Paul Grice: Philosopher and Linguist

Paul Grice, Philosopher and Linguist Siobhan Chapman

1403_902976_01_pre.qxd 12/13/04 6:23 PM Page i

Paul Grice, Philosopher and Linguist

1403_902976_01_pre.qxd 12/13/04 6:23 PM Page ii

Other books by Siobhan Chapman ACCENT IN CONTEXT PHILOSOPHY FOR LINGUISTS

1403_902976_01_pre.qxd 12/13/04 6:23 PM Page iii

Paul Grice, Philosopher and Linguist Siobhan Chapman

1403_902976_01_pre.qxd 12/13/04 6:23 PM Page iv

© Siobhan Chapman 2005 All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No paragraph of this publication may be reproduced, copied or transmitted save with written permission or in accordance with the provisions of the Copyright, Designs and Patents Act 1988, or under the terms of any licence permitting limited copying issued by the Copyright Licensing Agency, 90 Tottenham Court Road, London W1T 4LP. Any person who does any unauthorised act in relation to this publication may be liable to criminal prosecution and civil claims for damages. The author has asserted her right to be identified as the author of this work in accordance with the Copyright, Designs and Patents Act 1988. First published 2005 by PALGRAVE MACMILLAN Houndmills, Basingstoke, Hampshire RG21 6XS and 175 Fifth Avenue, New York, N.Y. 10010 Companies and representatives throughout the world PALGRAVE MACMILLAN is the global academic imprint of the Palgrave Macmillan division of St. Martin’s Press, LLC and of Palgrave Macmillan Ltd. Macmillan® is a registered trademark in the United States, United Kingdom and other countries. Palgrave is a registered trademark in the European Union and other countries. ISBN 1–4039–0297–6 hardback This book is printed on paper suitable for recycling and made from fully managed and sustained forest sources. A catalogue record for this book is available from the British Library. Library of Congress Cataloging-in-Publication Data Chapman, Siobhan, 1968– Paul Grice, philosopher and linguist / Siobhan Chapman. p. cm. Includes bibliographical references (p.) and index. ISBN 1–4039–0297–6 1. Grice, H. P. (H. Paul) I. Title. B1641.G484C48 2005 192–dc22 2004051259 10 14

9 8 7 13 12 11

6 10

5 09

4 08

3 07

2 06

1 05

Printed and bound in Great Britain by Antony Rowe Ltd, Chippenham and Eastbourne

1403_902976_01_pre.qxd 12/13/04 6:23 PM Page v

Contents Acknowledgements

vi

Preface

vii

1 The Skilful Heretic

1

2 Philosophical Influences

10

3 Post-War Oxford

31

4 Meaning

61

5 Logic and Conversation

85

6 American Formalism

114

7 Philosophical Psychology

138

8 Metaphysics and Value

157

9 Gricean Pragmatics

185

Notes

217

Bibliography

232

Index

243

v

1403_902976_01_pre.qxd 12/13/04 6:23 PM Page vi

Acknowledgements A number of people have made this book possible by giving generously of their time to talk about Paul Grice, his life and his work, whether by letter, by e-mail or in person. I would therefore like to thank the following, in an order that is simply alphabetical, and in the hope that I have not omitted anyone who should be thanked: Professor Judith Baker; the staff at the Bancroft Library, Berkeley, in particular David Farrell; Tom Gover of the Old Cliftonian Society; Kathleen Grice; Adam Hodgkin; Professor Robin Lakoff; The Rossallian Club; Michael Stansfield, college archivist at Corpus Christi and Merton colleges; Professor Sir Peter Strawson; Professor Richard Warner. I am especially grateful to Kathleen Grice for permission to use the photograph on the front cover and to quote from manuscript material; and to the Bancroft Library, Berkeley, which gave permission to quote from their manuscript holdings.

For my parents, Raymond and Patricia

vi

1403_902976_01_pre.qxd 12/13/04 6:23 PM Page vii

Preface

There is an inherent tension in writing about Paul Grice because two separate academic disciplines lay claim to his work, and do so with justification. Those who know him as a philosopher may not be aware of the full significance of his role in the development of present day linguistics, while those whose interest in him originates from within linguistics may be unfamiliar with the philosophical questions that are integral to his work. Grice would not have considered himself to be anything other than a philosopher, so the title of this book might seem odd or even provocative. But I chose it in order to reflect the profound influence he has had on a discipline other than his own. Grice’s work is of interest to philosophers and to linguists alike. I myself belong to the second group. My first encounter with Grice’s work was when I was introduced to his theory of conversation as a student, and was struck by the ambition and elegance of his proposed solution to a range of linguistic problems, as well as by the questions and difficulties it raised. I have since been pleased to recognise just these reactions to the theory in many of my own students. As I read more of Grice’s work, I became intrigued by the question as to whether there was a unity to be found in his thought that would incorporate the familiar work on conversation into a larger and more significant philosophical picture. This book is my attempt to answer that question. I certainly hope that it will be of interest to readers from both philosophical and linguistic backgrounds, with the following words of caution to the former. Grice’s work draws on a range of previous writings that often go unacknowledged, presumably because he assumed that his audience would be aware of its philosophical pedigree. The questions he is addressing can be less familiar to linguists, so on a number of occasions I have offered an overview of the history of an idea or an exegesis of a work. I run the risk that these sections may appear to be superficial, or alternatively to be labouring the obvious, to readers well versed in philosophy, but I have offered the degree of detail necessary to enable those unfamiliar with the relevant background to follow Grice’s arguments and appreciate the nature of his contribution. Further, my intention in vii

1403_902976_01_pre.qxd 12/13/04 6:23 PM Page viii

viii Preface

these sections is to sketch the relevant history of a philosophical debate, rather that to offer either a critique of or my own contribution to that debate. I can only ask for tolerance of these aspects of my book from readers already familiar with the relevant areas of philosophy.

1403_902976_02_cha01.qxd 12/13/04 6:24 PM Page 1

1 The Skilful Heretic

In 1986 Oxford University Press published a volume of essays drawing on the work of the philosopher Paul Grice, who was then 73. It was not formally described as a Festschrift, but Grice’s name was concealed as an acronym of the title, Philosophical Grounds of Rationality: Intentions, Categories, Ends, and many of those who contributed to the volume took the opportunity of paying tribute to his work and influence. Among these, Gordon Baker revealed that what he admired most was Grice’s ‘skilful advocacy of heresies’.1 In a similar vein, Grice’s colleague Richard Grandy once introduced him to an audience with the comment that he could always be relied on to rally to ‘the defence of the underdogma’.2 Given Grice’s conventional academic career together with his current status in philosophy and, particularly, in linguistics, these accolades may seem surprising. His entire working life was spent in the prestigious universities of Oxford and Berkeley, making him very much an establishment figure. Much of his philosophy of language, particularly his theory of conversational implicature, has for a number of decades played a central role in debates about the relationship between semantics and pragmatics, or meaning as a formal linguistic property and meaning as a process taking place in contexts and involving speakers and hearers. But the canonical status of Grice’s ideas masks their unconventional and even controversial beginnings. As Baker himself observes, Grice’s heresies have tended to be transformed by success into orthodoxies. In fact, Grice’s work was often characterised by the challenges it posed to accepted wisdom, and by the novel approaches it proposed to established philosophical issues. The theory of conversational implicature is a good example. It is often summarised as one of the earliest and most successful attempts to explain the fact, familiar to common sense, that 1

1403_902976_02_cha01.qxd 12/13/04 6:24 PM Page 2

2

Paul Grice, Philosopher and Linguist

what people literally say and what they actually mean can often be quite different matters. In particular, the theory is concerned with some apparent discrepancies between classical logic and natural language. More generally, it addresses the question of whether the meaning of everyday language is best explained in terms of formal linguistic rules, or in terms of the vagaries and variables of human communication. Grice’s theory developed against the background of a sharp distinction of approaches to these issues. Philosophers such as Bertrand Russell, who saw logic as the appropriate apparatus for explaining meaning, dismissed the differences displayed by natural language as examples of its inherent imperfection. Everyday language was just too messy and imprecise to form an appropriate topic for philosophical inquiry. The opposing view is perhaps best summed up by the slogan from Wittgenstein’s later work that ‘meaning is use’.3 Philosophers from this school of thought argued that if natural language diverges from logical meaning, this is simply because logic is not the appropriate philosophical tool for explaining language. Meaning in language is an important area of study in its own right, but can be considered only in connection with the variety of ways in which language is used by speakers. Grice’s approach to this debate was to argue that both views were wrong-headed. He proposed to think outside the standard disjunction of positions. Logic cannot give an adequate account of natural language, but nor is language simply a multifarious collection of uses, unamenable to logical analysis. Rather, logic can explain much, but not all, of the meaning of certain natural language expressions. More generally, formal semantic rules play a vital part in explaining meaning in everyday language, but they do not do the whole job. Other factors of a different but no less important type are also necessary for a full account of meaning. Most significantly, Grice argued that these other, nonsemantic factors are not a random collection entirely dependent on individual speakers and contexts, but can be systematised and explained in terms of general principles and rules. The principles that explain how natural language differs from logic can also explain a wide range of other features of human communication. In this novel attitude, Grice was certainly rejecting formal approaches, with their claim that the only meaning amenable to philosophical discussion was that which could be described in terms of truth-conditions, and could enter into truth-functional relationships. But also, and perhaps more surprisingly, he was rejecting a belief in everyday use as the chief, or indeed the only, location of meaning. This was a central tenet of philosophy at Oxford while Grice was developing his theory of

1403_902976_02_cha01.qxd 12/13/04 6:24 PM Page 3

The Skilful Heretic 3

conversation, or at least of that subsection of Oxford philosophy known as the philosophy of ordinary language. Grice himself was an active member of this subsection and is often referred to as a leading figure in the movement. However, his use of formal logic in explaining conversational meaning demonstrates that his heretical impulse extended even to ordinary language philosophy itself. Indeed, the success of his theory of conversational implicature has been credited with the eventual demise of this movement as a viable philosophical approach.4 Grice’s readiness to question conventions, even those of his own subdiscipline, makes him hard to categorise in terms of the usual philosophical distinctions. He does not sit easily on either side of the familar dichotomies; he is neither exactly empiricist nor exactly rationalist, neither behaviourist nor mentalist. This makes synoptic discussions of his work difficult, as is perhaps entirely appropriate for a philosopher who readily expressed his dislike of attempts to divide philosophy up into a series of ‘-ologies’ and ‘-isms’. The theory of conversation is undoubtedly the best known of Grice’s work. The particular, and often exclusive, emphasis on it can be explained perhaps in part by its intrinsic potential as a tool for linguistic analysis, and perhaps also, more accidentally, by the historical inaccessibility of much of his other writing. A notorious perfectionist, Grice was seldom happy that his work had reached a finished, or acceptable, state, and was therefore always reluctant to publish. Those who knew him have related this reluctance to the penetration of his own philosophical vision, turned as relentlessly on his own work as on that of others. His Oxford colleague Peter Strawson has suggested that this resulted in an uncomfortable degree of self-doubt: I suspect, sometimes, that it was the strength of his own critical powers, his sense of the vulnerability of philosophical argument in general, that partially accounted, at the time, for his privately expressed doubts about the ability of his own work to survive criticism. After all, if there were always some detectable flaws in others’ reasoning, why should there not be detectable, even though by him undetected, flaws in his own?5 Certainly, Grice’s account of conversation as an essentially cooperative enterprise is generally discussed, and frequently criticised, in isolation from the rest of this work, seen as a free-standing account of linguistic interaction. It is paraphrased, often in just a few pages, in most introductory text books on pragmatics and discourse analysis.6 It has

1403_902976_02_cha01.qxd 12/13/04 6:24 PM Page 4

4

Paul Grice, Philosopher and Linguist

been used or referred to in analyses of gendered language, children’s language, code switching, courtroom cross-examination, the language of jokes, oral narrative and the language of liturgy, along with many other topics.7 However, it is only one aspect of a large and diverse body of work from a career of over four decades. In the context of this work, it can be seen as an intrinsic part of the gradual development of Grice’s thinking on a range of philosophical topics. It also offers an analytic tool that Grice himself applied in his later work to a variety of philosophical problems, although never to the data of actual conversation. To some extent, then, Grice’s later use of the theory of conversation offers its own explanation of the significance, and the theoretical purpose, of that work. More than that, however, Grice’s less-known work merits attention in its own right. It ranges over a wide range of topics: not just the philosophy of language but epistemology, perception, logic, rationality, ethics and metaphysics. Nevertheless, it is possible to identify a number of running threads, and some striking continuities of theme and approach, throughout this diverse oeuvre. Grice himself argued that the disparate themes and ideas of his career displayed a fundamental unity.8 In general, however, as Richard Grandy and Richard Warner observe in their introduction to Philosophical Grounds of Rationality, ‘the systematic nature of his work is little recognised’.9 Throughout his work Grice focused on aspects of human behaviour, linguistic or otherwise, and the mental processes underlying them. Increasingly central in his work was the idea that an analysis of these processes revealed people to be rational creatures, and that this rationality was fundamental to human nature. In an overview of his own work, Grice himself suggested, with typical tentativeness, that: ‘It might be held that the ultimate subject of all philosophy is ourselves.’10 He also displayed a sophisticated respect for common sense as a starting point in philosophical inquiry. It is sophisticated in that he did not espouse the straightforward adoption of common sense attitudes and terminology into philosophy. Rather, he urged that the philosopher remain aware of how the themes of philosophical inquiry are usually understood by non-philosophers. In a large part, this meant paying serious attention to the language in which particular issues were ordinarily discussed, or to what Grice once described as ‘our carefree chatter’.11 In this focus at least he retained an approach recognisable from his background in ordinary language philosophy. However, he proved himself ready to go beyond the dictates of common sense where the complexity of the subject matter demanded. Philosophy, he argued, is a difficult subject

1403_902976_02_cha01.qxd 12/13/04 6:24 PM Page 5

The Skilful Heretic 5

that deals with difficult issues, and it is often necessary to posit entities not current in everyday thought: to construct theoretical, empirically unverifiable entities if they are explanatorily useful. Within these parameters, as he once suggested, ‘whatever does the job is respectable’.12 He was certainly not a simple egalitarian in philosophy. In the early 1980s he noted down, apparently for his own edification, the following complaint: It repeatedly astonishes me that people who would themselves readily admit to being devoid of training, experience, or knowledge in philosophy, and who have plainly been endowed by nature with no special gifts of philosophical intelligence should be so ready to instruct professional philosophers about the contents of the body of philosophical truths.13 Despite his readiness to add entities to philosophical explanations when this seemed appropriate, Grice was concerned that such explanations should remain as simple as possible, in the philosophical sense of simplicity as drawing on as few entities as possible. In the lectures on conversation he endorsed ‘Modified Occam’s Razor’, which he defined as ‘senses are not to be multiplied beyond necessity’.14 This philosophical parsimony was carried over into other areas of his work; he sought general explanations that would account for as wide a range of subject matter, for as few theoretical constructions, as possible. If he prized philosophical simplicity, Grice’s ideas were often far from reductive in more general terms. In particular, he did not shrink from explanations that saw people as separate and different: different from animals, in terms of communicative behaviour, or from automata, in terms of rational thought. More generally, Grice sustained an enthusiasm for the business of philosophy itself, accompanied by a conviction, unfashionable at the start of his career, that the works of philosophers from other schools of thought and different ages deserve thoughtful and continual reanalysis. His work reveals a belief that philosophical investigation and analysis are worthwhile ventures in their own right and a willingness to employ them to offer suggested approaches, rather than completed solutions, to philosophical problems. Further, Grice was always ready to hunt out and consider as many possible counter-arguments and refutations of his own ideas as he could envisage, often advancing some particular idea by means of a detailed discussion of an actual or potential challenge to it. These factors together lend a ‘discursive’ and at times a

1403_902976_02_cha01.qxd 12/13/04 6:24 PM Page 6

6

Paul Grice, Philosopher and Linguist

frustratingly tentative air to Grice’s work. But despite the earnestness of Grice’s interest in philosophy, and the rigour with which he pursued it, his tone was often playfully irreverent. This was, to some extent, the result of a deliberate policy that philosophy could, indeed should, be fun. ‘One should of course be serious about philosophy’, he argued, ‘but being serious does not require one to be solemn.’15 Grice was not a gifted public speaker, given to elaborate divergences in exegeses, and to stuttering and even mumbling in presentation.16 Nevertheless, his presentations could be entertaining occasions; tape recordings of him speaking about philosophy, both informally and formally, are often punctuated by laughter. Grice himself suggested that his search for the fun, even the funny in philosophy was prompted by ‘the wanton disposition which nature gave me’.17 But it had been reinforced ‘by the course of every serious and prolonged philosophical association to which I have been a party; each one has manifested its own special quality which at one and the same time has delighted the spirit and stimulated the intellect’. He had no time for those whose method of conducting philosophical debate was that of ‘nailing to the wall everything in sight’. In his view, philosophy was best when it was a collaborative, supportive activity. This idea was, to some extent, inherent in his philosophical background; ordinary language philosophy was frequently undertaken as a group activity. However, it is an idea that was peculiarly suited to his personality. Grice could be extremely convivial; he was often surrounded by people, and preferred to develop his ideas by discussion rather than by solitary composition. He was no ascetic; he ate, drank and smoked copiously. But conviviality is not the same as affability. His delight in philosophical discussion often spilled over into an intense desire to ‘win’ the argument. He could be curt in response to points of view with which he was not in sympathy. On those occasions when he sought intuitive responses to his linguistic examples, he could be dismissive of those who got it ‘wrong’. Moreover, those who know him have suggested that at times his tendency towards self-doubt could manifest itself in gloominess, even moroseness.18 Grice’s tendency was to become entirely absorbed with whatever he was engaged in, if it interested him intellectually. Often this was some philosophical issue of substance or of composition. His wife Kathleen recalls that he frequently stayed up late into the night, often pacing up and down as he battled with some problem.19 Sometimes other people were caught up in this absorption. During his time at Oxford, he conducted a long collaboration with Peter Strawson. In later life, Grice

1403_902976_02_cha01.qxd 12/13/04 6:24 PM Page 7

The Skilful Heretic 7

himself would recount how he was once phoned up by Strawson’s wife who told him to stop bothering her husband with late night phone calls.20 This obsessiveness was characteristic of all the activities to which he devoted his considerable energies. Kathleen remembers that he was always engaged on some project; if it was not philosophy then it was playing the piano, or chess, bridge or cricket. These last two were more absorbing than mere hobbies. He played both bridge and cricket competitively at county level. Cricket, in particular, was an obsession that almost rivalled philosophy; he played for a number of different clubs and ‘became an inelegant but extremely effective and prolific opening batsman’.21 During his Oxford career, he spent the large part of each summer on cricket tours. Grice’s immense energy in these different directions was undoubtedly aided by the amount of time he had at his disposal, in part the consequence of not having to concern himself too much with everyday practicalities; according to Kathleen, from childhood onwards he always had, or found, people to look after him. He certainly pursued his interests to the exclusion of the mundane, often neglecting food and sleep if a particular problem, or game, had his attention. In those areas where he did have practical responsibility, Grice was legendarily untidy. He took little concern over his clothes, or his personal appearance generally. His desk was constantly covered by huge and apparently unsorted piles of paper. However, he would not allow anyone else to touch these, claiming that he knew exactly where everything was. After he died, these papers were deposited in the Bancroft library at the University of California, Berkeley, as the H. P. Grice Papers. This archive, which amounts to 14 large cartons, contains whatever Grice had about himself at the end of his life, mainly heaped on his desk or crammed under his bed. It consists largely of papers from after his move to Berkeley in 1967. But it also includes those papers from his Oxford days, dating right back to the 1940s, that had seemed important enough to him to take and keep with him. The cartons contain a mixture of finished manuscripts, draft versions, lecture notes and odd jottings. These are often crammed with writing, but never with anything extraneous to the matter in hand; it seems that Grice never doodled. They offer some clue to at least one reason for Grice’s reluctance to publish. Grice seems to have been more struck than most by the essentially inconclusive nature of his own work; no project was, for him, ever really completed, or ever separate from the work that preceded and followed it. So notes and manuscripts were stored along with those from decades earlier if they were perceived to be on a related

1403_902976_02_cha01.qxd 12/13/04 6:24 PM Page 8

8

Paul Grice, Philosopher and Linguist

theme. Papers covered in Grice’s cramped hand in faint pencil, characteristic of his work from Oxford, were annotated and added to by notes in ball-point made years later in Berkeley. Ideas generally associated exclusively with his late work, such as those relating to rationality and to finality, are explored in notes dating back to the 1960s. Work often remained in manuscript form for so long that it needed to be updated as the years went by. In the original version of ‘Indicative conditionals’, part of the William James lectures of 1967, Grice uses the example: ‘Either Wilson or MacMillan will be P.M.’. At a later date ‘MacMillan’ has been crossed out and ‘Heath’ written over the top in a different coloured ink. This was the example used when the lecture was eventually published.22 Above all, the H. P. Grice papers confirm that for Grice there was no real distinction between life and work. He was not in the habit of keeping regular hours of work, or of distinguishing between philosophical and private concerns. In the same way, his current philosophical preoccupation would spill over into his everyday life. Any piece of paper that came to hand would be covered in logical notation, example sentences, trial lists of words, invented dialogues, or whatever else was preoccupying him at that moment, all written in what Grice himself described as ‘a hand which few have seen and none have found legible’.23 So the cartons contain not just the orthodox pages of ruled note paper, but also a miscellany of correspondence (often unopened), financial statements, menus, paper napkins, playing card boxes, cigarette packets and even airline sick bags that came to hand as ideas occurred to him during a flight. Some of these suggest that Grice’s writing style was not as spontaneous as it may sometimes appear. Certainly the manuscripts of articles and the lecture notes, always written out in full, were produced in longhand with little significant revision. But Grice was not averse to jotting down what he once described as ‘useful verbiage’ in preparation for these. In conjunction with some work on the place of value in human nature drawing on, but diverging from a number of previous philosophers including John Stuart Mill, he noted down the phrase: ‘so as not to be just Grice to your Mill . . .’.24 While working on an account of value and freedom drawing on both Aristotle and Kant, he tried out possible hybrids: ‘Ariskant? Kantotle?’.25 This book considers Grice’s work within a broadly chronological framework, following the course of his philosophical life. However, because Grice did not work on discreet topics in neat succession, it is sometimes necessary to group together strands of work on related topics even where they in fact extend over years or decades. Nevertheless, the

1403_902976_02_cha01.qxd 12/13/04 6:24 PM Page 9

The Skilful Heretic 9

chronological arrangement makes possible an understanding of the development, as well as the remarkable unity, of Grice’s thinking. It also allows some scope for considering the impact on it of the work of other philosophers, and of the various personal associations he formed throughout his life. The final chapter is concerned with the impact of Grice’s ideas on linguistics. It is concerned with the development of what has become known as ‘Gricean pragmatics’ and therefore concentrates on responses to the theory of conversation.

1403_902976_03_cha02.qxd 12/13/04 6:25 PM Page 10

2 Philosophical Influences

In a conversation recorded in 1983, Grice commented that fairly early in his career he stopped reading current philosophy.1 In the light of his constant engagement with the work of his contemporaries, this characteristically mischievous claim need not be taken at face value. Indeed, his own immediate elaboration somewhat modifies it. Instead of spending his time trying to keep up with all the philosophical journals, he concentrated increasingly on the history of philosophy, choosing his reading matter by strength of ideas rather than by date of composition. To this it might be added that he also devoted a great deal of time to face-to-face discussion with those he chose as his philosophical colleagues. Grice’s exaggerated claim therefore draws attention to two constant influences on his own philosophy: debate and collaboration with others engaged in similar fields of enquiry, coupled with what in the same conversation he calls ‘respect for the old boys’. It is tempting to identify the emergence of this tension between old and new ideas, or perhaps more accurately this dialogue between orthodox and challenging ways of thinking, throughout Grice’s early life. Born on 15 March 1913, he was the first son of a wealthy middle-class couple, Herbert and Mabel (née Felton) Grice, and grew up in Harborne, an affluent suburb of Birmingham. He was named after his father, but preferred his middle name, and was from an early age known generally as Paul. His early publications were credited to ‘H. P. Grice’, and are sometimes still cited as such, but in later years he published as simply ‘Paul Grice’. Herbert Grice is described in his son’s college register as ‘business, retd’. In fact he had owned a manufacturing business making small metal components that prospered during the First World War. When the business subsequently began to fail, Mabel stepped in to save the family finances by taking in pupils. She included Paul and his 10

1403_902976_03_cha02.qxd 12/13/04 6:25 PM Page 11

Philosophical Influences 11

brother Derek in this miniature school, meaning that the first few years of their education were conducted at home. This reversal of roles explains Herbert’s early ‘retirement’. After the failure of the manufacturing business, he did not attempt another venture, preferring instead to concentrate on his skills as a concert cellist. His musical talent was inherited by both his sons; Herbert, Paul and Derek would often perform domestically as a trio. Family life at Harborne was punctuated not just by musical recitals, and by innumerable games of chess, but also by controversy and debate, although the topic was generally theological rather than philosophical. Herbert had been brought up in a nonconformist tradition, while Mabel was a devout Anglo-Catholic. The third adult member of the household, an unmarried aunt, had converted to Catholicism. Grice recalled fondly the many doctrinal clashes he witnessed as a result of this mix, particularly those between his aunt and his father. He suggested that his father’s self-defence in these circumstances awakened, or at least reinforced, his own tendency towards ‘dissenting rationalism’; this tendency stayed with him throughout life.2 It would seem from this that he lent more towards his father’s way of thinking, but certainly by the time he reached adulthood he had abandoned any religious faith he may initially have held. He did not retain the Christianity with which he had been surrounded as a child, but he kept the enthusiasm for debate in general, and for the habits of questioning received wisdom and challenging orthodoxy on the basis of personal reasoning in particular. When Grice was 13 his education was put on to a more formal footing when he went as a boarder to Clifton College in Bristol. This public school, exclusively male at the time, provided excellent preparation for Oxford or Cambridge for the sons of those wealthy enough to afford it, or for those who, like Grice, passed the scholarship examination. Boys belonged to one of a number of ‘houses’, rather in the style of Oxbridge colleges, and received the education in classics that would equip them for entry to university. It seems that Grice excelled at this; at the end of his Clifton career, in July 1931, he won a classical scholarship to Corpus Christi College, Oxford. He had not concentrated exclusively on academic matters, however. He had been ‘Head of School’ during his final year, and had also kept up his musical interests. He performed a piano solo in the school’s 1930 Christmas concert. Joseph Cooper, Grice’s contemporary at Clifton who went on to become a concert pianist and then chairman of the BBC television programme ‘Face the Music’, played a piece by Rachmaninoff at the same concert. ‘We

1403_902976_03_cha02.qxd 12/13/04 6:25 PM Page 12

12 Paul Grice, Philosopher and Linguist

enjoyed Grice’s playing of Ravel’s “Pavane” ’, runs the school’s report on the concert, ‘its stateliness provided an effective contrast to the exuberance of the Rachmininoff.’3 Corpus Christi had a strong academic tradition, but was not as socially fashionable as some of the larger colleges. It therefore tended to attract students from more modest backgrounds than colleges such as Christ Church or Magdalen, where social success often depended on demonstrable affluence. However, despite such distinctions between individual colleges, Oxford in general was a place of privilege. Students were nearly all male, were predominantly from public schools, and were generally preparing to take their places as members of the establishment. Fashionable political opinions were left wing, and students of Grice’s generation tended to see themselves as rebelling against nineteenth-century notions of responsibility and propriety. Such rebelliousness, however, was of a very passive nature; it lacked the zeal and the active protest that was to characterise student rebellion in the 1960s. It did little to affect the day to day life in Oxford, where the university was legally in loco parentis, and where all undergraduates lived and dined in college. Some 50 years later, Grice recalled the atmosphere of the time with amused but affectionate detachment: We are in reaction against our Victorian forebears; we are independent and we are tolerant of the independence of others, unless they go too far. We don’t like discipline, rules (except for rules of games and rules designed to secure peace and quiet in Colleges), selfconscious authority, and lectures or reproaches about conduct. . . . We don’t care much to talk about ‘values’ (pompous) or ‘duties’ (stuffy, unless one means the duties of servants or the military, or money extorted by the customs people). Our watchwords (if we could be moved to utter them) would be ‘Live and let live, though not necessarily with me around’ or ‘If you don’t like how I carry on, you don’t have to spend time with me.’4 Grice was at Corpus Christi for four years. Classics was an unusual subject in this respect; most Oxford degree programmes were a year shorter. At that time, philosophy was taught only as part of the Classics curriculum, and it was in this guise that Grice received his first formal training in the subject. The style was conservative, based largely on close reading of the philosophers of Ancient Greece. This seems to have suited Grice’s meticulous and analytical style of thought well, as the tutorial teaching method suited his dissenting and combative

1403_902976_03_cha02.qxd 12/13/04 6:25 PM Page 13

Philosophical Influences 13

nature. Although there were lectures, open to all members of the university, teaching was based principally at tutorial, therefore college, level. Students would meet individually with their tutors to read, and then defend, an essay. In reflecting on his early philosophical education, Grice always emphasised what he saw as his own good fortune in being allocated as tutee to W. F. R. (Frank) Hardie. A rigorous classical scholar and an exacting tutor, Hardie might not have been everyone’s first choice, but Grice credited him with revealing the difficulty, but also the rewards of philosophical study. Grice also, it seems, relished the formalism that Hardie imposed on his already established appreciation of rational debate. Under Hardie’s guidance, this appreciation developed into a belief that philosophical questions are best settled by reason, or argument: I learnt also form him how to argue, and in learning how to argue I came to learn that the ability to argue is a skill involving many aspects, and is much more than an ability to see logical connections (though this ability is by no means to be despised).5 Grice even recounts with approval, but reluctant disbelief, the story that a long silence in another student’s tutorial was eventually broken by Hardie asking ‘And what did you mean by “of”?’.6 No doubt the contemporary detractors of ordinary language philosophy would have seen Grice’s enthusiasm for this anecdote of the 1930s as a sign that he was perfectly suited to the style of philosophy that would flourish at Oxford in the following two decades. This was crucially concerned with minute attention to meaning and to the language in which philosophical issues were traditionally discussed. Grice’s admiration for Hardie was later to develop into friendship; they were both members of a group from Oxford that went on walking holidays in the summers leading up to the Second World War, and Hardie taught Grice to play golf. The friendship was based in part on a number of characteristics the two had in common. There was their shared rationalism; Grice found welcome echoes of his own approach in Hardie’s ‘reluctance to accept anything not properly documented’.7 Moreover, Hardie never admitted defeat. Discussions with him, Grice later observed, never actually ended; they would simply come to a stop. In response, Grice learnt to adopt a similar stance in philosophical argument, clinging tendentiously to whatever position he had set out to defend. Once, when Grice was taught by a different philosophy tutor for one term, the new tutor reported back to Hardie that his student

1403_902976_03_cha02.qxd 12/13/04 6:25 PM Page 14

14 Paul Grice, Philosopher and Linguist

was ‘obstinate to the point of perversity’. To Grice’s enduring delight, Hardie received this report with thorough approval. However some may have judged his style of argument, Grice’s undergraduate career was an indisputable success. Students took ‘Moderations’ exams five terms into their degree, before proceeding to what was popularly known as ‘Greats’, the stage in the degree at which philosophy was introduced. Grice achieved First Class in his Moderations in 1933. In the summer of 1935 he took his finals, and was awarded First Class Honours in Literae Humaniores, as ‘Greats’ was officially called. His studies did not take up all of his time during these years, however, and he mixed academic success with extracurricular activities, particularly sporting ones. He played cricket, and in the summer of 1934 captained the college team. In the winter he followed up this success by becoming captain of the college football team. During this same period, he was president of the Pelican Philosophical Society and editor of the Pelican Record. Both took their name from the bird forming part of the Corpus Christi crest that had become the informal symbol of the college. After Grice completed his undergraduate studies, there was a brief hiatus in his academic career. There were at that time few openings at Oxford for those graduates who had not yet gained a University Lectureship or been elected to a College Fellowship. For the academic year 1935–6 Grice left Oxford, but not the elite education system of which he had been part for the past decade, when he went as Assistant Master to Rossall in Lancashire. At a time when there were very few Universities in the country, an Honours degree, especially from Oxford or Cambridge, was regarded as more than adequate a qualification to teach at a public school such as Rossall. It was not at all uncommon for Assistant Masters to stay in post for just a year or two before moving on to a more permanent position or embarking on a career in some other field. Indeed, of the 11 other Assistant Masters appointed between 1933 and 1937 by the same headmaster as Grice, only four stayed for more than a year; none of the other seven went on to careers as school masters.8 In Grice’s case, his appointment at Rossall filled the time between his undergraduate studies and his return to Oxford in 1936. It was also in 1936, unusually a year after the completion of his studies, that Grice was willing, or perhaps able, to pay the fee necessary for conferment of his BA degree. He was enabled to return to Oxford by two scholarships, both of which would have required him to sit examinations. First was an Oxford University Senior Scholarship, an award for postgraduate study at any

1403_902976_03_cha02.qxd 12/13/04 6:25 PM Page 15

Philosophical Influences 15

college and in any subject. In the same year, Grice was awarded a Harmsworth Senior Scholarship, a ‘closed’ scholarship to Merton college. The Harmsworth Scholarships, funded by an endowment from former college member Sir Hildebrand Harmsworth, were a relatively new venture, aimed specifically at raising the profile of postgraduate achievement at the college. A history of Merton notes that the scheme ‘brought to the college a succession of intelligent graduates from all colleges. In that way a window was opened on to the world of graduate research at a time when there were few opportunities at that level in the university as a whole.’9 The scholarships had been available only since 1931, and only three were awarded each year, across the range of academic subjects. Grice’s contemporaries were Richard Alexander Hamilton, also from Clifton College, in sciences, and Matthew Fontaine Maury Meiklejohn in modern languages. Grice was a Harmsworth Senior Scholar at Merton between 1936 and 1938, but no award or higher degree resulted from his time there. He was awarded an MA from Oxford in 1939 but this was a formality, available seven years after initial matriculation to students who had received a BA, upon payment of a fee. Nevertheless, the time spent at Merton was an important period in Grice’s philosophical development not least because it was during this time that he became aware of subjects and methodologies then current in philosophy, and indeed taking shape in Oxford itself. He added knowledge of contemporary and dissenting philosophy to his existing familiarity with established orthodoxy. Grice’s experiences of undergraduate study had been typical of the time. The philosophy taught and conducted at Oxford between the wars was extremely conventional, that is to say speculative and metaphysical in nature, with an almost institutionalised distrust of new ideas. The business of philosophy was centred on the exegesis of the philosophical, and particularly the classical ‘greats’. Indeed, until just before the Second World War, no living philosopher was represented on the Oxford philosophy syllabus. The picture was different at Cambridge. From early in the century, Russell had been developing his analytic approach to issues of knowledge and the language in which it is expressed, arguing that suitable rigorous analysis can reveal the ‘true’ logical form beneath the ‘apparent’ grammatical surface of sentences. Wittgenstein’s Tractatus, first published in English in 1922, had taken these ideas further, arguing that such analysis could reveal many of the traditional concerns of philosophy to be merely ‘pseudo-problems’. However, such was the insularity of academic institutions at the time that Oxford had not officially, and barely in practice, been aware of

1403_902976_03_cha02.qxd 12/13/04 6:25 PM Page 16

16 Paul Grice, Philosopher and Linguist

these developments in Cambridge, let alone of those in other parts of Europe. This isolation was fairly abruptly ended, at least for some of the younger philosophers at Oxford, during the period when Grice was at Merton. This change was due in a large part to A. J. Ayer, and to the publication of his controversial first book Language Truth and Logic in 1936. Ayer had himself been an undergraduate at Oxford between 1929 and 1932 but, unusually, had read many of Russell’s works, as well as the Tractatus. He had been encouraged in this by his tutor, Gilbert Ryle, a socially conservative but intellectually progressive don, who had even taken Ayer on a trip to Cambridge where he had met Wittgenstein personally. Ayer was impressed by the methods of analytic philosophy in general, and in particular by Wittgenstein’s ambitious claims that it could ‘solve’ age-old philosophical problems for good by analysing the language in which they were traditionally expressed. He was even more profoundly affected by the ideas with which he came into contact during a visit to Austria after finishing his studies. There he became acquainted with a number of members of the Vienna Circle, and attended several of their meetings. Then as now, this group of scientists and philosophers were referred to collectively as ‘the logical positivists’. They represented a number of separate interests and specialisms, but shared at least one common goal: the desire to refine and perfect language so as to make it an appropriately precise tool for scientific and logical discussion. They distinguished between meaningful statements, amenable to scientific study and discussion, and other statements, strictly meaningless but unfortunately occurring as sentences of imperfect everyday language. Their chief tool in this procedure was the principle of verification. The logical positivists divided the category of meaningful statements into two basic types. First, there are those expressed by analytic, or necessarily true, sentences. ‘All spaniels are dogs’ is true because the meaning of the predicate is contained within the meaning of the subject, or because ‘dog’ is part of the definition of ‘spaniel’. They added the statements of mathematics and of logic to the category of analytic sentences, on the grounds that a sentence such as ‘two plus two equals four’ is in effect a tautology because of the rules of mathematics. The second category of meaningful statements is those expressed by synthetic sentences, or sentences that might be either true or false, that are capable of being subjected to an identifiable process of verification. That is, there are empirically observable phenomena sufficient to determine the question of truth. All other statements, those neither expressed by

1403_902976_03_cha02.qxd 12/13/04 6:25 PM Page 17

Philosophical Influences 17

analytic sentences nor verifiable by empirical evidence, are meaningless. This includes statements concerning moral evaluation, aesthetic judgement and religious belief. For the logical positivists, such statements were not false; they merely had no part in scientific discourse. The doctrine of the Vienna Circle was radically empiricist and strictly anti-metaphysical. When Ayer returned to Oxford in 1933 as a young lecturer, he set himself the task of introducing the ideas of logical positivism to an English-speaking audience. Language, Truth and Logic draws heavily on the ideas of the Vienna Circle, but also includes elements of the British analytic philosophy Ayer had found persuasive as an undergraduate. In the opening chapter, provocatively entitled ‘The Elimination of Metaphysics’, he outlines the views of logical positivism on meaning. He also offers his own formulation of the principle of verification: We say that a sentence is factually significant to any given person if, and only if, he knows how to verify the proposition which it purports to express – that is, if he knows what observations would lead him, under certain conditions, to accept the proposition as being true, or reject it as being false.10 In subsequent chapters, Ayer applies this criterion uncompromisingly to the analysis of statements such as those of ethics and theology, those of our perceptions of the material world, and those relating to our knowledge of other minds and of past events. Language, Truth and Logic, which was widely read and discussed both within Oxford and beyond, tended to polarise opinions. Not surprisingly, many of the older, established Oxford philosophers reacted strongly against it and the explicit challenge it posed to philosophical authority. It dismissed many of the traditional concerns of philosophy on the grounds that they were based on the discussion of meaningless metaphysical statements; it proposed to solve many apparent philosophical problems simply by analysis of the language in which they were expressed and, of course, it rejected all religious tenets. However, to many of the younger philosophers it offered a welcome and exhilarating change for precisely these same reasons. Ayer became a central figure in a group of young philosophers who met to discuss various philosophical topics. The meetings took place in the rooms of Isaiah Berlin in All Souls College during term times for about two years from early 1937. Here Ayer first began to talk, and to disagree, with another young philosopher, J. L. Austin.

1403_902976_03_cha02.qxd 12/13/04 6:25 PM Page 18

18 Paul Grice, Philosopher and Linguist

In the decade immediately following the Second World War, Austin was to be instrumental in the development of ordinary language philosophy, which dominated Oxford, and was deeply influential on Grice personally. In the late 1930s, however, his philosophical position was yet to be fully articulated. Ayer’s biographer, Ben Rogers, has suggested that Austin was initially very impressed by Ayer, who was the only member of the group to have published, but that ‘during the course of the All Souls Meetings . . . Austin staked out, if not a set of doctrines, then an approach which was very much his own’.11 This approach brought him increasingly into conflict with Ayer. Initially their disagreements were amicable enough, but in later years their differences of opinion were to lead to an increasing and unresolved hostility. Grice perhaps identifies what it was that Austin found increasingly unsatisfactory about Ayer’s logical positivism when he observes in an unpublished retrospective that: ‘between the wars, in the heyday of Philosophical Analysis, when these words were on every cultured and progressive lip, what was thought of as being subjected to analysis were not linguistic but non-linguistic entities: facts or (in a certain sense) propositions, not words or sentences’.12 Analytic philosophy, whether in terms of Russell’s ‘logical form’, or of the verificationists’ ‘meaningfulness’, was concerned with explaining problematic expressions by translating statements in which such expressions occur into more elementary, logically equivalent statements in which they do not occur. It considered language as an important tool for the clarification of concepts and states of affairs, not as an appropriate target of analysis in its own right. Indeed language, as used in ordinary discourse, offered potential dangers that the philosopher needed to avoid; it could mislead the unwary into inappropriate metaphysical commitments. The major advocates of philosophical analysis between the wars, Russell, Wittgenstein and the Vienna Circle, all shared the conviction that everyday language was ‘imperfect’ and needed to be purified and improved before it could be ready for philosophical use. This disdain for ordinary language is at least in part the source of Austin’s growing discontent with analytic philosophy in general, and with its spokesman Ayer in particular. Austin became increasingly convinced that ordinary language was a legitimate focus of philosophical enquiry. Not only was it ‘good enough’ as a tool for philosophy, it was also possible that the way in which people generally talk about certain areas of experience might offer some important insights into philosophically problematic concepts. For instance, there is the problem of our knowledge of other minds. For Ayer, statements about the mental states of others were

1403_902976_03_cha02.qxd 12/13/04 6:25 PM Page 19

Philosophical Influences 19

meaningless and unacceptable as they stood. A statement such as ‘he is angry’ was not amenable to any process of verification, and committed its speaker to the existence of unempirical concepts such as minds and emotions. To avoid the unpleasant prospect of having to dismiss all such statements as simply meaningless, Ayer insisted in Language, Truth and Logic that they must be translatable into verifiable statements. Statements about others’ mental states are most appropriately translated into statements about their physical states: about their appearance and behaviour. Such statements, which concern phenomena that can be directly perceived, are amenable to verification. Assumptions about mental states are merely inferences from these. Austin’s defence of statements about other minds would run as follows. People use statements such as ‘he is angry’ all the time, with no confusion or misunderstanding. Therefore, it seems appropriate to take such statements seriously: to analyse them as they stand rather than to treat them as being in some way ‘really’ about perceptions and inferences. Furthermore, the fact that people ordinarily talk in this way might be seen as good grounds for accepting emotions and mental states as valid concepts. Grice had a lot in common with the group meeting in All Souls. He was much the same age; of the main members of the group, Austin was two years older than him, Ayer three and Berlin five, while Stuart Hampshire was a year younger. Moreover, like both Ayer and Austin he had studied Classics at Oxford, and been introduced to the study of philosophy in ‘Greats’. But college life was so insular that he was not aware of this group or party to its discussions. Merton was in many ways a similar college to Corpus Christi: small and relatively low in the social hierarchy of colleges. Indeed, it had a growing reputation for offering places to students without traditional Oxford credentials: to those from the Midlands and the North of England, and to grammar school pupils. In contrast, Christ Church and Magdalen, where Ayer and Austin were respectively, were both large and prestigious colleges. All Souls was exclusively a college of Fellows, offering no places to undergraduate or postgraduate students. Grice was later to account for his own absence from the meetings by explaining that he was ‘brought up on the wrong side of the tracks’.13 He did read Language Truth and Logic, and like many of his contemporaries was extremely interested in the methodology it suggested for tackling philosophical problems. However, he recalled in later life that from the beginning he had certain reservations about Ayer’s claims that were never laid to rest; ‘the crudities and dogmatism seemed too pervasive’.14

1403_902976_03_cha02.qxd 12/13/04 6:25 PM Page 20

20 Paul Grice, Philosopher and Linguist

Grice may have been unimpressed by some of Ayer’s more sweeping claims, but they opened up to him analytic possibilities that affected his own style of philosophy. An early typescript paper on negation has survived. This was a topic to which Grice returned in a series of lectures in the early 1960s, but on which he never published. The paper is not dated, but the use of his parents’ address in Harborne suggests that it was written before he obtained a permanent position at Oxford, therefore presumably before the Second World War. He tackles the epistemological problem posed by negative statements such as ‘this is not red’ or ‘I am not hearing a noise’. Such statements raise the question of how it is that we become acquainted with a negative fact. Grice argues that knowledge of negative facts is best seen as inferential. We are confident in our knowledge of a negative because we have evidence of some related positive fact. So our knowledge that a certain object is not red might be derived by inference from knowledge that it is green, together with an understanding that being green is incompatible with being red. The second example sentence is more problematic. It expresses a ‘privative fact’, an absence. If the inferential account of negative knowledge is to be maintained, it must be possible to suggest a type of fact incompatible with the positive counterpart of the negative, that is with the sentence ‘I am hearing a noise’, for which the speaker can have adequate evidence. Grice’s suggestion is that the incompatible fact offering a solution to this problem is the fact that the speaker entertains the positive proposition in question, without having an attitude of certainly to that proposition. If you think of a particular proposition, and are aware that you do not know it to be true, then you may be said to be inferring knowledge of a negative. If you entertain the proposition that you are hearing a noise, but do not know this proposition to be true, then you are in a position to assert ‘I do not hear a noise’. More generally, to state ‘I do not know that A is B’ is in effect to state ‘Every present mental process of mine has some characteristic incompatible with knowledge that A is B’.15 Although Grice does not acknowledge any such influences, his paper can be placed very much within the analytic tradition, and its main proposal draws on verificationist assumptions. The problem it tackles is the issue of how we are ‘really’ to understand negative statements: how we are to translate them so that they do not contain the problematic term ‘not’. Grice’s proposed analysis can be subjected to a process of verification, on the understanding that perception though the senses (‘it is green’) and introspection (‘every present mental process of mine . . .’) are empirical phenomena.

1403_902976_03_cha02.qxd 12/13/04 6:25 PM Page 21

Philosophical Influences 21

On completion of his Harmsworth studentship in 1938, Grice was appointed to a lectureship in philosophy at the third college of his Oxford career, St John’s. In the complex social hierarchy of Oxford colleges, this was definitely a step up; St John’s was larger, more affluent and more prestigious than either Merton or Corpus Christi. More importantly, it was a sign that Grice’s career was progressing well. Although it carried a relatively high teaching load, and brought with it no benefits of college membership, a lectureship was a good position from which to impress the existing Fellows of the college, who had control over the appointment of new members. In fact, Grice was elected to the position of full Fellow, tutor and lecturer in philosophy after just one year. He was to hold this post for almost 30 years, but initially he stayed at St John’s for only one because his career, like that of many of his contemporaries, was interrupted by war service. Grice was commissioned Lieutenant in the Navy in 1940. Initially, he was on active service in the North Atlantic. Then in March 1942 he joined Navy Intelligence at the Admiralty, where he stayed until the war ended. The war years saw the publication of Grice’s first article, ‘Personal Identity’, in Mind. Neither the topic nor the journal were particularly surprising choices for a young Oxford philosopher at that time. Mind, edited by G. E. Moore, was one of the leading, indeed one of the few, journals of contemporary philosophy, and had reflected the interest in logical positivism of the 1930s. After the war, under the editorship of Gilbert Ryle, it would become the acknowledged organ for much of the work from the new school of Oxford philosophy.16 The topic of personal identity was one of long-standing philosophical interest, and had received renewed attention during the 1930s; indeed, it was one of the topics discussed in the meetings at All Soul’s. It is ultimately concerned with a question that was to underlie Grice’s work throughout his life: the question of what it is to be a human being. Discussing personal identity means considering the relative importance of, and relationship between, physical and mental properties. Describing identity exclusively in terms of physical bodies presents problems, but so does describing it entirely in terms of mental entities. The former position would suggest that a single body must always be the location of a single identity, or person, regardless of personality change, memory loss or mental illness. Moreover, it would seem to entail that, since the composition of a physical body does not stay the same throughout a lifetime, someone must be regarded as having separate identities, or being different people as, say, a baby, an adult, and an old man. An account of personal identity based exclusively on mental sameness, however,

1403_902976_03_cha02.qxd 12/13/04 6:25 PM Page 22

22 Paul Grice, Philosopher and Linguist

would force us to accept that one mental state transferred into another body, perhaps by surgery or reincarnation, should result in no change of identity. Philosophers wrestled with hypothetical ‘problem cases’. If it were possible to take the brains out of two bodies and swap them over, would this be best described as a brain transplant or a body transplant? Ayer, Austin and their companions had concerned themselves with the status of Gregor Samsa from Kafka’s ‘Metamorphosis’: ‘was he an insect with the mind of a man, or a man with the body of an insect?’.17 In attempting to negotiate between these two opposing, equally problematic, accounts, Grice looks to ‘memory theories’ of personal identity. In this, he draws on work produced almost 250 years previously by John Locke, while acknowledging some of Locke’s critics and attempting to suggest a way of addressing them. In other words, Grice’s first published work shows the characteristic combination of old and new ways of thinking. He draws on work from philosophical ‘old boys’, such as Locke and his eighteenth-century critic Thomas Reid. At the same time, he looks for solutions to their problems in a much newer style of philosophy, one in general terms analytic, and in more particular terms concerned with the close analysis of individual linguistic examples. Locke argues that, unlike in the case of inanimate masses, the identity of living creatures must depend on more than bodily unity. As evidence he suggests the examples of an oak growing from a seedling into a great tree and then cut back, or a colt growing into a horse, being sometimes fat and sometimes thin; throughout these changes they remain the same oak, and the same horse. In search of an appropriate substitute for bodily sameness, he discusses what might be seen as a hierarchy of living beings, and considers where a notion of identity can be located in each case. In general, at each stage of the hierarchy, what is required is an account of function. The parts of a tree form a single entity not because they remain physically constant, but because they are arranged in such a way as to fulfil certain functions: the processes of nutrition and growth that together ensure the continued existence of the tree. In much the same way, the identity of an animal can be described in terms of the fulfilment of the functions that ensure its survival. Even ‘man’ is no different in this respect; human identity consists in the collection of parts functioning together over the course of a lifetime, to ensure the growth, maturation and survival of the individual human being. Various different particles of matter form a single human at different points in time precisely because they all participate in a single, continued, life.

1403_902976_03_cha02.qxd 12/13/04 6:25 PM Page 23

Philosophical Influences 23

Locke points out that this account of a man or human being, although it does not rely on simple bodily identity, does make necessary reference to the physical body. The parts of the body may change over time, but, united by common functions, they together form part of the definition of ‘a man’. If we relied only on mental identity, or ‘the identity of soul’, we would not be able to resist the idea that physically and temporally separate bodies might all belong to the same man. Locke suggests that an account based just on the soul could not rule out the unacceptable claim that the ill-assorted list: ‘Seth, Ismael, Socrates, Pilate, St Austin, and Caesar Borgia, . . . may have been the same man.’18 Furthermore, the form of a man is a necessary part of our understanding of the concept of a man, and of the meaning of the word ‘man’. Even if we were to hear a creature that looked like a parrot hold intelligent conversation and discuss philosophy, functions normally associated exclusively with people, we would hardly be tempted to say that it must be a man, but rather we would say that it was ‘a very intelligent rational parrot’.19 Locke’s next move is to suggest a categorial distinction between ‘man’ and ‘person’, a distinction he admits is at odds with normal understanding and speech. He offers a very specific definition of a person as ‘a thinking intelligent being, that has reason and reflection, and can consider itself as itself, the same thinking thing, in different times and places’.20 The point of this distinction is to explain the ‘extra’ properties persons are generally seen as possessing, beyond those of animals. It might be said variously to account for consciousness, self-awareness or the soul. The identity of a person, then, consists not just in the unity of functioning parts, the criterion that accounts alike for the identity of a plant, or an animal, or a ‘man’. The identity of a person depends on the continuation of the ‘reason and reflection’ across a range of different times and places. In Locke’s words, ‘as far as this consciousness can be extended backwards to any past action or thought, so far reaches the identity of that person’.21 Locke offers his own solution to what has come to be known as the ‘brain transplant’ problem, which he explains in terms of the transfer of souls. In Locke’s fanciful illustration, the soul of a prince is transferred into the body of a cobbler, the cobbler’s soul having been removed. We would say that what was left was the same person as the prince, since he would have the thoughts of the prince and the consciousness of the prince’s past life. However, we would at this point be forced to acknowledge the distinction Locke is drawing between ‘man’ and ‘person’, by saying that this was the same man as the cobbler, as demonstrated by the continuity of bodily functioning.

1403_902976_03_cha02.qxd 12/13/04 6:25 PM Page 24

24 Paul Grice, Philosopher and Linguist

There is one fairly obvious problem with an account of identity dependent on the extension back in time of a single consciousness. Locke in effect dismisses this problem by arguing that it is an error arising from a particular way in which language is generally used. The problem is concerned with memory loss. If he were completely and irrevocably to lose his memory, Locke suggests, we would still want to say that he was the same person as the one who performed the actions he has now forgotten. Yet on Locke’s own account, if his consciousness could no longer be extended back to these past actions, he could not be the same person as the one who performed them. Locke suggests that our reluctance to accept this conclusion, our inclination to insist that he is still the same person, can be traced to a lack of reflection about the use of the pronoun ‘I’. In this case, the amnesiac Locke would in fact be using ‘I’ to refer not to the person, but to the man. Locke does not himself offer any examples, but the following illustrates his point. If, having lost his memory and then been instructed about his own past life, he were to state ‘I was exiled by James II’, he would be using the pronoun to refer only to the same man, or living creature, as he is now, not to the same person. Our tendency to maintain that ‘I’ must in these examples stand for the same person is because of our failure to distinguish properly between ‘man’ and ‘person’: our habit of conflating the two. Once this conflation is recognised for what it is, the problem can be resolved; it may be possible for a single man to be more than one person during his life. Indeed, although our use of the pronoun sometimes seems to miss this, elsewhere in the language it is apparently acknowledged: when we say such a one is not himself, or is beside himself; in which phrases it is insinuated, as if those who now, or at least first used them, thought that self was changed; the selfsame person was no longer in that man.22 Grice’s account of personal identity relies on Locke’s theory, but draws more directly on two later philosophers. Thomas Reid, writing about a hundred years after Locke, produces a ‘problem example’. Ian Gallie, just a few years before Grice, offers a more developed discussion of the use of the first person pronoun. Reid points out that the notion of consciousness extending back in time can only make sense if it is understood as memory; Locke’s account can only be coherently understood as a memory theory of identity. Further, he cautions that although the expressions ‘consciousness’ and ‘memory’ are sometimes used inter-

1403_902976_03_cha02.qxd 12/13/04 6:25 PM Page 25

Philosophical Influences 25

changeably in normal speech, it is important for philosophers not to confuse the two. ‘The faculties of consciousness and memory are chiefly distinguished by this, that the first is an immediate knowledge of the present, the second an immediate knowledge of the past.’23 But if personal identity is dependent on knowledge of the past or on memory, the following case, which has become known as the ‘brave officer’ example, presents severe difficulties. A boy is whipped for stealing apples. When the boy grows up, he joins the army and, as a young officer, captures an enemy standard. In later life, the officer is made a general. It is perfectly possible that, when rescuing the standard, the officer was able to remember, even did actively remember, that he was whipped as a child. It is also possible that the elderly general can remember clearly the time when he captured the standard but, with memory fading, has forgotten all about being whipped as a boy. The memory is lost irrevocably; even if prompted about the incident, he cannot remember it. According to Locke’s account, we would have to say that the young boy was the same person as the brave officer, and the brave officer was the same person as the old general, but that the old general was not the same person as the young boy. Not only does this offend against our intuitive notions of identity, it also runs counter to basic logic. If A is identical to B, and B is identical to C, then it follows that C must be identical to A. In ‘Personal identity’, Grice acknowledges his considerable debt to Gallie’s article ‘Is the self a substance?’, which was published in 1936, also in Mind. It is a revealing choice of inspiration, perhaps reflecting Grice’s growing interest in the style of philosophy being practised by some of his Oxford contemporaries: the analysis of specific linguistic examples as a means of approaching more general philosophical problems. Gallie does not directly discuss the memory theory of identity. He assesses the case for the existence of a ‘self’, a metaphysical entity or substance, remaining constant across a series of temporally different mental experiences. As an enduring ego it forms a mental unit of identity, in addition to any purely physical phenomenon. From the outset, he proposes an account of ‘self’ in terms of ‘a certain class of sentences in which the word “I” (or the word “me”) occurs’.24 Such sentences can, he suggests, be grouped into two or perhaps three sets. The first of these is the set of sentences in which ‘my mind’ could be substituted for ‘I’. Gallie admits that in some cases the result may be ‘unusual English’, but maintains that it is enough that they would not be false. Examples include ‘I feel depressed’ and ‘I see it is raining’. The second set of sentences is defined negatively; such substitution is impossible in examples

1403_902976_03_cha02.qxd 12/13/04 6:25 PM Page 26

26 Paul Grice, Philosopher and Linguist

such as ‘I am under 6 feet in height’ and ‘I had dinner at 8 o’clock tonight’. Gallie acknowledges but dismisses from discussion a third but purely philosophical use of ‘I’; expressions such as ‘I hear a noise now’ and ‘I see a white expanse now’ are concerned with facts of which the speaker has introspective knowledge, rather than with any claims about properties of longer duration. Gallie’s starting point is the contention that examples from his first set offer evidence that the self must exist. If we use ‘I’ in these cases, and use it legitimately, there must be something to which we are referring; there must be a mental property that endures across a range of temporarily distinct experiences. Grice’s article also starts with a discussion of ‘I’ sentences, but his analysis of them is rather more subtle than Gallie’s, and his claims about their significance more ambitious. He divides Gallie’s two basic categories into three, although he implicitly dismisses the idea that Gallie’s ‘philosophical’ uses constitute a separate class. His starting point is not with minds but with bodies; the easiest sentences to define are those into which the phrase ‘my body’ can be substituted, sentences such as ‘I was hit by a golf ball’ and ‘I fell down the cellar steps’. A second class, one not distinguished by Gallie, seems to require something extra as well as a physical body to be involved. This includes sentences such as ‘I played cricket yesterday’ and ‘I shall be fighting soon’, in which substituting ‘my body’ for ‘I’ does not provide an exact or full paraphrase. These sentences in turn are to be distinguished from a further class in which substitution of ‘my body’ is even less satisfactory, sentences such as ‘I am hearing a noise’ and ‘I am thinking about the immortality of the soul’. Grice is not explicit about what would count as a suitable substitute for ‘I’ in these sentences. By implication, the ‘I’ of bodily identity is easier to paraphrase than the ‘I’ of mental identity, and Grice is not prepared as Gallie was to offer a paraphrase into ‘unusual English’. But he does suggest tentatively that, in the cricket and fighting sentences, reference to the physical body would need to be supplemented with ‘something about the sort of thoughts and intentions and decisions I had.’25 Grice’s ambitious claim is that the dispute about personal identity is really a disagreement about the status of certain ‘I’ sentences. Examples not concerned with bodily identity, sentences of the ‘I am hearing a noise’ category, are central to the type of identity philosophers have generally discussed. Philosophers over the centuries who have addressed the nature of personal identity have been concerned with the correct analysis of this type of ‘I’ sentence, whether they knew it or not. Moreover, Grice argues, they have also been concerned with the correct

1403_902976_03_cha02.qxd 12/13/04 6:25 PM Page 27

Philosophical Influences 27

analysis of a class of sentences closely related to these but containing the word ‘someone’ instead of the word ‘I’. An adequate account of personal identity must be able to offer an analysis both of ‘I am hearing a noise’ and of ‘someone is hearing a noise’. Grice’s theory of personal identity is based on Locke’s account, but attempts a modification to answer its critics. His analysis of the relevant ‘I’ and ‘someone’ sentences is in terms of a particular relationship between experiences divided by time and place. A series of such experiences can be said to belong to the same person. The relationship in question cannot be a simple one of memory, or possible memory, because this raises the ‘brave officer’ problem; a person having a particular experience may be quite incapable of remembering an earlier experience, but yet be the same person as the person who had that experience. As a maneuver to avoid this problem, Grice introduces the phrase ‘total temporary state’ (t.t.s.), as a term of art to describe all the experiences to which an individual is subject at any one moment. These t.t.s.’s can occur in series, and describing a person means describing what it is that makes any particular series the t.t.s.’s of one and the same person. Grice’s suggestion is that every t.t.s. in a ‘person’ series contains or could contain an element of memory relating it to some other t.t.s. earlier in the same series. In this way, the ‘brave officer’ problem is avoided. The t.t.s of the general contains a memory trace of the experience of the brave officer, and the t.t.s. of the brave officer contains a memory trace of the experience of the boy. The three t.t.s.’s are linked into a series and it does not matter that there is no direct link between that of the general and that of the boy. Grice analyses ‘someone hears a noise’, the example type he identified as central to personal identity, in terms of his notion of t.t.s.: a (past) hearing of a noise is an element in a t.t.s. which is a member of a series of t.t.s.’s such that every member of the series either would, given certain conditions, contain as an element a memory of some experience which is an element in some previous member, or contains as an element some experience a memory of which would, given certain conditions, occur as an element in some subsequent member; there being no subset of members which is independent from all the rest.26 Each t.t.s. contains at least one memory link, or potential memory link, either forward or backward in an appropriate series. Sentences containing the problematic ‘I’ and ‘someone’, the ‘I’ and ‘someone’ that cannot

1403_902976_03_cha02.qxd 12/13/04 6:25 PM Page 28

28 Paul Grice, Philosopher and Linguist

be paraphrased in terms of bodies, can be re-expressed in terms that remove the problem word and replace it with a suitably defined series of t.t.s.’s. Personal identity is shown to be a construction out of a series of experiences. Grice considers various possible objections that might be raised to his analysis, including the objection that it is just too complicated as an account of a set of apparently simple sentences. He is himself unsure that this is a particularly damning objection, but argues that his analysis of ‘self’-sentences is in any case: probably far less complicated than would be the phenomenalist’s analysis of any material object-sentence, if indeed a phenomenalist were ever to offer an analysis of such a sentence, and not merely tell us what sort of an analysis it would be if he did give it.27 This is an interesting dig. Phenomenalism was most closely associated at this time with the logical positivists, and therefore at Oxford with Ayer. It was a form of scepticism, holding that we do not have direct evidence of material objects; we have evidence only of various ‘sense data’, and infer the existence of material objects from these. In other words, phenomenalists claimed that statements about material objects need to be translatable into empirically verifiable statements about sense data. Grice’s objection is that the appropriate translations were never actually offered, merely discussed; his complicated analysis of ‘someone is hearing a noise’, on the other hand, offered something definite. This is closely related to the type of objection Austin was raising to Ayer’s method; his theories introduced technical terms such as ‘sense data’, but did not explain them in ordinary language. In the light of Grice’s sideswipe at sense data theories, it is perhaps surprising that John Perry, in his commentary, should find phenomenalist tendencies in Grice’s account of personal identity. His reason lies in the analytic approach Grice shares with phenomenalism. Both were concerned with analysing sentences that contain a problem expression (for the phenomenalists words referring to material objects, for Grice the relevant uses of ‘I’ and ‘someone’) in such a way that these expressions are removed and empirically observable phenomena are substituted. Sense data and total temporary states are both available to introspection. For phenomenalists, material objects are logical constructions based on the evidence of sense data; they need to be recognised as such and the ‘true’ form of the sentences expressing them needs to be understood. For Grice, persons are logical constructions from

1403_902976_03_cha02.qxd 12/13/04 6:25 PM Page 29

Philosophical Influences 29

experiences. Perry does, however, acknowledge an important difference between Grice’s enterprise and that of other analytic philosophers, of whom he takes Russell to be a prime example: In Russell’s view, the logical construction was the philosopher’s contribution to an improved conception of, say, a material object, free of the epistemological problems inherent in the ordinary conception. So analysis, for Russell, does not preserve exact meaning. But Grice intends to be making explicit, through analysis, the concept we already have.28 In other words, Grice sees ordinary ways of talking as adequate and valuable, if in need of clarification. For analytic philosophers such as Russell, ordinary language is simply not good enough; it needs to be purged of inappropriate existential commitments and vague terms before it can be a fit tool for philosophy or science. Grice’s account of personal identity has been generally well received in its field. Despite his reservations, Perry describes it as ‘the most subtle and successful’ attempt to rescue Locke’s memory theory.29 Similarly, Timothy Williamson cites Grice’s paper as the first in a succession of responses to the ‘brave officer’ problem in terms of a transitive relation between a set of spatio-temporal locations.30 In many ways it seems far removed from the work for which Grice is now best known. And indeed when he returned to philosophy after the war he was concerned with rather different topics. But this early article shows some of the traits that were to become characteristic of his work across a range of subjects, in particular a close attention to the nuances of language use. One theme in particular, the notion of what it is to be a person, and Locke’s idea of an ontological distinction between ‘man’ and ‘person’ was to prove central to much of Grice’s later philosophy. The publication of ‘Personal identity’ meant that Grice’s professional life was not entirely put on hold during the war years, and neither was his personal life. In 1943 he married Kathleen Watson. Kathleen was from London, but they had met through an Oxford connection. Her brother J. S. Watson had held a Harmsworth senior scholarship shortly after Grice and the two had become friends. James married during the war and, when his best man was killed on active service shortly before the ceremony, called on Paul at short notice to perform the duty. Paul and Kathleen met at the wedding. At the time of his own wedding, Paul still had two years of war service ahead of him, but it meant that when he returned permanently to Oxford he would not be living in St John’s,

1403_902976_03_cha02.qxd 12/13/04 6:25 PM Page 30

30 Paul Grice, Philosopher and Linguist

as he had done briefly as a bachelor. There were no rooms for married Fellows in college; indeed, women were not permitted on the premises, even at formal dinners. Paul and Kathleen settled in a flat on the Woodstock Road rented from his college. His Fellowship at St John’s was, of course, still open to him, as was the prospect of a closer involvement with the new style of philosophy taking shape in Oxford.

