Pleasure, Mind and Soul

Pleasure, Mind, and Soul

This page intentionally left blank

Pleasure, Mind, and Soul Selected Papers in Ancient Philosophy

C. C. W. Taylor

C L A RE N D O N P RE S S · OX F O RD 2008

1

Great Clarendon Street, Oxford ox2 6dp Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide in Oxford New York Auckland Cape Town Dar es Salaam Hong Kong Karachi Kuala Lumpur Madrid Melbourne Mexico City Nairobi New Delhi Shanghai Taipei Toronto With offices in Argentina Austria Brazil Chile Czech Republic France Greece Guatemala Hungary Italy Japan Poland Portugal Singapore South Korea Switzerland Thailand Turkey Ukraine Vietnam Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries Published in the United States by Oxford University Press Inc., New York © C. C. W. Taylor 2008 The moral rights of the author have been asserted Database right Oxford University Press (maker) First published 2008 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this book in any other binding or cover and you must impose the same condition on any acquirer British Library Cataloguing in Publication Data Data available Library of Congress Cataloging in Publication Data Data available Typeset by Laserwords Private Limited, Chennai, India Printed in Great Britain on acid-free paper by Biddles Ltd., King’s Lynn, Norfolk ISBN 978–0–19–922639–9 10 9 8 7 6 5 4 3 2 1

Contents Introduction Acknowledgements Abbreviations 1. Pleasure, Knowledge, and Sensation in Democritus (1967)

vii xii xv 1

2. ‘All Perceptions Are True’ (1980)

23

3. Plato, Hare, and Davidson on Akrasia (1980)

42

4. The End of the Euthyphro (1982)

62

5. The Arguments in the Phaedo Concerning the Thesis that the Soul is a Harmonia (1983)

73

6. Plato and Aristotle on the Criterion of Real Pleasures (1983)

91

7. Urmson on Aristotle on Pleasure (1988)

107

8. Popular Morality and Unpopular Philosophy (1990)

121

9. Socratic Ethics (1992)

134

10. Platonic Ethics (1998)

150

11. The Atomists (1999)

181

12. Aristotle on the Practical Intellect (2003)

204

13. Plato on Rationality and Happiness (2003)

223

14. Pleasure: Aristotle’s Response to Plato (2003)

240

15. The Hedonism of the Protagoras Reconsidered (2003)

265

16. Wisdom and Courage in the Protagoras and the Nicomachean Ethics (2004)

281

17. Socrates (2005)

295

vi

contents

18. Democritus and Lucretius on Death and Dying (2007)

316

19. Socrates under the Severans (2007)

327

General Bibliography Publications of C. C. W. Taylor Index of Passages Cited General Index

340 350 355 365

Introduction The papers contained in this volume span the forty years from 1967 to 2007, thus encompassing the greater part of my academic career, and in their content they represent my main interests during that period. Though a few of my published papers, particularly the early ones, are non-historical, and though I have also written on the history of early modern philosophy, specifically on the British Empiricists (see list of publications), most of my work has been in ancient philosophy. Within that broad rubric, I have focused on one principal and two subsidiary areas. The principal area is the ethics and moral psychology of Plato and Aristotle, with a particular concentration on pleasure and desire, but extending to moral epistemology. The subsidiary areas are first ancient atomism, both that of the fifth century bc (primarily Democritus) and Epicureanism, and secondly Socrates. Of the nineteen papers reprinted here twelve deal with ethics and moral psychology, four with atomism, and three with Socrates. I hope that a word of explanation of this spread of subjects will not be inappropriate, since I think that the explanation may be of some theoretical interest, beyond the purely biographical. Those who work in ancient philosophy are in my view always subject to, and are to various degrees conscious of two conflicting pressures, on the one hand to relate their studies to philosophical issues current at the time of writing, and on the other to relate the work of the ancient philosophers whom they study to the wider intellectual and cultural background against which those philosophers operated. On the one hand, why study ancient philosophers, as distinct from dramatists or historians, unless one is interested in the questions which those philosophers discussed? And if one is interested in those questions themselves, surely one must have some interest in seeking the answers to them. On the other hand, studying the work of any philosopher is an attempt to understand that philosopher’s thought, which requires detailed attention to the concepts in which it was framed, the language in which it was expressed, and the issues to which it was directly addressed. That is to say, study of a philosopher is inseparable from study of that philosopher’s intellectual world, and the more remote that world is

viii

introduction

from our own in time and in cultural presuppositions, the more demanding that study is likely to be. These pressures are potentially in conflict, since the fascination of either is capable of leading to the neglect, or even to the extinction of the other. Ideally, all one’s work should do justice to both, but in practice one or other is likely to predominate in a particular piece of work, and the question of how much weight is given to each in a single work calls for non-codifiable, Aristotelian-style discernment. It is the Heraclitean tension between these demands which gives the practice of the history of philosophy its occasionally exhilarating and invariably demanding character. These general considerations combine with some aspects of my personal biography in explaining the range of topics which dominate this volume. As a student of Classics at Edinburgh in the 1950s I had found my interests captured principally by the abstract speculations of the Presocratics and Plato (we studied virtually no Aristotle or later Greek philosophy) and had decided in consequence to widen my knowledge of philosophy by reading Greats at Oxford. As a pupil of R. M. Hare at Balliol I was immersed willynilly (though it was not in fact uncongenial to me) in moral philosophy and the debates about its foundations which raged between Hare on the one hand and Philippa Foot and Elizabeth Anscombe on the other. I was impressed by Anscombe’s argument that progress in moral philosophy was impossible without a more adequate moral psychology, and in particular that pleasure was a central concept badly in need of elucidation. I was also inspired by work being done at that time on pleasure and related concepts by Ryle, Kenny, and others to venture my own contribution, which led to my first published article. But at the same time I felt that proper understanding of pleasure itself required examination of the history of treatments of the concept, in which the Greeks were of course the pioneers. Hence I wrote my B.Phil. thesis on Plato’s treatment of pleasure, beginning with a brief survey of pre-Platonic discussions, including that of Democritus. The B.Phil. thesis eventually led on to my commentary on Plato’s Protagoras in the Clarendon Plato series and to The Greeks on Pleasure, coauthored with Justin Gosling. Some items in this volume are direct spin-offs from those larger works. Chapter 6 is a shortened version of part of a chapter of The Greeks on Pleasure, and chapters 14 and 15 are reconsiderations of discussions respectively in that work and in the Protagoras commentary,

introduction ix involving in each case some change of mind. Chapter 3 uses the Protagoras together with other texts to further the discussion of akrasia, a topic to which I had been introduced by Hare in the context of tutorial discussion of prescriptivism, and which I had pursued in reviewing his Freedom and Reason. A central theme in chapter 4, on the Euthyphro, is the application to that dialogue of the interpretation of the theory of the unity of the virtues which I had maintained in the Protagoras commentary. Chapter 16 uses the Protagoras in a different way, in making a direct comparison, with some philosophical implications, between the treatment of wisdom and courage in that dialogue and Aristotle’s handling of the same topics in the Nicomachean Ethics. Chapter 7, while making use of some textual material, is more directly philosophical than many of the other historical pieces, in that I there urge the philosophical thesis that in at least some cases the enjoyment of sensations is integral to the enjoyment of activity. I have indicated how some of the pieces in this volume arose from my prior philosophical preoccupations. Another important stimulus throughout my working life has been tutorial teaching. I deplore the dichotomy between teaching and research which is taken for granted in so much modern academic life, with its assumption that research alone constitutes serious academic activity, while teaching is an unrewarding chore to be hurried through with the mind in neutral. In my view teaching, especially by the tutorial method, is, or at least should be, an active and cooperative exploration of problems, in which the teacher is continually stimulated to reassess the subjects taught, and that process of exploration naturally leads to ideas which are subsequently developed for publication. While I should not be so absurdly vain as to suggest that my own teaching has always realized that ideal, I know that some of the pieces published here develop thoughts which originally came to me, or were suggested by others, in the context of teaching. That is true of the central themes of chapters 4, 5, and 18, and of some of the key ideas in chapters 8, 9, 10, 12, and 13. In general, the working out of my understanding, such as it is, of the ethics and moral psychology of Plato and Aristotle has been for me inseparable from my constant re-engagement with the major texts in the annual round of teaching. The subsidiary areas show the same interpenetration of philosophical preoccupations and input from teaching. I have mentioned that my early investigations of ancient treatments of pleasure included Democritus, and

x

introduction

I lectured on him in the 1960s. Another stimulus was tutorial teaching on Locke, which strengthened my interest in the ancient antecedents of the corpuscular philosophy of the seventeenth century. These factors combined to lead immediately to my discussion of the relation between Democritus’ ethics and his physical theories (chapter 1) and eventually to my commentary on Leucippus and Democritus in the Phoenix Presocratics series, of which the essentials are published here in chapter 11. An incidental product of this interest was my examination of Epicurus’ doctrine that all appearances are true (chapter 2), which I revisited both in discussing the Epicurean treatment of pleasure in The Greeks on Pleasure and in elucidating Democritus’ epistemology in the commentary; in chapter 2 I make use of texts of Hume and Locke in interpreting Epicurus. Considerably later, tutorials on De rerum natura III provoked the comparison of the treatments of death and dying by Lucretius and Democritus which appears as chapter 18. I originally approached Socrates, as almost everyone does, in the context of the ethical discussions of the early Platonic dialogues (chapter 9). A subsequent invitation to contribute a volume to the Past Masters series led me to set the Platonic figure in a wider historical context and to give some consideration to Socrates’ historical legacy; those aspects are represented by chapters 17 and 19. In terms of the contrast between conceptual and historical interests drawn above the chronological arrangement of the chapters, taken as a whole, manifests a certain shift in emphasis from the former in the direction of the latter. That was certainly not the result of any conscious methodological decision. It simply shows how my way of approaching the subject developed over the years, which may itself reflect developments in the subject as a whole, and consequently in the influences on me. As indicated above, I do not believe that there is a right way of pursuing the subject. In any case such a shift is a matter of degree. Granted that I may have become over the years somewhat more sensitive to historical considerations I hope, and believe, that I have not lost sight of the conceptual problems which focus our historical studies. The chapters of this volume are reprinted substantially as they originally appeared, subject to the following modifications: (i) the style of citations has been changed where necessary to ensure consistency throughout the volume; (ii) a few additional footnotes have been inserted, usually to provide cross-references to other chapters. In the few instances in which

introduction xi I have wished to record a change of mind, or have added references to material which has appeared since the original publication, I have appended an ‘Afterword’ to the text of the chapter. Two chapters were originally published in languages other than English. Of these, the original English text of chapter 12 was translated into German by Regine May, with additional input from Anselm M¨uller, for delivery in a lecture series in Munich in 2000. I made some changes to the German text, chiefly in the form of additional footnotes, for the published version, and what is printed here is my retranslation of that version. Chapter 16 was also written in English, for a conference in Naples in 2002, and was translated by Silvia Casertano for inclusion in the conference proceedings. Subsequently I delivered the English version at various venues, including Oxford, St Andrews, the University of Texas at Austin, and Cornell, and made some changes in the light of the discussions on those occasions. The version printed here includes those changes. In conclusion I thank all those colleagues and pupils who have contributed over the years to the work which is presented here. Though individual contributions can be identified only rarely (as signalled in the acknowledgements in individual chapters) their cumulative influence has created the intellectual environment without which that work would not have come into existence. As far as this volume is concerned I should like to thank the editorial staff of the Oxford University Press, and especially Peter Momtchiloff for his constant help and encouragement. C. C. W. T.

Acknowledgements The papers comprising this volume were originally published as follows. 1. ‘Pleasure, Knowledge and Sensation in Democritus’, Phronesis 12 (1967), 6–27. Reprinted by permission of Koninklijke Brill NV. 2. ‘ ‘‘All Perceptions are True’’ ’, in M. Schofield, M. Burnyeat, and J. Barnes, eds., Doubt and Dogmatism: Studies in Hellenistic Epistemology, Clarendon Press, Oxford, 1980, 105–24. 3. ‘Plato, Hare and Davidson on Akrasia’, Mind 89 (1980), 499–518. 4. ‘The End of the Euthyphro’, Phronesis 27 (1982), 109–18. Reprinted by permission of Koninklijke Brill NV. 5. ‘The Arguments in the Phaedo Concerning the Thesis that the Soul is a Harmonia’, in J. P. Anton and A. Preus, eds., Essays in Ancient Greek Philosophy II, State University of New York Press, Albany, 1983, 217–31. Reprinted by permission of the State University of New York Press. © 1983 State University of New York. All rights reserved. 6. ‘Plato and Aristotle on the Criterion of Real Pleasures’, in J. Harmatta, ed., Actes du VIIe Congr`es de la F.I.E.C., Akad´emiai Kiad´o, Budapest, 1983, 345–56. Reprinted by permission of the International Federation of the Societies of Classical Studies (FIEC). 7. ‘Urmson on Aristotle on Pleasure’, in J. Dancy, J. M. E. Moravcsik, and C. C. W. Taylor, eds., Human Agency: Language, Duty, and Value. Philosophical Essays in Honor of J. O. Urmson, Stanford University Press, Stanford, Calif., 1988, 120–32. Reprinted by permission of Stanford University Press. © 1988 by the Board of Trustees of the Leland Stanford Junior University. 8. ‘Popular Morality and Unpopular Philosophy’, in E. M. Craik, ed., ‘Owls to Athens’: Essays on Classical Subjects Presented to Sir Kenneth Dover, Clarendon Press, Oxford, 1990, 233–43. 9. ‘Socratic Ethics’, in B. S. Gower and M. C. Stokes, eds., Socratic Questions, Routledge, London 1992, 137–52. Reprinted by permission of Taylor and Francis Ltd.

acknowledgements xiii 10. ‘Platonic Ethics’, in S. Everson, ed., Companions to Ancient Thought 4: Ethics, Cambridge University Press, Cambridge, 1998, 49–76. Reprinted by permission of Cambridge University Press. 11. ‘The Atomists’, in A. A. Long, ed., The Cambridge Companion to Early Greek Philosophy, Cambridge University Press, Cambridge, 1999, 181–204. Reprinted by permission of Cambridge University Press. 12. ‘Aristotle on the Practical Intellect’ (in German, trans. R. May, with the assistance of A. W. M¨uller), in T. Buchheim, H. Flashar, and R. A. H. King, eds., Kann man heute noch etwas anfangen mit Aristoteles?, Meiner Verlag, Hamburg, 2003, 142–62. Reprinted by permission of Felix Meiner Verlag GmbH. 13. ‘Plato on Rationality and Happiness’, in J.Yu and J. J. E.Gracia, eds., Rationality and Happiness: From the Ancients to the Early Medievals, University of Rochester Press, Rochester, NY, 2003, 35–49. Reprinted by permission of the editors. 14. ‘Pleasure: Aristotle’s Response to Plato’, in R. Heinaman, ed., Plato and Aristotle’s Ethics (Proceedings of the Fourth Keeling Colloquium in Ancient Philosophy), Ashgate, Aldershot, 2003, 1–20. Reprinted by permission of the editor. 15. ‘The Hedonism of the Protagoras Revisited’, in A. Havl´ıcˇ ek and F. Karf´ık, eds., Plato’s Protagoras, Proceedings of the Third Symposium Platonicum Pragense, OIKOYMENH, Prague, 2003, 148–64. Reprinted by permission of OIKOYMENH. 16. ‘Wisdom and Courage in the Protagoras and the Nicomachean Ethics’ (in Italian, trans. S. Casertano), in G. Casertano, ed., Il Protagora di Platone: struttura e problematiche, Loffredo Editore, Naples, 2004, 716–28. Reprinted by permission of Loffredo Editore. 17. ‘Socrates’, in D. Borchert, ed., The Encyclopedia of Philosophy, 2nd edn., Macmillan Reference, USA, Detroit, 2005, vol. 9, 105–14. Reprinted by permission of the Gale Group. 18. ‘Democritus and Lucretius on Death and Dying’, in A. Brancacci and P.-M. Morel, eds., Democritus: Science, the Arts, and the Care of the Soul, Proceedings of the International Colloquium on Democritus (Paris, 18–20 September 2003) (Philosophia Antiqua 102), Brill, Leiden, 2007, 77–86. Reprinted by permission of Koninklijke Brill NV.

xiv

acknowledgements

19. ‘Socrates under the Severans’, in S. Swain, S. Harrison, and J. Elsner, eds., Severan Culture, Cambridge University Press, Cambridge, 2007, 496–507. Reprinted by permission of Cambridge University Press. Permission to republish these papers is gratefully acknowledged.

Abbreviations

A. Ancient sources

VH Aesch. Septem Aet. In Meta. In Top. De Plat. Ar. An. post. De an. EE EN GA GC MA Meta. MM PA Phys. Pol. Rhet. Top. Aristoph. Nub. Aspasius In EN

Aelian Varia historia Aeschylus Seven against Thebes Aetius Alexander Commentary on Meta. Commentary on Top. Apuleius On Plato Aristotle Posterior Analytics De anima Eudemian Ethics Nicomachean Ethics De generatione animalium De generatione et corruptione De motu animalium Metaphysics Magna moralia De partibus animalium Physics Politics Rhetoric Topics Aristophanes Clouds Commentary on EN

xvi

abbreviations

Athenaeus Deipn. Chalcidius Comm. Cic. De fin. De or. ND Tusc. Protr. Strom. Dem. DL Ep. Hdt. KD Eur. Supp. Eus. PE Isa. Isocr. Justin II Apol. Lactantius Div. inst. Lucr. Lys. Vita Plat. Cels. In De an. In GC

Deipnosophistai Commentaries Cicero De finibus De oratore De natura deorum Tusculan Disputations Clement Protrepticus Stromateis Demosthenes Diogenes Laertius Epicurus Letter to Herodotus Kuriai Doxai Euripides Supplices Eusebius Praeparatio Evangelii Isaeus Isocrates Second Apology Divina institutio Lucretius Lysias Olympiodorus Life of Plato Origen Against Celsus Philoponus Commentary on De an. Commentary on GC

abbreviations xvii In Phys. Pl. Ap. Charm. Crt. Euthphr. Euthyd. Gorg. Hp. Ma. Lach. Leg. Phd. Phaedr. Phil. Pol. Prot. Rep. Symp. Tht. Tim. Plut. Col. In Rep. Hist. eccles. Sext. M PH In Phys. Stob. In Phys. Ad nat.

Commentary on Phys. Plato Apology Charmides Crito Euthyphro Euthydemus Gorgias Hippias Major Laches Laws Phaedo Phaedrus Philebus Statesman Protagoras Republic Symposium Theaetetus Timaeus Plutarch Against Colotes Proclus Commentary on Rep. Rufinus Ecclesiastical History Sextus Empiricus Against the Mathematicians Outlines of Pyrrhonism Simplicius Commentary on Phys. Stobaeus Themistius Commentary on Phys. Tertullian Ad nationes

xviii

abbreviations

Apol. De an. Theod. Cure Theophr. De sens. Thuc. Xen. Ap. Mem. Symp.

Apologeticus De anima Theodoretus Cure of the Ills of the Greeks Theophrastus De sensibus Thucydides Xenophon Apology Memorabilia Symposium

B. Modern works

CQ Decleva Caizzi DK JHS Long/Sedley LSJ

Mullach OSAP PAS PR Usener

Classical Quarterly F. Decleva Caizzi, Antisthenis Fragmenta, Milan, 1966 H. Diels and W. Kranz, Die Fragmente der Vorsokratiker, 3 vols., Berlin, 1954 Journal of Hellenic Studies A.A. Long and D.N. Sedley, The Hellenistic Philosophers, 2 vols., Cambridge, 1987 H. G. Liddell and R. Scott, A Greek-English Lexicon, rev. H. J. Jones and R. McKenzie, Oxford, 1940 F. W. Mullach, Fragmenta Philosophorum Graecorum, ii, Paris, 1881 Oxford Studies in Ancient Philosophy Proceedings of the Aristotelian Society Philosophical Review H. Usener, Epicurea, Leipzig, 1887 (repr. Stuttgart, 1966)

1 Pleasure, Knowledge, and Sensation in Democritus While historians of philosophy, ancient and modern, have generally and rightly considered the main interest of Democritus to lie in his metaphysics and epistemology, the bulk of the fragments of his writings deal not with these but with ethical topics. It is, therefore, of obvious interest to enquire what connection, if any, may be discerned between the ethical writings and the main body of the atomistic system. Further, this enquiry, as undertaken by modern critics, has produced considerable divergence in its results. Thus on the one hand A. Dyroff¹ was unable to see any connection at all, while C. Bailey² is content with the conclusion that the ethical doctrine, which was in itself in no sense a coherent system, had only a loose connection with the main atomistic theory. In contrast, P. Natorp³ held that the ethical theory is closely integrated with the cosmological, a view which has been developed with impressive erudition by G. Vlastos.⁴ In this paper I attempt to show that while there certainly exists a close connection between the two main strands in Democritus’ philosophy, the exact nature of that connection has not been adequately outlined by either Natorp or Vlastos. To be more precise, their mistake seems to me to lie in looking for the connection in some description of the ultimate end of human action as conceived by Democritus, rather than in the relation of his accounts of moral and of theoretical knowledge. Natorp’s account presents an extremely Platonic picture of Democritus. He calls attention to those fragments, e.g. 37⁵—‘He who chooses the goods of the soul chooses the more divine; he who chooses those of ¹ Dyroff [1899], 41 ff. ² Bailey [1928], I. iii. 9–10. ³ Natorp [1893], ii. 3. ⁴ Vlastos [1945/6]. ⁵ The numbering of fragments is that of DK.

2

pleasure, knowledge, and sensation

its dwelling-place chooses human things’—and 187, where the soul is ranked above the body, and also to those, e.g. 189, 233–5, where the characteristically bodily pleasures, particularly eating, drinking, and sex, are denigrated on the grounds that a life given over to them ends by bringing more pain than pleasure, and that in any case these pleasures are inherently unsatisfactory, in that the satisfaction they give is only temporary, while the distress of e.g. hunger constantly recurs. His conclusion from this is that by his advice to seek happiness by cultivating the pleasures of the soul rather than those of the body Democritus is recommending that one should devote oneself to the highest activity of the soul, the study of the nature of things. And since the nature of things is revealed in the cosmological theory of Leucippus and Democritus, the link between atomic physics and ethics is simply that it is in the study of the former that man achieves his highest good. In drawing this conclusion Natorp puts considerable emphasis on a passage of Cicero (De fin. V. 8. 23; DK 68 A 169), where Democritus is described as having altogether withdrawn from worldly concerns ‘quid quarens aliud nisi vitam beatam? quam si etiam in rerum cognitione ponebat, tamen ex illa investigatione naturae consequi volebat, bono ut esset animo. ideo enim ille summum bonum εὐθυμίαν et saepe ἀθαμβίαν appellat, id est animum terrore liberum.’ Now this passage seems to me to say no more than that Democritus himself studied the nature of things with a view to achieving that freedom from anxiety which, according to Cicero, he identified with man’s highest good; there is no suggestion here that he thought that that was the only way of achieving it. Natorp’s reliance on this passage seems misconceived for a further reason; he maintains, rightly, as I hope to show later, that for Democritus εὐθυμία consisted not simply in freedom from disturbance but in pleasure unalloyed by any pain or unease, and that the study of the universe was the best sort of activity because the pleasure which one derives from that study is the best sort of pleasure. Yet this passage says explicitly that for Democritus the ultimate end was just to have one’s mind free from fear, and that the point of investigating the universe was not that it is pleasant, but simply that it gets rid of anxiety. If he is to retain his general conception of εὐθυμία, Natorp must hold that here Cicero misrepresents Democritus in an important particular. But it then seems that he is hardly justified in using this passage as the sole evidence for a conclusion about the nature of εὐθυμία. Natorp’s conclusion does not seem to me to be supported by any of the fragments; one might

pleasure, knowledge, and sensation 3 indeed claim that by ‘the goods of the soul’ in fr. 37 Democritus means cosmological speculation, but there seems no reason to suppose that the phrase must refer to that rather than to a quiet conscience (fr. 174) or to the joys of friendship (frr. 98–9). The two passages which Natorp himself cites both seem to give very doubtful support. The first is fr. 194, ~ ~ ἔργων γίνoνται, τὰ καλὰ των αἱ μεγάλαι τέρψεις ἀπὸ τoυ~ θεασθαι

which he takes in the sense ‘Great joys come from contemplating the wonderful works of nature’, i.e. from looking at the constitution of the universe as revealed by the atomic theory. But this is surely an extreme~ ἔργων. It seems more ly far-fetched sense for the phrase τὰ καλὰ των plausible to translate the whole ‘Great joys come from contemplating fine deeds’, perhaps in the sense that one source of pleasure is the knowledge that one has acted well (cf. fr. 174). Alternatively, the passage might be taken to refer to the pleasure of looking at works of art. The second passage is fr. 112, θείoυ νoυ~ τὸ ἀεί τι διαλoγίζεσθαι καλόν, which Natorp takes to mean that it is the mark of the splendid or ‘godlike’ intellect always to be thinking out scientific problems. This looks like a simple case of over-translation; the verb seems to have the quite unspecific sense of ‘consider’ or ‘think about’, which gives a sense which is both perfectly satisfactory and more in line with the general run of the fragments, viz. that it is a mark of the fine mind always to be thinking about something fine, as opposed, presumably, to mulling over such squalid topics as wine or chorus-girls. These fragments, then, do not support Natorp, nor, as far as I can see, do any others. Further, one fragment at least might reasonably be taken to contradict his theory, fr. 65, πoλυνoΐην, oὐ πoλυμαθίην ἀσκέειν χρή, which might be taken to say that for the good life one does not need formal learning, as one presumably would in order to master the atomic theory, but practical intelligence.⁶ Yet it would clearly be wrong to put too much weight on a single isolated sentence; it is sufficient to say that not only does Natorp’s view have no support in the fragments, but that from them there may just as plausibly be derived support for a directly contradictory theory. ~ and πoλυμαθίη (the latter covering cosmological speculation as ⁶ On the distinction between νoυς well as historical and mythological learning) v. Heraclitus DK 22 B 40.

