The essays collected in this volume, appearing between the years 1939 and 1958, include all of the late Professor J. L. Austin's published papers, and in addition two unpublished papers, ‘The Meaning of a Word’ (1940) and ‘Unfair to Facts’ (1954), as well as an unscripted talk, ‘Performative Utterances’, given in the Third Programme of the B.B.C. in 1956. The editors, J. O. Urmson and G. J. Warnock, have performed a real service in making these papers available in one volume; for the cumulative impact of the method Austin pursued in doing philosophy, and the questions this method raises, can be felt and appreciated more fully when a considerable assemblage of illustrative papers is brought together. The debt Austin owes to the tradition to which G. E. Moore and Wittgenstein contributed is quite clear, and the divergences from them are equally clear. In reviewing the contents of the ten articles in this volume I shall try to single out what is unique about his contribution, in particular what features of his procedure, falling as it does under the general classification ‘linguistic analysis’, were so distinctive as to win for it the attention accorded to a new departure.

The late John Austin's William James Lectures1 might well have borne the subtitle ‘In Pursuit of a Vanishing Distinction’. Although the chase is remorseless, glimpses of the quarry become increasingly equivocal and the hunter is left empty-handed at last. It is hard to know what has gone awry. Has the wrong game been pursued—and in the wrong direction?

The late Professor J. L. Austin was a man of the highest character and considerable intellectual ability. Those who heard him lecture testify that he made a strong and deep impression. He seemed to be completely master of his subject, his mind moved with great rapidity and he manifested a rare integrity of spirit. His death in early middle age is not only mourned by his colleagues; it is one of those events that make a deep gash, causing people to wonder why things are sometimes arranged so perversely in this world. He had so much to give.

This book was reconstructed by G. J. Warnock from notes Professor J. L. Austin prepared for a course of lectures he first gave in Oxford in Trinity Term, 1947, under the title ‘Problems of Philosophy’. The title was changed to ‘Sense and Sensibilia’ the following year. Mr Warnock deserves to be commended for a piece of work which must have been as difficult as its result is excellent. It is a considerable feat of sympathetic identification to have achieved the kind of continuity and order of thought as well as the stylistic continuity found in the book from the sketchy lecture notes a practised lecturer like Austin would need. Mr Warnock assures the reader in his Foreword that ‘…in all points of substance (and in many points of phraseology) his argument was the argument which this book contains’, and there can be no doubt whatever that Austin's thoughts have been recorded meticulously and with admirable clarity.

‘After it, the philosophy of perception cannot be discussed in ways it usually was discussed before.’ This is said about Sense and Sensibilia by Mr Bernard Williams in an article, ‘J. L. Austin's philosophy’, published in the Oxford Magazine of 6 December 1962. It is not quite clear what Mr Williams means by the remark. It might be understood as an endorsement of Austin's insistence that philosophers have lapsed into crudity and error through their neglect of distinctions marked by the rich variety of linguistic expressions in ordinary use. But Mr Williams himself in his article makes some effective criticisms of Austin's opinion that ordinary language is a reservoir of philosophically significant distinctions, and points out also that Austin's own practice in these lectures did not exhibit any close connection between his survey of the variety of the uses of such words as ‘look’, ‘seem’, ‘appear’, and his criticism of theories of perception.

The question of what we are doing when we give an explanation in terms of the agent's reasons for his behaviour is both familiar and complicated. The question I shall raise here, however, is somewhat more limited. It is that of what is involved in making use of such explanations once we have some cursory acquaintance with their structure. In particular, I wish to draw attention to some relations between explanations in terms of intended goals and relations in terms of the agent's reasons for an act.