anticipated. Austin himself doubted compatibilist manoeuvres such as those of Moore, believing the problem to be with what is meant by ‘determinism’, claiming never to have met a determinist even though having met people who insisted that they were determinists.
Much of Austin’s work, although itself preparatory, is often suggestive of new approaches; but his Sense and Sensibilia lectures, while displaying his fondness for novel examples, are largely negative, arguing against Ayer’s The Foundations of Empirical Knowledge, challenging Ayer’s use of the argument from illusion to show that we never directly perceive material objects, but only sense-data (coloured patches seen, hardnesses touched). Austin has fun, checking out our use of ‘deception’, ‘delusion’ and ‘illusion’ and when it is appropriate to use ‘real’, ‘directly’ and ‘certain’. He has no difficulty in showing that the frequent oddness of the claim that we do directly perceive material objects matches the oddness of the claim that we do not. We might well, though, rightly assert on occasions that we can, for example, directly see the speakers or directly hear their voices. In line with his ‘no modification without aberration’, Austin points out that when we assert that something is a cigarette, we should need specific reasons to generate legitimate questions such as whether the cigarette is real, is directly seen or is certainly a cigarette. Unless special doubt-raising reasons are given, the ordinary language user would rightly treat the sceptic’s implied doubt as nonsense. ‘If that’s not a real cigarette, I don’t know what is.’ Although Austin is right about ordinary expectations here, he appears about ordinary expectations here, he appears insensitive to the distinction between the ‘nonsense’ ascribed because of the pointlessness of a comment and that ascribed because of some meaninglessness. He also offers no justification why the only good reasons for doubt must be specific. The sceptic does offer reasons and the only reason they are not specific is because, according to the sceptic, they apply in most cases and sometimes in all cases of a certain type.
Austin’s appeals to ordinary language specifically miss their target where Ayer is concerned, for Ayer is highlighting the implications of statements such as ‘I hear a coach’ and, to achieve greater perspicuity, introduces the term ‘sense datum’ and revised understandings of the ordinary terms ‘directly’ and ‘certain’. Where Ayer remains vulnerable to Austin is in his justification for preferring sense-datum language over ordinary language. Whatever Ayer might say about choosing between languages, he writes as if we really only ever perceive sense-data. For Austin ‘real’ and ‘directly’ – as with ‘freely’ – secure whatever sense they have from contrasts with their opposites; their opposites wear the trousers. The issue of whether you are directly seeing the ship might hang on whether you are viewing it through a periscope; whether the duck in front of you is real might depend on whether some toy ducks are mingled with the flesh and blood. Even here it is arguable whether the hotchpotch of examples concerning ‘real’ – and Austin claims that no general criteria exist for distinguishing the real from the unreal – have much to do with the epistemological concern of what is going on when we distinguish what seems to be so from what really is so. Where Austin significantly challenges Ayer is over the incorrigible. Ayer uses his sense-datum language in the hope of finding some incorrigible statements, statements about which we cannot be mistaken. Austin questions whether first-person reports of sensations or sense-data are incorrigible. If we can be certain of anything, it is equally that, under the right conditions, there is a pig slap-bang in front of us. In such a case, we make no inference from evidence to justify our judgement. We just know things are thus and so. This returns us to Austin’s ‘Other Minds’, where he reminds us that while we might be asked how we know, we should not be asked why we know – contrasting with what can be asked of beliefs. Austin’s approach, as well as aiding the movement against sense-data, is in the spirit of J. Cook Wilson and Prichard, both of whom argued that knowledge is a kind of mental state distinct from belief, a view that reappeared at the end of the twentieth century in Timothy Williamson’s work.
Austin stretched philosophers’ eyes. The Austinian landscape has donkeys being shot – are they being shot by mistake or by accident? – and people with false teeth, yet artificial limbs. There are at least three different ways of spilling ink – and we meet suitors for young ladies’ hands being asked whether their intentions, not their purposes, are honourable. There are goldfinches – we are sure that they are real – but they might yet explode quoting Mrs Woolf; and we encounter thieves pretending to be cleaning the windows, while eyeing the jewellery, yet still cleaning those windows a treat. Austin’s writings are pregnant with vivid examples; yet his insistence on the dominance of ordinary language runs the risk of sterility. When new things need to be said – by a Freud explaining behaviour or a Schrödinger describing subatomic structures or even an Austin challenging philosophical platitudes – living initially with apparent conceptual confusion might be a price worth paying.
Ayer complained that Austin was like a greyhound who, fancying no running himself, bites the other greyhounds so they too give up the race. Austin’s promise has been seen by some as nothing but linguistic performance with no philosophical promise and no importance at all – though Austin himself mused that what is important is not importance but truth. And it is the truth that Austin sought.