4

pleasure, knowledge, and sensation

Vlastos’ theory has the advantage over Natorp’s of a much more intimate dependence on the texts, both of the fragments and of the secondary authorities. He begins by citing texts from the Hippocratic corpus to show that some medical theorists regarded psychical states, both normal and abnormal, as causally dependent on bodily states, and in particular on the dispositions of the elements composing the body. Then, drawing attention to the atomistic view of the soul as a physical structure which moves the body by virtue of the particularly dynamic character of the fiery, spherical soul-atoms,⁷ he suggests that in Democritus’ theory the causal dependence is reversed, states of the whole organism being dependent on the physical constitution of the soul. In particular, the ultimate end of human conduct, which as well as εὐθυμία Democritus is said to have called εὐεστώ, ἀταραξία, and ἀθαμβία, was a particular physical state of the soul, in which the atoms were in the proper arrangement, not subject to any of the violent physical disturbances consequent upon the intense stimulation afforded by sensual pleasures. Many of the terms in which the ethical theory is expounded or described refer directly, according to Vlastos, to the physical theory. Thus εὐεστώ, literally ‘well-being’, means ‘having one’s essential nature (ἐστώ) in a good state,’ that nature being one’s soul-atoms in the surrounding void. Then Diogenes Laertius’ description of εὐθυμία as ~ καὶ εὐσταθως ~ ἡ ψυχὴ διάγει,⁸ means the state the state καθ’ ἣν γαληνως in which the soul remains physically undisturbed like a calm sea. Again, fr. 191, which says that εὐθυμία comes from moderation in pleasure and balance (συμμετρίη) in one’s life, is taken in a physical sense. This fragment goes on, ‘Excess and deficiency tend to change and cause considerable movement in the soul, and souls which are subject to movement over a large interval are neither stable nor happy’; according to Vlastos the description is a literal account of physical motion. The striking fr. 33, ~ δὲ ϕυσιoπoιε ~ι, ἡ διδαχὴ μεταρυσμo ~ι τὸν ἄνθρωπoν, μεταρυσμoυσα

also fits Vlastos’ theory neatly; in imparting new thoughts to the mind teaching actually alters the physical pattern of the soul-atoms by providing new physical stimulation (for this account of thought v. Ar. Meta.  5, 1009b7 ff.; DK 68 A 112), and thus literally fashions a new ϕύσις for the individual. For Vlastos, as for Natorp, fr. 187, ⁷ Ar. De an. 404a5 ff.; DK 68 A 101.

⁸ IX. 45; DK 68 A 1(45).

pleasure, knowledge, and sensation 5 ~ μαλλoν ~ ἀνθρώπoις ἁρμόδιoν ψυχης ἢ σώματoς λόγoν ~ πoιε ισθαι,

is a key slogan of Democritus’ ethical programme, but the slogan is understood in a quite different sense. According to Vlastos, the λόγoς here referred to is a theory about the nature of the soul, of which his own atomic theory is of course the best example. Vlastos’ account also dovetails Democritus’ theory of knowledge neatly with the physical and ethical theories. Fr. 69 says that for all men the same thing is good and true, while what is pleasant is different for different men. This gives an obvious parallel with the famous fr. 125, νόμῳ χρoίη, νόμῳ γλυκύ, νόμῳ πικρόν, ἐτεᾗ δ’ ἄτoμα καὶ κενόν. What truly exists is atoms and void, while such qualities as colour, sweetness, and in general secondary qualities are mere shifting appearances. Pleasantness is ranked with the latter, as it obviously varies from person to person, while the good is independent of all changes in the perceiver or the environment. But we do not have a mere parallelism, for the good is identical with the real; ~ τωὐτὸν ἀγαθὸν καὶ ἀληθές. ἀνθρώπoις πασι

The good is εὐεστώ, the real stuff of the soul, viz. atoms and void, in the proper arrangement. The process of discovering the real nature of the world and that of discovering the ultimate end of human conduct is one and the same, that of penetrating the shifting screen of phenomena to the underlying reality. Vlastos’ account has very considerable attractions. Not only does it tie together a number of apparently disparate elements in the tradition, but it systematically applies to the ethics conclusions which follow from or are at least consistent with the materialistic premisses of Democritus’ cosmology. If he was a consistent materialist he must have held that all introspectively observable psychical states are at least causally dependent on physical states of the organism conceived as an aggregate of atoms in the void. Further, though not required by the theory, the suggestion that mental disturbance is produced by violent physical motion of atoms in the soul and happiness by a calm and settled state of the atoms would be a plausible hypothesis for an atomist. Again, Aristotle’s statement in Meta.  5 (v. supra) that for Democritus thought was identical with sensation and

6

pleasure, knowledge, and sensation

sensation with qualitative change suggests that Vlastos is right in his claim that according to the theory teaching operates by changing the disposition of the atoms in the soul. So if Vlastos had been content to advance his thesis as a conjectural account of what Democritus may have held, if he applied his principles without inconsistency, it could have been accepted as providing a useful insight into the possibilities of the atomic theory. Unhappily, however, he went further, claiming to find in the texts explicit support for the contention that Democritus in fact made the link between ethics and physics which his investigations had shown to be possible. Here his contention becomes unacceptable, for his interpretations of the crucial texts are in almost every case highly dubious and in some cases clearly impossible. The corner-stone of Vlastos’ account is his analysis of the meaning of the word εὐεστώ, which is given by Diogenes Laertius,⁹ Stobaeus,¹⁰ and Clement of Alexandria¹¹ as Democritus’ synonym for εὐθυμία, was reputed to be the title of one of the ethical works,¹² and also occurs in a single fragment, no. 257. Starting from the etymology of the word as εὐ + ἐστω, literally ‘well-being’, Vlastos draws attention to the use of the simple ἐστώ by Philolaus¹³ and the compound ἀειεστώ by Antiphon,¹⁴ where ἐστώ has the sense of ‘being’ or ‘substance’, and concludes that ‘To an atomist ἐστώ can mean only one thing; atoms and the void’. Hence for Democritus εὐεστώ means having the atoms and void of one’s soul in the proper arrangement. Besides the passage of Philolaus quoted by Vlastos, and its citation by Photius as a Pythagorean name for the dyad,¹⁵ ἐστώ occurs uncompounded only once, in a passage of Archytas preserved by Stobaeus.¹⁶ In Philolaus it has the sense of ϕύσις or oὐσία, while in Archytas it has the sense of the Aristotelian ὕλη, being contrasted with μoρϕή and with the efficient cause of change. Vlastos’ interpretation requires that the element ἐστώ should be used by Democritus in the compound εὐεστώ in one or other of these senses; to this suggestion there are serious objections. Firstly, the uncompounded word is not attested for a writer in any dialect other than Doric. To this Vlastos might reply that he does not have to claim that Democritus, writing in Ionic, used the Doric ἐστώ as a technical term, but merely that he took over the sense of the ⁹ IX. 45; DK 68 A 1. ¹⁰ II. 7. 3i; DK 68 A 167. ¹¹ Strom II. 130; DK 68 B 4. ¹² DL IX. 46; DK 68 A 33. Acc. Diogenes the title Eὐεστώ did not appear in Thrasyllus’ catalogue. ¹³ DK 44 B 6. ¹⁴ DK 87 B 22. ¹⁵ LSJ s.v. ii. ¹⁶ I. 41. 2.

pleasure, knowledge, and sensation 7 word as used in Doric philosophical writing to give a special sense to the standard Ionic εὐεστώ; the tradition¹⁷ of his association with Pythagoreans and with Philolaos in particular might be held to support this. Yet if this is to be more than an interesting, but unverifiable, hypothesis it must have some independent support; Vlastos’ attempt to provide this by his citation of Antiphon clearly fails, for his use (presumably a coinage) of ἀειεστώ confirms what is apparent from the standard uses of εὐεστώ, that when compounded in Attic and Ionic ἐστώ has the sense not of ϕύσις ~ or ὕλη but of ε ἰναι. ’Aειεστώ is being for ever, just as εὐεστώ is being in a good state. So far, then, from its being the case that ‘to an atomist ἐστώ can mean only one thing; atoms and the void,’ it appears that to Democritus, as much as to anyone else writing in Ionic, ἐστώ as an element in compounds would most naturally have the sense of the verbal substantive ‘being’, which is no more to be taken to refer to atoms and void than, say, the noun ‘running’, even though everything which is and everything which runs are alike composed of atoms and void. Further, εὐεστώ is a perfectly standard fifth-century word for well-being or prosperity;¹⁸ its use by Democritus as a synonym for εὐθυμία would not seem to call for the slightest special explanation. Vlastos’ account of the meaning of εὐεστώ must, then, be regarded as an unsupported conjecture which on ordinary scientific principles of simplicity it is safest to reject. It would be justifiable to reverse this verdict only if the other passages cited by Vlastos, or any other evidence, provided positive grounds for doing so. In fact most of the passages quoted by Vlastos give no independent support to his conjecture; since they may be understood without reference to that conjecture they support it only if one has already decided on other grounds to adopt it. Thus there is little independent probability in Vlastos’ analysis of Diogenes’ description of εὐθυμία as a state in which ~ ~ καὶ εὐσταθως ἡ ψυχὴ διάγει. From its original Homeric sense γαληνως of ‘well-built, firmly-based’, describing the sort of building not liable to be shaken by e.g. earthquake, εὐσταθής comes to have the regular sense of ‘tranquil’ or ‘settled’, in application to the weather, constitutions of states, bodily conditions or states of mind; similarly, γαληνής has regularly the metaphorical sense of ‘gentle’ or ‘calm’ in application to mental states.¹⁹ Admittedly, in fr. 191 Democritus says that excess and deficiency of pleasure ¹⁷ DL IX. 38.

¹⁸ LSJ.

¹⁹ LSJ s.vv.

8

pleasure, knowledge, and sensation

impart large movements to the soul, which prevent it from being εὔθυμoς . Prima facie this seems to support Vlastos, and to suggest that the passage of Diogenes should be taken in a corresponding sense, but closer attention to the wording of the fragment indicates a different conclusion. ‘Excess and deficiency’, says Democritus, ‘tend to be variable and to impart large movements to the soul; and souls which are moved over large intervals are neither stable nor happy.’ Vlastos interprets this as saying that souls whose atoms are in violent motion are not stable, but surely ‘souls which are moved over large intervals’ is a very odd way of describing souls in that state. A soul in such a state is not itself moved over a large interval any more than a city is when all its citizens run about the streets. One might regard this simply as a pettifogging objection, on the ground that Democritus clearly means ‘souls whose atoms are moved over large intervals’, but that notion too seems to fit very oddly into the general context of the atomistic account of the soul. On this account, the unhappy soul is distinguished from the happy one by the fact that its atoms move over greater intervals. But since according to atomic theory all atoms are in perpetual motion²⁰ and soul-atoms are the most mobile of all,²¹ it is hard to see why in terms of the theory the fact that in some mental states the soul-atoms move further than in others should be supposed to make the crucial difference between well-being and misery. Again, since all atoms are in constant motion, one atom could be said to move further than another only in the sense that it moved further in one direction before colliding with another atom. So excess and defect of pleasure must be supposed to space the atoms out more widely, so that each atom can travel further without hitting another. There seems to be neither any independent ground for the suggestion that anything like that was supposed to happen, nor any obvious reason why an atomist should assume that it must. Of course, none of these considerations show that it is impossible for Democritus to have believed something like that, but they show that what purported to be an obvious and illuminating interpretation of fr. 191 involves a good deal of unsupported and somewhat implausible reconstruction of the Democritean view of the soul. It would be simpler to treat the spatial terms in the fragment as metaphorical, taking μεγάλας κινήσιας as meaning ‘movements from one extreme to the other’, and interpreting the passage ²⁰ Ar. De caelo, 300b8–10; DK 67 A 16.

²¹ Ar. De an. 405a11–13; DK 68 A 101.

pleasure, knowledge, and sensation 9 as a whole as follows, that a soul which oscillates from one extreme of the pleasure-distress scale to the other cannot be stable, just as a pillar which shakes about, or weather which changes very rapidly, are not stable. Other passages cited by Vlastos are equally problematical. Unless one is already convinced of the truth of his theory, there is small temptation to understand fr. 187, ~ μαλλoν ~ ἢ σώματoς λόγoν ἀνθρώπoις ἁρμόδιoν ψυχης ~ πoιε ισθαι,

in the required sense, viz. ‘It is fitting for men to devise a theory of the soul rather than of the body’. It seems much better to translate ‘It is fitting for men to pay more attention to the soul than to the body’, a rendering which not only gives a standard sense to λόγoν πoιε ~ισθαι + gen., while Vlastos’ ~ ψυχης ~ πoιε ~ suggestion would seem to require λόγoν περὶ της ισθαι, but also fits more naturally the rest of the passage, ‘For perfection of soul remedies bodily defects, but strength of body without intelligence does not make the soul any better’. Nor does Vlastos’ interpretation of fr. 69, ἀνθρώπoις πᾶσι τωὐτὸν ἀγαθὸν καὶ ἀληθές · ἡδὺ δὲ ἄλλῳ ἄλλo, fare much better. His suggestion is that since what is real is atoms and void (fr. 125), and what is good for man is obviously εὐεστώ, we are here told that εὐεστώ is identical with some state of atoms and void. But every state of atoms and void that obtains is true for all men, while on Vlastos’ theory not every state is good. Thus there is no simple identity between what is real and what is good. But in any case the sense of the passage requires that the good and the true should be, not the same as one another but the same for everyone, as opposed to the pleasant, which varies from one individual to another. As far as the logic of the passage goes, this would leave it quite open for the good and the true to be two distinct entities, which both have the property of being the same for everyone, as e.g. red and green are the same for everyone while undeniably different from one another. This passage seems, then, as barren as the others of support for Vlastos’ theory. This leaves us, finally, with fr. 33, ~ δὲ ϕυσιoπoιε ~ι, ἡ διδαχὴ μεταρυσμo ~ι τὸν ἄνθρωπoν, μεταρυσμoυσα

10

pleasure, knowledge, and sensation

interpreted by Vlastos as saying that teaching alters the physical configuration of the soul-atoms and thus creates a new ϕύσις , ‘configuration’ being equivalent to ‘pattern’ or ‘arrangement’. The difficulty here is that ῥυθμός appears rarely if ever to have the sense of ‘spatial pattern’; apart from its central senses of ‘rhythm’ and ‘time’, its most common uses are equivalent to ‘condition’ and to ‘shape’.²² More importantly, it was in the latter sense that it was used as an atomistic technical term, meaning the shape of the individual atoms, while the word for their arrangement was διαθιγή.²³ It would, then, be highly inconsistent for Democritus to use the verb μεταρυσμόω with the sense of change in the spatial ordering of atoms. Furthermore, consistently with the technical atomistic usage, the ordinary, literal meaning of the verb is either generally ‘to change’ or particularly ‘to change shape’, not ‘to rearrange’. Moreover, like the English ‘re-form’, the verb has a common use in the sense of ‘amend’, with particular reference to conduct, which is exactly what is required in this context. If Vlastos is unwilling to accept that, he must hold that Democritus is here using the verb in a technical sense inconsistent with his own standard terminology, or that the verb does have the sense of physical reshaping. The former alternative is clearly undesirable, as the interpretation was originally held to be necessary to account for traces of the terminology of the physical theory in the ethical fragments. But if the latter alternative is accepted, the fragment must be taken to say that teaching changes the physical shape of the person taught. Clearly on the normal sense of ‘shape’, viz. the visible outline of a body, that doctrine is very implausible; while that is of course not to say that Democritus cannot therefore have held it, it seems perverse to attribute it to him on the strength of a fragment for which a commonsense interpretation is available. But to try to evade this conclusion by positing some special sense of ‘shape’, something like ‘structure’, is in effect to revert to the first alternative. Instead of being forced to an interpretation of the fragment by the sense of the words, one is positing unattested senses for the words in order to fit an already accepted interpretation. It is much simpler to take the fragment as enunciating the truism that nature and teaching are not altogether different, since teaching changes a man’s character etc., and in so doing makes his nature anew. The sense is thus akin to that of the traditional saying²⁴ ‘Habit is a second nature’, which no one has been ²² LSJ.

²³ Ar. Meta. 985b4 ff.; DK 67 A 6.

²⁴ LSJ s.v. ἔθoς .

pleasure, knowledge, and sensation 11 inclined to take in the sense that habit rearranges the physical elements of the human organism.²⁵ To sum up, Vlastos’ detailed investigations do not provide any significant degree of confirmation for what may reasonably be conjectured about Democritus’ view of the connection between his ethics and his cosmology as a whole. In particular, he does not add anything to the probability, which is on general grounds of consistency very considerable, that in atomistic theory all mental states, including εὐθυμία, are causally dependent on the shape and disposition of the atoms and void which are the ultimate physical constituents of the human person, a complex of body and soul. His claim to produce evidence of the sort of physical state on which εὐθυμία was held to depend is clearly unfounded. Other suggestions, of a like degree of plausibility, may readily be made, e.g. that εὐθυμία depends on a physical state in which the soul-atoms move in regular motion at moderate speed, in contrast with a state of violent sensory or other stimulation, in which they are subject to fast and irregular motion, confirmation for this being sought in the steadiness and placidity which characterizes the εὔθυμoς ,²⁶ and which is appropriate to one whose soul is in the former state. But all such speculation is clearly without foundation; we just do not know what, if anything, Democritus said about the physical state of the soul of the εὔθυμoς , nor how such a view, supposing him to have had one, related to his teaching as to how εὐθυμία ought to be attained. It would seem that we must accept this agnostic conclusion as the last word on the subject, were it not that some features of Democritus’ ethical writings show an interesting parallelism with his epistemology. Some of these features are indeed noticed by both Natorp and Vlastos, but tend to become obscured in the hunt for the nature of εὐθυμία. It seems useful, then, to attempt to isolate this parallelism from the rest of their theories, with a view to delineating it as precisely as possible. This parallelism is best illuminated via consideration of the fact that both the ethical and the epistemological theories contain prima facie contradictions. The contradiction in the ethical theory is put into the sharpest focus by the juxtaposition of fr. 74, ²⁵ For retractation of this criticism see appendix to chapter 11 of this volume. ²⁶ e.g. frr. 3, 191.

12

pleasure, knowledge, and sensation ἡδὺ μηδὲν ἀπoδέχεσθαι, ἢν μή συμϕέρῃ,

with fr. 188, ὅρoς συμϕόρων καὶ ἀσυμϕόρων τέρψις καὶ ἀτερπίη. The latter tells us that the criterion of whether or not something is useful is that it is pleasant, i.e. presumably that something is useful if and only if it is pleasant. The former, however, says that some things may be pleasant but not useful, which is a direct contradiction of our interpretation of fr. 188. This contradiction is to be resolved by regarding each of these fragments as dealing with a different aspect of pleasure; 74 is about the particular action or experience, whose pleasantness or unpleasantness may be considered without any consideration of its place in the broader context of the life of the individual, including its effect on the pleasantness and unpleasantness of other things. 188, on the other hand, is concerned with the pleasantness or unpleasantness, not of the single action or experience, but of one’s life considered as a whole; the criterion of whether something is useful or harmful is whether it is likely to make one’s life as a whole more or less pleasant, which now allows one to see fr. 74 as consistent with 188, in that something may obviously be pleasant in itself and yet tend to make one’s life as a whole unpleasant. The sense in which I speak of the pleasantness of one’s life as a whole is the familiar one in which one speaks e.g. of enjoying one’s life at university, or finding married life very pleasant. The relation of this kind of enjoyment to the enjoyment of the particular activities and experiences composing the whole is complicated; on the one hand it is clear that in order to have this ‘overall’ enjoyment one must enjoy a considerable proportion of the particular activities etc. which make up one’s life, while on the other ‘overall’ enjoyment is not a simple summation of particular enjoyments, since one does not necessarily increase one’s ‘overall’ enjoyment by increasing the number or intensity of one’s particular enjoyments. Even leaving aside questions of satiety, from the fact that one enjoyed each of twenty strawberries it does not seem to me to follow that one’s enjoyment of the dish would have been greater had it contained another one, even though, had there been another one, one would have enjoyed it too. If this principle holds for such a simple contrast as that between the enjoyment of a dish and the enjoyment of the individual parts of the dish, it seems more obviously to hold the more complicated the context becomes into which the particular enjoyment is fitted. ‘Overall’

pleasure, knowledge, and sensation 13 enjoyment is determined not simply by the number and intensity of one’s particular enjoyments, but also, in some way which is unclear to me, by the weight or value which certain particular enjoyments assume in one’s life as a whole. Despite the evidence of the doxographical tradition represented by Diogenes,²⁷ Cicero²⁸ and Strabo,²⁹ which I judge to have been unduly influenced by Epicureanism, this idea of overall enjoyment seems nearer to the sense of εὐθυμία than mere tranquillity. For that view of εὐθυμία allows us to account for sayings which are quite anomalous on the ‘pure tranquillity’ view, e.g. fr. 200, ~ ~ oὐ τερπόμενoι βιoτῃ,³⁰ ἀνoήμoνες βιoυσιν

and the very striking fr. 230, βίoς ἀνεόρταστoς μακρὴ ὅδoς ἀπανδόκευτoς where the word ἀνεόρταστoς suggests that feasting and merry-making and all the usual accompaniments of a religious festival (ἑoρτή), have a place in the good life. This view has the further advantage of being able to accommodate those fragments (e.g. 3, 174, 215) which stress the role of freedom from trouble and fear in the good life, for it is obvious that fear and worry prevent one from enjoying life. It also enables us to give a good account of two fragments which we may consider as expanding the advice of fr. 74 to avoid harmful pleasures, firstly fr. 71, ἡδoναὶ ἄκαιρoι τίκτoυσιν ἀηδίας , and secondly fr. 72, ~ εἰς τἆλλα τὴν ψυχήν. αἱ περί τι σϕoδραὶ ὀρέξεις τυϕλoυσιν

We have here two related reasons for avoiding such pleasures, firstly that they cause positive distress (e.g. a hangover), and secondly that they distract one from the kind of activity which produces εὐθυμία (including the pursuit of moderate pleasures, fr. 191). We are thus able to attribute to Democritus a doctrine which is not only consistent but which gives a good explanation of the relevant fragments, that while the worth of individual pleasures is judged by a further criterion, that criterion is provided by pleasure itself.³¹ It is in this sense, I suggest, that we should interpret the testimony of Stobaeus³² that εὐθυμία is produced by the distinguishing ²⁷ IX. 45; DK 68 A 1(45). ²⁹ I, p. 61; DK 68 A 168. ³² II. 7. 3i; DK 68 A 167.

²⁸ De fin. V. 8. 23: DK 68 A 169. ³⁰ Fr. 204 is a variant of this.

³¹ See McGibbon [1960].

