AUSTIN, JOHN LANGSHAW
(1911–1960)

John Langshaw Austin was White's professor of moral philosophy at Oxford from 1952 until his death in 1960. Educated at Shrewsbury School and Balliol College, Oxford, he became a fellow of All Souls College in 1933; in 1935 he moved to Magdalen College, where he taught with conspicuous success until elected to the White's chair. During World War II he served with distinction in the British Intelligence Corps; he attained the rank of lieutenant-colonel and was awarded the OBE and the Croix de Guerre, as well as being made an officer of the Legion of Merit.

In the years before the war Austin devoted a great deal of his time and energy to philosophical scholarship. He made himself an expert in the philosophy of Gottfried Wilhelm Leibniz and also did much work on Greek philosophy, especially Aristotle's ethical works. At this period his own thought, although notably acute and already distinctive in style, was largely critical and altogether lacked the positive approach that distinguished his postwar work. His one published paper belonging to this early period, "Are There A Priori Concepts?" very fairly represents the astringent style and outlook that gave him the reputation of being a rather terrifying person. According to Austin's own statements, it was not until the beginning of the war that he began to develop the outlook on philosophy and method of philosophizing that marked his mature work, and it is of this work alone that an account will be given.

Aims and Methods

The practical exigencies of lecturing and the traditions of paper reading (especially in symposia, to which some of his important papers were contributions) prevented some of the most characteristic features of Austin's preferred methods and aims from being clearly and fully exemplified in his written work. Lecturing is essentially a solo effort, whereas Austin believed that the best way of doing philosophy was in a group, and papers, especially in symposia, are almost inevitably on topics of traditional philosophical interest, whereas Austin preferred to keep the traditional problems of philosophy in the background. We shall therefore start by giving some account of the method and aims that Austin always advocated and practiced, most notably in meetings held regularly on Saturday mornings in the Oxford term with a group of like-minded Oxford philosophers.

language

Austin did not present his aims and methods as the only proper ones for a philosopher; whatever one or two uncautious remarks in his British Academy lecture "Ifs and Cans" may suggest to the contrary, he did not claim more than that his procedures led to definite results and were a necessary preliminary for anyone who wished to undertake other kinds of philosophical investigation. But he certainly considered them so valuable and interesting in their results, and so suited to his own linguistically trained capabilities and tastes, that he never felt it necessary to investigate for himself what else a philosopher might usefully do. What he conceived of as the central task, the careful elucidation of the forms and concepts of ordinary language (as opposed to the language of philosophers, not to that of poets, scientists, or preachers) was, as Austin himself was well aware, not new but characteristic of countless philosophers from Socrates to G. E. Moore. Nor were the grounds for this activity especially novel. First, he claimed, it was only common prudence for anyone embarking on any kind of philosophical investigation, even one that might eventually involve the creation of a special technical vocabulary, to begin with an examination of the resources of the terminology already at one's disposal; clarification of ordinary language was thus the "begin-all," if not the "end-all," of any philosophical investigation. Second, he thought that the institution of language was in itself of sufficient interest to make it worthy of the closest study. Third, he believed that in general a clear insight into the many subtle distinctions that are enshrined in ordinary language and have survived in a lengthy struggle for existence with competing distinctions could hardly fail to be also an insight into important distinctions to be observed in the world around us—distinctions of an interest unlikely to be shared by any we might think up on our own unaided initiative in our professional armchairs.