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 31

3 Post-War Oxford

The intellectual atmosphere of Oxford philosophy during the decade or so from 1945 was very different from that which Grice had known before the Second World War. As students and dons alike returned from war service, the process of change that had begun with a few young philosophers during the 1930s picked up momentum. Old orthodoxies and methods were overthrown as a host of new thinkers and new ideas took their place. The style of study was questioning, exploratory and cooperative. Many of the philosophers who were active in this period, such as Geoffrey Warnock and Peter Strawson, have since testified to the spirit of optimism, enthusiasm and sheer excitement that predominated.1 The reasons for these emotions were similar to those that had drawn A. J. Ayer to the work of the Vienna Circle just over a decade earlier. A new style of philosophy was going to ‘solve’ many of the old problems. Once again, this was to be achieved by close attention to and analysis of language. For logical positivists this involved ‘translating’ problematic statements of everyday language into logically rigorous, empirically verifiable sentences. For the new Oxford philosophers, however, no such translations would be necessary. A suitably rigorous attention to the facts of language was going to be a sufficient, indeed the only suitable, philosophical tool. In retrospect, Grice suggested the following causes of the differences after the war: the dramatic rise in the influence of Austin, the rapid growth of Oxford as a world-centre of philosophy (due largely to the efforts of Ryle), and the extraordinarily high quality of the many young philosophers who at that time first appeared on the Oxford scene.2 31

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 32

32 Paul Grice, Philosopher and Linguist

Gilbert Ryle, the tutor who had introduced Ayer to Wittgenstein and encouraged him to travel to Vienna, was a decade or more older than Grice and his contemporaries, having been born in 1900. However, his influence on Oxford philosophy during the 1940s and 1950s was considerable. His philosophy of mind, in particular, was widely known through his teaching, and was eventually published as The Concept of Mind in 1949. In this book, Ryle argues against the received idea of the mind as a mental entity separate from, but in some way inhabiting, the physical body: Descartes’s ‘ghost in the machine’. All that can legitimately be discussed are dispositions to act in a certain way in certain situations. The error of dualism can be attributed to the tendency to analyse minds as ‘things’ just as bodies are things. Taken together with a strict distinction between mental and physical, this gives rise to the error of defining minds as ‘non-physical things’. This error stems from a basic category mistake involving the type of vocabulary used in discussions of minds. Differences between the physical and the mental have been ‘represented as differences inside the common framework of the categories of “thing”, “stuff”, “attribute”, “state”, “process”, “change”, “cause” and “effect” ’.3 Only close attention to language reveals this tendency among philosophers to extend erroneously terms applicable to physical phenomena, and thus to enter into unproductive metaphysical commitments. Ryle was keen to see this philosophical approach adopted more widely, and after he took over as editor of Mind in 1947, began what Jonathan Rée has described as a ‘systematic campaign’ to take control of English philosophy. He drew together some of the more promising young Oxford philosophers and ‘by galvanising them into writing, especially about each other, in the pages of Mind, he gave English academic philosophy in the fifties an energy and sense of purpose such as it has never seen before or since.’4 The style of philosophy Ryle was deliberately promoting found resonances in the work of J. L. Austin, who was continuing to develop the notions about ordinary language with which he had confronted Ayer before the war. Ayer himself had returned to Oxford only briefly, before taking up a chair at University College London in 1946. The ideas of logical positivism were, in any case, losing their earlier glamour and appeal for younger philosophers, and Austin became the natural leader of the next wave of Oxford philosophy, both before and after his appointment as White’s Professor of Moral Philosophy in 1952. This promotion may seem surprising today; Austin had published just three papers by 1952, none of which were strictly about moral philosophy. It seems that credit was given to the importance of his ideas, and his rep-

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 33

Post-War Oxford 33

utation as an inspiring and charismatic teacher, a reputation achieved despite his rather austere personality. Few if any of his colleagues felt that they really got to know him personally. This was perhaps in part because, as Geoffrey Warnock has suggested, he was ‘a shy man, wary of self-revelation and more than uninviting of self-revelation by others’.5 It may also have been in part because, unusually among his peers, he eschewed college social life, preferring to spend his time quietly at home with his family in the Oxfordshire countryside. Nevertheless, most people who knew him had their favourite story to sum up what could be a formidable and confrontational, but also a humourous and humane personality. George Pitcher relates how Austin once attended a meeting at his college concerned with plans for the college buildings. The meeting got embroiled in a long and seemingly unreconcilable discussion about the President’s Lodgings. When finally asked for his opinion, Austin’s response was ‘Mr President, raze it to the ground.’6 Grice recalls how a colleague once challenged Austin with Donne’s lines ‘From the round earth’s imagined corners,/ Angels your trumpets blow’ as an example of non-understandable English. ‘It is perfectly clear what it means,’ replied Austin, ‘it means “Angels, blow your trumpets from what persons less cautious than I would call the four corners of the earth”.’7 Such anecdotes illustrate Austin’s ability to cut through protocol and rhetoric and reveal the plain facts underneath: often to reveal that things are just as simple as they seem, and a lot more simple than they are made out to be. This was the spirit in which he approached philosophy. He was undoubtedly lucky in this project in the number of dedicated and talented young philosophers who, as Grice puts it, ‘appeared’ at Oxford in the 1940s and 1950s. In the early post-war years, undergraduates and new tutors were generally older than had been the tradition, and eager to study, their university careers having been postponed or interrupted by the war. Austin played an active role in spotting and encouraging talent, building up what quickly came to be seen as his own ‘school’ of philosophy. This development seems rather at odds with his views on discipleship. Ben Rogers has suggested that, unlike Wittgenstein, Austin ‘discouraged anything like a cult of personality: he wanted to put philosophy on a collective footing’.8 However, there is some evidence that Austin was at least in part predisposed to encourage, and even to enjoy, his status as leader. Grice reports that Austin was once heard to remark to a colleague, ‘if they don’t want to follow me, whom do they want to follow?’9

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 34

34 Paul Grice, Philosopher and Linguist

The style of philosophy developed and practised by Austin and his followers has been variously labelled as ‘Oxford philosophy’, ‘linguistic philosophy’ and ‘ordinary language philosophy’. It is in fact far from uncontroversial that a single, identifiable approach united the philosophers working in Oxford at this time, even those in Austin’s immediate circle. Grice himself denies this assumption in a number of published commentaries, such as the uncompromising claim that: ‘there was no “School”; there were no dogmas which united us.’10 Perhaps the only common position of the Oxford philosophers, certainly the only one Grice acknowledges, was a belief in the value, indeed the necessity, of the rigorous analysis of the everyday use of words and expressions. They worked on a wide range of philosophical problems, and even when addressing the same issue they often disagreed on analysis or conclusion. But in approaching these problems they were in agreement that careful attention to the words in which they were conventionally expressed was the best path to understanding, tackling and perhaps eliminating the problems themselves. The origins of this approach are controversial. Commentators have looked beyond Ryle and Austin to suggest Ludwig Wittgenstein, G. E. Moore, Bertrand Russell and even Gotlob Frege as candidates for the title ‘father’ of ordinary language philosophy. There is certainly something in all these claims; each thinker was influential in the development of the analytic tradition from which ordinary language philosophy emerged. But under Ryle’s and Austin’s guidance it developed in directions not envisaged, and in some cases explicitly not approved, by its putative mentors. In the late nineteenth century Frege proposed analysis of language as a method for tackling philosophical problems. He was chiefly a mathematician, but in the course of his work addressed the language of logical expressions, and the potential problems inherent in it. Frege’s most influential contribution to the philosophy of language is his distinction between ‘sense’ and ‘reference’. Any name, whether a proper name or a description, that points to some individual in the world, does so by way of a particular means of identifying that individual. A name therefore has both a reference, the individual it denotes, and a sense, the means by which reference is indicated. In this way, Frege explains statements of identity. If we attend only to reference, an example such as ‘Sophroniscus is the father of Socrates’ would have to be dismissed as uninformative; it would simply offer a logical tautology, stating of an individual that he is identical with himself. If, however, we attend to sense as well as reference, we can analyse this example as making an informative statement to the effect that the two names may differ in

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 35

Post-War Oxford 35

sense but share the same reference. The example is significant because, although there is no difference between the denotation of the two names, there is nevertheless ‘a difference in the mode of presentation of the thing designated’.11 Russell proposed the ‘translation’ account that became characteristic of analytic philosophy. Suitably rigorous attention could reveal the logical structure beneath a grammatical form that was often misleading and imperfect. His main difference from Frege was in his analysis of descriptions. Russell observes in his 1905 article ‘On denoting’ that ‘a phrase may be denoting, and yet not denote anything; for example, “the present king of France” ’.12 For Frege such examples posed no problem. It was perfectly possible to say that such an expression had a sense but no reference. But the consequences of this position are unacceptable to Russell because of their impact on logic. If ‘the king of France’ simply has no reference, then any sentence of which it is subject, such as Russell’s famous example ‘the king of France is bald’, must also fail to refer to the world; in other words it must fail to be either true or false. For Russell this result is intolerable because it dispenses with classical, two-valued logic, in which every proposition must be capable of being judged either ‘true’ or ‘false’. Classical logic states that if a proposition is true then its negation must be false, and vice versa. But under Frege’s analysis ‘the king of France is bald’ and ‘the king of France is not bald’ are indistinguishable in terms of truth value; both fail to be either true or false. Not only is this unacceptable, argues Russell, it also goes against the facts of the matter; ‘the king of France is bald’ is a simple falsehood. Russell uses the label ‘definite descriptions’ for phrases such as ‘the king of France’, and claims that they are a particular type of expression demanding a particular type of analysis. Despite appearances, they are not denoting expressions, and they do not function as subjects of sentences in which they superficially appear to be in subject position. The subject/predicate form of a sentence containing a definite description, a sentence such as ‘the king of France is bald’, is misleading. That is, the logical structure of the sentence is at odds with its purely grammatical form; it is actually a complex of propositions relating to the existence, the uniqueness, and then the characteristics, of the individual apparently identified. Russell’s rendition of this example can be paraphrased as: ‘there exists one entity which is the king of France, and that entity is unique, and that entity is bald.’ Re-analysed in this way, it is possible to demonstrate that, in the absence of a unique king of France, one of these propositions is simply false. If it is not true that

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 36

36 Paul Grice, Philosopher and Linguist

there is a present king of France, the first part of the logical form is false, making the sentence as a whole also false. Russell’s theory of descriptions has been seen as the defining example of analytic philosophy. Ayer often repeated his admiration for it in these terms. His close attention to language as well as to logic further caused Russell to be credited with inspiring ordinary language philosophy.13 However, he would certainly not have been pleased by this latter accolade; he was still active in philosophy when the new approach was in the ascendant and he publicly and vociferously opposed it.14 His own motivation in ‘On denoting’, and elsewhere, was not to describe natural language for its own sake, but to explain away the apparent obstacles it offered to classical logic. If every statement containing a denoting phrase could be explained as a complex proposition not involving a denoting phrase as a constituent, bivalent logic could be maintained. Like Frege, Russell’s first and primary philosophical interest was in mathematical logic. His interest in the analysis of language and therefore in meaning grew out of a concern for analysing the language in which the ideas of logic are expressed, and ultimately for refining that language. Russell’s Cambridge colleague and almost exact contemporary G. E. Moore would perhaps have been less uncomfortable with being credited as genitor of ordinary language philosophy. Like many of the Oxford philosophers a generation later, he came to philosophy from a background in classics, and remained sensitive to details of meaning. This was coupled with a rather leisured approach to philosophical enquiry; a private fortune enabled him to spend seven years at the start of his career away from all professional responsibilities, pursuing his own philosophical interests. Certainly, the result was a thorough and meticulous attention to the language in which philosophical discussion is couched. Another common feature between Moore and many ordinary language philosophers, one identified by Grice, was an antimetaphysical, determinedly ‘common-sense’ approach to philosophical issues.15 An illustration of both the rigorous analysis and the confidence in intuitive understanding can be found in Moore’s article ‘A defence of common sense’, published in 1925. This was a forerunner of the infamous British Academy lecture of 1939 in which he held up his own hand as incontrovertible and sufficient ‘proof’ of the existence of material objects. In the published article, too, Moore addresses the question of the status of our knowledge of external objects, as well as our belief in the existence of other human beings, and indeed the viability of any

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 37

Post-War Oxford 37

physical reality independent of mental fact. In asserting such viability Moore is, as he explains, arguing against the sceptical view that all that we have access to are ‘ideas’ of objects, offered to us by means of our sense impressions, never the objects themselves. Such philosophical views were dominant in English philosophy when Moore himself was a young man, and were in part what he was reacting against in defending common sense. Moore’s ‘defence’ rests almost entirely on his assertion of what he knows to be the case, by introspection based on the dictates of common sense. He knows with certainty of the existence and continuity of his own body, as well as its distinction from other, equally real physical objects. Not only that, but all other human beings have a similar, but distinct set of knowledge relating to their own bodies. Further, to describe these as common-sense beliefs is actually to admit that they must be true, since the very expression ‘common sense’ implies that there is a set of other minds holding this belief and therefore that other minds, and other people, exist. Common sense should extend to analysis of the language of philosophy as well. In assessing the truth of a sentence such as ‘The earth has existed for many years past’, it is necessary to consider only ‘the ordinary or popular meaning of such expressions’. Moore comments dryly that the existence of such a unique, ordinary meaning ‘is an assumption which some philosophers are capable of disputing’.16 This capacity, he suggests, has led to the erroneous argument that a statement such as this is questionable, and perhaps even false. This insistence that philosophers attend to the ordinary use of the expressions with which they are dealing, together with the expression of bafflement that they should declare themselves unsure of the nature of such ordinary use, was to become as characteristic of Austin’s work as it was of Moore’s. Austin was widely known to be impressed by Moore’s work, and inspired by his methodology, although he was less convinced that intuitive convictions and commonplace orthodoxies could be accepted at face value without further exploration. Grice recalls him declaring, with a characteristic combination of scholarly enthusiasm and school-boy jocularity: ‘Some like Witters, but Moore’s my man.’17 This pithy summary of his own philosophical influences introduces one of the most controversial issues concerning the origins of ordinary language philosophy: the impact of Wittgenstein. Some commentators have argued that Wittgenstein’s later work was crucially important; Ernest Gellner, for instance, discusses ‘linguistic philosophy’ as if it were uncontroversially a product of Philosophical Investigations.18 But in his biographical sketch of Austin, Geoffrey Warnock argues cat-

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 38

38 Paul Grice, Philosopher and Linguist

egorically that Wittgenstein did not influence the philosophical method Austin promoted.19 The issue is a difficult one to settle. Ben Rogers suggests that: ‘Oxford philosophy of the 1940s and 1950s was more indebted to Wittgenstein than it liked to admit,’20 and it is certainly possible that his ideas informed discussion at the time, even if his intellectual presence was not directly acknowledged. Philosophical Investigations, presenting Wittgenstein’s later work on language, was not published until 1953, but earlier versions of it were circulated and read in Oxford in the immediate post-war years.21 Other commentators, such as Williams and Montefiore, emphasise what they see as the striking differences between Wittgenstein and Austin. The two men’s personalities, they suggest, were reflected in their different styles of philosophy: Wittgenstein’s intense and introspective, Austin’s scholarly and meticulous.22 Wittgenstein’s early work had been inspired by his former tutor Russell. Although highly original, it was very much within Russell’s formal, analytic tradition. Language operates by expressing propositions that may differ from apparent surface form but map on to the world by means of the basic properties of truth and falsity. ‘To understand a proposition means to know what is the case if it is true,’ Witttgenstein suggests in his Tractatus, prefiguring the logical positivism of the Vienna Circle.23 However, the work Wittgenstein produced when he returned to philosophy after a self-imposed break was notoriously different from that of his youth. It shows some striking similarities to ordinary language philosophy. In Philosophical Investigations, Wittgenstein argues that language is better defined in terms of the uses to which it is put than the situations in the world it describes. Words are tools for use and, just as a tool may not always be used for the same function, meaning is not necessarily constant. Coining one of the most enduring metaphors in the philosophy of language, he argues that a word is best described as having a loose association of meanings that are essentially independent but may display certain ‘family resemblances’. He argues that apparent philosophical problems are in fact issues of language use; correct investigation of the language in which problems are posed reveals them to be merely ‘pseudo problems’. Also, and perhaps even more strikingly, he comments that, in philosophical discussion as elsewhere, ‘When I talk about language (words, sentences, etc.) I must speak the language of every day.’24 Whatever positive influences may have acted on Austin, it is certain that he was at least in part reacting against some other philosophical styles. In particular, he was taking issue with logical positivism, and with

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 39

Post-War Oxford 39

the work of his old sparring partner A. J. Ayer. Soon after the war, Austin developed his response to Ayer on sense data. This was presented as a series of lectures, first delivered in 1947, and was published only posthumously, when Geoffrey Warnock edited the lecture notes under the title Sense and Sensibilia. Austin deals in detail and critically with Ayer’s Foundations of Empirical Knowledge, the book published in 1940 that expanded on and developed the implications of the verificationist account presented in Language, Truth and Logic. Ayer had argued in the earlier book that the operation of the principle of verification for an individual is not directly dependent on the physical world, but relies on the data we receive through our senses. These ‘sense data’ are the only evidence we have available by which to subject sentences making statements about the world to appropriate processes of verification. In The Foundations of Empirical Knowledge, Ayer addresses the question of whether our knowledge is determined by material objects themselves or by our personal sense data. He develops a phenomenalist approach to this question, arguing that beliefs and statements about material objects are only legitimate if they are translated into beliefs and statements about our sense data. In this way, he maintains that discussion of the status of material objects is primarily a dispute about language. It is a dispute about the best way to analyse statements of experience. Sentences concerned with material objects and those concerned with sense data are simply uses of two different types of language. They refer to the same things in two different ways, rather than describing ontologically different phenomena.25 Austin disliked the idea that philosophers need to ‘see through’ ordinary language before they can say anything rigorous about material objects. In his lectures on perception he argues that belief in sense data is an error into which philosophers have been led by the words they coin. The term itself is introduced artificially; ordinary people in everyday life find no need of it. Even the word ‘perception’ has an unnecessarily technical ring for Austin. People rarely talk about perceiving material objects, and still less frequently of experiencing sense data. They do talk about seeing things, and so philosophers should attend to the suggestions this offers as to how the world works. The ‘plain man’ can cope perfectly well with distinguishing appearance from reality, with describing rainbows, or with commenting that ships on the horizon may appear closer than they are. Philosophers need only look to the actual use of words such as ‘reality’, ‘looks’ and ‘seems’ to understand this. Ordinary language has its own, perfectly adequate way of dealing with the difference between appearance and reality. Near the

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 40

40 Paul Grice, Philosopher and Linguist

beginning of his lectures, Austin sums up his response to sense data theorists: My general opinion about this doctrine is that it is a typically scholastic view, attributable, first, to an obsession with a few particular words, the uses of which are over-simplified, not really understood or carefully studied or correctly described . . . Ordinary words are much subtler in their uses, and mark many more distinctions, than philosophers have realised.26 In other work, Austin objects to the assumption he finds in logical positivism and elsewhere that the most important or the only philosophically interesting function of language is to make statements about the world. Philosophers concerned with questions of truth and falsity, and with determining meaning in terms of verification, overlook the many different ways in which speakers actually use language, and fall into the fallacy of seeing it as merely a tool for description. ‘The principle of Logic that “Every proposition must be true or false” has too long operated as the simplest, most persuasive and most pervasive form of the descriptive fallacy.’27 Only a careful consideration of the way in which people actually use language enables philosophers to dispense with this fallacy. Even when people do make statements of fact, it is often necessary to consider ‘degrees’ of success, rather than simple truth. It is only with such a concept that we can make sense of various loose ways of talking, such as ‘the galaxy is the shape of a fried egg’, or explain why a statement may be true, but somehow ‘inept’, such as saying ‘all the signs of bread’ when clearly in the presence of bread. Austin’s approach, then, was based on a distrust of philosophical constructions and technical terminology, especially terminology he suspected of being unnecessary or poorly defined, but it was not entirely negative. He wanted to replace these with a serious and rigorous attention to the way in which any matter of philosophical interest is discussed in everyday language. He sets out these ideas particularly clearly and characteristically stridently in ‘A plea for excuses’, a paper delivered to the Aristotelian Society in 1956, which came to be seen almost as a manifesto for ordinary language philosophy. Philosophers have for too long conducted their discussion in real or feigned ignorance of the way in which key words and phrases are used in ordinary, everyday language. It is not just that the language of everyday is a worthy subject and adequate vehicle for philosophical enquiry. Rather, it is the only tool with a claim to being sufficient for philosophical purposes, honed

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 41

Post-War Oxford 41

as it is by generations of use to describe the distinctions and connections human beings have found necessary. These can reasonably be claimed to be more sound and subtle ‘than any that you or I are likely to think up in our arm-chairs of an afternoon – the most favoured alternative method’.28 Meanings, including the meanings of philosophical terms, can only coherently be discussed with reference to the ways in which words are used in ordinary discourse. More specifically in ‘A plea for excuses’, Austin is concerned with a rigorous examination of the circumstances in which excuses are offered, and the different forms they may take. For instance, excuses are only offered for actions performed freely. A consideration of the ordinary use of ‘freely’ to describe an action reveals that it does not introduce any particular property, but serves to negate some opposite, such as ‘under duress’. It would be used only when there is some suggestion that the opposite could or might apply. There would be something strange about applying the term ‘freely’ to a normal action performed in a normal manner. Austin sums up this claim in the slogan ‘no modification without aberration’.29 For many ordinary uses of many verbs, there is simply no modifying expression, either positive or negative that can appropriately and informatively be applied. The ‘natural economy of language’ dictates that we can only add a modifying expression ‘if we do the action named in some special way or circumstance’.30 Austin promoted his approach to philosophy through personal contact, including his own teaching, at least as much as through published papers such as ‘A plea for excuses’. Lynd Forguson has studied undergraduate Oxford philosophy examination papers and observed that questions reflecting ordinary language philosophy began to appear as early as 1947, and were increasingly frequent throughout the 1950s. For instance, students were asked: ‘Is it essential to begin by deciding the meaning of the word “good” before going on to decide what things are good?’31 Perhaps most significantly Austin’s ideas were disseminated through active collaboration in philosophy with his colleagues. This did not take the form of any co-authored papers or books. Indeed, both Grice and Stuart Hampshire have suggested that Austin’s independent character and idiosyncratic style would have made him an awkward writing companion.32 His approach to collaboration in philosophy was to draw together the responses and insights of as many different informed observers of language use as possible. In the late 1940s, he instituted a series of ‘Saturday Mornings’, meetings that ran in term time and brought together a number of like-minded Oxford philosophers.

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 42

42 Paul Grice, Philosopher and Linguist

Grice was present from the start. Other members included R. M. Hare, Herbert Hart, David Pears, Geoffrey Warnock and Peter Strawson. Grice recalls how Austin described these meetings, perhaps semi-seriously, as a weekend break from the regular work of ‘philosophical hacks’, enabling them to turn their attention to less narrowly philosophical topics.33 Perhaps with this explanation in mind, Grice labelled the meetings ‘The Play Group’. He seems to have been inordinately pleased with this name, and employed it in both spoken and written accounts of Oxford philosophy throughout his life. He used it at the time with some of his colleagues but never, it seems, with Austin himself. Austin was not the sort of man with whom to share such a joke.34 The American logician W. V. O. Quine spent the academic year 1953–4 at Oxford as visiting Eastman professor. In his memoir of this period he passes on the rumour that attendance at Saturday Mornings was controlled by rules devised by Austin to preclude particular individuals.35 There does not seem to be any explicit first-hand support from members of the group for this allegation. However, there is some evidence from them that Austin imposed criteria for attendance that would, in effect, have ruled out anyone older than himself who might prove a challenge to his leadership. Geoffrey Warnock remarks simply that attendance was ‘restricted to persons both junior to Austin and employed as whole-time tutorial Fellows’.36 Many years later, when Grice was thinking back to the days of the Play Group, he jotted down a list of Oxford philosophers of the time. Interestingly, he divides his list into three categories: ‘yes’ (Austin, Grice, Hampshire, Strawson, Warnock . . .), ‘no’ (Anscombe, Dummett, Murdoch . . .) and ‘overage’ (Ryle, Hardie . . .).37 Initially, at least, Austin seems to have stuck to his self-imposed remit. The Play Group engaged not so much in philosophical debate as in intellectual puzzles and exercises pursued for their own sake. A favourite occupation was to analyse and categorise the rules of games, or to invent games of their own. Kathleen Grice recalls that one term they seemed to spend most of their time drawing dots on pieces of paper. Another term they decided to ‘learn Eskimo’.38 However, they gradually turned their attention to more obviously philosophical occupations. Austin liked to pick a text per term for the group to read and discuss. However, some of the group did not find this a particularly fruitful philosophical method, largely because of Austin’s preferred pace, which was painstakingly slow. ‘Austin’s favoured unit of discussion in such cases was the sentence’, Warnock recalls, ‘not the paragraph or chapter, still less the book as a whole.’39 At other times, the Play Group put into practice

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 43

Post-War Oxford 43

Austin’s views about the correct business of philosophers by engaging in ‘linguistic botanising’. Austin admired the sciences, and regretted that his own education had given him little real knowledge or expertise in them. His idea was to apply the same techniques of rigorous observation and attention to detail as would be used in a discipline such as botany to the material of linguistic use. The language, and the way in which it was ordinarily used, were there as empirical facts to be observed by those with the sensitivity to do so. Only in this way could its categories and fine distinctions, honed over generations, become available to the philosopher. Examples of this philosophical method are to be found in many of Austin’s own writings. For instance, in ‘A plea for excuses’, where he advocates philosophical attention to ‘what we should say when, and so why and what we should mean by it’, he suggests some of the words and phrases related to excuses that may shed light on its use: abstract nouns such as ‘misconception’, ‘accident’, ‘purpose’ and verbs such as ‘couldn’t help’, ‘didn’t mean to’, ‘didn’t realise’.40 He proposes a philosophical methodology of ‘going through the dictionary’; only such a rigorous analysis of the language will give a thorough and objective overview of the subject matter, which can then be subjected to a process of introspective analysis. At the Play Group, linguistic botanising was no less painstaking, but it took a discursive, collaborative form. The members would, in effect, pool their linguistic resources in order to draw up lists of words related to the particular subject under discussion. They would then analyse the uses and nuances of these words, deciding which were suitable, and which unsuitable, in various different contexts. As Grice later explained it, they would examine a wide range of the relevant vocabulary, to ‘get a list of “okes and nokes” ’.41 J. O. Urmson has described the processes involved more fully: Having collected in terms and idioms, the group must then proceed to the second stage in which, by telling circumstantial stories and constructing dialogues, they give as clear and detailed examples as possible of circumstances under which this idiom is to be preferred to that, and that to this, and of where we should (do) use this term and where that. . . . At [the next] stage we attempt to give general accounts of the various expressions (words, sentences, grammatical forms) under consideration; they will be correct and adequate if they make it clear why what is said in our various stories is or is not felicitous, is possible or impossible.42

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 44

44 Paul Grice, Philosopher and Linguist

Austin argued that this process was both empirical and objective. Several philosophers conferring together would avoid the possible distortions and idiosyncrasies that might skew the findings of a single researcher, as well as bringing together a wide range of different experiences and specialisms. At a conference in 1958, he was challenged to say what set of criteria was adequate for judging a philosophical analysis of a concept. He is reported as replying, ‘If you can make an analysis convince yourself, no matter how hard you try to overthrow it, and if you can get a lot of cantankerous colleagues to agree with it, that’s a pretty good criterion that there is something in it.’43 Their task was not just to list and to categorise areas of vocabulary, but to analyse the conceptual distinctions these indicated. Only after selecting relevant terms and discussing their usage was it legitimate to compare the findings with what philosophers had traditionally said. Despite his frosty demeanour and austere habits, Austin displayed huge enthusiasm and optimism, attitudes that infected the members of the Play Group. Grice has commented that Austin clearly found philosophy tremendous fun, although he ‘would not be vulgar enough to say anything like that’.44 If Austin’s method attracted devotees, however, it also attracted critics. The practitioners of ordinary language philosophy saw it as an exciting new approach capable of overthrowing past orthodoxies and offering solutions to age-old problems, but its critics saw it as sterile and complacent, valuing lexicographical pedantry over real philosophical argument.45 Some of its more high profile opponents, such as Bertrand Russell and A. J. Ayer, targeted what they saw as the shallow insights, even the platitudes, to which the Oxford philosophers limited themselves. Maintaining his position that ordinary language is simply not good enough for rigorous philosophical enquiry, Russell derided what he saw as the triviality of ordinary language philosophy; ‘To discuss endlessly what silly people mean when they say silly things may be amusing but can hardly be important.’46 Other critics targeted the practitioners more directly, characterising them as spoilt and snobbish, ‘playing’ at philosophy without feeling the need to justify their leisurely activities. Ernest Gellner, a former student of Austin’s, published a book in 1959 exclusively concerned with attacking the Oxford philosophers and their method. Gellner finds his own way of repeating Russell’s mockery when he suggests that ordinary language philosophy suggests that ‘the world is just what it seems (and as it seems to an unimaginative man at about mid-morning)’.47 This style of philosophy ‘at long last, provided a philosophic form eminently suitable for gentlemen’ because it demanded immense leisure and ‘yet its message is that everything remains as it

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 45

Post-War Oxford 45

is’.48 C. W. K. Mundle argues that Austin’s elitism mars his philosophy. He detects prescriptivism beneath Austin’s apparently descriptivist concern for ‘what we should say’. ‘In the name of ordinary language, he demands standards of purity and precision of speech which are extraordinarily rare except among men who got a First in Classical Greats.’49 There is undoubtedly some substance in the accusations of elitism. The members of the Play Group shared an understanding that detailed analysis and meticulous enquiry were important ends in themselves, but also a certainty of leisure and means to pursue these ends. Their appointments did not carry the pressure to publish that to modern academics is a matter of course, a fact almost certainly crucial to the success of ordinary language philosophy. Linguistic botanising was often conducted self-consciously for its own sake, without expectations of specific, publishable ‘results’. Stuart Hampshire has admitted that Austin’s method depended on a comparatively careless attitude to time and publication: ‘you’re apt to get discouraged about publishing because you can’t fit it all neatly into a journal article.’50 At the time, the Oxford philosophers were in general reluctant to argue their own case. Partly in response to this, Grice later took on the task of offering a first-hand account of ordinary language philosophy in defence against some of its critics. The subject clearly exercised him from the mid-1970s on; he talked, both formally and informally on the topic, and wrote copious notes about it. In a move that would have seemed to the critics to have confirmed their worst fears about snobbery and self-importance, he compares post-war Oxford to the Academy at Athens, with Austin by implication playing the role of Aristotle. In fact the comparison is not as pompous as it may first appear; Grice is not suggesting that mid-twentieth-century Oxford was as important to the history of philosophy as ancient Athens. Rather, he sees certain similarities between the two enterprises that gave ordinary language philosophy a philosophical pedigree its critics claimed it lacked. He sees reflections of Austin’s method in what he describes as the ‘Athenian dialectic’: the interest in discourse in general, and in the facts of speech in particular. The connection is summed up for Grice in the two aspects of the Athenian interest in ta legomena. The phrase can refer to ‘what is spoken’, both in the sense of ordinary ways of talking, and of generally received opinions. Both schools were concerned not just with the analysis of language, but with seeing the way people speak as a reflection of how they generally understand the world, a reflection of the ‘common-sense’ point of view. The difference is, Grice argues, that Austin and his school concentrated more closely on the first aspect;

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 46

46 Paul Grice, Philosopher and Linguist

Aristotle and his on the second. Aristotle was concerned not so much with methodology as with picking out foundations of truth. Austin was interested in the ‘potentialities of linguistic usage as an index of conceptual truth, without regard for any particular direction for the deployment of such truth.’51 Grice also addresses critics of ordinary language philosophy more directly, accusing them of simple misrepresentation. Gellner’s book, in particular, he describes informally as ‘a travesty’. In effect, he accuses the critics of inverted snobbery: of assuming that ordinary language philosophy must be trivial and elitist because it was practised by a socially privileged group of people in an ancient university. He admits that the style of philosophy came easiest to those who had received a classical, therefore a public school, education. Education in the classics had fitted the Oxford philosophers to the necessary close attention to linguistic distinctions: ordinary language philosophy involved ‘the deployment of a proficiency specially accessible to the establishment, namely a highly developed sensitivity to the richness of linguistic usage’.52 Grice’s response is perhaps confused. At times he seems to accept the accusation of elitism but to argue against its derogatory implications. Why should high culture in philosophy be considered bad, he wonders, if it is prized in art, literature and opera? At other times, however, he seems simply to reject the accusation, arguing that anyone with suitably acute powers of observation could engage in this kind of analysis. This ambivalence is most apparent in some of his comments that did not make it into print. In a section of the handwritten draft of the ‘Reply to Richards’, but not of the published version, he comments on the accusation of elitism: To this the obvious reply would seem to be that if it were correct that philosophy is worth fostering, and that proficiency demands a classical education, then that would provide additional strength to the case for the availability of a classical education. But I doubt if the facts are as stated; I doubt if ‘linguistic sensitivity’ is the prerogative of either the well-to-do or of the products of any particular style of education.53 Grice’s attitude to his social environment may have been, or have become, somewhat ambivalent, but during the 1940s and 1950s he was central to the ordinary language philosophy movement. He attended Austin’s Saturday Mornings regularly and enthusiastically. Outside these meetings, he worked collaboratively with other members of the group.

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 47

Post-War Oxford 47

He even collaborated with Austin himself, although in teaching rather than writing. The two taught joint sessions on Aristotle. Grice later claimed that these were particularly easy teaching sessions for him because no preparation was necessary. They would simply arrive in class and Austin would start talking, freely and spontaneously. At some point Grice would find something with which he disagreed and say so, ‘and the class would begin’.54 It seems that their style of teaching often rested on each holding fast to an opposing position. Speaking long after he had left Oxford, Grice recalled Austin’s typical response when challenged on some point: ‘Well, if you don’t like that argument, here’s another one.’55 Grice may have been amused by Austin’s view of philosophy in terms of ‘winning’, but he himself was not likely to back down on a position once he had taken it up. However, this method had its successes. Their pupil J. L. Ackrill went on to become a Fellow of Brasenose College, and to publish his own translation of Aristotle’s Categories and De Interpretatione in 1963. Among his acknowledgements to those who had helped him in understanding these works, he comments: ‘I am conscious of having benefited particularly from a class given at Oxford in 1956–7 by the late J. L. Austin and Mr H. P. Grice.’56 For Grice himself, the most fruitful collaboration during this period was the joint work he undertook with Peter Strawson. Strawson had gone up to Oxford in 1939 and had been Grice’s tutee. He later commented that: ‘tutorials with [Grice] regularly extended long beyond the customary hour, and from him I learned more of the difficulty and possibilities of philosophical argument than from anyone else.’57 Strawson’s career also had been interrupted by the war, and he did not gain a permanent position until 1948, when he was appointed Fellow of University College. He was one of the younger members of the Play Group, but nevertheless a central figure in ordinary language philosophy, and unusually successful in completing work. In 1950 he published ‘On referring’, in which he offered a response to Bertrand Russell’s theory of descriptions, an account that had been widely praised, and practically unchallenged, since it first appeared in 1905. Russell had argued that natural language can be misleading as to the structure of our knowledge of the world; Austin and his followers that, on the contrary, it is the best guide we have to that structure. Strawson suggests that, preoccupied with mathematics and logic, Russell had overlooked the fact that language is first and foremost a tool in communication. Russell had ignored the role of the speaker: ‘ “mentioning”, or “referring”, is not something an expression does; it is something that someone can use an expression to do.’58 Once an appropriate distinction is drawn between

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 48

48 Paul Grice, Philosopher and Linguist

‘a sentence’ and ‘a use of a sentence’, it is possible to consider how most people would actually react on hearing the ‘king of France’ example (in which, for some reason, Strawson changes ‘bald’ to ‘wise’). When used in 1950, or indeed in 1905, the example is not false. Rather, the question of whether it is true or false simply does not arise, precisely because the subject expression fails to pick out an individual in the world, and therefore the expression as a whole fails to say anything. For Strawson, then, ‘the king of France is wise’ does not entail the existence of a unique king of France. Rather, it implies this existence in a ‘special’ way. In later work he uses the label ‘presupposition’ to describe this special type of implication.59 This was a term originally used by Frege in his discussion of denoting phrases and, like Frege, Strawson notes that presupposition is distinct from entailment because it is shared by both a sentence and its negation. Both ‘the king of France is wise’ and ‘the king of France is not wise’ share the presupposition that there exists a unique king of France, although their entailments contradict each other. On Russell’s analysis, the negative sentence is ambiguous. Both the existence and the wisdom of the king of France are logical entailments, so either could be denied by negating the sentence. Strawson argues that Russell has attempted, inappropriately, to apply classical logic to what are in fact specific types of natural language use. ‘Neither Aristotelian nor Russellian rules give the exact logic of any expression of ordinary language’; he concludes, ‘for ordinary language has no exact logic.’60 Stawson and Grice taught together, covering topics such as meaning, categories and logical form. W. V. O. Quine, who attended some of their weekly seminars in the spring of 1954, paints a picture of these sessions: Peter and Paul alternated from week to week in the roles of speaker and commentator. The speaker would read his paper and then the commentator would read his prepared comments. ‘Towards the foot of page 9, I believe you said . . .’ Considered judgement was of the essence; spontaneity was not. When in its final phase the meeting was opened to public discussion, Peter and Paul were not outgoing. ‘I’m not sure what to make of that question.’ ‘It depends, I should have thought, on what one means by . . .’ ‘This is a point that I shall think further about before the next meeting.’61 This style of debate was obviously not to Quine’s taste, and there is certainly something reminiscent of Grice’s former tutor W. F. R. Hardie in what he describes. Grice worked and talked very much within the

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 49

Post-War Oxford 49

formal and mannered social milieu of Oxford at the time. His lecture notes from the period, generally written out in full, contain many references to what ‘Mr Stawson proposed last week’ or ‘Mr Warnock has suggested’. But there is evidence also that the apparently prevaricating answers Quine ridicules were genuine responses to the complexity of the matter in hand. Grice often begins his lectures with a return to ‘a question raised last week to which I had no answer’. Indeed, he later argued explicitly that philosophy works best not as a series of rapid retorts to impromptu questions, but as a slow and considered debate: The idea that a professional philosopher should either have already solved all questions, or should be equipped to solve any problem immediately, is no less ridiculous than would be the idea that Karpov ought to be able successfully to defend his title if he, though not his opponent, were bound by the rules of lightening chess.62 This slow, thoughtful style was carried over to the deliberation and composition of Grice and Strawson’s writing and was, Grice later suggested, the reason why the collaboration eventually ceased. Strawson, who seems to have been much more interested in finished products than Grice was, eventually could no longer cope with the laborious process, in which complete agreement had to be reached over a sentence before anything could be written down. But it lasted long enough for Grice to be impressed by the ‘extraordinary closeness of the intellectual rapport’ they developed, such that ‘the potentialities of such joint endeavours continued to lure me.’63 One of their writing projects developed from their teaching on categories. Their topic dated back to Aristotle, and had also been discussed by Kant. It is a type of descriptive metaphysics, in that it involves analysis and classification towards an account of the very general notion of ‘what there is’. In metaphysics, this account is offered in terms of the nature and status of reality, rather than in terms of the types of description of individual aspects of that reality to be found in the sciences. In particular, it is concerned with reality as reflected in human thought. Kant had argued that the study of human thinking is important precisely because it is the only guide we have to reality. The facts of ‘things in themselves’, aside from our human perceptions of them, are unknowable. For Aristotle, these questions had been closely linked to his interest in the study of language. Human language is the best model we have available of human thought. We structure our language to reflect the structure of our thought. So, for instance, the division of sentences into subjects and predicates

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 50

50 Paul Grice, Philosopher and Linguist

reflects the way we cognitively divide the world into things and attributes, or particulars and universals. Grice and Strawson’s long, slow collaboration on this topic was not very productive, by modern standards at least, in that it resulted in no published work. But Grice himself valued it very highly, and commented late in his life that he had always kept the manuscripts with him.64 Unfortunately, only a fragment of these appears to have survived. However, this is accompanied by rough notes indicating something of their working method. They apply a distinctively ‘Austinian’ approach to Aristotle’s use of language, experimenting together to see which English words can successfully combine with which others, trying out combinations of possible subjects and predicates. Notes in Grice’s hand record that ‘healthy’ can be applied to, or predicated of, ‘person’, ‘place’, ‘occupation’, or ‘institution’, while ‘medical’ can be applied to ‘lecture’, ‘man’, ‘treatise’, ‘problem’, ‘apparatus’, ‘prescription’, and ‘advice’.65 Grice notes that the neutral term ‘employment’ can cover the importantly different terms ‘use’, ‘sense’ and ‘meaning’, and that here it is perhaps most appropriate to say that the predicate has the particular range of ‘uses’. This aspect of the work seems to take over Grice’s attention at one point. The notions of ‘sense’, ‘meaning’ and ‘use’ need to be distinguished not just from each other, but between discussions of sentences and of speakers. ‘Need to distinguish all these from cases where speaker might mean so-and-so or such-and-such, but wouldn’t say that of the sentence’ he notes, ‘ “Jones is between Williams and Brown” either spatial order or order of merit, but doubtful this renders it an ambiguous sentence.’ The notes also explore the difference in grammatical distribution between substantial and non-substantial nouns. ‘Mercy’ can be both referred to and predicated, whereas ‘Socrates’ can only be referred to. In other words, it seems that although both substantials and nonsubstantials can occupy subject position, substantials must be seen as the primary occupiers of this slot because they cannot occur as predicates. To demonstrate this point he distinguishes between establishing existence by referring to, and by predicating, expressions. The distinction is established by means of a comparison between two short dialogues. The following is perfectly acceptable; evidence is successfully offered for the existence of a universal by predicating it of a particular: A: Bunbury is really disinterested. B: Disinterested persons (real disinterestedness) does not exist. A: Yes they do (it does); Bunbury is really disinterested.

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 51

Post-War Oxford 51

In contrast descriptive statements do not guarentee the existence of their subjects, but rather stand in a ‘special relation’ to statements of existence. This special relation is, they note, elsewhere called ‘presupposition’; it seems that at this stage at least, Grice was prepared to endorse Strawson’s response to Russell. The following is ‘not linguistically in order’: A: Bunbury is really disinterested. B: There is no such person as Bunbury. A: Yes there is, he is really disinterested. Producing a sentence in which something is predicated of a substantial subject is not enough to guarentee the existence of that subject. They note that the situation would be quite different if ‘in the next room’ were substituted into A’s response. Here A’s choice of predicate does more than just offer a description of Bunbury; it points to a way of verifying his existence. Our use of language, then, very often presupposes the existence of substantial objects. Furthermore, substances are in a sense the basic, or primary objects of reference. Grice and Strawson admit that they have no conclusive proof of this latter claim, but they offer their own intuitions in support, presumably drawn from their experiments with substantial and non-substantial subjects. Further, substances are, in general, what we are most interested in talking about, asking about, issuing orders about; ‘indeed the primacy of substance is deeply embedded in our language’. Grice later suggested that the closest echoes of this joint work in published writing were in Strawson’s book Individuals, subtitled ‘An essay in descriptive metaphysics’, published in 1959. In this Strawson considers in much more detail the implications of various kinds of descriptive sentences for the theory of presupposition. But there were also resonances of the joint project in various aspects of Grice’s own later work. Two points from the end of the manuscript fragment, one a nonlinguistic parallel and one a consequence of their position, indicate these. The parallel comes from a consideration of the basic needs for survival and the attainment of satisfaction people experience as living creatures. Processes such as ‘eating’, ‘drinking’, ‘being hurt by’, ‘using’, ‘finding’, are entirely dependent on transactions with substances. Grice and Strawson suggest that it is not entirely fanciful to consider that the structure of language has developed to reflect the structure of these most basic interactions with the world.

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 52

52 Paul Grice, Philosopher and Linguist

The relevant consequence of their position relates to a familiar target for ordinary language philosophers: the theory of sense data. Their argument is in essence a version of the argument from common sense, although they do not explicitly acknowledge this. If substances are a primary focus of interest, and if this fact is reflected in the language, then this offers good evidence that the world must indeed be substantial in character. If the proponents of sense data were correct, if all we can accurately discuss are the individual sensations we receive through our sense, then ‘substantial terminology would have no application’. Of course, this argument is fundamentally dependent on faith in ordinary language. In effect it claims that, since people talk about, and indeed focus their talk on, material objects, we have adequate grounds for accepting that material objects exist. The second major collaboration between Grice and Strawson, which produced their only jointly published paper, was composed uncharacteristically rapidly. It was written in the same year as Quine observed their joint seminars, and indeed was in direct response to his visit. Quine introduced to Oxford an American style of philosophy. This had its roots in logical positivism, but had developed in rather different directions. In particular, Quine’s empiricism took the form of an extreme scepticism towards meaning. He argued that it was legitimate to look at meaning in terms of how words are used to refer to objects in the world, but not to discuss meanings as if they had some existence independent of the set of such uses. In an essay published in 1953 he argues that the error of attributing independent existence to meaning explains the tenacity of the doctrine of the distinction between analytic and synthetic sentences, even in empirical philosophies where it should have been abandoned. Quine argues that, although it is a basic doctrine of many empirical theories, the analytic/synthetic distinction is inherently unempirical. It relies on the notion that words have meaning independent of individual, observable instances of use. The sentence ‘no bachelor is married’ can be judged analytic and therefore necessarily true only on the assumption that ‘bachelor’ is synonymous with ‘unmarried man’. Yet such an assumption depends on a commitment to abstract meaning. It is legitimate only to consider the range of phenomena to which the two terms are applied, a process that must inevitably be open-ended. In other words, we are committed to the truth of so-called analytic sentences for exactly the same reason that we are committed to the truth of certain synthetic sentences, such as ‘Brutus killed Caesar’. Our past experience of the world, including our experience of how the words of

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 53

Post-War Oxford 53

our language are applied, has led us to accept them as true. However, our belief in any statement established on empirical grounds is subject to revision in the light of new experience. We may find it hard to imagine what experience could lead us to abandon belief in ‘no bachelor is married’, but that is simply because of the strength of our particular empirical commitment to it. The difference between analytic and synthetic statements, then, is not an absolute one, but simply a matter of degree. The distinction reflects ‘the relative likelihood, in practice, of choosing one statement rather than another for revision in the event of recalcitrant experience’.66 One weekend during Quine’s 1953–4 visit, Grice and Strawson put together their response to this argument, to be presented at a seminar on the Monday. In 1956 Strawson revised the paper on his own and sent it to The Philosophical Review, where it was published as ‘In defence of a dogma’. Their defence is exactly of the type that Quine might have expected to encounter in Oxford in the 1950s; it rests on the way in which some of the key terms are actually used. The terms ‘analytic’ and ‘synthetic’ have a venerable pedigree. Generations of philosophers have found them useful in discussions of language and, moreover, are generally able to reach consensus over how they apply to novel, as well as to familiar examples. But the argument does not end with philosophical usage. Grice and Strawson argue that closely related distinctions can be found in everyday speech. The notion of analyticity rests, as Quine had noted, on the validity of synonymy. The expression ‘synonymous’ may not be very common in everyday speech, but the equivalent predicate ‘means the same as’ certainly is. Borrowing an example from Quine’s essay, Grice and Strawson note that ‘creature with a heart’ is extensionally equivalent to ‘creature with kidneys’; it picks out the same group of entities in the world precisely because all creatures with a heart also have kidneys. However, Grice and Stawson argue that the relationship between these two expressions is not the same as that between ‘bachelor’ and ‘unmarried man’. Many people would be happy to accept that ‘ “bachelor” means the same as “unmarried man” ’ but would reject ‘ “creature with a heart” means the same as “creature with kidneys” ’. In effect, Grice and Strawson are defending the existence of exactly the type of meaning Quine dismisses. The element distinguishing the second pair of terms, but not the first, cannot be anything to do with the objects in the world they are used to label, because in this they are exactly equivalent. Rather, there must be some further notion of ‘meaning’ that people are aware of when they deny that ‘creature with heart’ and ‘creature with kidneys’ mean the same. It is this notion of

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 54

54 Paul Grice, Philosopher and Linguist

meaning, which forms part of the description of the language but does not directly relate to the world it describes, that Quine rules inadmissible. This also allows for the existence of analytic sentences. ‘A bachelor is an unmarried man’ is analytic precisely because its subject means the same as its predicate. This can be confirmed by studying just the language, not the world. In much the same way, the meaning of some sentences makes them impossible in the face of any experience. That is, we just cannot conceive of any evidence that would lead us to acknowledge them as true. Grice and Strawson contrast ‘My neighbour’s three-year-old child understands Russell’s Theory of Types’ with ‘My neighbour’s three-year-old-child is an adult’.67 The first may well strike us as impossible, but we can nevertheless imagine the sort of evidence that would force us, however reluctantly, to change our mind and accept it as true. But in the second case, unless we allow that the words are being used figuratively or metaphorically, there is just no evidence that could be produced to lead us to change our minds. The impossibility of the second example is entirely dependent on the meanings of the words it contains. It is clear that the nature of analytic and synthetic sentences, and of our knowledge of them, exercised Grice a good deal both at this time and later. Kathleen Grice recalls that, during the 1950s, he delighted in questioning his children’s playmates about ‘whether something can be red and green all over’, and enjoyed their subsequent confusion, insisting that spots and stripes were not allowed.68 As he observes in his own notes, ‘Nothing can be red and green all over’ is a supposed candidate for a statement that is both synthetic and a priori.69 Grice was presumably amusing himself by testing this claim out on some genuinely naive informants. In later years, he expressed himself still committed to the validity of the distinction between synthetic and analytic sentences, and indeed as seeing this as one of the most crucial of philosophical issues. However, he had grown unhappy with his and Strawson’s original defence of it. In particular, he no longer felt that it was enough to argue that the distinction is embedded in philosophical conscience. The argument indicates that the distinction is intelligible, in that it is possible to explain what philosophers mean by it, but not that it is theoretically necessary. In other words, ‘the fact that a certain concept or distinction is frequently deployed by a population of speakers and thinkers offers no guarantee that the concept in question can survive rigorous theoretical scrutiny’.70 Certainly their case in defence of the distinction may seem rather weak as a piece of ordinary language philosophy. Austin had rejected other terms, such as ‘sense data’, precisely because their

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 55

Post-War Oxford 55

use seemed to be restricted to philosophical discourse. Nevertheless, philosophers as distinguished as Anthony Quinton found their argument against Quine compelling, paraphrasing them as saying that: ‘An idea does not become technical simply by being dressed up in an ungainly polysyllable.’71 Grice never really returned to the task of mounting a new defence for the distinction, although he maintained that it was important and, crucially, ‘doable’.72 Grice’s other major collaboration of the 1950s was with Geoffrey Warnock, a regular member of the Play Group, and later Principal of Hertford College. As with Strawson, Grice’s collaboration with Warnock grew out of a joint teaching venture, and again this took the form of a series of evening lectures at which they alternated as speaker and commentator. The pace was leisurely and exploratory; a single topic might be developed over the course of a term, or even over successive years. Their theme was perception, the subject of Austin’s Sense and Sensibilia lectures. In these, Austin had applied his method of trying out examples to investigate shades of meaning for some of the relevant terminology. For instance, he argued that the terms ‘looks’, ‘appears’ and ‘seems’, which Ayer had assumed to be as good as synonymous, in fact displayed some striking differences in meaning when considered in a set such as ‘the hill looks steep’, ‘the hill appears steep’, ‘the hill seems steep’.73 Warnock has commented that, perhaps because of the intense treatment in these lectures, the vocabulary of perception was never discussed on Saturday mornings.74 However, this was an omission that he and Grice made up for in private; their surviving notes are full of lists of words and usages by means of which they hoped to discover the understanding of perception embedded in the language. For instance, they draw up a list headed ‘Syntax of Illusion’, considering the various constructions in which the word ‘illusion’ can appear, and the shades of meaning associated with these. ‘Be under the illusion (that)’ and ‘have the illusion (that)’ are grouped together, presumably as predicates that can attach to people. ‘Gives the illusion (that)’ and ‘creates the illusion (that)’ are listed separately, with the note that these might ‘apply exclusively to cases where anyone might be expected, in the circs, to handle the illusion’. The example ‘it is an illusion’ is accompanied by the unanswered question ‘what is it?’75 Grice and Warnock are concerned here with the so-called ‘argument from illusion’. This had been put forward by a number of philosophers in support of the need to explain perception in terms of sense data. Austin spends a considerable proportion of Sense and Sensibilia

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 56

56 Paul Grice, Philosopher and Linguist

attacking Ayer over exactly this issue, although he acknowledges that Ayer’s commitment to it is tentative. The argument draws on the fact that in a number of more or less mundane situations things appear to be other than they in fact are. For instance, a stick placed partially in water has the appearance of being bent, even though it is perfectly straight. When you look at your reflection in the mirror, your body appears to be located several feet behind the mirror when in fact it is located several feet in front of it. In these cases, the argument goes, there must be something that looks bent, or behind the mirror, that is distinct from the material object in question. The term ‘sense data’ is necessary in order to discuss this; the individual receives sense data that are distinct from the material object in question. From here, the argument is extended to cover all perception, even when no such illusions are involved. Seeing these types of illusion does not feel like a qualitatively different experience from seeing things as they actually are, hence it would be bizarre to maintain that the object of perception is of one kind in the cases of illusion and of another kind in other cases. It must be the case that all we ever do perceive are sense data, not material objects themselves. Austin spends several lectures addressing these issues, arguing that it is an unnecessary imposition to claim that there must be one type of object of perception. Ayer and others have illegitimately assumed that optical illusions, such as the ones discussed, along with dreams, mirages, hallucinations and various other phenomena can all be explained in the same way. Rather, we see a whole host of different entities in different situations. These different types of experience are in fact generally fairly easy to distinguish, as any reflection on the language available to discuss them will reveal. People ordinarily are quite capable of saying that ‘the stick looks bent but it is really straight’; it is philosophical sophistry to claim that there must be something, distinct from the stick itself, which has the ‘look’ in question. With typical bluntness, Austin dismisses the question of what you see when you see a bent stick as ‘really, completely mad’. And with perhaps some sleight of hand he dismisses the mirror example by arguing that what actually is several feet behind the mirror is ‘not a sense datum, but some region in the adjoining room’.76 In their lectures on perception, Grice and Warnock take up the question of whether or not it is legitimate to talk of any intermediary between material objects and our perceptions of them. They notice an asymmetry between the senses, or at least the vocabulary that describes them. The verbs ‘hear’, ‘smell’ and ‘taste’ can all take objects that do

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 57

Post-War Oxford 57

not describe material phenomena. It is possible to ‘hear a car’, but is equally possible to ‘hear the sound of a car’. In a similar way, it is possible to describe the ‘smell of a cheese’ or the ‘taste of a cake’. This is not the case for the other two senses. The objects of touch are always material and not, or not primarily, the ‘feeling’ of material things. Feeling, they note, perhaps entails having sensations, but it is not the case that these sensations are what we feel. Furthermore, there is no word at all to occupy this intermediary place in the case of sight. Certainly there is no general word analogous to ‘sounds’. This conclusion was not reached lightly. Notes in Grice’s hand on the back of an envelope try out a list of words in an attempt to find a suitable candidate for the role, but all are rejected. The following types of objects, he notes, fail to fit the bill because they are not universal objects of seeing; ‘surface, colour, aspect, view, sight, gleam, appearance . . .’ can all be used, but only in special circumstances. In most cases, we simply talk of seeing the object in question. In response to this lack of a word analogous to ‘sound’, Grice and Warnock do something rather surprising for philosophers of ordinary language; they make one up. They introduce the word ‘visa’ to describe the non-material intermediaries between material objects and sight. Commenting later on this enterprise, Grice was keen to point out that ordinary language philosophy did not in fact rule out the introduction of jargon, so long as it was what he rather enigmatically describes as ‘properly introduced’.77 Their manoeuvre was rather different from that of introducing a technical term such as ‘sense data’, and then basing a philosophy around the existence of sense data. Instead, they were giving a name to what they had discovered to be a gap in the language in order to explore the language itself: to consider why no such word, and therefore presumably no such concept, was necessary. Their suggestion is that sight and touch are the two essential senses. That is, they are the senses on which we rely most closely to tell us about the material world. Touch is primary. Material objects are most essentially present to us through touch, and this is the sense we would be least ready to disbelieve. However, because our capacity to touch material objects at any one time is limited, we rely to varying degrees on our other senses. Of these, sight correlates most closely with touch. It offers us the best indication of what an object would feel like. It is also more constant; an object capable of being heard may sometimes fail to produce its distinctive sound, but an object capable of being seen cannot fail to produce its ‘visum’. For precisely this reason, the word ‘visum’, to describe something separate from the object producing it, does not exist.

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 58

58 Paul Grice, Philosopher and Linguist

There are two possible accounts of the absence of a word such as ‘visum’. The stronger account is that there is no space in the language for such a word; it could not exist. The weaker claim is that such a word could exist, but there is simply no need for it, because we never need to talk about the perceptions of sight without talking about material objects themselves. Warnock held to the stronger view and Grice to the weaker. This difference seems to have been rather a fraught one; Grice begins one lecture with the following extended metaphor. I begin with an apology (or apologia). Last week it might well have appeared to some that W. was doing the digging while I was administering the digs. I should not like it to be thought that I am not in sympathy with his main ideas about ‘visa’; the only question in my mind is precisely what form an answer on his general lines should take; and if I am busier with the awl than with the spade, the reason (aside from ineptitude with the latter instrument) is that I am pretty sure that he is digging in the right place.78 It seems that the previous week Warnock had argued that if there were a word ‘visum’, then, analogous to sound, it would have to be possible to say that a material object could be ‘visumless’, but that this would not fit into our existing conceptual framework. Grice’s argument is that if we want to use such a word, we can always make use of one of his suggestions, such as ‘colour expanse’. It is probably correct to say that whenever we see a material object we see a colour expanse, but such an expression is reserved for those occasions when appearance and reality differ. It would not be strictly false to use them on other occasions, but: ‘it would be unnatural or odd to employ this expression outside a limited range of situations, just because the empirical facts about perception mentioned by W. ensure that normally we are in a position to use stronger and more informative language.’ Grice’s implication is that if a speaker is in a position to say something more informative, the less informative statement would not be used. There was no joint publication for Grice and Warnock. On his own, Warnock did publish in this general field, although he did not make direct use of the notion of ‘visa’. His 1955 paper to the Aristotelian Society, ‘Seeing’, is chiefly concerned with a close examination of ‘the actual employment of the verb “to see” ’.79 Warnock considers the implications of the various categories of object that the verb commonly takes. In a footnote he acknowledges his general debt to ‘Mr H. P. Grice’. Grice himself made even less direct use of their joint project in his own later

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 59

Post-War Oxford 59

work, although in the early 1960s he published two papers concerned with general questions about the status of perception.80 Later in that same decade he wrote a reflection on Descartes, although this was not published for more than 20 years.81 Descartes’s scepticism leads Grice to reconsider the claims of phenomenalism, and to argue for a clearer distinction between ‘truth-conditions’, ‘establishment-conditions’ and ‘reassurance-conditions’ for statements about material objects. It is evident that in drawing on the vocabulary of perception, Grice and Warnock were not just dutifully following Austin’s lead; both were enthusiastic about the enterprise, and impressed by its results. Warnock recalls that at one point during this process Grice exclaimed, ‘How clever language is!’, and goes on himself to gloss this remark by explaining: ‘We found that it made for us some remarkably ingenious distinctions and assimilations.’82 At this stage at least, Grice was committed to the close study of ordinary language as a productive philosophical enterprise, and to Austin’s notion of ‘linguistic botanising’ as a philosophical tool. He was very much a part of the social, as well as the philosophical, milieu of 1950s Oxford. He and Kathleen still lived on the Woodstock Road with their children Karen and Tim, born in 1944 and 1950. However, he took a full and active role in the life of St John’s College which meant dining there frequently. Kathleen remembers the college as a distant, hierarchical institution. Women were permitted to enter college only once a year, when the porters and college servants arranged a Christmas party for the wives and children of Fellows.83 Quine’s thumbnail sketch of his impressions of Grice in 1953, unflattering in the extreme, suggests that his concern for college proprieties was even stronger than his disregard for personal appearance: Paul was shabby . . . His white hair was sparse and stringy, he was missing some teeth, and his clothes interested him little. He was vicepresident of St John’s college, and when I came as guest to a great annual feast there I was bewildered to encounter him impeccably decked out in white tie and tails, strictly fashion-plate.84 The vice-presidentship in question was a one-year appointment, and seems to have been largely social in nature, concerned with the organisation and running of college events. However, Grice’s academic standing was also rising; in 1948 he was appointed to a University lectureship in addition to his college fellowship. This meant giving lectures open to all members of the university, in addition to the tutorial teaching duties at St. John’s. Grice was also making a growing number of visits

1403_902976_04_cha03.qxd 12/13/04 6:25 PM Page 60

60 Paul Grice, Philosopher and Linguist

away from Oxford to give lectures and seminars. Most significantly, he began to make fairly frequent trips to America, where he was invited to speak at various universities. He was in fact one of the few to talk about ordinary language philosophy outside Oxford at that time, giving the lecture later published as ‘Post-War Oxford Philosophy’ at Wellesley College, Massachusetts in 1958.85 However, during this time he was gradually moving away from some of the strictures and dogmas of that approach. In many ways, Grice never fully abandoned ordinary language philosophy; his notes testify to the fact that he frequently returned to it as a method, in particular using ‘linguistic botany’ to get his thinking started on some topic. However, his philosophy began to diverge from it during this period in some highly significant ways that were to make themselves manifest in his best known work.