14

pleasure, knowledge, and sensation

and differentiation of pleasures, and is the finest and most useful thing for men. While the differentiation of pleasures is of course the task of practical intelligence, the standard by reference to which they are distinguished from one another is their contribution to the overall enjoyment of life. As the most useful thing this provides the criterion for the assessment of the value of particular pleasures (fr. 74), and as the finest thing it provides the supreme criterion of the moral worth of actions.³³ The contradiction in Democritus’ epistemology is essentially the difficulty that troubled Russell,³⁴ viz. that while all knowledge of the external world is derived from sense-perception, the evidence of perception itself forces us to the conclusion that perception cannot be relied upon, from which it seems to follow that no knowledge of the external world is possible. This difficulty is vividly expressed in fr. 125, where Galen first of all quotes the familiar rejection of sensory information, νόμῳ χρoίη, νόμῳ γλυκύ, νόμῳ πικρόν, ἐτεῃ~ δ’ ἄτoμα καὶ κενόν, and then gives the reply which Democritus puts into the mouth of the personified senses, ~ τάλαινα ϕρήν, παρ’ ἡμέων λαβoυσα τὰς πίστεις ἡμέας καταβάλλεις ; ~ πτωμά τoι τὸ κατάβλημα. Now in what sense is the atomic theory based on empirical evidence? Certainly not in the sense that the atoms themselves, and a fortiori their numbers, movements, shapes, and dispositions, are observable entities. Yet there are two important ways in which the theory does depend on empirical observation. Firstly, the starting point of the theory was the attempt to account for the diversity of phenomena without either succumbing to the Eleatic elenchus or getting involved in logical difficulties about qualitative differentiation such as vitiated the similar attempt of Anaxagoras. To this end the atomists developed an elaborate system of explanations of physical phenomena by correlation with various dispositions of variously shaped atoms in the void. But unless one knows enough about the external world to be able to say what it is that is thus correlated with microscopic events, this procedure is obviously absurd. Aristotle emphasizes this point in

³³ While this view enables e.g. fr. 207 to be seen as an application of the hedonistic criterion, it is unlikely that Dem. applied it with perfect consistency (see. frr. 194, 174). ³⁴ An Inquiry into Meaning and Truth, London, 1940, introduction p. 15: ‘Naive realism leads to physics and physics, if true, shows that naive realism is false.’

pleasure, knowledge, and sensation 15 discussing Leucippus;³⁵ in contrast to the Eleatics, who held that perception is altogether illusory and that the only source of knowledge of the world is consideration of the logic of the verb ‘is’, Leucippus claimed to have a theory which agreed with the data of perception, and which accounted satisfactorily for such basic features of the world of sense as coming into being, ceasing to exist, motion and the multiplicity of particular things. Secondly, the atoms themselves and their motion and interaction were described in terms whose primary application is to the macroscopic world revealed to the senses, not only in that such adjectives as ‘round’, ‘sharp’, ‘impenetrable’, and ‘regularly-moving’, which derive their sense, directly or indirectly, from the world of experience, were applied to them, but that the mechanical processes observed to govern the macroscopic world were assumed to operate in the microscopic world also, e.g. the assumption that in the original cosmic whirl the larger atoms would collect together and the smaller apart from them depends on the assumption that their behaviour reproduces that of grains in a rotating sieve,³⁶ while the whirl was conceived on the analogy of an eddy of wind or water, in which the lighter atoms are thrown out to the circumference, while the others remain in the centre.³⁷ Unless, therefore, the atomic theory admitted sense-perception as a source of knowledge at least to the extent necessary to give a sense to its central concepts and to establish the facts which grounded by analogy its main hypotheses about the microscopic world, it was bound, as Democritus clearly saw, to refute itself.³⁸ One response to such a situation would be to relapse into complete scepticism, and a number of fragments might be taken to suggest that Democritus did indeed do so, e.g. fr. 117, ~ γὰρ ἡ ἀλήθεια, ἐτεῃ~ δὲ oὐδὲν ἴσμεν. ἐν βυθῳ

and frr. 6–10 preserved by Sextus Empiricus,³⁹ of which the most striking is no. 7, ~

δηλo ~ι μὲν δὴ καὶ oὑτoς ὁ λόγoς, ὅτι ἐτεῃ~ oὐδὲν ἴσμεν περὶ oὐδενός, ἀλλ’ ἐπιρυσμίη ἑκάστoισιν ἡ δόξις . Yet to regard Democritus as a sceptic is to ignore the evidence of the same passage of Sextus that he thought that he had found a way out of ³⁵ GC 325a23 ff.; DK 67 A 7. ³⁶ Fr. 164. ³⁷ Ar. De caelo 295a10–12; cf. DL IX. 31; DK 67 A 1. ³⁸ See von Fritz [1938], 19–30. ³⁹ M VII. 135–40.

16

pleasure, knowledge, and sensation

his difficulties in the distinction between σκoτίη and γνησίη γνώμη, the former being equivalent to ordinary, empirical observation of the world, while the latter is a theoretical account of things which supplements the inadequacy of the senses.⁴⁰ This has generally been taken, as indeed it is by Sextus, as showing that in the last resort Democritus rejected the senses as unreliable and thought that a true account of things could be given only by pure reason. We should thus have to interpret him, not as having believed that he had escaped from his own dilemma, but rather, after the manner of Wittgenstein in the Tractatus, as having consciously published a theory containing its own refutation. This interpretation is, however, open to objection on the grounds that it not only ignores certain evidence of Democritus’ views on knowledge and perception, especially that of Sextus himself and of the catalogue of Democritus’ works given by Diogenes,⁴¹ but also that it involves misinterpretation of the crucial fr. 11. To take the latter point first, the fragment presents a contrast, not between knowledge and ignorance, but between two sorts of γνώμη, of which the function of the superior is not altogether to discredit the other, but rather to supplement its inadequacy. It is when the senses are unable to proceed below a certain level of discrimination (ἐπ’ ἔλαττoν), and one needs a more precise method of investigation (ἐπὶ λεπτότερoν < δέῃ ζητε ~ιν >), that γνησίη γνώμη takes over. The trouble with the senses, according to this fragment, is not that they induce one to take illusion for reality, but that they give only a superficial account of the nature of things, as opposed to that insight into their real (i.e. microscopic) nature which the atomic theory provides. Now it is very probable that, like Eddington in his introduction to The Nature of the Physical World (London, 1935), Democritus at times used language indicative of a confusion between on the one hand the contrast between the view of the world given by common observation and that given by scientific investigation and on the other hand the quite distinct contrast between the real world and an illusory one. We can, nonetheless, find evidence that he did hold, perhaps with less than perfect consistency and clarity, that the senses did give correct information about the world, and further that their role was in some way central in his theory of knowledge. ⁴⁰ Fr. 11.

⁴¹ IX. 45–9; DK 68 A 33.

pleasure, knowledge, and sensation 17 The evidence from Sextus comes from the same passage. After quoting the fragment running νόμῳ γλυκύ etc., he continues, ‘And in the work entitled Kρατυντήρια, although he had undertaken to assign to the senses control over belief, nevertheless he is found to condemn them,’ illustrating this by citing another sceptical fragment, no. 9. So one of the themes of that work was to show that the senses had in some sense the last word in the acquisition of knowledge. The title of this work is significant; it would appear to mean literally either ‘strengthening, establishing securely’, or else ‘getting control’, while in the catalogue given by Diogenes the title has a note to the effect that it was ‘critical of what had been said before’. Two interpretations of this seem possible, either that it was a work of criticism of his predecessors, or else that it consisted of criticism of his own doctrines; to a work of the former character a title with the sense of overthrowing or refuting would appear more appropriate, while a work establishing one’s own views by criticism of one’s earlier writings might well be called ‘Strengthening Arguments’. In a work of either kind the vindication of the senses could naturally play an important role, either against the Eleatic attack, as in the work of Leucippus referred to by Aristotle,⁴² or against what he may have come to regard as somewhat misleading overstatements in his own works. In any case, the significance of this work for the present argument is that Democritus held the doctrine of the supremacy of senseperception with sufficient confidence to use it in criticism either of himself or of others. Yet how could that doctrine be consistent with the general principles of atomistic epistemology, and in particular with the contrast between γνησίη and σκoτίη γνώμη? I suggest that the complete story is as follows. We begin with the commonsense picture of the world, in which the information provided by the senses is accepted without question. Various considerations, including Eleatic puzzles about plurality and about comingto-be, and perhaps also considerations of the subjectivity of such sensory data as tastes and colours,⁴³ lead to dissatisfaction with this picture. A theory of the basic constitution of things is then developed which, taking the commonsense picture as its starting point, remedies its deficiencies by showing (a) how the phenomena simply presented by the commonsense ⁴² DK 67 A 7. ⁴³ Sext. PH II. 63; DK 68 A 134: Theophr. De sens. 63; DK 68 A 135.

18

pleasure, knowledge, and sensation

picture are explained and (b) how the difficulties of that picture are eliminated by the postulating of certain fundamental entities and natural laws. Yet though these fundamental entities are unobservable, or at least unobserved, this second, scientific picture of the world is still ‘under the control of the senses’, in two ways. Firstly, since one of the purposes of the theory is to explain the phenomena, any failure to take account of any phenomenon or set of phenomena counts against the theory, either in the sense that it is insufficiently general or, more seriously, that it is directly falsified. Secondly, since the behaviour of the fundamental entities is assumed to be governed by the physical laws of the phenomenal world, any explanation which required a breach of those laws, as discerned by sensory observation, would be illegitimate. Yet clearly, in carrying out this ‘controlling’ function the senses are subject to the familiar weaknesses which are the product of the dependence of the observer on the physical environment, and hence the theory as a whole can be asserted with only that degree of confidence which those weaknesses allow. This, it seems to me, is the explanation of how it is that Democritus, in maintaining the ‘control of the senses over belief’, can yet be represented as ‘condemning’ them. For this condemnation comes to no more than this, that we can never know anything with absolute certainty, but only what changes according to the interaction of the atoms within and external to us.⁴⁴ The point of this is not to deny altogether the possibility of knowledge, still less to refuse to recognize sense-perception as a source of veridical information; rather it is to point out the necessary limitations of knowledge which depends ultimately on that source. It is here, I think, that we have the explanation of the apparent inconsistencies in Aristotle’s account of Democritus’ epistemology. At De an. 404a 27 ff.⁴⁵ and again at Meta.  5, 1009b 12 ff.⁴⁶ he says that Democritus held that all sensation is veridical, while in the immediately preceding sentence of Meta.  he cites Democritus as saying that either nothing is true or that it (i.e. presumably everything) is unclear to us. In themselves these remarks might be taken to describe the same sceptical position, viz. that since there is no criterion of truth by which the data of sensation can be assessed, one may say indifferently that nothing at all is true or that whatever is given in sensation is true. Democritus’ position ⁴⁴ Fr. 9.

⁴⁵ DK 68 A 101.

⁴⁶ DK 68 A 112.

pleasure, knowledge, and sensation 19 would then be the same as that of Protagoras.⁴⁷ This cannot, however, be an adequate picture; not only do both Sextus⁴⁸ and Plutarch⁴⁹ say that Democritus argued against Protagoras, but Aristotle twice refers to Democritus in GC in a way inconsistent with this interpretation. The first passage is that already referred to,⁵⁰ in which Leucippus is credited with the construction of a theory which reconciled the phenomena revealed by perception with the logical requirements of Eleatic monism; it is reasonable to assume that, while crediting Leucippus with the invention of this theory, Aristotle means here to describe Democritus’ views also, since at the beginning of the chapter the theory later outlined as that of Leucippus is introduced as that of Leucippus and Democritus. The second passage,⁵¹ referring explicitly to both, says that since they held that all perception is veridical, they developed their theory to account for the fact that the data of perception are often contradictory. We can now see how they could combine belief in the truth of their theory with the doctrine that ‘truth is in the appearances’, viz. by the belief that conflict between the data of perception could not be resolved by showing that one perceptual judgement was truer than another, but only by showing how each of the conflicting perceptual claims arose from the interaction of the atoms of the observer and of his environment. Every individual has his own commonsense picture of the world, none of which is truer than any other; the only intersubjectively true picture is the scientific one, which can however claim to be true only insofar as it provides an explanation of every commonsense picture, an explanation moreover which depends for its verification on the same potentially conflicting data of perception. The theory is ‘under the control of the senses’ in that it ultimately relies on empirical confirmation, while at the same time being required to explain all sensory phenomena, none of which can be regarded as more veridical than another. Yet if no empirical judgment is truer than any other, there can be no empirical confirmation of any scientific theory. There is thus a fatal inconsistency in the theory,⁵² of which Democritus may perhaps have been at least dimly aware. The parallel with the treatment of pleasure should now be clear. In discovering the truth about the world the unreflective man naturally ⁴⁷ Philoponus In De an. p. 71, 19 ff.; DK 68 A 113. ⁴⁸ M VII. 389; DK 68 A 114. ⁴⁹ Col. 1108f; DK 68 B 156. ⁵⁰ 325a23 ff.; DK 67 A 7. ⁵¹ 314a21 ff.; DK 67 A 9. ⁵² Cf. Theophr. De sens. 69; DK 68 A 135.

20

pleasure, knowledge, and sensation

assumes that its nature is completely revealed by sense-perception; in the sphere of action his natural impulse is to pursue whatever is immediately pleasant. In each case, however, a rational theory intervenes, showing in the first case that the sensory picture of the world is not completely satisfactory and in the second that a life spent in the pursuit of every immediate pleasure will become unliveable. Yet in neither case is the original impulse so much abrogated as developed to embrace the insight of the new theory. In the cognitive field sense-perception finds its place in controlling the explanatory functioning of the theory in the ways sketched above, while in the sphere of action pleasure gains its position as the criterion of right conduct when its sense has been widened from the enjoyment of a particular action or experience to the enjoyment of life as a whole. Like Vlastos, I see confirmation of this parallelism in fr. 69, ~ τωὐτὸν ἀγαθὸν καὶ ἀληθές . ἡδὺ δὲ ἄλλῳ ἄλλo, ἀνθρώπoις πασι

which I, however, interpret in a rather reduced sense. The essential points seem to me to be, firstly, the contrast between qualities inherent in an object irrespective of the observer and those which vary in different observationsituations (the νόμῳ–ἐτεῃ~ contrast), and secondly, the conjunction in the first clause of ἀγαθόν and ἀληθές , the objects respectively of practical activity and of cognitive reasoning. The thought appears to be this, that in the practical and theoretical spheres the same contrast applies between the state of affairs as immediately (and misleadingly) apprehended and the true state of affairs which can be grasped only through reflection, and that the man who takes immediate pleasure as the only or the chief guide to action is making the same sort of mistake as the man who takes the commonsense picture of the world as revealed by the senses as adequate. If we bear in mind the close association on the part of some earlier thinkers of pleasure and pain with αἴσθησις ⁵³ (probably undifferentiated between the senses of ‘perception’ and ‘sensation’), this parallelism will not seem particularly far-fetched. A similar strain of thought may be discerned in the Phaedo, especially at 81b–83d, where bodily pleasure and sense-perception are inextricably interwoven; the effect of relying on these is that one comes to believe that the things which cause pleasure and pain are ‘clearest and truest’ (83c7), and thinks that ‘whatever the body says is true’ (d 6); in ⁵³ Theophr. De sens. 16; DK 31 A 86 (on Empedocles & Anaxagoras).

pleasure, knowledge, and sensation 21 contrast, philosophy, in freeing the soul from the tyranny of the body, shows that ‘investigation by means of the eyes and ears and the other senses is full of deception’ (a4–5). For Democritus, the unreflective man believes that whatever αἴσθησις tells him, whether that this apple is sweet or that the pleasure of drinking is worth pursuing, is true. Rational reflection, however, shows that one should rely, not on any αἴσθησις , but (a) on perceptual judgements which verify the atomic theory, and (b) on the overall enjoyment of one’s life. The parallel must not be pressed too far, since it also appears that Democritus held, inconsistently with the above, that every perceptual judgement was true, whereas there is no indication from the ethical fragments that he held that every judgement of the worth of an individual action was in the same way true. The ethical theory is therefore saved, perhaps by its very lack of sophistication compared with the physical theory, from the self-refutation to which the latter eventually succumbs. Further confirmation of this parallelism may, I suggest, be derived from the remarks on criteria with which Sextus closes the account of Democritus to which we have already referred.⁵⁴ Citing as his authority a certain Diotimus,⁵⁵ he says that Democritus recognized three criteria: ~ μὲν των ~ ἀδήλων καταλήψεως τὰ ϕαινόμενα − − − ζητήσεως της ~ τὰ πάθη. δὲ τὴν ἔννoιαν − − − αἱρέσεως δὲ καὶ ϕυγης

The first clause gives exactly the sense outlined above; ‘the conception of things unseen’, i.e. a theory of the unobservable ‘real nature’ of things, is judged adequate or inadequate according to its ability to account for sensible phenomena. The πάθη, by which is apparently meant pleasure and distress, are the criteria of choice and aversion in that Democritus’ practical theory is a hedonistic one. Finally (following the hint given by Sextus’ reference to Plato, Phaedr. 237b7–c1) the criterion of the worth of an investigation is one’s conception of the nature of its object. The other two criteria represent the application of this methodological principle to theoretical and practical investigations respectively. Our conclusion, then, is that the view of Dyroff and Bailey that there is no connection, or only a loose connection, between Democritus’ physical and ethical theories cannot be upheld. On the other hand, the attempts of ⁵⁴ DK 68 A 111.

⁵⁵ See DK c. 76.

22

pleasure, knowledge, and sensation

Natorp and Vlastos to locate the connection in Democritus’ conception of the good for man, while not without inherent plausibility, appear not to be substantiated by the evidence cited in their support. Rather the connection is that both theories may be regarded as examples of an epistemological method in which unreflecting reliance on the data of sensation is replaced by reliance on a rational theory, which yet depends on sensation in that (a) the physical theory is subject to empirical verification and (b) the good for man is identical with pleasure in the sense of the enjoyment of life. Afterword pp. 10–11 For an illuminating discussion of Democritus’ classification of the properties of atoms and of the terminology in which he designated them see Mourelatos [2005]. p. 14 ‘logical difficulties about qualitative differentiation such as vitiated the similar attempt of Anaxagoras’. In Taylor [1997], ch. 6, pp. 213–15, I argue that in fact Anaxagoras’ physical theory allows him to give an adequate account of the qualitative differentiation of physical types.

2 ‘All Perceptions Are True’1 Epicurus is reported by Diogenes Laertius (X. 31) as having said that perceptions (or perhaps ‘the senses’ or ‘sense-impressions’, which are also possible meanings of Epicurus’s own term aisth¯eseis) are among the criteria of truth; this report is confirmed by two passages of Epicurus himself, at DL X. 50–2 and 147. By itself this need imply nothing more than the merest common sense; of course perception and the senses must have some role in determining what is the case and what is not, and hence which statements are true and which are false. But the matter is not so straightforward. For, firstly, aisth¯esis and related words are used in a wider range of contexts than ‘perceptions’ and its cognates: e.g. cases of hallucination are sometimes said to involve aisth¯esis (see below). Secondly, Epicurus is also said to have maintained the much more obviously controversial thesis that every aisth¯esis is true. In this paper I shall try to establish what he meant by those statements, to clarify the relation between them, and to consider their wider implications for his epistemological and physical theory. I turn first to the doctrine that all aisth¯eseis are true, for which the evidence has been helpfully collected by Gisela Striker². No version of it occurs in the texts of Epicurus himself, but the following doctrines, or versions of the same doctrine, are attributed to him by other writers. 1. Every aisth¯eton is true (Sext. M VIII. 9: Ep. ta men aisth¯eta panta elegen al¯eth¯e kai onta ... pant¯on de t¯on aisth¯et¯on al¯eth¯on ont¯on; 63: Ep. elege men panta ta aisth¯eta einai al¯eth¯e, kai pasan phantasian apo huparchontos einai ...). Cf. M VIII. 355, every aisth¯eton is reliable (bebaion). ¹ I am indebted to the participants in the 1978 Oxford conference for pointing out serious defects in the original version of this paper. I trust that I have profited from their criticisms. I am especially grateful to Gisela Striker, from whose publications on this topic I have learned a great deal, and to Malcolm Schofield and David Sedley for their detailed suggestions for improvements. ² Striker [1977].

24

‘all perceptions are true’ 2. Every aisth¯esis and every phantasia is true (Usener, no. 248: Aetius IV. 9. 5; Usener, p. 349, 5–6: Aristocles apud Eus. PE XIV. 20. 9). 3. Every phantasia occurring by means of aisth¯esis is true (Plut. Col. 1109 a–b). 4. Every phantasia is true (M VII. 203–4, 210). 5. Aisth¯esis always tells the truth (M VIII. 9: t¯en te aisth¯esin ... dia pantos te al¯etheuein; 185: m¯edepote pseudomen¯es t¯es aisth¯ese¯os). Cf. the passages in Cicero referred to by Striker, to the effect that the senses are always truthful.

It appears likely that most of these formulations differ from one another only verbally. Thus Sextus is the only writer cited above to use the term aisth¯eton (‘sense-content’³), and his use of the term strongly suggests that it is interchangeable with phantasia (‘appearance’). This appears particularly from M VII. 203–4, where the thesis that phantasia is always true is supported by a number of examples from the various senses, e.g. ‘The visible (horaton) not only appears (phainetai) visible, but in addition is of the same kind as it appears to be’, which are summed up in the words ‘So all phantasiai are true.’ Here, then, what is true of aisth¯eta is taken to be true of phantasiai as a whole; further evidence that Sextus regards the terms as coextensive is given by M VIII. 63–4, where he represents Epicurus as counting Orestes’ hallucination of the Furies as a case of aisth¯esis, and therefore as true. Again, in the passage from Aristocles cited above, quoted by Eusebius, aisth¯esis is treated as interchangeable with phantasia, since the thesis introduced by means of both terms (i.e. 2 above), is expressed in the course of the passage firstly as the thesis that every aisth¯esis is true and then as the thesis that every phantasia is true. If these writers treat aisth¯eton and phantasia as strictly coextensive terms, they misrepresent Epicurus, who distinguishes phantasiai of the mind (e.g. appearances in dreams) from phantasiai of the senses (DL X. 50–1). In strict Epicurean doctrine, then, aisth¯eta are a species of phantasiai. But the misrepresentation is not crucial, ³ I use the term ‘sense-content’ as roughly equivalent to the Greek aisth¯eton, to indicate the informational content of a perceptual or quasi-perceptual act, without commitment to the objective reality of that content. For example, if some honey tastes sweet to me, I have an aisth¯eton of the sweet, whether or not the honey really is sweet. I interpret Epicurus as holding that in every instance of aisth¯esis we have some sense-contents, as regarding hallucinations as instances of aisth¯esis, and as saying of cases of perceptual illusion such as that of the oar’s looking bent that the sense-content is true, though some belief which we may form as a result of having that sense-content may be false.

‘all perceptions are true’ 25 since Epicurus clearly holds that both sensory and non-sensory phantasiai are always true (ibid.: for the Epicurean view of the ‘truth’ of dreams, see below). In the passages from Sextus cited under 5 above, where aisth¯esis is said always to tell the truth and never to lie, it is possible to render the word as ‘sense’ (equivalent to ‘the senses’), ‘perception’, ‘sensation’ (i.e. the faculties thereof), ‘the (particular act of) perception’, or ‘the (particular) sensation (occurring in the perceptual context)’. But however we render it, the thesis that aisth¯esis always tells the truth is presented either as following immediately from the central thesis that all aisth¯eta are true, or as entailing it, or as restating it. The precise logical relation of the two theses (if indeed they are two) is impossible to determine from these passages; by the same token, their intimate logical interconnection is displayed by both. Our evidence, then, indicates that ancient writers regularly attribute to Epicurus or the Epicureans the doctrines that every phantasia is true and that every aisth¯eton is true. Though strictly aisth¯eta are a species of phantasiai, differentiated by their causation via the sense-organs, some later sources appear to make no distinction between the terms. In some reports the doctrine that every aisth¯eton is true appears to be expressed as ‘Every aisth¯esis is true’; in others, where aisth¯esis may mean ‘sense’ or ‘faculty’, the thesis that aisth¯esis always tells the truth is inextricably interwoven with the thesis that all aisth¯eta are true. We find, then, in the texts of Epicurus the doctrine that sense-contents are among our criteria for discriminating what is true from what is false, and we find ascribed to Epicurus by later writers the doctrine that all sense-contents are true. This ascription is confirmed by Epicurus’ remarks at DL X. 50–2, though the doctrine is not explicitly stated there. Some of these writers clearly regard Epicurus as holding that the former doctrine implies the latter, and as maintaining the latter because he was committed to it by the former. Thus Aristocles (loc. cit.) says that people who hold that every aisth¯esis and every phantasia is true do so from the fear that, if they said that any aisth¯esis were false, they would have no criterion nor any sure or reliable standard (sc. of truth). Similarly Plutarch interprets the Epicurean thesis that all phantasiai arising by means of aisth¯esis are true as leading to the consequence that all perceptible objects are a mixture of all the qualities which they seem to have in any conditions of perception; the Epicureans accept that, he says, because if they didn’t they admit that their standards would go to pot and their criterion altogether disappear (Col.

26

‘all perceptions are true’

1109a–e). Did Epicurus in fact hold that if aisth¯eseis are among our criteria of truth, then aisth¯eseis must themselves be true? And if he did, can we see why? This second question may appear surprising. To some people it may just seem obvious that any criterion of truth must itself be true, and anyone who thinks so will naturally assume that it must have been obvious to Epicurus also. But so far from being obvious, it is not even true. By ‘criterion’ we understand ‘something used in discriminating, judging, or determining’.⁴ In general, it is plainly false that, whenever we distinguish F from G (where ‘F’ and ‘G’ stand in for names of properties), that by means of which we distinguish is either F or G. This may be so in some cases, e.g. we may discriminate straight from crooked lines by applying a straight ruler (kan¯on). But, while it is true that Epicurus used the term kan¯on as equivalent to krit¯erion (apud DL X. 129: his work dealing with criteria of truth was entitled Kan¯on, DL X. 31), it would be gratuitous to suppose that he reached his doctrine by generalizing from this single instance in the face of the multitude of counter-instances where the terms kan¯on and krit¯erion are equally at home. Thus, in the required sense of ‘criterion’ taste is the criterion of sweet and sour, but taste (i.e. the sense of taste) is itself neither sweet nor sour; sight is the criterion of black and white, but is itself neither black nor white. The general point holds also for the special case of distinguishing truth from falsity. Thus sometimes we find out the truth about some matter by looking, but it does not make sense to say that looking is true (or that it is false). Thus to attribute to Epicurus either the general assumption that any criterion of F-ness must itself be F, or the special assumption that any criterion of truth must itself be true, is to attribute to him a fairly obvious falsehood. There is, however, no need to attribute either assumption to him. Rather, the evidence indicates that he held that all aisth¯eseis must be true because of the particular sort of criterion that aisth¯eseis are, and that his reasons for holding that a criterion of that sort must itself be true were better than the untenable assumptions which we have just dismissed. The sort of criterion in question is evidence. Aisth¯eseis are used in discriminating truth from falsehood in that they provide evidence on the basis of which we judge (krinomen) what is true and what false, just as the ⁴ See Striker [1974].