It is not too soon to remove at this stage some common misconceptions about Austin's aims and methods. First, although he was not concerned with studying the technical terminology of philosophers, he had no objection in principle to such terms; he thought that many such technical terms had been introduced inappropriately and uncritically, as is clear from his discussion, in Sense and Sensibilia, of the sense-datum terminology, but he used much of the traditional technical vocabulary of philosophy and added many technical terms of his own invention—as almost any page of How to Do Things with Words will bear witness. Second, Austin did not think that ordinary language was sacrosanct; he certainly thought it unlikely that hopelessly muddled uses of languages would survive very long and felt that they were more likely to occur in rather specialized and infrequently used areas of our vocabulary, but there was never any suggestion that language as we found it was incapable of improvement; all he asked was that we be clear about what it is like before we try to improve it.

technique

We have seen that there was nothing essentially novel in Austin's philosophical aims; what was new was the skill, the rigor, and the patience with which he pursued these aims. Here we are dealing with Austin's own personal gifts, which cannot be philosophically dissected. Nor did Austin have any theory of philosophical method; what he had was a systematic way of setting to work, something on a par with a laboratory technique rather than with a scientific methodology. This technique, unlike the skill with which he followed it, was quite public and one that he was willing and eager to employ in joint investigations with others, so we can easily give an account of it.

A philosopher or, preferably, a group of philosophers using this technique begins by choosing an area of discourse in which it is interested, often one germane to some great philosophical issue. The vocabulary of this area of discourse is then collected, first by thinking of and listing all the words belonging to it that one can—not just the most discussed words or those that at first sight seem most important—then by looking up synonyms and synonyms of synonyms in dictionaries, by reading the nonphilosophical literature of the field, and so on. Alongside the activity of collecting the vocabulary one notes expressions within which the vocabulary can legitimately occur and, still more important, expressions including the vocabulary that seem to be a priori plausible but that can nonetheless be recognized as unusable.

The next stage is to make up "stories" in which the legitimate words and phrases occur; in particular, one makes up stories in which it is clear that one can appropriately use one dictionary "synonym" but not another; such stories can also be found ready made in documents. In the light of these data one can then proceed to attempt to give some account of the meaning of the terms and their interrelationships that will explain the data. A particularly crucial point, which is a touchstone of success, is whether one's account of the matter will adequately explain why we cannot say the things that we have noted as "plausible" yet that in fact we would not say. At this stage, but not earlier, it becomes profitable to examine what other philosophers and grammarians have said about the same region of discourse. Throughout (and this is why Austin so much preferred to work in a group) the test to be employed of what can and what cannot be said is a reasonable consensus among the participants that this is so. Such a consensus, Austin found, could be obtained in an open-minded group most of the time; where such agreement cannot be obtained the fact should be noted as of possible significance. Austin regarded this method as empirical and scientific, one that could lead to definitely established results, but he admitted that "like most sciences, it is an art," and that a suitably fertile imagination was all important for success.

It was the lack of thoroughness, of sufficient research before generalization, in previous investigations of language, whether by those who called themselves grammarians or by those who called themselves philosophers, that Austin most deplored. He seriously hoped that a new science might emerge from the kind of investigations he undertook, a new kind of linguistics incorporating workers from both the existing linguistic and the philosophical fields. He pointed to other "new" sciences, such as logic and psychology, both formerly parts of philosophy, as analogues and was indifferent about whether what he was doing "was really philosophy."

So much must suffice as an account of the method of work that Austin advocated. It has been based on a set of notes for an informal talk, characteristically titled "Something about One Way of Possibly Doing One Part of Philosophy." As Austin admitted in those notes, he had said most of this in his papers "A Plea for Excuses" and "Ifs and Cans," and to all who worked with him it was familiar from his practice. Although inevitably, as we have noted, this method could not be followed in writings (it is in any case a method of discovery and not of presentation), its use underlies and can be discerned in his published work. Thus, before writing "Words and Deeds" or How to Do Things with Words he went right through the dictionary making a list, which still survives, of all verbs that might be classed as "performative" in his terminology. The art of telling "your story" is amusingly illustrated over and over again in his paper "Pretending" and, indeed, in all his other published writings. His insistence that it is a mistake to dwell only on a few well-examined notions in a field of discourse is illustrated by his concentration on such notions as "mistake," "accident," and "inadvertence" (in "A Plea for Excuses") and on the use of "I can if I choose" (in "Ifs and Cans"), rather than on "responsibility" and "freedom," in his papers that have a bearing on the free-will problem. Similarly, when his Saturday morning group turned its attention to aesthetics Austin betrayed far more interest in the notions of dainty and dumpy milk jugs than in that of a beautiful picture.