1403_902976_05_cha04.qxd 12/13/04 6:26 PM Page 61

4 Meaning

Grice’s growing unease with ordinary language philosophy brought him into conflict with J. L. Austin. Speaking long after leaving Oxford, he mentioned that he himself had always got on well with Austin, at least ‘as far as you can get on with someone on the surface so uncosy’.1 But he clearly found that engaging with him in philosophical discussion could be a frustrating enterprise. Austin would adopt a stance on some particular aspect of language, often formed on the basis of a very small number of examples, and defend his position against a barrage of objections and counter-examples, however obvious. His tenacity in this would be impressive; Grice knew he himself had won an argument only on those occasions when Austin did not return to the topic the following week. There is something in this reminiscent of the criticisms of obstinacy levelled against Grice himself as an undergraduate. However, it seems that there was more to Grice’s frustration than simply encountering someone even less ready to back down in an argument than he was himself. He was becoming increasingly unhappy with the methodology supporting Austin’s hasty hypothesis formation: the absolute reliance on particular linguistic examples. It is not that Grice lost faith in the importance of ordinary language or the value of close attention to differences in usage. However, along with some of Austin’s harsher and more outspoken critics, Grice came to regard his approach as unfocused and even trivial. In his reminiscences of Oxford, he complained that Austin’s ability to distinguish the philosophically interesting from the mundane was so poor that the Saturday morning Play Group sometimes ended up devoting an inordinate amount of time to irrelevant minutiae. They once spent five weeks trying to discern some philosophically interesting differences between ‘very’ and ‘highly’, considering why, for 61

1403_902976_05_cha04.qxd 12/13/04 6:26 PM Page 62

62 Paul Grice, Philosopher and Linguist

instance, we say ‘very unhappy’ but not ‘highly unhappy’. The result was ‘total failure’. On another occasion, Austin proposed to launch a discussion of the philosophical concept of Pleasure with a consideration of such phrases as ‘I have the pleasure to announce . . .’. Stuart Hampshire remarked to Grice that this was rather like meeting to discuss the nature of Faith, and proceeding to discuss the use of ‘yours faithfully’. In a similar vein, Grice was far from convinced that Austin’s trusted method of ‘going through the dictionary’ had much to offer in the way of philosophical illumination. Speaking publicly in the 1970s, he commented that he once decided to put this method to the test by using it to investigate the vocabulary of emotions. He started at the beginning of a dictionary in search of words that could complement the verb ‘to feel’. He gave up towards the end of the Bs, on discovering that the verb would tolerate ‘Byzantine’, and realising that ‘the list would be enormous’.2 He never found any satisfactory way of narrowing such a list so as to include only philosophically interesting examples, and clearly believed that Austin had no such method either. Grice did not hesitate to explain his views to Austin, whose reply was typical of both his commitment to his methodology, and his intransigent response to opposition. Grice reports himself as commenting that he, for one, ‘didn’t give a hoot what the dictionary says’, to which Austin retorted, ‘and that’s where you make your big mistake’.3 The differences between Austin and Grice were not simply matters of personality or even of methodology. They reflected some important ways in which the two philosophers disagreed over what was appropriate in an account of language. Grice was increasingly convinced of the necessity for explanations that took the form of general theories, rather than the ad hoc descriptions that were Austin’s trademark. He believed that Austin should be prepared to go further in drawing out the theoretical implications of his individual observations. Indeed, it has been suggested that Austin was suspicious of broad theories, generally supposing them to be over-ambitious and premature.4 Grice also differed from Austin in his growing conviction that a distinction needed to be recognised between what words literally or conventionally mean, and what speakers may use them to mean on particular occasions. This conviction is not present in his earliest work. In ‘Personal identity’, for instance, he writes of ‘words’ and of ‘the use of words’ interchangeably. However, the distinction is the subject of the long digression in his notes for the ‘Categories’ project with Strawson. It was to play an increasingly prominent part in his work, and to set him apart most

1403_902976_05_cha04.qxd 12/13/04 6:26 PM Page 63

Meaning 63

sharply from the orthodoxies of ordinary language philosophy. For Austin, as for Wittgenstein in his later work, use was the best, indeed the only, legitimate guide to meaning. The privilege afforded to use left no room for a distinction between this and literal meaning. Grice did not claim that actual use is unimportant. Indeed, in many of his writings he studied it with as much zeal as Austin. Rather, he saw it as a reliable guide just to what people often communicate, not necessarily to strict linguistic meaning. Both the general interest in explanatory theories and the more particular focus on the distinction between linguistic meaning and speaker meaning are apparent in ‘Meaning’, a short but hugely influential article first published in 1957. It is tempting to trace the development of Grice’s ideas through his work during the 1950s, but in fact this is ruled out by the article’s history. Grice wrote the paper in practically its final form in 1948 for a meeting of the Oxford Philosophical Society but, as usual, he was hesitant to disseminate his ideas more widely. Almost a decade later, Peter Strawson was worried that the Oxford philosophers were not sufficiently represented in print. Unable to convince Grice to revise his paper and send it to a journal, he persuaded him to hand over the manuscript as it stood. Strawson and his wife Ann edited it and submitted it to the Philosophical Review. The paper therefore dates in fact from before Grice’s collaborations with Warnock and with Strawson, from before Austin’s development of speech act theory, and before even the publication of Ryle’s The Concept of Mind. ‘Meaning’ offers a deceptively straightforward-looking account of meaning, apparently based on a common-sense belief about the use of language. In effect, Grice argues that what speakers mean depends on what they intend to communicate. However, his project is much more ambitious, and its implications more far-reaching, and also more problematic, than this suggests. First, he introduces hearers, as well as speakers, into the account of meaning. The intention to communicate is not sufficient; this intention must be recognised by some audience for the communication to succeed. Hence the speaker must have an additional intention that the intention to communicate is recognised. More significantly, Grice proposes to broaden out his account to offer a theory of linguistic meaning itself. Linguistic meaning depends on speaker meaning, itself explained by the psychological phenomenon of intention; conventional meaning is to be defined in terms of psychology. In the opening paragraphs of ‘Meaning’, Grice does something Austin seems never to have attempted. He focuses the methodology of ordinary language philosophy on the concept of meaning itself, consider-

1403_902976_05_cha04.qxd 12/13/04 6:26 PM Page 64

64 Paul Grice, Philosopher and Linguist

ing and comparing some different ways in which the word ‘mean’ is used. His conclusion is that there are two different uses of ‘mean’, reflecting two different ways in which we generally conceive of meaning. The first use is exemplified by sentences such as ‘those spots mean measles’ and ‘the recent budget means that we shall have a hard year’. The second includes examples such as ‘those three rings on the bell (of the bus) mean that the bus is full’ and ‘that remark, “Smith couldn’t get on without his trouble and strife”, meant that Smith found his wife indispensable’. Grice draws up a list of five defining differences between these sets of examples, differences themselves expressed in terms of linguistic relations. For instance, the first, but not the second set of examples entail the truth of what is meant. It is not possible to add ‘but he hasn’t got measles’, but it is possible is to add ‘but the bus isn’t in fact full – the conductor has made a mistake’. Further, the examples in the first set do not convey the idea that anyone meant anything by the spots or by the budget. It is, however, possible to say that someone meant something by the rings on the bell (the conductor) and by the remark (the original speaker). And only in the second set of examples can the verb ‘mean’ be followed by a phrase in quotation marks. So ‘those three rings on the bell mean “the bus is full” ’ is possible, but not ‘those spots mean “measles” ’. Grice characterises the difference between his two sets of uses for ‘mean’ as a difference between natural and non-natural meaning. The latter, which includes but is not exhausted by linguistic meaning, he abbreviates to ‘meaningNN’. In pursuing the question of what makes ‘meaningNN’ distinctive, he considers but rejects a ‘causal’ answer. He offers the following as a summary of this type of answer, paraphrasing and partly quoting C. L. Stevenson: For x to meanNN something, x must have (roughly) a tendency to produce in an audience some attitude (cognitive or otherwise) and the tendency, in the case of a speaker, to be produced by that attitude, these tendencies being dependent on ‘an elaborate process of conditioning attending the use of the sign in communication’.5 This account makes the causal answer look very much like behaviourism, with its notion of meaning as, in effect, the tendency for certain responses to be exhibited to certain stimuli. However, in the work from which Grice quotes, a book called Ethics and Language published in 1944, Stevenson is careful to distance himself from straightforwardly behaviouristic psychology. He argues that such an account

1403_902976_05_cha04.qxd 12/13/04 6:26 PM Page 65

Meaning 65

would be too simplistic to explain a process as complex as linguistic meaning because it could refer only to overt, observable behaviour. Rather, meaning is a psychologically complex phenomenon. The ‘responses’ produced will include cognitive responses, such as belief and emotion. Therefore, any account in terms of responsive action must be allowed ‘to supplement an introspective analysis, not to discredit it’.6 Stevenson, like Grice after him, draws attention to a distinction between two ways in which ‘mean’ is used, although he does so only in passing. He, too, defines one class of meaning as ‘natural’, a class he illustrates by suggesting that ‘a reduced temperature may at times “mean” convalescence’.7 He contrasts these uses of ‘mean’ with linguistic meaning which, he argues, is dependent on convention. The ways in which speakers use words on individual occasions may vary considerably, but conventional meaning must be seen as primary and relatively stable, underlying any of the more specific meanings in context. A word may be used in context to ‘suggest’ many things over and above what by convention it ‘means’. So, for instance, the sentence ‘John is a remarkable athlete’ may have a disposition to make people think that John is tall, ‘but we should not ordinarily say that it “meant” anything about tallness, even though it “suggested” it’.8 The connection between ‘athlete’ and ‘tall’ is not one of linguistic rule, so tallness cannot be part of the meaning, however strongly it is suggested. What is suggested can sometimes be even more central, as in the cases of metaphors. Here, a sentence is ‘not taken’ in its literal meaning, but can be paraphrased using an ‘interpretation’ that ‘descriptively means what the metaphorical sentence suggests’.9 Grice’s critique of the ‘causal’ theory of meaning includes a response to Stevenson’s ‘athlete’ example. He argues that, in declaring ‘tallness’ to be suggested rather than meant, Stevenson is in effect pointing to the fact that it is possible to speak of ‘nontall athletes’: that there is no contradiction in terms here. This argument, Grice claims, is circular. He does not elaborate this point, but he presumably means that in establishing the acceptability of this way of speaking we need to refer to the conventions of linguistic behaviour, yet it is these very conventions we are seeking to establish. Grice adds that, on Stevenson’s account, it would be difficult to exclude forms of behaviour we would not normally want to regard as communicative at all. It may well be the case that many people, seeing someone putting on a tailcoat, would form the belief that that person is about to go to a dance, and that the belief would arise because that is conventional behaviour in such circumstances. But it could hardly be correct to conclude that putting on a tail-

1403_902976_05_cha04.qxd 12/13/04 6:26 PM Page 66

66 Paul Grice, Philosopher and Linguist

coat meansNN that one is about to go to a dance. To qualify by stating that the conventions invoked must be communicative simply introduces another circle: ‘We might just as well say “X has meaningNN if it is used in communication”, which, though true, is not helpful.’10 As a solution to the problems inherent in the notion of conventional behaviour, Grice introduces intention, doing so abruptly and without much elaboration. However, the concept would probably not have been much of a surprise to his original audience in 1948, having had a long pedigree in Oxford philosophy. Grice later commented that his interest in intention was inspired in part by G. F. Stout’s ‘Voluntary action’. Stout was a previous editor of Mind, and his article was published there in 1896. In it he argues that if voluntary action is to be distinguished from involuntary action in terms of ‘volition’, it is necessary to offer a unique account of volition, distinguishing it from will or desire. He suggests that in the case of volition, but not of simple desire, there is necessarily ‘a certain kind of judgement or belief’, namely that ‘so far as in us lies, we shall bring about the attainment of the desired end’.11 Indeed, if there is any doubt about the outcome of the volition, expressed in a conditional statement, this must refer to circumstances outside our control. This explains the difference between willing something and merely wishing for it. However, there is another crucial distinction between voluntary and involuntary action. In the case of the former, the action takes place precisely because of the relevant belief; a voluntary action happens because we judge that we will do it. An involuntary action may be one we judge is going to take place, but only because other factors have already determined this. Further light is shed on Grice’s interest in intention by a paper called ‘Disposition and intention’ he circulated among his colleagues just a few years after he had first written ‘Meaning’. The paper was never published, and has survived only in manuscript, accompanied by a number of written responses in different hands. Reading these, it is not hard to imagine how a writer who tended towards perfectionism and self-doubt could be further dissuaded from publication by the results of canvassing opinion. Oxford criticism might be couched in genteel terms, but it could be harsh. One commentary begins: ‘What it comes to, I think, is this; it seems to me that there is something wrong with the way Grice goes to work . . . .’12 In ‘Dispositions and intentions’, Grice declares that his purpose is to consider the best analysis of ‘psychological concepts’, concepts generally expressed in statements beginning with phrases such as ‘I like . . .’, ‘I want . . .’, ‘I expect . . .’. Such statements present analytical problems because they seem to refer to experience that is essen-

1403_902976_05_cha04.qxd 12/13/04 6:26 PM Page 67

Meaning 67

tially private and unverifiable. Grice proposes to compare two common approaches to this problem. One he labels the ‘dispositional’ account, whereby the relevant concept is seen as a disposition to act in certain ways in certain hypothetical situations. Thus a statement of ‘I like X’ could be seen as specifying that the subject would act in certain ways in the interests of X, should the situation arise. The alternative is to consider such statements as describing ‘special episodes’, in other words to concede that they can be understood as descriptive only of private sensations, directed simply at the individual concept in question. These sensations take the form of special and highly specific psychological entities, such as ‘liking-feelings’. Grice is dismissive of a third possibility, an explanation from behaviourism describing psychological concepts purely in terms of observable responses. He rejects such approaches as ‘silly’. Grice argues that although the dispositional account runs into difficulty in certain cases, such as ‘I want . . .’, ‘I expect . . .’ and, significantly, ‘I intend . . .’, it is not appropriate simply to switch to the ‘special episode’ account. The problems for the dispositional account are connected to their analysis in terms of hypotheticals. It would seem reasonable in the case of any genuine hypothetical to ask ‘how do you know?’ or ‘what are your grounds for saying that?’. But in the case of the psychological concepts mentioned above, no such evidence can generally be offered, or reasonably expected. Speakers cannot be expected to be judging such concepts from observation, not even from a special vantage point; they can only be judging from personal experience. In Grice’s metaphor, ‘I am not in the audience, not even in the front row of the stalls, I am on the stage.’13 The ‘special episode’ account does not fare much better. Every case of wanting, for instance, must create a separate psychological episode. The system of explanation quickly becomes unwieldy; wanting an apple, for instance, must be a discrete phenomenon, different from wanting an orange or a pear or a banana. The third alternative, any version of a behaviourist account, is simply doomed to failure in cases such as ‘want’. Grice singles Gilbert Ryle out for criticism. Ryle’s account of psychological concepts is at best not clearly distinguished from behaviourism because of its ambiguity over whether the evidential counterparts of dispositions can be private or must always be observable by other people. In The Concept of Mind, Ryle describes certain statements concerned with psychological concepts as acts characteristic of certain moods. This is in keeping with his refusal to distinguish between physical and mental entities or activities. When

1403_902976_05_cha04.qxd 12/13/04 6:26 PM Page 68

68 Paul Grice, Philosopher and Linguist

we feel bored, we are likely to utter ‘I feel bored’ in just the same way as we are likely to yawn. We do not observe our own mood and then offer an account of it; we simply act in accordance with it. Thus the statement is not produced as a result of assessing the evidence, but may itself form part of that relevant evidence, even to the speaker. ‘So the bored man finds out he is bored, if he does find this out, by finding that among other things he glumly says to others and to himself “I feel bored” and “How bored I feel”.’14 In response to such statements, we may want to ask ‘sincere or shammed?’, but not ‘true or false?’. This is because statements such as ‘I feel bored’, ‘ I want . . .’ or ‘I hope . . .’ act not as descriptive statements about ourselves but as ‘avowals’: unreflective utterances that may be treated as pieces of behaviour indicative of our frame of mind. In a passage strikingly reminiscent of Russell’s analysis of definite descriptions, Ryle argues that the grammatical form of such sentences ‘makes it tempting to misconstrue [them] as selfdescriptions’.15 In fact they are simply things said in the relevant frames of mind. However, unlike other forms of behaviour, speech is produced to be interpreted. Grice argues that the difference between speech and other forms of behaviour is much greater than Ryle allows. In particular, while it is possible to ‘sham’ behaviour such as yawning without being false, a ‘shammed’ statement of ‘I feel bored’ will be simply false. A statement of boredom is not of the same order as a yawn when it comes to offering information precisely because, as Ryle acknowledges, it is uttered voluntarily and deliberately. Grice adds another possible indicator of boredom into the comparison. When listening to a political speech, we might give an indication that we feel bored by saying ‘I feel bored’, by yawning, or indeed by making a remark such as ‘there is a good play coming on next week’. This last remark is also voluntary and deliberate, but it does not offer anything like the same strength of evidence of our state of boredom as the alleged ‘avowal’, precisely because it is not directly concerned with our state of mind. Only ‘I am bored’ is a way of telling someone that you are bored. Furthermore, although it might be possible to some extent to sustain a theory of avowals or other behaviour as offering information about the state of mind of others, it will hardly do for one’s own state of mind. A man does not need to wait to observe himself heading for the plate of fruit on the table before he is a position to know that he wants pineapple. Grice’s suggested solution to the apparent failure of the ‘dispositional’ and the ‘special event’ accounts, together with the unsustainability of a behaviourist approach, rests on intention. When analysing many

1403_902976_05_cha04.qxd 12/13/04 6:26 PM Page 69

Meaning 69

verbs of psychological attitude used to refer to someone else, we can produce a straightforward hypothetical, saying ‘if so and so were the case he would behave in such and such a way’. However, the same does not seem to be the case when talking about oneself. ‘If so and so were the case I would behave in such and such a way’ cannot really be understood as a statement of hypothetical fact, but as a statement of hypothetical intention, just so long as the behaviour in question can be seen as voluntary. Further, it is simply not possible to say ‘I am not sure whether I intend . . .’, in the way it is possible to doubt other psychological states. It might be possible to utter such a sentence meaningfully, but only to express indecision over the act in question, as when someone replies to the question ‘Do you intend . . . ?’ by saying ‘I’m not sure’. It cannot mean that the speaker does not know whether he or she is in a psychological state of intending or not. Grice argues that a correct analysis of intention can form an important basis for the analysis of a number of the relevant psychological concepts. He offers a twofold definition of statements of intention, which relies on dispositional evidence without divorcing itself completely from the privileged status of introspective knowledge. The first criterion, familiar from Stout’s original paper, is that the speaker’s freedom from doubt that the intended action will take place is not dependent on any empirical evidence. Grice’s second stipulation is that the speaker must be prepared to take the necessary steps to bring about the fulfilment of the intention. There would be something decidedly odd about stating seriously, ‘X intends to go abroad next month, but wouldn’t take any preliminary steps which are essential for this purpose’. Grice’s analysis of statements of intention ends his unpublished paper rather abruptly, without returning to the other psychological verbs discussed at the outset. Having established that a doubt over one’s own intention is something of an absurdity, he offers a characteristically tantalising suggestion that ‘we may hope that this may help to explain the absurdity of analogous expressions mentioning some other psychological concepts, though I wouldn’t for a moment claim that it will help to explain all such absurdities.’ Although his analysis of psychological concepts is here incomplete, he has established two significant commitments. First, he has justified the inclusion of psychological concepts in analyses; they do not need to be ‘translated’ away into behavioural tendencies or observable phenomena. Second, he has established the concept of intention as primary; it offers an illustration of how a psychological concept may be described, and is also in itself implicit in the analysis of other such concepts.

1403_902976_05_cha04.qxd 12/13/04 6:26 PM Page 70

70 Paul Grice, Philosopher and Linguist

Grice was not the only Oxford philosopher to revisit the topic of intention in the mid-twentieth century. In 1958, Stuart Hampshire and Herbert Hart published ‘Decision, intention and certainty’. This echoes Grice’s earlier, unpublished paper in arguing that there is a type of certainty about future actions, independent of any empirical evidence, that is characteristic of intention. They distinguish between a ‘prediction’ of future action and a ‘decision’ about it. A statement of the first is based on considering evidence, but a statement of the second on considering reason, even if this is not done consciously. People who say ‘I intend to do X’ may reasonably be taken to be committed to the belief that they will try to do X, and will succeed in doing X unless prevented by forces outside their control. Hampshire and Hart’s claims are similar to those of Stout, but there are interesting differences in approach and style, revealing of the distance between the philosophical methods characteristic of their times. Stout is representative of an older style of Oxford philosophy, complete with sweeping generalisations and moralistic musings. In discussing the tendency of voluntary determination to endure over time and against obstacles, he comments that: ‘If we are weak and vacillating, no one will depend on us; we shall be viewed with a kind of contempt. Mere vanity may go far to give fixity to the will.’16 After logical positivism, and after Austin’s insistence on close attention to the testimony of ordinary language, such unempirical excesses would not have been tolerated. Hampshire and Hart stick to more sober reflections on the language of intention and certainty. They come very close to one of Austin’s observations in ‘A plea for excuses’ when they comment that: ‘If I am telling you simply what someone did, e.g. took off his hat or sat down, it would normally be redundant, and hence misleading, though not false, to say that he sat down intentionally.’17 They describe the adverb ‘intentionally’ as usually used to rule out the suggestion that the action was in some way performed by accident or mistake. Also like Austin, they do not dwell on the possible consequences of the inclusion of the ‘redundant’ material. An interest in the psychological concept of intention, and a reaction against the behviouristic tendencies he identified in work by Stevenson and by Ryle, were not the only sources of inspiration for Grice’s ‘Meaning’. In the following brief paragraph from that article, he comments on theories of signs: The question about the distinction between natural and non-natural meaning is, I think, what people are getting at when they display an interest in a distinction between ‘natural’ and ‘conventional’ signs.

1403_902976_05_cha04.qxd 12/13/04 6:26 PM Page 71

Meaning 71

But I think my formulation is better. For some things which can meanNN something are not signs (e.g. words are not), and some are not conventional in any ordinary sense (e.g. certain gestures); while some things which mean naturally are not signs of what they mean (cf. the recent budget example).18 The mention of ‘people’ is not backed up by any specific references, but the unpublished papers Grice kept throughout his life include a series of lectures notes from Oxford in which he presents and discusses the ‘theory of signs’ put forward in the nineteenth century by the American philosopher C. S. Peirce. These suggest that Grice’s account of meaning developed in part from his reaction to Peirce, whose general approach would have appealed to him. Peirce was anti-sceptical, committed to describing a system of signs elaborate enough to account for the classification of the complex world around us, a world independent of our perceptions of it, and available for analysis by means of those perceptions. His theory of signs was therefore based on a notion of categories as the building blocks of knowledge; signs explain the ways we represent the world to ourselves and others in thought and in language. In a paper from 1867, Peirce reminds his readers ‘that the function of conceptions is to reduce the manifold of sensuous impressions to unity, and that the validity of a conception consists in the impossibility of reducing the content of consciousness to unity without the introduction of it’.19 Conceptions, and the signs we use to identify them, comprise our understanding of the world. Signs act by representing objects to the understanding of receivers, but there are a number of different ways in which this may take place. There are some representations ‘whose relation to their object is a mere community of some quality, and these representations may be termed likenesses’, such as for example the relationship between a portrait and a person. The relations of other representations to their object ‘consists in a correspondence in fact, and these may be termed indices or signs’; a weathercock represents the direction of the wind in this way. In the third case, there are no such factual links, merely conventional ones. Here the relation of representation to object ‘is an imputed character, which are the same as general signs, and these may be termed symbols’; such is the relationship between word and object.20 Peirce later extended the general term ‘sign’ to cover all these cases, and used the specific terms icon, index and symbol for his three classes of representation. In his lectures and notes on ‘Peirce’s general theory of signs’, Grice analyses Peirce’s use of the term ‘sign’, and proposes to equate it with

1403_902976_05_cha04.qxd 12/13/04 6:26 PM Page 72

72 Paul Grice, Philosopher and Linguist

a general understanding of ‘means’. His chief argument in favour of this equation is that Peirce is not using ‘sign’ in anything recognisable as its everyday or ordinary sense. In a piece of textbook ordinary language philosophy, he argues that ‘in general the use (unannounced) of technical or crypto-technical terms leads to nothing but trouble, obscuring proper questions and raising improper ones’.21 Restating Peirce’s claims about ‘signs’ in terms of ‘means’ draws attention to shared features of a range of items commonly referred to as having ‘meaning’, as well as highlighting some important differences between Peirce’s categories of ‘index’ and ‘symbol’. Grice does not seem to have anything to say about Peirce’s ‘icons’, perhaps because this type of sign is least amenable to being re-expressed in terms of meaning. Using his translation of ‘is a sign of’ into ‘means’, Grice reconsiders Peirce’s own example of an index. He observes that the sentence ‘the position of the weathercock meant that the wind was NE’ entails, first, that the wind was indeed NE; and second, that a causal connection holds between the wind and weathercock. This feature, he notes, seems to be restricted to the word ‘means’. You can say ‘the position of the weathercock was an indication that the wind was NE, but it was actually SE’; ‘was an indication that’ is not a satisfactory synonym of ‘means’ because it does not entail the truth of what follows. The causal connection also offers an interesting point of comparison to different ways in which ‘mean’ is used. Grice draws on an example that also appears in ‘Meaning’ when he suggests a conversation at a bus stop as the bus goes. The comment ‘Those three rings of the bell meant that the bus was full’ could legitimately be followed by a query of ‘was it full?’. At this early stage in the development of his account of meaning, Grice was obviously troubled by the relationship between sentence meaning and speaker meaning, and that between the meaning words have by convention and the full import they may convey in particular contexts. There is no fully developed account of these issues in the lectures, but there are a few rough notes suggesting that he was aware of the need to identify some differences between ‘timeless’ meaning and speaker meaning. At one point he notes to himself that the following questions will need answers: ‘what is the nature of the distinction between utterances with “conventional” meaningNN and those with non-conventional?’ and ‘nature of distinction between language and use language? (sic)’. Elsewhere, he ponders the difference between ‘what a sentence means (in general; timeless means; if you like the meaning of (type) sentence)’ and ‘what I commit myself to by the use of a sentence (if you like, connect this with token sentence)’. It is not clear from

1403_902976_05_cha04.qxd 12/13/04 6:26 PM Page 73

Meaning 73

this whether he is seeing ‘non-conventional’ aspects of meaning as separate from, or coterminous with, those aspects of meaning arising from particular uses in context. However, the first two questions suggest that ‘non-conventional’ and ‘use’ are at least potentially separate issues. When Grice collected his thoughts on meaning into the form in which they were eventually published, he produced a definition dependent on a complex of intentions. In the case of straightforwardly informative or descriptive statements, meaningNN is to be defined in terms of a speaker’s intention to produce certain beliefs in a hearer. It is also necessary that the speaker further intends that the hearer recognise this informative intention and that this recognition is itself the cause of the hearer’s belief. This definition excludes from meaningNN a case where ‘I show Mr X a photograph of Mr Y displaying undue familiarity to Mrs X’, but includes the related but crucially different case where ‘I draw a picture of Mr Y behaving in this manner and show it to Mr X’.22 In the former case, the effect would be produced in Mr X whether or not he paid any attention to the intention behind the behaviour. In the second example the recognition of this intention is crucial, and precisely this case would normally be seen as a piece of communication. A similar distinction can be made when the definition is extended to include communication intended to influence the action of others. A policeman who stops a car by standing in its way will produce his desired effect whether or not the motorist notices his intention. A policeman who stops a car by waving is relying on the motorist recognising the wave as intended to make him stop. The latter, but not the former, is an example of meaningNN. Avoiding the problem he identified in Stevenson’s work, that of being unable to offer a full account of particular, context-bound uses, Grice proposes to describe meaningNN in relation to a speaker and a particular occasion in the first place, and the meaningNN of words and phrases as secondary, and derived from this. He tentatively suggests that ‘x meansNN that so-and-so’ might be equated with some statement ‘about what “people” (vague) intend (with qualifications about “recognition”) to effect by x’.23 In the concluding paragraphs of his article, Grice alludes briefly to two further aspects of meaning that draw on the questions he was posing himself in his notes and that were to prove central to his later work. First, although he is dealing primarily with specific instances of meanings on particular occasions, his account of meaningNN is not intended to cover the total significance potentially associated with the use of an expression. He draws a distinction between the ‘primary’ intention of an utterance, relevant to its meaningNN, and any further

1403_902976_05_cha04.qxd 12/13/04 6:26 PM Page 74

74 Paul Grice, Philosopher and Linguist

intentions: ‘If (say) I intend to get a man to do something by giving him some information, it cannot be regarded as relevant to the meaningNN of my utterance to describe what I intend him to do.’24 Even though conventional meaning is to follow from intention, Grice hints that some types of intention, possibly highly relevant to how an utterance is received, cannot properly be described as part of the meaningNN of that utterance. In his published paper, Grice is noncommittal about the precise number of distinctions between ‘levels’ of significance needed, and his rough notes suggest that this was an area in which he was unsure. Second, the hearer may sometimes look to specific context to determine the precise intention behind an utterance; he may consider, for instance, which of two possible interpretations would be the most relevant to what has gone before, or would most obviously fit the speaker’s purpose. Grice notes that such criteria are not confined to linguistic examples: Context is a criterion in settling the question of why a man who has just put a cigarette in his mouth has put his hand in his pocket; relevance to an obvious end is a criterion in settling why a man is running away from a bull.25 Grice’s suggestion that conventional meaning can be defined in terms of intentions has been widely praised. The philosopher Jonathan Bennett suggests that the idea of meaning as a kind of intending was commonplace, but that ‘Grice’s achievement was to discover a defensible version of it.’26 Grice’s postgraduate student and later colleague Stephen Schiffer went further, describing ‘Meaning’ as ‘the only published attempt ever made by a philosopher or anyone else to say precisely and completely what it is for someone to mean something’.27 However, commentators were not slow to point out the problems they saw as inherent in such an account. Responses to ‘Meaning’ have been numerous, and ranged over more than 40 years, but some raising the most fundamental objections were from Oxford in the years immediately following its publication. Austin did not produce a formal response, and Grice has suggested that they never even discussed it informally. This, he suggests, was not at all unusual. Austin rarely discussed with Grice, or with other colleagues, views that had become established through publication. He preferred his conversations to be about new directions and new channels of thought.28 To those Oxford philosophers who did discuss ‘Meaning’ in print, a comparison with Austin’s account of ‘speech acts’ was almost inevitable.

1403_902976_05_cha04.qxd 12/13/04 6:26 PM Page 75

Meaning 75

Austin’s version of speaker meaning was developed during the 1940s and 1950s. Various versions of it appeared in lectures and articles during this period, but the most complete account was presented when Austin was invited to deliver the prestigious William James lectures at Harvard in 1955. An edited and expanded version of his notes from these was published posthumously as How to do Things with Words. Austin’s ideas changed and developed over the course of time, and even during the lectures themselves, but some basic fundamental principles can be identified. These drew on his notion of the ‘descriptive fallacy’, the idea that linguistic study is ill served by the traditional philosopher’s focus on language as a tool for making statements capable of being judged to be either true or false. Throughout the development of speech act theory, Austin’s assumptions were very different from Grice’s working hypothesis in ‘Meaning’ that language is primarily used to make informative, descriptive statements. For Austin, making statements of fact was only one, and not a particularly significant one, of the many different things people do with words. He proposed to analyse the range of functions language can be used to perform. Initially, Austin concentrates on explicit ‘performatives’, statements that appear to describe but in fact bring about some state of affairs, usually a social fact or interpersonal relationship. He considers ritualised performatives, which require a very particular set of circumstances to be successful, such as ‘I name this ship the Queen Elizabeth’ (during the appropriate ceremony), ‘I do’ (at the relevant point in a marriage service) or ‘I give and bequeath my watch to my brother’ (in a will). He then expands his field to include more every day examples, such as ‘I bet you sixpence it will rain tomorrow’, and ‘I promise to . . .’. Characteristically, Austin proposes ‘going through the dictionary (a concise one should do) in a liberal spirit’ in search of performative verbs.29 Later, however, he realises that even this could not give an exhaustive list of what people do with words. People do things with all sorts of different types of utterances. Someone who utters, ‘Can you pass the salt?’ may well succeed in making a request by means of apparently asking a question. In response to his observations about the possible differences between the apparent form of an utterance and its actual function, Austin proposes to divide the analysis of meaning into three levels, or three separate acts performed in the course of a single utterance. The ‘locutionary act’ is determined by the actual form of words uttered, established by their sense and reference: ‘ “meaning” in the traditional sense’.30 At the level of ‘illocutionary act’, the notion of force becomes relevant. It is

1403_902976_05_cha04.qxd 12/13/04 6:26 PM Page 76

76 Paul Grice, Philosopher and Linguist

determined by the matter of how the locution in question is being used, of what is being done in saying what is said. This can be seen as at least in part dependent on the individual intentions of the speaker, but these in turn depend on, and follow from, certain conventions of language use. So, in the example above, we can understand that the speaker intends to request that the hearer pass the salt, but only because there are conventions of use determining that questions of ability can be used with this force. The illocutionary act is not a simple consequence of the locutionary, but is concerned with conventions ‘as bearing on the special circumstances of the occasion of the issuing of the utterance’.31 Finally, and separately, it is necessary to consider the effect of the utterance. Austin is here perhaps making a concession to the ‘causal’ theories of meaning Grice considered but dismissed. Austin refers to the effect achieved by the utterance, but unlike Stevenson he sees this as just one component of meaning, not a sufficient definition. The ‘perlocutionary act’ is defined by this effect, which may coincide with what the speaker intended to bring about, but need not. The perlocutionary act, then, is not a matter of convention, and cannot be determined simply with reference to the words uttered. It is no surprise that Grice’s ‘Meaning’ should have been discussed in the light of Austin’s speech act theory. The two seem in many ways to cover similar ground, in that both oppose much of the traditional philosophical literature on meaning by insisting on the roles of context and of speaker intention. In fact, some commentators have been tempted to read Grice’s paper as an implicit response, or challenge, to speech act theory. As Grice himself has pointed out, there is no such direct link. ‘Meaning’ existed in almost its finished state before the ideas in How to do Things with Words were made public, or even in some cases developed. As he became aware of Austin’s ideas, Grice did recognise some connections to his own themes and preoccupations, but continued to find his own treatment of them more compelling. In particular, he remained convinced that an account of meaning needed to give prominence to psychological notions, and was unhappy with what he saw as Austin’s excessive dependence on conventions; Austin was ‘using the obscure to explain the obscure’.32 From this, it would seem that his main objection to speech act theory was that it did not offer much explanation of meaning. For Grice, the observable facts of the conventions of use are in need of explanation, and this is best done with reference to individual instances of intention. Austin, on the other hand, was seeking to explain these conventions simply by drawing attention to another, more complex level of conventions.

1403_902976_05_cha04.qxd 12/13/04 6:26 PM Page 77

Meaning 77

Peter Strawson picks up on exactly this difference between Grice and Austin when he reflects on ‘Meaning’ in his 1964 article ‘Intention and convention in speech acts’. His chief concern is the validity of Austin’s notion of illocutionary act, based as it is on conventions, both the conventions of the language that define the locutionary act, and the further conventions of use that mediate between this and illocutionary force. However, Strawson argues, there are many cases where the difference between locutionary and illocutionary act cannot be explained simply with reference to conventions. In certain circumstances, to say ‘the ice over there is very thin’ may well have the force of a warning, but there is no particular convention of language use specifying this. Similarly, the difference between an order and an entreaty may depend not on any identifiable convention, but on a range of factors concerned with the speaker’s state of mind and consequent presentation of the utterance. Strawson looks to Grice’s notion of non-natural meaning to explain this additional, non-conventional set of factors. His paraphrase of Grice’s account, expressed in terms of three separate intentions, was to prove instructive to future commentators: S non-naturally means something by an utterance x if S intends (i1) to produce by uttering x a certain response (r) in an audience A and intends (i2) that A shall recognise S’s intention (i1) and intends (i3) that this recognition on the part of A of S’s intention (i1) shall function as A’s reason, or part of his reason, for his response (r).33 On Strawson’s interpretation, one of the defining intentions is focused on Grice’s indication that the speaker provides the hearer with a ‘reason’ to think or act in a certain way. He proposes to modify Grice’s account before enlisting it to explain the notion of illocutionary force, because of certain counter-examples to which it is subject as it stands. It is possible to envisage contexts in which S utters x with exactly the three intentions described above, but in which it does not seem appropriate to say that S was trying to communicate with A. Strawson does not illustrate this problem with a specific example, but he suggests a schematic outline. S may intend to get A to believe something, fulfilling (i1). S may arrange fake evidence, or grounds for this belief, such that A can see the evidence, and indeed such that A can see S fabricating it. S knows that A will be perfectly aware that the evidence is faked, but assumes that A will not realise that S has this awareness. The intention behind this is for A to have the relevant belief precisely because he recognises that S intends him to have this belief; this is S’s (i2). Further, it is S’s intention

1403_902976_05_cha04.qxd 12/13/04 6:26 PM Page 78

78 Paul Grice, Philosopher and Linguist

that the recognition of (i1) will be the reason why A acquires the belief; indeed, A will have no other reason to form this belief, knowing that the evidence for it is faked. This fulfils (i3). In effect, an intentional account of meaning would have to include an extra layer of intention, because for S to be telling something to, or communicating with A, S must be taken to intend A to recognise the intention to make A recognise the intention to produce the relevant belief in A. In other words, S must have a further intention (i4) that A should recognise intention (i2). If S’s intention (i4), and hence intention (i2) is fulfilled, then A may be said to have understood the utterance, a definition Strawson links to Austin’s notion of ‘uptake’, and hence to the successful accomplishment of illocutionary force. Strawson argues that the complex intention (i4) is the necessary defining characteristic of saying something with a certain illocutionary force. He hints that his additional layer of intention may not in fact prove sufficient, and that ‘the way seems open to a regressive series of intentions that intentions should be recognised’.34 Grice’s definition of meaning as determined entirely by speaker intention is at odds with, or according to Strawson can be complementary to, Austin’s reliance on convention. But it also opposes him to the notion, common in philosophy and later in linguistics, that meaning is determined by the rules of the language in question, defined independently of any particular occasion of utterance. Austin does allow some space to this type of meaning, specifying that the usual sense and reference of the words uttered constitute the locutionary act. He does not, however, develop this notion, or explain its relationship to his more central concern with speaker meaning. Strawson seems to downplay the importance of linguistic meaning for Austin when he places him on the same general side as Grice in what he describes elsewhere as the ‘Homeric struggle’ between ‘the theorists of communicationintention and the theorists of formal semantics’.35 Unlike Austin, Grice’s more extreme arguments that meaning is to be explained entirely in terms of ‘communication-intention’ do not presuppose any notion of semantic meaning. This is an aspect of Grice’s theory on which other commentators, such as John Searle, have focused. Searle was a student at Oxford during the 1950s, where he was taught by both Austin and Strawson. He completed a doctoral dissertation on sense and reference in 1959 before returning to his native America. Well versed in both the traditional philosophy of meaning and Austin’s concentration on communication, he worked on the theory of speech acts, publishing a book on the subject in 1969 in which he classifies possible illocutionary acts into a restricted number of general categories. He

1403_902976_05_cha04.qxd 12/13/04 6:26 PM Page 79

Meaning 79

also introduces the notion of indirect speech acts, in which one illocutionary act is performed by means of the apparent performance of a different illocutionary act. Austin’s example ‘Can you pass the salt?’ succeeds as a request both because the apparent act, that of asking for information about the hearer’s ability, is not appropriate in the context, and because rules of language use specify that this form of words can be used to make a request. For Searle, language is essentially a rulegoverned form of behaviour. It is centred on the speech act, the basic unit of communication, but its rules include the specific semantic properties of its words and phrases. This particular point gives rise to Searle’s objections to Grice’s total reliance on individual speaker intention in explaining meaning. Searle does allow for intentions in his account; indeed he is interested in basing it on just such psychological concepts and proposes to develop Grice’s account in order to do so. In a later discussion he comments that his analysis in Speech Acts ‘was inspired by Grice’s work on meaning’.36 However, like Austin before him, he sees semantic rules as underlying the range of intentions with which a speaker can appropriately produce any particular utterance. Searle is much more explicit than Austin in specifying that linguistic rules come first and restrict intentional possibility. Unlike Grice, he therefore singles out linguistic acts as separate from general communication; in the case of language, meaning exists independent of intention, or of occasion of use. Searle illustrates this with his own counter-example to Grice. An American soldier is captured by Italian troops in the Second World War. Neither the soldier nor his captors speak German, and the soldier speaks no Italian. The soldier wishes to make the Italian troops believe that he is German. He utters the only German sentence he knows, which happens to be ‘Kennst du das Land wo die Zitronen blühen?’, a line of poetry remembered from childhood that translates as ‘Knowest thou the land where the lemon trees bloom?’. He hopes that his captors will guess that he must be saying ‘I am a German soldier’. The soldier intends to get his captors to believe he is a German soldier. He intends that they should recognise his intention to get them to believe this, and that this recognition should serve as their reason for entertaining that belief. All the circumstances are in place for a case of Grice’s meaningNN. Yet, Searle argues, it would surely be unreasonable to maintain that when the soldier says ‘Kennst du das Land wo die Zitronen blühen?’, he actually non-naturally means ‘I am a German soldier’. The rules of German do not allow this; the words mean ‘Knowest thou the land where the lemon trees bloom?’. Searle concludes that Grice’s account is inadequate as it

1403_902976_05_cha04.qxd 12/13/04 6:26 PM Page 80

80 Paul Grice, Philosopher and Linguist

stands because ‘Meaning is more than a matter of intention, it is also at least sometimes a matter of convention.’37 Like Strawson, Searle proposes to amend Grice’s intentional account in the light of his own counter-example. In effect, he argues that the notion of intention must be expanded to include convention. When speakers use words literally, part of their intention is that hearers recognise the communicative intention precisely because of the pre-existing rules for using the words chosen. Searle attempts to balance the twin effects of intention and convention in his account of meaning. The result is rather further removed from Grice than Searle acknowledges when he suggests that he is offering an ‘amendment’. It returns to the traditional idea of conventional, or linguistic, meaning as the basic location of meaning on which intention-based speaker meaning is dependent.38 The magnitude of the consequences of Searle’s attempt to introduce conventions into an intentional account is perhaps illustrative of the gulf between semantic and psychological approaches to meaning. It also highlights the scale of Grice’s undertaking in producing an account of meaning relying entirely on the latter. Stephen Schiffer was a graduate student in the 1960s when, as he has since described, he was ‘much taken with Grice’s program’.39 He was impressed by what he saw as Grice’s reduction of the semantic to the psychological, but also believed he had identified the inadequacies of this particular account of speaker meaning. In the 1972 book in which he attempted to develop and improve on Grice’s account, he notes that it is not possible to claim priority for conventional meaning and at the same time define meaning in terms of intention. To maintain that speaker meaning, ‘what S meant by x’, is primary, but also that speaker meaning could be in any way dependent on the meaning of x would be simply circular. Schiffer does not see the exclusive concentration on intention as necessarily problematic, or at least not for this reason: ‘To say this does not commit Grice to holding that one can say whatever one likes and mean thereby whatever one pleases to mean. One must utter x with the relevant intentions, and not any value of ‘x’ will be appropriate to this end: I could not in any ordinary circumstances request you to pass the salt by uttering “the flamingoes are flying south early this year”.’40 Schiffer’s own response is that conventional meaning does exist, but is secondary to intentional meaning, because it is built up in a speech community over time, resultant on the fulfilling of certain communicative intentions. The problems Schiffer identifies for Grice are not problems with the notion that meaning is best given a psychological rather than a seman-

1403_902976_05_cha04.qxd 12/13/04 6:26 PM Page 81

Meaning 81

tic definition, an assumption with which he agrees. Rather, Schiffer questions the adequacy of intention as the relevant psychological state. He echoes Strawson’s schematic counter-example to argue that it can be possible for all the intentions Grice defines to be in place, but for no act of communication to be achieved. Strawson’s example indicated that another layer of intention (Strawson’s i4) must be added to the definition. Schiffer argues explicitly for the consequence Strawson only suggested, namely that the need for regressive intentions will in fact prove to be infinite. He constructs a series of more and more elaborate counterexamples to show that the levels of intention necessary fully to define meaning can never be exhausted. In the first such example, S produces a cacophonous rendition of ‘Moon Over Miami’ with the intention of driving A out of the room. S’s intention is that A will leave the room not just because A cannot stand the noise, but because A recognises that S is trying to get rid of A. Further, S intends that A recognise the intention that A recognise the intention to get A to leave the room, because S wishes A to be aware of S’s disdain. ‘In other words’, Schiffer explains, ‘while A is intended to think that S intends to get rid of A by means of the repulsive singing, A is really intended to have as his reason for leaving the fact that S wants him to leave.’41 In order to rule out the unacceptable suggestion that by singing ‘Moon Over Miami’ S meant that A was to leave the room, Schiffer proposes another level of intention. Further, more complex, examples call for further such layers. For Schiffer, the infinitely regressive series of intentions that intentions be recognised that this causes makes for an unacceptable definition of meaning. For Schiffer the best hope for a psychological account of meaning lies in an appeal to a special type of knowledge. This type of knowledge occurs when two people are in the following relation to a piece of information, or to a proposition p. S and A both know that p; S knows that A knows that p; A knows that S knows that p; A knows that S knows that A knows that p; S knows that A knows that S knows that p; and so on without limit. Such a pattern also leads to an infinite regress but, Schiffer argues, a harmless one. It is just the sort of situation we do in fact find in everyday life, as when S and A are sitting facing each other with a candle in between. Schiffer coins the phrase ‘mutual knowledge*’ for the recursive set of knowing he describes. S and A mutually know* that there is a candle on the table. Once Grice’s account of meaning is adapted to refer to mutual knowledge*, the apparent counter-examples are all ruled inadmissible, and therefore no longer pose a problem. The intentions S has in producing an utterance must be ‘mutual known*’ by S and A for a true instance of meaning to take place. Mutual knowl-

1403_902976_05_cha04.qxd 12/13/04 6:26 PM Page 82

82 Paul Grice, Philosopher and Linguist

edge* with its infinite regress, is missing in each of the counterexamples. ‘Meaning’ might appear to be offering a definition of the concept of meaning by proposing to substitute the concept of intention: in the vein of the logical positivists to be ‘translating’ problematic sentences into those that do not contain the problem term. This was certainly Schiffer’s interpretation when he described Grice’s account as ‘reductionist’. However, more recent commentators have been keen to defend Grice against this interpretation. For instance, Giovanna Cosenza has suggested that the analysis would be reductionist ‘only if Grice had seriously considered speakers’ intentions and psychological states as more fundamental, more basic, either epistemologically or metaphysically, than all the concepts related to linguistic meaning’, and claimed that Grice did not hold this view.42 Anita Avramides has also argued that Grice’s account of meaning need not be seen as reductionist.43 Grice was certainly distancing himself from those accounts of what he would call ‘non-natural meaning’ in which the notion of convention was primary. However, he was somewhat tentative about the precise relation between the two central concepts; his belief that convention could be entirely subsumed within an intentional account is expressed more as a hope than as a conviction. He seems, further, to have been troubled by the exact relationships between the different messages potentially conveyed by a single utterance. Even if the definition of convention is to be dependent on intention, some messages are more closely or more obviously related to conventional meaning than others. He as yet had no formal account of how the ‘full significance’ of an utterance might be derived or calculated. Nor had he yet drawn a clear distinction between messages conveyed by the words uttered and messages conveyed by the very act of utterance. This is a distinction Schiffer describes by differentiating between ‘S meant something by (or in) producing (or doing) x’ and ‘S meant something by x’.44 A very similar point is made by Paul Ziff in his 1967 response to ‘Meaning’. Ziff is decidedly dismissive of Grice’s ‘ingenious intricate discussion’ of meaning and the ‘fog’ to which it gives rise.45 He argues that the meanings of words and phrases, and indeed the lack of meaning of nonsense ‘words’, are quite independent of any individual’s intention. Like Schiffer, he argues that ‘Grice seems to have conflated and confused “A meant something by uttering x” . . . with the quite different “A meant something by x”.’46 The responses and criticisms of his peers were to feed into Grice’s own thinking about meaning over the following years. His published output

1403_902976_05_cha04.qxd 12/13/04 6:26 PM Page 83

Meaning 83

during the 1950s was restricted to ‘In defence of a dogma’ and ‘Meaning’, hardly an impressive record even by the standards of the time. Peter Strawson was responsible for seeing both these papers into print, while those he left alone fell victim to Grice’s perfectionism. Some, like ‘Intentions and dispositions’ were never published, while others waited in manuscript for 30 or more years. Of these, ‘G. E. Moore and philosopher’s paradoxes’ and ‘Common sense and scepticism’ indicate a development in Grice’s thinking on the need to distinguish between what our words literally mean and what we mean by using those words. They also suggest something of Grice’s ambivalence towards the place of conventional meaning. Like much of his work, these papers developed from his teaching: the former from a lecture delivered in 1950 and the latter from joint classes with A. D. Woozley at the same time or even earlier. Grice’s project in these papers was a typically ‘ordinary language’ enterprise: the desire to tackle a familiar philosophical problem by means of a rigorous examination of the terms characteristically employed to discuss it. The problem in question was how to address scepticism. In response to G. E. Moore’s ‘defence from common sense’, the American philosopher Norman Malcolm had suggested that Moore was in effect approaching the problem via an appeal to ordinary language, although apparently without knowing it. We use expressions to refer to and to describe material objects, and we do so in a ‘standard’ or ‘ordinary’ way. To claim, as some philosophers have done, that there are no material objects, is therefore to ‘go against ordinary language’.47 In effect, such philosophers are claiming that some uses of ordinary language are self-contradictory, or always incorrect, an inadmissible claim because ‘ordinary language is correct language’.48 Grice challenges Malcolm’s interpretation of Moore. Malcolm draws an unwarranted polarity between ‘ordinary’ and ‘self-contradictory’ in language use, he suggests. The two properties need not be incompatible; people do routinely say things that are literally self-contradictory or absurd. Furthermore, not every meaningful sentence would actually find a use in ordinary language. In ‘Common sense and scepticism’, he offers a striking example of what he means. ‘It would no doubt be possible to fill in the gaps in “The — archbishop fell down the — stairs and bumped — — like —,” with such a combination of indecencies and blasphemies that no one would ever use such an expression’, but we would not therefore want to treat the expression as self-contradictory.49 In ‘G. E. Moore and philosopher’s paradoxes’, he suggests that it is often necessary to acknowledge a difference between ‘what a given expression means (in

1403_902976_05_cha04.qxd 12/13/04 6:26 PM Page 84

84 Paul Grice, Philosopher and Linguist

general)’ and ‘what a particular speaker means, or meant, by that expression on a particular occasion’.50 Figurative or ironic utterances, for instance, involve ‘special’ uses of language. As a general definition, ‘what a particular speaker means by a particular utterance (of a statement-making character) on a particular occasion is to be identified with what he intends by means of the utterance to get his audience to believe.’51 For this intention to be successful, the speaker must at least rely on the audience’s familiarity with ‘standard’ or ‘general’ use, even if individual occasion meaning is to differ from this. In this way Grice is able to offer his own challenge to the sceptic. Faced with everyday statements about material objects, the sceptic is forced to claim either that, on particular occasions, people use expressions to mean things they have no intention of getting their audience to believe, or that people frequently use language in a way quite unlike its proper (‘general’) meaning, leaving the success of everyday communication unexplained. Although in general agreement with Malcolm’s aim in refuting scepticism, Grice is unhappy with the apparent simplifications, and some of the implications, of his argument. He replaces it with his own more sophisticated argument from ordinary language, an argument depending in particular on a detailed examination of the nature of ‘meaning’.

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 85

5 Logic and Conversation

As Grice’s enthusiasm for ordinary language philosophy became increasingly qualified during the 1950s, his interest was growing in the rather different styles of philosophy of language then current in America. Recent improvements in communications had made possible an exchange of ideas across the Atlantic that would have been unthinkable before the war. W. V. O. Quine had made a considerable impression at Oxford during his time as Eastman Professor. Grice was interested in Quine’s logical approach to language, although he differed from him over certain specific questions, such as the viability of the distinction between analytic and synthetic statements. Quine, who was visiting England for a whole year, and who brought with him clothes, books and even provisions in the knowledge that rationing was still in force, travelled by ship.1 However, during the same decade the rapid proliferation of passenger air travel enabled movement of academics between Britain and America for even short stays and lecture tours. Grice himself made a number of such visits, and was impressed by the formal and theory-driven philosophy he encountered. Most of all he was impressed by the work of Noam Chomsky. It may seem surprising that the middle-aged British philosopher interested in the role of individual speakers in creating meaning should cite as an influence the young American linguist notorious for dismissing issues of use in his pursuit of a universal theory of language. But Grice was inspired by Chomsky’s demonstration in his work on syntax of how ‘a region for long found theoretically intractable by scholars (like Jesperson) of the highest intelligence could, by discovery and application of the right kind of apparatus, be brought under control’.2 Less formally, he expressed admiration for an approach that did not offer ‘piecemeal reflections on language’ but rather where ‘one got a picture 85

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 86

86 Paul Grice, Philosopher and Linguist

of the whole thing’.3 Chomsky’s first and highly influential book Syntactic Structures was well known among Grice’s Oxford contemporaries. The Play Group worked their slow and meticulous way through it during the autumn of 1959.4 Austin, in particular, was extremely impressed. Grice characterised and perhaps parodied him as revering Chomsky for his sheer audacity in taking on a subject even more sacred than philosophy: the subject of grammar.5 Grice’s own interest was focused on theory formation and its philosophical consequences. Chomsky was taking a new approach to the study of syntax by proposing a general theory where previously there had been only localised description and analysis. He claimed, for instance, that ideally ‘a formalised theory may automatically provide solutions for many problems other than those for which it was explicitly designed’.6 Grice’s aim, it was becoming clear, was to do something similar for the study of language use. Meanwhile, ordinary language philosophy itself was in decline. As for any school of thought, it is difficult to determine an exact endpoint, and some commentators have suggested a date as late as 1970. However, it is generally accepted that the heyday of ordinary language philosophy was during the years immediately following the Second World War. The sense of excitement and adventure that characterised its beginning began to wane during the 1950s. Despite his professed dislike of discipleship, Austin seems to have become anxious about what he perceived as the lack of a next generation of like-minded young philosophers at Oxford. It became an open secret among his colleagues that he was seriously contemplating a move to the University of California, Berkeley.7 No final decision was ever made. Austin died early in 1960 at the age of 48, having succumbed quickly to cancer over the previous months. Reserved and private to the last, he hid his illness from even his closest colleagues until he was unable to continue work. His death was certainly a blow to ordinary language philosophy, but it would be an exaggeration to say that it was the immediate cause of its demise. Grice, who seems to have been regarded as Austin’s natural deputy, stepped in as convenor of the Play Group, which met under his leadership for the next seven years. Individuals such as Strawson, Warnock, Urmson and Grice himself continued to produce work with recognisably ‘ordinary language’ leanings throughout the 1960s. Grice’s interests at this time were not driven entirely by philosophical trends in Oxford and America; he was also turning his attention to some very old logical problems. In particular, he was interested in questions concerning apparent counterparts to logical constants in natural language. For instance, in the early 1960s he revisited a theme he had

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 87

Logic and Conversation 87

first considered before the war, when he gave a series of lectures on ‘Negation’. In these, he concerns himself with the analysis of sentences containing ‘not’, and with the extent to which this should coincide with a logical analysis of negation. Consideration of a variety of example sentences leads him to reject the simple equation of ‘not’ with the logical operation of switching truth polarity, usually positive to negative. He argues that ‘it might be said that in explaining the force of “not” in terms of “contradictory” we have oversimplified the ordinary use of “not”.’ In another lecture from the series he suggests that the lack of correspondence between ‘not’ and contradiction ‘might be explained in terms of pragmatic pressures which govern the use of language in general’.8 Grice was hoping to find not just an account of the uses of this particular expression, but a general theory of language use capable of extension to other problems in logic. He would have been familiar enough with such problems. The discussion of some of them dates back as far as Aristotle, in whose work he was well read even as an undergraduate. In Categoriae, Aristotle describes not just categories of lexical meaning, but also the types of relationships holding between words. To the modern logician, the use of terms in the following passage may be obscure, but the relationship of logical entailment is easily recognisable. One is prior to two because if there are two it follows at once that there is one whereas if there is one there are not necessarily two, so that the implication of the other’s existence does not hold reciprocally from one.9 The relationship between ‘two’ and ‘one’, or indeed between any two cardinal numbers where one is greater than the other, is one of asymmetrical entailment. ‘Two’ entails ‘one’, but ‘one’ does not entail ‘two’. A similar relationship holds between a superordinate and any of its hyponyms, or between a general and a more specific term. To use Aristotle’s example: ‘if there is a fish there is an animal, but if there is an animal there is not necessarily a fish.’10 The asymmetrical nature of this relationship means that use of the more general term tells us nothing at all about the applicability of the more specific. Aristotle also considers the relative acceptability of general and specific terms, and in doing this he goes beyond a narrowly logical focus. For if one is to say of the primary substance what it is, it will be more informative and apt to give the species than the genus. For example,

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 88

88 Paul Grice, Philosopher and Linguist

it would be more informative to say of the individual man that he is a man than that he is an animal (since the one is more distinctive of the individual man while the other is more general).11 Applying the term ‘animal’ to an individual tells us nothing about whether that individual is a man or not. Therefore, if the more specific term ‘man’ applies it is more ‘apt’, because it gives more information. This same point arises in a discussion of the applicability of certain descriptions later in Categoriae. Aristotle suggests that: ‘it is not what has not teeth that we call toothless, or what has not sight blind, but what has not got them at the time when it is natural for it to have them.’12 A term such as ‘toothless’ is only applied, because it is only informative, in those situations when it might be expected not to apply. Here, again, the discussion of what ‘we call’ things goes beyond purely logical meaning to take account of how expressions are generally used. Logically speaking a stone could appropriately be described as toothless or blind; in actual practice it is very unlikely to be so described. Grice’s self-imposed task in considering the general ‘pragmatic pressures’ on language use was, at least in part, one of extending Aristotle’s sensitivity to the standard uses of certain expressions, and examining how regularities of use can have distorting effects on intuitions about logical meaning. He was by no means the first philosopher to consider this. For instance, John Stuart Mill, in his response to the work of Sir William Hamilton, draws attention to the distinction between logic and ‘the usage of language’.13 He reproaches Hamilton for not paying sufficient attention to this distinction, and suggests that this is enough to explain some of Hamilton’s mistakes in logic. Mill glosses Hamilton as maintaining that ‘the form “Some A is B” . . . ought in logical propriety to be used and understood in the sense of “some and some only” ’.14 Hamilton is therefore committed to the claim that ‘all’ and ‘some’ are mutually incompatible: that an assertion involving ‘some’ has as part of its meaning ‘not all’. This is at odds with the observations on quantity in Categoriae and indeed, as Mill suggests, with ‘the practice of all writers on logic’. Mill explains this mistake as a confusion of logical meaning with a feature of ‘common conversation in its most unprecise form’. In this, he is drawing on the extra, non-logical but generally understood ‘meanings’ associated with particular expressions. In a passage that would not be out of place in a modern discussion of linguistics, Mill suggests that:

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 89

Logic and Conversation 89

If I say to any one, ‘I saw some of your children to-day,’ he might be justified in inferring that I did not see them all, not because the words mean it, but because, if I had seen them all, it is most likely that I should have said so.15 Mill draws a distinction between what ‘words mean’ and what we generally infer from hearing them used. In what can be seen as an extension of Aristotle’s discussion of ‘aptness’, he argues that it is a mistake to confuse these two very different types of significance. A more specific word such as ‘all’ is more appropriate, if it is applicable, than a more general word such as ‘some’. Therefore, the use of the more general leads to the inference, although it does not strictly mean, that the more specific does not apply. ‘Some’ suggests, but does not actually entail ‘not all’. Besides his interest in logical problems with a venerable pedigree, Grice was also concerned with issues familiar to him from the work of recent or contemporary philosophers. In both published work and informal notes he frequently lists these and arranges them in groups. Part of his achievement in the theory he was developing lay in seeing connections between an apparently disparate collection of problems and countenancing a single solution for them all. For instance, in Concept of Mind, Ryle argues that, although the expressions ‘voluntary’ and ‘involuntary’ appear to be simple opposites, they both require a particular condition for applicability, namely that the action in question is in some way reprehensible. If they were simple opposites, it should always be the case that one or other would be correct in describing an action, yet in the absence of the crucial condition, to apply either would be to say something ‘absurd’. Similarly, although if someone has performed some action, that person must in a sense have tried to perform it, it is often extremely odd to say so. In cases where there was no difficulty or doubt over the outcome, it is inappropriate to say that someone tried to do something: so much so that some philosophers, such as Wittgenstein, have claimed that it is simply wrong. Another related problem is familiar from Austin’s work; it is the one summed up in his slogan ‘no modification without aberration’. For the ordinary uses of many verbs, it does not seem appropriate to apply either a modifying word or phrase or its opposite. Austin was therefore offering a generalisation that includes, but is not restricted to, Ryle’s claims about ‘voluntary’ and ‘involuntary’. For many everyday action verbs, the act described must have taken place in some non-standard way for any

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 90

90 Paul Grice, Philosopher and Linguist

modification appropriately to apply. Austin offers no theory based on this observation, and indeed Grice was unimpressed by it even as a generalisation; he claimed in an unpublished paper that it was ‘clearly fraudulent’. ‘No “aberration” is needed for the appearance of the adverbial “in a taxi” within the phrase “he travelled to the airport in a taxi”; aberrations are needed only for modifications which are corrective qualifications.’16 Grice’s general account of language, conceived with the twin ambitions of refining his philosophy of meaning and of explaining a diverse range of philosophical problems, gradually developed into his theory of conversation. Like his project in ‘Meaning’, this draws on a ‘common-sense’ understanding of language: in this case, that what people say and what they mean are often very different matters. This observation was far from original, but Grice’s response to it was in some crucial ways entirely new in philosophy. Unlike formal philosophers such as Russell or the logical positivists, he argued that the differences between literal and speaker meaning are not random and diverse, and do not make the rigorous study of the latter a futile exercise. But he also differed from contemporary philosophers of ordinary language, in arguing that interest in formal or abstract meaning need not be abandoned in the face of the particularities of individual usage. Rather, the difference between the two types of meaning could be seen as systematic and explicable, following from one very general principle of human behaviour, and a number of specific ways in which this worked out in practice. In effect, the use of language, like many other aspects of human behaviour, is an end-driven endeavour. People engage in communication in the expectation of achieving certain outcomes, and in the pursuit of those outcomes they are prepared to maintain, and expect others to maintain, certain strategies. This mutual pursuit of goals results in cooperation between speakers. This manifests itself in terms of four distinct categories of behaviour, each of which can be summarised by one or more maxims that speakers observe. The categories and maxims are familiar to every student of pragmatics, although in later commentaries they are often all subsumed under the title ‘maxims’. Category of Quantity 1. 2.

Make your contribution as informative as is required (for the current purposes of the exchange). Do not make your contribution more informative than is required.

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 91

Logic and Conversation 91

Category of Quality 1. 2.

Do not say what you believe to be false. Do not say that for which you lack adequate evidence.

Category of Relation 1.

Be relevant.

Category of Manner 1. 2. 3. 4.

Avoid ambiguity of expression. Avoid ambiguity. Be brief. Be orderly.17

Grice uses the simple notion of cooperation, together with the more elaborate structure of categories, to offer a systematic account of the many ways in which literal and implied meaning, or ‘what is said’ and ‘what is implicated’, differ from one another. In effect, the expectation of cooperation both licenses these differences and explains their usually successful resolution. Speakers rely on the fact that hearers will be able to reinterpret the literal content of their utterances, or fill in missing information, so as to achieve a successful contribution to the conversation in hand. The noun ‘implicature’ and verb ‘implicate’ (as used in relation to that noun) are now familiar in the discussion of pragmatic meaning, but they were coined by Grice, and coined fairly late on in the development of his theory. In early work on conversation he suggested that a ‘special kind of implication’ could be used to account for various differences between conventional meaning and speaker meaning. He ultimately found this formulation inadequate, together with a host of other words such as ‘suggest’, ‘hint’ and even ‘mean’, precisely because of their complex pre-existing usage both within and outside philosophy. The difference between saying and meaning had an established philosophical pedigree, as well as a basis in common sense. A literature dating back to G. E. Moore, but largely concentrated in the 1950s and 1960s, offers a significant context to Grice’s work. Little of this is now read and none of it attempts anything as ambitious as a generalising theory of communication. With typical laxness, Grice does not refer to this body of work in any of his writings on the topic, although he must have been aware of it; much of it originated from the tight circle of Oxford philosophy. Writing for his fellow philosophers, and in the conventions of

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 92

92 Paul Grice, Philosopher and Linguist

his time, he was able to refer vaguely and generally to the debate in the expectation that the relevant context would be recognised. G. E. Moore had argued that when we use an indicative sentence we assert the content of the sentence, but we also imply personal commitment to the truthfulness of that content; ‘If I say that I went to the pictures last Tuesday, I imply by saying so that I believe or know that I did, but I do not say that I believe or know this.’18 This distinction was picked up by a number of philosophers in the 1940s and 1950s. In 1946, Yehoshua Bar-Hillel discussed the sense of ‘imply’ identified by Moore, proposing to describe it as ‘pragmatical’.19 Bar-Hillel identifies himself as a supporter of ‘logical empiricism’ (he quotes approvingly a comment from Carnap to the effect that natural languages are too complex and messy to be the focus of rigorous scientific enquiry) and his article is aimed explicitly at rejecting philosophy of the ‘analytic method’ in general and of Moore in particular. However, he suggests that by using sentences that are ‘meaningless’ to logical empiricists, such as the sentences of metaphysics or aesthetics, ‘one may nevertheless imply sentences which are perfectly meaningful, according to the same criteria, and are perhaps even true and highly important’.20 A full account of the pragmatic sense of ‘implies’ might, he predicts, prove highly important in discussing linguistic behaviour. A few years later D. J. O’Connor contrasted the familiar ‘logical paradoxes’ of philosophy with less well known ‘pragmatic paradoxes’. O’Connor draws attention to certain example sentences (such as ‘I believe there are tigers in Mexico but there aren’t any there at all’) which, although not logically contradictory or necessarily false, are pragmatically unsatisfactory. His purpose is to distinguish between logic and language and to urge philosophers to attempt to ‘make a little clearer the ways in which ordinary language can limit and mislead us’.21 The philosophical significance of implication was also discussed by those more sympathetic to ordinary language, in particular by members of the Play Group. Writing in Mind in 1952, J. O. Urmson acknowledged the ‘implied claim to truth’ in the use of indicative statements, but felt the need to clarify his terms: ‘The word “implies” is being used in such a way that if there is a convention that X will only be done in circumstances Y, a man implies that situation Y holds if he does X.’22 Urmson suggests the addition of an ‘implied claim to reasonableness’; ‘it is, I think, a presupposition of communication that people will not make statements, thereby implying their truth, unless they have some ground, however tenuous, for those statements.’23 Read with hindsight, Urmson’s suggestion looks like a hint towards Grice’s second maxim of

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 93

Logic and Conversation 93

Quality. Another member of the Play Group, P. H. Nowell-Smith, appears to add something like a maxim of Relevance, a notion that Grice discusses as early as the original version of ‘Meaning’. In his book Ethics, Nowell-Smith suggests that there are certain ‘contextual implications’ that generally accompany the use of words, but that are dependent on particular contexts. Without reference to Urmson, he phrases these as follows: When a speaker uses a sentence to make a statement, it is contextually implied that he believes it to be true. . . . A speaker contextually implies that he has what he himself believes to be good reasons for his statement. . . . What a speaker says may be assumed to be relevant to the interests of his audience.24 For Nowell-Smith the contextual rules are primary, and do not operate with reference to logical meaning; in effect there is no ‘what is said’. In fact, he specifies that logical meaning is a subclass of contextual implication, because it is meaning we are entitled to infer in any context whatsoever. Other philosophers concerned with implication, however, did identify different levels of meaning. In The Logic of Moral Discourse, 1955, Paul Edwards distinguishes between the referent of a sentence (the fact that makes a sentence true or, in its absence, false: in effect the truth-conditions) and ‘what the sentence expresses’ (anything it is possible to infer about the speaker on the basis of the utterance).25 Some attempted syntheses of thinking about implication lacked much in the way of generalising or explanatory force. In his 1958 article ‘Pragmatic implication’, C. K. Grant’s account of a wide variety of linguistic phenomena leads him to the claim that ‘a statement pragmatically implies those propositions whose falsity would render the making of the statement absurd, that is, pointless.’26 What is implied by a statement of p, he insists, is not just the speaker’s belief in p, but some further proposition. Isobel Hungerland, writing in 1960, reviews the range of attempts to mediate between asserted and implied meaning and exclaims ‘What a range of rules!’.27 Her suggestion of a generalisation across these rules is that ‘a speaker in making a statement contextually implies whatever one is entitled to infer on the basis of the presumption that his act of stating is normal.’28 Grice was working on his own generalisation, which was to link together the range of principles of language use into one coherent account, and to give this a place in a more ambitious theory of language. If the maxims of Quality and of Relation were familiar from con-

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 94

94 Paul Grice, Philosopher and Linguist

temporary work in Ethics, those of Quantity and Manner drew on Grice’s long-standing concern about stronger and weaker statements. This was the idea that had interested him in his rough notes on ‘visa’ for his work on perception with Geoffrey Warnock. The earliest published indication appeared in 1952, in a footnote to Peter Strawson’s Introduction to Logical Theory. In discussing the relationship between the statement ‘there is not a book in his room which is not by an English author’ and the assumption ‘there are books in his room’, Strawson draws attention to the need to distinguish between strictly logically relations and the rules of ‘linguistic conduct’.29 He suggests as one such rule: ‘one does not make the (logically) lesser when one could truthfully (and with equal or greater linguistic economy) make the greater, claim.’ It would be misleading, although not strictly false, to make the less informative claim about English authors if in a position to make the much more informative claim that there are no books at all. Strawson acknowledges that ‘the operation of this “pragmatic rule” was first pointed out to me, in a different connection, by Mr H. P. Grice.’ It was typical of Grice that he did not publish his own account of this pragmatic rule until almost ten years after Strawson’s acknowledgement. ‘The causal theory of perception’ appeared in the Proceedings of the Aristotelian Society, after Grice delivered it as a symposium paper at the meeting of the Society in Cambridge in July 1961. His main focus is his long-standing interest in the status of our observations of material objects. Grice offers tentative support for a causal account of perception. Theories of this type date back at least to John Locke. They maintain that we directly perceive through our faculties of sense a series of information, given the title ‘sense data’ by later philosophers, and infer that these are caused by material objects in the world. For Locke, the inference was justifiable, but his account of perception, and causal theories ever since, had been dogged by accusations of scepticism. If we have no direct access to material objects, but only to the sense data they putatively cause, the way is left open to doubt the independent existence of such objects. Grice’s defence is tentative, even by his own circumspect standards. He proposes to advance a version of the causal theory ‘which is, if not true, at least not obviously false’.30 Some of the defence itself is presented in quotation marks, as a fictional ‘opponent’ argues with a fictional ‘objector’ to the causal theory, implicitly distancing Grice from the arguments on either side. This tentativeness might in part be explained by the extreme heresy of Grice’s enterprise for a philosopher of ordinary language. Some of Austin’s most strident remarks about ordi-

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 95

Logic and Conversation 95

nary language had appeared in his attack in Sense and Sensibilia on philosophical theories of perception, and their reliance on the obscure and technical notion of ‘sense data’. Austin’s chief target here had been phenomenalism which, in claiming that sense data are the essence of what we think of as material objects, is often seen as a direct competitor to the causal theory of perception. Although defending one of phenomenalism’s rivals, Grice was suggesting reasons for retaining Austin’s particular bugbear, ‘sense data’, as a term of art. Indeed, he even goes so far as to hint that his own, speculative version of the causal theory ‘neither obviously entails nor obviously conflicts with Phenomenalism’.31 Grice defends sense data because, as he argues, the sceptic who wants to question the existence of the material world needs to make use of them and the sceptic’s position at least deserves to be heard. He defends it also because of his own ideas about language use. Indeed, his essay on perception seems at times in danger of being hijacked by his interest in language. This is not entirely out of keeping with the subject; through Locke and Berkeley to Ayer and Austin, the philosophy of perception had focused on the correct interpretation of sentences in which judgements of perception are expressed. However, Grice makes no secret of his hope that his interpretation of such sentences will fit with views of language and use motivated by other considerations. Much of the essay is taken up with a discussion of general principles governing the use of language. This is inspired by a standard argument for retaining ‘sense data’ as a technical term; it would appear to underlie ordinary expressions such as ‘so-and-so looks F [e.g. blue] to me’. In effect, a suitably rigorous account of the meaning of such expressions would need to include reference to something such as a sense datum that corresponds to the subjective experience described. A potential objector might point out that such expressions are not appropriate to all instances of perception; they would only in fact be used in contexts where a condition of either doubt or denial (D-or-D) as to the applicability of F holds. ‘There would be something at least prima facie odd about my saying “That looks red to me” (not as a joke) when I am confronted by a British pillar box in normal daylight at a range of a few feet.’32 Grice considers two possible ways of explaining this oddity. It could be seen as ‘a feature of the use, perhaps of the meaning’ of such expressions that they carry the implication that the D-or-D condition is fulfilled. Use of such expressions in the absence of the condition would therefore constitute a misuse, and would fail to be either true or false.