‘all perceptions are true’ 27 evidence of witnesses in a court is used by the judge to determine the truth of the matter in dispute. And just as the evidence of the witnesses must itself be true in order that a sound verdict be arrived at, so the evidence of aisth¯eseis must be true if we are to attain to knowledge of the world. The choice of forensic terminology is not, of course, accidental, but rather reflects Epicurus’ own descriptions of the role of aisth¯eseis. They are treated as witnesses in a legal action, whose word may be challenged or accepted, and whose evidence may be used to establish the claims of other parties as true, or convict them as false. So at DL X. 146: ‘If you fight against all aisth¯eseis, you will not have anything by reference to which you can pass judgement on those which you say have spoken falsely.’ Here aisth¯eseis say things, and the sceptic is inclined to condemn them as lying witnesses. (Anag¯og¯e, ‘referral’, can also have a legal sense, that of referring a dispute to an arbitrator, as in a third-century treaty from Delphi.⁵) The forensic analogy is reflected in the Epicurean terminology for the process by which beliefs are confirmed or refuted by the evidence of aisth¯eseis; a belief is true if it is ‘witnessed for’ (epimartureitai) or ‘not witnessed against’ (ouk antimartureitai), false if it is witnessed against or not witnessed for (DL X. 50–1; cf. M VII. 210–16). Aisth¯eseis, then, are reliable witnesses; nothing can convict them of falsehood (dielengksai). Since what they say is true, we may use it as evidence (s¯emeia) from which to arrive by inference at conclusions about what we are not directly aware of (ibid. hothen kai peri t¯on ad¯el¯on apo t¯on phainomen¯on chr¯e s¯emeiousthai; cf. DL X. 38, 87, 97, 104). It is, then, the forensic analogy of the evidential role of aisth¯esis which commits Epicurus to the doctrine that aisth¯eseis are true, rather than the implausible assumptions which I dismissed earlier. That analogy itself does not support the thesis that all aisth¯eseis must be true: judicial procedures depend on testimony, though witnesses frequently lie, misremember etc. The stronger thesis rather depends on the basic character of the evidence of aisth¯eseis. Aisth¯eseis have to be true if we are to have any knowledge or well-founded beliefs about the world; for all such knowledge and belief is inferential, justified ultimately by appeal to the evidence which aisth¯eseis provide. If we challenge any piece of that evidence itself, we can do so only by appeal to further evidence of the same kind, but we can give no reason ⁵ E. Schwyzer, Dialectorum Graecorum Exempla Epigraphica Potiora (Leipzig, 1923), 328a, II A 17.

28

‘all perceptions are true’

why we should prefer one piece of evidence to another. Hence the only alternative to total scepticism is undifferentiated acceptance of aisth¯eseis as true (DL X. 146–7). Epicurus might then seem to present an early version of the familiar empiricist doctrine of the foundations of knowledge; sensecontents provide the immediate, i.e. non-inferential data by inference from which we arrive at, or at least justify, beliefs about physical objects, other persons etc. But in fact the empiricist doctrine aligns rather awkwardly with the forensic analogy; for on the empiricist view, as represented by Locke or Berkeley, we have immediate and complete knowledge of our sense-contents themselves, and proceed to more or less risky inferences from that knowledge to knowledge and/or belief about the external world. In terms of the forensic analogy, the fundamental empiricist thesis that we have immediate and incorrigible knowledge of our sense-contents (‘ideas’ in some versions, ‘sense-data’ in others) would surely have to be expressed as follows, that our sense-contents are totally reliable witnesses as to what they themselves say. So if a sense-content says ‘My evidence is as follows: p’, there can be no doubt that that is what its evidence is, namely that p. But what we want from a witness is not just to be sure what his evidence is, but in addition to be sure that his evidence is true. We want a guarantee not merely that he says that p, but that it is true that p, and it is clear that the empiricist thesis that we have incorrigible knowledge of what his evidence is cannot provide such a guarantee. So if Epicurus’ thesis that all aisth¯eseis are true is interpreted just as the thesis that we cannot be mistaken about what aisth¯eseis we have, it cannot do justice to the evidential role which he assigns to aisth¯eseis. This brings me to the central question of this paper, viz. what did Epicurus mean by his thesis that all aisth¯eseis are true? We have seen that a superficially plausible account of that thesis, viz. that we have incorrigible awareness of our sense-contents, is insufficient to account adequately for the fundamental status of the thesis in Epicurus’s epistemology. We might, of course, be forced to conclude that Epicurus failed to recognize the inadequacy; but at least it provides us with a motive to look for another interpretation which will represent the thesis as complying more closely with Epicurus’ intentions. In this connection we must take notice of two connected points which I have hitherto ignored. Firstly, I have taken it for granted that Epicurus’ thesis was that all aisth¯eseis are true, and have raised the question how that is to be understood. But, as is well known,

‘all perceptions are true’ 29 al¯eth¯es may in context mean ‘real’ rather than ‘true’, or may be ambiguous between the two senses. Moreover, there is some evidence to suggest that the Epicurean thesis may have been that all aisth¯eseis are in some sense real. Secondly, Epicurus appears to have insisted that every aisth¯esis, in contrast with belief, is alogos. Whatever that means precisely, there is a prima facie inconsistency in the claim that what is lacking in logos is nonetheless true, since truth and falsity surely apply in the first instance to logos or logoi, to speech or things said. This difficulty might be thought to support the suggestion that al¯eth¯es should in this context be understood as ‘real’ rather than as ‘true’. My next task, therefore, is consideration of that suggestion; in the course of that consideration I hope to explain what Epicurus had in mind in describing aisth¯eseis as alogoi. The suggestion is derived from two ancient sources, DL X. 32 and Sext. M VIII. 9. Diogenes’ statement comes at the end of his summary, paraphrased above, of the role of aisth¯eseis as a criterion. After describing how thoughts (epinoiai) are formed by the operation of various mental processes on aisth¯eseis, he adds ‘And the visions (phantasmata) of madmen and those which occur in dreams are al¯eth¯e, for they have effects (kinei); but that which is not has no effect’ (to de m¯e on ou kinei). Here the thought seems plainly to be that only something real can have any effect on anything. But dreams and visions manifestly have effects, most obviously on the behaviour of those subject to them. So dreams and visions are real things. Sextus is even more explicit. ‘Epicurus’, he says, ‘said that all aisth¯eta are al¯eth¯e and things that there are (onta); for he made no distinction between calling something al¯ethes and calling it existent’ (huparchon). It is noteworthy that both contexts also contain a reference to the thesis that aisth¯esis is alogos. In Diogenes this occurs in 31 immediately after the list of criteria, including aisth¯eseis. Diogenes’ words are ‘For every aisth¯esis, he (i.e. Epicurus) says, is alogos and receptive of no memory; for it is not stimulated (kineitai) by itself, nor, being stimulated by something else, can it add or take away anything.’ Sextus’ report is similar: ‘And aisth¯esis, which is receptive of the things presented to it, neither taking away nor adding nor altering anything, since it is alogos, always tells the truth and receives what there is just as it is in reality (kai hout¯o to on lambanein h¯os eiche phuse¯os auto ekeino). But while all the aisth¯eta are al¯eth¯e, the things we believe (ta doxasta) differ, and some are true and some false ...’ It is tempting to offer the following explanation of the thesis that aisth¯esis is al¯eth¯es and of its connection with

30

‘all perceptions are true’

the doctrine that aisth¯esis is alogos. The thesis is simply the thesis that every instance of aisth¯esis, including dreams and hallucinations under aisth¯eseis since those states consist in the reception of impressions similar to those of perception proper, is a real event, consisting in the reception by the physical organism of a pattern of physical stimulation (in terms of the atomistic theory of mental functioning, a bombardment of large numbers of eid¯ola in very rapid succession). Aisth¯esis is alogos in that aisth¯esis is the purely physical event of the reception of stimulation; being purely physical it lacks any mental content, and hence cannot be true or false. It merely reproduces the external stimulation without addition or subtraction, as the wax reproduces the impression of the signet which is pressed into it. Truth and falsity belong to the judgements which we make about the physical process of stimulation, characterizing it as the perception of this or that. This interpretation is tempting insofar as it explains how aisth¯eseis could sensibly be described both as al¯etheis and as alogoi, insofar as it does justice to the evidence which suggests that al¯etheia should be taken in this context as ‘reality’, and insofar as it frees Epicurus’ theory from such absurdities as the claim that the madman’s hallucinations are true. But these advantages are bought at too high a price. This account suffers from the basic defect of having altogether abandoned the evidential role of aisth¯eseis. For on this account aisth¯eseis themselves don’t say anything; they are purely physical events whose occurrence is reported, correctly or incorrectly, in thought. But a witness who says nothing is no witness; and if anything in this area is clear, it is that Epicurus regards aisth¯eseis as witnesses to the reality of things. Further, this interpretation is inadequate even with respect to those very passages adduced in its support. The identification of its inadequacy will, I hope, lead us to a more correct understanding of the thesis that all aisth¯eseis are al¯etheis. If we look again at M VIII. 9, we see that Sextus’ first point is that Epicurus made no distinction between ‘true’ and ‘real’; consequently he defined ‘true’ as ‘that which is as it is said to be’ and ‘false’ as ‘that which is not as it is said to be’. This amounts to a conflation of the notions of ‘true’ and ‘real’, and of ‘false’ and ‘unreal’. ‘True’ and ‘false’ apply properly to what is said or thought; consequently an account of truth along these lines ought to take the form of the Aristotelian formula ‘Truth is saying of what is that it is, and of what is not that it is not’ (Meta. 1011b 26). ‘Real’ and ‘unreal’, on the other hand, characterize entities or states of affairs as

‘all perceptions are true’ 31 in the Epicurean formula, but characterize them independently of what is thought or said. The conflation of the two pairs of notions in the Epicurean formula suggests that Epicurus is not interested in any contrast between a report and what it reports, but in the systematic correlation between a true report and the reality of what it reports. Consequently Sextus’ report moves immediately from the claim that all aisth¯eta are al¯eth¯e kai onta to the assertion that aisth¯esis always tells the truth, since it represents the reality which impinges on it exactly as it is, without addition or diminution, the whole being finally summed up as ‘While all the aisth¯eta are true, some things that we believe are true and some false.’ It is clear that the contrast, central to the interpretation we are considering, between purely physical aisth¯esis, which is reported, and thought, which reports it, is inconsistent with Sextus’ evidence in this passage. For aisth¯esis itself, Sextus tells us, reports the physical stimulation which gives rise to it. The special status of aisth¯eseis as witnesses is due to the fact that the reports which aisth¯esis gives represent that physical stimulation with perfect accuracy. And the accuracy of those reports is guaranteed by the fact that aisth¯esis is alogos. What this seems to mean is that aisth¯esis lacks the capacity to form any judgement about the pattern of stimulation presented to it; its function is restricted to representing it as it is. But that representation has a content to which truth and falsity are applicable; the special feature of that representation is that in fact it is always true. We might perhaps be inclined to identify aisth¯eta with the physical stimulation, which is real, as distinct from aisth¯esis, which reports that stimulation, and which is true. But that would be to impose on the passage a sharper distinction than the text warrants. In view of the conflation of the concepts of truth and reality which occurs in this passage (see above), it is much more plausible that no clear distinction is drawn between the putative bearers of those predicates, aisth¯esis and aisth¯eton. This is supported by the evidence presented earlier that Sextus uses the term aisth¯eton interchangeably with phantasia. The latter term denotes an act of mental representation or the content of such an act, ‘something’s appearing so and so’. It is, therefore, unlikely that in this passage Sextus uses the term aisth¯eton to denote a purely physical entity, in contrast with a kind of mental representation. This leaves us with the evidence of DL X. 32. Now it is undisputed that the point there is that the phantasmata of madmen and dreamers are real, since they have effects, and only real things can have effects. But

32

‘all perceptions are true’

this is ambiguous as it stands. For it is unclear whether what is meant is that the states of having a hallucination or having a dream are real states, since those states have effects, e.g. they produce movements and utterances on the part of the people who have them, or that the objects present to the mind in those states are real objects, since those objects have effects, e.g. they induce beliefs on the part of those to whom they are present. The use of the word phantasmata, rather than phantasiai, might suggest the latter, but the lack of any systematic distinction between terminology for mental acts and that for objects of such acts makes it unsafe to rely on the terminological point alone. The evidence of two further passages dealing with hallucinations does indeed support the latter interpretation against the former, but indicates at the same time that Diogenes’ very compressed report omits a central feature of Epicurus’ view of such phenomena. The passages in question are Plut. Col. 1123b–c and Sext. M VIII. 63–4. Plutarch chides Epicureans for holding that none of the ‘sights and extraordinary things’ which occur in dreams and visions is ‘an optical illusion or false or insubstantial’ (parorama ... oude pseudos oude asustaton), but that they are all phantasiai al¯etheis and physical objects (s¯omata) and shapes which come from the surrounding environment. Here the conflation between act and object is very marked: a class of phenomena is said to be neither false nor insubstantial, but actually to consist of physical objects and real/true phantasiai. It is clear that the physical reality of the objects of dreams and visions is a central Epicurean dogma. Like the objects of any other mental processes, these objects are a sort of physical objects, viz. films of atoms emitted from the surfaces of three-dimensional aggregates of atoms. But the conflation between act and object apparent in this passage should prompt us to ask whether the Epicurean thesis that such objects are al¯etheis was simply the thesis that they are real (i.e. physical) objects. Is it not likely that Epicurus also held some view about the way these items were represented in the aisth¯esis of the dreamer or the visionary, a view which was also expressed in the thesis that those objects are al¯etheis? The passage from Sextus clearly indicates that this is the case. ‘Epicurus,’ he says, ‘said that all the aisth¯eta are al¯eth¯e, and that every phantasia arises from something that there is, and is just as the thing which stimulates the aisth¯esis’ (toiaut¯en hopoion esti to kinoun t¯en aisth¯esin). This sentence gives us both aspects of Epicurus’ theory of the al¯etheia of aisth¯eseis: the things

‘all perceptions are true’ 33 which stimulate aisth¯esis are real, physical things, namely eid¯ola, and they are represented in aisth¯esis exactly as they are. We should recall the evidence presented earlier that aisth¯esis is incapable of altering in any way the things which it represents. This general theory of the al¯etheia of aisth¯esis is then applied to what would ordinarily be considered a case of false aisth¯esis, viz. Orestes’ hallucination of the Furies. ‘So in the case of Orestes ... his aisth¯esis which was stimulated by eid¯ola was al¯eth¯es (for the eid¯ola were there), but his mind believed falsely when it thought that the Furies are solid objects.’ Given the general theory just stated the explanatory clause ‘for the eid¯ola were there’ (hupekeito gar ta eid¯ola) should be understood as ‘for the eid¯ola which stimulated his aisth¯esis were there just as they were represented in his aisth¯esis.’ We see, then, that Diogenes’ testimony of the reality of dreams and visions is seriously inadequate. The phantasmata, i.e. things appearing to the mind, which occur in those states are al¯eth¯e not merely in that they are real physical objects, but also in that they really are exactly as they are represented in aisth¯esis. As we should expect, the reality of the things represented is inseparable from the truth of the representation. Diogenes’ evidence is, therefore, insufficient to support the suggestion that the Epicurean doctrine that aisth¯eseis are al¯etheis merely reduces to the claim that aisth¯eseis are real, physical events. On the contrary, given Epicurus’ epistemological requirements it is an indispensable part of that doctrine that aisth¯esis is not merely a response to physical reality but a faithful representation of that reality. That this is a general feature of aisth¯esis, and not merely a feature of the special cases of dreams and hallucinations, is amply attested by the sources. In addition to the evidence of Sextus cited earlier (M VIII. 9) to the effect that aisth¯esis is ‘receptive of the things presented to it, neither taking away nor adding nor altering anything’, we may cite M VII. 210, ‘It is proper to aisth¯esis merely to receive what is present and stimulates it, e.g. colour, but not to discriminate that the thing here is one thing and the thing here another. Therefore all the appearances which arise in this way are true, but the beliefs are not all true ...’ (adopting Usener’s supplementation of the text, p. 181). The obscure expression ‘not to discriminate that the thing here is one thing and the thing here another’ refers to the application of this theory to phenomena of perceptual illusion, perspective etc. The large rectangular tower looks small and round when seen from a distance, because when the eid¯ola reach the eye they actually are small and round, the large, rectangular

34

‘all perceptions are true’

eid¯ola emitted by the tower having been reduced in size and altered in shape by friction with the atoms in the space between the tower and the observer. Aisth¯esis faithfully reproduces the actual state of the eid¯ola when they reach the eye, but it is no part of the business of aisth¯esis to distinguish the small round eid¯olon which you get when you look at the tower from here (i.e. from a distance) from the large rectangular one which you get when you look at it from here (i.e. from close up). That is the work of opinion, not of h¯e alogos aisth¯esis; opinion is distorted when it gives rise to the belief that the same object is visually presented (phantaston) under both conditions, and is hence mispresented when one looks from a distance.⁶ Two passages from Plutarch’s Adversus Colotem, dealing with different kinds of perceptual phenomena, present essentially the same picture. In 1121a–b he presents the Epicurean theory of illusion in much the same way as Sextus; when the rectangular tower looks round in the distance, and the straight oar in water looks bent, then the sense-organ comes into contact with an eid¯olon which is round, and one which is bent, and aisth¯esis receives an accurate imprint of that (t¯en ... aisth¯esin al¯eth¯os tupousthai). At 1109a–e the phenomena are those of perceptual relativity, e.g. that the same water feels warm to one person and cold to another. Here the explanation is once again that each person’s experience faithfully reproduces the actual character of the eid¯ola which he receives; the water itself contains atoms of both kinds, which are registered by observers whose sense-organs are in the appropriate condition to receive them. Disagreement on whether the water really is warm or cold reflects failure to distinguish the eid¯ola which make contact with the sense-organs from the aggregates of atoms which emit them. So far, then, the evidence suggests that the Epicurean thesis that all aisth¯eseis are al¯etheis is not to be interpreted as the thesis that we have incorrigible acquaintance with our sense-contents. Nor is it merely the thesis that all aisth¯eseis are real, either in its trivial version, viz. that every occurrence of aisth¯esis is a real event, or, in the more substantial version, viz. ⁶ This seems to give the best account of what is meant by ‘discriminating that the thing here is one thing, and the thing here another’. Other suggestions, e.g. that Sextus is saying that while aisth¯esis registers that this is, e.g. green and this blue, it is not aisth¯esis but logos which identifies the green thing as grass and the blue thing as the sky, introduce ideas irrelevant to the purpose of the passage, which is to give the Epicurean explanation of the source of the (erroneous) belief that misperception occurs. This source is the failure to distinguish distinct perceptual objects in distinct conditions of perception, a failure attributable to opinion, not to perception itself.

‘all perceptions are true’ 35 that every instance of aisth¯esis consists in the stimulation of the sense-organ by a real, physical object. Rather, it is the thesis that every instance of aisth¯esis consists in the stimulation of the sense-organ by a real object which is represented in aisth¯esis exactly as it is in reality. If we need further evidence in favour of that interpretation, and against the ‘incorrigible acquaintance’ interpretation, it is provided by the context of the first passage from Plutarch just mentioned, Col. 1120c ff. Here Plutarch is dealing with Colotes’ critique of the scepticism of the Cyrenaic school; he represents him as rejecting, and indeed as making fun of the fundamental sceptical thesis that, while we do indeed have incorrigible acquaintance with our own sense-contents, that amounts merely to knowledge of an internal state, providing no ground for any justifiable inference about anything external to the percipient. Plutarch states this position in its own technical terms: the percipient is ‘sweetened’ when something tastes sweet to him, ‘embittered’ when something tastes bitter to him, ‘illumined’ when he seems to see light etc., ‘since each of these experiences has its own particular incontestable clarity in itself’ (t¯on path¯on tout¯on hekastou t¯en enargeian oikeian en haut¯oi kai aperispaston echontos).⁷ Colotes apparently made fun of this curious terminology, pointing out that the sceptic ought not to say that something was a man or a horse, but rather that he himself was ‘manned’ or ‘horsed’, i.e. was in a state of seeming to see a man or a horse. Plutarch attempts to show that Colotes’ own position does not differ from the one which he is attacking; he urges that just as the sceptics admit infallible knowledge of sense-contents but do not allow any knowledge of anything beyond them, so the Epicureans must admit that, while they have knowledge of the eid¯ola which impinge on them, they can have no knowledge of the external objects which, they claim, emit those eid¯ola. This is undeniably a difficulty for Epicurean epistemology, as for any representational theory of perception, but the point which concerns us here is that Plutarch’s attempt to assimilate the Epicurean position to the sceptical one itself throws into relief the basic difference between the two. For the sceptic’s starting point (and his finishing point too, for that matter) is knowledge of one’s own perceptual states, ‘perceptual sweetening’ etc., whereas the Epicurean starts from direct acquaintance with physical objects impinging on the senses. The sceptic declares insoluble the problem of justifying the inference from ⁷ Cf. Sext. M VII. 190–8, PH I. 10.

36

‘all perceptions are true’

descriptions of perceptual states to statements about external objects. For the Epicurean, on the other hand, descriptions of perceptual states are already descriptions of a percipient in contact with the physical world. The inference which he has to justify is not that from experience to the external world, but the inference from descriptions of eid¯ola to descriptions of their causes. This is put beyond doubt by Plutarch’s description of the Epicurean position: When a round eid¯olon comes into contact with us, or again a bent one, they say that our aisth¯esis is truly imprinted, but they don’t allow us to affirm as well that the tower is round and the oar bent; they guarantee their experiences and appearances (phantasmata) as reliable, but are not willing to admit that the external things are that way. And just as they [i.e. the sceptics] ought to say ‘horsed’ and ‘walled’, not ‘horse’ and ‘wall’, so these people have to say that their sight is ‘rounded’ and ‘bended’, not that the oar is bent or the tower round. For the eid¯olon by which the sight is affected is bent, but the oar from which the eid¯olon comes is not bent. (1121a–b)

The evidence which I have presented seems to me to establish my contention that the Epicurean thesis that all aisth¯eseis are al¯etheis is not the familiar empiricist axiom that we have complete and incorrigible acquaintance with our sense-contents, as expressed e.g. by Hume (Treatise I. iv. 2): ‘Since all actions and sensations of the mind are known to us by consciousness, they must necessarily appear in every particular what they are, and be what they appear’. I have not, of course, argued that Epicurus denied that axiom, or maintained anything inconsistent with it; I have merely argued that that is not what he means by ‘All aisth¯eseis are al¯etheis.’ But our incidental treatment of the thesis that aisth¯esis is alogos has shown that Epicurus accepted another dogma of empiricism, viz. the view that, in contrast with its active role in forming concepts and making judgements, in perception the mind is passive, merely reproducing data which are presented to it. A paradigm expression of this view is provided by Locke (Essay II. i. 25): These simple ideas, when offered to the mind, the understanding can no more refuse to have, nor alter when they are imprinted, nor blot them out and make new ones itself, than a mirror can refuse, alter, or obliterate the images or ideas which the objects set before it do therein produce. As the bodies that surround us

‘all perceptions are true’ 37 do diversely affect our organs, the mind is forced to receive the impressions; and cannot avoid the perception of those ideas that are annexed to them.

Allowing for the fundamental difference that Locke is talking about the passive reception of mental impressions, whereas the Epicureans were concerned with the passive reception of physical eid¯ola, this passage is strongly reminiscent of the sources for the Epicurean doctrine that aisth¯esis is alogos. All our sources for that doctrine (DL X. 31, Sext. M VII. 210, VIII. 9, Aristocles, loc. cit.) associate it closely with the assertion that aisth¯esis cannot alter the data presented to it in any way, but merely reproduces them as it were photographically. The analogy of the camera is, though anachronistic, quite an apt expression of the Epicurean view. Their thought seems to have been that, like the camera, aisth¯esis cannot lie, since aisth¯esis puts no construction on what it ‘sees’ nor compares it with what it remembers (DL loc. cit. mn¯em¯es oudemias dektik¯e), but, like the camera, merely records what is before it. But it is precisely this passivity in the face of stimulation which gives aisth¯esis its evidential value; after all, what better evidence could there be that Brutus stabbed Caesar than a photograph of him stabbing Caesar? The paradox that aisth¯esis is both a witness and alogos is thus resolved. Aisth¯esis is alogos in that it can’t think about what it sees, and thus can’t misrepresent it. But it tells us what it sees, in just the way that a camera does, by presenting it, and thus has maximum evidential value. The photographic analogy has the further advantage of emphasizing a fundamental difficulty in the Epicurean doctrine that aisth¯eseis are a criterion of truth. For we are able to treat photographs as evidence only because we know how cameras work; we know what kind of casual process is necessary for the production of a photograph, and so are able to infer back from the image to the reality which must have produced it. Similarly, we are entitled to treat aisth¯eseis as reliable witnesses to reality only if we already accept the Epicurean account of the physical processes which give rise to them.⁸ For if we don’t already accept that perception consists in the physical impact on the sense-organs of films of atoms emitted by physical objects, and in the passive mirroring of those films in consciousness, then ⁸ This may perhaps be the force of the obscure statement (DL X. 32) that the reality of perceptions guarantees the truth of aisth¯eseis (to ta epaisth¯emata d’ huphestanai pistoutai t¯en t¯on aisth¯ese¯on al¯etheian).