Work

It is not possible to give a systematic account of Austin's "philosophy," for he had none. His technique lent itself rather to a set of quite independent inquiries, the conclusions of none of which could serve as premises for a further inquiry; his discussions of the language of perception (in Sense and Sensibilia ), the concept of pretending, the notion of truth, and the terminology of excuses were all based on the study of speech in those fields and not on any general principles or theories. Nor would it serve any useful purpose to attempt to summarize his various investigations one by one, since they depend so much for their interest and force on the detailed observations about language that they contain. It will be more useful to discuss, first, what he thought of as his main constructive work—the doctrine of illocutionary forces that arose out of his earlier distinction of performative and constative utterances, contained in How to Do Things with Words —and, second, the application of his technique to the criticism of some traditional theories about perception as found in his Sense and Sensibilia.

theory of illocutionary forces

Austin's theory of illocutionary forces arose from his observation that a considerable number of utterances, even those in the indicative mood, were such that in at least some contexts it would be impossible to characterize them as being true or false. Examples are "I name this ship the Saucy Sue" (which is part of the christening of a ship, and not a statement about the christening of a ship), "I promise to meet you at two o'clock" (which is the making of a promise and not the report of a promise or a statement about what will happen), and "I guarantee these eggs to be new-laid" (which is the giving of a guarantee and not a report of a guarantee). These utterances Austin called "performative," to indicate that they are the performance of some act and not the report of its performance; he did not speak as some do who purport to discuss his views, of "performative verbs," for the verb promise can well occur in reports—for example, "I promised to meet him." To provide the necessary contrast, Austin coined the technical term constative to apply to all those utterances that are naturally called true or false; he thought that statement and similar words often used by philosophers roughly as he used constative had in ordinary use too narrow a meaning to serve the purpose.

For a time Austin appears to have been fairly satisfied with this distinction, which he gave in print in his "Other Minds" article in 1946, using it to illuminate some features of utterances beginning "I know.…" But although the distinction is clearly useful at a certain level, Austin began to doubt whether it was ultimately satisfactory. He found it impossible to give satisfactory criteria for distinguishing the performative from other utterances. The first person of the present indicative, which occurs in the three examples given above, is clearly not a necessary feature; "Passengers are warned to cross the tracks only by the bridge" is an act of warning as much as "I warn you to cross.…" Further, in a suitable context "Don't cross the tracks except by the bridge" may also be an act of warning (as in another context it might be an act of commanding); this makes it necessary to distinguish the primative performative from the explicit performative, the latter, but not the former, making clear what act was being performed in its formulation.

Still more important, the constative seemed to collapse into the performative. Let us consider the four utterances "I warn you that a train is coming," "I guess that a train is coming," "I state that a train is coming," and "A train is coming." The first of these is an act of warning, the second is surely one of guessing, the third apparently one of stating, while the fourth may be any of these as determined by context. Thus, the various forms of constatives—stating, reporting, asserting, and the rest—seem to be merely a subgroup of performatives. It might seem that still one crucial difference remains, that while performative utterances may be in various ways unhappy (I may say "I promise to give you my watch" when I have not got a watch, or am speaking to an animal, or have no intention of handing the watch over), the characteristic and distinctive happiness or unhappiness of constatives is truth and falsehood, to which the other performatives are not liable.