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 96

96 Paul Grice, Philosopher and Linguist

Alternatively, such statements could be seen as true whenever the property F applies but, in the absence of the D-or-D condition, severely misleading. Use in the absence of the condition would therefore constitute a more general error, because of ‘a general feature or principle of the use of language’.33 Grice describes how until recently he had been undecided as to which of these two explanations to favour. He had, however, lent towards the second because it ‘was more in line with the kind of thing I was inclined to say about other linguistic phenomena which are in some degree comparable.’34 Towards the end of his discussion, Grice hesitantly suggests a first attempt at formulating the general principle of language use: ‘One should not make a weaker statement rather than a stronger one unless there is a good reason for so doing.’35 This is, of course, remarkably similar to the observation attributed to him by Strawson a decade earlier. It seems that during the intervening years Grice had become increasingly impressed by how many different types of meaning this rule could explain, in other words by how general and simple an account of language use it suggested. In two interesting respects the formulation of the rule had changed. First, it had changed from a statement about what ‘one does’ in language use to what ‘one should do’. This brought it nearer to the first maxim of Quantity it was eventually to become. Second, the new formulation of the rule introduces possible exceptions, by alluding to ‘good reasons’ for breaking it. Again, Grice does not elaborate on this qualification, but the consideration of reasons for breaching the maxims, together with a discussion of the consequences of doing so, was to be vital to the subsequent development of his work. It was what was to give it explanatory and generalising abilities beyond those of a simple list of ‘rules’ of linguistic behaviour. Grice’s claims about the use of language offer support to the defender of sense data. They allow the defender to describe statements of the ‘so-and-so looks F to me’ type as strictly true regardless of whether the D-or-D condition holds. If such statements call for the use of sense data as a technical term to explain their meaning, then sense data are applicable to any description of perception. It is simply that in most non-controversial contexts such statements will be avoided because, although perfectly true, they will introduce misleading suggestions. Here, perhaps, was Grice’s greatest heresy. Austin’s rejection of sense data relied on an appeal to what people ordinarily say, together with an assumption that this is the best guide to meaning and truth. Grice’s tentative support for it relies on a distinction between what people can truthfully say, and what they actually, realistically do say. Despite its

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 97

Logic and Conversation 97

emphasis on linguistic use, ‘The causal theory of perception’ was well received as a contribution to the philosophy of perception. It was anthologised in 1965 in Robert Swartz’s Perceiving, Sensing and Knowing. In 1967, Grice’s sometime collaborator Geoffrey Warnock included the essay in his volume on The Philosophy of Perception, singling it out for praise as an ‘exceedingly ingenious and resourceful contribution’.36 As was the case throughout his working life, Grice made no firm distinction between his teaching and the development of his own philosophical ideas. A series of lecture notes from the 1960s on topics relating to speaker meaning and context therefore provide an insight into the lines along which his thinking was developing, as do some rough notes from the same period. What his students gained from the fresh ideas in Grice’s lectures, however, they paid for in his cavalier attitude to the topic and aims of the class. He begins one lecture, which is entitled simply ‘Saying: Week 1’: Although the official title of this class is ‘Saying’, let me say at once that we are unlikely to reach any direct discussion of the notion of saying for several weeks, and in the likely event of our failing to make any substantial inroads on the title topic this term, my present intention is to continue the class into next term.37 Grice’s interest in these lectures was in finding out what can be learnt about speaker meaning, specifically about what sets it apart from linguistic meaning, from close attention to its characteristics and circumstances. The opening of another of the lectures, entitled ‘The general theory of context’, tells rather more of his purpose and method than is made explicit in much of the later, published work. Philosophers often say that context is very important. Let us take this remark seriously. Surely, if we do, we shall want to consider this remark not merely in its relation to this or that problem, i.e. in context, but also in itself, i.e. out of context. If we are to take this seriously, we must be systematic, that is thorough and orderly. If we are to be orderly we must start with what is relatively simple. Here, though not of course everywhere, to be simple is to be as abstract as possible; by this I mean merely that we want, to begin with, to have as few cards on the table as we can. Orderliness will then consist in seeing first what we can do with the cards we have; and when we think that we have exhausted this investigation, we put another card on the table, and see what that enables us to do.

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 98

98 Paul Grice, Philosopher and Linguist

It is not hard to discern Austin’s influence here, in the insistence that a particular philosophical commonplace, if it is to be accepted as a useful explanation, must first be subject to a rigorous process of analysis. The call for system and order, however, is Grice’s own. He had reacted against precisely the tendency towards unordered, open-ended lists in Austin’s work. He argues in these lectures that thinking seriously about context means thinking about conversation; this is the setting for most examples of speaker meaning. He proposes, therefore, to compile an account of some of the basic properties common to conversations generally. His method of limiting his hand was to result in certain highly artificial simplifications, but he made these simplifications deliberately and knowingly. For instance, the relevant context was to be assumed to be limited to what he calls the ‘linguistic environment’: to the content of the conversation itself. Conversation was assumed to take place between two people who alternate as speaker and hearer, and to be concerned simply with the business of transferring information between them. A number of the lectures include discussion of the types of behaviour people in general exhibit, and therefore the types of expectations they might bring to a venture such as a conversation. Grice suggests that people in general both exhibit and expect a certain degree of helpfulness from others, usually on the understanding that such helpfulness does not get in the way of particular goals, and does not involve undue effort. If two people, even complete strangers, are going through a gate, the expectation is that the first one through will hold the gate open, or at least leave it open, for the second. The expectation is such that to do otherwise without particular reason would be interpreted as deliberately rude. The type of helpfulness exhibited and expected in conversation is more specific because of a particular, although not a unique, feature of conversation; it is a collaborative venture between the participants. At least in the simplified version of conversation discussed in these lectures, there is a shared aim or purpose. However, an account of the particular type of helpfulness expected in conversation must be capable of extension to any collaborative activity. In his early notes on the subject, Grice considers ‘cooperation’ as a label for the features he was seeking to describe. Does ‘helpfulness in something we are doing together’, he wonders in a note, equate to ‘cooperation’? He seems to have decided that it does; by the later lectures in the series ‘the principle of conversational helpfulness’ has been rebranded the expectation of ‘cooperation’. During the Oxford lectures Grice develops his account of the precise nature of this cooperation. It can be seen as governed by certain regu-

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 99

Logic and Conversation 99

larities, or principles, detailing expected behaviour. The term ‘maxim’ to describe these regularities appears relatively late in the lectures. Grice’s initial choices of term are ‘objectives’, or ‘desiderata’; he was interested in detailing the desirable forms of behaviour for the purpose of achieving the joint goal of the conversation. Initially, Grice posits two such desiderata: those relating to candour on the one hand, and clarity on the other. The desideratum of candour contains his general principle of making the strongest possible statement and, as a limiting factor on this, the suggestion that speakers should try not to mislead. The desideratum of clarity concerns the manner of expression for any conversational contribution. It includes the importance of expectations of relevance to understanding and also insists that the main import of an utterance be clear and explicit. These two factors are constantly to be weighed against two fundamental and sometimes competing demands. Contributions to a conversation are aimed towards the agreed current purposes by the principle of Conversational Benevolence. The principle of Conversational Self-Love ensures the assumption on the part of both participants that neither will go to unnecessary trouble in framing their contribution. Grice suggests that many philosophers are guilty of inexactness in their use of expressions such as ‘saying’, ‘meaning’ and ‘use’, applying them as if they were interchangeable, and in effect confusing different ways in which a single utterance can convey information. For instance, Grice refers back to the discussion at a previous class he had given jointly with David Pears, when the exact meaning of the verb ‘to try’ was discussed. This, of course, was one of the specific philosophical problems he was interested in accounting for by means of general principles of use. Stuart Hampshire had apparently claimed that if someone, X, did something, it is always possible to say that X tried to do it. This was challenged; in situations when there is no obvious difficulty or risk of failure involved it is inappropriate to talk of someone’s trying to do something. Grice’s answer had been that, while it is always true to say that X tried to do something, this may sometimes be a misleading way of speaking. If X succeeded in performing the act, it would be more informative and therefore more cooperative to say so. Therefore, an utterance of ‘X tried to do it’ will imply, but not actually say, that X did not succeed. In his consideration of the desiderata of conversational participation, Grice initially compiles a loose assemblage of features. By the later lectures these appear in more or less their final form under the categories Quantity, Quality, Relation and Manner (or, sometimes, Mode). In

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 100

100 Paul Grice, Philosopher and Linguist

arranging the desiderata in this way, Grice was presumably seeking to impose a formal arrangement on a diverse set of principles. But it seems that he had other motives: semi-seriously to echo the use of categories in such orthodox philosophies as those of Aristotle and Kant, and more importantly to draw on their ideas of natural, universal divisions of experience. The regularities of conversational behaviour were intended to include aspects of human behaviour and cognition beyond the purely linguistic. Grice’s collaborative work with Strawson had been concerned with Aristotle’s division of experience into ‘categories’ of substances. Aristotle’s original formulation of the list of such properties allows that they can take the form of ‘either substance or quantity or qualification or a relative or where or when or being-in-a-position or having or doing or being-affected’.38 He concentrates mainly on the first four, and these received most attention in subsequent developments of his work. They were the starting-point for Kant’s use of categories to describe types of human experience, and his argument that these form the basis of all possible human knowledge. In the Critique of Pure Reason, Kant proposes to divide the pure concepts of understanding into four main divisions: ‘Following Aristotle we will call these concepts categories, for our aim is basically identical with his although very distinct from it in execution.’39 These are categories ‘Of Quantity’, ‘Of Quality’, ‘Of Relation’ and ‘Of Modality’, with various subdivisions ascribed to each. Kant’s claims for both the fundamental and the exhaustive nature of these categories are explicit: This division is systematically generated from a common principle, namely the faculty for judging (which is the same as the faculty for thinking), and has not arisen rhapsodically from a haphazard search for pure concepts, of the completeness of which one could never be certain.40 Kant goes so far as to suggest that his table of categories, containing all the basic concepts of understanding, could provide the basis for any philosophical theory. These, therefore, offered Grice divisions of experience with a sound pedigree and an established claim to be universals of human cognition. Early in 1967, Grice travelled to Harvard to deliver that year’s William James lectures, the prestigious philosophical series in which Austin had put forward his theory of speech acts 12 years earlier. Grice’s entitled his lectures ‘Logic and conversation’. He was presenting his current thinking about meaning to an audience beyond that of his students and

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 101

Logic and Conversation 101

immediate colleagues and was clearly aware of the different assumptions and prejudices he could expect in an American, as opposed to an Oxford, audience. ‘Some of you may regard some of the examples of the manoeuvre which I am about to mention as being representative of an out-dated style of philosophy’, he suggests in the introductory lecture, ‘I do not think that one should be too quick to write off such a style.’41 Addressing philosophical concerns by means of an attention to everyday language was still a highly respectable, even an orthodox approach in Oxford. In America it was seen by at least some as belonging to an unsuccessful, and now rather passé, school of thought. In pleading its cause, Grice argues that it still has much to offer: in this case, the possibility of developing a theory to discriminate between utterances that are inappropriate because false, and those that are inappropriate for some other reason. Despite the difficulties inherent in such an ambitious scheme, and the well-known problems with the school of thought in question, he does not give up hope altogether of ‘systematizing the linguistic phenomena of natural discourse’. Grice’s ultimate aim in the lectures is ambitious and uncompromising; his interest ‘will lie in the generation of an outline of a philosophical theory of language’.42 He argues for a complex understanding of the significance of any utterance in a particular context; its meaning is not a unitary phenomenon. Conventional meaning has a necessary, but by no means a sufficient role to play. Indeed conventional meaning is itself not a unitary phenomenon. Some aspects of it involve the speaker in a commitment to the truth of a certain proposition; this is ‘what is said’ on any particular occasion. Other aspects may be associated by convention with the words used, but not be part of what the speaker is understood literally to have said. The examples ‘She was poor but honest’ and ‘He is an Englishman; he is, therefore, brave’ convey more than just the truth of the two conjuncts, more than would be conveyed by ‘She was poor and honest’ or ‘He is an Englishman and he is brave’. An idea of contrast is introduced in the first example and one of consequence in the second. These ideas are attached to the use of the individual words ‘but’ and ‘therefore’, but do not contribute to the truth-conditions of the sentences. We would not want to say that the sentences were actually false if both conjuncts were true, but we did not agree with the idea of contrast or of consequence. We might, rather, want to say that the speaker was presenting true facts in a misleading way. These examples demonstrate implicated elements associated with the conventional meaning of the words used, elements Grice labels ‘conventional implicatures’.

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 102

102 Paul Grice, Philosopher and Linguist

There is another level at which speaker meaning can differ from what is said, dependent on context or, for Grice, on conversation. In ‘conversational implicatures’ meaning is conveyed not so much by what is said, but by the fact that it is said. This is where the categories of conversational cooperation, and their various maxims, play their part. The onus on participants in a conversation to cooperate towards their common goal, and more particularly the expectation each participant has of cooperation from the other, ensures that the understanding of an utterance often goes beyond what is said. Faced with an apparently uncooperative utterance, or one apparently in breach of some maxim, a conversationalist will if possible ‘rescue’ that utterance by interpreting it as an appropriate contribution. In this way, Grice offers a more detailed account of the idea he explored in ‘Meaning’, and in his notes from that time: that there are three ‘levels’ of meaning, or three different degrees to which a speaker may be committed to a proposition. His model now includes, ‘what is said’, ‘conventional meaning’ (including conventional implicatures) and ‘what is conversationally implicated’. The presentation of the norms of conversational behaviour in the William James lectures is rather different from Grice’s handling of them in his earlier work. The maxims, or the categories they fall into, are no longer presented as the primary forces at work. Instead, all are assumed under a general ‘Principle of Cooperation’. The principle appeared late in the development of Grice’s theory. It enjoins speakers to: ‘Make your conversational contribution such as is required, at the stage at which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged.’43 The name ‘Cooperative Principle’ was even later; it was added using an omission mark in a manuscript copy of the second William James lecture. Grice may well have been attempting to give a name, and an exact formulation, to his previously rather nebulous idea of cooperation or ‘helpfulness’. However, the effect was to change what was presented as a series of ‘desiderata’, features of conversational behaviour participants might expect in their exchanges, to something looking like a powerful and general injunction to correct social behaviour. In the development of his theory of conversation, Grice was much exercised by the status of the categories as psychological concepts. He questioned whether the maxims were the result of entering into a quasicontract by engaging in conversation, simply inductive generalisations over what people do in fact do in conversation, or, as he suggested in one rough note, just ‘special cases of what a decent chap should do’. He remained undecided on this matter throughout the development of

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 103

Logic and Conversation 103

the theory, content to concentrate on the effects on meaning of the maxims, whatever their status. By the time of the William James lectures, however, he seems to be closer to an answer. He is ‘enough of a rationalist’ to want to find an explanation beyond mere empirical generalisation.44 The following suggestion results from this impetus: So I would like to be able to show that observation of the Cooperative Principle and maxims is reasonable (rational) along the following lines: that anyone who cares about the goals that are central to conversation/communication (such as giving and receiving information, influencing and being influenced by others) must be expected to have an interest, given suitable circumstances, in participation in talk exchanges that will be profitable only on the assumption that they are conducted in general accordance with the Cooperative Principle and the maxims.45 This is a wordy explanation, and also a troublesome one. It seems to create a loop linking the aim of explaining cooperation to an account of conversation as dependent on cooperation, a loop from which it does not successfully escape. The link between reasonableness and cooperation is far from explicit. Nevertheless, this passage offers Grice’s account of his own preferences in seeking an answer to the question over the status, and hence the motivation, for the Cooperative Principle. His preference, particularly his reference to ‘rational’ behaviour, was to prove important in the subsequent development of his work. However derived, the maxims operate to produce conversational implicatures in a number of different ways. In many cases, they simply ‘fill in’ the extra information needed to make a contribution fully cooperative. A says ‘Smith doesn’t seem to have a girlfriend these days’ and B replies, ‘He has been paying a lot of visits to New York lately’. B’s remark does not, as it stands, appear relevant to the preceding remark. But it is easy enough to supply the missing belief B must hold for the remark to be relevant. B conversationally implicates that Smith has, or may have a girlfriend in New York.46 In other cases the speaker seems to be far less cooperative, at least at the level of ‘what is said’. In order to be rescued as cooperative contributions to the conversation, such examples need to be not so much filled out as re-analysed. Because of the strength of the conviction that the speaker will, other things being equal, provide cooperative contributions, the other participant will put in the work necessary to reach such an interpretation. In perhaps his most famous example of con-

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 104

104 Paul Grice, Philosopher and Linguist

versational implicature, Grice suggests the case of a letter of reference for a candidate for a philosophy job that runs as follows: ‘Dear Sir, Mr X’s command of English is excellent, and his attendance at tutorials has been regular. Yours, etc.’ The information given is grossly inadequate; the writer appears to be seriously in breach of the first maxim of Quantity, enjoining the utterer to give as much information as is appropriate. However, the receiver of the letter is able to deduce that the writer, as the candidate’s tutor, must know more than this about the candidate. There must be some reason why the writer is reluctant to offer the extra information that would be helpful. The most obvious reason is that the writer does not want explicitly to comment on Mr X’s philosophical ability, because it is not possible to do so without writing something socially unpleasant. The writer is therefore taken conversationally to implicate that Mr X is no good at philosophy; the letter is cooperative not at the level of what is literally said, but at the level of what is implicated. In examples such as this a maxim is deliberately and ostentatiously flouted in order to give rise to a conversational implicature; such examples involve exploitation. These examples, and others Grice discusses in the second William James lecture, are all specific to, and entirely dependent on, the individual contexts in which they occur. Grice labels all such example ‘particularised conversational implicatures’. There are other types of conversational implicature in which the context is less significant, or at least can operate only as a ‘veto’ to implicatures that arise by default unless prevented. These are implicatures associated with the use of particular words. Unlike conventional implicatures, they can be cancelled: that is explicitly denied without contradiction. These ‘generalised conversational implicatures’ account for many of the differences between the logical constants and the behaviour of their natural language counterparts. In effect, Grice claims that there simply is no difference between, say ‘~’, ‘Ÿ’, ‘⁄’ and ‘not’, ‘and’, ‘or’ at the level of what is said. The well-known differences are generalised conversational implicatures often associated with the use of these expressions, implicatures determined by the categories and maxims he has established. Part of Grice’s motivation for this proposal was the desire for a simplification of semantics. The alternative to such an account was to posit a semantic ambiguity for a wide range of linguistic expressions. Grice argues against this, proposing a principle he labels ‘Modified Occam’s Razor’, which would rule against it in decisions of a theoretical nature. The principle states that ‘senses are not to be multiplied beyond necessity’.47 Grice’s reference was to William of Occam, or Ockham, the

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 105

Logic and Conversation 105

fourteenth-century philosopher credited with the dictum ‘entities are not to be multiplied beyond necessity’. This is known as ‘Occam’s razor’, although it is not clearly attributable to any of his writings, and it is not at all uncommon for philosophers to discuss it in isolation from Occam’s actual work. It is taken as a general injunction not to complicate philosophical theories; the best theory is the simplest theory, invoking the fewest explanatory categories. The preference for simple philosophical theories that do not add complex and potentially unnecessary categories was one with an obvious appeal to philosophers of ordinary language. Indeed, when Gilbert Ryle published his collected papers in 1971, he commented on the ‘Occamising zeal’ particularly apparent in the earlier articles.48 Another contemporary philosopher to draw on an Occam-type approach to discussions of meaning was B. S. Benjamin, whose article ‘Remembering’ is referred to in the first William James lecture. He does not draw an explicit comparison to Occam’s razor, but he does pose himself the question of whether the verb ‘remember’ should be analysed as multivocal or univocal. For Benjamin, a ‘universal core of meaning is preserved in its use in different contexts’.49 Grice himself did not develop the connection between conversational implicature and the logical constants in any great depth, either in the William James lectures or elsewhere. This is perhaps surprising, given that he introduces his theory in terms of the question of the equivalence, or lack of equivalence, between certain logical devices and expressions of natural language. The implications of this question, together with the specific answers offered by conversational implicature, are treated in detail by others.50 A. P. Martinich has suggested that the initial concentration on, and subsequent abandonment of, the logical particles is a serious flaw in the construction of the second, and most widely read, of the William James lectures. In a book aimed at describing and promoting good philosophical writing, Martinich identifies this as ‘one of the greatest articles of the twentieth century’, but argues that the more general theory of ‘linguistic communication’ ought to have been made the focus from the outset. He comments that on first reading Grice’s article he was unimpressed by what he saw as an unacceptably complex mechanism to solve a very particular logical problem: ‘Once I realised that the solution was a minor consequence of his theory I was awed by its elegance and simplicity.’51 Grice’s discussion of logic is mainly restricted to the fourth William James lecture, which he later labelled ‘Indicative conditionals’ after the chief, but not the only, logical constant it discusses. Indicative condi-

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 106

106 Paul Grice, Philosopher and Linguist

tions had been a central theme of some lectures on logical form Grice delivered at Oxford while working on his theory of conversation. There he had commented extensively on Peter Strawson’s treatment of this topic in his 1952 book Introduction to Logical Theory. Grice does not mention Strawson at all in this fourth William James lecture. He does, however, discuss the views of what he calls a ‘ “strong” theorist’, views that accord with Strawson’s in the insistence that the logical implication, ‘p … q’ is different in meaning from various expressions in natural language, most notably ‘if p then q’. Strong theorists, Grice suggests, have tended to stress the difference between what he calls natural or ordinary language conditionals and ‘artificial’ conditionals, defined by logic and determined by truth-conditional properties. Strawson argues that, while logical conditionals can be given a full definition in terms of a truth table involving the two simple propositions involved (‘p’ and ‘q’), such an account will not be sufficient for natural language expressions such as ‘if p then q’. The logical account specifies that if p is true, q must also be true. If p is false, however, nothing can be predicted about the truth value of q; a false antecedent coupled with a false consequent is assigned the overall value ‘true’, but so is a false antecedent coupled with a true consequent. This truthconditional account seems insufficient as a definition of natural language ‘if . . . then . . .’ in at least two ways. There is a suggestion in most actual instances that there is some causal connection between the antecedent and the consequent. One of Strawson’s own examples is: ‘If it rains, then the party will be a failure.’ Such examples, he notes, are not about linguistic elements or logical relations: ‘they suggest connections between different things in the world, discovered by experience of these things.’52 Later, he comments that most examples of the ‘if . . . then . . .’ construction would suggest either that the truth of the antecedent is in some doubt, or that it is already known to be false. It is only in such cases that we would be likely to label the resulting statement true, or perhaps ‘reasonable’. Strawson’s suggestion is that ‘if p then q’ entails its logical counterpart ‘p … q’. However, ‘a statement of the form “p … q” does not entail the corresponding statement of the form “if p then q”.’53 There are aspects of the meaning of ‘if p then q’ that go beyond, and are not included in, the meaning of the logical conditional. In his Oxford lecture notes, Grice singles Strawson out as an example of a philosopher who has not paid sufficient attention to the different ways in which a natural language expression can convey significance when it occurs in context. For Grice, this oversight has serious consequences for an understanding of the meaning of various crucial exam-

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 107

Logic and Conversation 107

ples. Certainly, as well as discussing the ‘meaning’ of particular expressions, Strawson writes, apparently without distinction, of the ‘use’, ‘employment’, ‘sense’ and ‘implication’ of ‘if . . . then . . .’ statements.54 He describes the speaker’s relation to the relevant meaning as one of ‘suggesting’, ‘using’, ‘indicating’, ‘saying’ and even ‘making’.55 Grice’s contention, although not explicitly expressed in ‘Indicative conditionals’, is that there is a great deal of difference between the members of the sets that Strawson uses interchangeably. In particular, sufficient attention to the difference between saying and implicating can explain away the apparent differences between ‘p … q’ and ‘if p then q’. He begins the fourth William James lecture uncompromisingly enough by proposing to consider the supposed divergence between ‘if’ and ‘…’ and to demonstrate that ‘no such divergence exists’.56 He suggests that ‘if p then q’ is distinguished from ‘p … q’ by the ‘indirectness condition’ associated with it: the commitment to some causal or rational link between p and q. Grice’s claim, in effect, is that the literal meaning of ‘if p then q’, ‘what is said’ on any particular occasion of utterance, is simply equivalent to the logical meaning of ‘p … q’. The indirectness condition is a generalised conversational implicature. Like all generalised conversational implicatures, it can be cancelled without contradiction, either by circumstances in the context or by explicit denial. In a game of bridge: ‘My system contains a bid of five no trumps, which is announced to one’s opponents on inquiry as meaning “If I have a red king, I also have a black king”.’57 Here, Grice suggests, the indirectness condition simply does not arise because it is blocked by the context. There is no suggestion that having a black king is causally linked to having a red king, and the meaning is the same as that of the logical condition. To say ‘if Smith is in the library he is working’ would normally suggest the indirectness condition. But it is perfectly possible to say ‘I know just where Smith is and what he is doing, but all I will tell you is that if he is in the library he is working’, in a situation where the speaker had just looked in the library and seen Smith there working.58 In such a case also, the indirectness condition is not attached to the use of the expression. Grice suggests what he describes as two separate possibilities as to how the indirectness condition might be produced as an implicature, although it is not in fact clear that these are necessarily mutually exclusive. The first relies on his original ‘general principle’ about the tendency towards stronger rather than weaker statements, and the first maxim of Quantity that results from it. It is less informative to say ‘if p then q’ than it is to say ‘p and q’, because the former does not give

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 108

108 Paul Grice, Philosopher and Linguist

definite information about the truth values of p and q. Therefore, any utterance of ‘p … q’ will, unless prevented by context, give rise to the implicature that the speaker does not have definite information about the truth values of p and q. The same explanation can be extended to utterances of the form ‘p or q’. These too seem to differ systematically from the apparently equivalent logical disjunction, ‘p ⁄ q’. The truthconditional meaning of logical disjunction states simply that at least one of the simple propositions involved must be true. It is perfectly consistent, therefore, for both propositions to be true. Yet in natural language there is something distinctly odd about saying ‘p or q’ if you know for certain that both p and q are true. In Grice’s terms, ‘p or q’ shares the logical meaning of ‘p ⁄ q’, but in addition carries a generalised implicature that they are not both true. If the speaker were in a position to offer the more informative form ‘p and q’, then it would be conversationally more helpful to do so. Grice’s second suggestion is that implicated meanings of such expressions may follow from their role in conversation, and in human interaction and thought more generally. The familiar logical constants enable people to work out the problems presented to them by everyday life. In particular, disjunction enables people to consider alternatives and eliminate the untenable. It enables people to give partial answers to ‘Wh’-questions such as ‘Who killed Cock Robin?’. The most helpful answer would be a single subject (‘the sparrow’, ‘the wren’), but if the speaker is not in a position to offer one, a series of disjuncts is a way of setting out the possibilities. Conditionals, on the other hand, enable people to ponder the consequences of certain choices. They are, therefore, necessary to the successful operation of reasoning beings. It would simply not be rational to use a conditional in certain contexts: contexts where there is no doubt about the truth of the antecedent, for instance. For this reason, on the assumption that our interlocutors are rational beings, we tend to interpret a conditional as indicating that a simple coordination will not do. Similarly, it would not be rational to use a disjunction in a context where we know that one of the disjuncts is true; we would in effect be attempting to solve a problem that had already been solved. Grice suggests, with typical tentativeness, that: It might be that either generally or at least in special contexts it is impossible for a rational speaker to employ the conditional form unless, at least in his view, not merely the truth-table requirements are satisfied but also some strong connection holds. In such a case a speaker will nonconventionally implicate, when he uses the

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 109

Logic and Conversation 109

conditional form in such a context, that a strong connection does hold.59 Grice offers this account in terms of the behaviour of a ‘rational’ speaker in at least potential opposition to an account drawing on the first maxim of Quantity. Elsewhere, however, he discusses the maxims as describing individually rational forms of behaviour. It is therefore conceivable that it might be possible to integrate the two suggestions, seeing conditionals as used rationally to introduce the implication of strong connection, precisely because their ‘tentative’ state does not offer the information that would be cooperative if available. In the later ‘Logic and conversation’ lectures, Grice continues his programme for a philosophical theory of language by returning to older preoccupations, particularly to the problems identified in, and raised in response to, ‘Meaning’. In effect, the theory of conversation offers a much fuller account of the notion of speaker meaning than had been developed in ‘Meaning’, together with a principled system linking this to conventional meaning. When first introduced in the second lecture, conventional meaning itself receives rather cursory treatment. Grice suggests that ‘what is said’ can be roughly equated with conventional meaning, including assigning of reference to referring expressions and any necessary disambiguation. He almost immediately complicates this definition by stipulating that some aspects of conventional meaning can be implicated. Grice returns to the notion of ‘what is said’ in the fifth William James lecture, later published under the title ‘Utterer’s meaning and intentions’. Grice’s contention is still that intentions on individual occasions must be the primary criteria in determining meaning. However, simply referring to what some individual meant by some action on some particular occasion is not sufficient to arrive at an account of ‘what is said’. It does not rule out a host of examples that have nothing at all to do with saying, such as flashing your headlights to indicate that the other driver has right of way. Grice’s solution is in effect to do something close to the manoeuvre suggested by Searle in Speech Acts. He proposes to import into the definition of ‘meaning’ a specification that the utterer’s action must constitute a unit in some linguistic system that has, or contains, the meaning intended by the utterer. In other words, ‘U did something x which is an occurrence of a type S part of the meaning of which is “p”.’60 Furthermore, he introduces a notion of what U ‘centrally meant’ in doing x, in order to distinguish a core meaning from other aspects of what may be conveyed by an utterance, in particular from its con-

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 110

110 Paul Grice, Philosopher and Linguist

ventional implicatures. In effect, Grice is introducing the notion of conventional meaning into utterer meaning, making it serve as part of the definition of what a speaker can legitimately intend by an utterance, while still maintaining that individual speaker intention is the primary factor in meaning. Grice extends his discussion of the different ‘levels’ of meaning to allow for four separate levels, labelled ‘timeless meaning’, ‘applied timeless meaning’, ‘occasion meaning of utterance type’ and ‘utterer’s occasion meaning’. In the same lecture, Grice responds to some of the criticisms of ‘Meaning’. Most of these, such as Schiffer’s identification of an untenable infinite regress, were not published at the time. Grice is replying to objections put to him ‘in conversation’.61 His response to Schiffer is, in effect, that the mental exercises required in recognising the infinite regress in intentions quickly become ‘baffling’. Whatever they may demonstrate in theory, they are impossibly complex for any real-life situation: At some early stage in the attempted regression the calculations required of A by U will be impracticably difficult; and I suspect the limit was reached (if not exceeded) in the examples which prompted the addition of a fourth and fifth condition. So U could not have the intention required of him in order to force the addition of further restrictions. Not only are the calculations he would be requiring of A too difficult, but it would be impossible for U to find cues to indicate to A that the calculations should be made, even if they were within A’s compass. So one is tempted to conclude that no regress is involved.62 Searle’s ‘American soldier’ counter-example was already published in article form, and Grice responds to it by arguing that Searle is unjustified in claiming that the soldier did not mean ‘I am a German officer’. Regardless of what the utterance conventionally meant in German (the question about lemon blossom), if the soldier used it with the intention of getting his captors to believe that he was a German officer, this was what he meant by the utterance. Grice offers a revised definition of his account of meaning. Intention is still the central notion, which Grice discusses in terms of ‘M-intentions’, emphasising the importance of the speaker’s desire for some response from the audience. The remaining two lectures in the ‘Logic and conversation’ series develop further this notion of M-intending. At the start of the penultimate lecture, published under the title ‘Utterer’s meaning, sentence

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 111

Logic and Conversation 111

meaning and word-meaning’, Grice displays a typical mixture of caution and ambition. He is concerned with the relationship between what a particular person meant on a particular occasion, and what a sentence or word means. What he has to say about this ‘should be looked upon as an attempt to provide a sketch of what might, I hope, prove to be a viable theory, rather than as an attempt to provide any part of a finally acceptable theory’.63 Speakers may have one of two attitudes to an utterance, which will determine their M-intentions. They may wish to get hearers to believe something, or they may wish to get hearers to do something. These two attitudes, or moods, relate to indicative or assertive utterances and to imperative utterances. The two moods are represented by the symbols ⵫ and ! respectively. Grice proposes that the symbol * stand as a ‘dummy operator’ in the place of either of these. Thus a general statement of the form ‘Jones meant that *p’ could be filled out using a specific mood operator and an indicative sentence to give either ‘Jones meant that ⵫ Smith will go home’ or ‘Jones meant that ! Smith will go home’. A further expansion, substituting a clause in indirect speech, yields ‘Jones meant that Smith will go home’ and ‘Jones meant that Smith is to go home’, both instantiations of the original formula. M-intentions in relation to imperative types are now defined as the intention that hearers should intend to do something, rather than the earlier, simpler definition that they should do something. The Mintended effect of an indicative-type utterance is no longer necessarily that the hearer should believe something ‘but that the hearer should think that the utterer believes something’.64 The effect of these two changes, particularly the first, is that all possible M-intentions are defined in terms of some psychological notion. Grice explicitly defends his use of what he calls ‘intensional’ concepts: those defining meaning not just in relation to how language maps on to the world, but also to how it relates to individual minds. Psychological concepts should not be ruled out, he argues, if they have something explanatory to offer to a theory of language. In this lecture, Grice indicates an account of how timeless meaning in a single idiolect may develop to become timeless meaning for a group. In effect, he is arguing that individual speaker meaning can, over time, yield conventional meaning. The exact status of this conventional meaning is perhaps still uncertain. Meaning is best explained in terms of the practices of a group; drawing on this, conventional meaning comes from ‘the idea of aiming at conformity’.65 Grice recognises a tension, or an ‘unsolved problem’ in the relationship between these two types of meaning. Briefly, ‘linguistic rules’ can be

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 112

112 Paul Grice, Philosopher and Linguist

identified such that our linguistic practice is ‘as if we accepted those rules and conscientiously followed them’.66 But the desire to see this as an explanation rather than just an interesting fact about our linguistic practice leads us to suppose that there is a sense in which ‘we do accept these rules’. This then leaves the problem of distinguishing ontologically between acceptance of the rules and existence of the practice. In the final William James lecture, Grice attempts a restatement of his position on meaning. To say that someone, U, means something, p, by some utterance of C, is to talk about ‘U’s disposition with regard to the employment of C’. This disposition ‘could be (should be) thought of as consisting of a general intention (readiness) on the part of U; U has the general intention to use C on particular occasions to means that p’ (Grice is here using ‘means’ for ‘speaker meaning’).67 Although he still wants to claim that all meaning is ultimately dependent on intention, Grice is no longer claiming that all meaning is derived from speaker meaning, therefore specifically from M-intentions. His extensive modifications of his theory of meaning have themselves been the subject of criticism. In 1973 Max Black, responding to the lectures so far published, commented on Grice’s ‘almost unmanageable elaborations’ of his own theory.68 He likens the tenacity with which Grice clings to the original idea to the latter fate of the ‘Principle of Verifiability’, with its numerous qualifications and reservations in the face of counter-examples. Employing an accusation that had long haunted Grice, he suggests that this tenacity goes beyond ‘an admirable stubbornness’, amounting to ‘something that might be called a “philosophical fixation” ’. The exact nature of conventional meaning and its relation to the more central notion of speaker’s meaning are perhaps the least well resolved aspects of the William James lectures. They are, however, crucial to Grice’s central programme. A fully integrated philosophy of language relating abstract linguistic meaning to what individual speakers do in particular contexts needs at least a clear account of the status of linguistic meaning. Grice seemed uncomfortable with its presence, but increasingly to be aware of his need to account for it. Towards the end of his life, he commented that the relationship between word meaning and speaker meaning was the topic ‘that has given me most trouble’.69 However, some later commentators have been more optimistic on his behalf. Anita Avramides has argued that it is possible to reconcile a psychological account of meaning with a conventional account, urging that ‘Grice’s work on meaning is, I believe, of a hearty nature and can endure alteration and modification without becoming obsolete.’70 Brian Loar has dismissed possible criticisms that Grice’s account of speaker meaning must always presuppose linguistic

1403_902976_06_cha05.qxd 12/13/04 6:26 PM Page 113

Logic and Conversation 113

meaning, claiming that ‘the basic illumination shed by Grice’s account of speaker’s meaning does not depend on such precise conceptual explication.’71 The problems surrounding conventional meaning prompted some of the earliest published responses to the William James lectures, perhaps in part because the last two lectures in the series, both concerned with this general topic, were the first to appear in print: ‘Utterer’s meaning, sentence meaning, and word-meaning’ in 1968 and ‘Utterer’s meaning and intentions’ in 1969. In ‘Meaning and truth’, his discussion of the ‘Homeric struggle’ between formal semanticists and communicationtheorists, Strawson suggests that the latter must always acknowledge that in most sentences ‘there is a substantial central core of meaning which is explicable either in terms of truth-conditions or in terms of some related notion quite simply derivable from that of a truthcondition.’72 Strawson singles out Grice’s 1968 article as an example of a communication theorist implicitly making such as acknowledgement. Another published response to these early articles came from Chomsky. It is clear from his 1976 book Reflections on Language that Grice’s admiration for his work was not reciprocated. Chomsky detects a return to behaviourism in ‘Utterer’s meaning, sentence meaning, and word-meaning’. This is rather surprising, given Grice’s own explicit rejection of behaviourism, and given his spirited defence of the need to allow psychological concepts into the theory of language. Chomsky picks up on Grice’s reliance on having ‘procedures’ to produce certain effects as an account of timeless meaning for a particular individual. Such an account, Chomsky argues, simply cannot explain the creativity of language use: the ability of speakers to produce and hearers to interpret an infinity of new sentences. He is uncomfortable with the loose ends and pointers to unresolved problems typical of Grice’s work. In particular, and not surprisingly, he picks up on Grice’s discussion of the status of linguistic rules. This question, Chomsky insists, is ‘the central problem, not a marginal one’.73 His own answer is, of course, that speakers really do follow linguistic rules in constructing utterances, and that these rules can tell us everything about the linguistic meaning of an utterance, and nothing about what an individual speaker meant in producing the utterance. In this book, as elsewhere, Chomsky is advocating a complete divorce between the discussion of linguistic meaning and the study of communication. Communication, he argues, is not the only, and by no means a necessary, function of language. Here is the source of the ultimately irresolvable difference between Chomsky and Grice. For Grice, communication is primary; the best chance for explaining language is by way of explaining communication.

1403_902976_07_cha06.qxd 12/13/04 6:27 PM Page 114

6 American Formalism

Grice returned to Oxford after the William James lectures, but only for one term. In the autumn of 1967 he took up an appointment at the University of California, Berkeley. The move was permanent; during the rest of his life he was to make only a handful of brief return visits to England. It was also in many ways incongruous. A thoroughgoing product of the British elite educational system, Grice delighted in the rituals and formalities of college life. And even Kathleen was amazed that he was ready to turn his back on cricket, concluding that it was only because at 54 he was too old to play for county or college that he was prepared to move at all. But there was no reluctance in Grice’s acceptance of his new appointment. In fact, he engineered the move himself. While at Harvard, he had made his interest known generally in the American academic community. When the offer came from Berkeley he had accepted it enthusiastically and, somewhat naively, without negotiation. Grice’s only published comment on the reasons for his move to America is in the philosophical memoir in his Festschrift, part of his description of the development of the theory of conversation: During this time my philosophizing revealed a distinct tendency to appear in formal dress; indeed, the need for greater contact with experts in logic and in linguistics than was then available in Oxford was one of my main professional reasons for moving to the United States.1 Certainly, the later William James lectures show an increasing formalism. For instance, in ‘Utterer’s meaning, sentence meaning and wordmeaning’ Grice uses symbols to generalise over variables and produce 114

1403_902976_07_cha06.qxd 12/13/04 6:27 PM Page 115

American Formalism 115

generic logical forms to be instantiated in particular utterances. He had found this way of working with language appealing on his earlier visits to America, and two of his major philosophical influences at this time, Chomsky and Quine, were American. There were other, more personal reasons for the move. Grice had become a confirmed admirer of all things American. He appreciated the comparatively relaxed social mores; the informality of the 1960s, which had not yet fully penetrated the Oxford colleges, suited his own disregard for personal appearance. He loved the Californian sunshine. And he was enthusiastic about the ready availability of what he saw as the most important commercial commodities. Asked by one colleague what was attracting him away to America he replied, ‘Alcohol is cheap, petrol is cheap, gramophone records are cheap. What more could you want?’ Perhaps above all, he liked the openness and warmth of American society. One incident in particular, from a few years earlier, had strengthened him in this opinion. He had recently returned to Oxford after a stay of several months in Stanford. He ran into a relatively close colleague, who followed up a platitudinous declaration that they must meet up some time with the comment, ‘I don’t seem to have seen you around. Have you been away?’ For Grice, this was indicative of a certain coldness and distance in British, or at least Oxford, society that contrasted poorly with his recent experiences in America. Certainly, he took to his new lifestyle with ease and enthusiasm. Soon after the move he bought a house in the Berkeley Hills, with a balcony overlooking San Francisco Bay. Kathleen, who had stayed in England to oversee their two children’s university entrances, moved to Berkeley only in 1969, by which time the matter of the house was all settled. She took a job teaching children with special needs, and would often return after a day’s work to find every window sill covered with dirty glasses, some covered in paint smudges. Grice was in the habit of extending his hospitality to every visitor to the house, whether colleague, student or decorator, but not of tidying up afterwards.2 Berkeley in 1967 was a centre for the anti-Vietnam War movement, for the student protests and for the associated questioning of social and political orthodoxy. However, such challenges were not necessarily felt within the institution of the university itself. The philosophy department had maintained a sharp hierarchical distinction between faculty and students, even at graduate level. Grice, who specified that he would teach only graduate students, had an impact on this. For all its formality and traditionalism, the Oxford college tutorial system had the effect of promoting regular contact, and free exchange of ideas, between tutors

1403_902976_07_cha06.qxd 12/13/04 6:27 PM Page 116

116 Paul Grice, Philosopher and Linguist

and students. Grice had greatly benefited from this, first as a student of W. F. R. Hardie, and later as a tutor to younger philosophers such as Peter Strawson. At Berkeley he taught his students in small groups, and encouraged them to talk to him outside class times about philosophical ideas, including their own. At his instigation, the weekly philosophy research seminars were followed by an evening for both faculty and students in a local restaurant or at his house. In turn, he was a regular and enthusiastic attender of student parties. His motives were, of course, a mixture of the intellectual and the social, but the effect was to offer students an unprecedented access to the time and ideas of a member of faculty. Just as he retained a distinctly ‘Oxford’ ethos in his teaching, despite his enthusiasm for his new post, so Grice remained unmistakably British, and more specifically old-school British establishment. Perhaps even more strikingly after 1967 than before, his supposedly ‘everyday’ linguistic examples were peopled by majors, bishops and maiden aunts, set at tea parties and college dinners. He continued to produce manuscripts in British English spelling, to be painstakingly changed to American spellings by the copy editors. And in tape recordings even from 20 years after the move, his conservative RP accent stands out in sharp contrast to those of his American colleagues. Nor did Grice entirely leave behind the methods and orthodoxies of Oxford philosophy when he moved to Berkeley to be in closer contact with American formalism. Commentators on twentieth-century philosophy have drawn attention to the different ways in which analytic philosophy developed in America and in Britain. For instance, writing in 1956, Strawson describes the American tradition as turning away from everyday language in order to concentrate on formal logic, a tendency he sees represented in the work of Carnap and of Quine. This he contrasts with the British, or rather English, tradition, exemplified by Austin and Ryle, with its continued close attention to common speech. He sees the former approach as representing a desire to create something new, and the latter a desire to understand what already exists; he even goes so far as to suggest that these ‘national preferences’ are indicative of ‘a characteristic difference between the New World and the Old’.3 Grice could be seen as making a strong statement of academic intent by moving to America, but even in the work he produced immediately after that move, perhaps the most formal work of his career, he retained a distinctively ‘Old World’ sensitivity to ordinary language. Grice’s first publication after the move, apart from the two William James lectures that appeared in print in the late 1960s, was ‘Vacuous

1403_902976_07_cha06.qxd 12/13/04 6:27 PM Page 117

American Formalism 117

names’, published in 1969. This was firmly located in a formalist tradition, appearing in a volume of essays responding to the work of Quine and including contributions by linguists and logicians such as Chomsky, Davidson and Kaplan. It was concerned with issues of reference, in particular the question of how best to analyse sentences containing names that fail to refer, sentences such as ‘Pegasus flies’. Grice had in fact been working on reference for some time, both before and after his move. He was lecturing on the topic in Oxford in October 1966, and gave a paper closely related to ‘Vacuous names’ at Princeton in November 1967. He comments at the start of the published paper on his own inexperience in formal logic: ‘I have done my best to protect myself by consulting those who are in a position to advise me; they have suggested ideas for me to work on and have corrected some of my mistakes.’4 It seems that he had undertaken, quite self-consciously, to educate himself in predicate logic, so as to be in a position to answer the question in which he had become interested. The quantity of notes that survive from this period, and the wide range of materials they cover, testify to the perseverance, and also the relentlessness, with which he pursued this end. Tiny, scrawled logical symbols cover headed note-paper from both St John’s College and the Berkeley department, and envelopes addressed to Grice both at Woodstock Road and at various temporary addresses from his early months in Berkeley. In some of the notes from early in his Berkeley career, Grice devotes a lot of attention to the way in which expressions such as ‘reference’ and ‘referring’ are used in everyday language. He considers expressions such as ‘in saying p, S was making a reference’ and ‘in what S said p occurred referentially’. Such locutions, he notes to himself, are ‘principally philosophical in character; ordinary chaps don’t often say this sort of thing’.5 Instead, he investigates expressions that do occur in ordinary language, such as the phrase ‘when he said . . . he was referring to . . .’, for instance ‘when he said “The Vice President has resigned” he meant/was referring to the secretary’. People do regularly use the expression ‘refer’ to describe not just what phrases literally denote, but what people intend to pick out, or point to. Grice tackles the apparent paradox introduced by such an example. If what the speaker meant was that the secretary (Jones) had resigned, and it is true that the secretary had resigned, what he said is true. However, what he said entails that there is (was) a vice-president. In a situation where it is not true that there is (was) a vice-president, what is said is not true. It seems at least possible that the speaker’s remark must be both true (because the secretary has resigned) and not true (because there is no vice-president) at

1403_902976_07_cha06.qxd 12/13/04 6:27 PM Page 118

118 Paul Grice, Philosopher and Linguist

the same time. Grice notes that the ‘truth of what is said (in suitable cases) must turn not only on denotation but also on reference’: on what the speaker intends to pick out as well as on what the words literally indicate. In ‘Vacuous names’, Grice declares himself keen to uphold if possible a number of intuitively appealing dogmas. These should ideally inform the construction of a predicate calculus to explain the semantics of natural language. The dogmas include a classical view of logic, in which if Fa is true then ~Fa must be false, and vice versa. They also include the closely related claim that ‘if Pegasus does not exist, then “Pegasus does not fly” (or “It is not the case that Pegasus flies”) will be true, while “Pegasus flies” will be false’.6 If these truth relations hold, ‘Pegasus flies’ logically entails that Pegasus exists. In this, Grice is proposing to sustain a Russellian view of logic. This may seem like an unexpected or contradictory proposal from a philosopher of ordinary language who had collaborated with Strawson and employed his ‘presuppositional’ account of such examples in their joint work on categories. However, Grice had been moving away from a straightforwardly presuppositional account for some time. As early as the notes for the lectures on Peirce from which ‘Meaning’ developed, he had pondered the idea that examples Strawson would describe as presuppositional might provide illustrations of the difference he was investigating between sentence meaning and speaker meaning.7 Grice does not argue that Strawson’s account of presupposition does not work, or does not explain accurately how people understand utterances in context. But he suggests that it is not necessary to use Strawson’s observation as an explanation of sentence logic as well as speaker meaning. In the years since he made these notes, he had of course refined this notion in much more detail, developing in the William James lectures the idea that logical form may be quite different from context-bound interpretation, with general principles of language use mediating between the two. He adopts this approach in ‘Vacuous names’. He considers two different uses of the phrase ‘Jones’s butler’ to refer to an individual. On one occasion, people might say ‘Jones’s butler will be seeking a new position’ when they hear of Jones’s death, even if they do not know who the butler is; they might equally well say ‘Jones’s butler, whoever he is, will be seeking a new position’. On a different occasion, someone says ‘Jones’s butler got the hats and coats mixed up’, describing an actual event, but mistakenly applying the name ‘Jones’s butler’ to a person who is actually Jones’s gardener; Jones does not in fact have a butler. In effect Grice wields Modified Occam’s Razor, although he does

1403_902976_07_cha06.qxd 12/13/04 6:27 PM Page 119

American Formalism 119

not name it, when he insists that there is no difference in the meaning of the descriptive phrase in these two instances, only in the use to which it is put. ‘I am suggesting that descriptive phrases have no relevant systematic duplicity of meaning; their meaning is given by a Russellian account.’8 So in the second case, when ‘Jones’s butler’ fails literally to refer, ‘What, in such a case, a speaker has said may be false, what he meant may be true (for example, that a certain particular individual [who is in fact Jones’s gardener] mixed up the hats and coats).’9 For Grice, our responses to speakers’ everyday use of expressions may not be the best guide to logic. Grice was genuine in his desire to find out more about both logic and linguistics. His notes from the years immediately following the move to Berkeley reveal that, as well as setting himself the task of learning predicate calculus, he was reading current linguistic theory. Along with his own extensive notes on the subject, he kept a copy of Chomsky’s ‘Deep structure, surface structure and semantic interpretation’, in a version circulated by the Indiana University Linguistics Club early in 1969.10 He also had a number of articles on general semantics, on model-theoretic semantics, and on the semantics of children’s language. The notes show him at work on the interface between semantics and syntax. There are tree diagrams and jottings of transformational rules mapping one diagram on to the next. There are sketches of phrase structure rules, such as ‘NP Æ Art + N’ and ‘N Æ Adj + N’, accompanied by the list ‘N: boy, girl, tree, hill, dog, cat’.11 He seems to have been interested for a while in reflexive pronouns, listing verbs with which they were optional, or verb phrases from which they could be deleted (‘shave myself’/‘shave’, ‘dress myself’/‘dress’) and those from which they could not (‘cut myself’/‘cut’, ‘warm myself’/‘warm’). He also dabbled in phonetics, making notes to remind himself of words representative of the various vowel sounds, and drawing up a table of the distinctive features of the consonants [p], [t] and [k]. Given this flurry of interest in linguistics, and his stated intention of putting himself into closer connection with practising linguists, it is striking that Grice does not seem to have sought any contact with linguists working at Berkeley. This is all the more surprising given that the interests of some Berkeley linguists at the time might be seen as corresponding remarkably closely to his own. If Berkeley was a hub of student uprising in 1967, it was to become a centre for a different, quieter revolution in the early 1970s. The generative semanticists were in revolt against what were now the orthodoxies of Chomskyan linguistics. Chomsky was maintaining his hard line on the difference between

1403_902976_07_cha06.qxd 12/13/04 6:27 PM Page 120

120 Paul Grice, Philosopher and Linguist

linguistic and non-linguistic phenomena. He argued that generative grammar was concerned with syntactic structure, and with semantic meaning to the extent that it depended on deep structure. It had nothing whatsoever to say about how speakers use language in context. The generative semanticists argued that, if grammar really was to offer an account of the human mind, it should be able to explain all aspects of meaning, including the obvious fact that meaning is often indirect, or non-literal. Robin Lakoff has described how, in making such claims, linguists such as Paul Postal, George Lakoff, James McCawley and John Ross were seen as directly opposing Chomsky. Their attempts to find ways of incorporating various aspects of utterance meaning into a formal theory of language led to more and more complicated linguistic rules, and eventually to the failure of generative semantics as an enterprise. Along the way, however, they picked up on work from ordinary language philosophy as ‘the magic link between syntax and pragmatics’.12 Speech act theory offered a way of incorporating utterance function into grammar. Generative semanticists attempted to account for implicit performatives, where no overt performative verb is present, in terms of a ‘performative marker’ in deep structure.13 Grice’s theory of conversation suggested possibilities for incorporating context-sensitive rules into grammar, and thereby explaining a wide range of non-literal or indirect meaning.14 However, there was little personal contact between Grice and the linguists who were making such enthusiastic use of his work. Programmes of the Berkeley linguistics colloquium, sent to Grice in the philosophy department, provided him with paper for his incessant jottings and listings, but were left unopened. McCawley’s work on the performative hypothesis gets a mention in a handout for Grice’s students from 1971, but references in his more public lectures and in his published work are always to philosophers rather than linguists. Aware of what linguists were doing, but not in active dialogue with them, he worried privately in his notes over whether logicians and linguists actually mean the same by their apparently shared vocabulary such as ‘syntax’ and ‘semantics’. ‘I have the feeling’, he confesses, ‘that when I use the word “semantic” outside logical discussion, I am using the word more in hope than in understanding.’15 There are, perhaps, two explanations for Grice’s silence on the topic of generative semantics in particular, and of linguistics more generally, although there is no evidence that Grice himself endorsed either of these. One is to do with the specific differences between Grice’s enterprise and that of his contemporaries in linguistics, the other with the

1403_902976_07_cha06.qxd 12/13/04 6:27 PM Page 121

American Formalism 121

more general differences between linguistics and philosophy as disciplines.16 Although one of Grice’s central interests at the time was that of linking semantics and syntax into a coherent theory of language, he saw semantics as primary and syntax as derived from it. As he suggested in one talk on the subject: What I want to do is in aid of the general idea that one could work out for a relatively developed language a syntax such that all the way down any tree which would terminate in the [sic] sentence of the language would be a structure to which a semantical rule was attached.17 His theory of meaning was the driving force behind his theory of language. The generative semanticists, on the other hand, took syntax as primary. Syntax, with a much more precise definition than Grice was using, was the governing factor in language. Their aim was to use it to explain as many aspects of meaning as possible. More generally, as Robin Lakoff has pointed out, the disciplines of philosophy and linguistics have different expectations of a ‘theory’ of language and make different assumptions about the role of examples.18 Linguists, at least those within a broadly ‘Chomskyan’ framework, are generally committed to the scientific validity of their theories. Examples are used to demonstrate a point and, in the face of counterexamples, the theory must be modified or in the worst case abandoned. Philosophers in Grice’s style are concerned with producing philosophically revealing accounts of the relevant phenomena, and examples are used to illustrate these accounts. Some of the more recent reactions to the theory of conversation from within linguistics highlight these different expectations. For instance, Paul Simpson has complained that ‘Grice’s own illustrations are all carefully contrived to fit his analytic model; as a result there emerge from the theory too few explicit criteria to handle the complexity of naturally occurring language.’19 For Grice’s purposes, the close fit of illustration and model is no problem; to a linguist it is a major fault. In 1970, Grice gave a series of lectures and seminars at the University of Urbana under the uncompromisingly ambitious title ‘Lectures on language and reality’. He explains at the start of these that his overarching theme will be reference, a topic that has interested him for some time. His interest follows on his belief that a clear account of reference is essential to a full understanding of many other central topics in the philosophy of language. More specifically and recently, he has come to

1403_902976_07_cha06.qxd 12/13/04 6:27 PM Page 122

122 Paul Grice, Philosopher and Linguist

see its significance to the connections between grammar and logic, between syntax and semantics. In the first lecture he launches into an elaborately coded account of language, based on a pseudo-sentence invented by Rudolf Carnap. In the introduction to The Logical Syntax of Language, Carnap presents a typically logical positivist account of the philosopher’s reason for taking language seriously. A suitably rigorous language will provide the necessary tools for logical and scientific exposition. This language is to be a formal system, concerned with types and orders of symbols but paying no attention to meaning. ‘The unsystematic and logically imperfect structure of the natural word-languages’ makes them unamenable to such formal analysis, but in a ‘wellconstructed language’ it is possible to formulate and understand syntactic rules.20 For instance, given an appropriate rule, it can be proved that the word-series ‘Pirots karulize elatically’ is a sentence, provided only that ‘Pirots’ is known to be a substantive (in the plural), ‘karulize’ a verb (in the third person plural), and ‘elatically’ an adverb . . . The meaning of the words is quite inessential to the purpose, and need not be known. Carnap does no more with his invented example, proposing instead to stick to symbolic languages. Grice’s approach and purpose in ‘Lectures on language and reality’ are very different from Carnap’s, although he does not refer directly to these differences, and indeed mentions Carnap only in passing. Still at heart a philosopher of ordinary language, Grice is convinced that natural language is a worthwhile focus of philosophical investigation in its own right, and that its apparent logical imperfections are not insurmountable. Moreover, it is philosophically important because of the purposes to which it is put, not because of its potential for logical expression. Perhaps with an eye to the subversiveness of his undertaking, he borrows from Austin’s paper ‘How to talk: some simple ways’ in suggesting that his programme might be subtitled ‘How pirots carulize elatically: some simpler ways’.21 Grice uses Carnap’s nonsense words, and others like them, to build up a fragment of a fantasy world. So his audience is treated to pieces of information such as ‘a pirot a can be said to potch of some obble x as fang or feng; also to cotch of x, or some obble o, as fang or feng; or to cotch of one obble o and another obble o1 as being fid to one another’.22 Some way into the first lecture he offers the audience the key to this

1403_902976_07_cha06.qxd 12/13/04 6:27 PM Page 123

American Formalism 123

code. Pirots are much like ourselves, and inhabit a world of obbles very much like our own world. To potch is something like to perceive, and to cotch something like to think. Feng and fang are possible descriptions, much like our adjectives. Fid is a possible relation between obbles. Part of the reason for the elaborate story of obbles, he suggests, is because: it seems to me very important that, when one is considering this sort of thing, one should take every precaution to see that one isn’t taking things for granted and that the concepts which one is going to use have, as their basis, concepts which will only bring in what is required for them to do whatever job it is that one wants them to do. Carnap wanted to use semantically opaque forms to hint at how an analysis of syntax might proceed without reference to meaning. Grice’s intention in borrowing his example is to consider what concepts might be necessary to the discussion of meaning and reference, freed from the normal preconceptions of such a discussion. It is when the behaviour of pirots is considered, when a group of pirots in a particular gaggle interact together, that the notion of reference becomes important. Situations in which pirots want to communicate about obbles are when language ‘gets on to the world’. In effect, Grice is encouraging the dispassionate consideration of what rational beings are likely to do with a communication system. The regularities of syntax are based on language’s function of referring to the world; the types of meanings the pirots need to express determine the structures of the language. A successful language is one able to offer true descriptions of the world. The business of language, its driving force, is communication. Grice may appear to have moved rather a long way from the ideals of ordinary language philosophy, in constructing an artificial code, or language fragment. However, he emphasises that he see this as a necessary simplification, as a way of modelling and defamiliarising natural language in order to study it more clearly. His interest remains with the issue of how language maps on to reality: how it exists principally as a system for communicating about the world shared by a community of speakers. He is interested in the workings of natural language rather than the regularities of the constructed, purified language of logic. The transcriptions of the Urbana seminars show one participant asking him whether it would not be better to stick to a logical language, with its simple, familiar structures: ‘why bother with ordinary

1403_902976_07_cha06.qxd 12/13/04 6:27 PM Page 124

124 Paul Grice, Philosopher and Linguist

language?’ Grice’s simple reply is ‘I find ordinary language more interesting.’23 Grice returns to the debate between Russell and Strawson over denoting expressions that fail to refer in the fourth Urbana lecture, the only one to be published (some years later, as ‘Presupposition and conversational implicature’). Here again, his inclination is in favour of Russell’s theory of descriptions to account for examples such as ‘the king of France is bald’, and to find alternative explanations for the ‘presuppositional’ effects associated with them. Characteristically tentative, Grice does not explicitly endorse the Russellian position, but proposes to investigate it and test the extent of its applicability. This time, he states explicitly that he will apply the ‘dodge’ of implicature. On Russell’s interpretation of ‘the King of France is bald’, both the unique existence and the baldness of the king of France are logical entailments of the sentence; ‘the king of France is not bald’ is ambiguous, with both entailments of the positive sentence equally possible candidates for denial. Strawson objected that a person who utters ‘the king of France is not bald’ will generally be taken to be committed to the existence of the king and to be denying his baldness. The supposed ambiguity is rarely if ever a feature of the use of such an expression. Grice concurs with Strawson’s suggestion that the commitment to existence seems to be shared by the assertion and its denial. However, he argues, it is not necessary to assume that the commitment is of the same order, or to the same degree, in both cases. If Russell’s interpretation is retained, it follows that the denial must be semantically ambiguous between a negation with scope over the whole sentence, and negation with scope simply over the baldness. The latter does, whereas the former does not, retain semantic commitment to the existence of the king of France. On this interpretation the existence of the king of France is an entailment of the positive assertion, and of one (strong) reading of the denial. However, ‘without waiting for disambiguation, people understand an utterance of “the king of France is not bald” as implying (in some fashion) the unique existence of the king of France.’24 In response to this, it is possible to argue that on the weak reading of the denial the commitment to the existence of the king, although not a logical entailment, arises as a conversational implicature in use. The implicature envisaged by Grice in this latter case is a generalised conversational implicature; it arises by default when the expression is used, unless cancelled explicitly or by features of the context. It is present because of the speaker’s choice of the form of expression. For this reason Grice proposes to explain it in terms of the general category

1403_902976_07_cha06.qxd 12/13/04 6:27 PM Page 125

American Formalism 125

of Manner. In his original account of this category he suggested four maxims, and hinted that it might be necessary to add more. Here, he suggests one such addition, namely the maxim ‘Frame whatever you say in the form most suitable for any reply that would be regarded as appropriate’, or ‘Facilitate in your form of expression the appropriate reply’.25 It is reasonable to expect that conversational contributions will be expressed in a manner conducive to the most likely possible reply or range of replies. The facilitation of likely responses is also advanced to counter another of Strawson’s objections to Russell. On Russell’s account, an utterance of ‘the king of France is bald’ is simply false. Yet, Strawson observes, our most likely response on hearing it uttered would not be to say ‘you’re wrong’ or ‘that’s false’, but to be in some sense lost for an answer. According to Grice, the speaker in such a situation has not framed the utterance in a way that makes a reply easy. One possible likely response to a statement is a denial. In a conversational contribution entailing several pieces of information it is possible that any one of these pieces may be denied, but some particular aspect is often a most likely candidate. If all pieces of information are equally likely to be denied, the most cooperative way of phrasing the utterance would be as a simple conjunction of facts. The speaker would be expected to utter something close to the Russellian expansion, something such as ‘there is one unique king of France, and this individual is bald’. In producing the contracted form ‘the king of France is bald’, however, the speaker is apparently assuming that possible responses do not include the denial of the unique existence of the king of France. Grice suggests that it is generally information with ‘common-ground’ status that is not considered a likely basis for denial. This may be because the information is commonly known, and known to be commonly known between speaker and hearer. But this explanation will not always serve: For instance, it is quite natural to say to somebody, when we are discussing some concert, My aunt’s cousin went to that concert, when we know perfectly well that the person we are talking to is very likely not even to know that we have an aunt, let alone know that our aunt has a cousin. So the supposition must be not that it is common knowledge but rather that it is noncontroversial, in the sense that it is something that we would expect the hearer to take from us (if he does not already know). That is to say, I do not expect, when I tell someone that my aunt’s cousin went to a concert, to be questioned whether I have an aunt and, if so, whether my aunt has a cousin.