38

‘all perceptions are true’

we have no reason to accept the fundamental premiss that aisth¯esis, being in the required sense alogos, must be inerrant. Why should aisth¯esis not process the physical stimuli which it receives, so that the same physical data are perceived differently by observers in different perceptual conditions? That suggestion does, after all, provide a perfectly plausible account of many cases of perceptual relativity, e.g. the same wine’s tasting sweet to the healthy man and sour to the sick man. One might respond that in that situation the observers have different data, since things do appear differently to them. But that is to interpret ‘data’ as ‘sense-contents’ instead of as ‘physical stimuli’; the Epicurean theory requires the latter interpretation, since the nerve of that theory is precisely that sense-contents exactly mirror the physical stimuli which excite them. If that claim is abandoned we have to retreat to the claim that we cannot be mistaken about what sense-contents we have. I hope to have shown that that is a retreat from the Epicurean position, not a restatement of it. It is no accident that the examples which our sources rely on in elucidating that position concentrate on the spatial properties of the eid¯ola, viz. size and shape. For the thesis that eid¯ola are represented in perception exactly as they are in reality is most readily applicable to just those properties. Eid¯ola are physical structures, composed of atoms whose only intrinsic properties are size, shape, weight, and ‘the necessary accompaniments to shape’ (e.g. propensity for certain sorts of motion: DL X. 54). Such structures themselves have objective size and shape, which may be reproduced faithfully or inaccurately in perception. But since they are composed of individual atoms which lack such secondary properties as colour, taste, and smell, and since they are individually imperceptible, a succession of eid¯ola being required to stimulate the sense-organ, they cannot themselves be said to possess colour, taste, or smell. So if the same wine tastes sweet to the healthy man and sour to the sick man, what becomes of the Epicurean thesis that the aisth¯esis of each man is true? It cannot sensibly be maintained that the healthy man’s eid¯ola really are sweet, and the sick man’s really are sour, for imperceptible things have no taste. Perhaps, then, the thesis is that the aggregate of eid¯ola which each receives really is as each tastes it. But here ‘really is’ seems to come to no more than ‘really tastes’, and that in turn to no more than ‘really tastes to him’, in which case the thesis that every aisth¯esis is true reduces to the triviality that whatever tastes etc. a certain way to someone really tastes etc. that way to him. As far as I

‘all perceptions are true’ 39 can see, the theory can be saved from this trivialization only by means of providing a theoretical identification of e.g. real sweetness as a phenomenal property associated with a specific physical structure S (e.g. a structure of smooth particles, Lucr. IV. 622–4). Then the thesis that the perception of sweetness is true would be the substantial thesis that, whenever anything tastes sweet to a percipient, that percipient is being stimulated by eid¯ola of structure S. The Epicurean theory certainly provides the materials for such identifications (see especially Lucretius’ account of hearing, smell, and taste, IV. 522–721): it is less clear that either Epicurus or his followers grasped the theoretical necessity for them. The Epicurean thesis that all aisth¯eseis are true provides, then, no independent support for the physical theory. Rather it presupposes that theory, in two ways. As regards the perception of shape and size, the theoretical account of the mirroring of eid¯ola in perception is required to explain the sense in which perception is true. As regards the perception of secondary qualities, the role of the theory is to identify a certain physical structure as what is truly perceived. The theory predicts that whatever is perceived as a phenomenal property, e.g. sweetness, will in fact be an instance of that structure; the truth of the perception consists in its fulfilling that prediction. In advance of the theory as a whole, the claim that all perception is true requires independent support. Epicurus seems to have thought that that support could be provided by epistemological considerations, in that the only alternative to the acceptance of the claim is total scepticism, which he regarded as an absurd or self-defeating doctrine. Thus at DL X. 146–7 he asserts that rejecting all aisth¯eseis together or any particular one alike lead to the abandonment of any criterion of what is true and false. It is implied that that is an unacceptable position, though it is not stated why it is unacceptable. Perhaps his thought is that scepticism itself requires the conceptual distinction between truth and falsehood, and that that distinction itself presupposes the ability to tell which things are true and which false, which in turn presupposes a means of telling, i.e. a criterion. That there was an Epicurean argument along these lines is proved by Lucr. IV. 473–9: Yet even if I were to concede that he [i.e. the sceptic] does know this [i.e. that he knows nothing], let me ask him this: since he has previously seen nothing true in things, how does he know what knowing and again not knowing are? What

40

‘all perceptions are true’

has produced his conception of the true and the false? What has proved that the doubtful differs from the certain? You will find that your conception of the true was produced by the senses first of all, and that the senses cannot be refuted.

Perhaps an expansion of this point, or perhaps a separate one, is the argument that reason cannot give grounds for rejecting the evidence of perception, since reason itself originates from or depends on perception. This is certainly Epicurean, as it is reported both in the same passage of Lucretius, ‘Moreover, what must be held to be more trustworthy than the senses? Will reason, arising from false sense, be strong enough to speak against them, reason which arises solely from the senses? Unless they are true, all reason also becomes false’ (482–5), and in the corresponding summary in DL X. 32: ‘Nor again can reason [refute aisth¯eseis], for all reason depends on aisth¯eseis.’ This is obviously a descendant of Democritus’ famous ‘Complaint of the Senses’ (frag. 125 DK), ‘O wretched mind, you receive your evidence from us, and do you overthrow us? Our overthrow is a fall for you’, and can be developed into an argument to show that scepticism is self-refuting. Epicurus had, then, some good arguments, or at least the materials of such arguments, which he could advance against scepticism without presupposing his physical theory. His method thus displays a subtle interaction of epistemological and metaphysical considerations. The fundamental epistemological requirement is that every aisth¯esis should be true, i.e. that whatever seems to be the case should in some sense or other actually be the case. It then becomes part of the task of the general theory of nature to specify the sense or senses in which what seems to be actually is. It is an astonishing achievement of atomism, both in its fifth-century and in its Epicurean version, to have provided an even reasonably plausible account of the satisfaction of this requirement as part of a comprehensive account of the world. But problems remain. For the sceptic can reasonably claim that the account of how things always are as they seem is, in the last resort, empty. For example, how is the claim that sweetness is always the taste of a structure of smooth atoms to be tested? Suppose microscopic examination revealed that in some cases the atoms were smooth, but in others spiky. If both the microscopic and the gustatory observations are, in the theoretical sense, ‘true’, then we have two sets of atoms instead of one. No doubt we could add to the theory a description of how a structure of smooth

‘all perceptions are true’ 41 atoms emits a structure of spiky ones, but the problem of verification arises there again, and so on at every level. The basic difficulty is that a theory of objective reality which is not subject to any constraint by experience must be empty of actual content.⁹ Epicurus could have avoided this difficulty only by abandoning his fundamental epistemological requirement and facing up to the sceptical challenge to find a way of discriminating veridical from non-veridical experience. The subsequent history of philosophy to the present indicates the formidable nature of that undertaking; the Epicurean alternative, though ultimately unsuccessful, was well worth exploring. ⁹ For discussion of the same problem in fifth-century atomism see chapter 1 of this volume.

3 Plato, Hare, and Davidson on Akrasia1 My starting point is Donald Davidson’s discussion of the problem of akrasia in ‘How is Weakness of the Will Possible?’² In Davidson’s formulation, the problem consists in the fact that three propositions, which taken individually are each persuasive (indeed he goes so far as to say that they seem selfevident, p. 95), appear to form an inconsistent triad. These propositions are (using Davidson’s formulation and numbering): P1. If an agent wants to do x more than he wants to do y and he believes himself free to do either x or y, then he will intentionally do x if he does either x or y intentionally; P2. If an agent judges that it would be better to do x than to do y, then he wants to do x more than he wants to do y; P3. There are incontinent actions. Davidson’s solution of the problem is to argue that these three propositions are not in fact inconsistent. Specifically, the incontinent agent (i.e. the man who, given a choice between doing x intentionally and doing y intentionally, does y though his judgement is in favour of doing x) does not judge that it would be better to do x than to do y. Rather, he judges that, all things considered, it would be better to do x than to do y, which, Davidson argues, is compatible with its not being the case that he judges that it would be better to do x than to do y. Thus while from P1 and P2 we may derive by syllogism the following proposition (true in Davidson’s view), P4. If an agent judges that it would be better to do x than to do y and he believes himself free to do either x or y then he will intentionally do x if he does either x or y intentionally, ¹ I am grateful to Jennifer Hornsby for her helpful comments.

² Davidson [1969].

plato, hare, and davidson on akrasia 43 P4 does not contradict P3. For given Davidson’s analysis of what it is that the incontinent man judges, P3 should be formulated as P3
. Some agent judges that, all things considered, it would be better to do x than to do y, and believes himself free to do either x or y, and does y intentionally. And given that there is no incompatibility between ‘A judges that, all things considered, it would be better to do x than to do y’ and ‘It is not the case that A judges that it would be better to do x than to do y’, P4 and P3
can both be true. In this paper I shall not directly discuss Davidson’s proposed solution. I merely remark that, even if his argument is correct, it does not seem to succeed in solving the problem. For if we have adequate grounds for accepting P2 as true, we surely have equally good grounds for accepting the following: P2
. If an agent judges that, all things considered, it would be better to do x than to do y, then he wants to do x more than he wants to do y. And given Davidson’s analysis of the incontinent man’s judgement, as expressed in P3
, we now have the triad P1, P2
, P3
, which is inconsistent. Davidson does not give reasons for believing P2; in fact his view of it is somewhat obscure, since he says both that P1–3 seem self-evident and that it is easy to doubt whether they are true in their form as stated (particularly P1 and P2: paraphrasing pp. 95–6). I interpret him as meaning that, while there may be questions as to the precisely correct wording, the general principles expressed in P1–P3 are certainly true (and would perhaps if precisely formulated be self-evident). Against this unargued acceptance (in principle) of P2, it is a sufficient ad hominem argument to assert that P2
has an equal claim to unargued acceptance. More seriously, those who incline towards accepting P2 do so on the general ground that there is some necessary connection between what an agent judges it better to do and what he wants more to do. Whatever the nature of that connection, it must hold as well for what an agent judges it better to do, all things considered, as for what he judges it better to do, sans phrase. For if practical judgements give expression to desires, then the judgement ‘All things considered, it is better to do x than y’ is a paradigm instance of a judgement fitted for that role. Or, on the other hand, if practical judgements are conceived of as generating

44

plato, hare, and davidson on akrasia

the desire for what is recognized in the judgement, how could ‘All things considered, it is better to do x than to do y’ fail to exercise that mysterious force? After all, Davidson regards that proposition as expressing the agent’s judgement in the light of all available considerations (pp. 110–11); and if that kind of judgement lacks the capacity to generate desire, then how could any judgement have that capacity? But Davidson’s approach to the problem is not merely (a) obscure and (b) likely to be unsuccessful even if capable of clarification. It is also misconceived. The solution to the problem lies in the approach rejected by Davidson at the beginning of his paper, viz. that of giving up one or more of the principles P1–P3. In the remainder of this paper I shall (i) argue that P2 is false, (ii) show why some philosophers have accepted P2 or some similar principle, (iii) attempt to show what was wrong with their reasons for accepting that principle, and (iv) enunciate and defend a principle which is sufficiently close to P2 to be confused with it, but which is (a) distinct from P2, (b) true, and (c) consistent with P1 and P3.

I Since P2 is a universal proposition, it will be shown to be false by a counter-example. Any typical instance of action against one’s better judgement provides such a counter-example. For example, consider a man who is tempted to sleep with someone else’s wife, has the opportunity to do so, but judges that all things considered, it would be better not to sleep with her than to do so. Then, giving in to temptation, and acting against his better judgement, he sleeps with her. It seems to me plainly true of this man both that he judges it better not to sleep with the woman than to sleep with her, and that it is not the case that he wants not to sleep with her more than he wants to do so. (In fact, though this is irrelevant to the status of the case as a counter-example, he wants to sleep with her more than he wants not to do so.) We have thus satisfied the requirement of producing a counter-example which shows P2 to be false. But, it will be objected, this is to go too fast. No one who, like Davidson, accepts both P2 and P3 will agree that this case is a counter-example to P2, since this sort of case is just the kind of case which P3 admits. Such a philosopher, then, maintains that the case of the akratic man described

plato, hare, and davidson on akrasia 45 above is not a case of an agent who both judges that doing x is better than doing y and who does not want to do x more than he wants to do y. That is to say, he maintains either that the akratic man does not judge that doing x is better than doing y, or that he wants to do x more than he wants to do y. Let us consider these positions in turn. The first position is a highly paradoxical one for someone to hold who also believes that there are akratic actions, for the phenomenon of akrasia is nothing other than that of action against one’s better judgement. Davidson’s actual position is, indeed, a version of this. The akratic man does not judge that doing x is better than doing y: rather he judges that, all things considered, doing x is better than doing y. But, as I have already pointed out, this provides no defence against the counter-example, since Davidson is committed to P2
, If an agent judges that, all things considered, it would be better to do x than to do y, then he wants to do x more than he wants to do y, and the instance of akrasia described above will serve as a counter-example to this principle too. If Davidson wishes, as he must, to resist the claim that this is a counter-example, he cannot now maintain that the agent does not judge that, all things considered, it would be better to do x than to do y, since on his own hypothesis that is just what the akratic agent does judge. Hence he must maintain that he wants to do x more than he wants to do y. That is, this attempted defence of the first position (the only one with any appearance of plausibility), has turned out to depend on the second position, which I shall now consider. According to this position, the man in my example, who succumbs to the temptation of adultery while judging it better not to do so, wants not to sleep with the woman more than he wants to sleep with her. It is important to keep in mind first of all that the desires in question are desires relative to one particular action, viz., the desire to sleep with (and the desire to refrain from sleeping with) this woman on this occasion. For it may well be the case that the man wants (dispositionally, and as a long-term aim) to be the sort of man who avoids adultery more than he wants (dispositionally, and as a long-term aim) to be a Don Juan, this being manifested by his general pursuit of the former aim and his general eschewing of the latter (allowing for occasional lapses). But this is simply irrelevant to whether he wants not to sleep with her now more than he wants to sleep with her

46

plato, hare, and davidson on akrasia

now. Again, to avoid any possible irrelevance, it is not in dispute that the akratic man does want both to commit adultery with her now and to avoid it: characteristically, the akratic situation is one of conflicting desires. The question, to repeat at tedious length, is which of these particular things he wants more than the other. What then decides the question whether A wants to do a particular thing x more than he wants to do a particular thing y? One way of answering this question is by observing which of the two A actually does: there is one use of the expression ‘ ... wants ... more than ... ’ according to which, given the choice between doing x and doing y but not the possibility of doing both, A wants to do x more than he wants to do y, iff A does x and does not do y. In this use ‘ ... wants ... more than ... ’ is equivalent to ‘ ... prefers ... to ... ’; the question of what A prefers to do is settled by consideration of what he does, irrespective of whether he acts eagerly or reluctantly. The only requirements are that A should act freely and intentionally. Davidson’s P1 gives a weaker version of this criterion. Given either version, if A intentionally did y in preference to x it follows that he did not want to do x more than he did y. Since Davidson presents P1 as a principle which is true generally, and not merely given a particular use of the expression ‘ ... wants ... more than ... ’ he leaves himself defenceless against the counter-example I have given. But more detailed consideration of the phenomena of wanting shows that Davidson has a defence, albeit not one which can ultimately save him. By contrast with the use mentioned above, there is another use of the expression ‘ ... wants ... more than ... ’ where the agent’s choice of x in preference to y is neither necessary nor sufficient for the truth of ‘A wants x more than y’. Given this use, what is decisive is the agent’s attitudes to x and y, e.g. whether he is eager to do one, reluctant to do the other, whether the thought of one or the other fills him with enthusiasm etc. With this conception of wanting in mind (which we might call wanting as inclination to distinguish it from wanting as preferential choice) there is no contradiction in such a statement as ‘Of course I wanted to go fishing much more than I wanted to go to the meeting, but all the same I chose to go to the meeting, much against my inclinations’. Davidson’s P1 is false given this conception of wanting and may therefore be taken as restricted in its application to wanting as preferential choice. This would then allow Davidson to say that from the fact that A intentionally did y in preference

plato, hare, and davidson on akrasia 47 to x it does not follow that he wanted to do y more than he wanted to do x, provided that ‘wanted to do y more than he wanted to do x’ is here understood as ‘had a stronger inclination to do y than he had to do x’. Hence Davidson could still hang on to his claim that, contrary to my proposed counter-example, A in fact wanted to do x more than he wanted to do y. But this defence leaves Davidson in a worse position than ever, for his claim that the agent in the example wants to avoid adultery with her now more than he wants to commit it has now to be understood as the claim that he has a stronger inclination to avoid adultery with her now than to commit it. Yet a situation of conflict between desire for a long-term good and desire for immediate pleasure is a paradigm instance of conflict between reasoned desire on the one hand and inclination on the other. The fact that the agent found the idea of adultery so intensely attractive as to have great difficulty in adhering to his judgement about what it was best to do is a sufficient condition of the truth of the proposition that his inclination to commit adultery was stronger than any inclination he may have had to refrain. Had he nonetheless resisted this inclination, it would not have ceased to be true that his inclination toward adultery was stronger. We should then have had a situation where the question ‘Did A want to commit adultery more than he wanted to refrain from it?’ has no single answer. Given the conception of wanting as preferential choice, he wanted to avoid it more than he wanted to commit it, but given the conception of wanting as inclination he wanted to commit it more than he wanted to avoid it. But in the situation as described, the question has the same answer on either conception of wanting; whether we are asking about the agent’s preferential choice or his inclinations, in either case he wanted to commit adultery more than he wanted to avoid it. On either conception of wanting, then, the counter-example to P2 stands. A defender of P2 has now only one recourse, viz. to claim that there is some further conception of wanting such that an agent who both feels a stronger inclination to do y than to do x and actually does y in preference to x may still truly be said to want to do x more than he wants to do y. This would be a counsel of desperation. I at least am clear that I have no such conception, and I doubt whether anyone has. I can understand the proposition ‘A wants x more than he wants y’ in terms of A’s preferential choices, or in terms of his inclinations, but otherwise I have no idea what it might mean. But suppose that this is merely conceptual inadequacy on

48

plato, hare, and davidson on akrasia

my part. Even if we were to grant that there is some as yet unelucidated conception of wanting which allows us to say that our akratic man wanted to avoid adultery more than he wanted to commit it, that would not save the Davidsonian position, defined as the position that P1–P3 are all true. For it is now admitted that, on this new (supposed) conception of wanting, A may want to do x more than he wants to do y, and believe himself free to do either x or y and do y intentionally and not do x. That is, the defender of the Davidsonian position can save P2 only at the cost, not only of introducing an as yet unexplained conception of wanting, but also of giving up P1. He would do better to give up P2 and have done with it. I have now shown by means of a counter-example that on an ordinary understanding of what it is to want x more than y P2 is false. It has also emerged incidentally that P1 holds only where A wants x more than y in the sense that A chooses x in preference to y, and that it too is false where A wants x more than y in the sense that A has a stronger inclination for x than he has for y. In the next sections I shall consider the reasons why two other philosophers, Plato and R. M. Hare, have accepted P2 or a related principle.

II For evidence that Plato accepted a version of P2 we may turn to Prot. 358b–e. Here Plato says (i) If pleasure is the good, then if anyone judges that x is better than y, and he is able to do either x or y, he does x; (ii) If A judges that x is good and y bad, he is not willing to go for y in preference to x; (iii) If A judges that (a) x is bad, (b) y is bad, and (c) y is worse than x, and if A must choose either x or y, he will choose x. Plato’s explicit words fall short of the full generality of P2, since (i), which asserts that an agent will always choose what he takes to be the better of two alternatives, is dependent on the assumption that pleasure is the good, whereas (ii) and (iii), which are apparently independent of that assumption, make the lesser claims that an agent will always prefer anything good to anything bad, and that he will always prefer the lesser of two evils. In fact, I doubt whether the restriction is of any significance.

plato, hare, and davidson on akrasia 49 For, firstly, in the Protagoras at any rate, Plato argues that pleasure is in fact the good,³ which, together with the conditional expressed in (i) above, gives the generalization that if anyone judges x better than y, and believes himself able to do x or y, he will do x, which amounts to P2 (given the conception of wanting as preferential choice). Secondly, the discussion of the mechanics of choice at 356a–c shows that Plato sees the same principles applying to the choice between two good alternatives as to those between two bad and a bad and good, which are the cases explicitly dealt with by (ii) and (iii). I therefore interpret Plato as holding that it is as foreign to human nature to fail to choose the better of two alternatives as it is to choose a bad alternative when a good is available, or to choose the greater evil when only bad alternatives are available. While this thesis is asserted rather than argued for in the Protagoras it is fairly clear that it rests ultimately on the two assumptions (i) that for an agent to judge one alternative better than another is for him to judge that alternative more in his interest than the other, (ii) every agent always does what he thinks will best promote his interest. (i) The ultimately self-interested ground of choice of alternatives as better or worse appears clearly e.g. in the argument with Polus in the Gorgias. Polus maintains that while it is more shameful or disgraceful to treat someone else unjustly than to be treated unjustly it is worse to be treated unjustly (474c), Socrates maintaining, on the contrary, that acting unjustly is both more shameful and worse. The immediately preceding discussion has made it clear that what is at issue is whether acting unjustly tends to bring it about that the agent has a completely satisfactory life, as Polus maintains by appeal to the example of successful tyrants, or whether it tends to make the agent wretched and unfortunate, i.e. to give him an unsatisfactory life. In order to provide the necessary proof that the unjust man will have an unsatisfactory life Socrates establishes the following principle (474d–475b) about the opposition kalon (beautiful, creditable, praiseworthy)—aischron (ugly, shameful, discreditable): (x) (y) (If x is more kalon than y then x is pleasanter than y or x is more useful than y and if x is more aischron than y then x is more unpleasant than y or x is worse than y). ³ For a defence of this controversial view see Taylor [1976], 208–9, and for a reconsideration see chapter 15 of this volume.