In a brilliant, if not always immediately convincing, discussion (Lecture XI of How to Do Things with Words ) Austin tried to break down even this distinction. First, we cannot contrast doing with saying, since (in addition to the trivial point that in stating one is performing the act of uttering words or the like) in constative utterances one is stating, describing, affirming, etc., and these acts are on a par with warning, promising, and so on. Second, all constatives are liable to all those kinds of infelicity that have been taken to be characteristic of performatives. Just as I should not promise to do something if I do not intend to do it, so I should not state that something is the case unless I believe it to be so; just as my act of selling an object is null and void if I do not possess it, so my act of stating that the king of France is bald is null and void if there is no king of France; just as I cannot order you to do something unless I am in a position to do so, so I cannot state what I am not in a position to state (I cannot state, though I can hazard a guess about, what you will do next year). Further, even if we grant that "true" and "false" are assessments specific to constatives, is not their truth and falsity closely parallel to the rightness and wrongness of estimates, the correctness and incorrectness of findings, and so on? Is the rightness of a verdict very different from the truth of a statement? Further, to speak of inferring validly, arguing soundly, or judging fairly, is to make an assessment belonging to the same class as truth and falsehood. Moreover, it is only a legend that "true" and "false" can always be appropriately predicated of constatives; "France is hexagonal" is a rough description of France, not a true or false one, and "Lord Raglan won the battle of Alma" (since Alma was a soldiers' battle in which Lord Raglan's orders were not properly transmitted) is exaggerated—it is pointless to ask whether it is true or false. It was on the basis of such considerations as these that Austin felt himself obliged to abandon the distinction between the performative and the constative.

To replace the unsatisfactory distinction of performatives and constatives Austin introduced the theory of illocutionary forces. Whenever someone says anything he performs a number of distinguishable acts, for example, the phonetic act of making certain noises and the phatic act of uttering words in conformity with grammar. Austin went on to distinguish three other kinds of acts that we may perform when we say something: First, the locutionary act of using an utterance with a more or less definite sense and reference, for example, saying "The door is open" as an English sentence with reference to a particular door; second, the illocutionary act, which is the act I may perform in performing the locutionary act; third, the perlocutionary act, which is the act I may succeed in performing by means of my illocutionary act. Thus, in performing the locutionary act of saying that a door is open I may be performing an illocutionary act of stating, or hinting, or exclaiming; by performing the illocutionary act of hinting I may succeed in performing the perlocutionary act of getting you to shut it. In the same way, by performing the locutionary act of saying "Down with the monarchy" I may succeed in the perlocutionary act of bringing about a revolution, whereas in performing the locutionary act I would be inciting to revolution (successfully or unsuccessfully).

We now see that the constatives, along with performatives, can be construed as members of one particular subclass of illocutionary forces. Thus, in his provisional classification of illocutionary forces Austin had a subclass of expositives, which included the "constative" acts. In performing a locutionary act we may be affirming, denying, stating, describing, reporting, agreeing, testifying, rejoining, etc., but in performing a locutionary act we may also perform an act with commissive force, as when we promise, bet, vow, adopt, or consent; with verdictive force, as when we acquit, assess, or diagnose; with exercitive force, as when we appoint, demote, sentence, or veto; or with behabitive force, as when we apologize, thank, or curse.

Such is the crude outline of Austin's theory of illocutionary forces. Though his own exposition is of course much more full and rewarding, he said of it (How to Do Things with Words, p. 163): "I have purposely not embroiled the general theory with philosophical problems (some of which are complex enough almost to merit their celebrity); this should not be taken to mean that I am unaware of them." We may be permitted to illustrate the philosophical importance of bearing in mind the distinctions Austin made with one example of our own. Very often in recent years philosophers have set out to explain the meaning of the word good or of sentences containing the word good. Some of them have done so by saying that in such sentences the speaker expresses his own feelings (attitudes) and evokes similar feelings (attitudes in others). It might well seem that here they have set out to give an account relevant to locutionary force and that they have instead given one possible illocutionary force ("In saying that it was good I was expressing my favorable attitude toward it") and, alongside it, one possible perlocutionary force ("By saying that it was good I evoked in him a favorable attitude"). It should be clear in the light of Austin's work that such an account will not do. But Austin said very little about locutionary force in detail, and one of the most pressing general questions that arise from his work is that of the relationship between illocutionary force and locutionary force; while recognizing that they are different, and that locutionary force is in some way prior, can we, for example, conclude that the locutionary force of utterances containing the word promise can be explained without reference to the typical illocutionary force of "I promise"? This is far from clear.