1403_902976_07_cha06.qxd 12/13/04 6:27 PM Page 126

126 Paul Grice, Philosopher and Linguist

This is the sort of thing that I would expect him to take from me, that is, to take my word for.26 ‘The king of France is bald’ presents the baldness as available for discussion, as possibly controversial. The existence, however, is not treated as something likely to be challenged. This explains the difficulty encountered by anyone who does in fact want to challenge it. Similarly, the negative statement is interpreted as a challenge to some actual or potential statement in which the baldness, not the existence, of the king of France is at issue. This explains why it is generally taken to commit the speaker to the belief that the king of France exists, even if it does not in fact logically entail this. Grice’s treatment of presuppositional phenomena in ‘Presupposition and conversational implicature’ is sketchy and partial. He offers a fairly detailed account of examples containing definite descriptions, but his rapid survey of other types of presuppositional phenomena suggests that his account may not be easily extendable. The paper offers a further example of an attempt to apply the phenomenon of conversational implicature to a philosophical problem of long standing, and in particular to preserve a classical account of logic in the face of apparently intractable facts of language use. In considering the link between grammatical structure and the status of information, it draws on Grice’s general interest in ways in which meaning may drive syntax. He does not generalise from the particular consideration of referring expressions in this paper. However, an unpublished paper from much later in his life does suggest the directions in which his thoughts may have developed. In 1985, Grice was seeking to reconcile the Oxford school of philosophy with Aristotle’s idea that philosophy is about the nature of things, rather than language. He proposes to adopt the hypothesis that opinion is generally reflected in language, with different ‘levels’ representing different degrees of commitment. Some aspects of knowledge receive the deepest level of embedding within syntax, residing in what Grice describes as the ‘deep berths’ of language. It is not possible for a speaker even to use the language without being committed to these. The deepest levels are at a premium, so it is in the interests of speakers to reserve these for their deepest commitments. People might challenge these, Grice suggests, but it would be dangerous to do so. If we subscribe to this account, we might be tempted to argue that first principles of knowledge are to be found in the syntactic structure, rather than the vocabulary, of language: ‘how we talk ought to reflect our most solid, cherished and generally accepted opinions.’27 In this discussion of what

1403_902976_07_cha06.qxd 12/13/04 6:27 PM Page 127

American Formalism 127

is presented as uncontroversial and what as available for denial, Grice might be described as interested in the ways in which different syntactic devices available for conveying information bring with them different existential and ontological commitments. The fourth Urbana lecture, taken on its own, offers an account of how a possible addition to the category of Manner suggests a new approach to the problems posed by Russell’s theory of descriptions. In the somewhat abridged form in which it was published as ‘Presupposition and conversational implicature’ this is, of course, how it is generally read. In the context of the Urbana lectures as a whole, however, it can also be seen as part of a much more ambitious project to describe how language, as a system to serve the communicative needs of rational beings, ‘gets on to’ the world, or functions to refer and describe. In this context, it is also apparent that Grice saw the question of the nature of reference as central to the philosophy of language, and particularly to his current interest in the division between syntax and semantics, or logic and grammar. Ultimately, it might be hoped that the structure of language could be shown to be revealing of the structure of human knowledge, precisely because language is dependent on knowledge. The theory of reference was not the only topic in the philosophy of language to which Grice returned during the early 1970s. His new interest in the relationship between semantics and syntactic form, and hence in the expression of human thought, also prompted him to revisit the topic of intention. He had been elected Fellow of the British Academy in 1966, and in the summer of 1971 he made one of his rare return visits to England in order to give an Academy lecture. For his topic, he chose to return to the themes of his discussion paper ‘Intentions and dispositions’ some 20 years earlier. In the mean time, Hampshire and Hart had published their ‘Decision, intention and certainty’. Gilbert Harman has repeated the rumour that this paper, containing a similar idea to that in ‘Intentions and dispositions’, ‘prompted Grice to develop his alternative analysis’.28 In his British Academy lecture, ‘Intention and uncertainty’, Grice refers rather vaguely to ‘an unpublished paper written a number of years ago’, and a thesis advanced in it that he now wants to criticise. He reformulates his original two-pronged account as a three-pronged one by making independence from empirical evidence into a separate criterion. ‘X intends to do A’ can truthfully be used only in cases where X is ready to take any preparatory steps necessary to do A. Further, such a statement implies or suggests that X is not in any doubt that X will in fact do A. Finally, this certainty on the part of the speaker must be independent of any empirical evidence. Drawing on

1403_902976_07_cha06.qxd 12/13/04 6:27 PM Page 128

128 Paul Grice, Philosopher and Linguist

the old ordinary language technique of devising a dialogue as context for a particular word, Grice offers the following illustration: X: Y: X: Y: X:

I intend to go to that concert on Tuesday. You will enjoy that. I may not be there. I am afraid I don’t understand. The police are going to ask me some awkward questions on Tuesday afternoon, and I may be in prison by Tuesday evening. Y: Then you should have said to begin with, ‘I intend to go to the concert if I am not in prison’, or, if you wished to be more reticent, something like, ‘I should probably be going’, or, ‘I hope to go’, or, ‘I aim to go’, or, ‘I intend to go if I can’. Grice seems unconcerned by the highly artificial nature of this ‘conversation’, compared to the structure and flow of naturally occurring talk. He comments that ‘it seems to me that Y’s remarks are reasonable, perhaps even restrained.’29 His point is that it seems legitimate to object to a use of ‘I intend . . .’ in cases where the speaker is not sure that it will be possible to fulfil the intention. The omission of such an addition as ‘. . . if I can’ is possible only in cases where the doubt is perfectly obvious to both speaker and hearer, as when the speaker is making plans to keep ducks in extreme old age, or to climb Everest. Grice’s criticism of his own earlier analysis focuses on the implication that an intention is a type of belief, and the secondary notion that it is a belief independent of evidence. He argues that there are simply no parallels to be found for this type of belief; neither a hunch nor a memory will fit the bill. A belief such as an intention would have to be – one that could not be true, false, right or mistaken – simply does not properly bear the title of belief at all. This means that intention lacks a clear definition, leaving it open to a sceptical attack challenging the coherency of ever saying ‘I intend . . .’. To use the statement ‘I intend to do A’ is to involve oneself in some degree of factual commitment and there must be something giving one the right to do this. Usually that something is evidence, but in the case of intention it has been established that there is nothing that would count as adequate evidence. Grice responds to this possible sceptical position by rejecting the assumption that in stating an intention one is involved in a factual commitment. Characteristically, he develops this objection by means of a careful analysis of language. He suggests that there is a difference between two uses of the ‘future tense’ in English, a difference not

1403_902976_07_cha06.qxd 12/13/04 6:27 PM Page 129

American Formalism 129

marked linguistically, or marked only in careful use in the distinction between ‘I will . . .’ and ‘I shall . . .’. In most normal uses, ‘I will . . .’ can be used either with a future intentional or a future factual meaning. Strictly speaking, if we intend to go to London tomorrow, we can only honestly use the expression ‘I will go to London tomorrow’ in its future intentional meaning; to do otherwise would be to suggest that it is somehow beyond our control. And for these future intentional statements it is simply not legitimate to ask for an evidential basis, any more than it would be in response to ‘Oh for rain tomorrow!’. In each case to ask for evidence would be ‘syntactically’ inappropriate. Having drawn attention to this distinction, Grice ponders the links it suggests between intending and believing with respect to a future action. He proposes the term ‘acceptance’ as one that can express ‘a generic concept applying both to cases of intention and to cases of belief’.30 Using the notion of acceptance in the original three-pronged analysis might, he suggests, rescue it from the criticisms based on its reliance on belief as central to intention. In the case of intention but not of belief, X accepts that X will do A and also accepts that X doing A will result from that acceptance. No external justification is required. In the case of belief, however, some external justification, or evidence, is needed for the acceptance. At the start of ‘Intention and uncertainty’, Grice acknowledges a debt to Kenny’s concept of ‘voliting’. He does not refer directly to Kenny’s work, but its influence appears in relation to the notion of ‘acceptance’. Anthony Kenny, an Oxford philosopher who was President of the British Academy, had published a collection of essays entitled Action, Emotion and Will since Grice’s first attempt at analysing intention. Kenny argues that expressions of desire and of judgement both display a similar complexity; both take as object not a thing but a state of affairs, despite surface appearances to the contrary. ‘Wanting’ always specifies a relation to a proposition; so, for instance ‘wanting X’ is in fact ‘wanting to get X’. For this reason, we might ‘expect that an analysis of a report of a desire should display the same structure as the analysis of a report of a judgement’.31 Kenny introduces the artificial verb ‘volit’ to generalise over the range of attitudes a speaker can adopt to a proposition that express approval, whether this be approval of actual states, as in ‘x is glad that p’ or of possible states as in ‘x wishes that p’. This positive attitude distinguishes ‘voliting’ verbs from simple judgement, which may be neutral or even disapproving. Kenny’s ‘volition’ also covers intention. Later, Kenny proposes to borrow Wittgenstein’s term ‘sentence-radical’ to describe the propositional form that can serve as

1403_902976_07_cha06.qxd 12/13/04 6:27 PM Page 130

130 Paul Grice, Philosopher and Linguist

the object of any verb of ‘voliting’. So ‘You live here now’ and ‘Live here now!’ share the same sentence-radical as a common element of meaning, and it is necessary to distinguish this from the ‘modal component, which indicates what function the presentation of this state of affairs has in communication’.32 By means of this idea, Kenny is able to identify a connection between his verbs of volition and those of judgement. There is some relation which holds between a man’s Idea of God and his Idea of England no matter whether he judges that God punishes England or he wants God to punish England. Any theory of judgement or volition should make this common element clear.33 In ‘Intention and uncertainty’, Grice employs Kenny’s notion of ‘voliting’ in drawing a connection between intention and willing, but he does not make much use of Kenny’s analysis in terms of ‘sentenceradicals’. This clearly influenced his thinking in this area, however, and its effects can be seen in ‘Probability, desirability and mood operators’, a paper written and circulated in 1972, but never published. The connection between probability and desirability, or at least between probability statements and desirability statements, had been discussed by Donald Davidson in a paper Grice heard at the Chapel Hill Colloquium in Philosophy at the University of North Carolina in the autumn of 1967. Davidson was another of the highly formal American philosophers of logic and language influential on Grice’s thinking and methodology at this time. In this paper, eventually published in 1970, Davidson uses pr and pf, qualitative operators with scope over pairs of propositions that express attitudes of the probable and the obligatory respectively. Thus, ‘pr (Rx, Fx)’ can be instantiated as ‘That the barometer falls probabilizes that it will rain’, or, more naturally ‘If the barometer falls, it almost certainly will rain’. A moral judgements such as ‘lying is wrong’ is to be understood, again as a relation between two propositions, as something such as ‘That an act is a lie prima facie makes it wrong’. This can be symbolised as ‘pf (Wx, Lx)’.34 Grice considers attitudes towards probability and desirability in terms of the functions they serve in human cognition. This is reminiscent of his work in ‘Indicative conditionals’ on the purpose of logical connectives such as ‘⁄’ and ‘…’ in terms of their roles in reasoning processes. Here, he argues that probabilistic argument functions in order to reach belief in a certain proposition. Moreover:

1403_902976_07_cha06.qxd 12/13/04 6:27 PM Page 131

American Formalism 131

The corresponding function of practical argument should be to reach an intention or decision; and what should correspond to saying ‘p’, as an expression of one’s belief, is saying ‘I shall do A’, as an expression of one’s intention or decision.35 Going further than Davidson, Grice argues that structures expressing probability and desirability are not merely analogous; they can both be replaced by more complex structures containing a common element. Grice proposes two types of operators, OpA and OpB. In combination, these replace Davidson’s pf and pr. The operators grouped together as OpB represent moods close to ordinary indicatives and imperatives. They can be divided into two types: OpB1 and OpB2, corresponding to ⵫ and ! respectively. A-type operators, on the other hand, represent some degree or measure of acceptability or justification. They can take scope over either of the B-type operators, yielding ‘OpA1 + OpB1 + p’, or ‘OpA1 + ⵫ + p’ for an expression of ‘it is probable that p’ and ‘OpA1 + OpB2 + a’, or ‘OpA1 + ! + p’ for an expression of ‘it is desirable that a’. Moving on from operators to consider the psychological aspect of reasoning, Grice proposes two basic propositional attitudes: J-acceptance and V-acceptance, to be considered as more or less closely related to believing and wanting. Generalising over attitudes using the symbol ‘y’, he proposes ‘X y1 [p]’ for J-accepts and ‘X y2 [p]’ for V-accepts. There are further, more complex attitudes y3 and y4. These are reflexive: attitudes that x can take to J-accepting or V-accepting. y3 is concerned with an attitude of V-accepting towards either J-accepting [p] or J-accepting [~p]; x wants to decide whether to believe p or not. y4 is concerned with an attitude of V-accepting towards either x V-accepts [p] or x V-accepts [~p]; x wants to decide whether to will p or not. On the understanding that willing p gives an account of intending p, this offers a formalisation of intending. Grice notes that for each attitude there are two further subdivisions, depending on whether the attitude is focused on an attitude of x or of some other person. Therefore ‘x y3A [p]’ is true just in case ‘x y2 [x y1 [p] or x y1 [~p]]’ is true; ‘x y3B [p]’ is true just in case ‘x V-accepts (y2) [y V-accepts (y2) [x J-accept (y1) [p] or x J-accept (y1) [~p]]]’ is true. Grice suggests an operator, Opia, corresponding to each particular propositional attitude y3B, where ‘i’ is a dummy taking the place of either 1, 2, 3, or 4, and where ‘a’ is a dummy taking the place of either ‘A’ or ‘B’. He now has four sets of operators, corresponding to four sets of propositional attitudes, which he describes as follows:

1403_902976_07_cha06.qxd 12/13/04 6:27 PM Page 132

132 Paul Grice, Philosopher and Linguist

Op1a Judicative (A cases Indicative, B cases Informative) Op2a Volitive (A cases Intentional, B cases Imperative) Op3a Judicative Interrogative (A cases Reflective, B cases Imperative) Op4a Volitive Interrogative (A cases Reflective, B cases Inquisitive) For all ‘moods’ except the second, the syntax of English does not reflect the distinction between the A and B cases. Grice notes that ‘in any application of the scheme to ordinary discourse, this fact would have to be accommodated’. This suggests that he was thinking of his scheme as at least in principle applicable to an analysis of everyday language. For present purposes, he proposes to concentrate on the simple judicative and volitive operators. This is presumably because these express his basic psychological categories of J-accepting and V-accepting. He has, however, established that it is possible to discuss more complex operators, and therefore more complex psychological attitudes, a topic to which he was to return. The attitudes expressed by ⵫, and by the pair and !A and !B (‘I shall do A’ and ‘Do A’) can be expressed by a general psychological verb of ‘accepts’. So, for instance, ‘x J-accepts [p]’ is ‘x accepts [⵫p]’ and ‘x V-accepts [p]’ is ‘x accepts [!Ap]’. Grice makes it clear that, in discussing meaning, he is still describing a way of speaking, a type of procedure for achieving a goal. The semantic characterisation of the operators consists in specifying the associated procedures. This position has an obvious kinship with views of Searle; one place where we differ is that I am not prepared to treat as primitive the notion of rule (whether constitutive or otherwise), nor that of convention (indeed I am dubious about the relevance of the latter notion to the analysis of meaning and other semantic concepts). The notion of a rule seems to me extremely unclear, and its clarification would be one result which I hope might be obtained, ultimately, from the line of investigation which I am pursuing in this paper. The written dialogue between Davidson and Grice continued. In 1978 Davidson published a paper entitled simply ‘Intending’, part of which

1403_902976_07_cha06.qxd 12/13/04 6:27 PM Page 133

American Formalism 133

he devoted to a response to Grice’s ‘Intention and uncertainty’. Like his earlier paper, ‘Intending’ was the result of various drafts and conference papers, including one presented at the Chapel Hill Colloquium in Philosophy in October 1974. There, Davidson has commented, ‘it received a thorough going over by Paul Grice.’36 In ‘Intending’, Davidson suggests, or rather presupposes, that it is reasonable to ‘want to give an account of the concept of intention that does not invoke unanalysed episodes or attitudes like willing, mysterious acts of the will, or kinds of causation foreign to science’.37 He sees a problem here in accounting for ‘pure intending’, a type not accompanied by any observable behaviour or consequence. He proposes instead a definition, or a path towards a definition, in terms of beliefs and attitudes that can be taken as rationalising an intention. Davidson acknowledges Grice’s point that intentions are often conditional on certain issues, and that the conditions are often not expressed, or not articulated, for various reasons. He quotes Grice’s dialogue about the fact that X might miss the concert by being in prison. His argument is that, however misleading X’s original statement of intention may have been, it was not strictly incorrect. It is not inaccurate to say ‘I intend to go to the concert’ when you know that there may well be a reason why you cannot go, anymore than it is contradictory to say ‘I intend to be there, but I may not be there’. Indeed, stating every condition that might prevent the fulfilment of an intention is not practicable, even if it were desirable. ‘We do not necessarily believe we will do what we intend to do, and . . . we do not state our intentions more accurately by making them conditional on all the circumstances in whose presence we think we would act.’38 Davidson considers intention in terms of wanting, but not as a subclass of wanting. Indeed, he argues that it is possible for someone to say that he wants to go to London next week, but he does not intend to go because there are other things he wants to do more. Rather, both intending and wanting are separate sub-parts of a general psychological proattitude, and they are expressed by value judgements. Grice’s ‘thorough going over’ of Davidson’s paper was never published, but it was tape recorded and subsequently transcribed. In it he outlines the position he took in ‘Intention and uncertainty’. ‘X intends to do A’ entails that X believes X will do A; X’s belief depends on evidence, but the relevant evidence derives from X’s own psychological state. For Davidson the relationship between intention and belief is not one of entailment, but rather by saying ‘I intend . . .’ in certain circumstances you suggest that you believe you will do it. Grice wonders whether it might be possible to explain Davidson’s position in terms of

1403_902976_07_cha06.qxd 12/13/04 6:27 PM Page 134

134 Paul Grice, Philosopher and Linguist

a conversational implicature from ‘I intend . . .’ to ‘I believe . . .’. But he comments wryly that ‘I have never been a foe to the idea of replacing alleged entailments by implicatures; but I have grave doubts whether this manoeuvre is appropriate in this case.’39 Grice builds his case around a distinction between extensional and non-extensional intentions. In effect, a non-extensional intention is one that the subject in question would recognise. This is the one most commonly considered in discussions of intention, Grice argues. If ‘The Dean intends to ruin the department (by appointing Snodgrass chairman)’ is true on a non-extensional reading, then the Dean would, if forced, be in a position to aver ‘I intend to ruin the Department’. If the Dean says ‘I shall ruin the Department’ he is expressing, rather than stating his intention. It would, contra Davidson’s claim, be inconsistent for the Dean to say ‘I shall ruin the Department, but maybe I won’t in fact ruin it’. The apparent counter-examples can be explained in terms of ‘disimplicature’. In effect, context sometimes means that normal entailments are suspended. If we say that Hamlet saw his father on the ramparts at Elsinore, in a context where it is generally known that Hamlet’s father is dead, then we are not committed to the usual entailment that Hamlet’s father was in fact on the ramparts. In such a context, the speaker ‘disimplicates’ that Hamlet’s father was on the ramparts. In the same way, when context makes it quite apparent that there may be forces that will prevent us from fulfilling an intention, we are not committed to the usual entailment that we believe we will fulfil it. A speaker who says ‘Bill intends to climb Everest next week’ disimplicates that Bill is sure he will climb Everest, just because everyone knows of the possibly prohibitive difficulties involved. The notion of disimplicature suggests some interesting possible extensions to Grice’s theory of conversation, but it does not seem to be one to which he returned. As it is presented in ‘Logic and conversation’, and therefore as it is generally known, implicature is a matter of adding meaning to ‘what is said’: to conventional or entailed meaning. With the notion of disimplicature, Grice appears to be conceding that the meaning conveyed by a speaker in a context may in fact be less than is entailed by the linguistic form used. In the cases under discussion, there is some particular element, some entailment, that is ‘dropped’ in context. However, he also hints that disimplicature can be ‘total, as in “You are the cream in my coffee” ’. This remark appears in parentheses and is not elaborated, but Grice is presumably suggesting that, in a context which makes the whole of ‘what is said’ untenable, it can be replaced with an implicated, metaphorical meaning. The mechanisms

1403_902976_07_cha06.qxd 12/13/04 6:27 PM Page 135

American Formalism 135

of disimplicature are not discussed at all, but they could presumably be explained in terms of the assumption that the speaker is abiding by the maxims of conversation, particularly the first maxim of Quality. If one or all the entailments of ‘what is said’ are plainly false, they can be assumed not to arise on that particular occasion of use. If the disimplicature is total, the hearer is forced to seek a different interpretation of the utterance. There are undoubtedly problems inherent in the notion of disimplicature, which would provide at least potential motivation for not pursuing the idea. In a 1971 article, Jonathan Cohen suggests a ‘Semantical Hypothesis’ as an alternative explanation of the phenomena of logical particles that Grice explains by means of his ‘Conversationalist Hypothesis’. Cohen’s suggestion is that some natural language expressions, such as ‘either . . . or’, differ from their apparent logical counterparts. In this particular case, ‘either . . . or’ differs from ‘⁄’ by entailing, as part of its dictionary meaning, that there is indirect evidence for the disjunction. In examples such as ‘The prize is either in the garden or in the attic, but I’m not going to tell you which’, however, this meaning does not survive. This is because ‘it is deleted or cancelled in certain contexts, just as the prefixing of “plastic” to “flower” deletes the suggestion of forming part of a plant.’40 The idea of ‘disimplicature’ seems remarkably close to Cohen’s looser notion of ‘deletion’, laying open the possibility that the very natural language expressions Grice was seeking to explain might after all differ semantically from their logical equivalents. In his reply to Davidson, as in his earlier unpublished paper, Grice considers the role of intentions in actual everyday cognition. ‘First we are creatures who do not, like the brutes, merely respond to the present; we are equipped to set up, in the present, responses in the more remote future to situations in the less remote future.’ Grice pictures our view of the future as an appointment book in which various entries are inscribed. We are aware of a distinction between entries we have no control over (inscribed in black) and those that are there by our control, or depend on us for their realisation (inscribed in red). These latter entries are intentions. If we can foresee an unpleasant consequent of such entries, we will try to avert it if we can by deleting a red entry. The Dean derives the conclusion from the existing entries that ‘Early December Dean gets fired’. He is understandably unhappy about this, but is in a position to do something about it if he can find that one of the entries from which he drew the conclusion is in red. He is able to delete, for instance, ‘Dean appoints Snodgrass’ inscribed in red in mid-November. The conclusion Grice draws from this is:

1403_902976_07_cha06.qxd 12/13/04 6:27 PM Page 136

136 Paul Grice, Philosopher and Linguist

that we need a concept which is related to that of being sure in the kind of way which I have suggested intending is related to being sure; and if we need such a concept we may presume that we have it. This sentence contains a large leap in the argument, but the recording of his original presentation reveals that Grice got a big laugh from the audience at this point. He presents this conclusion gleefully, provocatively ‘trying it on’, suggesting that ‘intend’ seems to him a good candidate for the name of the concept in question. This was, however, to prove a serious manoeuvre in his methodology. In the field of philosophical psychology in particular, he was to argue that the need for a particular concept in rational creatures was good enough reason to assume its existence, or at least posit it for the purpose of theorising. Grice challenges Davidson’s views on wanting, describing his picture of a person’s mental life as one of a series of wants competing to be assigned the status of intention. For Grice this is just too neat and organised to be plausible. It does not even reflect how the matter is described in ordinary language. There is in fact a vast range of terms in the ‘wanting-family’ of verbs, and these demand careful attention. Contra Davidson’s claim, we would be very unlikely to say ‘I want to go to London next week but I don’t intend to because there are other things I want more’. We are far more likely to use some phrase such as ‘I would like to go to London but . . .’. In the following elaborately extended metaphor, Grice expresses his unease with the notion of a series of competing wants, arguing that rationality is necessary to impose some order or restraint on pre-rational emotions and impulses: It seems to me that the picture of the soul suggested by D’s treatment of wanting is remarkably tranquil and, one might almost say, computerised. It is a picture of an ideally decorous board meeting, at which the various heads of sections advance, from the stand-point of their particular provinces, the case for or against some proposed course of action. In the end the chairman passes judgement, effective for action; normally judiciously, though sometimes he is for one reason or another over-impressed with the presentations made by some particular member. My soul doesn’t seem to me, a lot of the time, to be like that at all. It is more like a particularly unpleasant department meeting, in which some members shout, won’t listen, and suborn other members to lie on their behalf; while the chairman, who is often himself under suspicion of cheating, endeavours to impose some kind of order; frequently to no effect, since some-

1403_902976_07_cha06.qxd 12/13/04 6:27 PM Page 137

American Formalism 137

times the meeting breaks up in disorder, sometimes, though it appears to end comfortably, in reality all sorts of enduring lesions are set up, and sometimes, whatever the outcome of the meeting, individual members go off and do things unilaterally. In the final section of his reply, Grice returns to the discussion of methodology. He takes issue with Davidson’s aim of avoiding the mysterious and the unexplained in accounting for intention. On the contrary, in philosophy of mind, ‘a certain kind of mysteriousness’ is to be prized. A concept is useful to the extent that it can explain behaviour, and this consideration outweighs some degree of opaqueness in the concept itself, as in the case of Grice’s J-accepting and V-accepting. A psychological theory will be more or less rich, depending on the complexity of the behaviour, and therefore of the necessary psychological states, of the creature under investigation. In the seven years from the time of his move to America, Grice had developed his interest in the philosophy of language to include the relationship between semantics and syntax. This had involved him in considerations of how the structure of language may mirror the structure of thought, which had in turn led him back with new insights to some old interests, including the nature of intention, and the role of psychological concepts such as this in linguistic meaning. Interest in how language is used to express thought about the world prompted a consideration of psychological concepts and ultimately a return to a topic from much earlier in his career: the philosophical status of rationality.

1403_902976_08_cha07.qxd 12/13/04 6:28 PM Page 138

7 Philosophical Psychology

In 1975 Grice was made full professor at Berkeley, and in the same year he was elected president of the Pacific division of the American Philosophical Association. Now in his sixties, there was no apparent diminution in his restless energy and intellectual enthusiasm. He travelled frequently to speak at conferences, deliver lecture series and lead discussions, mainly within the United States. His reputation benefited from this exposure, and also from the wider dissemination of his work on implicature. He finally published the second and third William James lectures belatedly and apparently reluctantly during the 1970s, under the titles ‘Logic and conversation’ and ‘Further notes on logic and conversation’.1 Meanwhile, Grice was cutting an increasingly striking figure around the Berkeley campus. He was always a large man and, apparently unmoved by California’s growing interest in healthy eating, he had recently been putting on weight. In his youth he had been proud of his blond curls, but now his hair was thinning and whitish, and grew down over his collar. His clothes were chosen for comfort and familiarity rather than for style or fashion. He was most offended, however, when it was once suggested that his trousers were held up by a piece of string. Wasn’t it obvious, he demanded, that he was using two old cricket club ties? Grice did not, perhaps, deliberately cultivate his image as an eminent but eccentric Englishman abroad, but he does seem to have taken some pleasure in the effect he produced. Drinking one evening in Berkeley’s prestigious Claremont Hotel, he was approached by a total stranger who wanted to know who he was: ‘I don’t recognise you, but I’m sure you must be someone distinguished.’ Perhaps more telling than the anecdote itself is the fact that Grice was immensely pleased by the stranger’s comment. 138

1403_902976_08_cha07.qxd 12/13/04 6:28 PM Page 139

Philosophical Psychology 139

Grice’s chaotic untidiness extended from his person to his immediate surroundings. His desks at home and on campus were piled high with an apparent miscellany of notes, manuscripts and correspondence. He drove a series of old cars, all bought cheaply and constantly breaking down, a particular problem for Grice who had neither ability nor inclination to fix them. His disregard for appearances may have been the result of preoccupation or absent-mindedness, but it did not represent an embracing of the laissez-faire, or ‘hippy’ attitude of his students’ generation, as has sometimes been suggested.2 In fact, Grice seems to have been wary of the values this represented. Very early in the 1970s he jotted down an extensive list, apparently simply for his own benefit, of opposing ‘good’ and ‘bad’ values in this new ethos. The list headed ‘Good Things’ includes ‘doing your own thing (self-expression)’, ‘independence (cf. authority)’, ‘new (unusual) experiences (self-expansion)’ and ‘equality’. The ‘Bad (or at least not good) Things’ include ‘(self) discipline’, ‘loyalty (except political)’, ‘tolerance (except towards what we favour)’ and ‘culture (except popular)’. Some items on the list are marked with as asterisk, which Grice annotates as ‘some v. bad’. These include ‘discrimination’, ‘getting unfair advantage’ and ‘authority’.3 There is no mistaking Grice’s disapproval for the new ethos, which may seem strange in the light of his own affectionate reminiscences of ‘pinko Oxford’, with its avowed dislike of discipline and authority. In fact, the rebelliousness of Grice’s Oxford contemporaries was, as he himself half admits in those reminiscences, more a pose or a show of dissent than a real revolt. It drew on a lifestyle that depended on privilege maintained by the status quo. Grice was made uneasy by the more businesslike rebellion of the 1960s and 1970s, with its challenge to authorities, political, social and intellectual. If Grice was somewhat wary of his students’ values, this did not diminish his enthusiasm and success in teaching them. He was exceptionally bad at the paper work and administration accompanying his teaching duties. Repeated requests from administrative staff for course reading lists went unanswered and sometimes unopened, course descriptions in student handbooks were often brief to the point of evasion, and lists of grades were submitted late or not at all. However, he seems to have taken the business of teaching itself much more seriously. Those who have heard him generally agree that he was not a talented lecturer.4 Tapes of his lectures bear this out; they are full of hesitations, false starts and labyrinthine digressions. However, the few surviving tapes of Grice’s seminars suggest that his approach lent itself more naturally to this form of teaching. A tape from 1978 of a seminar

1403_902976_08_cha07.qxd 12/13/04 6:28 PM Page 140

140 Paul Grice, Philosopher and Linguist

on theories of truth records him patiently drawing out responses from his students, encouraging even faltering attempts to form answers. Above all, he was interested in nurturing sensitivity to philosophical method, and in reinforcing students in the habit of forming their own judgements. In the same seminar, as he helps along the discussion of truth, he draws out the significant steps in the argument: consider what criticisms might be made of our position, then think about how we might get out of them.5 In teaching, he still emphasised linguistic intuition. Former student Nancy Cartwright has recalled a seminar on metaphysics in the mid-1970s in which Grice encouraged his students to consider the theoretical claims of physics by determining where they would incorporate ‘as if’.6 A note from a graduate student from the late 1970s includes the comment, ‘I’d say that you’ve already increased my belief in paying attention to our linguistic intuitions.’7 As a supervisor he was encouraging and attentive. Judith Baker, one of Grice’s doctoral students at this time, has commented that, ‘When we discussed my thesis, Grice completely entered into my research and questions. He had a great gift for looking at what another philosopher wished to understand.’8 It does seem that, despite his rather cavalier attitude to some of his professional duties, Grice was genuinely concerned about the welfare of his Berkeley students. Some of his notes for the Urbana lectures are written on the back of a draft letter in which he contributes to a faculty discussion of ‘Graduate Programme Revision’. Recycling a favourite piece of terminology, he suggests a series of ‘desiderata’ for the graduate programme. These are concerned not so much with syllabus content as with the more general student experience. They include comments about increasing student access to faculty, as well as points such as ‘reduce or eliminate 3rd year doldrums’ and ‘foster a sense of personal adequacy, together with the idea that philosophy can be an exciting and enjoyable activity’. He makes a final and particularly impassioned point about ‘financial discrimination’, interesting in the light of his rather wary attitude to new notions of equality. ‘A pass at QE should not be nullified by poverty’, he argues ‘and the added anxiety would be bad in itself, and bad for student harmony and for the flow of applications for admission to the program.’9 As well as a growing academic reputation, and popularity among his students, Grice enjoyed at Berkeley productive intellectual friendships of the sort he had valued at Oxford. These were largely formed with younger philosophers such as George Myro, Richard Warner and Judith Baker. He collaborated with them in teaching and in writing, but above

1403_902976_08_cha07.qxd 12/13/04 6:28 PM Page 141

Philosophical Psychology 141

all he enjoyed long, informal discussions. He seemed to find it easier to put together and explore his own ideas in conversation with others than in solitude. These discussions often took place at the house in the Berkeley hills, with Kathleen on hand to cook whatever meal was appropriate to the time of day. Grice commented: To my mind, getting together with others to do philosophy should be very much like getting together with others to make music: lively yet sensitive interaction is directed towards a common end, in the case of philosophy a better grasp of some fragment of philosophical truth; and if, as sometimes happens, harmony is sufficiently great to allow collaboration as authors, then so much the better.10 Perhaps the most significant collaboration of this period was that with Judith Baker. They produced copious notes on ethics, particularly drawing on Kant and on Aristotle. They completed a book-length manuscript called ‘Reflections on morals’, credited to ‘Paul Grice & Judy Baker or Judy Baker & Paul Grice’ and planned and partly wrote another called ‘The power structure of the soul’.11 These were intended for publication, but have not yet appeared.12 The only published work to result from their long collaboration was one article, ‘Davidson on “Weakness of the will” ’, in which they return to some of the themes discussed by Grice in the unpublished ‘Probability, desirability and mood operators’.13 Grice’s chief solo interest during the 1970s was not very far removed from the themes of his collaborative work with Judith Baker. Nor, typically, did it represent a completely new direction in his thinking. Throughout his philosophical career, Grice’s choices of topic tended to follow on from and build on each other, rather than forming discreet units. In this case, he returned to ideas that had interested him at least since his early work towards the William James lectures, namely the psychological property of rationality, its status as a defining feature of human cognition, and its consequences for human behaviour. Grice suggested that he drew out, and perhaps saw, the significance of rationality to his theory of conversation more clearly in later commentary on the work than in his original formulation of it.14 He did mention rationality in the William James lectures, insisting that to follow the Cooperative Principle was to do no more than behave rationally in relation to the aims of a conversation. However, looking back over his work at the end of his life, he goes further, remarking on the essentially psychological nature of his theory of conversation. It was aimed not at

1403_902976_08_cha07.qxd 12/13/04 6:28 PM Page 142

142 Paul Grice, Philosopher and Linguist

describing the exact practices and regularities of everyday conversation, but at explaining the ways in which people attribute psychological states or attitudes to each other, in conversation or in any other type of interaction. Crucially, it saw conversation, an aspect of human behaviour, as essentially a rational activity, and sought to explain this rationality. Grice defends himself against criticisms that his theory overlooks argumentative, deceitful or idle conversations: ‘it is only certain aspects of our conversational practice which are candidates for evaluation, namely those which are crucial to its rationality rather than to whatever other merits or demerits it may possess.’15 Properties of rationality, he argues, can legitimately be abstracted away from more specific conversational details. Some of Grice’s notes from the year or so immediately preceding the William James lectures suggest that even at that time he was concerned with the analysis of reason for its own sake. The back of a letter about the 1966 AGM of the North Oxford Cricket Club, for instance, is devoted to an imaginative weighing up of the respective pros and cons of spending July teaching philosophy at the University of Wigan, and of spending it on a cricket tour. Grice is looking for an analysis of ‘reasons for doing’, perhaps for the type of language in which people usually express such arguments. On the same paper he tries out different uses of the modal verb ‘should’.16 In other notes from the same year he starts considering the distinction between ‘reasons for’ and ‘reasons why’.17 There seems then to have been a break in this line of thought; Grice did little with the idea of reason during the time of his move to America, while he was concentrating on syntax and semantics. However, he returned to the topic in the early 1970s, somehow finding the earlier notes in his idiosyncratic filing system and annotating and extending them. Returning to the distinction between ‘reasons for . . .’ and ‘reasons why . . .’, he adopts a distinctively ‘ordinary language philosophy’ approach as he tries out different uses and occurrences of the phrases. One note is actually headed ‘botanizing’ and lists, under the heading ‘Reasons (practical)’, such phrases as ‘the reason why he j’d (action) was . . .’, ‘there was (a) reason to j’, ‘there was (a) reason for him to j (action, attitude: fire x, distrust x)’, ‘he had reason to j’, ‘he had a reason for j-ing’, ‘his reasons for j-ing’. Grice makes notes on the distinction between ‘the reason why x . . . was that p’ and ‘x’s reason for . . . was that/so that/to p’. He notes that for Aristotle: Reasons for believing, if accepted, culminate in my believing something (where what I believe is the conclusion of the argument asso-

1403_902976_08_cha07.qxd 12/13/04 6:28 PM Page 143

Philosophical Psychology 143

ciated with these reason). Reasons for doing, if accepted, culminate in my doing something, where what I do is the conclusion of the ‘practical’ aspect associated with these reasons. i.e. the conclusion of a practical argument is an action.18 Grice presented the results of these deliberations in a series of lectures called ‘Aspects of reason’, as the Immanuel Kant lectures at Stanford in 1977. In 1979 he used them again as the John Locke lectures in Oxford, this time adding an extra lecture to the series. The lectures were eventually edited by Richard Warner and published in 2001. The invitation to present the John Locke lectures was the occasion of one of Grice’s rare trips to England. He made conscientious and surprisingly neat packing lists, including careful reminders to himself of the notes and books he would take with him to work on when not lecturing. His list of notes includes ‘Kant notes – Aristotle notes – Incontinence – Ethics (Judy) – Presupposition’. The books he was packing represent some heavy, if predictable, reading: ‘Kant, Aristotle, Leibniz, Russell’. Perhaps less predictably, the list starts with the anomalous and unspecified ‘Zoology’. Grice’s interest in the study and classification of living beings was to show itself faintly in the lectures on reason, and more clearly in the work to follow. At the start of the John Locke lectures, Grice comments on his pleasure at being back at Oxford: I find it a moving experience to be, within these splendid and none too ancient walls, once more engaged in my old occupation of rendering what is clear obscure. I am, at the same time, proud of my mid-Atlantic status, and am, therefore, delighted that the Old World should have called me in, or rather recalled me, to redress, for once, the balance of my having left her for the New.19 He was no doubt highlighting the distinction between Old and New World for rhetorical effect. But there was something genuine in his claim to ‘mid-Atlantic status’; just as he was always too prominently British fully to integrate into Californian life, he was no longer comfortably at home in Oxford. Notes sent and received at the time, inevitably covered with scribbled jottings to do with reason, indicate that he stayed in his old college, St Johns, and arranged to meet many of his former colleagues for drinks and dinners. But he had to reaccommodate to the formalities of Oxford. The back of a programme of a match between Oxford University Cricket Club and Hampshire shows

1403_902976_08_cha07.qxd 12/13/04 6:28 PM Page 144

144 Paul Grice, Philosopher and Linguist

him reminding himself to order ‘dress shirt, black tie’. He also needs to remember ‘shoe polish/brushes OR get cleaned’ and ‘button, needle, thread’. He poses himself a question that would not have troubled him for some years: ‘gown or not gown?’20 There is no doubt that Grice knew the rituals of Oxford life well enough. However, the act of readapting to them seems to have involved him in some deliberate and rather strained effort after more than a decade in the very different social milieu of Berkeley. Some of his attitudes had changed as radically as his dress, although these changes were more carefully concealed from his former colleagues. Some years previously, soon after his move to America, he had confided in Kathleen that he had discovered that very few of the students at Berkeley knew any Greek; most read Aristotle in translation. Further, and rather to Kathleen’s surprise, he expressed himself quite happy with this state of affairs; there were now some very good translations available. Kathleen was well aware that he would never have admitted to such an opinion in the presence of any of his former Oxford colleagues.21 The first of the John Locke lectures is titled ‘Reasons and reasoning’. Grice describes his desire to clarify the notion of reason, a notion deemed worthy of attention by both Aristotle and Kant. Typically circumspect in his proposal, he suggests that such a clarification may make some interesting philosophical conclusions possible, but that the actual conclusions are beyond the scope of these lectures. He prefaces his consideration of the technicalities of reasoning with a fairly lengthy discussion of a rather more abbreviated notion. This is the definition of reasoning with which most ordinary people are familiar. It is tempting here to see a parallel with Grice’s theory of conversation. The study of the literal meanings of words is crucially important, and shows up the striking parallels between natural language and logic, but it does not tell us everything about how people actually use words. In the same way, the logical processes of reason are of supreme philosophical importance, but nevertheless need to be seen as distinct from people’s everyday experience of reasoning. Introspection is of limited use in establishing the processes of reasoning because people frequently take short cuts and make inferential leaps in their thinking. To do so is often beneficial in terms of the time and energy saved, and indeed might constitute a definition of ‘intelligence’.22 In his treatment of reason, Grice draws on and develops the line of argument from the unpublished ‘Probability, desirability and mood operators’. The earlier paper was circulated in mimeograph, although far less widely than ‘Logic and conversation’.23 It may have been at least

1403_902976_08_cha07.qxd 12/13/04 6:28 PM Page 145

Philosophical Psychology 145

in part because much of it was subsumed into these lectures that it was not itself published. Grice’s method of examining and clarifying the notion of reason is one Austin would have recognised and approved. He presents the results of the dusting down of the methods of linguistic botanising apparent in his notes from the previous few years. He considers different ways in which the word ‘reason’ is used, classifies these uses into different categories, and illustrates these categories with examples. A sentence such as ‘The reason why the bridge collapsed was that the girders were made of cellophane’ exemplifies what he describes as ‘explanatory reasons’.24 The same type of meaning is conveyed by variants such as ‘the reason for the collapse of the bridge was that . . .’ and ‘the fact that the girders were made of cellophane was the reason why the bridge collapsed’. In all such examples the fact about the girders is offered as an explanation, or a causal account, of the collapse of the bridge. They are elaborate versions of Grice’s original ‘reason for’ category. Such examples are ‘factive’ with respect to both events; the speaker implies the truth of both the fact about the bridge and the fact about the girders. Second, there are ‘justificatory reasons’, exemplified by ‘the fact that they were a day late was a reason for thinking that the bridge had collapsed’ and ‘the fact that they were a day late was a reason for postponing the conference’. There are many possible variants on these patterns, including ‘he had reason to think that . . . (to postpone . . .) but he seemed unaware of the fact’ and ‘the fact that they were so late was a reason for wanting to postpone the meeting’. These are all variations of Grice’s original category of ‘reason why’. They offer some support, although not inevitably necessary or sufficient justification, for a psychological state (‘thinking’, ‘wanting’) or an action (‘postponing’). Such examples are factive with respect to the reason given, but do not guarantee the truth of the other event (the collapse of the bridge, the postponement of the conference). Grice labels the third type of use of ‘reason’ ‘JustificatoryExplanatory’, because of their hybrid nature; they draw on aspects of both the preceding classes.25 Examples include ‘John’s reason for thinking Samantha to be a witch was that he had suddenly turned into a frog’ and ‘John’s reason for denouncing Samantha was to protect himself against recurrent metamorphosis’. These reasons are relative to a particular person. As the examples illustrate, they may be introduced by either ‘that . . .’ or ‘to . . .’. In the first case, the resulting sentence is doubly factive unless ‘X thought that’ is inserted before the reason. In the second case, the sentence is only singly factive. In each case, if the

1403_902976_08_cha07.qxd 12/13/04 6:28 PM Page 146

146 Paul Grice, Philosopher and Linguist

relevant individual believes that the reason gives sufficient justification for the attitude (‘thinking’) or action (‘denouncing’) then the reason can be seen as an explanation for that attitude or action. Grice in fact suggests that this type of reason is a special case of explanatory reason; ‘they explain, but what they explain are actions and certain psychological attitudes’.26 Having spent considerable time and gone into some complexity in distinguishing these classes of ‘reason’, both in the published lectures and in his preparatory notes, Grice appears to drop them rather abruptly. He turns to a consideration of practical and non-practical, or alethic reasons: in effect the reasons to do and reasons to believe that had concerned him in ‘Probability, desirability and mood operators’. Grice does little to smooth the transition between these two topics, but does mention in passing that he will be concentrating exclusively on justificatory reasons. In seems that concentrating on ‘reasons to’ would enable him to look at the bases for intentional actions and for psychological attitudes. These two aspects of human psychology had provided an underlying thread in his work at least since ‘Meaning’. Grice refers to Kant’s theory that there is one faculty of reason, a faculty that manifests itself in a practical and a non-practical aspect. For Grice, the nature and identity of a faculty is prohibitively difficult to investigate in itself. The language used to describe reason, however, is amenable to rigorous analysis. He therefore proposes to approach Kant’s theory by addressing the question of whether ‘reason’ has the same meaning or different meanings in ‘practical reason’ and ‘non-practical reason’. He notes that a variety of ordinary language expressions used in reasoning can apparently describe both. The word ‘reason’ itself is one example, but so too are ‘justification’, modal verbs such as ‘ought’ and ‘should’, and phrases such as ‘it is to be expected’. In effect, he is seeking to apply his Modified Occam’s Razor to the vocabulary of reason by avoiding systematic ambiguities between separate meanings for such a range of natural language vocabulary. In doing so, he draws on his own earlier response to Davidson’s suggestion of the operators ‘pr’ and ‘pf’. He proposes that all reasoning statements have a common component of underlying structure, as well as an indication of the semantic subtype to which they belong. He posits as the common component a ‘rationality’ operator, which he labels ‘Acc’, and for which he suggests the interpretation ‘it is reasonable that’. This is then followed by one of two mood-operators, symbolised by ‘⵫’ for alethic statements and ‘!’ for practical statements. These two operators precede, and have scope over, the ‘radical’ (‘r’). So the structure for ‘John should

1403_902976_08_cha07.qxd 12/13/04 6:28 PM Page 147

Philosophical Psychology 147

be recovering his health now’ is ‘Acc + ⵫ + r’, while the structure for ‘John should join AA’ is ‘Acc + ! + r’.27 Grice suggests that the mood-operators be interpreted as two different forms of acceptance, technically labelled ‘J-acceptance’ and ‘V-acceptance’, but informally equated with judging and willing, or with ‘thinking (that p)’ and ‘wanting (that p)’. The mood-operators place conditions on when it is appropriate for a speaker to use a reasoning statement, in other words when speakers feel justified in expressing acceptance of a proposition. This raises the question of what is to count as relevant justification. Alethic statements are clearly justified by evidence. We have reason for making a statement about our knowledge or belief of a certain proposition just in case we have evidence for the proposition. Ideally, we are justified in using such statements if the proposition in question is true. But truth is not a possible means of validating statements of obligation, command, or intention. To use some concrete examples (which Grice does not do), ‘That must be the new secretary’, interpreted alethically, would be optimally legitimate in a context where its ‘radical’, namely ‘that is the new secretary’, is true. The legitimacy of ‘You must not kill’, interpreted practically, however, cannot be assessed in terms of the truth of ‘you don’t kill’. There must be some other grounds on which such statements can be judged. Grice’s suggests that, in reasoning generally, we are interested in deriving statements we perceive as having some sort of ‘value’. Truth, certainly, is one type of value, and alethic reasoning is aimed at deriving true statements. But in practical reasoning we are concerned with another type of value, which might be described as ‘goodness’. Grice’s own summary of this difference, in which he implicitly links the two types of reasoning to his two broad psychological categories of mood operator, runs as follows: We have judicative sentences (‘⵫’-sentences) which are assigned truth (or falsity) just in case their radicals qualify for radical truth (or radical falsity); and we have volitive sentences (‘!’-sentences) which are assigned practical value (or disvalue) just in case their radicals qualify for radical goodness (or badness). Since the sentential forms will indicate which kind of value is involved, we can use the generic term ‘satisfactory’.28 In his discussion of practical value, Grice is drawing on themes he presumably expects to be familiar to his audience from Aristotle’s Nico-

1403_902976_08_cha07.qxd 12/13/04 6:28 PM Page 148

148 Paul Grice, Philosopher and Linguist

machean Ethics, although he does not make any explicit reference. Grice himself was well versed in this work, which was a particular focus for his old tutor W. F. R. Hardie, and on which he himself worked collaboratively with Austin. Aristotle identifies rationality as a defining characteristic of human beings, distinguishing them from all other life forms. The rational nature of human beings ensures that all their actions are aimed at some particular end or set of ends. Although these ends differ depending on the specific field of activity, ‘if there is any one thing that is the end of all actions, this will be the practical good’.29 Similarly, Grice considers practical value (he says very little more about ‘goodness’ here) as an end for which people strive in what they wish to happen, and therefore in what they choose to do. Here, again, he returns to the type of rationality he discussed in the William James lectures. Rational beings have an expectation that actions will have certain consequences, and are therefore prepared to pursue a line of present behaviour in the expectation of certain future outcomes. They are prepared to conform to regularities of communicative behaviour in the expectation of being able to give and receive information. This, however, is just one type of a much more general pattern of human behaviour. All actions are ultimately geared towards particular ends. The extra lecture Grice added to the series at Oxford is ‘Some reflections about ends and happiness’, originally delivered as a paper at the Chapel Hill colloquium in 1976. Here too he draws on Aristotle’s Nicomachean Ethics, this time more explicitly. Aristotle suggested that the philosopher contemplating the range of different ends for different actions is confronted with the question of the nature of the ultimate end, or the ‘supreme good’ to which all actions aim. The answer is not hard to find, but the significance of that answer is more complex. ‘ “It is happiness”, say both ordinary and cultured people; and they identify happiness with living well or doing well.’30 For ordinary people, however, living well involves something obvious such as wealth, health or pleasure. For the philosopher the nature of happiness is much more complex, and is linked to the proper function of a man. Individual men have different specific functions, and are judged according to those specific functions; a good flautist is one who plays the flute well and a good artist one who paints well. More generally, a good man is to be defined as one who performs the proper functions of a man well. Because rationality is the distinguishing feature of mankind, the function of a man is to live the rational life, and a good man is one who lives the rational life well. Therefore the good for man, or that which results in happiness, is ‘an activity of soul in accordance with virtue, or if there are more

1403_902976_08_cha07.qxd 12/13/04 6:28 PM Page 149

Philosophical Psychology 149

kinds of virtue than one, in accordance with the best and more perfect kind.’31 For Grice, happiness is a complex end; that is, it is constituted by the set of ends conducive to happiness. The precise set of ends in question may vary from individual to individual, but they can be seen a conducive to happiness if they offer a relatively stable system for the guidance of life. Grice finishes on a leading but inconclusive note. It seems intuitively plausible that it should be possible to distinguish between different sets of ends in respect of some external notion. Otherwise it would, uncomfortably, be impossible to choose between the routes to happiness selected ‘by a hermit, by a monomaniacal stamp-collector, by an unwavering egotist, and by a well-balanced, kindly country gentleman’.32 Grice’s cautious proposal of the basis for a notion of ‘happinessin-general’, abstracted from individuals’ idiosyncrasies, draws on the essential characteristics of a rational being. He concludes his lecture series with the following, highly suggestive outline: The ends involved in the idea of happiness-in-general would, perhaps, be the realization in abundance, in various forms specific to individual men, of those capacities with which a creatureconstructor would have to endow creatures in order to make them maximally viable in human living conditions, that is, in the widest manageable range of different environments. As Richard Warner indicates in a footnote, the original Chapel Hill lecture ends with the further comment: ‘But I have now almost exactly reached the beginning of the paper which, till recently, you thought I was going to read to you tonight, on the derivability of ethical principles. It is a pity that I have used up my time.’ The rather enigmatic reference to a ‘creature-constructor’, and the link to ethics with which Grice tantalised his Chapel Hill audience, draw on ideas developed over a number of years but not discussed in these lectures. Grice was working on the notion of a ‘creature-constructor’ back in the mid-1960s. He in fact originally planned to include discussion of this topic in the William James lectures. Papers from Oxford from 1966 include a jotted note entitled ‘Provisional Grand Plan for James Lectures and Seminars’. There are two bullet points: 1. Use ‘God’ as expository device. Imagine creature, to behaviour of which ‘think’, ‘know’, ‘goal’, ‘want’, ‘purpose’ can be applied. But no

1403_902976_08_cha07.qxd 12/13/04 6:28 PM Page 150

150 Paul Grice, Philosopher and Linguist

communication. Goals continuation of self (or of species). Imagine God wishing to improve on such creatures, by enabling them to enlarge their individual worlds and power over them, by equipping them for transmission of beliefs and wants (why such change ‘advance’; because of decreased vulnerability to vicissitudes of nature). 2. Mechanisms for such development. Elements in behaviour must in some way ‘represent’ beliefs and wants (or their objects) or rather states of the world which are specially associated with them.33 Grice’s private note in fact contains the germs of most of his complex ‘creature-constructor’ programme, and of the topics to which he attempted to link it throughout the rest of his life. It also explains his interest in zoology. He was studying life in terms of the structures into which it has been classified. His idea was that it might, theoretically at least, be possible to devise a ladder or hierarchy of creatures, such that each rung on the ladder represents a creature incorporating the capacities of the creature on the lower rung, and adding some extra capacity. As a presentational convenience, Grice proposes an imaginary agent or designer of this process, the expository device ‘God’ of his early notes. In later notes, and certainly in all published work, he prefers ‘creatureconstructor’ or ‘Genitor’. The Genitor’s task is to decide what faculties to bestow on each successive creature so as better to equip it to fulfil its ends. The chief end of all life forms is to ensure their own continued survival, or the survival of the species to which they belong. Grice’s hierarchy could be seen as an extension of Aristotle’s division of living creatures in the Nicomachean Ethics into those that experience nutrition and growth (including plants), those that are sentient (including horses and cattle) and those that sustain ‘practical life of the rational part’ (only human beings).34 Grice does not mention this, or indeed Locke’s discussion of personal identity, familiar to him from his work on the topic more than 30 years earlier. Locke posited a hierarchy of living beings, with the identity of beings at each stage depending on the fulfilment of functions necessary to the continued existence of that being. Grice would presumably have assumed that the philosophical pedigree of his ideas would be transparent to his audience. He goes much further than Locke by employing the hierarchy of beings in a mental exercise to assess the status and function of human rationality. In order to describe, or to represent, the place of human beings on the ladder of existence, Grice recycles the ‘pirots’ that had appeared in the ‘Lectures on language and reality’. There, pirots had branched out

1403_902976_08_cha07.qxd 12/13/04 6:28 PM Page 151

Philosophical Psychology 151

from carulising elatically in order to model the psychological processes of communication. Now, they stand in for human beings in the Genitor’s deliberations. They appeared in this guise as early as 1973 when Grice was teaching a course at Berkeley called ‘Problems in philosophical psychology’. One of his students took the following notes in an early seminar in this series: ‘one of the goals of the pirot programme is to discover what parallels, if any, emerge between the laws governing pirots and human psychology’.35 Grice was returning to his earlier interest in psychological states. He had addressed these back in the early 1950s when he had considered the best way of describing intentions. In ‘Dispositions and intentions’ he had criticised dispositional accounts that described intentions in terms of dispositions to act. In particular he saw Ryle’s version of this as coming dangerously near to behaviourism and the denial of any privileged access to our own mental states. His attitude seems to have softened somewhat in the intervening twenty years. The student’s notes describe the pirot programme as an ‘offspring of dispositional analyses of metal states from Ryle C of M. But behaviourists unable to provide conditions for someone being in particular mental state purely in terms of extensionally described behaviour.’ The pirot programme was designed to offer the bridge behaviourism could not provide, by describing underlying mental states derived from, but ontologically distinct from, observable behaviour. It was a functional account, in that each posited mental state was related to a specific function it played in the creature’s observable behaviour; ‘in functional account, functional states specified in terms of their relation to inner states as well as to behaviour’. Further, the student noted, ‘a functional theory is concerned not only with behavioral output and input of a system, but of [sic] the functional organization of inner states which result in such outputs.’ ‘Philosophical psychology’, then, was Grice’s label for the metaphysical enquiry into the mental properties and structures underlying observable human behaviour. In contrast to a purely materialist position such as that of behaviourism, Grice was arguing in favour of positing mental states in so far as they seemed essential to explaining observable behaviour. Such explanatory power was sufficient to earn a posited mental property a place in a philosophical psychology. This openness to the inclusion of mental phenomena in philosophical theories was apparent in his earlier work on intentions, and also in the William James lectures. He was later to describe such an approach as ‘constructivist’. The philosopher constructs the framework of mental structure, as dictated by the behaviour in need of explanation. The

1403_902976_08_cha07.qxd 12/13/04 6:28 PM Page 152

152 Paul Grice, Philosopher and Linguist

philosopher is able to remain ambivalent as to the reality or otherwise of the posited mental state, so long as they are explanatorily successful. In fact, the ability to play an active role in explaining behaviour is the only available measure of reality for mental states. In the pirot programme, the fiction of the Genitor provided a dramatised account of this process, with decisions about what mental faculties would promote beneficial behaviours playing directly into the endowment of those faculties. Grice had presented some of his developing ideas along these lines in a lecture delivered and published before the John Locke lectures, but in which the place of rationality in ‘creature-construction’ is more fully developed. ‘Method in philosophical psychology’ was Grice’s presidential address to the Pacific division of the American Philosophical Association in March 1975. In general, the lecture is concerned with establishing a sound philosophical basis for analysing psychological states and processes. He sums up his attitude to the use of mental states in a metaphor: ‘My taste is for keeping open house to all sorts of conditions and entities, just so long as when they come in they help with the housework’.36 In keeping with this policy of tolerance towards useful metaphysical entities, Grice proposes two psychological primitives, chosen for their potential explanatory value. These are the predicateconstants J and V which, as he was to do in the John Locke lectures, he relates loosely to the familiar notions of judging, or believing, and willing, or wanting. Concentrating on V, we can see ways in which positing such a mental state can help explain simple behaviour in a living creature. An individual may be said to belong to class T of creatures if that individual possesses certain capacities which are constitutive of membership in T, as well perhaps as other, incidental capacities. These constitutive capacities are such that they require the supply of certain necessities N. Indeed the need for N in part defines the relevant capacity. Prolonged absence of some N will render the creature unable to fulfil some necessary capacity, and will ultimately threaten its ability to fulfil all its other necessary conditions. Such an absence will therefore threaten the creature’s continued membership of T. In these circumstances, we can say that the creature develops a mental attitude of V with respect to N. This mental attitude is directed at the fulfilment of a necessary capacity; its ultimate end is therefore the continued survival of the creature. A concrete example of this process, a simplified version of the one offered by Grice, is to think of T as representing the class ‘squirrel’ and N as representing ‘nuts’. One necessary condition for a creature to be a

1403_902976_08_cha07.qxd 12/13/04 6:28 PM Page 153

Philosophical Psychology 153

squirrel is, let us say, its capacity to eat nuts. This capacity is focused around its own fulfilment; it is manifested in the eating of nuts. If the creature experiences prolonged deprivation of nuts, its ability to fulfil this capacity, and indeed all other squirrel-capacities, will be threatened (it will be in danger of death by starvation). In such a situation, it is legitimate to say that the squirrel wants nuts. This posited mental state explains the squirrel’s behaviour (it will eat nuts if it can find them) in relation to a certain end (survival).37 A metaphysical theory that offers a model of why a squirrel eats nuts may not seem particularly impressive, but it is a necessary first step in modelling behaviour in much more complex creatures: ultimately, in describing human psychology. In making this transition Grice draws on the notion of ‘creature-construction’. The genitor is bound by one rule; every capacity, just like the squirrel’s capacity to eat nuts, must be useful or beneficial to the creature, at least in terms of survival. This rule is sufficient to ensure that any mental properties introduced into the system are there for a reason, fulfilling Grice’s criterion of the usefulness of metaphysical entities. It also, perhaps, allows Grice to imply that such mental capacities might be evolutionarily derived, without having to commit to this or any other biological theory. The final stage in the mental exercise is to consider what further, incidental features may arise as side-effects of these capacities for survival. Grice argues that it is not unreasonable to suggest that, in designing some of the more complex creations, the genitor might decide to endow creatures with rationality. This capacity would qualify as having survival benefits. For creatures living in complex, changing environments, it is more beneficial to have the capacity to reason about their environment and choose a strategy for survival, than to be endowed with a series of separate, and necessarily finite, reflex responses. Just as with the more primitive capacity of eating, it is necessary for the complex creature to utilise the capacity of reason in order to retain continued use of it, and indeed of the set of all its capacities. But once the creature is reasoning, the subjects to which it applies this capacity will not be limited to basic survival. A creature with the capacity to hypothesise about the relationship between means and end will not stop at deciding whether to hide under or climb into a tree, for instance. A creature from the top of the scale of complexity will be able to consider the hypothetical processes of its own construction. That creature will assess how each capacity might be in place simply to promote survival, and might confront a bleak but inevitable question: what is the point of its own continued survival?

1403_902976_08_cha07.qxd 12/13/04 6:28 PM Page 154

154 Paul Grice, Philosopher and Linguist

The creature’s capacity for reason is exactly the property to help it survive this question. Rather than persuading the creature to give up the struggle for survival, the capacity is utilised to produce a set of criteria for evaluating ends. These criteria will be used to attribute value to existing ends, which promote survival, and to further ends that the creature will establish. The creature will use the criteria to evaluate its own actions and, by extension, the actions of other rational creatures. The justification the creature constructs for the pursuit of certain ends, dependent on the set of criteria, will provide it with justification for its continued existence. Thus the capacity of rationality both raises and answers the question ‘Why go on surviving?’38 Grice hints at the possible beginnings, and status, of ethics, when he suggests that the pursuit of these ends might be prized by the collective rational creatures to such an extent that they are compiled into a ‘manual’ for living. The creatures’ purposes in existing are no longer just survival, but following certain prescribed lines of conduct. Grice tantalisingly ends the section of his lecture on ‘creature construction’ with a nod at some possible further implications of this idea. Such a manual might, perhaps, without ineptitude be called an IMMANUEL; and the very intelligent rational pirots, each of whom both composes it and from time to time heeds it, might indeed be ourselves (in our better moments, of course).39 Somewhat at odds with his more specific usage elsewhere, Grice has here been using the term ‘pirot’ to refer to any creature on the scale. The ‘very intelligent rational pirots’ are the psychologically most complex of such creatures. In adopting this term, which is added in the margin of his preparatory notes, Grice would appear to be making a pun on Locke’s ‘very intelligent rational parrot’, the imaginary case used to argue that bodily form must have some bearing on personal identity.40 Other than this knowing reference, Grice does not make any attempt to link his work on creature construction to Locke, or to his own earlier interest in personal identity. Richard Warner has criticised Grice for ignoring the connections between rationality, happiness and personal identity, arguing that a notion of self will determine what counts as an essential end for an individual. If, he suggests, we are to pursue Grice’s strategy – and it is a promising one – of replacing Aristotle’s ‘activity in conformity with virtue’ with ‘suitability for the direction of life’, I would suggests that an appeal to

1403_902976_08_cha07.qxd 12/13/04 6:28 PM Page 155

Philosophical Psychology 155

the notion of personal identity should play a role in the explication of suitability.41 As Grice’s reference to a list of ‘rules for living’ suggests, the next direction in which he explored these ideas was in the area of philosophical ethics. He did not, however, return to the implications of his pun on ‘manual’ and ‘Immanuel’. It is possible that he was referring to Immanuel Kant, whose work on reason was central to his own thinking on this subject. But the pun may also contain a hint towards an explanation of religious morality, and perhaps even religious belief, as facts of human existence, necessary consequences of human nature. In his own notes on the subject this aspect of the ‘Immanuel’ figures more prominently. In general, the plausibility of explaining religion within philosophical psychology seems to loom larger in Grice’s thinking than his published work of the time would suggest. For instance, on the back of an envelope from the bank, in faint, almost illegible hand he wonders: ‘perhaps Moses brought other “objectives” from Sinai, besides 10 comms’. Here he may be thinking about other systems for living, suggesting that ‘perhaps we might . . . “retune” them’. In the same notes he seems to be pondering the differences between ‘God’ and ‘good’, between ‘conception’ and ‘reality’. Perhaps, he speculates, the analytic/synthetic distinction can be treated along these lines; ‘real or pretend doesn’t matter provided . . . does explanatory job’.42 Grice seems here to be experimenting with the idea of ‘constructivism’ as a general philosophical method, not just as an approach to philosophical psychology. An entity, state or principle is to be welcomed if it introduces an increase in explanatory power. Beyond that it is not suitable, and indeed not possible, to enquire into its ‘reality’. A system of explanations is successful entirely to the extent that it is coherent, relatively simple, and explains the facts. It is not hard to discern the influence of Kant in this epistemology, and his notion that ‘things in themselves’ are inevitably unreachable, but that our systems of human knowledge can nevertheless seek to model reality. In his final decade, Grice turned such a philosophical approach in the direction of his interest in metaphysics, in philosophical psychology and, derived from this, in ethics. Anthony Kenny suggests that since the Renaissance the main interest from philosophers in human conduct has been epistemological: Moral philosophy has indeed been written in abundance: but it has not, for the most part, been based on any systematic examination of

1403_902976_08_cha07.qxd 12/13/04 6:28 PM Page 156

156 Paul Grice, Philosopher and Linguist

the concepts involved in the description and explanation of those human actions which it is the function of morals to enjoin or forbid, to criticise and to praise.43 Grice was perhaps unusual in approaching his work on ethics from the basis of many years’ work on precisely those concepts.