50

plato, hare, and davidson on akrasia

It is clear that the intention of this principle is to establish creditableness/discreditableness as a function of the two factors, pleasure/pain and usefulness/harmfulness. The positive side of the latter opposition is designated by the term translated ‘usefulness’ (¯ophelia), the negative by the term ‘bad’ (kakon), which is universally the opposite of ‘good’ (agathon). It is clear that in this argument the pairs ‘good-bad’ and ‘useful-harmful’ function as interchangeable. Socrates uses the above principle to argue that since acting unjustly is admitted to be more disgraceful than being treated unjustly, and since it is clearly not more unpleasant, it must be worse (i.e. more harmful), and immediately (475d–e) forces Polus to admit that neither he nor anyone else would prefer what is worse to what is less bad. It is plain that what is meant is that no one will prefer what is worse for himself, since Plato is well aware that many people actively pursue what is worse for others. Similar moves are made in the next stage of the argument, where Socrates tries to show that if one has done wrong it is better for one to be punished than to go unpunished. The man who is justly punished is treated in a creditable way. But by the previous principle whatever is creditable is either pleasant or beneficial. Punishment is not pleasant for the person who undergoes it. Therefore the man who undergoes just punishment ‘undergoes good things’ (has good things happen to him). Therefore, he is benefited (476e–477a). Similarly, if vice is the most shameful thing it must be the most painful or the most harmful; but it is not painful, therefore ‘since it exceeds in the greatest harm it is the greatest evil (kakon) of all things’ (477e). It is clear, then, that throughout this argument the question of whether one alternative action is better or worse, as opposed to more or less creditable or honourable, than another is the question of whether that action tends more to promote or hinder the interest of the agent, which is identified as his achievement of a fully satisfactory life (eudaimonia). Assumptions (i) and (ii) are both apparent in one of the few passages where Socrates gives an argument to support any version of P2. The passage in question is Meno 77b–78b, where Socrates is examining Meno’s suggestion that excellence (aret¯e) is to be defined as ‘Desiring fine things and being able to get them’ (77b). Socrates first of all secures Meno’s immediate agreement that to desire fine things is to desire good things, and then argues that the inclusion of ‘desiring good things’ in Meno’s specification of aret¯e is redundant since it is impossible for anyone to desire anything

plato, hare, and davidson on akrasia 51 except what he takes to be good. Hence his proposed definition can be reduced to ‘the ability to acquire good things’ (78c). The first point to note is that the distinction between what is fine or honourable (kalon) and what is good (agathon), which was essential to Polus’ position, is completely ignored in this argument, in that Socrates and Meno move immediately from ‘desiring fine things’ to ‘desiring good things’ and thereafter conduct the discussion wholly in terms of good and bad, finally substituting ‘the ability to acquire good things’ for Meno’s ‘being able to get fine things’. This might be taken to suggest that the concept of goodness employed here is not, as in the previous argument, identical with the concept of what promotes the interest of the agent, but the development of the argument shows that this inference would be mistaken. Meno starts by maintaining that some people want good things but others want bad things. The latter class of persons consists of two sub-classes, those who want bad things in the mistaken belief that the things which they want are good things, and those who want bad things in the knowledge that the things they want are bad (77b–c). Socrates asks (d1–3) whether those who want bad things (of either sub-class) think that the bad things will benefit them, or whether they recognize (gign¯osk¯on) that the bad things (sc. which they want) harm whoever gets them; Meno replies that some (sc. those in the first sub-class) think that the bad things will benefit them, while others (sc. the members of the second sub-class) recognize that they will harm them. The assertion that every bad thing harms the person who has or gets it shows that in this argument too things are judged good or bad iff they promote or hinder the interest of the agent, since it would be commonplace that no one could be harmed otherwise than by a bad thing, or benefited otherwise than by a good thing. So far then, this argument is seen after all to embody assumption (i), viz. that for an agent to judge one alternative better than another is for him to judge that alternative more in his interest than the other. The conclusion of the argument gives assumption (ii), at least in the weaker form that no one so acts as to do what he thinks will damage his interests. The argument continues as follows. Those who want bad things in the mistaken belief that they are good things are not strictly to be described as wanting bad things, but rather as wanting good things, though the things they want are in fact bad. (It would be clearer to say that the description under which they specify what they want is not ‘something bad’ but ‘something good’

52

plato, hare, and davidson on akrasia

or, adopting G. Santas’s terminology,⁴ that the intended object of their desire is not something bad but something good.) The members of the other sub-class, however, are assumed to want bad things in the knowledge that they will harm them. But they also believe that anyone who is harmed is made wretched to the extent to which he is harmed. But no one wants to be wretched and unfortunate. Hence no one wants to be harmed and hence no one wants bad things. One might object to this argument on a number of different counts. Firstly, one might challenge the truth of either of the premisses of the concluding stage, viz. that to be harmed is to be made wretched and unfortunate to the extent to which one is harmed or that no one wants to be wretched and unfortunate. One objection to the first premiss would clearly be irrelevant: this would be the objection that, since one can be harmed without being aware of it, being harmed doesn’t imply that one is thereby made wretched, in the sense of subjectively miserable. This misses the mark because the Greek terms athlios and kakodaim¯on imply that the state of the person to whom they apply is a deplorable one, but not necessarily that that person is aware that his state is such. More to the point is the objection that some sorts of harm don’t seem significant enough to count towards making the person harmed wretched and unfortunate, descriptions which imply that his overall state is a deplorable one. Thus if I kick someone in the shin, causing pain and a bruise, I may perhaps be said to have harmed him, but it is doubtful if I have really contributed to making him wretched and unfortunate. One line of reply to that objection is to concede that it succeeds in differentiating trivial from serious harm, but to maintain that the kinds of conduct which Socrates is trying to explain away (drinking to excess, dissipating one’s fortune, etc.) fall under the latter description. Another is to take the notion of ‘being unfortunate’ as ‘having one’s interests adversely affected’: this gives an unexceptionable first premiss, that anyone who is harmed to any extent has his interests adversely affected to that extent, and requires that the second premiss be reformulated as ‘No one wants to have his interests adversely affected’. The truth of this premiss might be attacked whether it is expressed in this new formulation or in the original one; surely some people do want their interests to be adversely affected, or even (in an extreme case) want to ⁴ Santas [1964].

plato, hare, and davidson on akrasia 53 be wretched and unfortunate. For example, some people want to suffer for their sins, not necessarily with a view to the improvement of their subsequent lot, but just because they have been so deplorably wicked that it is right that they should suffer extremes of misfortune. Others again might want their interests to be sacrificed for the attainment of some ideal. It is indeed possible to defend this premiss against this objection, though the defence lacks plausibility, in my view. I shall not, however, pursue that topic, for even if the truth of both premisses is granted, the major objection still stands that those premisses do not entail the conclusion. Even if it is true that to be harmed is to be made, to some extent, wretched and unfortunate, it does not follow that no one wants things he knows to be harmful. This argument requires the principle that if A wants to do x, and knows that doing x will bring about the existence of state S, then he wants the existence of S; the argument consists in fact of the application of modus tollendo tollens to this principle. But the principle is false; for A may want to do x despite the fact that he knows that doing x will bring about the existence of S, a state which he wants not to occur; or he may be quite indifferent to the fact that doing x will bring about S, and hence it will not be the case that he wants S to occur. If he actively wants S not to occur, then this may induce him not to do x in fact, but he may still want to do x, which of itself falsifies the principle. And if he does x, thereby bringing about S, the principle is still false. For if he does x, in total indifference to the fact that doing x will have the foreseen consequence of bringing about S, or bitterly regretting the fact that it will have that inevitable but deplored consequence, in neither case does he want to bring about S. That is to say, there is nothing which he wants under the description ‘object which will bring about S’. The concept of wanting an object under a description figures in a true principle about the relation of desired objects and their foreseen consequences, viz. ‘If A wants to do x under the description ‘‘action which will bring about the existence of S’’, then A wants the existence of S’, or alternatively ‘If A wants to do x because he believes that doing x will bring about the existence of S, then A wants the existence of S’. But we cannot save Plato’s argument by substituting this principle for the false one on which he actually relies. For given the true principle and Plato’s two premisses all that follows is that no one wants anything because he believes that that thing will harm him, whereas Socrates’ argument requires the conclusion that no one wants anything

54

plato, hare, and davidson on akrasia

which he believes will harm him. Obviously, most of the cases which Socrates is trying to exclude are cases of wanting something although one believes it will harm one, not of wanting something because one believes it will harm one. I thus conclude that the only argument which Plato gives in support of any version of P2 is inadequate. Nor do I believe that any argument can be adequate which endeavours to support P2 by appeal to self-interest as the basis of evaluation. For if P2 is formulated explicitly in terms of the agent’s interest, i.e. If an agent judges that it would be more in his interest to do x than to do y, then he wants to do x more than he wants to do y, counter-examples can readily be supplied, as Davidson himself shows by his demonstration (pp. 101–2), that one may as readily want to act, and actually act, against one’s better self-interested judgement as against one’s better judgement when supported by other considerations. Why might one incline to accept the self-interested version of P2? Perhaps because one holds the plausible belief that every agent wants the promotion of his own interest more than he wants anything else. That belief would license the further reformulation of P2 as If an agent judges that doing x would promote what he most wants more than doing y would, then he wants to do x more than he wants to do y. If that principle is read throughout as a principle of preferential choice, then it becomes If an agent judges that doing x would promote the attainment of what he chooses in preference to anything else more than doing y would, he chooses to do x in preference to doing y. That is to say, every agent adopts what he takes to be the best means of attaining his ultimate goal. That is plainly false: given the choice of better or worse means of attaining his goal an agent might choose the worse, against his better judgement, for all sorts of reasons, such as squeamishness or laziness. Read as a principle about inclinations, or a mixed principle relating inclination to preferential choice, it is just as plainly false. For example consider the following:

plato, hare, and davidson on akrasia 55 If an agent judges that doing x would tend to promote the attainment of what he chooses to do in preference to anything else more than the doing of y would, he has a stronger inclination to do x than to do y. This is false, since an agent may find that the taking of the means necessary to his preferred end runs strongly counter to his inclinations; consider the case of a resistance fighter whose preference for the freeing of his country from occupying forces obliges him to conquer his natural revulsion and kill an enemy soldier. Moreover, the belief which leads to this reformulation (viz. that every agent wants the promotion of his own interest more than he wants anything else), though plausible, is itself false, as is shown by the counter-examples on the one hand of self-sacrifice and on the other of action against one’s better judgement. In the first sort of case the agent wants to do something else (e.g. sacrifice his life for an ideal) more than he wants to promote his interest because he judges it better to do so, in the latter he wants to do something else (typically pursue some short-term pleasure) more despite the fact that he judges it better to pursue his long-term interest. That concludes what I have to say about Plato. I hope that I have shown (a) that Plato accepted a version of P2 because he believed that every agent’s ground of evaluation was self-interested, and (b) that that belief is both false and, even if true, insufficient for the truth of P2. I turn now to some brief consideration of Hare.

III Hare’s commitment to P1 and P2 arises from his theory of the prescriptive nature of evaluative judgement and from the close connection which he sees between evaluative judgements and desires. I shall take it that the former is sufficiently familiar to make exposition otiose, but the latter requires some discussion. He maintains that if A wants to do x more than he wants to do y he assents to the prescriptive judgement ‘Let me do x in preference to y’⁵ and in terms of his general theory sincere assent to that judgement requires that he act on it. Hence Hare accepts P1. I am not clear ⁵ Hare [1963], 70–1, 91.

56

plato, hare, and davidson on akrasia

whether he maintains that if A wants to do x more than he wants to do y, A is committed also to the judgement ‘It is better for me to do x than to do y’. The latter, of course, would also commit him to accepting ‘Let me do x in preference to y’, but from the fact that he is already committed to that we can’t of course infer that he is committed to what entails it. But it is clear that the converse implication from evaluative judgement to desire does hold, in Hare’s view: if we use the word ‘desire’ in a wide sense, we can say that any evaluation, just because it is prescriptive, incorporates the desire to have or do something rather than something else. The wide sense in which we are here using ‘desire’ is that in which any felt disposition to action counts as a desire ... ⁶

It isn’t altogether clear whether Hare means that any evaluation whatever incorporates some desire, or whether the point is restricted to evaluations by an agent of courses of action etc. to be undertaken by himself. If the former, then his thought is that an evaluative judgement such as ‘Fischer ought to move his bishop’, since it entails the judgements ‘Let Fischer move his bishop’ and ‘Let anyone, in circumstances similar to those in which Fischer is now, move his bishop’, includes the desire that those prescriptions be fulfilled. The sense in which assent to such prescriptions incorporates a desire is presumably explained by the statement quoted above that ‘any felt disposition to action counts as a desire’. The assent to such prescriptions is seen as something like the acceptance of a commitment that the specified state of affairs should occur, which, at least in the case of the universalized prescription, is in turn explicable as a commitment to action of the appropriate sort, should the relevant circumstances ever apply to the prescriber. This commitment may in turn be described without undue distortion as the recognition or perhaps formation of a disposition to act. Since it is quite unclear how anyone could have such an attitude to someone else’s action (e.g. no meaning attaches to the statement that I am committed to Fischer’s moving his bishop) it is fairest to Hare to assume him to mean that desires are incorporated primarily in first-person evaluative judgements, and that they are derivatively incorporated in second and third person judgements insofar as the latter, being universalizable, entail universal prescriptions which in turn entail first person prescriptions. Thus ⁶ Hare [1963], 170.

plato, hare, and davidson on akrasia 57 my judgement ‘Fischer should move his bishop’ does not incorporate the desire on my part that Fischer should move his bishop, but does incorporate the desire on my part that were I in a position relevantly similar to that in which Fischer now is, I should move my bishop.⁷ This restriction does not affect the main issue, since it is clearly Hare’s view that in making an evaluative judgement on a course of action to be undertaken by himself, the agent incorporates a desire to do whatever it is that he evaluates more favourably, i.e. Hare accepts P2: If an agent judges that it would be better to do x than to do y, then he wants to do x more than he wants to do y. As I have already argued that P2 is false, I shall confine myself to consideration of Hare’s reasons for accepting it. Fundamental to his reasoning is his thesis that evaluative judgements entail prescriptions, e.g. the judgements ‘It would be better for me to do x than y’, ‘I ought to do x rather than y’ and ‘It would be all right for me to do x but not all right for me to do y’, all entail the prescription ‘Let me do x rather than y’. Hare’s language in the passage quoted above suggests that we have in such a prescriptive judgement an explanatory concept which will explain how it is that evaluative judgements have action-guiding force. Evaluative judgements incorporate desire, i.e. felt dispositions to action, because they are prescriptive. But on the contrary, the sense of such a prescription as ‘Let me in situation S do x’ has itself to be explained as follows, that if an agent judges that in situation S he ought to do x or that it would be permissible for him to do x, then he will be inconsistent if he fails to do x should S obtain. That is to say, the entailment of the prescriptive judgement has to be explained in terms of the commitment to action on the part of the person making the evaluation, and cannot therefore itself explain that commitment. Hare has therefore to give independent grounds for positing that commitment to action in the first place. It clearly will not do to say that the commitment arises because evaluative judgements express desires, and desires carry a commitment to action. For firstly it is just false that, on any ordinary understanding of the word ‘desire’, all evaluative judgements express some desire of the ⁷ But in a later paper Hare implicitly maintains that the judgement ‘Let it be the case that A φs’ expresses the desire of the utterer, not merely that he should φ were he in a case similar to that in which A now is, but that A should φ in the actual case. The paper is Hare [1971a], reprinted in Hare [1971b]; the crucial passage occurs on p. 88 of the former volume and on p. 50 of the latter.

58

plato, hare, and davidson on akrasia

person making them. And secondly, in trying to elucidate his notion of the incorporation of desires in evaluative judgements, Hare is obliged to explain his ‘wide sense’ of ‘desire’ as ‘any felt disposition to action’, and we in turn have been obliged to elucidate ‘felt disposition’ as commitment. So once again, the explanans (i.e. that evaluative judgements express desires) turns out to contain the explanandum (i.e. that evaluative judgements carry a commitment to action).

IV In my concluding section I propose an alternative principle to P2. I am happy to leave P1 to stand, provided that it is understood as a principle of preferential choice. I have already pointed out that it is false when applied to inclinations. A number of alternatives to P2 might be proposed. Thus the following has at least the merit of truth: If an agent judges that it would be better to do x than to do y he believes that he has better reasons for doing x than for doing y. But it could reasonably be objected that, while true, this is uninformative, since it merely links one belief or judgement to another, and moreover to a judgement of the same basic sort, since the judgement that one action is better than another is linked to the judgement that one set of reasons is better than another. For these reasons I propose instead the following: P2

. If an agent judges that it would be better to do x than to do y he ranks the doing of x by him on this occasion higher than he ranks the doing of y by him on this occasion.⁸ This has the merit of bringing out the fact, which those who defend P2 in its original form have grasped, that judging it better to do x than to do ⁸ Cf. Watson [1977]. Watson accepts (p. 321) the principle ‘If one wants to do x more than one wants to do y, one prefers x to y, or ranks x higher than y on some scale of values, or ‘‘desirability matrix’’ ’, and claims that that supports the thesis that P2 is ‘true if understood in the language of evaluation’ (ibid.). Thus understood, P2 is (apparently) expressed as ‘If a person judges x to be better than y, then he or she values x more than y’. But that is a version of P2

, not P2. P2 and P2

are not equivalent, since P2

is true and P2 false. P2

does indeed follow from P2, when the latter is taken together with Watson’s principle, but to derive P2 from P2

one requires not that principle but its converse, which, as I argue below (pp. 59–60), is false.

plato, hare, and davidson on akrasia 59 y is not purely a matter of recognizing that the doing of x possesses some property to a higher degree than the doing of y, but also has some necessary connections with the agent’s motivation and behaviour. But it avoids the error of P2, which was to focus those connections on the agent’s desires, when in fact they are scattered over a wider range of the agent’s attitudes and behaviour. For those connections are mediated by the truth of ‘A ranks the doing of x by him on this occasion higher than he ranks the doing of y by him on this occasion’. And the truth of that proposition is loosely linked, in the way characteristic of cluster-concepts, to the satisfaction of an open disjunction of conditions of which the following are typical: A does x spontaneously and unhesitatingly in preference to y; A feels pleased that he has done x in preference to y; A feels remorse that he has not done x in preference to y; A regards this as a typical case of choice between doing x and doing y, and admires people who in such cases do x in preference to doing y. No single condition is either necessary or sufficient, since any condition may be absent, provided that some other is satisfied, or may be present but be overridden by some contrary condition. Yet the satisfaction of some disjunction of conjunctions of those conditions is sufficient, and that of the disjunction of conditions necessary, for the truth of ‘A ranks the doing of x by him on this occasion higher than he ranks the doing of y by him on this occasion’. Analogous conditions may be specified for the truth of ‘A ranks the doing of actions of type E higher than he ranks the doing of actions of type F’. The force of P2

is then to specify that the conditions necessary for the truth of ‘A ranks the doing of x by him on this occasion higher than he ranks the doing of y by him on this occasion’ are in turn necessary for the truth of ‘A judges that it would be better to do x than to do y’. It follows that the truth of ‘A judges that it would be better to do x than to do y’ is sufficient for those conditions, i.e. for the satisfaction of the disjunction of the open set of conditions which I have indicated, though not for the satisfaction of any single condition. Now it is true that some of these conditions imply certain propositions about the agent’s desires; thus if A is ashamed of himself for having done y rather than x, A wishes that he had not done y, but had done x instead. Again, if A is glad that people whom he likes do x rather than y, A in general wants it to be the case that people whom he likes should do x rather than y. But it is a gross oversimplification to allow the inference from (i) ‘A ranks the doing of x

60

plato, hare, and davidson on akrasia

by him on this occasion above the doing of y by him on this occasion’ to (ii) ‘A wants to do x on this occasion more than he wants to do y on this occasion’. For instance, (ii) may be false, since A’s inclination was to do y and he acted on it, but yet it may be true that after he had done y he wished that he had done x. Or again, (ii) may be false, but it may be true that A in general wants it to be the case that people whom he likes should do x rather than y, and also true that A regards this as a typical case of choice between x and y, which may be sufficient for the truth of (i). This account of P2

seems to me to do justice to the truth of Davidson’s remark (p. 99) that ‘if someone really (sincerely) believes he ought, then his belief must show itself in his behaviour’, while at the same time showing the error in the parenthesis which concludes that sentence ‘(and hence of course in his inclination to act, or his desire)’. For I hope that I have shown that the connections between the judgement that x is better than y and behaviour are too subtle to be adequately expressed in any single statement of connection between judgement and desire, such as that given by the original P2. Of course, we might take P2 just as a schematic formulation of the complicated web of connections between judgement and desire which I have so roughly sketched. But given that re-reading P2 will fail to satisfy another requirement on which Davidson rightly insists (ibid.) as essential to his account of the problem, viz. that ‘want’ or any expression which is substituted for it should be univocal between P1 and P2. For if ‘wants to do x more than he wants to do y’ in P2 is read as ‘ranks the doing of x higher than the doing of y’, and if that is elucidated as I have attempted, then it must have the same reading and elucidation in P1. And it will, I hope, be clear without further discussion that, given that elucidation of ranking the doing of x higher than the doing of y, P1 will be false. P2

, then, is (a) true and (b) distinct from the original P2. It is also consistent with P1 (interpreted as a principle of preferential choice) and P3. And it is the fact that these principles are consistent which constitutes, in my view, the solution to the problem of akrasia. For cases of akrasia are cases where an agent judges that it is better to do x than to do y, but does y, and this is possible because the agent’s judging it better to do x than to do y does not imply either that he does x in preference to doing y or that he wants to do x more than he wants to do y, but rather that he ranks the doing of x by him on this occasion higher than

plato, hare, and davidson on akrasia 61 he ranks the doing of y by him on this occason. That ranking does indeed have certain implications concerning what he does and what he wants, but those are not the simple implications specified by P1 and P2. The mistakes of the eminent philosophers whom I have been criticizing seem to me to arise from a common source, namely failure to appreciate the complexities of the concept of desire and of its relations with judgement on the one hand and action on the other. Afterword For discussion see Schueler [1983], with reply by Taylor [1984]. In discussing Meno 77b–78b in this chapter I take for granted the traditional ‘subjectivist’ interpretation of Socrates’ claim that no one desires anything other than good things: i.e. I take his claim to be that no one desires anything other than what he or she takes to be good. Penner and Rowe [1994] argue that, on the contrary, the Socratic view is that no one desires anything except what is actually good. Segvic [2000] maintains a related view. Wolfsdorf [2006] defends the traditional view; his paper includes a useful survey of the literature. Irwin [1995] suggests (pp. 138–9) that Socrates may mean no more than that no one ever desires bad things because they are bad, and hence that ‘the good is one object of everyone’s desire, not that it is the object of all desire’. Reasons for rejecting this suggestion are cogently stated by Scott [2006], 219–20.

4 The End of the Euthyphro1 Most of the recent discussions of Plato’s Euthyphro have concentrated either on the so-called ‘Euthyphro dilemma’ (i.e. the problem of whether divine commands create moral values or presuppose independently existing values) or on the intricacies of the argument in 10a–11b in which a version of that problem is posed. The concluding section of the dialogue from 11b to the end, though much discussed by earlier writers,² is virtually ignored in the modern literature. In this paper, I renew the discussion of this section of the dialogue, in the belief that some of the insights of the earlier generation of commentators may prove to have something to contribute to the problem of the unity of virtue in the early Platonic dialogues. A resum´e of the section in question will be helpful. At 11b it is agreed that Euthyphro’s proposed account of to hosion as what is pleasing to all the gods has been refuted, since it has been shown (by the argument of 10a–11b) to lead to contradiction: accordingly, Socrates invites Euthyphro to suggest another account. But Euthyphro is now (not surprisingly) in a state of complete bewilderment, so Socrates volunteers (11e2–4) to help him out with a suggestion. ‘Think of this, now’, he says, ‘doesn’t it seem to you necessary that whatever is hosion is just?’ (e4–5). Euthyphro agrees, whereupon Socrates asks whether he also thinks that whatever is just is hosion, or whether his view is that, while everything hosion is just, not everything which is just is hosion (11e6–12a2). On Euthyphro’s failing to grasp the point, Socrates spells it out by means of another example (12a3–c9) and Euthyphro eventually settles for the thesis that ¹ An earlier version of this paper was read to the Southern Association for Ancient Philosophy in 1981. I have made several changes in the light of the discussion and wish to acknowledge my debt to all those whose criticisms have led me to improve the paper. ² In addition to the works cited later in this paper, there are useful discussions in the following: Jowett [1871], i. 295–9; Adam [1890], xiv–xvi; Heidel [1900], 163–81; Burnet [1924], esp. 56–7; Croiset [1946], i. 181–3; Friedl¨ander [1964], ii. 83–9. For further references see Rabinowitz [1958], 113.

the end of the euthyphro

63

everything hosion is just but not vice versa, i.e. that ‘The hosion is a part of the just’ (c10–d3). In response to Socrates’ question ‘Which part?’ (d5–7), Euthyphro replies that it is that part which is concerned with the service (therapeia) of the gods, while the remainder of the just is that part which is concerned with the service of men (e5–8). At this point a few words of elucidation are in order. Firstly, the subsumption of to hosion under justice should remind us that the subject of the dialogue is hosiot¯es as an attribute of persons and their actions. Of course, things of other kinds can be hosia, e.g. a grove or a temple, to which the Greek adjective hosios is as readily applicable as its English renderings ‘holy’ and ‘sacred’. The rejected account of to hosion as what is beloved of or pleasing to the gods has at least this to be said for it that it applies alike to non-personal and to personal instances; the gods may love woods and rivers as well as men and their doings. But the opening of the dialogue makes it clear that agents and their actions are the primary cases: the quest for an account of to hosion arises out of the question whether it is hosion or anosion of Euthyphro to prosecute his father for homicide and Euthyphro’s first answer to the question ‘What do you say to hosion is?’ is ‘It’s what I’m doing now’ (5d6–9). The adjective hosios is treated as interchangeable with euseb¯es, meaning ‘well-disposed towards the gods’, ‘reverent’, or ‘religious’: Socrates first asks Euthyphro for his account of to hosion in the words ‘What sort of things do you say to eusebes and to asebes are, in cases of homicide and in the other types of case?’ (5c9–d1). To hosion is then the virtue of being properly disposed both in thought and action towards the gods: it is hosion for Euthyphro to prosecute his father if and only if so doing manifests the proper relationship to the gods, which is to say that it fulfils a religious obligation. No one English word conveys this sense exactly: ‘religious’, as applied (in a slightly archaic usage) to persons and their actions perhaps comes closest, while its opposite ‘irreligious’ comes close to capturing the sense of aseb¯es. The suggested connection between hosiot¯es and justice is then straightforward. Justice (dikaiosun¯e) is the primary social virtue, the standing disposition to respect and treat properly all those with whom one enters into social relations. Human individuals have social relations, not only with other human individuals, but also with the gods. Hosiot¯es is then the name of that particular aspect of the basic social virtue which is directed towards the gods: every hosion act is a just act, i.e. a just act directed towards a god or