criticism of traditional philosophy

We have examined in outline an example of Austin's work on a piece of clarification of language without any reference, save incidental, to the traditional problems of philosophy. We shall now turn to Sense and Sensibilia, which is emphatically a polemical discussion of one of the central problems of epistemology. But we shall find the essential features of Austin's method still present, the presentation only being different. Austin had recommended that when the method is used as one of inquiry the vocabulary and phrases, natural and odd, that occur to us should be studied and conclusions drawn before the conclusions of traditional philosophy are compared with them. Here, however, when he presents results he at each stage presents first the traditional philosophical theses and then shows their errors by confronting them with the actual facts, linguistic and otherwise.

In Sense and Sensibilia, Austin examines the doctrine that we never directly perceive material things but only sense data (or ideas, or sense contents, etc.), insofar as that doctrine is based upon the so-called argument from illusion. He maintains that it is largely based on an obsession with a few words "the uses of which are oversimplified, not really understood or carefully studied or correctly described" (Sense and Sensibilia, p. 3). With special reference to A. J. Ayer and Price, he shows how illusions are traditionally confused with delusions, are defined in terms of belief that one sees a material thing when in fact one does not (whereas some illusions, such as one hatched line appearing to be longer than another of equal length, involve nothing of the sort), and are taken to include such phenomena as sticks looking bent in water, which are not illusions at all. A portion of the argument that clearly exhibits his method at work is where he contrasts the actual complexities and differences in our use of "looks," "appears," and "seems" with the traditional confusion of these terms in traditional philosophy. Especially interesting is the discussion of the traditional accounts of "reality"; these he contrasts with the multifarious uses of the word real, which takes its significance only from the implied contrast in context with artificial, fake, bogus, toy, synthetic, and so on, as well as with illusory and apparent.

But it is perhaps more important now for us to notice another element in the argument that is very characteristic but that we have as yet given little notice, which is Austin's care to avoid oversimplification and hasty generalization of nonlinguistic, as well as linguistic, fact. The ordinary man does not, as is so often stated or implied in accounts of the argument from illusion, believe that he always sees material things; he knows perfectly well that he sees shadows, mirror images, rainbows, and the like. The number of kinds of things that we see is large and to be settled by scientific investigation, not by philosophy; the question whether the invariable object of perception is a material thing or a sense datum is thus absurd. Again, it is not true that a straight stick in water normally looks like a bent stick out of water, for we can see the water; an afterimage does not look like a colored patch on a wall; a dream is distinguished by the dreamlike quality that occasionally, but only occasionally, we attribute to some waking experience. Again, he points out that situations in which our perception is queer may arise because of defects in sense organs or peculiarities of the medium or because we put a wrong construction on what we (quite normally) see, and it is a mistake to attempt to give a single account of all perceptual error. None of these are linguistic points, and Austin had no purist, theoretical notion that he was prohibited as a philosopher from any attention to nonconceptual issues; he thought that philosophical error did arise from empirical error.

Once again, it would be pointless to attempt to reconstruct the whole argument of Sense and Sensibilia here; we must be content with noticing the few points made that perhaps have some bearing on a general understanding of his general position. But it should perhaps be stressed that Austin in these lectures discussed only one theory of perception as based on one particular kind of argument; although one may expect to get help from it in study of other problems in the field of perception, it would be a mistake to suppose that the book contains a full study of all problems of perception or to criticize it because it leaves many difficult problems unanswered.