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 157

8 Metaphysics and Value

Grice retired from his post at Berkeley in 1980, at the age of 67. There was no question, however, that he would cease full-time philosophical activity. He continued to teach, both part-time at Berkeley and, for a few years in the early 1980s, as a visiting professor at the University of Washington in Seattle. He also continued to work on new ideas, to write and, increasingly, to publish. His reluctance ever to consider a piece of work finished or ready for scrutiny had meant that only 12 articles had been published during a career of over 40 years. And in almost all cases he had been either effectively obliged to include a paper in the volume of proceedings to which it belonged, or otherwise cajoled into publishing. It had, however, long been agreed that the William James lectures would be published in full by Harvard University Press and now finally Grice began to take an active interest in this process. He perhaps realised that it was not feasible to continue delaying the revisions and reformulations that he considered necessary. Perhaps, also, the enthusiasm with which his ideas were received in Berkeley, and the frank admiration of students and colleagues, had slowly worn away at his formidable self-doubt. Writing in Oxford, Strawson has suggested that Grice’s inhibiting perfectionism was ‘finally swept away by a warm tide of approbation such as is rarely experienced on these colder shores’.1 Certainly, he began to draw up numerous lists of papers on related themes that he would like to see published with the lectures. These included previously published articles such as ‘The causal theory of perception’, but also a number that still remained in manuscript form, such as ‘Common sense and scepticism’, and ‘Postwar Oxford philosophy’. This planned volume was focused on his philosophy of language, but Grice was also turning his attention to publishing in other areas. After a visit to Oxford in 1980, the year in which he was made an Honorary 157

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 158

158 Paul Grice, Philosopher and Linguist

Fellow of St John’s College, he received a letter from a representative of Oxford University Press. This followed up a meeting at which Grice had clearly been eager to discuss a number of different book projects. The letter refers to ‘the commentary on Kant’s Ethics which you have been working on with Professor Judy Baker’, and suggests that this might be the project closest to being in publishable form. However, the publisher expresses the hope that work on this ‘will not deflect you from also completing your John Locke lectures on Reason’, and a further work on Ethics.2 Grice’s desire to see some of his major projects through publication coincided with his retirement, and so could not have been driven by the demands of his post. Rather, it reflected a wish for some of the strands of his life’s work to reach a wider audience than the select number of students and colleagues who had attended his classes and heard his lectures. Grice’s interests in reason and in ethics increasingly absorbed him during the 1980s. They were prominent in his teaching at Seattle and at Berkeley. He himself summarised this change of emphasis as a move away from the formalism that had dominated his work for a decade or more. The main source of the retreat for formalism lay, he suggested, ‘in the fact that I began to devote the bulk of my attention to areas of philosophy other than philosophy of language’.3 Certainly, the discussion of reason and, particularly, of ethics in Grice’s later work takes a more metaphysical and discursive tone than that of some of the linguistic work from his early years in America. However, the break with the philosophy of language was not absolute; both new topics draw on questions about the correct analysis of various sentence types. In Grice’s treatment of reason, close attention to the use of certain key terms led him to posit a single faculty of rationality, yielding both practical and non-practical attitudes. From there, his discussion of expressions of practical reasoning led to the question of the appropriate type of value to serve as the end point of such reasoning, analogous to alethic value in non-practical reasoning. A simplistically defined notion of a system for the direction of life raised the awkward suggestion that it might be impossible to discriminate between a wide variety of modes of living. People are generally in agreement that the actions of Grice’s ‘wellbalanced, kindly country gentleman’, for instance, are to be afforded greater value than those of his ‘unwavering egotist’, but an account of value based simply on reasoning from means to ends seems unable to account for this. The study of ethics has long included analysis of statements of value. Sentences such as ‘stealing is wrong’ appear to draw on moral concepts;

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 159

Metaphysics and Value 159

it can therefore be argued that an accurate understanding of the meaning of such sentences is fundamental to any explanation of those moral concepts. Very broadly, there are two positions on this question: the objectivist and the subjectivist. One objectivist approach to ethics argues that moral values exist as ontologically distinct entities, external to any individual consciousness or opinion. Human beings by nature have access to these values, although perceptions of them may be more or less acute, observance of them more or less rigorous, and understanding of their implications more or less perceptive. Perhaps the most influential advocate of this moral objectivism was Kant, who described ethical statements as categorical imperatives; they are to be observed regardless of personal preference or individual circumstance. These he distinguishes from hypothetical imperatives, also often expressed in terms of what one ‘should’ do, but dependent on some desired goal or end. A canonical subjectivist position is that there are no moral absolutes; expressions of value judgements are always particular to subject and to context. Perhaps the most extreme version of this position was that put forward by A. J. Ayer, who followed his verificationist programme through to its logical conclusion by arguing in Language, Truth and Logic that sentences such as ‘stealing money is wrong’ have no factual meaning. Although apparently of subject/predicate form, they do not express a proposition at all. They merely indicate a personal, emotional response on the part of the speaker, much the same as writing ‘stealing money!!’, where ‘the shape and thickness of the exclamation marks show, by a suitable convention, that a special sort of moral disapproval is the feeling which is being expressed’.4 A less reductionist, but equally subjectivist position, is to analyse such statements as in fact having the form of ‘hypothetical imperatives’. That is, moral statements attempt to impose restrictions on behaviour, but do so in relation to certain actual or potential outcomes. They stipulate means to reach certain ends. These ends may be made explicit, as in sentences such as ‘if you want to be universally popular, you should be nice to people’. Often, however, they are suppressed, but can be understood as stipulating a general condition of success or advancement. So ‘stealing is wrong’ is hypothetical in nature, but this is hidden by its surface form. It can be interpreted as something such as ‘if you want to get on in life, you should not steal’. In his work on this subject, Grice draws particularly on two subjectivist analyses of statements of value as hypothetical imperatives: Philippa Foot’s 1972 essay ‘Morality as a system of hypothetical imperatives’ and J. L. Mackie’s 1977 book Ethics: Inventing Right and Wrong.

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 160

160 Paul Grice, Philosopher and Linguist

These choices are not surprising; Grice knew both Foot and Mackie personally from Oxford. In discussing the view that morality consists of a system of hypothetical imperatives, not of absolute values, Foot considers its implications for freedom and responsibility. Dispensing with absolute value raises the fear that people might some day choose to abandon accepted codes of morality. Perhaps this fear would be less, she suggests, ‘if people thought of themselves as volunteers banded together to fight for liberty and justice and against inhumanity and oppression’, rather than as being bound to conform to pre-existing moral imperatives.5 Mackie concedes that objectivism seems to be reflected in ordinary language. When people use statements of moral judgements, they do so with a sense that they are referring to objective moral values. Moreover, ‘I do not think that it is going too far to say that this assumption has been incorporated in the basic, conventional meanings of moral terms.’6 However, he argues that this in itself is not sufficient to establish the existence of objective value. The question is not linguistic but ontological. Mackie rehearses two fairly standard arguments to support the break with common sense he is proposing: the argument from relativity and the argument from queerness. The argument from relativity draws attention to the variety of moral codes and agreed values that have existed at different times in human history, and exist now in different cultures. This diversity, Mackie suggests, argues that moral values reflect ways of life within particular societies, rather than drawing on perceptions of moral absolutes external to those societies. However, he sees the argument from queerness as far more damaging to the notion of absolute value. It draws attention to both the type of entity, and the type of knowledge, this notion seems to demand. Objective values would be ‘entities or qualities or relations of a very strange sort, utterly different from anything else in the universe’, while knowledge of them ‘would have to be by some special faculty of moral perception or intuition, utterly different from our ordinary ways of knowing everything else’.7 The supporter of objective value is forced to advocate types of entity and types of knowledge of a ‘queer’ and highly specific kind. Despite Mackie’s concession that a commitment to objective value is apparent in ordinary language, it was far from necessary that a philosopher of ordinary language would take an objectivist stance on ethics. Indeed, at least one of Grice’s Oxford contemporaries who published on the topic presented a subjectivist argument. In 1954 P. H. Nowell-Smith published Ethics, the book in which he advanced rules of ‘contextual

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 161

Metaphysics and Value 161

implication’ that foreshadowed some of Grice’s maxims. Grice does not refer to this book in his own writings either on conversation or on value. Nowell-Smith defends a subjectivist account of moral value, enlisting the argument described by Mackie as ‘the argument from queerness’. In the tradition of post-war Oxford philosophy, Nowell-Smith employs different uses of words, and even fragmentary conversational ‘scripts’ to support his arguments. The meanings of individual words, he argues, often depend on the context in which they occur. For Nowell-Smith, this observation strengthens his subjectivist position. Since words do not remain constant, but are always changing their meanings, there can be no sentences that are true simply because of the meanings of the words they contain; in effect there can be no analytic sentences. Hence statements of moral value cannot be necessarily, but only contingently true.8 In 1983, Grice was working on another major lecture series, which he entitled ‘The conception of value’. He notes in the introductory lecture that the title is ambiguous, and deliberately so. The lectures are concerned both with ‘the item, whatever it may be, which one conceives, or conceives of, when one entertains the notion of value’, and with ‘the act, operation, or undertaking in which the entertainment of that notion consists’.9 In effect, they address the two areas in which Mackie argues objectivism introduces ‘queerness’. Grice suggests that the nature and status of value are of fundamental philosophical importance, while the way in which people think about value, their mental relationship to it, is deeply and intrinsically linked to the very nature of personhood. Value, as suggested in the John Locke lectures, is the ultimate end point and the measure of rationality. The value of any piece of alethic reasoning is dependent on the facts of the matter. The value of practical reasoning, however, is dependent on a notion of goodness. Grice does not state his plan for addressing this issue, or for linking it to an account of ethics. However, as the lectures progress it becomes clear that he is supporting a means–end account of practical reasons, and of assigning some motives and systems for living higher value than others. This, indeed, seems a necessary corollary of his account of reasoning in terms of goals, and of the theory of the derivability of ethical principles hinted at in the closing remark of ‘Some reflections on ends and happiness’. However, in one of his more flamboyant heresies, Grice proposes to retain the notion of absolute value. This notion, generally associated with Kant’s idea that moral imperatives are categorical, is canonically seen as being in direct conflict with an account of value concerned with goals, or with reasoning from means to ends.

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 162

162 Paul Grice, Philosopher and Linguist

‘The conception of value’ was written for the American Philosophical Association’s Carus lecture series, which Grice was to deliver in the same year. Progress was slow, particularly on the third and final lecture, in which his account of objective value was to be presented and defended. Grice had always preferred interaction to solitary writing, and as he got older he relied increasingly on constructive discussion and even on dictation. Unlike in earlier years, he produced very few draft versions of his writings during the 1980s; the papers that survive from this period consist only of very rough notes and completed manuscripts. Now, finding composition even more laborious than usual, Grice enlisted the help of Richard Warner. Warner, who was then working at the University of California, Los Angeles, offers a vivid account of the process: To work on the Carus lecture, I drove up from LA on Friday arriving at noon. We began work, drinking inexpensive white wine as the afternoon wore on; switched to sherry as we broke off to play chess before dinner; had a decent wine from Paul’s wine cellar with dinner and went back to work after eating. We worked all day Saturday with copious coffee in the morning (bread fried in bacon grease for breakfast) and kept the same afternoon and evening regime as the previous day. Paul did not sleep at all on Saturday night while he kept turning the problems over in his head. After breakfast on Sunday, he dictated, without notes, an almost final draft of the entire lecture.10 The result of this process was the third Carus lecture, ‘Metaphysics and value’. Grice addresses the properties common to the two types of value by which alethic and practical arguments are assessed, and the centrality of both types to the definition of a rational creature. In this extremely complex but also characteristically speculative discussion, he returns to the idea that every living creature possesses a set of capacities, the fulfilment of each of which is a necessary condition for the continued possession of that capacity, and indeed of all others. Here, the results of Grice’s interest in biology are apparent. On the plane on the way to Oxford in 1979 he had been reading books on this topic. In 1981 he made a note to himself to ‘Read “Selfish Gene”, “chimp” lit’.11 In other notes from this period he lists ‘mandatory functions’, including ‘Reproduction, Ingestion, Digestion, Excretion, Repair, Breath (why?), Cognition’.12 In ‘Metaphysics and value’, these thoughts are given a more formal shape as Grice considers again the distinction between essential and accidental properties of creatures. Essential prop-

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 163

Metaphysics and Value 163

erties establish membership of a certain class of creature: ‘the essential properties of a thing are properties which that thing cannot lose without ceasing to exist (if you like, ceasing to be identical with itself).’13 He speculates that rationality is in fact an accidental, rather than an essential property for the class of human being. However, as he argued in ‘Method in philosophical psychology’, once endowed with rationality, the creature will use the capacity to consider questions not just about how ends are to be achieved, but about the desirability of those ends. The creature in possession of rationality, the human being, is endowed with a concern that its attitudes and beliefs are well grounded, or validated. Grice speculates that, as a result of the possession of this property, the human being performs a process of ‘Metaphysical Transubstantiation’ on itself. It endorses this concern for validation with the status of an end in itself, and it transforms itself into a creature for whom this is the purpose: into a person. The properties of the human being and the person are the same, but they are differently distributed. Crucially, for this metaphysically constructed entity ‘person’, but not for the human being, rationality is an essential property. There is an unacknowledged link back to the work of John Locke in this suggestion, particularly to the work Grice drew on in his early article ‘Personal identity’. Locke proposed a distinction between ‘man’ and ‘person’ to take account of the fact that people, as essentially rational beings, are uniquely distinguished from other life forms by something like self-awareness or consciousness. For Locke, the identity of a person depends on the continuation of a single consciousness, while the identity of a man, like that of any living creature, depends on unity of function. Grice’s suggestion is similar to Locke’s in its metaphysical distinction between ‘man’ or ‘human being’ and ‘person’. However, he goes beyond Locke in attributing the distinction to the operation of rational creatures themselves, and in the link he proceeds to make between the transformed creature and the nature of value. For Grice, the rational creature is able to devise and evaluate answers to the questions about value it is raising. This will initially take the form of comparing different possible answers in terms of effectiveness and coherence. By means of raising questions about the justification of ends, and assessing possible answers to those questions, the person takes on the function of attributing value. A rational creature becomes a creature that, by metaphysical definition, has the capacity to attribute value. Value, even the type of value known as ‘goodness’, can be discussed objectively, because a creature uniquely endowed with, and indeed defined by, the capacity to judge it, exists. The creature is ‘a being whose

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 164

164 Paul Grice, Philosopher and Linguist

essence (indeed whose métier) it is to establish and to apply forms of absolute value’.14 Values exist because valuers exist. By means of this status afforded to the nature of personhood, it is perhaps possible to reassess Grice’s sketchy reference in ‘Aspects of reason’ to the devising of systems for living. People devise systems for living from the application of the capacity of reason and the resultant ability to ascribe value to certain courses of action and withhold it from others. The development and maintenance of basic moral principles can therefore be seen as a necessary product of rational human nature. In this way, Grice constructs his account of value as being dependent on analyses from means to ends, yet at the same time having objective existence. The process of Metaphysical Transubstantiation operates primarily to turn human beings into people and therefore valuers, but it thereby makes value viable. The process, and therefore the existence of valuers and value, derives from the capacity, incidentally present in human beings, for rationality. Neither value itself nor the process of valuing depend on any novel or ‘queer’ properties or processes. The nature of value, and the apprehension of value, can both be explained in terms of rationality, a phenomenon with an acceptable philosophical pedigree of its own, and one used in other areas of explanation. In Grice’s terms, value is embedded in the concept of people as rational creatures; ‘value does not somehow or another get in, it is there from the start’.15 In the years following the delivery of the Carus lectures, Grice began to justify his claim that the study of value could have illuminating and far-reaching consequences in different areas of philosophical investigation. He had already considered the implications of value for his own account of meaning in the paper ‘Meaning revisited’. As so often with Grice’s work, the date of publication is misleading. It was published in 1982 in the proceedings of a colloquium on mutual knowledge held in England, but an earlier version of part of the paper had been delivered at a colloquium in Toronto in March 1976. Grice offers a tantalisingly brief suggestion of what might happen if the very intelligent rational creature began to speak: an outline of an account of meaning within the framework of creature-construction to which, it seems, he did not return. If the creature could talk it would be desirable, and judged so by the creature, that its utterances would correspond with its psychological states. This would enable such beneficial occurrences as being able to exchange information about the environment with other creatures; it would be possible to transmit psychological states from one creature

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 165

Metaphysics and Value 165

to the next. Further, the creature will judge this a desirable property of the utterances of other creatures; it will assume, if possible, that the utterances of others correspond with their psychological states. In a sentence that is at once complex, hedged and vague, Grice proposes to retain some version of an intentional account of meaning: It seems to me that with regard to the possibility of using the notion of intention in a nested kind of way to explicate the notion of meaning, there are quite a variety of plausible, or at least not too implausible, analyses, which differ to a greater or lesser extent in detail, and at the moment I am not really concerned with trying to adjudicate between the various versions.16 In an addition to the 1976 paper for the later colloquium he suggests that intentional accounts of meaning, including his own, have left out an important concept. This is heralded at the start of the article as ‘The Mystery Package’, and later revealed to be the concept of value. Grice hints that he has firm theoretical reasons for wanting to introduce value as a philosophical concept. Many philosophers regard value with horror, yet it can answer some important philosophical problems. It makes possible a range of different expressions of value in different cases. Grice suggests the word ‘optimal’ as a generalisation over these different usages. This offers solutions to some problem areas in the intentional account of meaning, most significantly to the supposed ‘infinite regress’ identified by critics such as Schiffer. Grice expresses some scepticism over whether any such problem can in fact be found, but allows that it could at least be argued that his account entailed that ‘S want A to think “p, because S want A to think ‘p, because S wants . . .’ ” ’. In effect, this demands a mental state of S that is both logically impossible, because it is infinite, and yet highly desirable. Grice argues that just because it cannot in practice be attained, this mental state need not be ruled out from philosophical explanation. Indeed, ‘The whole idea of using expressions which are explained in terms of ideal limits would seem to me to operate in this way’, a procedure Grice links to Plato’s notion of universal Ideals and actual likenesses.17 The state in which a speaker has an infinite set of intentions is optimal in the case of meaning. This optimal state is unattainable; strictly speaking S can never in fact mean p. Nevertheless, actual mental states can approximate to it. Indeed, for communication to proceed successfully other people can and perhaps must deem the imperfect actual mental

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 166

166 Paul Grice, Philosopher and Linguist

state to be the optimal state required for meaning. Grice uses the story of the provost’s dog to illustrate the notion of ‘deeming’. It is a story of which he was apparently fond; he used it again in his later paper ‘Actions and events’. The incoming provost of an Oxford college owned a dog, yet college rules did not allow dogs in college. In response, the governing body of the college passed a resolution that deemed the provost’s dog to be a cat. ‘I suspect’, suggests Grice, ‘that crucially we do a lot of deeming, though perhaps not always in such an entertaining fashion.’18 In the second version of the paper the ‘mystery package’ seems to be the focus of Grice’s interest, but commentators at the colloquium on mutual knowledge were more interested in the possible ‘evolutionary’ implications of his ideas: the suggestion that non-natural meaning might be a descendant of natural meaning.19 This is not an idea that Grice seems to have pursued further. However, Jonathan Bennett had already postulated a possible Gricean origin for language, although carefully not committing himself to it.20 Grice revisited ‘Meaning’, but not ‘Logic and conversation’. Others, however, have drawn connections between rationality, value and cooperation. Judith Baker has suggested that, ‘when we cooperate with other humans on the basis of trust, our activity is intelligible only if we conceive of ourselves and others as essentially rational.’21 Marina Sbisa has gone further, making a more explicit link between Grice’s notions of ‘person’ and of ‘cooperative participant’ and arguing that ‘we might even envisage Metaphysical Transubstantiation as an archetypal form of ascription of rationality on which all contingent ascriptions of cooperativity are modelled.’22 Richard Grandy, tackling the question of why a hearer might conclude that a speaker is being cooperative, suggests tentatively that Grice ‘would have argued that the Cooperative Principle ought to be a governing principle for rational agents’ and that such agents would follow it ‘on moral, not on practical or utilitarian grounds.’23 However, writing long before the publication of Grice’s work on rationality and value, Asa Kasher suggested that the Cooperative Principle should be replaced by a more general and basic rationality principle. He argues that the individual maxims do not display any clear connection with the Cooperative Principle, but that they follow from the principle that one should choose the action that most effectively and efficiently attains a desired end. In this way the maxims ‘are based directly on conclusions of a general rationality principle, without mediation of a general principle of linguistic action’.24 Further, in interpreting an utterance, a hearer assumes that the speaker is a rational agent.

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 167

Metaphysics and Value 167

It might be that the discussion of value in Grice’s later work offers a new way of interpreting the rather problematic formulation of his earlier work. Rationality and mutual attribution of rationality may be better descriptions of the motivation for the maxims described in the William James lecturers than is ‘cooperation’. Speakers follow the rules of conversation in pursuit of particular communicative ends. On the assumption that other speakers are also rational creatures, the maxims also offer a good basis for interpreting their utterances. As rational creatures, speakers attribute value to ends that are consistent and coherent. An implicature may be accepted as the end point of interpretation if it accords with the twin views that the other speaker is a rational being, and that their utterances correspond to their psychological states. On this view, the Cooperative Principle and maxims are not an autonomous account of conversational behaviour precisely because they follow from the essential nature of speakers as rational creatures. Cooperation is, in effect, just one manifestation of the much more general cognitive capacity of rationality. Further, introducing the notion of ‘value’ into the theory of conversation offers a defence against the charge of prescriptivism. The maxims are not motivated by a duty to adhere to some imposed external ethical code concerned with ‘helpfulness’ and ‘considerateness’. Their imperative form belies the fact that they are a freely chosen but highly rational course of action. Like moral rules they are means towards certain ends: ends that are attributed value by creatures uniquely endowed to do so. Both the tendency to adhere to the maxims and the tendency to prize certain courses of action as ‘good’ follow as consequences of the same essential property of people as rational beings: the property, indeed the necessity, of attributing value. As such, the theory of conversation can be argued to be one component in Grice’s overarching scheme: a description of what it is to be human. The allocation of time and the completion of projects became a pressing concern for Grice as the 1980s wore on because his health was failing. He had been a heavy smoker throughout his adult life, and tape recordings from the 1970s onwards demonstrate that his speech was frequently interrupted by paroxysms of coughing. In response he gave up cigarettes suddenly and completely in 1980. He insisted to Kathleen that his last, unfinished, packet of one hundred Player’s Navy Cut stayed in the house, but he never touched it. This drastic action was taken, as he himself was in the habit of commenting, too late. In December 1983 he gave a farewell lecture at the end of his time at the University of Washington, in which he referred to ‘my recent disagreement with

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 168

168 Paul Grice, Philosopher and Linguist

the doctors’. Even before this had taken place, he explains, he had decided to offer this talk as a thank you on his and Judith Baker’s behalf, ‘for making the time spent by us here such fun’.25 Whether or not the doctors in question were insisting that Grice finish his demanding shuttle between Berkeley and Seattle, they certainly seem to have been recommending that he slow down. It was around this time that he was first diagnosed with emphysema, the lung condition that was presenting him with breathing difficulties and was to cause increasing ill health over the following years. If his doctors were urging him to burden himself with fewer commitments, they did not check his enthusiasm for work. Stephen Neale, then a graduate student at Stanford, spent a day with Grice in Berkeley in 1984. He found that Grice ‘was already seriously weakened by illness, but he displayed that extraordinary intellectual energy, wit, and eye for detail for which he was so well known, and I left his house elated and exhausted’.26 Grice’s response to his illness seems to have been if anything an increased impetus towards advancing the projects then in hand, and a renewed drive towards seeing his many unpublished works into print. Early in 1984 he began making lists of things to do. Some of these include private and financial business which, mixed in with academic matters, have survived with his work papers; there was, by this time at least, no real distinction between the personal and the philosophical. He planned to ‘collect “best copies” of unpublished works and inform “executors” of their nature and location and of planned “editing” ’.27 On the same list appears the resolution to ‘go through old letters and dispose’. In the mid-1980s, then, Grice was looking back over the work of more than four decades. This was prompted partly by his concern to tie up loose ends and partly by the demands of selecting the papers to appear along with the William James lectures in the projected book from Harvard University Press. There was also another spur to this retrospective reflection. Richard Grandy and Richard Warner were collecting and editing the essays eventually published as Philosophical Grounds of Rationality. They were themselves writing an introductory overview of Grice’s work. Grice called his response to this ‘Reply to Richards’, fusing, as he puts it, the editors into a multiple personality. In this he develops some of the themes of his current work. But he begins with a short philosophical memoir: ‘Life and opinions of Paul Grice’. The paper he offered at the end of his time in Seattle was, it seems, a preliminary sketch for this. The original, informal title was ‘Prejudices and predilections. Which becomes life and opinions’.

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 169

Metaphysics and Value 169

One consequence of this enforced nostalgia was a renewed interest in the practices of ordinary language philosophy. Grice looked back at the influence of Austin with the amused detachment afforded by a distance of over 20 years, or perhaps more accurately with fond memories of the amused detachment with which he viewed Austin at the time. He was earnest, however, in the guarded enthusiasm with which he recounted the art of ‘linguistic botanising’. He had never fully abandoned the practice in his own philosophy. For instance, in his notes on psychological philosophy, he had started making a list of ‘lexical words’ relevant to the topic. These included ‘Nature, Life, Matter, Mind, Sort, End, Time’. Perhaps catching himself in the act, next to this list he had posed the self-mocking question ‘ “Going through Dictionary”??’28 In ‘Reply to Richards’, he suggests that this method has merits that might recommend it even to contemporary philosophers: ‘I continue to believe that a more or less detailed study of the way we talk, in this or that region of discourse, is an indispensable foundation for much of the most fundamental kind of philosophizing.’29 This was a theme he repeated, often rather more forcefully, in a number of other papers, talks and informal discussions at that time. Nevertheless, Grice was well aware of how far he had moved away from Oxford philosophy in taking on metaphysical topics such as value. In 1978, in the introductory seminar to his course ‘Metaphysics’, he explained to his students that not very long before the word itself could be used as a term of abuse. At Oxford in the 1950s, metaphysics was viewed as ‘both non-existent and disgusting’. Oxford philosophers, he suggests, would have said: ‘All we need is tasteful and careful botanising and we’ll have all any gentleman needs to know’. Twenty-five years later, Grice admitted, he was convinced that a more systematic treatment of philosophy was needed.30 Grice’s enthusiasm for the old-fashioned method of linguistic botanising was accompanied, and perhaps reinforced, by a strong dislike and distrust of technology. This attitude applied to attempts to use technological tools or analogies in philosophy itself, and extended into the more personal domain. He was scathing about what he saw as attempts to ‘reduce’ the understanding of human beings to the understanding of computers. He may have had in mind recent debates about the possibility and nature of Artificial Intelligence, although he does not refer to them explicitly. In 1985, he wrote an overview of his earlier work on the philosophy of language, titled ‘Preliminary valediction’ and presumably serving as an early draft of the retrospective that appeared in Studies in the Way of Words. In this he makes an explicit link between his philosophical interest in ordinary language and his dislike of

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 170

170 Paul Grice, Philosopher and Linguist

technological explanations in philosophy. In a single sentence, perfectly well formed, but long even for him, he tackles the need for a carefully considered argument to justify attention to ordinary language: Indeed it seems to me that not merely is such argument required, but it is specially required in the current age of technology, when intuition and ordinary forms of speech and thought are all too often forgotten or despised, when the appearance of the vernacular serves only too often merely as a gap-sign to be replaced as soon as possible by jargon, and when many researchers not only believe that we are computers, but would be gravely disappointed should it turn out that we are, after all, not computers; is not a purely mechanical existence not only all we do have, but also all we should want to have?31 In arguing for the centrality of consciousness, that is in arguing against the possibility of Artificial Intelligence, Grice was again affirming his commitment to the significance of psychological concepts to rationality in general and to meaning in particular. The processes of the interpretation and production of meaning are significant for a human being in ways that are perhaps mysterious, but crucially cannot be observed in computers. Personally, too, Grice would have nothing to do with computers. When he retired from his full-time post at Berkeley he lost the secretarial support he had always relied on and was forced to get a PC at home. But it was Kathleen who learnt to use the computer and mastered word processing. Grice could never get beyond his horror of the spell checker which, he complained, rejected ‘pirot’ and questioned ‘sticky wicket’. Grice was not just looking back, reflecting on the philosophy of earlier decades and planning the publication of his manuscripts. He was also working on new projects. His obsession with these was such that even his interest in his own health seemed to be predicated on the desire to have time to finish them. Cancelling a lecture he had been scheduled to give in 1986, he wrote ‘my doctor has just advised me that I am taking on too much and that unless I reduce my level of activity I shall be jeopardising the enterprises which it is most important to me to complete.’32 From the notes and the work plans he was making at the time, it seems that Grice had two main enterprises in mind: producing a finished version of his lectures on value, and developing a project these had prompted. This new project was characteristic of the pattern of Grice’s work in two ways. It was daringly ambitious, in that he foresaw for it a wide range of implications and applications. But it was also not entirely

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 171

Metaphysics and Value 171

new, in that the ideas from which it developed had long been present in his work. In effect, it was nothing less than the foundations of a theory to encompass the whole of philosophy, with all its apparent branches and subdivisions. Grice was working on a metaphysical methodology to explain how the theories that make up human knowledge are constructed. Or, in one of his own favourite phrases from this time, he was doing ‘theory-theory’. Grice had long been of the opinion that the division of philosophy into different fields and disciplines was an artificial and unproductive practice. This opinion was certainly reflected in the course of his own work, in which he followed what seemed to him the natural progression of his interests, without regard for the boundaries he was crossing between the philosophies of mind and of language, logic, metaphysics and ethics. In 1983, he recalled with some pride a mild rebuke he had received years earlier from Gilbert Harman, to the effect that when it came to philosophy, ‘I seemed to want to do the whole thing myself’.33 To some extent, especially in the latter part of his career, Grice had actually defined his own philosophical interests in these terms. He contributed a brief self-portrait to the Berkeley Philosophy Department booklet in 1977, in which he described himself as working recently chiefly in the philosophy of language and logic; ‘takes the view, however, that philosophy is a unity and that no philosophical area can, in the end, be satisfactorily treated in isolation from other philosophical areas’.34 He comments further on this attitude in his ‘Reply to Richards’: When I visit an unfamiliar university and (as occasionally happens) I am introduced to ‘Mr Puddle, our man in Political Philosophy’ (or in ‘Nineteenth-Century Continental Philosophy’ or ‘Aesthetics’, as the case may be), I am immediately confident that either Mr Puddle is being under-described and in consequence maligned, or else Mr Puddle is not really good at his stuff. Philosophy, like virtue, is entire. Or, one might even dare to say, there is only one problem in philosophy, namely all of them.35 Grice’s distrust of subdivisions and labels extended to the classification of different schools of thought or philosophical approaches. Labelling philosophers according to the style of their subject and their approach to it was, he felt, unhelpful and even stultifying. In preparatory note for ‘Reply to Richards’ he lists some of the ‘cons’ of philosophy as being ‘fads’, ‘band wagons’, ‘sacred cows’ (the ‘pros’ are simply ‘fun’ and ‘col-

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 172

172 Paul Grice, Philosopher and Linguist

laboration’). He then goes on to produce a list of ‘-isms’ that might bear out his objections, including ‘naturalism’, ‘phenomenalism’, ‘reductionism’, ‘empiricism’ and ‘scepticism’.36 In a later note, he worked out a series of nicknames for the dedicated followers of different schools of thought. ‘Constructivists’ were ‘Egg-heads’; ‘Realists’ were ‘Fat-heads’; ‘Idealists’ were ‘Big-heads’; ‘Subjectivists’ were ‘Hard-heads’ or ‘Nutheads’; and ‘Metaphysical Sceptics’ were ‘No-heads’, or ‘Beheads’.37 It is not clear whether Grice intended these jottings for any purpose other than his own amusement. It seems that he made public use of only one of these labels, when speaking to the American Philosophical Association, and drawing attention to his own belief in the crucial connection between value and rationality: To reject or to ignore this thesis seems to me to be to go about as far as one can go in the direction of Intellectual Hara-Kiri; yet, I fear, it is a journey which lures increasingly many ‘hard-headed’ (even perhaps bone-headed) travellers.38 In ‘Reply to Richards’, and elsewhere, Grice is characteristically cagey about the exact consequences of his views on the entirety of philosophy. He is certain, however, that progress is to be made through his view of metaphysics as essentially a constructivist enterprise, in which categories and entities are added if they are explanatorily useful. In ‘Method in philosophical psychology’ he used the metaphor of offering houseroom to even unexplained conditions and entities if they were able to help with the housework. In the Carus Lectures, arguing that any account of value should be given metaphysical backing, he suggests that the appropriate metaphysics will be constructivist, not reductionist. In another arresting metaphor, he tells his audience that, unlike reductionist philosophers, ‘I do not want to make the elaborate furniture of the world dissolve into a number of simple pieces of kitchenware.’ The world is rich and complex, and the philosopher should offer an explanation that retains these qualities. The constructivist would begin with certain elements that might be seen as metaphysically primary, and would then ‘build up from these starting-points, stage by stage, a systematic metaphysical theory or concatenation of theories’.39 First, it would be necessary to define the suitable form for any theory to take, to engage in theory-theory. Grice does not here enter into a discussion of theory-theory itself, but it seems that his account of value is an example fitting the general constructivist approach he has in mind. The simple psychological attitudes of J-accepting and V-accepting are

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 173

Metaphysics and Value 173

posited, together with the programme of creature construction, as primary elements in the theory. Grice’s elaborate account involving rationality as a non-essential property of human beings, the process of Metaphysical Transubstantiation, the creation of people as value-givers and the derivation of absolute value, are built up from these starting points. There is no finished written account of metaphysical constructivism and theory-theory, although they are outlined in ‘Reply to Richards’. Grice never wrote the work ‘From Genesis to Revelation (and back): a new discourse on metaphysics’, which he seems to have contemplated.40 Probably the fullest record of his thoughts on this matter is a lengthy taped conversation from January 1983, in which he talks to Judith Baker and Richard Warner, setting out the ideas he hopes to have time to develop. The discipline of theory-theory, he explains, can be discussed in both general and applied terms. In general terms, it is concerned with examining the features any adequate theory would have to exhibit: with laying out the ground rules of theory construction. In its more applied aspect, it is concerned with detailing the theories formed by these rules, and with classifying them into types. Theory-theory includes in its scope, but is not restricted to, philosophical theory. From this starting point, Grice goes on to suggest a hierarchical structure of theories, the scope and subject matter at each level being successively more restricted and specific. He offers an example of a theory from each of these levels; in effect each theory is a component of the one before. If philosophical theory is one type demarcated by theorytheory, then metaphysics is the pre-eminent form of philosophical theory. Concerned as it is with the fundamental questions of what there is, and how it is to be classified, it deserves the title of ‘First Philosophy’. General theory-theory is concerned with specifying the appropriate subject matter and procedures of metaphysics, while applied theory-theory is concerned with the specific categories needed to delineate the subject matter, with investigating what kind of entities there are. Within metaphysics, each sub-branch has its own general and applied theory-theory. One such branch that particularly interests Grice is philosophical psychology. General theory-theory is concerned with the nature and function of psychological concepts, applied theorytheory with finding a systematic way of defining the range of psychological beings. Rational psychology is a special case. Here, general theory-theory is concerned with the essence of rational beings. Applied theory-theory widens the scope to consider features or properties other than the essential. In this way, it leads to the characterisation of a prac-

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 174

174 Paul Grice, Philosopher and Linguist

tical theory of conduct, in part an ethical system. Finally, and rather briefly, Grice comments that philosophy of language is to be seen as a sub-theory of rational psychology. Grice is even more tentative about the operation of metaphysical construction itself, in other words the procedures of metaphysics specified in general theory-theory. The relevant principles, he insists, are not fully worked out, but two ideas about the forms they might take have been with him for some time. The first idea he calls the ‘bootstrap principle’, in response to the realisation that a theory constructed along the lines he is describing must be able to ‘pull itself up by its bootstraps’, or rely on its own primitive concepts to construct potentially elaborate explanations. A metaphysical system contains both its primitive subject matter, and the means by which that subject matter is discussed and explained. The means are meta-systematic. In constructing the theory, it must be possible to include in the meta-system any materials that are needed. The only restriction on this, introduced in the interests of economical conduct, is that any ideas or entities used in the meta-system should later be introduced formally into the system. The legitimate routines or procedures for construction include ‘Humean projection’ and ‘nominalisation’. First, if a particular psychological state is central to a theory, it is legitimate to use the object of those states in the metasystem. The objects of psychological states are the ‘radicals’ of thought described in Grice’s work on rationality, Second, it is legitimate to introduce entities corresponding to certain expressions used in the relevant area of investigation. The second general idea does not have a specific name, but can be summarised by saying that the business of metaphysics in any area of investigation is not just the production of one theory. Grice maintains that he wants to be free to introduce more than one metaphysical theory, because the production of different systems of explanation may be differently motivated. This principle, he observes, ‘fits in nicely with my ecumenical tendencies’. A metaphysical system does not have to be unitary or monolithic. In fact, Grice doubts the existence of ‘a theory to be revealed on The Day of Judgement which will explain everything’. Rather than aiming at some ultimate truth, or correct version, Grice sees any theory as crucially a revision of some prior theory, or, if there is no such precedent, as building on common sense and intuition. Asked by one of his colleagues as to whether it is legitimate to reject a previous theory, or the prompting of common sense, once it has been revised, Grice expresses himself strongly opposed to the idea. The best understanding of some entity comes from an appreciation of all the guises in

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 175

Metaphysics and Value 175

which it appears during the development of systems. The most appropriate theory on any given occasion is not always the most elaborate one available. Grice referred to his metaphysical interests in some of the articles he wrote, or at least completed, during these final but productive years. In ‘Actions and events’, published in 1986, he responds to Donald Davidson’s logical analysis of action sentences that included actions in the category of entities.41 The question relates, he argues, to the fundamental issue of metaphysics: the question of what the ‘universe’ or the ‘world’ contains. Grice is inclined to agree with Davidson over this act of inclusion, but to do so from a very different metaphysical position. He suggests that Davidson, like Quine, is a ‘Diagnostic Realist’; the actual nature of reality is unavailable for inspection, but the job of philosophers is to make plausible hypotheses about it. Grice’s own metaphysical approach, however, is one of constructivism, of building up an account from primitives, rather than of forming theories consistent with scientific explanations. ‘One might say that, broadly speaking, whereas the Diagnostic Realist relies on hypothesis, the Constructivist relies on hypostasis.’42 ‘Actions and events’ brings together many of Grice’s apparently disparate philosophical interests into one article of broad scope and remarkable complexity. Still not recognising traditional philosophical boundaries, he moves from metaphysics, intention and language to rationality and the status of value. In a few concluding paragraphs, Grice turns to what he sees as the essential link between action and freedom. The link is treated briefly and cautiously in this published work, but was in fact a major concern in his thinking about value. Grice’s notes from the early 1980s show him applying the familiar techniques of ‘linguistic botanising’ to the concept of freedom. He jotted down phrases such as ‘alcohol-free’, ‘free for lunch’ and ‘free-wheeling’, and listed many possible definitions, including ‘liberal’, ‘acting without restriction’ and ‘frank in conversation’.43 Richard Warner has commented that their discussions in preparation for the third Carus lecture were concerned almost exclusively with freedom as a source of value, and that he himself encouraged Grice to pursue this line of enquiry rather than that of creature construction as being the more interesting and productive.44 Essentially, Grice saw in the nature of action a means of justifying the problematic concept of freedom, and from this a defence of his objective conception of value. He introduces a classification of actions that is reminiscent of his hierarchy of living creatures. Some actions are caused by influences external to a body, as is the case

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 176

176 Paul Grice, Philosopher and Linguist

with inanimate objects. Next, actions may have causes that are internal to the body but are simply the outcome of a previous stage in the same process, as in a ‘freely moving’ body. Then there are causes that are both internal and independently motivated. Actions provoked by these causes may be said to be based on the beliefs or desires of the creature in question and as functioning to provide for the good of the creature. Finally, there is a stage exclusive to human creatures at which the creature’s conception of something as being for its own good is sufficient to initiate the creature in performing an action. ‘It is at this stage that rational activity and intentions appear on the scene.’45 The particular nature of human action suggests the necessity of freedom. That is, humans act for reasons. These reasons may be more than mere desires and beliefs motivated by ones own survival; they may be freely adopted for their own sake. This in turn suggests that people must be creatures capable of ascribing value to certain ends, independent of external forces or internal necessity. In Grice’s terms, the particular type of freedom demanded by the nature of human action ‘would ensure that some actions may be represented as directed to ends which are not merely mine, but which are freely adopted or pursued by me’.46 In this way, a consideration of action suggests the necessity of freedom, and freedom in turn offers an independent justification for people as ascribers of value. If people are inherently capable of ascribing value, then value must have an objective existence. Independent of the programme of creature construction, the concept of freedom may offer another version of the argument that value exists because valuers exist. In ‘Metaphysics, philosophical escatology and Plato’s Republic’ completed in 1988 and published only in Studies in the Way of Words, Grice suggests a distinction between ‘categorial’ and ‘supracategorial’ metaphysics. The first is concerned with the most basic classifications, the second with bringing together categorially different items into single classifications, and specifying the principles for these combinations. This second type of metaphysics, closely related to his informal account of constructivism, he tentatively labels ‘Philosophical Escatology’.47 The study of escatology, or of last or final things, belonged originally to Christian theology. Grice is laying claim to an exclusively philosophical version of this discipline, and thereby linking metaphysical definitions to finality, or purpose. The paper ‘Aristotle on the multiplicity of being’ was published in 1988, but there is evidence that he had been working on it at least since the late 1970s. Alan Code mentions a paper ‘Aristotle on being and

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 177

Metaphysics and Value 177

good’ that Grice read at a conference at the University of Victoria in 1979, and that sounds in many ways similar to the published version.48 Grice offers an explicit defence of metaphysics, contrasting it particularly with the rigid empiricism of logical positivism: A definition of the nature and range of metaphysical enquiries is among the most formidable of philosophical tasks; we need all the help we can get, particularly at a time when metaphysicians have only recently begun to re-emerge from the closet, and to my mind are still hampered by the aftermath of decades of ridicule and vilification at the hands of the rednecks of Vienna and their adherents.49 Grice saw the ‘revisionary character’ of metaphysical theorising as a topic to which he would need to return. In fact, it is closely linked to another late area of interest, a topic he often described as ‘the Vulgar and the Learned’. In this terminology, and indeed in the ideas it described, it seems likely that Grice was borrowing from Hume, one of the ‘old boys’ of philosophy he particularly admired. Hume had written about the relationship between philosophical terminology and everyday understanding in the following terms: When I view this table and that chimney, nothing is present to me but particular perceptions, which are of a like nature with all other perceptions. This is the doctrine of philosophers. But this table, which is present to me, and that chimney, may and do exist separately. This is the doctrine of the vulgar, and implies no contradiction. There is no contradiction, therefore, in extending the same doctrine to all perceptions.50 Grice argued in his account of theory-theory that pre-theoretical intuitions, or common-sense accounts, are often used as the foundations of metaphysical theories. In later notes he was particularly preoccupied by the role of such ‘everyday’ or ‘vulgar’ understanding in metaphysical construction. Very generally, he urges that theorists of metaphysics are well advised to take account of common-sense understanding. It is not that common sense is ‘superior’ to metaphysical accounts, or can replace them. The complex accounts offered by philosophers serve a different function from those of everyday life, and can coexist with them because they are appropriate to different purposes. In his notes from around this time, he compares the vulgar and the learned with reference to what he calls ‘Eddington’s table’ and ‘the

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 178

178 Paul Grice, Philosopher and Linguist

vulgar table’. Grice’s reference here is to Arthur Eddington’s The Nature of the Physical World, first published in 1928. Eddington begins his book, originally delivered as a series of lectures, with the assertion that as he sat down to write he was confronted with not one table but two. There is the ordinary, familiar table: ‘a commonplace object of that environment which I call the world’. There is also a scientific table, an object of which he has become aware only comparatively recently. Whereas the ordinary table is substantial, the scientific table is mostly emptiness, while ‘sparsely scattered in that emptiness are numerous electric charges rushing about with great speed; but their combined bulk amounts to less than a billionth of the bulk of the table itself’.51 Eddington argues that the two different descriptions of the table are discreet and serve distinct purposes. Although they might ultimately be said to describe the same object, the scientist must keep the two descriptions separate, in effect ignoring the ‘ordinary table’ and concentrating only on the ‘scientific table’. Grice’s brief notes suggest that he is happy to accept both the vulgar and the learned description of the table. There is, he notes, ‘no conflict . . . Scientific purposes and everyday purposes are distinct’.52 Grice was advocating philosophical respect for common sense, and as such aligning himself with a philosophical tradition stretching back through Moore and Berkeley to Aristotle. Although he developed this position explicitly only towards the end of his life it is clear that, as so often in his work, he was returning to a theme, and to a philosophical outlook that had been with him for many years. It is not difficult to relate this idea to Austin’s respect for ordinary language, and the indication of everyday concepts and distinctions it contains. Indeed, notes from much earlier in Grice’s career show that, while he was still at Oxford, he was linking a philosophical interest in common sense to a respect for ordinary language. His jotted ‘Aspects of respect for vulgar’ include ‘protection against sceptic’, ‘proper treatment of Folk Wisdom’, ‘protection of speech from change and impropriety’ and, perhaps most tellingly ‘proper position treatment of ordinary ways of speech (speech phenomena)’.53 The same form of respect can also be detected, although more problematically, in his later work on value. Subjectivists such as Mackie had suggested that they were themselves necessarily arguing against common sense, rejecting the idea apparent in ordinary language that values ‘exist’ as objective topics of knowledge. Grice himself rather irreverently sums up a position such as Mackie’s in his notes from this period

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 179

Metaphysics and Value 179

as follows: ‘value-predicates (e.g. “good”, “ought”) signify attributes which ordinary chaps ascribe to things. It is however demonstrable, on general grounds, that these attributes are vacuous. So position attribution by the vulgar are systematically false’.54 According to this account, in upholding objective value, Grice would be maintaining a straight forwardly common-sense position. In fact Grice himself does not accept this simple identification of objectivism with common sense. He argues in the Carus lectures that ‘it seems to me by no means as easy as Mackie seems to think to establish that the “vulgar valuer”, in his valuations, is committed to the objectivity of value(s).’55 In the taped conversation he suggests that the need to work on the status of common sense, intuition and pre-theoretical judgement is particularly acute for anyone who wants to take a constructivist approach to a topic such as value. It is not that the philosopher should simply take on board all ‘vulgar’ pronouncements on such a topic, but rather that a constructivist should attempt at least to use them as a starting point for any theory of value. Grice’s ‘common sense’ philosophy is, therefore, a more sophisticated approach than one relying simply on the understanding apparent in everyday language. Years earlier he defended common sense but attacked Moore’s simplistic application of it in ‘G. E. Moore and philosophers’ paradoxes’. In the later work, he argues that although common sense may be the starting point, exactly how a theory is to be constructed from there is a matter for theory-theory: ‘Moore-like proclamations cut no ice.’ Grice was also turning his attention to another loose end in his work: the study of perception. In particular, he returned to the area in which he had worked with Geoffrey Warnock 30 or more years before, even planning a ‘Grice/Warnock retrospective’. He sketched out a number of topics for this, but it does not appear to have progressed beyond note form. The first topic was to be ‘The place of perception as a faculty or capacity in a sequence of living things’. Thinking about the old issue of perception in relation to his more recent interest in creatureconstruction, he wondered: ‘at what point, if any, is further progress up the psychological ladder impossible unless some rung has previously been assigned to creatures capable of perception?’56 He dwells on the advantages of perception in terms of survival, the crucial factor for adding any capacity during creature-construction, and assesses the possible support this might offer for common sense against philosophers of sense data. If perception is to be seen as an advantage, providing knowledge to aid survival in a particular world, ‘the objects revealed by

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 180

180 Paul Grice, Philosopher and Linguist

perception should surely be constituents of that world’. It might be possible to say that sense data do not themselves nourish or threaten, but constitute evidence of things that do. However, Grice notes that he is more tempted by an alternative explanation. ‘Flows of impressions’ do not so much offer evidence of what is in the world, as ‘prompt, stimulate or excite certain forms of response to the possibility that such states of affairs do obtain’. These responses may be either verbal or behavioural. Both types of response suggest commitment on the part of the responding creature to the state of affairs in question. Here again, Grice seems to be offering a defence of a common sense approach to philosophical questions, but a more sophisticated one than that of simple realism. The existence of the material object is indicated not by the sense impressions themselves, but by the responses these elicit in the perceiving creature. From the middle of the 1980s onwards, Grice was increasingly preoccupied with various publishing projects, mainly with book plans. His notes from the time suggest an explosion of energy in this area. A list from 1986 includes ‘ “Method: the vulgar and the learned” with ?OUP. “From Genesis to Revelations” ?HUP’. Other notes are uncompromisingly labelled ‘work program’, and show him carefully dividing up time by year, by month or even by day, including the injunction that he should set aside part of December 1986 to ‘CATCH UP on prior topics’.57 For most of Grice’s potential audience, however, it was the William James lectures that were most eagerly awaited. With the exception of the few lectures already published, these remained as ageing manuscripts in Grice’s idiosyncratic filing system. Even with his new enthusiasm for publishing, it took some persuasion from Harvard University Press, and from Kathleen, to revisit these. It is clear that he was not averse in principle to publishing the lectures; his flurry of lists from the early 1980s testify to this. But he had set himself an exacting programme of revising and adding to the them. Grice was adamant from the start that, whatever else the book contained, the William James lectures should be kept together and distinct from other papers. In a letter to his publisher he wrote, ‘the link between the two main topics of the James lectures is not loose but extremely intimate. No treatment of Saying and Implying can afford to omit a study of the notion of Meaning which plainly underlies both these ideas.’58 Before the book had a name, he was planning that it should contain a Part A entitled ‘Logic and conversation’. It seems that this was to include a number of postscripts to the lectures. Part B was provisionally entitled ‘Other essays on language and related matters’, and

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 181

Metaphysics and Value 181

was to contain a number of further postscripts and reassessments. Of all the proposed postscripts, only ‘Conceptual analysis and the province of philosophy’, concerned mainly with revisiting the themes of ‘Postwar Oxford philosophy’, was ever written. The inclusion of ‘In defence of a dogma’ in Part B was a matter of some debate, Grice remaining resolutely in favour. When a reader for the publisher questioned this, Grice responded with a robust defence: ‘the analytic/synthetic distinction is a crucially important topic about which I have said little in the past thirty years, and must sometime soon treat at some length.’ In his original plan the book was also to include a transcription of a talk on this same topic between himself, George Myro, Judith Baker, George Bealer and Neil Thomason. Indeed, Grice suggested initially that Studies in the Way of Words should be published as authored by ‘Paul Grice with others’. Grice revisits the analytic/synthetic distinction in the ‘Retrospective epilogue’. He introduces a number of inhabitants of ‘philosophical Never-Never-Land’, M*, A*, G* and R*, the fairy godmothers of Moore, Austin, Grice and Ryle. These fairy godmothers are useful to philosophical discussion because they hold explicitly all the views that their godchildren hold, without regard for whether these views are explicit or implicit in their actual work. R* suggests that the division of statements of human knowledge into ‘analytic’ and ‘synthetic’ categories, far from being an artificial and unnecessary complication by philosophers, is in fact an insightful attempt by theorists to individuate and categorise the mass of human knowledge. This view offers a particular challenge to Quine’s attack on the distinction, with which Grice had taken issue in his joint article with Strawson over 30 years earlier. It is in tacit reference to this that Grice concludes his epilogue: Such consideration as these are said to lie behind reports that yet a fifth fairy godmother, Q*, was last seen rushing headlong out of the gates of Never-Never-Land, loudly screaming and hotly pursued (in strict order of seniority) by M*, R*, A* and G*. But the narration of these stirring events must be left to another and longer day.59 It is tempting to read a note of wistful resignation into the manner in which Grice chose in 1987 to conclude Studies in the Way of Words. However, letters he was writing at the time, which affirm his intention to revisit these topics at length, argue that he at least in part believed he had time to do so. Moreover, he submitted applications for Faculty Research Grant support for both the academic years 1986–7 and 1987–8,

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 182

182 Paul Grice, Philosopher and Linguist

for a project on ‘Metaphysical foundations of value and the nature of metaphysics’. Kathleen recalls that he was never successful in his bids for such research grants. Despite this, he does not seem to have made much effort to accommodate to the demands of the application process. In answer to a question about the significance of his research project and the justification of the proposed expenditure, he wrote in 1986 ‘I hope the character of the enterprise will be sufficient evidence of its significance.’ In 1987 he offered an even more terse account of his plans: ‘their importance seems to me to be beyond question’.60 He was applying for funding to cover the expenses of secretarial support in transcribing tapes on to which he had been dictating his notes and typing up manuscripts, and for duplication, supplies and mailing. His intention was to produce two books for Oxford University Press, one a revised version of the Carus lectures and one on ‘Methodology in Metaphysical Theory’, a study to include a consideration of ‘the nature and degree of respect due from metaphysical theorists towards the intuitions and concerns of the common man’. The plan was to complete at least first drafts during 1987–8. Whatever his official predictions, Grice was aware that his condition was worsening and that the prognosis was not good. In the summer of 1987 he underwent eye surgery in an attempt to rescue his failing sight, and by the end of the year he needed a hearing aid. In addition, his mobility became further restricted. The entrance to his house, and the balcony on which he loved to sit, were on the first floor over the garage; a further flight of stairs led to the second floor. By early in 1988, a stair lift was needed. Even so, Grice was effectively housebound, and received visits from publishers, colleagues and even students on his balcony. From there he also continued to work on new ideas, apparently with a sense of urgency that was only increased by his condition. He jotted on the back of an envelope, ‘Now that my tottering feet are already engulfed in the rising mist of antiquity, I must make all speed ere it reaches, and (even) enters, my head.’61 He dealt with the business of seeing Studies in the Way of Words to completion. Lindsay Waters from Harvard University Press, who worked with Grice throughout the preparation of the book, has commented on the determination with which he did this, suggesting that ‘He knew the fight with the manuscript editor was his last.’62 Paul Grice died on 28 August 1988, at the age of seventy-five. He had, perhaps, as many projects on the go as ever. Studies in the Way of Words was on the way to press; Grice had gone over the final changes with the manuscript editor just days before. His ‘work in progress’ was still on

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 183

Metaphysics and Value 183

his desk, and included notes on ‘the vulgar and the learned’ that indicate that he was turning his thoughts back to the topic of perception in relation to this idea. The causal theory of perception could, the notes suggest, be seen as a natural way of reconciling the vulgar and the scientific. Perceptions could be connected either with scientific features or with vulgar descriptions, as they were in his article ‘The causal theory of perception’. The causal theory explains the way things appear on the basis of the way they are. It rules out the account from sense data theory, which explains the way things are on the basis of the ways things appear. Scientific theory, Grice had noted ‘cuts us off from objects’.63 The proofs of ‘Aristotle on the multiplicity of being’, an article he had at one point considered including in Studies in the Way of Words, were posted to him on August 17 and remained untouched. Studies in the Way of Words, containing the long-awaited complete William James lectures and Grice’s other most influential papers in the philosophy of language, was published by Harvard University Press in 1989. A number of reviewers took the opportunity, before engaging with the details of Grice’s philosophy, to pay tribute to him. Peter Strawson described his ‘substantial and enduring contribution to philosophical and linguistic theory’.64 Tyler Burge praised his ‘wonderfully powerful, subtle, and philosophically profound mind’.65 Charles Travis argued that, ‘H. P. Grice transformed, often deeply, the problems he touched and, thereby, the terms of philosophical discussion.’66 Robert Fogelin suggested that ‘even if Grice was one of the chief and most gifted practitioners of OLP, he was also, perhaps, its most penetrating critic.’67 As Fogelin notes, ‘Meaning revisited’, with its enigmatic reference to ‘The Mystery Package’ cannot be fully interpreted without reference to Grice’s work on values and teleology. When Grice died, the Carus lectures were still in manuscript form, and he had asked Judith Baker to see them through publication with Oxford University Press. She prepared the lectures for publication, together with part of the ‘Reply to Richards’ and ‘Method in philosophical psychology’. The Conception of Value appeared in 1991. Reviewers were generally ambivalent, welcoming the contribution to debate, but expressing some reservations as to the details of Grice’s position. For instance, Gerald Barnes describes Grice’s defence of absolute value as ‘unconventional, richly suggestive, often fascinating, and, by Grice’s own admission, incomplete at a number of crucial points’.68 The John Locke lectures had to wait a further decade before they, too, appeared from Oxford University Press, edited and introduced by Richard Warner under the title Aspects of Reason. Again, reviewers had reservations. A. P. Martinich declared

1403_902976_09_cha08.qxd 12/13/04 6:29 PM Page 184

184 Paul Grice, Philosopher and Linguist

himself unconvinced by the appeal to the ordinary use of expressions such as ‘reason’, and criticised Grice’s style of presentation for containing ‘too many promissory and subjunctive gestures’.69 The body of published work, and the legacy of his teaching and collaborations, formed the basis of Peter Strawson and David Wiggin’s assessment of Grice for the British Academy: Other anglophone philosophers born, like him, in the twentieth century may well have had greater influence and done a larger quantity of work of enduring significance. Few have had the same gift for hitting on ideas that have in their intended uses the appearance of inevitability. None has been cleverer, or shown more ingenuity and persistence in the further development of such ideas.70 Perhaps an even more succinct summary of Grice’s philosophical life is the one implied by a comment he himself made as it became clear to him that he would not recover his failing health. Despite the indignities of his illness and treatment, he confided in Kathleen that he did not really mind about dying. He had thoroughly enjoyed his life, and had done everything that he wanted to do. Grice’s career was characterised by years of self-doubt, by almost ceaseless battles with the minutiae of philosophical problems, and by a final frenzied struggle to finish an ever-multiplying list of projects. Yet it seems that, when he reviewed the effort in comparison with the results, he was content with the deal.

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 185

9 Gricean Pragmatics

By all the obvious measures of academic achievement, the big success story for Grice has been the response in linguistics to his theory of conversation. The ‘Logic and conversation’ series, particularly the individual lecture published under that title, has been cited widely and frequently from the first. Beyond frequency of reference, perhaps the greatest accolade is for a writer’s name to be coined as an adjective, a process with few precedents in linguistics. There is ‘Saussurian linguistics’, the ‘Chomskyan revolution’ and ‘Hallidayan grammar’. From the mid-1970s on, linguists could be confident that a set of assumptions about theoretical commitments could be conveyed by the phrase ‘Gricean pragmatics’. The term has been used as a label for a disparate body of work. It includes relatively uncritical applications of Grice’s ideas to a wide range of different genres, as well as attempts to identify flaws, omissions or full-blown errors in his theory. It also includes a variety of attempts to develop new theories of language use, based to varying degrees on Grice’s insights. These different types of work do not all agree on the efficacy of the Cooperative Principle, on the exact characterisation of the maxims, or even on the nature of conversational implicature itself. What they have in common, qualifying them for the title ‘Gricean’, is an impulse to distinguish between literal meaning and speaker meaning, and to identify certain general principles that mediate between the two. The latter is perhaps Grice’s single greatest intellectual legacy. Stephen Neale has suggested that ‘in the light of his work any theory of meaning that is to be taken at all seriously must now draw a sharp line between genuinely semantic facts and facts pertaining to the nature of human interaction.’1 Grice’s understanding that linguistic meaning and meaning in use may differ was shared with a variety of previous philosophers and indeed 185

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 186

186 Paul Grice, Philosopher and Linguist

with common sense. But he went much further in his claim that the differences were amenable to systematic, independently motivated explanation, and in his ambitious proposal to provide such an explanation. Whatever its flaws, imprecisions and even errors, the theory of conversation has provided a definite starting point for those interested in discussing the relevant layers of meaning. Stephen Levinson has gone so far as to suggest that ‘Grice has provided little more than a sketch of the large area and the numerous separate issues that might be illuminated by a fully worked out theory of conversational implicature.’2 The success of the theory of conversation was achieved despite its publication history. Demand was such that copies of the William James lectures were in fairly wide circulation around American and British universities soon after 1967 in the blurred blue type of a mimeograph, and many early responses to them refer directly to this unpublished version. Some commentators did not have even this degree of access to the lectures. In an article published in 1971, Jonathan Cohen comments that he is responding to ‘the oral publicity of the William James Lectures’.3 When the lectures began to appear in print, they were scattered and piecemeal. ‘Utterer’s meaning, sentence meaning, and word-meaning’ appeared in Foundations of Language in 1968 and ‘Utterer’s meaning and intentions’ in The Philosophical Review in 1969. ‘Logic and conversation’, the second in the original lecture series, appeared in two separate collections, both published in 1975. Its publication was clearly the source of some confusion, with both sets of editors claiming to offer the essay in print for the first time. The Logic of Grammar, edited by Donald Davidson and Gilbert Harman, seems to have been Grice’s preferred publication, and was the one he always cited. There, his lecture appears alongside essays by past and contemporary philosophers such as Frege, Tarski and Quine. The editor’s introduction places Grice in a tradition of philosophers interested in the application of traditional theories of truth to natural language. In Speech Acts, the third volume in the series Syntax and Semantics, on the other hand, ‘Logic and conversation’ is placed in the context of contemporary linguistics. Accompanying articles by, for instance, Georgia Green and Susan Schmerling discuss conversational implicature itself. The editors, Peter Cole and Jerry Morgan, hail the first publication of ‘the seminal work on this topic’.4 It is rumoured that Grice signed the contract for his lecture to appear in this linguistics series while at a conference at Austin, Texas, and only after a long evening spent in the bar. Peter Cole included ‘Further notes on logic and conversation’ in Volume 9 of the same series in 1978.