64

the end of the euthyphro

gods (e.g. making a sacrifice, mentioning a divine name with respect) and to be hosios is simply to be just in one’s dealings with the gods. Hosiot¯es is then a kind of justice, viz. justice vis-`a-vis the gods, just as parricide is a kind of murder, viz. murder of one’s father. On the position which Euthyphro finally adopts (12e5–8) justice towards other humans has no special name, being described merely as ‘the remaining part of the just’ (i.e. what’s left when hosiot¯es has been distinguished); the term ‘justice’ is treated as generic, the social virtue as such, exercised either towards gods or humans. Ordinary Greek idiom would naturally appropriate the term dikaiosun¯e as the name for the virtue of social relations with human agents, and it is in accordance with that usage that the good man is described at Gorg. 507b as one who would do right by men, i.e. justice, and by the gods, i.e. religion (lit. ‘concerning men he would do the fitting just things, and concerning the gods (the fitting) religious things’).³ It is unnecessary to suppose any difference of doctrine between that passage and the Euthyphro; rather the difference is to be accounted for by a natural shift in application of the term dikaiosun¯e between the genus and its principal species parallel to that illustrated by the pair of sentences ‘Parricide is the worst kind of murder’ and ‘John was guilty of murder, Peter of the far worse crime of parricide’.⁴ What has been said so far leaves the content of hosiot¯es quite open; thus it has still to be determined whether proper behaviour towards the gods is restricted to the area of prayer, ritual, etc., or whether (as suggested by the problem raised by Euthyphro’s prosecution of his father) it extends more widely into what we should regard as the sphere of morality. Euthyphro’s description of hosiot¯es as justice concerned with the service of the gods might suggest that he has the narrower conception in mind, as therapeia is regularly used in the sense ‘religious observance’, but we should not read too much into the use of this word, since the Greek construction requires that it be understood also in the account of ‘justice towards men’ (to de peri t¯en t¯on anthr¯op¯on to loipon einai tou dikaiou meros, 12e7–8). Here the word has to be given a fairly attenuated sense, such as ‘treatment’, which itself gives a good contrast with ‘justice towards the gods’: religion is a matter of treating the gods correctly, the remainder of justice is a ³ Cf. Dover [1974], 247. ⁴ Cf. Aesch. Seven against Thebes, 597–612 (cited by Dover [1974], 203).

the end of the euthyphro

65

matter of treating other people correctly. Of course, correct treatment of the gods will include ritual as a prominent aspect, which is no doubt why the word therapeia is a natural one for Plato to use in this context, but we should not seize on the word to restrict the scope of hosiot¯es in Euthyphro’s formulation to ritual alone. I now revert to the text. Socrates asks Euthyphro to explain what he means by therapeia of the gods. One ordinary sense of the word is ‘care, looking after’, e.g. care of horses, but it is soon agreed that h¯e t¯on the¯on therapeia is not caring for the gods, since the aim of caring for A is to benefit or improve the condition of A, but the aim of h¯e t¯on the¯on therapeia is not to benefit the gods (12e9–13d3). Rather it is the sort of therapeia which a servant (therap¯on) provides for his master, viz. service or assistance (another ordinary sense of the word) (13d4–8). But the giving of assistance presupposes that the person assisted has a goal, the achievement of which is furthered by the assistance: thus a boat-builder’s assistant helps him to build a boat (d9–e5). What, then, is the goal towards the achievement of which the gods are assisted by the therapeia of their worshippers (13e6–14a10)? Euthyphro seems to be misled by the wording of Socrates’ final question ‘What, then, are the many fine things which the gods bring about? How should we sum up their work?’, for his answer misses the train of thought which Socrates has been following. He answers that if a man knows how to please the gods by worshipping them aright with prayer and sacrifice they preserve his house and his city, but if he displeases them by failure in these respects then his irreligion will bring about total ruin (presumably of house and city) (14a11–b7). This misses Socrates’ drift: as Socrates later remarks (14d6–7), Euthyphro now conceives of the gods as standing in a quasi-commercial relationship with men, in which the gods either reward men for attentions paid to them, or punish them for failure to pay those attentions. But in such an exchange of services the worshipper cannot be described as assisting the god to promote the worshipper’s well-being, since ex hypothesi the god has no pre-existing purpose of promoting the worshipper’s good, but acquires an interest in that good only when the relation is set up. The situation of the worshipper is similar to that of a professional entertainer: he is paid for providing a service but it would be absurd for him to claim that by providing that service he assists his patron in the patron’s task of enabling him (the entertainer) to earn his living. If the concept of assistance applies at all to this case it applies not to the

66

the end of the euthyphro

entertainer but to the patron, who, by hiring the entertainer, assists him in his pre-existing task of earning a living. On that analogy it is the god who assists the worshipper, not vice versa. The notion of divine aid to the pious is indeed intelligible and prominent in many religions, including Greek religion: but in giving such a case as illustration of assistance to the god by the worshipper Euthyphro is represented by Plato, presumably intentionally, as confused. That Plato was aware of the confusion is apparent from Socrates’ reply (14b7–c6). He says that Euthyphro has turned away when he was on the point of giving an adequate account of what hosiot¯es is. That is to say, Euthyphro has now taken a wrong turning, which Socrates, like a lover pursuing his beloved, has to follow. The first stage in the examination of this wrong turning is the characterization of Euthyphro’s conception of hosiot¯es successively as knowledge of prayer and sacrifice (c5–6), knowledge of how to ask for things from the gods and give things to them (d1–2), and finally as commercial expertise between gods and men, each side getting from the other what it wants while in return satisfying the wants of the other (d9–e7). What good, then, do the gods get from their dealings with men? Since Euthyphro maintains (15a5–6) the previous position that the gods can’t be benefited, i.e. improved, by the actions of men, the good that they get is honour and pleasure (a9–10). Hence the wrong turning leads back again to the answer rejected earlier that to hosion is what is pleasing to the gods (b1–c10). Socrates offers to try a fresh start, but Euthyphro has to leave to keep an appointment, and the dialogue ends. As is well known, it is characteristic of the early dialogues to end in aporiai, and not unknown for them to contain fairly clear hints of a conclusion which is not explicitly drawn (e.g. Charm. 174d–175a). The Euthyphro gives us a clearer hint than most in Socrates’ complaint that Euthyphro has gone astray when he was on the point of giving the right answer to the question ‘What is to hosion?’ As we saw, Euthyphro’s false move was to misunderstand the question ‘What are the fine things which the gods accomplish?’ Socrates had brought him to the point of seeing hosiot¯es as service of the gods, i.e. as assistance to them in their work. Now a competent craftsman doesn’t use assistants unnecessarily: hence, assuming the competence of the gods (which is not questioned in the dialogue), if ‘assistance to the gods’ is to be a correct account of hosiot¯es there

the end of the euthyphro

67

must be some good purpose for the achievement of which the gods need human assistance. Plainly the gods don’t need human help in creating and maintaining the natural world, assuming those to be divine tasks. But there is one good product which they can’t produce without human assistance, namely, good human souls. For a good human soul is a self-directed soul, one whose choices are informed by its knowledge of and love of the good. A good world must contain such souls and hence, if the beneficent divine purpose is to be achieved, human beings must play their part by knowing (and hence loving) the good and acting in accordance with that knowledge. True hosiot¯es, the real service of the gods, turns out then to be nothing other than aret¯e itself.⁵ It is, however, aret¯e under a certain aspect; just as Aristotle describes ‘general justice’ as aret¯e in relation to another (pros heteron EN 1130a10–13) so the conception of hosiot¯es which we have identified in the Euthyphro is that of aret¯e pros ton theon, goodness of soul seen as man’s contribution to the divine order of the universe.⁶ Now this is not a novel suggestion: indeed some version of it is maintained by most of the writers on the dialogue whom I have consulted. Gomperz ([1905], 367) provides an illuminating parallel from Kant’s Religion Within the Limits of Unassisted Reason: Religion is the recognition of all our duties as divine commands ... In a universal religion there are no special duties towards God ... If anyone finds such a duty in the reverence due to God, he does not reflect that this is no particular act of religion, but a religious temper accompanying all our acts of duty without distinction.

Yet interesting though the parallel is, the fact that Kant had a certain thought is not itself evidence that Plato had a similar one, nor does a cloud of modern witnesses provide direct confirmation for a speculative interpretation of an ancient author. For support we must rather turn again ⁵ For evidence in popular Greek thought of a wide conception of hosiot¯es, in which much of social morality was seen as falling within the sphere of interest of the gods, see Dover [1974], 250–4. ⁶ Compare Leg. X. 903b–904b: ‘He who provides for the world has disposed all things with a view to the preservation and perfection of the whole, wherefore each several thing also, as far as may be, does and has done to it what is meet ... Thine own being also, fond man, is one such fragment, and so, for all its littleness, all its striving is ever directed towards the whole ... it is not made for thee, but thou for it. ... He (sc. who provides for the world) contrived where to post each several item so as to provide most utterly, easily and well for the triumph of virtue and rout of vice throughout the whole’ (tr. A. E. Taylor [1934]). If I am right in my account of the Euthyphro, Plato’s last thoughts on the cosmic role of the moral agent had much in common with what may have been his first.

68

the end of the euthyphro

to the text, not expecting explicit confirmation or refutation (for the interpretation is speculative) but rather looking for things in the text which provide a good fit with the interpretation, and also for things which clash with it. The first point in favour of the interpretation (noticed by A. E. Taylor⁷ and I. M. Crombie⁸ is that it allows even Euthyphro’s wrong turning, when correctly interpreted, to provide a correct account of hosiot¯es. Following the lead given by Euthyphro’s assertion that the man who knows how to please the gods by prayer and sacrifice wins rewards (14b2–5), Socrates summarizes his view in the characteristically Socratic suggestion that hosiot¯es is a sort of knowledge of prayer and sacrifice (14c5–6). Prayer being construed as asking for things from the gods and sacrifice as giving things to the gods, this formula becomes ‘holiness is knowledge of how to ask from and give to the gods’ (d1–2). But that knowledge is not a distinct sort of knowledge from knowledge of how to live, i.e. (given the Socratic thesis that virtue is knowledge) from goodness (aret¯e) itself. Someone who knows what human good consists in knows that a life lived in the pursuit of that good is the only thing which the gods need from human beings (see above) and, as Socrates points out, it is not an intelligent thing to give someone what he doesn’t need (e2–4). Again, he knows that such a life, being the supreme object of value, is that which above all we should ask from the gods: hence it is no accident that when Socrates prays in the dialogues he almost always prays for wisdom and goodness.⁹ When the knowledge of how to live which is aret¯e is related to the gods, that knowledge itself becomes knowledge of how to live in relation to the gods, which includes knowledge of how to give to the gods and how to ask from them. And one gives correctly to them just by living well and asks correctly from them by asking to live well. That was not indeed what Euthyphro meant, but Plato surely intends us to see that even the primitive notion of religion as a commercial transaction, when interpreted in the light of the Socratic doctrine that virtue is knowledge, contains an aspect of the truth. This leads us to the firmest confirmation of the interpretation, viz. that it applies to hosiot¯es the doctrine of the unity of the virtues, explicitly argued for in the Protagoras and implicit in the Meno and Laches. Hosiot¯es is not ⁷ Taylor [1926], 154–6. ⁸ Crombie [1962], 211. ⁹ See especially Phaedr. 279b–c, and for fuller details and discussion Jackson [1971].

the end of the euthyphro

69

an element contributing along with others to total goodness, in the way that a succession of different courses, each good, makes up a good dinner. Rather hosiot¯es is goodness seen under a particular aspect, viz. the relation of man to gods. Hence whatever goodness is, that (whatever it is) will be what hosiot¯es is. Now the Protagoras and Meno tell us that goodness is a sort of knowledge, viz. knowledge of how to live or how to achieve the good (identified in the former but not the latter as the maximization of pleasure). (I pass over as irrelevant the complication that in the Meno Plato apparently accepts orth¯e doxa as an alternative account of goodness.) Hence knowledge of the good will be what hosiot¯es is. Now at the end of the Protagoras (361b) Socrates says that each of the virtues (courage, s¯ophrosun¯e, justice, and hosiot¯es) has turned out to be knowledge: the argument has been given for courage only (viz. that knowledge of what is good and bad is necessary and sufficient to make a man brave) but is plainly intended to be applicable to the others. That same account of courage is derived in the Laches, only to be rejected on the ground that if it were correct courage would be identical with goodness, whereas it is ex hypothesi a ‘part’ of total goodness. The implication is clear that it is the hypothesis which is to be rejected, rather than the account of courage. As is spelled out in the Protagoras the traditional view of the specific aretai as parts of total aret¯e is to be rejected. Rather each aret¯e is identical with every other and with goodness as such. By this I understand the doctrine that what makes a man a good man overall is the very same thing as what makes him a courageous man, as what makes him a right-minded or self-controlled man, as what makes him a just man and as what makes him a religious man. The thing in question is his knowledge of what is good. The relation represented by the verb ‘makes’ here is not that of formal causation, as in ‘What makes this figure a triangle is its having three straight sides’. Rather it is a relation of efficient causality, as in ‘What makes him a first-rate tennis player is his perfect balance, excellent coordination and exceptional stamina’. This illustration is helpful up to a point, in that it suggests a parallel for the thesis of the unity of the virtues. Suppose that all and only those qualities listed above are causally necessary and sufficient, not only for their possessor’s being a first-rate tennis player, but also for his being first-rate at badminton and at squash. Then it would be true that what makes him excellent at any one of those games is the very same thing (i.e. the same set of attributes) as

70

the end of the euthyphro

what makes him excellent at both the others. Of none of the three games is it the case that his pre-eminence at that game derives even in part from some attribute which is exercised exclusively in the playing of that game. No doubt we should still refuse to say ‘His being good at tennis is the very same attribute as his being good at squash’, since the reference of ‘his being good at tennis’ includes his actually playing tennis excellently, which is a different activity from his actually playing squash excellently. But this merely brings out the difference between our concept of ‘being good at ϕ-ing’ and Plato’s concept of ‘the aret¯e appropriate to ϕ-ing’: being good at ϕ-ing is possessing and exercising the capacity to ϕ well, whereas the aret¯e appropriate to ϕ-ing is what accounts for the possession of that capacity. This is seen by the fact that Plato accepts that many types of aretai, including the principal moral virtues, are kinds of knowledge, whereas knowing how to ϕ is never a sufficient account of what being good at ϕ-ing is. As thus understood the doctrine does not embrace the thesis that the names of the specific virtues or the adjectives formed from those names are synonymous. ‘Just’ and ‘religious’ are not synonymous, nor are ‘justice’ and ‘religion’ though ‘justice’ and ‘religion’ name the same state in a man, since what makes a man just is the very same thing as what makes him religious, viz. his knowledge of the good. ‘Just’ means roughly ‘treating others properly’ and ‘religious’ means ‘treating the gods properly’: hence the description of someone as just and religious is not a mere hendiadys. The fact that a good man is correctly described by the various nonsynonymous expressions ‘brave’, ‘religious’, ‘just’, etc. allows a sense for the traditional talk of the ‘parts’ of total aret¯e; nothing in the doctrine requires the abandonment of that way of speaking. Hence it is no surprise to find, e.g. in the Meno, the doctrine of the unity of virtue combined with talk (e.g. 78–9) of s¯ophrosun¯e, hosiot¯es, etc. as parts of aret¯e. What has to be abandoned is the substantial claim that the specific aretai are different states of an agent, e.g. that what makes a man courageous is something, a force say or a motivation, different from what makes him just, and that again something different from what makes him religious. In general, then, there is no incompatibility between the doctrine of the unity of virtue and the treatment of the particular virtues as ‘parts’ of total virtue. There is, however, a particular difficulty arising from the talk of ‘parts’ for the interpretation of the Euthyphro which I have been defending. On that interpretation, the concluding section of the dialogue conveys the

the end of the euthyphro

71

doctrine that hosiot¯es is goodness, described as related to the gods. Hosiot¯es thus includes justice, since a good man is just: in the terminology of the dialogue, all his just acts will be among the things he gives to the gods, i.e. every just act is a hosion (religious) act. But the hypothesis which led to this account was that to hosion is a part of to dikaion, and an implication of this is carefully spelled out by Socrates (12d1–3): ‘where there is something hosion, there is something dikaion, but where there is something dikaion there is not everywhere something hosion’, i.e. ‘Whatever is hosion is dikaion, but it is not the case that whatever is dikaion is hosion’. Moreover, this is not Euthyphro’s hypothesis, but Socrates’, and must therefore be assumed to have Plato’s approval. Can it have been Plato’s intention to reach from this hypothesis a conclusion which entails the denial of the hypothesis? (The difficulty is emphasized by Vlastos [1973], 228.) I do not think that this difficulty is particularly serious. We can reconstruct Plato’s thought as follows. In looking for an account of the nature of religion one begins from the insight that fulfilling one’s obligations to the gods is a special instance of fulfilling one’s obligations. One expresses this by the description of religion as a part of justice, and makes the natural assumption that the extension of the part is less than that of the whole, i.e. that religious acts are a subset of the set of just acts. One then asks what part of justice religion is and identifies it as the part to do with the service of the gods. But when the concept of the service of the gods is examined, it becomes clear that the only satisfactory account of it is that it consists in being a good man. Hence while the initial insight that fulfilling obligations to the gods is a special case of fulfilling obligations stands, the inference that religious acts are a subset of just acts falls: since what one has to do to fulfil religious obligations is nothing other than to be good, including being just, it follows that every religious act is a just act, and vice versa. This is not itself the thesis that justice is the same thing as hosiot¯es, which we have identified as the thesis that what makes a man just is the same thing as what makes him religious, viz. knowledge of what is good. Rather it is a consequence of that thesis, derived via the premiss that what the gods need from men is goodness. The identity of justice and hosiot¯es is argued for, by a woefully bad argument, in the Protagoras (330–1): we have now discovered that a stronger case for it emerges in the Euthyphro from the hypothesis that hosiot¯es is a part of justice. That hypothesis is not so much contradicted as superseded: what remains of it is the analytic truth that fulfilling religious

72

the end of the euthyphro

obligations is a special instance of fulfilling obligations. But Plato’s aim is to establish, not an analytic thesis, but the substantial thesis of the identity of justice and hosiot¯es, which is in turn a part of the doctrine of the unity of virtue. That doctrine, it must be emphasized, has nothing to do with the meanings of words. It is a psychological doctrine, which asserts that all human excellence derives, via a single universal motivation, viz. the desire for the agent’s good, from a single intellectual state, the knowledge of what the agent’s good consists in.¹⁰ The analytic hypothesis serves its function by introducing the argument which leads up to the thesis of the identity of justice and hosiot¯es, but which stops short at the point where that thesis should be stated. I conclude then, that the traditional account of the concluding section of the Euthyphro is supported by the fact that it represents Plato as doing in this dialogue what he also does in the Laches and Charmides. Each of these three dialogues seeks for an account of a single virtue and ostensibly fails to find it. Yet each conveys the implicit message that the virtue in question is nothing else than goodness itself, understood as knowledge of the good. That implicit doctrine, with its corollary that each virtue is the same thing as each of the others, is explicitly argued for in the Protagoras. The prima facie objection presented by the fact that in the Euthyphro Socrates argues from the hypothesis that hosiot¯es is a part of justice is not an obstacle to this interpretation; that hypothesis contains an analytic truth which is, therefore, consistent with the substantial doctrine of the unity of virtue. As to the falsehood in the hypothesis, that religious acts are a subset of just acts, that may be regarded as a ladder to be thrown away once it has led to the truth that religion is nothing other than goodness seen as man’s cooperation with the gods. Afterword For further discussion see Calef [1995a] and [1995b] and McPherran [1995] and [1999], ch. 2. ¹⁰ See Penner [1973].

5 The Arguments in the Phaedo Concerning the Thesis that the Soul is a Harmonia1 At Phaedo 85e–86d Simmias puts forward an argument to show that the soul cannot be immortal. The premisses of the argument are firstly that the soul is a harmonia of the elements that compose the body, the hot, cold, wet, dry, and so on (86b6–c2), and secondly that no harmonia can exist unless the elements of which it is a harmonia maintain the proper interrelation. (This point is made in 85e3–86b5 with reference only to a particular case, the harmonia of a lyre, but is clearly to be taken generally.) It follows that when the interrelation of the bodily elements has been dissolved by death, the soul-harmonia cannot exist apart. This argument is presented in the dialogue as posing a major objection to the thesis of the immortality of the soul; those who had been convinced by Socrates’ previous arguments are now thoroughly dismayed (88c). It is, therefore, worth some consideration, particularly since the premiss that the soul is a harmonia expresses a philosophical doctrine whose sense is far from clear. Furthermore, the counter-arguments by which Socrates claims to refute Simmias have provoked considerable disagreement among commentators as to their interpretation, while questions may be raised as to their validity. I propose, then, first to ask what is the meaning of the thesis that the soul is a harmonia and second to examine Socrates’ arguments against Simmias. ¹ An earlier version of this chapter was written in 1970 and delivered (in the author’s absence) at the meeting of the eastern division of the American Philosophical Association in December of that year (see Journal of Philosophy 67 (1970), 724). It is that version which is mentioned by Gallop [1975], 157. In preparing this revision I have profited not only from Gallop’s valuable discussion but also from the comments of Gary Matthews on the 1970 version.

74

the thesis that the soul is a harmonia

The ambiguity of the thesis that the soul is a harmonia emerges from consideration of the different shades of meaning that the word harmonia may have. Formed from the verb harmozein (‘fit together’) it expresses the idea of things being fitted together in an exact arrangement to make an integrated whole, but particular uses express various aspects of the basic sense. Thus the word sometimes means ‘proportion’, particularly in contexts where elements are mingled in proportion, as where Empedocles describes painters mixing their colours ‘mixing them in proportion, some more and some less’ (DK 31 B 23, line 4), and sometimes ‘arrangement’ or ‘organization’ (conveying the idea of the proper relation of parts), as when Heraclitus refers to the harmonia of opposite forces in a bow or a lyre (DK 22 B 51). Or again a harmonia may be identical with a complex of parts in a certain order or arrangement; this is the sense in which the word can mean ‘joint’ or ‘framework’ (see LSJ). Aristotle’s discussion of the soul-harmonia thesis at De an. 1. 4 deals certainly with the first two senses, and perhaps also with the third. In one sense a harmonia is the logos of a mixture, i.e. the ratio of the elements, which may be expressed mathematically. In another it is a combination (synthesis) of physical objects, probably in the sense of the arrangement of a number of physical parts but perhaps also as the complex of those parts in that arrangement. There appears also to be a fourth sense of harmonia that Aristotle ignores, in which a harmonia is something causally dependent on a certain disposition of materials; e.g. a melody is distinct from the strings that produce it, and equally from the tuning of the strings, though without strings there could be no tuning, and without tuning no melody. The word has this sense especially in musical contexts, meaning variously ‘scale’, ‘mode’, or generally ‘music’ (see LSJ). Given, then, that the elements in question are those that compose the human body, the hot, the cold, etc. (which are presumably thought of as different kinds of stuff), there appear to be four possible interpretations of the thesis that the soul is a harmonia of those elements: 1. the soul is identical with the ratio or formula according to which the elements are combined to form the living man; 2. the soul is identical with the mixture or combination of those elements according to that formula; 3. the soul is some entity produced by the combination of those elements according to that formula, but distinct alike from them and from the formula itself;

the thesis that the soul is a harmonia

75

4. the soul is identical with a state of the bodily elements, viz., the state of being combined according to that formula. It might be objected at this point that the third alternative is illusory, since even when the harmonia is a scale or melody it must be considered identical with a mixture of elements. This seems implausible on the assumption that the elements in question are strings or other physical objects composing the instrument that produces the music, but that assumption is mistaken. Just as the elements of a physical organism such as a living human body are the hot, the cold, etc., so the elements of a piece of music are the high and the low, which are conceived of as being mixed together in the proper proportions to give the right notes, either in the sense that each note is thought of as consisting of so much of the high mixed with so much of the low, or in the sense that each mode or scale is produced by combining so many high notes in fixed ratios with so many low notes. The elements, therefore, of a musical harmonia are themselves musical entities, the high and the low, not the physical objects that produce the sounds. This theory is clearly expressed for instance in the pseudo-Aristotelian treatise De mundo 396b7 ff. (DK 22 B 10): ‘Music makes a single harmonia out of different sounds by mixing together high and low, long and short notes.’ On this view of musical harmonia, then, the harmonia cannot be separated from its elements, and so this view does not admit interpretation 3 as an alternative to the others. But while this view of the nature of musical harmonia appears to have been the standard view of musical theory and gives the most exact parallel to other kinds of harmonia (e.g. the formation of physical substances out of the elements), it is emphatically not the view of musical harmonia that Simmias uses to illustrate his thesis. For his presentation of that thesis involves positing a parallelism between two relations, of each of which the terms are (a) a physical object and (b) a non-physical entity causally dependent on that object. Thus corresponding to the incorporeal soul we have the musical harmonia, which is ‘invisible and incorporeal and all-beautiful and divine’ (85e5–6), while corresponding to the physical body we have not the high and the low but the physical strings and pegs of the lyre, which can be broken apart and left lying around after the harmonia has vanished. It is true that Simmias slightly distorts the parallel when he says (85b5–c1) that the soul is a harmonia of the hot, cold, etc. in the body, since a more