It is hardly imaginable that anyone would ever deny that Austin displayed a very great talent in the kind of work he chose to do. Some have criticized him on the ground that there are more important things for philosophers to do than this; on that point Austin always refused to argue, simply saying that those who preferred to work otherwise should do so and asking only that they not do what he did in the traditional slipshod way. To those who said that philosophers should work with an improved scientific language he replied flatly that the distinctions of ordinary language were of interest in their own right and that one should not modify what one does not fully understand, but he offered no theoretical objections to such projects. He was content to work in a way which he felt he understood and found rewarding. As for the assertion sometimes made, that Austin's kind of work is private to his own peculiar gifts and that it was therefore a mistake for him to recommend the method to others, time alone can decide.

A final word should be said about Austin's relation to other philosophers. He greatly admired G. E. Moore, but it is a mistake to view his work as an offshoot of Cambridge philosophy. Moore, like Austin and unlike most Cambridge philosophers, had a linguistic and classical background rather than a scientific one. Austin owed no special debt to Bertrand Russell and was far more unlike Wittgenstein than is sometimes recognized. For Ludwig Wittgenstein an understanding of ordinary language was important because he believed that the traditional problems of philosophy arose from misunderstandings of it, but Wittgenstein had in mind gross category mistakes, and he wished to study ordinary language only so far as was essential for eliminating these. Austin was interested in fine distinctions for their own sake and saw the application of his results to the traditional problems of philosophy as only a by-product. He was uninterested in the party conflicts of philosophy, following always his individual bent.

See also Aristotle; Ayer, Alfred Jules; Language; Leibniz, Gottfried Wilhelm; Moore, George Edward.

Bibliography

primary works

For brevity, first publication of individual papers (collected in Philosophical Papers ) is omitted.

Foundations of Arithmetic. Oxford: Blackwell, 1950. Translation of Frege's Grundlagen der Arithmetik.

Critical notice of J. Łukasiewicz's Aristotle's Syllogistic: From the Standpoint of Modern Formal Logic Mind 61 (1952).

"Performatif-constatif" and contributions to discussion in La philosophie analytique. Paris: Minuit, 1962.

How to Do Things with Words, 2nd ed. Edited by J. O. Urmson and Marina Sbisà. Cambridge: Harvard University Press, 1975.

Sense and Sensibilia. Edited by Geoffrey J. Warnock. London: Oxford University Press, 1975.

Philosophical Papers, 3rd ed. Edited by J. O. Urmson and Geoffrey J. Warnock. Oxford: Oxford University Press, 1979.

secondary works

Berlin, Isaiah, et al. Essays on J. L. Austin. Oxford: Clarendon Press, 1973.

Cavell, Stanley. Philosophical Passages: Wittgenstein, Emerson, Austin, Derrida. Bucknell Lectures in Literary Theory, 12. Oxford: Blackwell, 1995.

Di Giovanna, Joseph J. Linguistic Phenomenology: Philosophical Method in J. L. Austin. American University Studies. Series V, Philosophy, Vol. 63. New York: Lang, 1989.

Fann, K. T., ed. Symposium on J. L. Austin. New York: Humanities Press, 1969.

Furberg, Mats. Saying and Meaning: A Main Theme in J. L. Austin's Philosophy. Totowa, NJ: Rowman and Littlefield, 1971.

Graham, Keith. J. L. Austin. Atlantic Highlands, NJ: Humanities Press, 1977.

Holdcraft, David. Words and Deeds. Oxford: Oxford University Press, 1978.

Rorty, Richard, ed. The Linguistic Turn: Recent Essays in Philosophical Method. Chicago: University of Chicago Press, 1967.

Searle, John. Expression and Meaning: Essays in the Theory of Speech Acts. Cambridge, U.K.: Cambridge University Press, 1979.

Searle, John. Speech Acts: An Essay in the Philosophy of Language. Cambridge, U.K.: Cambridge University Press, 1969.

Warnock, Geoffrey J. J. L. Austin. London: Routledge, 1989.

J. O. Urmson (1967)

Bibliography updated by Philip Reed (2005)