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 187

Gricean Pragmatics 187

One catalyst for the emergence of pragmatics, and Gricean pragmatics in particular, was the interest in Grice’s work from within generative semantics, the movement fronted by the young linguists working just across campus from Grice in the years immediately following his move to Berkeley. Attempting to integrate meaning, including contextual meaning, into their study of formal linguistics, these refugees from transformational orthodoxy discovered what they saw as a promising philosophical precedent in ‘Logic and conversation’. Asa Kasher has suggested that Grice opened a ‘gap in the hedge’, allowing linguists to see the possibilities of ‘extra-linguistic’ explanations.5 Gilles Fauconnier has found a more startling metaphor; at a time when linguistics was largely concerned with syntax, and with meaning only in so much as it related to deep grammatical structure, ‘Grice unwittingly opened Pandora’s box’.6 When linguists working within transformational grammar replied to the generative semanticists’ accusation that they were artificially ignoring meaning and context, they too drew on ‘Logic and conversation’. In effect, they cited Grice’s characterisation of ‘what is said’ and ‘what is implicated’ as a licence to draw a sharp distinction between the linguistic and the extra-linguistic. In this way, they argued that it was legitimate and indeed necessary for a linguistic theory such as transformational grammar to be concerned with meaning only in terms that were insensitive to context. Features of context were to be dealt with in a separate, complementary theory, perhaps even by Grice’s theory of conversation itself. Writing in the mid-1970s, transformational grammarians Jerrold Katz and Thomas Bever accuse the generative semanticists of a form of linguistic empiricism more extreme than Bloomfield. By attempting to fit every aspect of interpretation, however specific to context, into linguistic theory, the generative semanticists were failing to distinguish between grammatical and nongrammatical regularities: failing to respect the distinction between competence and performance. Katz and Bever commend Grice’s account of conversational implicature by way of contrast, as an example of a type of meaning properly and systematically handled outside the grammar by independently motivated cognitive principles.7 In the same volume, Terence Langendoen and Thomas Bever praise Grice for correctly distinguishing between grammatical acceptability and behavioural comprehensibility.8 Even Chomsky himself put aside his worries about behaviourism to draw on Grice’s work. In a 1969 conference paper concerned with defending his version of grammar against criticisms from generative semanticists he suggests that if certain aspects of meaning ‘can be explained in terms of general “maxims of discourse” (in the

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 188

188 Paul Grice, Philosopher and Linguist

Gricean sense), they need not be made explicit in the grammar of a particular language’.9 In the 1970s, then, the theory of conversation was seen by some as a potential means of incorporating contextual factors into grammatical theory, and by others as a reason for refusing to do so. Generative semantics eventually failed, perhaps a victim of its own ambition, and of the increasingly complex syntactic explanations needed to account for contextual meaning. Former generative semanticist Robin Lakoff has attributed its eventual decline to a growing realisation that syntax was not in fact the place where meaning could best be explained; they came to understand that ‘to deal with the phenomena we had uncovered, the relationships between form and function that our work had made manifest and unavoidable, we needed to shift the emphasis of the grammar from syntax to semantics and pragmatics.’10 Generative semantics had succeeded in putting the discussion of contextual meaning on the formal linguistic map, and had paved the way for the emergence of pragmatics as a separate discipline. The sheer size of Grice’s impact on this discipline makes anything approaching a comprehensive survey prohibitive, but a few key collections can provide a representative overview. By the end of the decade, pragmatics had its own international conference. In the proceedings of the 1979 conference held at Urbino Grice is cited, or at least alluded to, in 15 of the 37 papers.11 These papers range over topics such as language acquisition, second language comprehension, rhetoric and intonation, as well as theories of meaning, reference and relevance. In 1975, Berkeley Linguistics Society, a student-run organisation, began hosting an annual conference, bringing together linguists from across America. At the first meeting, Grice’s work formed the basis of two very different papers. Cathy Cogan and Leora Herrmann tackled colloquial uses of the phrase ‘let’s just say’, arguing that it ‘serves as an overt marker on sentences which constitute an opting out’ from straightforward observance of the maxims.12 At the same conference, Lauri Karttunen and Stanley Peters presented a much more formal paper on conventional implicata, arguing that they could be used to explain various so-called presuppositional phenomena.13 Since then, meetings of the society have included an ever-increasing number of papers on Gricean topics. In 1990 a parasession at the annual meeting was devoted to papers on ‘The Legacy of Grice’. In effect, the delegates all interpret this title in relation to the theory of conversation, which they applied to topics as diverse as the language of jokes, the grammaticalisation of the English perfect between Old English and Modern English, the teach-

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 189

Gricean Pragmatics 189

ing of literacy, and the forms of address adopted by mothers and fathers towards their children.14 By the time the The Concise Encyclopedia of Pragmatics was published in 1998, Grice’s work had earned mention in entries on 29 separate topics, including advertising, cognitive science, humour, metaphor and narrative. Again, the majority of these references are to ‘Logic and conversation’, and in her entry on ‘conversational maxims’ Jenny Thomas comments that ‘it is this work – sketchy, in many ways problematical, and frequently misunderstood – which has proved to be one of the most influential theories in the development of pragmatics.’15 The Journal of Pragmatics, which began in the mid1970s, has included numerous articles with Gricean themes, and in 2002 published a special focus issue entitled ‘To Grice or not to Grice’. In his introductory editorial, Jacob Mey argues that the range of articles included, both supportive of and challenging to Gricean implicature, form ‘a living testimony to, and fruitful elaboration of, some of the ideas that shook the world of language studies in the past century, and continue to move and inspire today’s research’.16 Nevertheless, Kenneth Lindblom suggests that the importance and implications of the theory of conversation tend to be underestimated, because it has been used by scholars in a range of fields with significant methodological differences.17 Certainly, Gricean analysis has appeared in discussions of a wide variety of linguistic phenomena, taking it well beyond the confines of its ‘home’ discipline of pragmatics. For instance, the relationship between Grice’s theory and literary texts has long been a source of interest to stylisticians. As early as 1976, Teun van Dijk suggested that the maxims ‘partially concern the structure of the utterance itself, and might therefore be called “stylistic” ’.18 He argues that a set of principles different from, but analogous to Grice’s Cooperative Principle and maxims are needed to explain literature. For Mary Louise Pratt, however, the most successful theory of discourse would be one that could explain literature alongside other uses of language. In the case of literature, where the Cooperative Principle is particularly secure between author and reader, ‘we can freely and joyfully jeopardize it’.19 Flouting and conversational implicatures are therefore particularly characteristic of literary texts but do not set them apart as intrinsically different from other types of language use. Other commentators have concentrated on Gricean analyses of discourse within literary texts, such as soliloquies in Hamlet, Macbeth and Othello and dialogue in Waiting for Godot and Finnegans Wake.20 Conversational implicature has also proved suggestive in historical linguistics as an explanation for a variety of meaning shifts. The basic

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 190

190 Paul Grice, Philosopher and Linguist

premiss of this research is that meaning that is initially associated with a word or phrase as an implicature can, over time, become part of its semantic meaning. Peter Cole describes this process as ‘the lexicalization of conversational meaning’.21 Elizabeth Traugott’s work in this area has drawn on a wide range of historical data. For instance, in a 1989 paper she traces a general trend in the history of English for changes from deontic to epistemic meanings; both words and grammatical forms that started out as weakly subjective become strongly subjective. For instance, ‘you must go’ changes its meaning from permission in Old English to expectation in Present Day English. This process she describes as ‘the conventionalising of conversational implicature’ based on ‘pragmatic strengthening’ and derived ultimately from Grice’s second maxim of Quantity.22 Debra Ziegeler has followed up this study by suggesting that another process of conventionalisation can account for the hypothetical and counterfactual meanings associated with some modal verbs in the past tense, such as would, this time a process explained by the first maxim of Quality.23 It would be wrong to suggest that Grice’s presence in linguistics has been greeted with constant, or universal, enthusiasm. The Concise Encyclopedia of Pragmatics is representative here, too. As well as the albeit modified praise from Thomas, the encyclopaedia includes an entry on ‘Discourse’ in which Alec McHoul accuses Grice of idealising conversation. According to McHoul, Grice ‘attempted to generalise from how he thought conversation operated’, and ended up giving ‘an ideal-rational version of socially pragmatic events so that it seems virtually irrelevant that the objects of analysis . . . constitute a form of social action’.24 McHoul is far from being a lone voice. The theory of conversation has frequently been accused of presenting an unrealistic picture of human interaction, and even of attempting to lay down rules of appropriate conversational etiquette. For instance, Dell Hymes suggests that Grice’s account of conversational behaviour incorporates ‘a tacit ideal image of discourse’,25 and Sandra Harris identifies in it ‘prescriptive and moral overtones’.26 For Rob Pope, an individual’s willingness to accept the maxims as an accurate account of communication is dependent on ‘the problem of how far you and I, in our respective world-historical positions, find it expedient or desirable to espouse a view of human interaction based on consensus and/or conflict’.27 Geoffrey Sampson even claims that ‘it would be ludicrous to suggest that as a general principle people’s speech is governed by maxims’ such as those proposed in the theory of conversation.28 However, he rather undermines his own argu-

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 191

Gricean Pragmatics 191

ment by presenting the observation that ‘people often manifestly flout his maxims’ as if it were a problem for Grice’s theory. A number of linguists have been quick to defend Grice from the suggestion that his maxims prescribe ‘ideal’ conversation. For instance, Penelope Brown and Stephen Levinson explain that the maxims ‘define for us the basic set of assumptions underlying every talk exchange. But this does not imply that utterances in general, or even reasonably frequently, must meet these conditions, as critics of Grice have sometime thought’.29 Robin Lakoff argues that strict adherence to the maxims is not usual and that in many cases: ‘an utterance that fails to incorporate implicature when it is culturally expected might be uncooperative and so liable to misunderstanding, hardly “ideal”.’30 There are a number of features of the presentation of the theory, particularly read in isolation from the rest of the lectures, that may contribute to the accusation of idealising conversation. First, Grice never defines what he means by ‘conversation’. It is clear from his earlier, unpublished writings that he adopted this term simply as a label for language in context, or as a reminder to his Oxford students that utterances do not generally occur in isolation from each other. However, in linguistics the term has acquired a much more specific and technical use. During the early 1970s, just as Grice’s theory was becoming widely known, work by linguists such as Harvey Sacks was revealing the characteristics and regularities of naturally occurring casual conversation, dwelling on features such as turn taking and topic organisation.31 Linguists familiar with this work have perhaps been taken aback by Grice’s apparently cavalier, unempirical generalisations. Second, Grice introduces the theory of conversation as an approach to specific philosophical problems, but his own enthusiasm for the particular conversational explanations to which it extends obscures this purpose. In the single lecture ‘Logic and conversation’ in particular, casual conversational examples appear to be the main focus. Further, Grice expresses both the Cooperative Principle and the individual maxims as imperatives. Despite his intention of generalising over the assumptions that participants bring to interactive behaviour, this grammatical choice gives his ‘rules’ a prescriptive appearance. Perhaps most damagingly of all, Grice does not comment on his own methodology and purpose in ‘Logic and conversation’. He does, however, give some indication of these in the earlier, unpublished lectures on language and context, for instance when he defends his method of working with as few ‘cards on the table’ as possible, in the

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 192

192 Paul Grice, Philosopher and Linguist

knowledge that this will result in certain deliberate simplifications. In another lecture, proposing to address some questions put to him in a previous discussion about the nature of his undertaking, he suggests how he would like his ideas to be evaluated. His theory building is different from his methodology in ‘Meaning’, which proceeded through the description of specific cases. I am here considering what is (or may be) only an ideal case, one which is artificially simplified by abstracting from all considerations other than those involved in the pursuance of a certain sort of conversational cooperation. I do not claim that there actually occur any conversations of this artificially simplified kind; it might even be that these could not be (cf frictionless pulleys). My question is on the (artificial) assumption that their only concern is cooperation in the business of giving and receiving information, what subordinate maxims or rules would it be reasonable or even mandatory that participants should accept as governing their conversational practice? Since the object of this exercise is to provide a bit of theory which will explain, for a certain family of cases, why it is that a particular implicature is present, I would suggest that the final test of the adequacy and utility of this model should be a) can it be used to construct explanations of the presence of such implicatures, and is it more comprehensive and more economical than any rival [I have tried to make a start on the first part of a)] b) Are the no doubt crude, pretheoretical explanations which one would be prompted to give of such implicatures consistent with, or better still favourable pointers towards, the requirements involved in the model?32 Grice’s idealisations are of the scientific type, necessary to establish underlying principles of the subject matter, rather than of the type specifying that it always does, or should, conform to certain optimal standards. Another criticism of Grice’s theory that seems off-beam in the light of his unpublished justification, but registers a legitimate concern for linguists, is that it works for only one particular discourse type: roughly for conversations between social equals who are mutually interested in exchanging information on a particular topic. Some of the earliest commentaries on the theory of conversation to consider the social and institutional contexts of language were, perhaps surprisingly, from philosophers rather than from linguists. In a 1979 edition of The Philo-

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 193

Gricean Pragmatics 193

sophical Quarterly, David Holdcroft and Graham Bird contribute to a discussion of whether the Cooperative Principle is best seen as applying to all sequences of speech acts, or simply to informal conversations. Holdcroft argues that the Cooperative Principle is not unproblematically universal across all sequences; in cases where participants do not ‘have equal discourse rights’ the maxims do not straightforwardly apply.33 He offers the examples of the proceedings of a court of law and a viva exam, suggesting that different sets of maxims are accepted by participants in different cases. Bird argues that situations in which participants have unequal rights do not pose the problems to Grice that Holdcroft envisages; maxims may be suspended by mutual agreement without offering counter-examples to the theory.34 These ideas were later addressed by linguists using actual conversational data rather than theorised ‘types’ of sequences. Norman Fairclough pioneered critical discourse analysis, an approach concerned with relating the formal properties of texts to interpretation and hence to their relationship to social context. He suggests that mainstream discourse analysis ‘has virtually elevated cooperative conversation between equals into an archetype of verbal interaction in general’.35 He holds Grice at least in part responsible for this because his emphasis on the exchange of information assumes equal rights to contribute for both parties. Fairclough analyses conversational extracts from police interviews of a woman making a complaint of rape and of a youth suspected of vandalism to argue that many interactions do not conform to this model. He notes that Grice’s own proviso that in some types of conversation influencing others may be more important has too often been overlooked. Other critical discourse analysts offer qualified support for Grice, for instance arguing that the theory of conversation need not be limited to casual conversation, but can be generalised to explain discourse in institutional contexts only once it has been made to take account of societal factors.36 John Gumperz uses a variety of data to argue that interactions such as cross-examinations, political debates and job interviews demonstrate that ‘conversational cooperation becomes considerably more complex than would appear on a surface reading of the Cooperative Principle.’37 William Hanks echoes many of the criticisms of the theory of conversation familiar to ethnolinguists, but argues that it should not be abandoned altogether within this discipline because ‘some parts of Grice’s approach are of telling interest to ethnography’, in particular the active role assigned to the hearer in attributing conversational meaning. For Grice, as for sociolinguists, ‘meaning arises out of the relation between parties to talk’.38

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 194

194 Paul Grice, Philosopher and Linguist

If it is to some extent contrary to Grice’s abstract, philosophical intentions that his theory should be applied to real language data, it is perhaps even more incongruous that it should be subjected to clinical testing. Nevertheless, the theory of conversation has generated much interest in the area of clinical linguistics, in particular the branch sometimes described as ‘neuropragmatics’. Chomsky was interested in the theory of conversation because it suggested to him a principled way of distinguishing between centrally linguistic meaning that might be described as belonging to a speaker’s competence or semantic ability, and context-dependent meaning belonging to performance or pragmatics. Clinical linguists have looked for evidence of a neurological basis for such a distinction. Many such studies have concentrated on patients with some degree of language impairment, in an attempt to discover whether semantics and pragmatics can be selectively impaired. For instance, Neil Smith and Ianthi-Maria Tsimpli, working within a broadly Chomskyan framework, have studied an institutionalised braindamaged patient with exceptional linguistic abilities. Their findings suggest that his ‘linguistic decoding’ is as good as anyone else’s but that he is unable to handle examples such as irony, metaphor and jokes, a distinction that they suggests points to a separation between linguistic and non-linguistic impairment.39 Such a suggestion seems to be supported by findings from researchers with a more clinical and less linguistic interest, who have suggested that core linguistic functions might be controlled by the left hemisphere of the brain while ‘pragmatic’, context-dependent meaning, together with emotions, is controlled by the right.40 Other clinicians have adopted a more overtly Gricean framework. Elisabeth Ahlsén uses the Cooperative Principle as a tool for describing aphasic communication in which, she argues, ‘implicatures based on the maxims of quantity and manner cannot be made in the same way as in the case of nonaphasic speakers.’41 The theory of conversation has also been used as a tool for describing more general communication disorders.42 Ronald Bloom and a group of fellow-experimenters argue that ‘Gricean pragmatics has provided a theoretically meaningful framework for examining the discourse of individuals with communication disorders’.43 Clinicians using Gricean analyses tend to be reluctant to draw definite conclusions about the division of labour between the hemispheres. This is argued perhaps most explicitly by Asa Kasher and his team, who have investigated the processing of implicatures by using an ‘implicature battery’ for both right brain-damaged and left braindamaged stroke patients. They use both verbal implicatures and non-

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 195

Gricean Pragmatics 195

verbal implicatures based on visual images, and conclude that ‘both cerebral hemispheres appear to contribute to the same degree to processing implicatures.’44 This scepticism about the localisation of pragmatic processing in the right hemisphere is shared by Joan Borod and others, who use a series of pragmatic features, including some based on Gricean implicature, to test language performance in right-hand braindamaged, left-hand brain-damaged and normal control subjects. They conclude that ‘deficits in pragmatic appropriateness are shared equally by both brain-damaged groups.’45 Elsewhere, Kasher has argued explicitly against the tendency to make sweeping generalisations over pragmatic phenomena that locate them all in the right hemisphere. Implicature, he argues, is part of a larger competence, ‘hence it would be implausible to assume that some domain-specific cognitive system produces conversational implicature’.46 As Kasher’s position illustrates, the positing of a separate pragmatic faculty, to which implicature belongs, is not dependent on the faculty being physically autonomous. It can be impaired selectively from language, an observation that in itself legitimises the hypothesis of a physical basis in the brain. Another field in which the theory of conversation has been used enthusiastically is Artificial Intelligence, a striking application given Grice’s personal animosity to computers. Computer programmers have realised that formal linguistic theories can go a long way towards specifying the rules that make up grammatical sentences, but have little to say about how those sentences can be used to form natural-sounding conversations. For instance, Ayse Pinar Saygina and Ilyas Ciceki are not impressed by the pragmatic skills of computers. They analyse the output of various programmes entered for the Loebner Prize, an annual competition to find the programme closest to passing the Turing Test, in which a computer is deemed to be intelligent if a human observer cannot distinguish its participation in a conversation from that of a human. Saygina and Ciceki argue that pragmatic acceptability, rather than the simple ability to generate natural language, is the biggest challenge to such programmes. Moreover, the programmes that are the most successful are those that adhere most closely to Grice’s maxims, suggesting future applications for these, specifically that programmers ‘will inevitably have to find ways of “making computers cooperate” ’.47 Various programmers have in fact made some attempts in this direction. Robert Dale and Ehud Reiter consider an algorithm for naturalsounding referring expressions in computer–human interaction. They note that such an algorithm must produce expressions that successfully pick out the referent without leading ‘the human hearer or reader to

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 196

196 Paul Grice, Philosopher and Linguist

make false conversational implicatures’.48 Perhaps not surprisingly, they consider Grice’s maxims to be ‘vague’ compared to the principles needed for computational implementation. Nevertheless, they base their algorithm on an interpretation of the maxims, claiming that the simpler interpretation produces a result that is ‘faster in computational terms and seems to be closer to what human speakers do when they construct referring expressions’.49 Michael Young tackles the problem of intelligent systems designed to form plans to direct their own activities or those of others. ‘There is a mismatch between the amount of detail in a plan for even a simple task and the amount of detail in typical plan descriptions used and understood by people’.50 In order to address this issue, Young proposes a computational model that draws on Grice’s maxim of Quantity, which he labels the Cooperative Plan Identification architecture. He argues that this produces instructions that are much more successful in communicating a plan to a human subject than instructions produced by alternative methods. Grice would have recognised discussion of irony and metaphor as more closely related to his work, although he would not necessarily have welcomed the form some of this debate has taken. The explanation of such ‘non-literal’ uses has generally been taken as a central task for the theory of conversation, and indeed Grice himself seems to encourage this view. In a rather hurried exegesis in the William James lectures, he explains irony and metaphor as floutings of the first maxim of Quality. ‘X is a fine friend’ said just after X has been known to betray the speaker cannot be intended to be taken as a true expression of belief. Instead, the speaker must intend some other related proposition; ‘the most obviously related proposition is the contradictory of the one he purports to be putting forward.’51 Grice later restricts his definition of irony, claiming that it is ‘intimately connected with the expression of a feeling, attitude or evaluation’.52 As an example of metaphor, ‘You are the cream in my coffee’ is also clearly false, but cannot be interpreted as meaning its simple opposite, which is a truism; ‘the most likely supposition is that the speaker is attributing to his audience some feature or features in respect of which the audience resembles (more or less fancifully) the mentioned substance.’53 Some linguists working on such figures of speech have accepted an account along the lines suggested by Grice.54 Others, however, have taken issue with the apparent ease with which Grice’s hearer in each case derives the interpretation. Their argument is that nothing within the Cooperative Principle or the maxims as they stand specifies why the hearer should reach that interpretation rather than some other. Robert

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 197

Gricean Pragmatics 197

Harnish wonders what tells us that the hearer in the irony example reaches the conclusion that ‘X is a cad’; ‘why should we suppose that the contradictory is the most obviously related proposition?’ Similarly, if Grice’s metaphor example is to be explained, ‘we need some principles to guide our search for the correct supposition.’55 Dan Sperber and Deirdre Wilson argue that Grice’s description of the process by which irony is interpreted ‘is virtually free of rational constraints’.56 There is no reason why the negation of what is said should serve as the implicature, rather than some other closely related assumption. Wayne Davis worries about how a hearer is expected to identify accurately exactly what sort of interpretation is intended when, for instance, the maxim of Quality is flouted.57 Some claim that experimental work also poses problems for Grice’s account. Rachel Giora argues that the ‘salient’ meanings of metaphors and irony ‘enjoy a privileged status: they are always accessed, and always initially, regardless of context’.58 Raymond Gibbs goes so far as to claim that ‘the results of many psycholinguistic experiments have shown the traditional, Gricean view to be incorrect as a psychological theory.’59 His argument is that a figurative example does not require more processing time than a literal one; speakers often understand metaphorical or ironic meaning straight away without first having to analyse and reject literal meaning. Gibbs sees this as a threat to the whole distinction between the said and the implicated. Not all experimental linguists would subscribe to Gibbs’s conclusions. Thomas Holtgraves points out that many metaphorical figures of speech such as ‘he spilled the beans’ have ‘a conventional, nonliteral meaning that is comprehended directly’ because in such cases ‘the default meaning is clearly the nonliteral meaning’.60 Gerald Winer and his team compare the reactions of children and adults to a set of questions and observe that children tend to respond to literal but trivial meanings, adults to metaphorical interpretation. They argue that these findings are quite consistent with Grice’s theory, which ‘holds that people respond to the intended or inferred, rather than the literal, meaning of utterances’.61 Anne Bezuidenhout and J. Cooper Cutting suggest that experiments can never bear directly on the debate over the accuracy of Grice’s theory because ‘Grice was interested in giving a conceptual analysis of the concepts of saying, meaning, and so on, and not in giving a psychological theory of the stages in utterance processing.’62 They also express some concerns about the validity of results based on simple measurements of processing time, and the readiness of experimental linguists to reach generalised conclusions from these.

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 198

198 Paul Grice, Philosopher and Linguist

Despite his own philosophical and analytical interests, then, Grice’s theory has been debated in terms of its social application, its computational implementation and its psychological and even neurological plausibility. It has also been discussed by those whose studies extend across cultures, some of whom have criticised it as ethnocentric, as drawing unquestioningly on the mores of one society. In an echo of the accusations of elitism levelled at Austin and his ‘empirical’ approach to language, Grice has even been accused of concentrating on one privileged rank of that society. This particular criticism dates back to Elinor Ochs Keenan’s article ‘The universality of conversational postulates’, first published in 1976. Keenan’s basic tenet is that Grice’s analysis is implicitly but inappropriately offered as applying universally. Her specific claim is that the maxims, in particular those of Quantity, do not hold among the indigenous people of the plateaus area of Madagascar. Hence the maxims are culturally specific and the assumption that Grice’s account demonstrates a pragmatic universal is flawed. In Malagasy society, speakers regularly breach the maxim to ‘be informative’, by failing to provide enough information to meet their interlocutor’s needs. So if A asks B ‘Where is your mother?’, B might well reply ‘She is either in the house or at the market’, even when fully aware of the mother’s location. Further, the reply would not generally give rise to the implicature that B is unable to provide a more specific answer because in that society ‘the expectation that speakers will satisfy informational needs is not a basic norm’.63 Keenan suggests some reasons for this specific to Malagasy society. New information is such a rare commodity that it confers prestige on the knower, and is therefore not parted with lightly. Further, speakers are reluctant to commit themselves explicitly to particular claims, and therefore to take on responsibility for the reliability of the information. For these reasons, the expectations about conversational practice and the implicatures following from particular utterances are different from those assumed by Grice to be universal norms. Keenan complains that ‘the conversational maxims are not presented as working hypotheses but as social facts’ and argues that much more empirical work on conversation must be conducted before true paradigms of conversational practice can be drawn up.64 Her criticism, then, is chiefly the familiar one that Grice has attempted to describe ‘conversation’ without having considered the mechanics of any actual conversation, with the added claim that such a consideration would reveal some significant variations between cultures. However, she welcomes Grice’s attempt to posit universal conversational principles, because of

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 199

Gricean Pragmatics 199

the possibility it suggests for those engaged in fieldwork to move away from specific empirical observations ‘and to propose stronger hypotheses related to general principles of conversation’.65 She suggests that the particular maxims may apply in different domains and to different degrees in different societies. In Malagasy society, determining factors include the importance of the information in question, the degree of intimacy between the participants, and the sex of the speaker. A number of linguists have defended Grice against Keenan’s accusation. Georgia Green has suggested that even the discovery that a particular maxim did not apply at all in one society would not invalidate Grice’s general enterprise. Since the maxims are individual ‘special cases’ of cooperation, ‘discovering that one of the maxims was not universal would not invalidate claims that the Co-operative Principle was universal’.66 Robert Harnish argues that Grice nowhere claims that all conversations are governed by all the maxims; in many types of conversation, as in Keenan’s examples, one or more of the maxims ‘are not in effect and are known not to be in effect by the participants’.67 Geoffrey Leech points out that ‘no claim has been made that the CP applies in an identical manner to all societies.’68 Keenan’s criticisms of Grice, and indeed these responses to those criticisms, consider meaning as a social and interactive, rather than a purely formal phenomenon. This is a project shared to varying degrees by works that have attracted the title ‘politeness theory’. Such works date back to the mimeograph days of ‘Logic and conversation’, and indeed can be seen as one area in which the newly emerging discipline of pragmatics took its inspiration and basic premisses from Grice. Looking back to the early 1970s, politeness theorists Penelope Brown and Stephen Levinson comment that at that time the division of meaning into semantics and pragmatics was motivated chiefly by ‘the basic Gricean observation that what is “said” is typically only part of what is “meant”, the proposition expressed by the former providing a basis for the calculation of the latter.’69 In studying politeness phenomena, linguists such as Brown and Levinson were seeking an explanation not so much of the mechanics of this calculation as of its motivation. John Gumperz observes in the foreword to Brown and Levinson’s study that politeness theory concentrated on the social functions of language, thereby bringing Grice’s work into the field of sociolinguistics for the first time. Theorised conversational politeness, like cooperation, is neither a description of ideal behaviour nor a guide to etiquette. Indeed, politeness theorists are eager to stress that they are positing hypotheses about conventionalised and meaningful patterns of behaviour, rather than

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 200

200 Paul Grice, Philosopher and Linguist

describing a general tendency in people to be ‘considerate’ or ‘nice’ to each other. The conventions of politeness are not accounted for in the maxims, and some have seen this as a weakness in Grice’s account. However, the tendency towards cooperation and that towards politeness are means towards different sets of ends: the former to informing and influencing and the latter to the maintenance of harmonious social relationships. Norms of politeness are therefore appropriate complements to, not subcategories of, norms of cooperation. This certainly seems to have been the view Grice himself took in his only reference to politeness in ‘Logic and conversation’: There are, of course, all sorts of other maxims (aesthetic, social, or moral in character), such as ‘Be polite’, that are also normally observed by participants in talk exchanges, and these may also generate nonconventional implicatures. The conversational maxims, however, and the conversational implicatures connected with them, are specially connected (I hope) with the particular purposes that talk (and so, talk exchange) is adapted to serve and is primarily employed to serve.70 Those who have taken a closer interest than Grice in politeness have generally done so for one of two reasons. Some have given what for Grice are secondary purposes of conversation a more central importance, arguing either explicitly or by implication that the maintenance of social relationships is at least as important a function of conversation as the ‘rational’ goals of informing and influencing. Others have seen in the phenomenon of politeness a possible answer to a question Grice has been criticised for never fully addressing: the question of why speakers convey meaning through implicatures at all, rather than through straightforward assertion. In other words, they have seen politeness not as a supplementary norm of conversation, but as a potential key to the operation of the Cooperative Principle itself. Some of the earliest work in politeness theory came from Robin Lakoff. In common with other generative semanticists, she was dissatisfied with the tenet that judgements of grammaticality were autonomous and sufficient: that a fully successful transformational grammar would be able to legislate absolutely between ‘good’ and ‘bad’ strings. Rather, she argued, phenomena entirely dependent on context, and on aspects of context as specific as the relationship between two people, could effect such judgements and should be taken into account. She has since argued that the notion of ‘continuousness’, the idea that

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 201

Gricean Pragmatics 201

judgements of acceptability must be arranged along a continuous scale rather than being binary and polar, is the single greatest contribution of generative semantics to the understanding of human language.71 In articles published in the early 1970s, Lakoff defended the importance of pragmatic explanations. In ‘Language in context’, for instance, she argues that transformational grammar had ‘explicitly rejected’ contextual aspects from consideration.72 Her particular claim is that certain features of English sentences can be explained only in terms of the relationship between the speaker and hearer. In this sense they are no different from, for instance, honorifics in Japanese; it is just that their context-bound nature is not immediately apparent because they are expressed using linguistic devices that have other, more centrally semantic, functions as well. So Lakoff discusses the use of modal verbs in offers and requests, considering why, for instance, ‘You must have some of this cake’ counts as a politer form than ‘You should have some of this cake’. Their social meaning is often dependent not just on the form used but the implications this triggers. The verb ‘must’ implies, generally contrary to the facts of the case, that the hearer has no choice but to eat the cake, hence that the cake is not in itself desirable, and that the speaker is politely modest about the goods she is offering. Lakoff draws an analogy between these politeness phenomena and some other conversational features ‘not tied to concepts of politeness’, such as those discussed in Grice’s mimeograph.73 Like politeness phenomena, these features are not part of the grammar but can nevertheless have an effect on the form of utterances. Lakoff notes that expressions such as ‘well’ and ‘why’ may sometimes be labelled ‘meaningless’, but that they in fact have specific conditions for use. Because they can be used either appropriately or inappropriately, a linguistic theory needs to be able to account for them, yet the conditions governing their appropriateness are often dependent on a purely contextbound phenomenon such as the breaking of a conversational maxim. They signal, respectively, that the utterance to follow is in some way incomplete, breaking the maxims of Quantity, and that a preceding utterance was in some way surprising, and so considered a possible breaking of the maxims of Quality. The attitudes of speakers to the information conveyed, just like their awareness of their relationships to each other, is sometimes communicated by the actual form of words used. Therefore, any sufficient account of language must pay attention to contextual as well as syntactic factors. In ‘The logic of politeness’, published the following year, Lakoff reiterates her belief in the necessity of contextual explanations to comple-

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 202

202 Paul Grice, Philosopher and Linguist

ment syntactic ones, but this time goes further in arguing that pragmatics should be afforded a status equal to syntax or semantics. Ultimately, she would like to see pragmatic rules ‘made as rigorous as the syntactic rules in the transformational literature (and hopefully a lot less ad hoc)’.74 The basic types of rules to be accounted for fall under two headings: ‘Be clear’ and ‘Be polite’. These sometimes produce the same effect in a particular context and therefore reinforce each other, but more often place conflicting demands on the speaker, meaning that one must be sacrificed in the interests of the other. Grice’s theory of conversation provides an account of the rules subsumed under ‘Be clear’. It remains to be explained why speakers so often depart from these norms in their conversations and here, Lakoff implies, the second type of pragmatic rule comes into play. In cases when the interests of clarity are in conflict with those of politeness, politeness generally wins through. This is perhaps because: ‘in most informal conversations, actual communication of important ideas is secondary to merely reaffirming and strengthening relationships.’75 Lakoff tentatively suggests three ‘rules of politeness’, analogous to the rules of conversation: ‘don’t impose’, ‘give options’ and ‘make A feel good – be friendly’.76 The first rule explains why people choose lengthy or syntactically complex expressions in apparent breach of clarity, such as ‘May I ask how much you paid for that vase, Mr Hoving?’ and ‘Dinner is served’. The second rule explains the use of hedges (‘Nixon is sort of conservative’) and euphemisms (‘I hear that the butler found Freddy and Marion making it in the pantry’), expressions leaving hearers the superficial option of not interpreting an expression in a way they may find offensive. The third rule explains nicknames and particles expressing feeling and involving the hearer, particles such as ‘like’, ‘y’know’, ‘I mean’. All these conversational features might appear to be in breach of one or more maxim; in effect the more powerful rules of politeness explain why and when the conversational maxims are breached. Lakoff even suggests that Grice’s rules of conversation are perhaps best seen as ‘subcases’ of Rule 1; in other words that the tendency towards clarity is just part of a more general tendency not to impose on your addressee. The other two types of politeness, particularly ‘be friendly’, take precedence over clarity when they come into conflict with it. For Lakoff politeness, perhaps including Grice’s maxims, is a subsection of the more general category of ‘cooperation’. Brown and Levinson’s theory of politeness is perhaps the most fully articulated and successful of the attempts to integrate sociolinguistic

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 203

Gricean Pragmatics 203

notions of politeness with Grice’s account of cooperation. The theory itself dates from just after the time when Lakoff was publishing on the need for linguistics to take account of contextual factors. Brown and Levinson first wrote up their ideas in the summer of 1974; these were eventually published as a long essay late in the 1970s, and in book form almost a decade later.77 They claim that their account is based on just the same basic premiss as Grice’s: ‘there is a working assumption by conversationalists of the rational and efficient nature of talk.’78 Brown and Levinson draw on sociology by introducing Erving Goffman’s notion of ‘face’ into the discussion of politeness. In the 1950s, Goffman defined face as the positive social value a person claims, and argued that ‘a person tends to conduct himself during an encounter so as to maintain both his own face and the face of the other participant.’79 Brown and Levinson refine Goffman’s idea by distinguishing between ‘positive’ and ‘negative’ face. Participants in social situations are aware, both in themselves and in others, of the needs of both positive and negative face. Positive face is, generally, the desire to be liked, to be held in high regard by others. Negative face is the desire not to be imposed upon, to retain freedom of choice. Certain types of act can be classified as inherently impolite because they are threatening to one or other type of face for the addressee. Criticisms are inherently impolite because they threaten hearers’ positive face by showing that they are in some way not held in high regard. Orders or requests are inherently impolite because they threaten negative face by seeking to impose the speaker’s will on others and to restrict freedom of choice. In such circumstances the speaker has the option of producing the impolite act ‘bald onrecord’.80 That is, of producing it without any attempt to disguise it: to say ‘You look fat in that dress’, or ‘Give me a lift to the station’. Brown and Levinson observe that in doing this, the speaker adheres precisely to Grice’s maxims; ‘You look fat in that dress’ is clear, informative and truthful. The speakers may be adhering to the maxims of conversation in these examples, but they are in breach of the rules of politeness prescribing maintenance of the other’s face. Utterances such as ‘That isn’t the most flattering of your dresses’ and ‘I wonder if I could trouble you for a lift to the station’ are less clear and less efficient than their ‘on-record’ counterparts. However, the very fact that speakers are putting more effort into producing these longer and more complex utterances ensures that they are seen to be striving to satisfy the hearers’ face wants. Such ways of conveying meaning are very common in conversation, Brown and

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 204

204 Paul Grice, Philosopher and Linguist

Levinson note. Successive commentators have been simply wrong in assuming that Grice specifies that utterances must meet the conditions of cooperation. Their purpose here, in fact, is ‘to suggest a motive for not talking in this way’.81 Like Lakoff, they note that hedges are generally concerned with the operation of the maxims. They emphasise that cooperation is met, indicate that it may not be met, or question whether it has been met.82 Further, ‘off-record’ strategies are often accompanied by a ‘trigger’ indicating that an indirect meaning is to be sought. The individual conversational maxims determine the appropriate type of trigger. Work on politeness in naturally occurring conversation has produced both supportive and more pessimistic responses to the Gricean framework. Susan Swan Mura draws on the transcriptions of 24 dyadic interactions to consider factors that might license deviation from the maxims. She concludes that for all the maxims the relevant factors provide for problem-free interaction in contexts where there is a threat of some sort to the coherence of the interaction. She concludes enthusiastically that ‘this study also provides preliminary evidence that the Cooperative Principle represents a psychologically real construct, modeling the behavioural (in a non-Skinnerian sense of the term) basis for conversational formulation and interpretation.’83 Yoshiko Matsumoto, on the other hand, considers politeness phenomena in Japanese and notes that ‘a socially and situationally adequate level of politeness and adequate honorifics must obligatorily be encoded in every utterance even in the absence of any potential face threatening act’, arguing that this is a serious challenge to the universality of both Grice’s and Brown and Levinson’s theories.84 In his contribution to the literature on politeness, Geoffrey Leech draws together concepts from contemporary pragmatics and adds some of his own, to present what he describes as an ‘Interpersonal Rhetoric’.85 He claims that the Cooperative Principle, although an important account of what goes on in conversation, is not in itself sufficient to explain interpersonal rhetoric. Principles of politeness and also of irony must be added on an equal footing with cooperation; indeed, they can sometimes ‘rescue’ it from apparently false predictions. The result is a vastly more complex account of conversation than Grice had envisaged, as well as a more complex pragmatics in general. For example, Leech suggests that if, in response to ‘We’ll all miss Bill and Agatha, won’t we?’, a speaker were to say ‘Well, we’ll all miss BILL’, thereby implicating that she will not miss Agatha, the speaker has apparently failed to observe the needs of Quantity. The speaker ‘could have been more

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 205

Gricean Pragmatics 205

informative, but only at the cost of being more impolite to a third party: [the speaker] therefore suppressed the desired information in order to uphold the P[oliteness] P[rinciple]’.86 Leech’s PP itself contains an elaborate list of additional conversational maxims. He initially describes these as comprising Tact, Generosity, Approbation, Modesty, Agreement and Sympathy, but later suggests that further maxims of politeness may prove necessary, including the Phatic maxim, the Banter maxim and the Pollyanna maxim.87 He seems, in fact to be prepared to add maxims in response to individual examples as they present themselves. This stance is open to the criticism that the choice of maxims becomes ad hoc, and the number arbitrary, significantly weakening the explanatory power of the theory. Indeed Brown and Levinson suggest that Leech is in danger of arguing for an ‘infinite number of maxims’, making his addition of the politeness principle in effect an elaborate description, rather than an explanatory account, of pragmatic communication.88 Leech’s ‘Interpersonal Rhetoric’ is unusual among theories of politeness for advocating such a plethora of principles in addition to Grice’s maxims of cooperation. It is certainly unusual in the more general framework of Gricean pragmatics. Here, the tendency in responses to the theory of conversation has been to attempt to reduce the number of maxims, to streamline the theory and therefore make it more powerfully explanatory and psychologically plausible. In their overview of the place of pragmatics in linguistics during the last quarter of the twentieth century, Hartmut Haberlard and Jacob Mey criticise Leech for multiplying pragmatic principles and types, while most responses to Grice have been reductive in tendency. The general feeling that Grice’s maxims need tidying up and clarifying is, they suggest, not particularly surprising ‘if one considers the strong a priori character of Grice’s maxims’.89 His main motivation in their choice and titles was to echo Kant’s table of categories. Perhaps the most extreme of such theories was Relevance Theory (RT), developed by a number of linguists, but principally by Dan Sperber and Deirdre Wilson. It offered an account of communication and cognition drawing on Grice’s theories of meaning and of conversation, especially on his distinction between what is said and what is meant, while at the same time departing from Grice significantly. From the mid-1980s and into the 1990s, RT was perhaps the most influential school of thought in pragmatics and spawned a large number of articles, conference papers and doctoral theses, especially at its ‘home’ at University College, London. It was used as a framework for the analysis of various semantic and pragmatic phenomena in a variety

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 206

206 Paul Grice, Philosopher and Linguist

of languages, as well as an approach to texts as diverse as poetry, jokes and political manifestos. In Relevance, Sperber and Wilson argue that Grice’s account is too cumbersome and too ad hoc, consisting in a series of maxims each intended to account for some particular, more or less intuitive response. Gricean pragmatics had for too long been bound by the outline provided by ‘Grice’s original hunches’.90 The dangers of this are apparent: ‘It might be tempting to add a maxim every time a regularity has to be accounted for.’91 Rather, the various maxims and submaxims can be subsumed into a suitably rich notion of relevance. This does not consist of a maxim, such as Grice’s ‘Be relevant’, which speakers may adhere to or not, and which in the later case may lead to implicature. There is no place in Sperber and Wilson’s theory for the notion of ‘flouting’ a rule of conversation. Rather, they posit a ‘Principle of Relevance’, a statement about communication; ‘communicators do not “follow” the principle of relevance; and they could not violate it even if they wanted to’.92 The principle applies ‘without exception’. It is a fact of human behaviour explaining how people produce, and how they respond to, communicative acts. The principle states that every act of ostensive communication is geared to the maximisation of its own relevance; people produce utterances, whether verbal or otherwise, in accordance with the human tendency to be relevant, and respond to others’ utterances with this expectation. Relevance is defined as the production of the maximum number of contextual effects for the least amount of processing effort. These notions are all entirely dependent on context; just as there is no place for flouting in Sperber and Wilson’s theory, so there is no place for any notion of generalised implicature. Every implicature is entirely dependent on and unique to the context in which it occurs. Sperber and Wilson take issue with Grice over what they see as his simplistic division of the semantic and the pragmatic, or the encoded and the inferenced. It is simply not possible, they argue, to arrive at a unique, fully articulated meaning from decoding alone: to produce a semantic form on which separate, pragmatic principles can then work. In assuming that this was possible, Grice took an unwarranted step back towards the code model he so tellingly critiqued in his earlier article ‘Meaning’, returning to an outmoded reliance on the linguistic code to deliver a fully formed thought. He abandoned a ‘common-sense’ approach with psychological plausibility: the attempt to formalise and explain the notion that ‘communication involves the publication and recognition of intentions’.93 Cognitive principles such as relevance will be involved even at the stage of producing a Gricean ‘what is said’,

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 207

Gricean Pragmatics 207

determining features such as reference assignment and disambiguation. Unlike the Gricean maxims, the Principle of Relevance is not an autonomous rule applying after decoding has issued a unique linguistic meaning. Rather, both ‘explicit’ and ‘implicit’ meaning rely on the Principle of Relevance. Sperber and Wilson argue, contra Grice, for the need to consider explicitness as a matter of degree. Unlike the ethnologists and politeness theorists who claim common ground with Grice through an interest in social interaction, relevance theorists see their approach, and also Grice’s maxims, as essentially cognitive in nature. For instance, Diane Blakemore has argued that RT responds to Grice’s own dismissal of ‘the idea that the maxims had their origin in the nature of society or culture on the grounds that this did not provide a sufficiently general explanation’ and his warning that ‘the key to a general, psychological explanation lay in the notion of relevance’.94 There is some evidence that Grice was reluctant to be drawn on his views on RT. His only published comment is very brief and rather damning. In the retrospective overview of his own work in Studies in the Way of Words, he comments on the possible problem of the interdependence of his maxims, especially Quantity and Relation. However, he argues that Sperber and Wilson’s account, far from clarifying this matter, in fact loses the significance of the link between relevance and quantity of information by considering relevance without specification of direction. The amount of information appropriate in any given context is determined by the particular aim of the conversation, or of what is deemed relevant to that aim. So success in conversation is not simply a matter of accumulating the highest possible number of contextual effects. Further, Grice argues that as well as direction of relevance, information is necessary concerning the degree of a concern for a topic, and the extent of opportunity for remedial action. In sum, the principle of relevance may well be an important one in conversation, but when it comes to implicature it might be said ‘to be dubiously independent of the maxim of Quantity’.95 This last suggestion seems to align Grice more with another branch of pragmatics, again one attempting to refine the notion of implicature and reduce the number of conversational maxims: the work of the so-called ‘neo-Griceans’. The term ‘neo-Gricean’ has been applied to a number of linguists, both American and British, and to a body of work produced since the early 1980s. Across the range of their published work, these linguists have revised their ideas many times, and used a sometimes bewildering array of terminology. Nevertheless, the work all remains located firmly

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 208

208 Paul Grice, Philosopher and Linguist

within the tradition of Gricean pragmatics, seeking to elaborate a theory of meaning based on levels of interpretation. Neo-Griceans have tended towards two types of revision of Grice’s theory of conversation, one much more radical than the other. First, very much in keeping with Grice’s own speculations about the interdependence of Quantity and Relation, they have questioned the need for as many separate maxims as Grice originally postulated, and investigated ways of reducing these. Second, and rather more at odds with Grice’s own original conception, they have questioned the sharpness of the distinction between the levels of meaning. They have been concerned with what Laurence Horn has described as a problem raised by the Cooperative Principle, ‘one we encounter whenever we cruise the transitional neighbourhood of the semantics/pragmatics border: how do we draw the line between what is said and what is implicated?’96 The neo-Griceans have upheld the importance of making some such distinction, but they have in various ways and with different terminologies challenged the viability of separating off entirely the ‘literal’ from the ‘implied’. In this, they share the anxiety expressed by Sperber and Wilson in terms of the appropriate division between ‘explicit’ and ‘implicit’ meaning. Sperber and Wilson argue that purely explicit meaning is generally not enough to render a full propositional form; implicit aspects of meaning must be admitted at various stages, leading to ‘levels’ of explicitness. Neo-Griceans have made similar claims, leading them not actually to abandon the distinction between semantics and pragmatics, but to question the autonomy of the two, and the possibility of their applying to interpretation in series. The chief ways in which the neo-Griceans differ from relevance theorists, and stay closer to Grice, is in their maintenance of separate maxims of conversation that can be exploited and flouted, and the use they make of generalised conversational implicatures (GCIs). These latter, rejected by RT because they take minimal account of context, are crucial to the development, and presentation, of much of the neo-Gricean literature. In his 1989 book A Natural History of Negation, Laurence Horn comments that ‘much of neo- and post-Gricean pragmatics has been devoted to a variety of reductionist efforts’.97 He suggests that a premiss for him, and perhaps for many of the neo-Griceans, setting them apart from relevance theorists, is that ‘Quality . . . is primary and essentially unreducible.’ What needs further to be said about conversation can be summarised in two separate, sometimes conflicting, driving forces within communication. These are, in general terms, the drive towards clarity, putting the onus on speakers to do all that is necessary to make

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 209

Gricean Pragmatics 209

themselves understood, and the drive towards economy, putting the onus on hearers to interpret utterances as fully as necessary. Horn labels the principles resulting from these drives Q and R respectively (standing for ‘quantity’ and ‘relation’, but with no straightforward correlation to Grice’s maxims). He traces the development of this reductive model to early work by Jay Atlas and Stephen Levinson, and by Levinson on his own. But an important part of the account of ‘Q-based implicatures’ draws on work almost a decade earlier by Horn himself on the notion of ‘scalar implicatures’. Horn’s development of the Gricean notion of Quantity to account for the interpretation of a range of vocabulary in terms of ‘scalar implicature’ is arguably one of the most productive and interesting of the expansions of ‘Logic and conversation’. It is in many ways true to Grice’s original conception of implicature, not least because it divides the ‘meaning’ of a range of terms into ‘said’ and ‘implicated’ components, removing the need to posit endless ambiguities in the language, and offering a systematic explanation of a wide range of examples. It is perhaps in keeping that such a thoroughly Gricean enterprise should remain in manuscript for many years, and then be published only partially (in A Natural History of Negation). The main source for Horn’s account of scalar implicature remains his doctoral thesis, On the Semantic Properties of Logical Operators in English, presented to the University of California, Los Angeles, in 1972. In his later commentary on scalar implicatures, Horn describes them as the ‘locus classicus’ of GCIs based on his own notion of Quantity.98 Indeed, Grice himself seems to have seen examples of this type as particularly important, although he does not use the term ‘scalar’, or identify them as a separate category. In ‘Logic and conversation’, he introduces the notion of GCI as follows: Anyone who uses a sentence of the form X is meeting a woman this evening would normally implicate that the person to be met was someone other than X’s wife, mother, sister, or perhaps even close platonic friend. . . . Sometimes, however, there would be no such implicature (‘I have been sitting in a car all morning’) and sometimes a reverse implicature (‘I broke a finger yesterday’). I am inclined to think that one would not lend a sympathetic ear to a philosopher who suggested that there are three senses of the form of expression an X. . . . [Rather] when someone, by using the form of expression an X, implicates that the X does not belong to or is not otherwise closely connected with some identifiable person, the implicature is present

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 210

210 Paul Grice, Philosopher and Linguist

because the speaker has failed to be specific in a way in which he might have been expected to be specific, with the consequence that it is likely to be assumed that he is no in a position to be specific. This is a familiar implicature and is classifiable as a failure, for one reason or another, to fulfil the first maxim of Quantity.99 The most easily identifiable type of GCI is based on the first maxim of Quantity, which calls for as much information as is necessary. In other words, it draws on Grice’s earliest observation about the regularities of conversation: that one should not make a weaker statement when a stronger one is possible. Horn observes that there are many areas of vocabulary where similar apparent problems can be resolved by just such an analysis. A more general term often implicates that a more specific alternative does not apply, by means of the assumption that the speaker is not in a position to provide the extra information. Horn discusses this informativeness in terms of the ‘semantic strength’ of a lexical item. Such items can be understood as ranged on ‘scales’ of informativeness, each scale consisting of two or more items. In all cases, it is possible to describe the different meanings associated with a particular item in terms not of a semantic ambiguity, but of a basic semantic meaning and of an additional implicated meaning that arises by default, but can be cancelled in context. The first scale Horn discusses is that of the cardinal numbers. The numbers can be seen as situated on a scale, ranging from one to infinity: ·1, 2, 3, 4, 5, . . . Ò. Each item semantically entails all items to its left, and pragmatically implicates the negation of all items to its right. As Horn describes it: Numbers, or rather sentences containing them, assert lowerboundedness – at least n – and given tokens of utterances containing cardinal numbers may, depending on the context, implicate upperboundedness – at most n – so that the number may be interpreted as denoting an exact quantity.100 This explains the apparent ambiguity of a cardinal number in use. ‘John has three children’, for example, will generally be interpreted as meaning ‘John has exactly three children’, but is not logically incompatible with John’s having four or more. Horn points to the fact that ‘John has three children, and possibly even more’ is acceptable, suggesting that ‘no more than three’ cannot be part of the semantic meaning of ‘three’. If it were, this example should be simply logically contradictory. Rather than forcing the intolerable conclusion that all

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 211

Gricean Pragmatics 211

cardinal numbers are semantically multiply ambiguous, Horn’s account allows that their semantic meaning is ‘at least . . .’ but that they implicate ‘at most . . .’. This is because, if the speaker were able to use an item further to the right on the scale, it would be more informative, hence more cooperative, to do so. Taken together, these two meanings give the interpretation ‘exactly . . .’. Horn proposes similar scales for vocabulary items from across the range of the lexicon, including adjectives, ·warm, hotÒ, ·good, excellentÒ; verbs, ·like, loveÒ, ·dislike, hateÒ; and quantifiers, ·all, someÒ. As Horn observes in his own later commentary, in all these examples ‘since the implicature relation is contextdependent, we systematically obtain two understandings for each scalar value p (‘at least p’, ‘exactly p’) without needing to posit a semantic ambiguity for each operator’.101 In this later work, Horn describes scalar implicatures as examples of Q-based implicatures, dependent on our expectations that speakers will seek to be as clear as possible. However, Q-based implicatures need to be balanced against R-based ones, drawing on the notion of economy. The tension between these can offer a clear account of the issue to do with articles Grice originally identified: Maxim clash, for example, arises notoriously readily in indefinite contexts; thus, an utterance of I slept in a car yesterday licenses the Qbased inference that it was not my car I slept in (or I should have said so), while an utterance of I broke a finger yesterday licenses the Rbased inference that it was my finger I broke (unless I know that you know that I am an enforcer for the mob, in which case the opposite, R-based implicatum is derived).102 There is some experimental support for the psychological plausibility of Horn’s scalar implicatures, and hence for the Gricean framework on which they are based. Ira Noveck concludes from a number of experiments that although adults react to the enriched pragmatic meaning of scalars, young but linguistically competent children respond chiefly to their logical meaning. ‘The experiments succeeded in demonstrating not only that these Gricean implicatures are present in adult inferencemaking but that in cognitive development these occur only after logical interpretations have been well established.’103 There have been some attempts to extend the range of Horn’s notion of scalar implicature. Horn himself has offered a scalar analysis of the ‘strengthening’ of ‘if’ to mean ‘if and only if’.104 Earlier, Jerrold Sadock sought to explain the meaning of ‘almost’ in terms of a semantic scale

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 212

212 Paul Grice, Philosopher and Linguist

·almost p, pÒ. Sadock observes that the argument for a relationship of implicature between the use of ‘almost p’ and the meaning ‘not p’ is considerably weakened by the difficulty of finding contexts in which it would be cancellable. His explanation is that this is because ‘the context-free implicature in the case of almost is so strong’.105 In discussing the one example he does see as cancellable, he introduces a qualification that is again to do with the ‘strength’ of an implicature. Surely, Sadock argues, if a seed company ran a competition to breed an ‘almost black marigold’, and a contestant turns up with an entirely black marigold, it would be in the interests not just of fair play but of correct semantic interpretation to award him the prize. However, he cautions: In the statement of the contest’s rules, the word almost should not be read with extra stress. Stressing a word that is instrumental in conveying a conversational implicature can have the effect of emphasizing the implicature and elevating it to something close to semantic content.106 Sadock does not go so far as claiming that an implicature can actually become part of semantic content, but his rather equivocal statement certainly suggests a blurring of what for Grice is a clearly defined boundary. This blurring of the boundaries is more apparent, or more explicitly advocated, in other neo-Gricean work, initially from Gazdar and later from Atlas and Levinson. Gerald Gazdar is perhaps the least clearly ‘Gricean’ of the neo-Griceans, in that he departs most radically from Grice’s bipartite conception of meaning. Nevertheless, he bases his pragmatic account at least in part on a conception of implicature drawing on ‘Logic and conversation’. He proposes a ‘partial formulization’ of GCIs, drawing on Horn’s original account, which he describes as ‘basically correct’.107 For Gazdar, the semantic representation of certain expressions carry ‘im-plicatures’, potential implicatures that will later become actual if not cancelled by context. The relationship between context, implicature and eventual interpretation is a formal process. Stephen Levinson, sometimes in collaboration with other linguists, has developed a revised account of GCIs. Levinson’s conception of the processes of interpretation is summarised in his exegesis of ‘Logic and conversation’ in his 1983 Pragmatics. He describes how the assumptions about conversation expressed in the maxims ‘arise, it seems, from basic rational considerations and may be formulated as guidelines for the efficient and effective use of language’.108 These goals in conversation are

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 213

Gricean Pragmatics 213

closely related to Horn’s ‘economy’ and ‘clarity’, respectively, and in Levinson’s work too they prompt the search for a reductive reformulation of the maxims. He is also concerned with developing an ‘interleaved’ account of the relationship between semantics and pragmatics. Along with RT, but differing overtly from it, Levinson’s account offers perhaps the most developed alternative to the view of semantics and pragmatics operating serially over separate inputs. In a joint article published in 1981, Atlas and Levinson lay claim to an explicitly Gricean heritage, but argue that ‘Gricean theories need not and do not restrict their resources to Grice’s formulations.’109 They argue that it is necessary to pay attention to semantic structure that exists above and beyond truth-conditions, and that in turn interacts with pragmatic principles to produce ‘informative, defeasible implicata’.110 This concern with layers of meaning not found in Grice’s formulation has remained in Levinson’s work. His 2000 book Presumptive Meanings is a sustained defence of the notion of GCI, based on Grice’s original conception but reformulating and elaborating it. As in the defence from 20 years earlier, GCIs are seen as informative and defeasible. Their essential purpose is to add information to what is offered by literal meaning; as Levinson explains, they help overcome the problem of the ‘bottleneck’ created by the slowness of phonetic articulation relative to the possibilities of interpretation. The solution, he suggests, takes the following form: ‘let not only the content but also the metalinguistic properties of the utterance (e.g., its form) carry the message. Or, find a way to piggyback meaning on top of the meaning.’111 As part of this programme, Levinson offers a detailed characterisation of three separate levels of meaning in utterance interpretation. As well as ‘sentence meaning’ and ‘speaker meaning’, he argues the need for ‘statement-meaning’ or ‘utterance-type meaning’, intermediate between the two. It is at this level, he suggests, that ‘we can expect the systematicity of inference that might be deeply interconnected to linguistic structure and meaning, to the extent that it can become problematical to decide which phenomena should be rendered unto semantic theory and which unto pragmatics.’112 He places GCIs at this intermediate level; for him they are ambiguously located between semantics and pragmatics. Part of his argument here hinges on the idea, not dissimilar from that put forward by Sperber and Wilson, that the very first level is often not enough to give a starting point of meaning at all. Various factors in what Grice calls ‘what is said’ need to be determined with reference to precisely those inferential features described as GCIs. Levinson describes this process as ‘pragmatic intrusion into truth conditions’.113 Pragmatic

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 214

214 Paul Grice, Philosopher and Linguist

inference plays a part in determining even basic sentence meaning, suggesting that pragmatics cannot be simply a ‘layer’ of interpretation applying after semantics.114 Neo-Griceans have sometimes been described as engaging in ‘radical pragmatics’ because, as Levinson’s model illustrates, they claim much richer explanatory powers for their pragmatic principles than is the case in traditional Gricean pragmatics. In ‘Logic and conversation’, the layers of meaning before the Cooperative Principle becomes effective are somewhat shady. Conventional meaning, always a problem area for Grice, is added to by various uninterrogated processes of disambiguation and reference assignment, to give ‘what is said’. This may be enriched by conventional implicature, depending on the individual words chosen. It is only then that Grice’s elaborate mechanism for deriving conversational implicatures is brought into play. For neo-Griceans such as Levinson, conventional meaning is a highly restrictive, although perhaps no less shady, phenomenon. It needs assistance from pragmatic principles before it can even offer a basis from which interpretation can proceed. In John Richardson and Alan Richardson’s instructive extended metaphor: Grice allowed himself considerable pragmatic machinery with which to try to bridge the gap between intuitive and posited meanings, whereas radical pragmaticians have been eagerly scrapping much of Grice’s machinery while frequently allowing gaps between intuitive and posited meanings greater than anything Grice proposed – a geometrically more ambitious program.115 A number of recent articles reassess the claims that implicature must play a part in determining literal content. These generally take issue with linguists such as the relevance theorists and Levinson, and show a renewed interest in Grice’s distinction between what is said and what is implicated, although they are by no means unanimously uncritical of his original characterisation of these. Kent Bach argues that ‘what is said in uttering a sentence need not be a complete proposition’; if we observe what Bach sees as Grice’s stipulation that ‘what is said’ must correspond directly to the linguistic form of the sentence uttered, it may be that what is said is not informationally complete.116 However, this need not be a problem, because in the Gricean framework what is communicated is not entirely dependent on what is said but is derived from it by a process of implicature. Bach accuses some of Grice’s commentators of confusing their intuitions about what is communicated, in which

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 215

Gricean Pragmatics 215

implicatures play a part, with judgements about semantic fact. Directly taking issue with RT, he argues that ‘I have had breakfast’ might be literally true even if the speaker had not had breakfast on the day of utterance; it is simply that intuitive judgements focus on a more limited interpretation. These intuitive judgements are not a good guide to the theoretical question of the constitution of ‘what is said’: ‘the most natural interpretation of an utterance need not be what is strictly and literally said.’117 Mira Ariel has also dwelt on intuitive responses in her discussion of ‘privileged interactional interpretation’.118 What people respond to as the speaker’s chief commitment does not always correspond either to Grice’s ‘what is said’ or to some enriched form such as the relevance theorists’ ‘explicatures’. Jennifer Saul suggests that Grice’s distinction between what is said and what is implicated has been dismissed too hurriedly by his critics. To argue against Grice on the grounds of how people generally react to utterances is seriously to misrepresent his intention. Further, within Grice’s original formulation there is no reason why what is said needs to be a cooperative contribution recognised by either speaker or hearer: ‘a speaker can perfectly well be cooperative by saying something trivial and implicating something non-trivial.’119 Grice’s work was often characterised by the speculative and openended approaches he took to his chosen topics. At times the result can be frustratingly tentative discussions of crucial issues that display a knowing disregard for detail or for definitions of key terms. Ideas and theories are offered as sketches of how an appropriate account might succeed, or how a philosopher – by implication either Grice or some unnamed successor – might eventually formulate an answer. In the case of the theory of conversation, these characteristics are perhaps at least in part responsible for its success. Certainly, Grice offered enough for his description of interactive meaning to be applied fruitfully in a number of different fields across a vast range of data. But the sketchiness of some aspects of the theory, and the failure fully to define terms as central as ‘what is said’, have also indicated possibilities for theoretical adjustments and reformulations. There is something of an air of ‘work in progress’ about even the published version of ‘Logic and conversation’ that indicates problems to be solved rather than simply ideas to be applied. Grice himself mounted a spirited defence of problems at the end of his own philosophical memoir: If philosophy generated no new problems it would be dead, because it would be finished; and if it recurrently regenerated the same old

1403_902976_10_cha09.qxd 12/13/04 6:31 PM Page 216

216 Paul Grice, Philosopher and Linguist

problems it would still not be alive because it could never begin. So those who still look to philosophy for their bread-and-butter should pray that the supply of new problems never dries up.120 A ready supply of problems is as valued by linguists as it is by philosophers. The original, highly suggestive, but somehow unfinished nature of the lectures on logic and conversation is a telling example of a Gricean heresy, and one that may explain much of the story of their success.

1403_902976_11_note.qxd 12/13/04 6:33 PM Page 217

Notes 1 The skilful heretic 1. Baker (1986: 277). 2. Festschrift etc. H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 3. See, for instance, Wittgenstein (1958: 43). 4. See Grice’s obituary in The Times, 30 August 1988. 5. In Strawson and Wiggins (2001: 516–17). 6. E.g. Brown and Yule (1983: 31–3), Coulthard (1985: 30–2), Graddol et al. (1994: 124–5), Wardhaugh (1986: 281–4). 7. Respectively, Rundquist (1992), Perner and Leekam (1986), Gumperz (1982: 94–5), Penman (1987), Schiffrin (1994: 203–26), Raskin (1985), Yamaguchi (1988), and Warnes (1990). 8. See, for instance, Grice (1987b: 339). 9. Grandy and Warner (1986a: 1). 10. Grice (1986a: 65, original emphasis). 11. Grice (1986b: 10). 12. ‘Philosophy at Oxford 1945–1970’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 13. Misc. notes, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 14. Grice (1978) in Grice (1989: 47). 15. Grice (1986a: 61). 16. See, for instance, Burge (1992: 619). 17. Grice (1986a: 61–2). 18. See, for instance, Strawson and Wiggins (2001: 527). 19. Kathleen Grice, personal communication. 20. Richard Warner, personal communication. 21. Grice’s obituary in The Times, 30 August 1988. 22. Grice (1967b) in Grice (1989: 64). 23. Handwritten version of Kant lectures, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 24. ‘Miscellaneous “Group” notes 84–85’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 25. ‘Richards paper and notes for “Reply”’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley.

2 Philosophical influences 1. Tape, 29 January 1983, Grice in conversation with Richard Warner and Judith Baker, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley.

217

1403_902976_11_note.qxd 12/13/04 6:33 PM Page 218

218 Notes 2. 3. 4. 5. 6. 7.

8. 9. 10. 11. 12. 13.

14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.

Grice (1986a: 46). Quotation courtesy of the Old Cliftonian Society. Grice (1991: 57–8, n1). Grice (1986a: 46–7). Ibid. Quotations in this paragraph from Tape, 29 January 1983, Grice in conversation with Richard Warner and Judith Baker, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Information from The Rossall Register, Rossall School, Lancashire. Martin and Highfield (1997: 337). Ayer (1971: 48). Rogers (2000: 144). ‘Preliminary Valediction, 1985’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Tape, Oxford Philosophy in the 1930s and 1940s, APA, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Grice (1986a: 48). ‘Negation and Privation’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Rogers (2000: 255). Rogers (2000: 146). Locke (1694: 37) in Perry (1975b). Locke (1694: 38) in Perry (1975b). Locke (1694: 39) in Perry (1975b). Ibid. Locke (1694: 47) in Perry (1975b). Reid (1785: 115) In Perry (1975b). Gallie (1936: 28). Grice (1941) in Perry (1975b: 74). Grice (1941) in Perry (1975b: 88). Grice (1941) in Perry (1975b: 90). Perry (1975a: 155n). Perry (1975a: 136). Williamson (1990: 122).

3 Post-war Oxford 1. 2. 3. 4. 5. 6. 7. 8. 9.

Warnock (1963) in Fann (1969: 10), Strawson in Magee (1986: 149). Grice (1986a: 48). Ryle (1949: 19). Rée (1993: 8). Warnock (1963) in Fann (1969: 20). Pitcher (1973: 20). Grice (1986a: 58). Rogers (2000: 255). Manuscript: ‘Philosophy and ordinary language’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley.

1403_902976_11_note.qxd 12/13/04 6:33 PM Page 219

Notes 219 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32.

33. 34. 35. 36. 37. 38. 39. 40. 41.

42. 43. 44.

45. 46. 47. 48. 49.

Grice (1986a: 50). Frege (1892: 57). Russell (1905: 479). For instance, Magee (1986: 10). For instance in the introduction to Gellner (1979), written and first published in 1959. For instance Grice (1987b: 345). Moore (1925) reprinted in Moore (1959: 36). Grice (1986a: 51). Gellner (1979: 17, and elsewhere). Warnock (1963) in Fann (1969: 11). Rogers (2000: 254). Warnock (1963) in Fann (1969: 11). Williams and Montefiore (eds) (1966: 11). Wittgenstein (1922: 4.024). Wittgenstein (1958: 120). Ayer (1964). Austin (1962a: 3). Austin (1950) in Austin (1961: 99). Austin (1956) in Austin (1961: 130). Austin (1956) in Austin (1961: 137). Austin (1956) in Austin (1961: 138, original emphasis). Forguson (2001: 333 and n16). Tape, Oxford Philosophy in the 1930s and 1940s, APA, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Grice (1986a: 51), Grice (1987a: 181). Grice (1986a: 49). Quine (1985: 249). Warnock (1963) in Fann (1969: 14). Notes for Ox Phil 1948–1970, APA, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Kathleen Grice (personal communication). Warnock (1973: 36). Austin (1956) in Austin (1961: 129, original emphasis). Tape, Oxford Philosophy in the 1930s and 1940s, APA, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Urmson et al. (1965: 79–80). ‘Philosophers in high argument’, The Times Saturday 17 May 1958. Tape, 29 January 1983, Grice in conversation with Richard Warner and Judith Baker, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. See Strawson in Magee (1986: 149), and Gellner (1979, passim) as examples of these two points of view. Russell (1956: 141). Gellner (1979: 23). Gellner (1979: 264). Mundle (1979: 82).

1403_902976_11_note.qxd 12/13/04 6:33 PM Page 220

220 Notes 50. Tape, Oxford Philosophy in the 1930s and 1940s, APA, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 51. Ibid. 52. Grice (1986a: 51). 53. Manuscript draft of ‘Reply to Richards’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 54. Tape, Oxford Philosophy in the 1930s and 1940s, APA, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 55. Ibid. 56. Ackrill (1963: vi). 57. Strawson (1950: 326). 58. Strawson (1998: 5). 59. For instance, in Strawson (1952: 175ff). 60. Strawson (1950: 344). 61. Quine (1985: 247–8). 62. Grice (1986a: 47). 63. Grice (1986a: 49). 64. Tape, 29 January 1983, Grice in conversation with Richard Warner and Judith Baker, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 65. Notes for Categories with Strawson, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 66. Quine (1953: 43). 67. Grice and Strawson (1956) in Grice (1989: 205). 68. Kathleen Grice, personal communication. 69. ‘The Way of Words’, Studies in: Notes, offprints and draft material, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 70. Grice (1986a: 54). 71. Quinton (1963) in Strawson (1967: 126). 72. Tape, 29 January 1983, Grice in conversation with Richard Warner and Judith Baker, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 73. Austin (1962a: 37). 74. Warnock (1973: 39). 75. Perception Papers, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 76. Austin (1962a: 30–1). 77. Tape, 29 January 1983, Grice in conversation with Richard Warner and Judith Baker, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 78. ‘More about Visa’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 79. Warnock (1955: 201). 80. Grice (1961) and Grice (1962). 81. Grice (1966). 82. Warnock (1973: 39, original emphasis).

1403_902976_11_note.qxd 12/13/04 6:33 PM Page 221

Notes 221 83. Kathleen Grice, personal communication. 84. Quine (1985: 248). 85. Grice (1958).

4 Meaning 1. Tape, 29 January 1983, Grice in conversation with Richard Warner and Judith Baker, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 2. Tape, Oxford Philosophy in the 1930s and 1940s, APA, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 3. Tape, 29 January 1983, Grice in conversation with Richard Warner and Judith Baker, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 4. See, for instance, Warnock (1973: 40). 5. Grice (1957: 379). 6. Stevenson (1944: 66). 7. Stevenson (1944: 38). 8. Stevenson (1944: 69). 9. Stevenson (1944: 74, original emphasis). 10. Grice (1957: 380). 11. Stout (1896: 356). 12. Notes on intention and dispositions – HPG and others in Oxford, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 13. ‘Dispositions and intentions’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 14. Ryle (1949: 103). 15. Ryle (1949: 183). 16. Stout (1896: 359). 17. Hampshire and Hart (1958: 7). 18. Grice (1957: 379). 19. Peirce (1867: 1). 20. Peirce (1867: 7). 21. Lectures on Peirce, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 22. Grice (1957: 382). 23. Grice (1957: 385). 24. Grice (1957: 386). 25. Grice (1957: 387). 26. Bennett (1976: 11). 27. Schiffer (1972: 7). 28. Tape, 29 January 1983, Grice in conversation with Richard Warner and Judith Baker, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 29. Austin (1962b: 149). 30. Austin (1962b: 108).

1403_902976_11_note.qxd 12/13/04 6:33 PM Page 222

222 Notes 31. Austin (1962b: 114). 32. Tape, 29 January 1983, Grice in conversation with Richard Warner and Judith Baker, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 33. Strawson (1964) in Strawson (1971: 155). 34. Strawson (1964) in Strawson (1971: 163). 35. Strawson (1969) in Strawson (1971: 171–2). 36. Searle (1986: 210). 37. Searle (1969: 45). 38. A similar point is made by Sperber and Wilson (1995: 25). 39. Schiffer (1987: xiii). 40. Schiffer (1972: 13). 41. Schiffer (1972: 18–19). 42. Cosenza (2001a: 15). 43. Avramides (1989), Avramides (1997). 44. Schiffer (1972: 3). 45. Ziff (1967: 1, 2). 46. Ziff (1967: 2–3). 47. Malcolm (1942: 349). 48. Malcolm (1942: 357, original emphasis). 49. Grice (c. 1946–1950: 150). 50. Grice (c. 1953–1958: 167). 51. Grice (c. 1953–1958: 168).

5 Logic and conversation 1. See Quine (1985: 235). 2. Grice (1986a: 59–60). 3. Tape, 29 January 1983, Grice in conversation with Richard Warner and Judith Baker, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 4. Warnock (1973: 36). 5. Tape, 29 January 1983, Grice in conversation with Richard Warner and Judith Baker, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 6. Chomsky (1957: 5). 7. See Warnock (1963) in Fann (1969: 21). 8. ‘Lectures on negation’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 9. Aristotle Categoriae, ch. 12, 14a, in Ackrill (1963). 10. Aristotle Categoriae, ch. 13, 14b, in Ackrill (1963). 11. Aristotle Categoriae, ch. 5, 2b, in Ackrill (1963). 12. Aristotle Categoriae, ch. 10, 12a, in Ackrill (1963). 13. Mill (1868:188). Mill’s work in this area is discussed by Horn (1989), who refers to a number of other nineteenth and twentieth century logicians who made similar observations. 14. Mill (1868: 210, original emphasis). 15. Ibid.

1403_902976_11_note.qxd 12/13/04 6:33 PM Page 223

Notes 223 16. ‘PG’s incomplete Phil and Ordinary Language paper’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 17. Grice (1975a: 45–6). 18. Moore (1942: 541, original emphasis). 19. Bar-Hillel (1946: 334). 20. Bar-Hillel (1946: 338, original emphasis). 21. O’Connor (1948: 359). 22. Urmson (1952) in Caton (1963: 224). 23. Urmson (1952) in Caton (1963: 229). 24. Nowell-Smith (1954: 81–2). 25. Edwards (1955: 21). 26. Grant (1958: 320). 27. Hungerland (1960: 212). 28. Hungerland (1960: 224). 29. Strawson (1952: 178–9). 30. Grice (1961: 121). 31. Grice (1961: 152). 32. Grice (1961: 124). 33. Grice (1961: 125). 34. Grice (1961: 126). 35. Grice (1961: 132). 36. Warnock (1967: 5). 37. ‘Logic and conversation (notes 1964)’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 38. Aristotle Categoriae, ch. 4, 1b, in Ackrill (1963). 39. Kant (1998: 212). 40. Kant (1998: 213). 41. Grice (1967a: 4). 42. Grice (1967a: 21). 43. Grice (1975a: 45). 44. Grice (1975a: 48). 45. Grice (1975a: 49). 46. Many of my students have pointed out to me that A could equally well understand B as implicating that Smith is just too busy, with all the visits he has to make to New York, for a social life at the moment. Green (1989: 91) argues that many different particularised conversational implicatures are possible in this example. 47. Grice (1978) in Grice (1989: 47). 48. Ryle (1971: vol. II: vii). 49. Benjamin (1956: 318). 50. Horn (1989), ch. 4. See also Levinson (1983), ch. 3 and McCawley (1981), ch. 8. This issue will be discussed in the final chapter. 51. Martinich (1996: 178). 52. Strawson (1952: 36). 53. Strawson (1952: 83). 54. Strawson (1952: 83, 86, 87n, 91). 55. Strawson (1952: 36, 37, 89, 85). 56. Grice (1967b: 58).

1403_902976_11_note.qxd 12/13/04 6:33 PM Page 224

224 Notes 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73.

Grice (1967b: 60). Grice (1967b: 59). Grice (1967b: 77). Grice (1969a) in Grice (1989: 88). Grice (1969a) in Grice (1989: 93, 94, 95). Grice (1969a) in Grice (1989: 99). Grice (1968) in Grice (1989: 117). Grice (1968) in Grice (1989: 123, original emphasis). Grice (1968) in Grice (1989: 127). Grice (1968) in Grice (1989: 136, original emphasis). Grice (1967c: 139, original emphasis). Black (1973: 268). Grice (1987b: 349). Avramides (1989: 39). See also Avramides (1997). Loar (2001: 104). Strawson (1969) in Strawson (1971: 178). Chomsky (1976: 75).

6 American formalism 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.

Grice (1986a: 60). All details from Kathleen Grice, personal communication. Strawson (1956: 110, original emphasis). Grice (1969b: 118). ‘Reference and presupposition’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Grice (1969b: 119). Lectures on Peirce, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Grice (1969b: 143). Grice (1969b: 142). Later published in Chomsky (1972a). Notes, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Lakoff (1989: 955). See, for instance, Sadock (1974). See, for instance, Gordon and Lakoff (1975). ‘Notes for syntax and semantics’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. I am grateful to Robin Lakoff for a very valuable discussion of the ideas in this paragraph. ‘Urbana seminars’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Lakoff (1995: 194–5). Simpson (1997: 151). Carnap (1937: 2). Lecture 1, ‘Lectures on language and reality’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. The

1403_902976_11_note.qxd 12/13/04 6:33 PM Page 225

Notes 225

22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40.

spelling ‘carulize’ may be Grice’s own, a result of his not referring directly to Carnap, or may be simply an error of transcription; the typescript from which this quotation was taken was made from a recording of the lectures. Lecture 1, ‘Lectures on language and reality’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. ‘Urbana seminars’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Grice (1981: 189). Ibid. Grice (1981: 190). ‘Philosophy at Oxford 1945–1970’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Harman (1986: 368). Grice (1971: 265). Grice (1971: 273). Kenny (1963: 207). Kenny (1963: 224). Kenny (1963: 229). Davidson (1970) in Davidson (1980: 37–8). ‘Probability, desirability and mood operators’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Davidson (1980: vii). Davidson (1978) in Davidson (1980: 83). Davidson (1978) in Davidson (1980: 95). ‘Reply to Davidson on “Intending” ’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Cohen (1971: 66).

7 Philosophical psychology 1. Grice (1975a) and Grice (1978). 2. For instance, although perhaps only semi-seriously, by Quine (1985: 248). 3. ‘Odd notes: Urbana and non Urbana’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 4. For instance, the writer of Grice’s obituary in The Times, 30 August 1988. Also Robin Lakoff and Kathleen Grice (personal communication). 5. ‘Tapes’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 6. Cartwright (1986: 442). 7. ‘Knowledge and belief’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 8. Judith Baker, personal communication. 9. ‘Urbana Lectures’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 10. Grice (1986a: 62). 11. ‘Reflections on morals’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley.

1403_902976_11_note.qxd 12/13/04 6:33 PM Page 226

226 Notes 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43.

Judith Baker is currently editing ‘Reflections on morals’ for publication. Grice and Baker (1985). Grice (1987b: 339). Grice (1987b: 369). ‘Reasons 1966’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. ‘Practical reason’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. ‘Kant lectures 1977’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Grice (2001: 3). ‘John Locke lectures miscellaneous notes 1979’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Kathleen Grice, personal communication. Grice (2001: 17). It is cited, for instance, in Levinson (1983). Grice (2001: 37). Grice (2001: 41). Grice (ibid., original emphasis). Grice (2001: 50). Grice (2001: 88, original emphasis). Aristotle Nicomachean Ethics, book 1 1097, in Thomson (1953). Aristotle Nicomachean Ethics, book 1 1095, in Thomson (1953). Aristotle Nicomachean Ethics, book 1 1098, in Thomson (1953). Grice (2001: 134). ‘Reference 1966/7’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Aristotle Nicomachean Ethics, book 1 1097, in Thomson (1953). ‘Steve Wagner’s notes on 1973 seminar’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Grice (1975b) in Grice (2001: 131). See Grice (1975b) in Grice (2001: 134–8) for a much fuller account of this process. Grice (1975b) in Grice (2001: 144). Grice (1975b) in Grice (2001: 145). Locke in Perry (1975b: 38). See Chapter 2 of this book for fuller discussion. Warner (1986: 493). ‘Reflections on morals, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Kenny (1963: 1).

8 Metaphysics and value 1. Strawson quoted in Strawson and Wiggins (2001: 517). 2. Misc. K. notes, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 3. Grice (1986a: 61). 4. Ayer (1971: 142). 5. Foot (1972) in Foot (1978: 167).

1403_902976_11_note.qxd 12/13/04 6:33 PM Page 227

Notes 227 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33.

34.

35. 36. 37. 38.

39.

Mackie (1977: 35). Mackie (1977: 38). For example, Nowell-Smith (1954: 79). Grice (1991: 23). Warner in Grice (2001: xxxvii, n). ‘Odd notes Spring 1981’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. ‘Reflections on morals’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Grice (1991: 79). Grice (1991: 90). Grice (1991: 67, original emphasis). Grice (1982) in Grice (1989: 283). Grice (1982) in Grice (1989: 301). Grice (1982) in Grice (1989: 302). See Isard (1982), Cormack (1982). Bennett (1976: 206–10). Baker (1989: 510). Sbisa (2001: 204). Grandy (1989: 524). Kasher (1976: 210). ‘Prejudices and predilections’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Neale (2001: 139). ‘Richards paper and notes for “Reply” ’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. ‘Reflections on morals’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Grice (1986a: 58). Tape, ‘PG seminar I Metaphysics S ‘78’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. ‘Preliminary valediction’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Unsorted notes, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Tape, 29 January 1983, Grice in conversation with Richard Warner and Judith Baker, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. ‘Course descriptions and faculty information, fall quarter 1977’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Grice (1986a: 64). ‘Festschrift notes (miscellaneous)’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. ‘Recent work’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. ‘PG’s incomplete “Phil and ordinary language” paper’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Grice (1991: 70).

1403_902976_11_note.qxd 12/13/04 6:33 PM Page 228

228 Notes 40. ‘Miscellaneous notes ’84–’85’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 41. See Davidson (1967). 42. Grice (1986b: 3, original emphasis.). 43. ‘Notes with Judy’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 44. Richard Warner (personal communication). 45. Grice (1986b: 34). 46. Ibid. 47. Grice (1988b: 304). 48. Code (1986: 413). 49. Grice (1988a: 176). 50. Hume (1740) in Perry (1975b: 174). 51. Eddington (1935: 5–6). 52. ‘Notes on “vulgar” and “learned” ’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 53. ‘To be sorted’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 54. Notes, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 55. Grice (1991: 41). 56. ‘Notes for Grice/Warnock retrospective’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 57. Unsorted notes, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 58. ‘The Way of Words, Studies In: Notes, offprints and draft material’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 59. Grice (1987b: 384–5). 60. ‘Copy of Faculty Research Grant application 1987–8’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 61. ‘Folder’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 62. Lindsay Waters quoted in The Times Higher Education Supplement, June 19, 1998. 63. ‘Notes dictated to Richard Warner on perception’, H. P. Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. 64. Strawson (1990: 153). 65. Burge (1992: 621). 66. Travis (1991: 237). 67. Fogelin (1991: 213). 68. Barnes (1993: 366). 69. Martinich (2002: 273). 70. Strawson and Wiggins (2001: 527–8).

1403_902976_11_note.qxd 12/13/04 6:33 PM Page 229

Notes 229

9 Gricean pragmatics 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46.

Neale (1992: 509). Levinson (1983: 118). Cohen (1971: 68n). Cole and Morgan (eds) (1975: xii). Kasher, cited in Mey (1989: 825). Fauconnier (1990: 391). Katz and Bever (1976: 49). Langedoen and Bever (1976: 253). Chomsky (1972b: 113). Lakoff (1989: 981). Parrett et al. (eds) (1981). Cogan and Herrmann (1975: 60). Karttunen and Peters (1975). Attardo (1990), Carey (1990), Elster (1990) and Rundquist (1990). Thomas in Mey (ed.) (1998: 170). Mey (2001: 911). Lindblom (2001: 1602). van Dijk (1976: 44). Pratt (1977: 215). Gilbert (1995), Gautam and Sharma (1986), Herman (1994). Cole (1975: 273). Traugott (1989: 50). Ziegeler (2000). McHoul in Mey (ed.) (1998: 227–8). Hymes (1986: 49). Harris (1995: 118). Pope (1995: 129, original emphasis). Sampson (1982: 203, original emphasis). Brown and Levinson (1987: 95). Lakoff (1995: 191). See, for instance, Sacks, Schegloff and Jefferson (1974), Sacks (1992). ‘Logic and conversation (notes 1964)’, H.P.Grice Papers, BANC MSS 90/135 c, The Bancroft Library, University of California, Berkeley. Holdcroft (1979: 131). Bird (1979). Fairclough (1985: 757). Sarangi and Slembrouck (1992). Gumperz (1990: 439). Hanks (2001: 208). Smith and Tsimpli (1995: 69, 74). Moscovitch (1983), Gardner et al. (1983: 173), Caplan (1987: 358). Ahlsén (1993: 61). For instance, Prutting and Kirchner (1987). Bloom et al. (1999: 554). Kasher et al. (1999: 586). Borod et al. (2000: 118). Kasher (1991: 139).

1403_902976_11_note.qxd 12/13/04 6:33 PM Page 230

230 Notes 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95.

Saygina and Ciceki (2002: 252). Dale and Reiter (1995: 234). Dale and Reiter (1995: 262). Young (1999: 216). Grice (1975a: 53). Grice (1978) in Grice (1989: 53). Grice (1975a: 53). For example, Martinich (1984). Harnish (1976: 346, 347). Sperber and Wilson (1995: 201). Davis (1998: 65). Giora (1999: 919). Gibbs (1999: 468). Holtgraves (1998: 21), Holtgraves (1999: 520). Winer et al. (2001: 486). Bezuidenhout and Cooper Cutting (2002: 443). Keenan (1976: 70). Keenan (1976: 79). Ibid. Green (1989: 95n). Harnish (1976: 340n). Leech (1983: 80). Brown and Levinson (1987: 49). Grice (1975a: 47). Lakoff (1989: 960). Lakoff (1972: 926n). Lakoff (1972: 916). Lakoff (1973: 296). Lakoff (1973: 298). Ibid. Brown and Levinson (1978, 1987). Brown and Levinson (1987: 4). Goffman (1955) reprinted in Laver and Hutcheson (1972: 323). Brown and Levinson (1987: 94). Brown and Levinson (1987: 95). Brown and Levinson (1987:164). Mura (1983: 115). Matsumoto (1989: 207). Leech (1983: xi). Leech (1983: 81). Leech (1983: 141, 144, 147). Brown and Levinson (1987: 4). Haberlard and Mey (2002: 1674). Sperber and Wilson (1995: 38). Sperber and Wilson (1995: 36). Sperber and Wilson (1995: 162). Sperber and Wilson (1995: 24). Blakemore (1990: 364). Grice (1987b: 372).

1403_902976_11_note.qxd 12/13/04 6:33 PM Page 231

Notes 231 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120.

Horn (2000: 321). Horn (1989: 194). Horn (1989: 204). Grice (1975a: 56–7). Horn (1972: (41). Horn (1989: 266–7). Horn (1989: 196). Noveck (2001: 183). Horn (2000). Sadock (1981: 265). Sadock (1981: 265n). Gazdar (1979: xi, 55). Levinson (1983: 101). Atlas and Levinson (1981: 11). Atlas and Levinson (1981: 56). Levinson (2000: 6). Levinson (2000: 23). Levinson (2000: 164). Levinson (2000: 217). Richardson and Richardson (1990: 500). Bach (2001: 18). Bach (2001: 30). Ariel (2002: 1003). Saul (2002: 365). Grice (1986a: 106).

1403_902976_12_bibl.qxd 12/13/04 6:33 PM Page 232

Bibliography Ackrill, J. L. (1963) (trans.) Aristotle’s Categories and De Interpretatione, Oxford: Clarendon Press. Ahlsén, Elisabeth (1993) ‘Conversational principles and aphasic communication’, Journal of Pragmatics 19.1: 57–70. Ariel, Mira (2002) ‘Privileged interactional interpretations’, Journal of Pragmatics 34.8: 1003–44. Atlas, Jay and Stephen Levinson (1981) ‘It-clefts, informativeness, and logical form: radical pragmatics (revised standard version)’, in Peter Cole (ed.) (1981): 1–61. Attardo, Salvatore (1990) ‘The violation of Grice’s maxims in jokes’, in Kira Hall et al. (eds) (1990): 355–62. Austin, J. L. (1950) ‘Truth’, Proceedings of the Aristotelian Society. Reprinted in J. L. Austin (1961): 85–101. Austin, J. L. (1956) ‘A plea for excuses’, Proceedings of the Aristotelian Society. Reprinted in J. L. Austin (1961): 123–52. Austin, J. L. (1961) Philosophical Papers, Oxford: Clarendon Press. Austin, J. L. (1962a) Sense and Sensibilia, Oxford: Clarendon Press. Austin, J. L. (1962b) How to do Things with Words, Oxford: Clarendon Press. Avramides, Anita (1989) Meaning and Mind, Cambridge, MA: MIT Press. Avramides, Anita (1997) ‘Intention and convention’, in Bob Hale and Crispin Wright (eds) (1997) A Companion to the Philosophy of Language, Oxford: Blackwell: 60–86. Ayer, A. J. (1964) Foundations of Empirical Knowledge, London: Macmillan [1st edn 1940.] Ayer, A. J. (1971) Language, Truth and Logic, Harmondsworth: Penguin [1st edn 1936; 2nd edn 1946.] Bach, Kent (2001) ‘You don’t say?’, Syntheses 128: 15–44. Baker, Gordon (1986) ‘Alternative mind-styles’, in Richard Grandy and Richard Warner (eds) (1986b): 277–314. Baker, Judith (1989) ‘The metaphysical construction of value’, The Journal of Philosophy 86.10: 505–13. Bar-Hillel, Yehoshua (1946) ‘Analysis of “correct” language’, Mind 55: 328–40. Barnes, Gerald (1993) ‘Review of The Conception of Value by Paul Grice’, Mind 102: 366–70. Benjamin, B. S. (1956) ‘Remembering’, Mind 65: 312–31. Bennett, Jonathan (1976) Linguistic Behaviour, Cambridge: Cambridge University Press. Berlin, Isaiah et al. (1973) Essays on J. L. Austin, Oxford: Clarendon Press. Bever, Thomas G. et al. (eds) (1977) An Integrated Theory of Linguistic Ability, Hassocks: The Harvester Press. [First published in the USA by Thomas Y. Cromwell, 1976.] Bezuidenhout, Anne and J. Cooper Cutting (2002) ‘Literal meaning, minimal propositions, and pragmatic processing’, Journal of Pragmatics 34.4: 433–56. 232

1403_902976_12_bibl.qxd 12/13/04 6:33 PM Page 233

Bibliography 233 Bird, Graham (1979) ‘Speech acts and conversation – II’, Philosophical Quarterly 29.115: 142–52. Black, Max (1973) ‘Meaning and intention: an examination of Grice’s views’, New Literary History 4: 257–79. Blakemore, Diane (1990) ‘Constraints on interpretations’, in Kira Hall et al. (eds) (1990): 363–70. Bloom, Ronald et al. (1999) ‘Psychometric aspects of verbal pragmatic ratings’, Brain and Language 68: 553–65. Borod, Joan C. et al. (2000) ‘Verbal pragmatics following unilateral stroke: emotional content and valence’, Neuropsychology 14.1: 112–24. Brown, Gillian and George Yule (1983) Discourse Analysis, Cambridge: Cambridge University Press. Brown, Penelope and Stephen Levinson (1978) ‘Universals in language usage: politeness phenomena’, in E. N. Goody (ed.) (1978) Questions and Politeness: Strategies in Social Interaction, Cambridge: Cambridge University Press: 56–289. Brown, Penelope and Stephen Levinson (1987) Politeness: Some Universals in Language Use, Cambridge: Cambridge University Press. Burge, Tyler (1992) ‘Review of Studies in the Way of Words by Paul Grice’, The Philosophical Review 101: 619–21. Caplan, David (1987) Neurolinguistics and Linguistic Aphasiology, Cambridge: Cambridge University Press. Carey, Kathleen (1990) ‘The role of conversational implicature in the early grammaticalization of the English perfect’, in Kira Hall et al. (eds) (1990): 371–80. Carnap, Rudolf (1937) The Logical Syntax of Language, London: Routledge & Kegan Paul. Cartwright, Nancy (1986) ‘Fitting facts of equations’, in Richard Grandy and Richard Warner (eds) (1986b): 441–53. Caton, Charles (1963) Philosophy and Ordinary Language, Urbana: University of Illinois Press. Chomsky, Noam (1957) Syntactic Structures, The Hague: Mouton. Chomsky, Noam (1972a) Studies on Semantics in Generative Grammar, The Hague: Mouton. Chomsky, Noam (1972b) ‘Some empirical issues in the theory of transformational grammar’, in Stanley Peters (ed.) (1972) Goals of Linguistic Theory, Englewood Cliffs, NJ: Prentice-Hall: 63–127. Chomsky, Noam (1976) Reflections on Language, London: Temple Smith. Code, Alan (1986) ‘Aristotle: essence and accident’, in Richard Grandy and Richard Warner (eds) (1986b): 411–39. Cogan, Cathy and Leora Herrmann (1975) ‘Interactions of the expression “let’s just say” with the Gricean maxims of conversation’, Proceedings of the 1st Annual Meeting of the Berkeley Linguistics Society: 60–8. Cohen, L. Jonathan (1971) ‘Some remarks on Grice’s views about the logical particles of natural language’, in Yehoshua Bar-Hillel (ed.) (1971) Pragmatics of Natural Languages, Dordrecht-Holland: D. Reidel Publishing: 50–68. Cole, Peter (1975) ‘The synchronic and diachronic status of conversational implicature’, in Peter Cole and Jerry L. Morgan (eds) (1975): 257–88. Cole, Peter (ed.) (1981) Radical Pragmatics, New York: Academic Press. Cole, Peter and Jerry L. Morgan (eds) (1975) Syntax and Semantics 3: Speech Acts, New York: Academic Press.

1403_902976_12_bibl.qxd 12/13/04 6:33 PM Page 234

234 Bibliography Cormack, Annabel (1982) ‘Comments on Grice’s paper’, in Neil Smith (ed.) (1982): 251–7. Cosenza, Giovanna (2001a) ‘Some limits and possibilities of Grice’s account of meaning and communication’, in Giovanna Cosenza (ed.) (2001b): 7–32. Cosenza, Giovanna (ed.) (2001b) Paul Grice’s Heritage, Turnhout: Brepols. Coulthard, Malcolm (1985) An Introduction to Discourse Analysis (2nd edn), London: Longman. [1st edn 1977.] Dale, Robert and Ehud Reiter (1995) ‘Computational interpretations of the Gricean maxims in the generation of referring expressions’, Cognitive Science 18: 233–63. Davidson, Donald (1967) ‘The logical form of action sentences’, in Nicholas Rescher (ed.) (1967) The Logic of Decision and Action, Pittsburgh: University of Pittsburgh Press. Reprinted in Donald Davidson (1980): 105–22. Davidson, Donald (1970) ‘How is weakness of the will possible?’, in Joel Feinberg (ed.) (1990) Moral Concepts, Oxford: Oxford University Press. Reprinted in Donald Davidson (1980): 21–42. Davidson, Donald (1978) ‘Intending’, in Yirmiaku Yovel (ed.) (1978) Philosophy of History and Action, The Hebrew University: D. Reidel and The Magnes Press. Reprinted in Donald Davidson (1980): 83–102. Davidson, Donald (1980) Essays on Actions and Events, Oxford: Clarendon Press. Davis, Wayne (1998) Implicature, Cambridge: Cambridge University Press. Eddington, Arthur (1935) The Nature of the Physical World, London: J. M. Dent and Sons. [1st edn 1928]. Edwards, Paul (1955) The Logic of Moral Discourse, Glencoe, IL: The Free Press. Elster, Charles (1990) ‘Gricean maxims and reading instruction’, in Kira Hall et al. (eds) (1990): 381–9. Fairclough, Norman (1985) ‘Critical and descriptive goals in discourse analysis’, Journal of Pragmatics 9: 739–63. Fann, K. T. (ed.) (1969) Symposium on J. L. Austin, London: Routledge. Fauconnier, Gilles (1990) ‘Invisible meaning’, in Kira Hall et al. (eds) (1990): 390–404. Fogelin, Robert (1991) ‘Review of Studies in the Way of Words, Paul Grice’, The Journal of Philosophy 88: 213–19. Foot, Philippa (1972) ‘Morality as a system of hypothetical imperatives’, The Philosophical Review 81.3. Reprinted in Philippa Foot (1978) Virtues and Vices and Other Essays in Moral Philosophy, Oxford: Blackwell: 157–73. Forguson, Lynd (2001) ‘Oxford and the “epidemic” of ordinary language philosophy’, The Monist 84.3: 325–45. Frege, Gottlob (1892) ‘On sense and meaning’, in Peter Geach and Max Black (eds) (1980) Translations from the Philosophical Writings of Gottlob Frege, Oxford: Blackwell [1st edn 1952]: 56–78. Gallie, Ian (1936) ‘Is the self a substance?’, Mind 45: 28–44. Gardner, Howard et al. (1983) ‘Missing the point: the role of the right hemisphere in the processing of complex linguistic materials’, in Ellen Perecman (ed.) (1983): 169–92. Gautam, Kripa K. and Manjula Sharma (1986) ‘Dialogue in Waiting for Godot and Grice’s conception of implicature’, Modern Drama 29: 580–6. Gazdar, Gerald (1979) Pragmatics, New York: Academic Press.

1403_902976_12_bibl.qxd 12/13/04 6:33 PM Page 235

Bibliography 235 Gellner, Ernest (1979) Words and Things, London: Routledge & Kegan Paul. [1st edn 1959.] Gibbs, Raymond (1999) ‘Interpreting what speakers say and implicate’, Brain and Language 68: 466–85. Gilbert, Anthony J. (1995) ‘Shakespearean self-talk, the Gricean maxims and the unconscious’, English Studies 76: 221–37. Giora, Rachel (1999) ‘On the priority of salient meanings: studies of literal and figurative languages’, Journal of Pragmatics 31.7: 919–29. Goffman, Erving (1955) ‘On face-work: an analysis of ritual elements in social interaction’, Psychiatry 18: 213–31. Reprinted in John Laver and Sandy Hutcheson (eds) (1972) Communication in Face to Face Interaction, Harmondsworth: Penguin: 319–46. Gordon, David and George Lakoff (1975) ‘Conversational postulates’, in Peter Cole and Jerry L. Morgan (eds) (1975): 83–106. Graddol, David, Jenny Cheshire and Joan Swann (1994) Describing Language (2nd edn), Buckingham: Open University Press. Grandy, Richard (1989) ‘On Grice on language’, The Journal of Philosophy 86.10: 514–25. Grandy, Richard and Richard Warner (1986a) ‘Paul Grice: a view of his work’, in Richard Grandy and Richard Warner (eds) (1986b): 1–44. Grandy, Richard and Richard Warner (eds) (1986b) Philosophical Grounds of Rationality, Oxford: Clarendon Press. Grant, C. K. (1958) ‘Pragmatic implication’, Philosophy XXXIII: 303–24. Green, Georgia M. (1989) Pragmatics and Natural Language Understanding, New Jersey: Lawrence Erlbaum Associates. Grice, Paul (1941) ‘Personal identity’, Mind 50: 330–50. Reprinted in John Perry (ed.) (1975): 73–95. Grice, Paul (c. 1946–1950) ‘Common sense and skepticism’, in Paul Grice (1989): 147–53. Grice, Paul (c. 1953–1985) ‘G. E. Moore and the philosopher’s paradoxes’, in Paul Grice (1989): 154–70. Grice, Paul (1957) ‘Meaning’, The Philosophical Review 66: 377–88. Reprinted in Paul Grice (1989): 213–23. Grice, Paul (1958) ‘Postwar Oxford philosophy’, in Paul Grice (1989): 171–80. Grice, Paul (1961) ‘The causal theory of perception’, Proceedings of the Aristotelian Society, Supplementary Volume 35: 121–52. Reprinted in Robert Swartz (ed.) (1965): 438–72. Reprinted in G. J. Warnock (ed.) (1967): 85–112. Reprinted (abridged) in Paul Grice (1989): 224–47. Grice, Paul (1962) ‘Some remarks about the senses’, in R. J. Butler (ed.) Analytical Philosophy, Oxford: Blackwell. Reprinted in Paul Grice (1989): 248–68. Grice, Paul (1966) ‘Descartes on clear and distinct perception’, in Paul Grice (1989): 186–95. Grice, Paul (1967a) ‘Prolegomena’, in Paul Grice (1989): 3–21. Grice, Paul (1967b) ‘Indicative conditionals’, in Paul Grice (1989): 58–85. Grice, Paul (1967c) ‘Some models for implicature’, in Paul Grice (1989): 138–43. Grice, Paul (1968) ‘Utterer’s meaning, sentence meaning, and word-meaning’, Foundations of Language 4: 225–42. Reprinted in Paul Grice (1989): 117–37. Grice, Paul (1969a) ‘Utterer’s meaning and intentions’, Philosophical Review 78: 147–77. Reprinted in Paul Grice (1989): 86–116.

1403_902976_12_bibl.qxd 12/13/04 6:33 PM Page 236

236 Bibliography Grice, Paul (1969b) ‘Vacuous names’, in Donald Davidson and Jakko Hintikka (eds) (1969) Words and Objections, Dordrecht: D. Reidel Publishing: 118–45. Grice, Paul (1971) ‘Intention and uncertainty’, Proceedings of the British Academy 57: 263–79. Grice, Paul (1975a) ‘Logic and conversation’, in Peter Cole and J. L. Morgan (eds) (1975): 41–58. Also published in Donald Davidson and Gilbert Harman (eds) (1975) The Logic of Grammar, Encino, Dickenson: 64–75. Reprinted in Paul Grice (1989): 22–40. Grice, Paul (1975b) ‘Method in philosophical psychology (from the banal to the bizarre)’, Proceedings and Addresses of the American Philosophical Association 48: 23–53. Reprinted in Paul Grice (1991): 121–61. Grice, Paul (1978) ‘Further notes on logic and conversation’, in Peter Cole (ed.) (1978) Syntax and Semantics 9: Pragmatics, Academic Press: New York. Reprinted in Paul Grice (1989): 41–57. Grice, Paul (1981) ‘Presupposition and conversational implicature’, in Peter Cole (ed.) Radical Pragmatics, New York: Academic Press. Reprinted in Paul Grice (1989): 269–82. Grice, Paul (1982) ‘Meaning revisited’, in Neil Smith (ed.) (1982): 223–43. Reprinted in Paul Grice (1989): 283–303. Grice, Paul (1986a) ‘Reply to Richards’, in Richard Grandy and Richard Warner (eds) (1986b): 45–106. Grice, Paul (1986b) ‘Actions and events’, Pacific Philosophical Quarterly 67: 1–35. Grice, Paul (1987a) ‘Conceptual analysis and the province of philosophy’, in Paul Grice (1989): 181–5. Grice, Paul (1987b) ‘Retrospective epilogue’, in Paul Grice (1989): 339–85. Grice, Paul (1988a) ‘Aristotle on the multiplicity of being’, Pacific Philosophical Quarterly 69: 175–200. Grice, Paul (1988b) ‘Metaphysics, phiosophical eschatology, and Plato’s Republic’, in Paul Grice (1989): 304–38. Grice, Paul (1989) Studies in the Way of Words, Harvard: Harvard University Press. Grice, Paul (1991) The Conception of Value, Oxford: Clarendon Press. Grice, Paul (2001) Aspects of Reason, Oxford: Clarendon Press. Grice, Paul and Judith Baker (1985) ‘Davidson on “Weakness of the will” ’, in Bruce Vermazen and Merrill Hintikka (eds) (1985) Essays on Davidson: Actions and Events, Oxford: Clarendon Press: 27–49. Grice, Paul and Peter Strawson, (1956) ‘In defence of a dogma’, The Philosophical Review 65: 141–58. Reprinted in Paul Grice (1989): 196–212. Gumperz, John (1982) Discourse Strategies, Cambridge: Cambridge University Press. Gumperz, John (1990) ‘Conversational cooperation in social perspective’, in Kira Hall et al. (eds) (1990): 429–41. Haberlard, Hartmut and Jacob Mey (2002) ‘Linguistics and pragmatics, 25 years after’, Journal of Pragmatics 34: 1671–82. Hall, Kira et al. (eds) (1990) Proceedings of the Sixteenth Annual Meeting of the Berkeley Linguistics Society: General Session and Parasession on the Legacy of Grice, Berkeley: Berkeley Linguistics Society. Hampshire, Stuart and H. L. A. Hart (1958) ‘Decision, intention and certainty’, Mind 67: 1–12.

1403_902976_12_bibl.qxd 12/13/04 6:33 PM Page 237

Bibliography 237 Hanks, William (2001) ‘Exemplary natives and what they know’, in Giovanna Cosenza (ed.) (2001b): 207–34. Harman, Gilbert (1986) ‘Willing and intending’, in Richard Grandy and Richard Warner (eds) (1986b): 363–80. Harnish, Robert (1976) ‘Logical form and implicature’, in Thomas Bever et al. (eds) (1977): 313–91. Harris, Sandra (1995) ‘Pragmatics and power’, Journal of Pragmatics 23: 117–35. Herman, David (1994) ‘The Mutt and Jute dialogue in Joyce’s Finnegans Wake: some Gricean perspectives’, Style 86: 505–13. Holdcroft, David (1979) ‘Speech acts and conversation – II’, The Philosophical Quarterly 29.115: 125–41. Holtgraves, Thomas (1998) ‘Interpreting indirect replies’, Cognitive Psychology 37: 1–27. Holtgraves, Thomas (1999) ‘Comprehending indirect replies: when and how are their conveyed meanings activated?’, Journal of Memory and Language 41: 519–40. Horn, Laurence (1972) On the Semantic Properties of Logical Operators in English, PhD thesis, University of California, Los Angeles. Horn, Laurence (1989) A Natural History of Negation, Chicago: University of Chicago Press. Horn, Laurence (2000) ‘From if to iff: conditional perfection as pragmatic strengthening’, Journal of Pragmatics 32.3: 289–329. Hume, David (1740) ‘Appendix’ to Book III, Treatise of Human Nature. Reprinted as ‘Second thoughts’ in John Perry (ed.) (1975): 173–6. Hungerland, Isabel (1960) ‘Contextual implication’, Inquiry 3: 211–58. Hymes, Dell (1986) ‘Discourse: scope without depth’, International Journal of the Sociology of Language 57: 49–89. Isard, S. D. (1982) ‘Comments on Grice’s paper’, in Neil Smith (ed.) (1982): 246–50. Kant, Immanuel (1998) Critique of Pure Reason, trans. and ed. Paul Guyer and Allen Wood, Cambridge: Cambridge University Press [1st German edn 1781.] Karttunen, Lauri and Stanley Peters (1975) ‘Conventional implicature in Montague Grammar’, Proceedings of the 1st Annual Meeting of the Berkeley Linguistics Society: 266–78. Kasher, Asa (1976) ‘Conversational maxims and rationality’, in Asa Kasher (ed.) (1976) Language in Focus: Foundations, Methods and Systems, Dordrecht: D. Reidel Publishing: 197–216. Kasher, Asa (1991) ‘Pragmatics and Chomsky’s research program’, in Asa Kasher (ed.) (1991) The Chomskyan Turn, Oxford: Blackwell: 122–49. Kasher, Asa et al. (1999) ‘Effects of right- and left-hemisphere damage on understanding conversational implicature’, Brain and Language 68: 566–90. Katz, Jerrold and Thomas G. Bever (1976) ‘The fall and rise of empiricism’, in Thomas G. Bever et al. (eds) (1977): 46–58. Keenan, Elinor Ochs (1976) ‘The universality of conversational postulates’, Language in Society 5: 67–80. Kenny, Anthony (1963) Action, Emotion and Will, London: Routledge & Kegan Paul. Lakoff, Robin (1972) ‘Language in context’, Language 48: 907–27.

1403_902976_12_bibl.qxd 12/13/04 6:33 PM Page 238

238 Bibliography Lakoff, Robin (1973) ‘The logic of politeness; or, minding your p’s and q’s’, in Papers from the Ninth Regional Meeting of the Chicago Linguistics Society, Chicago: Chicago Linguistics Society: 292–305. Lakoff, Robin (1989) ‘The way we were; or; the real actual truth about generative semantics: a memoir’, Journal of Pragmatics 13: 939–88. Lakoff, Robin (1995) ‘Conversational logic’, in Jef Verschueren et al. (eds) (1995) Handbook of Pragmatics, Amsterdam: John Benjamins: 190–8. Langendoen, D. Terence and Thomas G. Bever (1976) ‘Can a not unhappy person be called a not sad one?’, in Thomas G. Bever et al. (eds) (1977). Leech, Geoffrey (1983) Principles of Pragmatics, London: Longman. Levinson, Stephen (1983) Pragmatics, Cambridge: Cambridge University Press. Levinson, Stephen (2000) Presumptive Meanings, Cambridge, MA: The MIT Press. Lindblom, Kenneth (2001) ‘Cooperating with Grice: A cross-disciplinary metaperspective on uses of Grice’s Cooperative Principle’, Journal of Pragmatics 32: 1601–23. Loar, Brian (2001) ‘The supervenience of social meaning on speaker’s meaning’, in Giovanna Cosenza (ed.) (2001b): 101–13. Locke, John (1694) ‘Of identity and diversity’, in Essay Concerning Human Understanding (2nd edn). Reprinted in John Perry (ed.) (1975b): 33–52. Mackie, J. L. (1977) Ethics: Inventing Right and Wrong, Harmonsdsworth: Pelican Books. Magee, Bryan (1986) Modern British Philosophy, Oxford: Oxford University Press. [1st edn 1971.] Malcolm, Norman (1942) ‘Moore and ordinary language’ in Paul Schilpp (ed.) (1968): 345–68. Martin, G. H. and J. R. L. Highfield (1997) A History of Merton College, Oxford: Oxford University Press. Martinich, A. P. (1984) ‘A theory of metaphor’, Journal of Literary Semantics 13.1: 35–56. Martinich, A. P. (1996) Philosophical Writing (2nd edn), Oxford: Blackwell. [1st edn 1989.] Martinich, A. P. (2002) ‘Review of Paul Grice’s Aspects of Reason’, Journal of Philosophy XL.2: 273–4. Matsumoto, Yoshiko (1989) ‘Politeness and conversational universals: observations from Japanese’, Journal of Pragmatics 8.2/3: 207–21. McCawley, James (1981) Everything that Linguists have Always Wanted to Know about Logic, Oxford: Blackwell. Mey, Jacob (1989) ‘The end of the copper age, or: pragmatics 121/2 years after’, Journal of Pragmatics 13: 825–32. Mey, Jacob (ed.) (1998) The Concise Encyclopedia of Pragmatics, Amsterdam: Elsevier. Mey (2001) ‘To Grice or not to Grice’, Journal of Pragmatics 34: 911. Mill, John Stuart (1868) An Examination of Sir William Hamilton’s Philosophy, vol. 2, Boston, MA: William V. Spencer. Moore, G. E. (1925) ‘A defence of common sense’, in J. H. Muirhead (ed.) (1925) Contemporary British Philosophy, 2nd series. Reprinted in G. E. Moore (1959) Philosophical Papers, London: George Allen & Unwin: 32–59. Moore, G. E. (1942) ‘A reply to my critics’, in Paul Schilpp (ed.) (1968): 535–677.

1403_902976_12_bibl.qxd 12/13/04 6:33 PM Page 239

Bibliography 239 Moscovitch, M. (1983) ‘The linguistic and emotional functions of the normal right hemisphere’, in Ellen Perecman (ed.) (1983): 57–82. Mundle, C. W. K. (1979) A Critique of Linguistic Philosophy, 2nd edn, London: Glover and Blair. [1st edn 1970.] Mura, Susan Swan (1983) ‘Licensing violations: legitimate violations of Grice’s conversational principle’, in Robert T. Craig and Karen Tracy (eds) (1983) Conversational Coherence, Beverley Hills: Sage Publications: 101–15. Neale, Stephen (1992) ‘Paul Grice and the philosophy of language’, Linguistics and Philosophy 15: 509–59. Neale, Stephen (2001) ‘Implicature and colouring’, in Giovanna Cosenza (ed.) (2001b): 139–84. Noveck, Ira (2001) ‘When children are more logical than adults: experimental investigations of scalar implicature’, Cognition 78: 165–88. Nowell-Smith, P. H. (1954) Ethics, Harmondsworth: Penguin. O’Connor, D. J. (1948) ‘Pragmatic paradoxes’, Mind 57: 358–9. Parrett, Herman, Marina Sbisa and Jef Verschueren (eds) (1981) Possibilities and Limitations of Pragmatics, Amsterdam: John Benjamins. Peirce, C. S. (1867) ‘On a new list of categories’, in Nathan Houser and Christian Kloesel (eds) (1992) The Essential Peirce, vol. 1, Bloomington and Indianapolis: Indiana University Press: 1–10. Penman, Robyn (1987) ‘Discourse in courts: co-operation, coercion and coherence’, Discourse Processes 10.3: 201–18. Perecman, Ellen (ed.) (1983) Cognitive Processing in the Right Hemisphere, New York: Academic Press. Perner, Josef and Susan Leekam (1986) ‘Belief and quantity: three-year-olds’ adaption to listener’s knowledge’, Journal of Child Language 13: 305–15. Perry, John (1975a) ‘Personal identity, memory, and the problem of circularity’, in John Perry (ed.) (1975b): 135–55. Perry, John (ed.) (1975b) Personal Identity, Berkeley: University of California Press. Pitcher, George (1973) ‘Austin: a personal memoir’, in Isaiah Berlin et al. (1973): 17–30. Pope, Rob (1995) Textual Intervention, London: Routledge. Pratt, Mary Louise (1977) Toward a Speech Act Theory of Literary Discourse, Bloomington: Indiana University Press. Prutting, C. and D. Kirchner (1987) ‘A clinical appraisal of the pragmatic aspects of language’, Journal of Speech and Hearing Disorders 52: 105–9. Quine, W. V. O. (1953) ‘Two dogmas of empiricism’, in W. V. O. Quine (1953) From a Logical Point of View, Harvard: Harvard University Press: 20–46. Quine, W. V. O. (1985) The Time of My Life, Cambridge, MA: The MIT Press. Quinton, Anthony (1963) ‘The a priori and the analytic’, Proceedings of the Aristotelian Society 64: 31–54. Reprinted in Peter Strawson (ed.) (1967) Philosophical Logic, Oxford: Oxford University Press: 107–28. Raskin, Victor (1985) Semantic Mechanisms of Humour, Dordrecht: D. Reidel Publishing. Rée, Jonathan (1993) ‘English philosophy in the fifties’, Radical Philosophy 65: 3–21. Reid, Thomas (1785) ‘Of Mr Locke’s account of our personal identity’, from Essays on the Intellectual Powers of Man. Reprinted in John Perry (ed.) (1975b): 113–18.

1403_902976_12_bibl.qxd 12/13/04 6:33 PM Page 240

240 Bibliography Richardson, John and Alan Richardson (1990) ‘On predicting pragmatic relations’, in Kira Hall et al. (eds) (1990): 498–508. Rogers, Ben (2000) A. J. Ayer A Life, London: Vintage. Rundquist, Suellen (1990) ‘Indirectness in conversation: Flouting Grice’s maxims at dinner’, in Kira Hall et al. (eds) (1990): 509–18. Rundquist, Suellen (1992) ‘Indirectness: a gender study of flouting Grice’s maxims’, Journal of Pragmatics 18: 431–49. Russell, Bertrand (1905) ‘On denoting’, Mind 14: 479–99. Russell, Bertrand (1956) ‘The cult of “common usage” ’, in Robert Egner and Lester Denonn (eds) (1961) The Basic Writings of Bertrand Russell, London and New York: Routledge. Ryle, Gilbert (1949) The Concept of Mind, London: Hutchinson. Ryle, Gilbert (ed.) (1956) The Revolution in Philosophy, London: Macmillan. Ryle, Gilbert (1971) Collected Papers, London: Hutchinson. Sacks, Harvey, E. A. Schegloff and G. Jefferson, (1974) ‘A simplest systematics for the organization of turn-taking for conversation’, Language 50: 696–735. Sacks, Harvey (1992) Lectures on Conversation (G. Jefferson ed.) 2 vols, Oxford: Blackwell. Sadock, Jerold (1974) Towards a Linguistic Theory of Speech Acts, New York: Academic Press. Sadock, Jerold (1981) ‘Almost’, in Peter Cole (ed.) (1981): 257–71. Sampson, Geoffrey (1982) ‘The economics of conversation: comments on Joshi’s paper’, in Neil Smith (ed.) (1982): 200–10. Sarangi, Srikant and Stefaan Slembrouck (1992) ‘Non-cooperation in communication: a reassessment of Gricean pragmatics’, Journal of Pragmatics 17: 117–54. Saul, Jennifer (2002) ‘What is said and psychological reality: Grice’s project and relevance theorists’ criticisms’, Linguistics and Philosophy 25: 347–72. Saygina, Ayse Pinar and Ilyas Cicekli (2002) ‘Pragmatics in human-computer conversations’, Journal of Pragmatics 34.3: 227–58. Sbisa, Marina (2001) ‘Intentions from the other side’, in Giovanna Cosenza (ed.) (2001b): 185–206. Schiffer, Stephen (1972) Meaning, Oxford: Clarendon Press. Schiffer, Stephen (1987) Remnants of Meaning, Cambridge, MA: The MIT Press. Schiffrin, Deborah (1994) Approaches to Discourse, Oxford: Blackwell. Schilpp, Paul (ed.) (1968) The Philosophy of G. E. Moore, 3rd edn, La Salle, IL: Open Court. [1st edn 1942.] Searle, John (1969) Speech Acts, Cambridge: Cambridge University Press. Searle, John (1986) ‘Meaning, communication and representation’, in Richard Grandy and Richard Warner (eds) (1986b): 209–26. Simpson, Paul (1997) Language through Literature, London: Routledge. Smith, Neil (ed.) (1982) Mutual Knowledge, London: Academic Press. Smith, Neil and Ianthi-Maria Tsimpli (1995) The Mind of a Savant, Oxford: Blackwell. Sperber, Dan and Deirdre Wilson (1995) Relevance, 2nd edn, Oxford: Blackwell. [1st edn, 1986.] Stevenson, Charles (1944) Ethics and Language, New Haven: Yale University Press. Stout, G. F. (1896) ‘Voluntary action’, Mind 5: 354–66. Strawson, Peter (1950) ‘On referring’, Mind 59: 320–44. Strawson, Peter (1952) Introduction to Logical Theory, London: Methuen.

1403_902976_12_bibl.qxd 12/13/04 6:33 PM Page 241

Bibliography 241 Strawson, Peter (1956) ‘Construction and analysis’, in Gilbert Ryle (ed.) (1956): 97–110. Strawson, Peter (1959) Individuals, London: Methuen. Strawson, Peter (1964) ‘Intention and convention in speech acts’, The Philosophical Review LXXIII. Reprinted in Peter Strawson (1971): 149–69. Strawson, Peter (1969) ‘Meaning and truth’, Inaugural Lecture at the University of Oxford, Oxford University Press. Reprinted in Peter Strawson (1971): 170–89. Strawson, Peter (1971) Logico-Linguistic Papers, London: Methuen. Strawson, Peter (1990) ‘Review of Paul Grice, Studies in the Way of Words’, Synthese 84: 153–61. Strawson, Peter (1998) ‘Intellectual autobiography’, in Lewis Edwin Hahn (ed.) The Philosophy of P. F. Strawson, Chicago, IL: Open Court. Strawson, Peter and David Wiggins (2001) ‘Herbert Paul Grice 1913–1988’, Proceedings of the British Academy 111: 515–28. Swartz, Robert J. (ed.) (1965) Perceiving, Sensing and Knowing, Garden City, NY: Doubleday Anchor. Thomson, J. A. K. (trans.) (1953) Aristotle’s Ethics, London: Allen & Unwin. Traugott, Elizabeth (1989) ‘On the rise of epistemic meanings in English: an example of subjectification in semantic change’, Language 65.1: 31–55. Travis, Charles (1991) ‘Critical notice: annals of analysis’, Mind 100.2: 237–64. Urmson, J. O. (1952) ‘Parenthetical verbs’, Mind LXI: 480–96. Reprinted in Charles Caton (ed.) (1963): 220–40. Urmson, J. O., W. V. O. Quine and Stuart Hampshire (1965) ‘A symposium on Austin’s method’, 62nd annual meeting of the American Philosophical Association, Eastern Division. In K. T. Fann (ed.) (1969): 76–97. Van Dijk, Teun (1976) ‘Pragmatics and poetics’, in Teun van Dijk (ed.) (1976) Pragmatics of Language and Literature, Amsterdam: North-Holland Publishing: 23–57. Wardhaugh, Ronald (1986) An Introduction to Sociolinguistics, Oxford: Blackwell. Warner, Richard (1986) ‘Grice on happiness’, in Richard Grandy and Richard Warner (eds) (1986b): 475–93. Warnes, Martin (1990) ‘Philosophy, implicature and liturgy’, in David Jasper and R. C. D. Jasper (1990) Language and the Worship of the Church, London: Macmillan: 147–73. Warnock, G. L. (1955) ‘Seeing’, Proceedings of the Aristotelian Society 55: 201–18. Warnock, G. L. (1963) ‘John Langshaw Austin: a biographical sketch’, Proceedings of the British Academy 49. Reprinted in K. T. Fann (ed.) (1969): 3–21. Warnock, G. L. (ed.) (1967) The Philosophy of Perception, Oxford: Oxford University Press. Warnock, G. L. (1973) ‘Saturday mornings’, in Isaiah Berlin et al. (1973): 31–45. Williams, Bernard and Alan Montefiore (eds) (1966) British Analytical Philosophy, London: Routledge & Kegan Paul. Williamson, Timothy (1990) Identity and Discrimination, Oxford: Blackwell. Winer, Gerald et al. (2001) ‘Are children more accurate than adults? Spontaneous use of metaphor by children and adults’, Journal of Psycholinguistic Research 30.5: 485–96. Wittgenstein, Ludwig (1922) Tractatus Logico-Philosophicus, trans. C. K. Ogden, London: Routledge & Kegan Paul.

1403_902976_12_bibl.qxd 12/13/04 6:33 PM Page 242

242 Bibliography Wittgenstein, Ludwig (1958) Philosophical Investigations, trans. G. E. M. Anscombe, Oxford: Blackwell. [1st edn 1953.] Yamaguchi, Haruhiko (1988) ‘How to pull strings with words: deceptive violations in the garden-path joke’, Journal of Pragmatics 12.3: 323–37. Young, R. Michael (1999) ‘Using Grice’s maxim of Quantity to select the content of plan descriptions’, Artificial Intelligence 115: 215–56. Ziegeler, Debra (2000) ‘The role of quantity implicatures in the grammaticalisation of would’, Language Sciences 22.1: 27–61. Ziff, Paul (1967) ‘On H. P. Grice’s account of meaning’, Analysis 28: 1–8.

1403_902976_13_ind.qxd 12/13/04 6:34 PM Page 243

Index Ackrill, J. L., 47 Ahlsén, Elisabeth, 194 analytic/synthetic distinction, 16–17, 52, 161 Grice’s views on, 53–5, 85, 155, 181 aphasia, see clinical linguistics Ariel, Mira, 215 Aristotle, 8, 45–6, 47, 141, 142–3, 144, 178 Categories, 49–50, 87–8, 100 Nicomachean Ethics, 147–9, 150 Artificial Intelligence applications of Grice’s work in, 195–6 Grice’s views on, 169–70 Atlas, Jay, and Stephen Levinson, 209, 212, 213 Austin, John Langshaw, 47, 59, 98, 116, 122, 148, 178, 181, 198 career, 31–4 character, 33 death, 86 on the ‘descriptive fallacy’, 40, 75 philosophical development, 17, 19, 22, 28, 37 philosophical method and ‘Saturday Mornings’, 41–6, 55, 61–3, 145, 169 ‘A plea for excuses’, 40–1, 70, 89–90 Sense and Sensibilia, 39–49, 54–6, 94–6 speech acts, 74–9 Avramides, Anita, 82, 112 Ayer, A. J., 16–19, 22, 28, 31, 32, 36, 39, 44, 55–6, 95, 159 Bach, Kent, 214–15 Baker, Gordon, 1 Baker, Judith, 140, 141, 158, 166, 168, 173, 181, 183

Bancroft Library, H. P. Grice Papers, 7–8 Bar-Hillel, Yehoshua, 92 Barnes, Gerald, 183 Bealer, George, 181 behaviourism, 64–5, 67, 70, 113, 151 Benjamin, B. S., 105 Bennett, Jonathan, 74, 166 Berkeley, University of California, 86, 114, 115–16, 119, 138, 157 see also Bancroft Library Berkeley Linguistics Society, 188–9 Berlin, Isaiah, 17, 19 Bezuidenhout, Anne, and J. Cooper Cutting, 197 Bird, Graham, 193 Black, Max, 112 Blakemore, Diane, 207 Bloom, Ronald, and colleagues, 194 Borod, Joan, and colleagues, 195 Brown, Penelope, and Stephen Levinson, 191, 199, 202–4, 205 Burge, Tyler, 183 Carnap, Rudolf, 92, 116, 122–3 Cartwright, Nancy, 140 Chomsky, Noam, 86, 117, 119–20 Grice’s views on, 85–6, 114 views on Grice, 113, 187–8, 194 Clifton College, 11–12, 15 clinical linguistics, 194–5 Code, Alan, 176 Cogan, Cathy and Leora Herrmann, 188 Cohen, Jonathan, 135, 186 Cole, Peter, 186, 190 common sense, 4–5, 36–7, 52, 178–80 constructivism, 151–2, 155, 172–5, 176, 179 conventional implicature, 101, 102, 109–10, 188, 214 243

1403_902976_13_ind.qxd 12/13/04 6:34 PM Page 244

244 Index conversational implicature and disimplicature, 134–5 generalised, 104, 107–9, 206, 208–14 particularised, 103–4 philosophical background, 2–3, 87–97 and presupposition, 124–7 status of the theory of, 1, 3–4, 185–6, 188–91 universality of, 198–9 cooperation, 90–1, 98–9, 102–4, 166–7, Chapter 9 Cosenza, Giovanna, 82 creature-construction and meaning, 164 and perception, 179–80 and value, 149–54, 173, 175 critical discourse analysis, 193 Dale, Robert, and Ehud Reiter, 195 Davidson, Donald, 117, 186, 130–7, 146, 175 Davis, Wayne, 197 Descartes, René, 32, 59 discourse rights, 192–3 Eddington, Arthur, 177–8 Edwards, Paul, 93 ethics, 17, 141, 143, 154, 155–6, 158–61 see also value face, see politeness Fairclough, Norman, 193 Fauconnier, Gilles, 187 Fogelin, Robert, 183 Foot, Philippa, 159–60 Forguson, Lynd, 41 freedom, 160, 175–6 Frege, Gotlob, 34–5, 36, 48, 186 Gallie, Ian, 24, 25–6 Gazdar, Gerald, 212 Gellner, Ernest, 37, 44, 46 generative semantics, 119–21, 187, 200 Gibbs, Raymond, 197 Giora, Rachel, 197

Goffman, Erving, 203 goodness, see value Grandy, Richard, 1, 4, 166, 168 Grant, C. K., 93 Green, Georgia, 186, 199 Grice, Herbert Paul attitude to America, 143–4, 85, 115 appearance, 7, 59, 138 career, 14, 21, 30, 59, 114, 138, 157 character, 6–7, 139, 168 childhood, 10–11 death, 182 education, 11–14 family, 10–11, 59 ill health, 167–8, 170, 182 attitude to linguistics, 114, 119–21 marriage, 29 attitude to philosophy, 3, 5–6, 171–2 attitude to publishing, 3, 7–8, 157–8, 168, 180 teaching style, 97, 116, 139–40 attitude to technology, 169–70, 195 war service, 21 working method, 6, 8, 10, 49, 141, 162 Grice, Kathleen (née Watson), 6–7, 29–30, 42, 54, 59, 114, 115, 141, 144, 167, 170, 180, 182, 184 Gumperz, John, 193, 199 Haberlard, Hartmut, and Jacob Mey, 205 Hamilton, Sir William, 88 Hampshire, Stuart, 19, 41, 45, 62, 70, 99, 127 Hanks, William, 193 happiness, 148–9 Hardie, W. F. R., 13–14, 42, 48, 116, 148 Hare, R. M., 42 Harman, Gilbert, 127, 171, 186 Harnish, Robert, 196–7, 199 Harris, Sandra, 190 Hart, Herbert, 42, 70, 127 historical linguistics, 188, 189–90 Holdcroft, David, 193 Holtgraves, Thomas, 197 Horn, Laurence, 208–12

1403_902976_13_ind.qxd 12/13/04 6:34 PM Page 245

Index 245 Hume, David, 174, 177 Hungerland, Isobel, 93 Hymes, Dell, 190 intention, see meaning irony, 84, 194, 196–7, 204 Kant, Immanuel, 8, 49, 100, 141, 143, 144, 146, 155, 159, 161, 205 Karttunen, Lauri, and Stanley Peters, 188 Kasher, Asa, 166, 187, 194–5 Katz, Jerrold, and Thomas Bever, 187 Keenan, Elinor Ochs, 198–9 Kenny, Anthony, 129–39, 155–6 Lakoff, George, 120 Lakoff, Robin, 120, 121, 188, 191, 200–2, 203 Langendoen, D. Terence, and Thomas Bever, 187 Leech, Geoffrey, 199, 204–5 Levinson, Stephen, 186, 209, 212–14 see also Atlas, Jay, and Stephen Levinson; Brown, Penelope, and Stephen Levinson Lindblom, Kenneth, 189 linguistic philosophy, see ordinary language philosophy literature, 189 Loar, Brian, 112–13 Locke, John, 94, 95 on personal identity, 22–5, 27, 29, 150, 154, 163 logic, 35 Grice’s familiarity with, 117 indicative conditionals and, 105–9, 130 and natural language, 2, 48, 86–9, 104–5, 118–19, 122, 135, 144 logical positivism, 16–20, 28, 31, 32, 38, 40, 52, 70, 82, 90, 122, 177 Mackie, J. L., 159–61, 178–9 Malcolm, Norman, 83–4 Martinich, A. P., 105, 183–4 Matsumoto, Yoshiko, 204

maxims of conversation, 90–1, 99–100, 102, 107–8, chapter 9 development of, 92–4, 96 psychological status of, 102–3 McCawley, James, 120 McHoul, Alec, 190 meaning conventional v. speaker, 50, 62–3, 72–4, 78–80, 82–4, 90, 97, 100, 109–13, 118–19, 185–6 and creature construction, 164–6 and intention, 63, 66, 68–70, 73–82, 84, 109–12, 127–37, 165, 206 levels of, 74, 93, 102, 110, 208, 213–14 nonnatural (meaningNN), 64, 66, 72–3, 79, 82, 166 see also reference metaphor, 134–5, 189, 194, 196–7 ‘Metaphysical Transubstantiation’, 163, 164, 166, 173 Mey, Jacob, 189 see also Haberlard, Hartmut, and Jacob Mey Mill, John Stuart, 8, 88–9 mind, philosophy of, 18–19, 32 mood, see reasons Moore, G. E., 21, 34, 36–7, 83, 91–2, 178, 179, 181 Mundle, C. W. K., 45 Mura, Susan Swan, 204 Myro, George, 140, 181 Neale, Stephen, 168, 185 negation, 20, 48, 87, 124 neo-Griceans, 207–14 neuropragmatics, see clinical linguistics Noveck, Ira, 211 Nowell-Smith, P. H., 93, 160–1 Occam’s razor, modified, 5, 104–5, 118–19, 146 O’Conor, D. J., 92 ordinary language philosophy, 3, 4, 6, 13, 18, 57, 86, 160–1 criticisms of, 44–5

1403_902976_13_ind.qxd 12/13/04 6:34 PM Page 246

246 Index ordinary language philosophy – continued Grice’s attitude to, 45–7, 59, 60, 61–3, 72, 83–4, 85, 101, 116, 122–4, 128, 142, 145, 168, 175, 178–9 origins of, 33–41 see also Austin, John Langshaw Oxford, 12, 14, 15–16, 17, 31, 33, 59, 114–16, 143–4 All Souls’ College, 17, 19, 21 Christ Church College, 12, 19 Corpus Christi College, 11, 12, 14, 19, 21 Magdalen College, 12, 19 Merton College, 15, 19 St John’s College, 21, 29–30, 59, 158 see also ordinary language philosophy Pears, David, 42, 99 Peirce, Charles Sanders, 71–2 perception, philosophy of, 55–9, 83–4, 94–7, 179–80, 183 see also sense data Perry, John, 28–9 personal identity, philosophy of, 21–9, 150, 154–5 phenomenalism, 28, 39, 59, 95 Pitcher, George, 33 Plato, 165 politeness (theory), 199–205 Pope, Rob, 190 Postal, Paul, 120 Pratt, Mary Louise, 189 presupposition, 48, 51, 118, 124–7, 143 psychological concepts, 111, 113, 136, 137, 146, 151, 152, 172–3 Quine, Willard Van Orman, 42, 48–9, 59, 85, 114, 116, 117, 175, 181, 186 semantic scepticism, 52–4 Quinton, Anthony, 55 rationality, philosophy of, 4, 8, 141–50, 153–4, 163–4, 173, 175–6

and the Cooperative Principle, 166–7 reasons practical v. non-practical, 146–8, 158, 161 reasons for v. reasons why, 142, 145 see also rationality Rée, Jonathan, 32 reference Grice on, 117–19, 121–7 see also Frege, Gotlob; Russell, Bertrand Reid, Thomas, 22, 24–5 Relevance Theory, 205–8, 213, 215 religion, 11, 149–50, 155 Richardson, John, and Alan Richardson, 214 Rogers, Ben, 18, 33, 38 Ross, John, 120 Rossall School, 14 Russell, Bertrand, 2, 15, 18, 29, 34, 38, 44, 90 theory of descriptions, 35–6, 47–8, 68, 118, 124–5, 127, 143 Ryle, Gilbert, 16, 21, 31–2, 34, 42, 67–8, 70, 89, 105, 116, 151, 181 Sacks, Harvey, 191 Sadock, Jerold, 211–12 Sampson, Geoffrey, 190 Saul, Jennifer, 215 Saygina, Ayse Pinar, and Ilyas Cicek, 195 Sbisa, Marina, 166 Schiffer, Stephen, 80–2, 110, 165 Searle, John, 78–80, 109, 110, 132 Seattle, University of Washington, 157, 158, 167–8 sense data, 28, 39–40, 52, 54, 56, 57, 94–6, 179–80, 183 Simpson, Paul, 121 Smith, Neil, and Ianthi-Maria Tsimpli, 194 sociolinguistics, 199 speech acts, see Austin, John Langshaw; Searle, John Sperber, Dan, and Deirdre Wilson, 197, 205–7, 213 Stevenson, Charles, 64–5, 70, 73, 76

1403_902976_13_ind.qxd 12/13/04 6:34 PM Page 247

Index 247 Stout, G. F., 66, 69, 70 Strawson, Peter Fredrick, 3, 31, 42, 63, 77–8, 81, 83, 86, 94, 106–7, 113, 116, 157, 183 collaborative work with Grice, 6–7, 47–54 debate with Russell on referring, 47–8, 124–5 and David Wiggin, 184 Studies in the Way of Words, 169 Grice’s views on, 180–1 preparation of, 157, 168, 182 publication of, 183 stylistics, see literature Swartz, Robert, 97 theory-theory, 170–5, 177 Thomas, Jenny, 189, 190 Thomason, Neil, 181 transformational grammar, 187–8 Traugott, Elizabeth, 190 Travis, Charles, 183 truth-conditions, see logic Turing Test, 195 Urmson, J. O., 43, 86, 92 value, 147–8, 154, 158 and cooperation, 166–7

Grice’s theory of, 161–4, 172, 175–6, 178–9, 183 and meaning, 164–6 see also ethics Van Dijk, Teun, 189 verification, see logical positivism Vienna Circle, see logical positivism ‘vulgar’ understanding, 177–9, 183 Warner, Richard, 4, 140, 143, 149, 154–5, 162, 168, 173, 175, 183 Warnock, Geoffrey, 31, 33, 37–8, 39, 42, 86 collaborative work with Grice on perception, 55–9, 94, 179 Williams, Bernard, and Alan Montefiore, 38 Williamson, Timothy, 29 Winer, Gerald, and colleagues, 197 Wittgenstein, Ludwig, 2, 15, 16, 31, 33, 34, 37–8, 63, 89, 129 Woozley, A. D., 83 Young, Michael, 196 Ziegeler, Debra, 190 Ziff, Paul, 82

1403_902976_13_ind.qxd 12/13/04 6:34 PM Page 248