76

the thesis that the soul is a harmonia

exact parallel to the strings, etc., of the lyre would be provided by the limbs and organs of the body than by their microscopic elements. But the essential point is the contrast of the incorporeal product with its physical cause, and for that contrast it is unnecessary for Simmias to distinguish the macroscopic parts of the body from their own elements, which are no doubt conceived of as minute but equally corporeal parts. The relation of musical harmonia to its elements, which Simmias is using, cannot therefore be that between a scale or tune and the musical elements of high and low, etc., but must be that between a musical instrument and some non-physical entity produced by a certain state of the instrument. This enables us immediately to eliminate interpretation 2 above. For it would clearly be absurd to make a sharp contrast between the physical elements and the non-physical harmonia, if the latter were just the elements themselves in a certain arrangement. One might as sensibly contrast the invisible, incorporeal plum pudding with the gross, earthy suet, raisins, flour, etc., that compose it. This still leaves three of the original four alternatives, that the relation of the soul to the body is (using the original numbering) (1) that of the ratio of the tuned strings to the strings themselves, or (3) that of the music produced by the instrument to the instrument itself, or (4) that of the state of being in tune to the strings. There is no conclusive evidence from the text which alternative Plato had in mind, or indeed whether he had distinguished the three. Various phrases give some hints, but these are conflicting and inconclusive. Thus the description of musical harmonia as ‘all-beautiful and divine’ might seem most readily applicable to the music produced by the instrument; but when we reflect that the speaker is a pupil of the Pythagorean Philolaus, who might therefore be expected to have a lively reverence for numbers as the source of all things, this argument seems to have little force as between alternatives 1 and 3. Rather stronger is the argument from Simmias’ statement at 92d2 that the soul-harmonia doctrine is accepted by most people; surely, it may be argued, this indicates that the soul is something distinct from a mathematical ratio, since such an obscure theory can never have been held by the majority. On the other hand, the view that the soul is something non-physical, which is yet dependent on a certain state of the body, so that when that state is disrupted the soul is dissipated, might seem to be quite congenial to common sense. But against this we have the comparison of the soul at 86c6–7 to ‘harmoniai in sounds and in all the works of the craftsmen.’

the thesis that the soul is a harmonia

77

‘All the works of the craftsmen’ must include statuary and painting, and probably carpentry and house-building as well. Where, in the products of these arts, are we to look for the non-physical product of the physical elements? Surely in the harmony or proportion of the constituent parts, as exemplified by the proportions between different amounts of different pigments, or by the relations between the dimensions of the various parts of a statue or a piece of furniture. It would be too fantastic to suggest that to every well-made table there corresponds a non-physical entity that is related to the disposition of its parts as the non-physical soul is to the disposition of the bodily elements, or as the non-physical music is to the disposition of the strings. This comparison, then, tends to support alternatives 1 and 4, rather than 3, which was suggested by the claim of popular acceptance for the harmonia thesis. Further difficulty is created by the description of the soul at 86b9 as a mixture (krasis) of the bodily elements. The word krasis, which is regularly used as a synonym for harmonia (e.g. Aristotle, De an. 408a30–1), commonly occurs, like the English ‘mixture’ in contexts that leave it open whether the word refers to the compound of elements that are mixed, or to the state of those elements of being mixed. We have seen that the former alternative is unacceptable, but what about the latter? Can Plato mean that the soul is identical neither with a ratio nor with any non-physical product of a ratio, but rather with a certain state of the body, viz., the state in which the elements of the body are in a certain ratio? While on the one hand this would give a fair account of the comparison of the soul with harmoniai in works of art, on the other hand it fits rather ill with the sharp contrast between the invisible, divine musical harmonia and the physical instrument, while again it might well seem very dubious that most people believe that the soul is nothing other than a bodily state. There appear, then, to be hints of support in the dialogue for all three possible interpretations of the soul-harmonia thesis, which might suggest that Plato has failed to distinguish these alternatives. Before leaving this question, however, we should look at some evidence from other sources, to see whether these throw any light on Plato’s meaning. First, there is the fact already mentioned that in De an. 1.4 Aristotle ignores the possibility that on the harmonia theory the soul might be a non-physical entity causally dependent on the ratio of bodily elements, while explicitly mentioning the possibilities of its being identical with that ratio and of its being identical with an arrangement of parts, which is itself

78

the thesis that the soul is a harmonia

ambiguous between ‘identical with the parts as arranged’ and ‘identical with the state of being arranged which characterizes the parts’. Not only does he not give the first-mentioned possibility as a possible interpretation of the thesis, but he appears to introduce it as an alternative view of the soul, which would be unacceptable to an adherent of the harmonia theory. After producing objections first to the suggestion that the soul is a combination of limbs and then to the suggestion that it is the ratio of the mixture of elements he adds (408a20–1) ‘Is it the ratio which is the soul, or is it rather something separate which comes to be in the parts of the body?’ The implication is that he is here suggesting a more plausible alternative theory, not giving an interpretation that a supporter of the theory would accept as expressing his meaning. Yet in order to take this as a basis for the interpretation of the thesis of the Phaedo we should have to have some reason to believe that Aristotle’s target in the De anima is specifically the thesis proposed in that dialogue, rather than any other version. Such reason is lacking. Aristotle describes the theory as widely current (407b27–8), but does not ascribe it to anyone in particular, and does not mention the Phaedo in his discussion. We know that a version of the theory was held by Aristotle’s follower Aristoxenus² and it is possible that a version different from that of the Phaedo was maintained by Philolaus (see below). The harmonia theory was, then, a current theory of the soul,³ and we are safest to suppose Aristotle to be attempting to expose the fundamental errors of the theory as such, rather than give an exact exposition of any version of it. One might hope to throw some light on the precise sense of the theory by considering its origins, but here too it is impossible to reach any positive conclusions. None of the speakers in the dialogue attributes it to any named philosopher, but since Simmias says that ‘we’ hold the soul to be a harmonia (88b6–7) and Echecrates that he has always been very impressed by that thesis (88d3–4), it would be natural to assume that it was current in the Pythagorean circle to which they belonged. Though they are described as pupils of Philolaus (61d–e, cf. DL VIII. 46), the theory itself is not ascribed to Philolaus by any writer earlier than Macrobius (fourth to fifth centuries ad), who says that Pythagoras and Philolaus held that the soul ² See Gottschalk [1971]. ³ Lucretius discusses the thesis that the ‘sensus animi’ is a harmonia at III. 98–135, attributing it merely to ‘(the) Greeks’.

the thesis that the soul is a harmonia

79

is a harmonia (DK 44 A 23). It is not clear how much reliance can be put on this testimony, since there is obviously a possibility that it may derive ultimately from this very passage of the Phaedo. But whatever the truth about that, Philolaus’ view of the soul cannot be reconciled with the harmonia theory as expounded by Simmias. For at 61a–62b it is implied that Philolaus taught that suicide is wrong on the ground that the soul is put by the gods in the body as a prison for a set time, and must not seek to escape before the time of its release, though a philosopher will welcome death, presumably because his soul will have a better existence in separation from the body. This is supported by a quotation from Philolaus given by Clement of Alexandria (Strom. III. 3.17.1 (DK 44 B 14) ): ‘The soul is yoked to the body and as it were buried in this tomb as a punishment.’ The conclusion is plain that unlike his pupils who take part in the dialogue Philolaus believed that the soul exists independently of the body. It is not impossible that he may have held some version of the theory, in which the soul was a non-physical entity whose association with the body depended on the maintainance of the proper bodily ratio, but the divergence from the view expressed by Simmias is so great that it is obviously fruitless to attempt to interpret the latter in such a way as to assimilate it to some conjectural reconstruction of Philolaus’ views.⁴ I conclude, then, that not only is there no evidence that the soul-harmonia thesis definitely identifies the soul either with a ratio of bodily elements or with the state of being in that ratio or with some entity dependent on the possession of that ratio, but that we can best account for what is said on the assumption that Plato did not clearly distinguish the three possibilities. In considering the arguments against the thesis we shall therefore regard them as concerned with a thesis containing those three alternatives in undifferentiated form. Socrates’ first counter-argument requires little comment. He points out that the thesis is inconsistent with the doctrine accepted earlier that all knowledge is in fact recollection of what the soul had learned in a previous, disembodied existence. No harmonia can exist unless the elements of which it is a harmonia are already in existence, and hence if the soul is a harmonia of the bodily elements it cannot have had a previous non-bodily existence (91e–92e). This argument is cogent against any interpretation of the thesis; ⁴ For a fuller discussion (leading to similar conclusions) see Gottschalk [1971].

80

the thesis that the soul is a harmonia

obviously a bodily state cannot exist unless some body exists of which it is a state, and equally obviously a non-physical entity causally dependent on a ratio of bodily elements cannot exist before those elements have been combined in that ratio. A defender of the thesis might, however, argue that it is not cogent against the identification of the soul with the mathematical ratio itself. For a ratio, being a timeless mathematical entity, cannot itself be said to come into existence whenever it is embodied in some particular material. Since it exists equally at all times, it may truly be said to have existed before a certain body came into being, and hence the argument from recollection does not refute this version of the thesis. This defence is not, indeed, adopted by Simmias, who agrees that his thesis is inconsistent with the doctrine that knowledge is recollection. Nor is it difficult to see why. For it is possible to use this defence only at the cost of making the soul-harmonia a universal; if a pair of elements combines in the ratio 3/4, then indeed that ratio existed before the combination of those elements, but what existed was the ratio 3/4, i.e., the very same ratio that is exemplified whenever three units are related to four units. Thus anyone holding this theory would have to admit that many things must have the same soul, including things generally reckoned inanimate, e.g. geometrical diagrams, since the same ratio that is embodied in a human being and is his soul may also hold between certain lines and angles. It is not, of course, impossible that anyone might have believed something like this; it might, for instance, provide a theory to account for transmigration. Empedocles would on this view have been a bush, a fish, etc. (DK 31 B 117) because one and the same ratio would have been embodied in bush, fish, and Empedocles; i.e. they all had the same soul. Simmias, however, will have none of this; if his version of the theory is interpreted as making the soul a mathematical entity, it must be such an entity individuated by being embodied in these bodily elements. As such it cannot exist independently of the elements by reference to which it is individuated, any more than Socrates’ height can exist independently of Socrates, though in the sense in which Socrates’ height is a universal, say 5 feet 6 inches, that length may be said always to have existed whether or not Socrates exists. This way of looking at the thesis has the advantage of preserving as a necessary truth the thesis that different persons have numerically different souls, whereas on the other interpretation it might be discovered as the result of physiological investigation that two different people had the same soul. It leaves the

the thesis that the soul is a harmonia

81

thesis open, however, to attack on the grounds of inconsistency with the doctrine that knowledge is recollection; whether one considers it refuted on that ground will naturally depend on the strength of one’s conviction of the truth of that doctrine. The remaining arguments are more problematical, in that commentators have disagreed not merely on their conclusiveness but also on how many arguments are employed, and precisely what these arguments are. Like Hicken [1954], Bluck [1955], and Gallop [1975], I discern two arguments, as opposed to the four specified by Philoponus in his commentary on De an. I.4.⁵ These arguments are not, however, presented consecutively; at 92e4–93a10 Socrates gives a set of propositions A1–A4 that are not immediately used in the argument.⁶ Instead at 93a11–12 he begins a new argument by formulating a principle B1 that is to some extent independent of A1–A4. This argument (argument B) continues to its conclusion at 94a12–b3. Then at 94b4 Socrates returns to A1–A4, which he uses to construct the second argument (argument A), whose conclusion is reached at 95a2. While I shall deal first with argument B, it is necessary first to look at A1–A4 in order to determine their relation to B1. Socrates begins by securing Simmias’ acceptance of the proposition that the properties of a harmonia are determined by those of its elements (92e4–93a2; A1). We then have three successive applications of this principle, first to all activities and passivities of the harmonia (a4–5; A2) and then to some particular activities and passivities that are excluded by the principle. It is impossible for a harmonia to lead or control its elements, but it must rather be controlled by them (a6–7; A3), and it is impossible for it to be affected or behave in any way contrary to that which its elements determine (93a8–9; A4). At 93a11–12 we have the principle that marks the beginning of argument B: ‘Well now, doesn’t every harmonia have to be the kind of harmonia that corresponds to the way that it is ordered?’ (B1).⁷ It is not easy to find a translation that is both exact and comprehensible but the next sentence, giving an application of the principle, makes the ⁵ The grounds for rejecting Philoponus’ analysis, which is followed, not without incoherence, by Archer-Hind [1883] and Hackforth [1955], are cogently stated by Hicken [1954], 17–18 and by Gallop [1975], 170–1. ⁶ See appendix. ⁷ The verb translated ‘order’ (harmozein) may equally properly be rendered ‘attune’, ‘arrange’, or ‘organize’. I adhere to ‘order’ throughout.

82

the thesis that the soul is a harmonia

meaning fairly clear; if a harmonia is more ordered it is more (of) a harmonia, and if it is less ordered it is less (of) a harmonia (93a14–b2; B2). The sense of B1 itself can then best be expressed formally, as follows, that where F stands for an adjective that can apply to a harmonia, and where F-ly stands for the adverb formed from the adjective for which F stands, then for all X, if X is a harmonia, if X is ordered F-ly, X is an F harmonia. While this certainly goes beyond anything that is said in A1–A4, this ‘formal’ account of the dependence of a harmonia on what gives rise to it may nonetheless be seen as continuing the line of thought begun there. The crucial difference, emphasized by Gallop [1975], 158, is that whereas in A1–A4 we are concerned with the dependence of the harmonia on its elements, B1 states the dependence of the harmonia on the order or arrangement of the elements. Argument B proceeds by way of three further premisses, B3 that a soul is no more or less (of) a soul than any other (b4–6), B4 that there are some good souls and some bad (b8–c2), and B5 that a good soul is in order and a bad soul out of order (c3–10). None of these premisses is felt to require any justification or explanation; the sense of the third is clearly that the good man is not a prey to the conflicting desires and impulses that are the mark of the bad man, but has all his wants properly under control with a view to the attainment of the right ends. We now come to one of the most problematical passages in the argument: at d1–5 Socrates says that premiss B3 is the same as the proposition (B7) that no harmonia is more or less (of) a harmonia than any other, and Simmias agrees. Of course B3 is not as it stands equivalent to B7, and the question is what additional assumptions Plato must have used in order to produce what he regarded as a valid equivalence. Clearly we cannot derive such an equivalence simply by making the most obvious assumption, viz., the assumption under examination in this argument, that the soul is a harmonia (B6), since taken together with B3 that would still allow that some harmoniai might be more or less harmoniai than others. But did Plato see that? I am inclined to think that he did not, but rather, assuming the soul to be a harmonia, took this to imply that whatever is true of soul is also true of harmonia (using the terms in the unquantified style familiar from Aristotle). In effect this is to confuse implication with equivalence, which seems a possible error for Plato to commit at this stage in his philosophical development. (See Gallop [1975], 162–3.)

the thesis that the soul is a harmonia

83

The standard modern interpretation of this sentence, adopted by ArcherHind, Bluck, Hackforth and Hicken (but not by Burnet [1925] or Gallop), differs from the above in taking Socrates to be asserting not a general proposition about all harmoniai but a specific proposition about the sort of harmonia that souls are, viz., that no soul-harmonia is more or less a harmonia than any other. As this requires an admittedly unnatural reading of the text as it stands, many scholars (see Hackforth’s note, [1955], 116) have suggested deleting the word harmonias from d4, thus making the sentence read ‘And this (namely, the admission that no soul is more or less a soul than any other) is the admission that no (soul) is more or less a harmonia than any other.’ But since this emendation lacks any manuscript authority and destroys what looks like a very emphatic and deliberate parallelism of sentence construction, it is worth asking whether there are cogent grounds for emending the text or for reading the received text in a sense other than its natural one. The strongest ground appears to be that urged by Hicken, that since the argument is to depend on the assumption that some harmoniai (in particular, goodness) admit of degrees, it would be flatly inconsistent if Plato also relied on the assumption that no harmonia admits of degrees. I doubt the cogency of this argument, which seems to depend on a confusion in the notion of ‘degrees of harmonia’. For the thesis that some harmoniai (e.g. goodness) admit of degrees comes to this, that some things, e.g. the parts of the soul, may be so arranged as to approximate more or less closely to some norm that represents the perfect arrangement of those things. But that is in no way incompatible with the thesis that I take Plato to be asserting at 93d1–5, viz., that if what a thing is is a harmonia, it cannot be more or less a harmonia than anything else that is a harmonia. This amounts to an extension of the truism ‘Everything is what it is,’ and applies alike to perfect and to imperfect orderings. Every ordering of parts of the soul, at whatever remove from the norm, is equally an ordering of parts of the soul. There is, then, no general inconsistency between the theses ‘No harmonia is more or less a harmonia than any other’ and ‘Some things are more ordered (in Platonic terms ‘‘partake more of order’’) than others.’ Plato, however, thinks that contradiction is generated if one says that one harmonia is more ordered than another; that he is wrong even in this restricted thesis will be seen once the argument is viewed as a whole. The next step (93d6–8; B8) is that something that is neither more nor less a harmonia is neither more nor less ordered; this follows directly by

84

the thesis that the soul is a harmonia

contraposition from B2. Another problematic sentence follows (d9–11; B9): ‘And does that which is neither more nor less ordered partake more or less of order, or to just the same extent? To the same extent.’ At first sight it might appear that this is the converse of B8. But, firstly, in contrast to the previous sentence, where the subject is ‘that which is neither more nor less a harmonia’ the predicate of B9 is ‘partakes of (i.e. is characterized by) harmonia more or less.’ While one might indeed see here a confusion between predication and identity, it is more charitable to take the shift in terminology as indicating that a new point is being made by the introduction of a further premiss. Secondly, if B9 is read as the converse of B8, it has no subsequent role in the argument; whereas if it is read as ‘Something which is neither more nor less ordered has neither more nor less order,’ we have a straightforward argument, as will be seen immediately. Socrates next concludes (d12–e2; B10) that no soul is more or less ordered than any other, giving as premiss B3. In fact B10 follows, not from B3, but from B6, 7, and 8 (see appendix). Socrates’ derivation of B10 from B3 presumably indicates that he is relying on the fallacious derivation of B7 from B3 (together with B6). From this point the argument is straightforward. From B9 and B10 it follows that no soul is more or less ordered than any other (e4–5; B11), and hence by B5 and B11 that no soul is better or worse than any other (e7–94a10; B12). It is agreed (94a12–b3) that this conclusion is absurd, and hence one of the premisses from which it is derived must be false; obviously, the premiss to be rejected is the assumption that the soul is a harmonia. It appears, therefore, that we have in argument B a single argument that is, despite some obscurities, clear in its main lines and (perhaps not so clearly) fallacious. The flaw is not simply the fallacious equivalence of B3 and B7, since one might patch this up by introducing B7 as a premiss; it is perfectly plausible to suggest that, where F is a predicate saying what kind of thing its subject is, if A and B are both Fs, A can’t be more (of) an F than B. That emendation still leaves a fallacious argument, though the text leaves room for more than one account of the fallacy. One possibility is the following. The argument is invalid through Plato’s failure to recognize that ‘... is a harmonia (ordering)’ and ‘... is ordered’ are incomplete predicates, requiring to be completed as ‘... is an ordering of elements of type E
(hereafter ‘... is an E-ordering’) and ‘... is ordered

the thesis that the soul is a harmonia

85

with respect to elements of type E’ (hereafter ‘... is E-ordered’). The distinction is crucial, since one and the same thing may consist of elements of two types E and E
, such that that thing is an ordering of elements of type E, while it neither is an ordering of elements of type E
nor is characterized by order in respect of elements of that type. Imagine a university composed of independent multi-disciplinary colleges, which also has a faculty structure crossing collegiate boundaries. Imagine further that, while the relations of the faculties to one another and to the university are fully organized, intercollegiate relations and relations between the university and the colleges are anarchic. (Readers familiar with the Universities of Oxford and Cambridge will note elements both of verisimilitude and of exaggeration.) The questions ‘Is the university an organization?’ (I treat the term organization as interchangeable for the purpose of this illustration with ordering) and ‘Is the university organized?’ (= ‘ordered’) have no determinate answers. It is an organization of faculties, but (despite being composed of colleges) it is not an organization of colleges, nor is it organized in respect of the colleges. The situation of the soul, on the assumptions of the harmonia theory, is parallel. It is ex hypothesi an ordering of the bodily elements (allowing for the ambiguity of that thesis as discussed above). In virtue of that relation of bodily elements the soul consists of psychic elements, desires, intellect, etc., which may themselves be organized in a coherent way (the state of a good soul), or may lack organization (the state of a bad soul). No soul, good or bad, is more an ordering of bodily elements than any other, for an ordering of bodily elements is just what a soul is. But orderings of bodily elements may be more or less ordered in respect of psychic components. The various steps of the argument have now to be rephrased in terms of the complete predicates ‘... is an E-ordering’ etc. The crucial changes are as follows: for B2 substitute B2
—Any E-ordering that is more E-ordered is more of an E-ordering, and any that is less E-ordered is less of an E-ordering; for B5 substitute B5
—A good soul is ps-ordered (possesses psychic order), a bad soul is ps-disordered (lacks psychic order); for B6 substitute B6
—The soul is a ph-ordering (i.e. an ordering of physical elements); for B7 substitute B7
—No E-ordering is more an E-ordering than any other;

86

the thesis that the soul is a harmonia

for B8 substitute B8
—That which is neither more nor less an E-ordering is neither more nor less E-ordered; for B9 substitute B9
—That which is neither more or less E-ordered has neither more nor less E-order. It is now clear that the argument fails to lead to a reductio. From B6
,
7 , and 8
we derive B10
: ‘No soul is more or less ph-ordered than any other’. And from B9
and B10
we derive B11
: ‘No soul has more or less ph-order than any other’. But from B5
and B11
it is impossible to derive B12. For by B5
bad souls possess less ps-order than good souls, while by B11
all souls have equal amounts of ph-order. B12 does not follow, and hence there is no contradiction with B4 and no reductio. It is, then, not inconsistent to maintain that an entity that arises from the organization of bodily elements may itself contain parts or elements of another sort that lack organization, or alternatively that a certain organizational state of bodily elements (i.e. the state of being ensouled) may be in a particular case further characterizable as a state of psychic disorganization. An alternative diagnosis of the fallacy is provided by Gallop [1975], 163–6. On this view the flaw is not the failure to supply different completions for an incomplete predicate at different stages of the argument but rather a failure to observe an ambiguity in the term harmonia between ‘tuning’ or ‘order’ (‘attunement1 ’ in Gallop’s terminology) and ‘correct tuning’ or ‘good order’ (‘attunement2 ’). While every attunement1 is equally an attunement1 different attunements1 may be characterized by attunement2 to different extents. (Essentially the same point is made by pointing out that the sentence ‘A is more ordered than B’ is ambiguous between ‘A is more of an ordering than B’ and ‘A is better ordered than B.’) Hence B12, ‘No soul is better or worse than any other,’ does not follow from B11: ‘No soul has more or less order than any other’. For the predicate in B12 is an instance of ‘... is characterized by attunement2 ’ whereas that in B11 is ‘... is characterized by attunement1 .’ As far as I can see, this suggestion fits the text as closely as that given above. Since Plato is arguing from the premiss that all souls share equally in the ordering of physical elements to the (absurd) conclusion that all share equally in the good ordering of psychic elements, it is not surprising that the text should leave it open whether the failure of the argument turns on the slip from ‘ph-order’ to ‘ps-order’ (my

the thesis that the soul is a harmonia

87

suggestion) or on that from ‘order’ to ‘correct order’ (Gallop’s suggestion). Either is sufficient to generate a fallacy, while the text allows for both. Argument A is resumed immediately. It is agreed that in a sensible man the soul controls and opposes bodily inclinations such as hunger and thirst (94b4–c1; A5). Socrates now (c3–e6) recalls steps A3 and A4, to the effect that a harmonia can never control or be opposed to its elements. Hence (94e8–95a2) the soul cannot be a harmonia. This concludes the discussion of the thesis. In this case it is clear that the crucial slip is that from ph-order to ps-order. (This may be some small reason for preferring that account of the fallacy in argument B.) If the harmonia thesis is to be refuted by this argument, the soul’s control of bodily inclinations must be a case of what is denied by A3, viz., the soul’s controlling its elements. But according to the thesis, the elements of which the soul is a harmonia are not any sort of inclinations but bodily elements, the hot, the cold, the wet, the dry, etc. On Simmias’ statement of the thesis the activity of the soul is determined by the interrelations of those elements, so that, for instance, a certain mixture of hot and cold will produce anger in the soul, or a certain proportion of dry to wet the desire for a drink. But in argument A Socrates treats such events as being angry or wanting a drink as themselves impulses of the physical elements of which the soul is ex hypothesi a harmonia, and insists on the incompatibility of A3 and A4 with the view that the soul opposes these impulses, in the sense that the reason often opposes such desires. Clearly, there is no inconsistency. All that a defender of the harmonia thesis need say is that in the case of such a conflict of reason and desire we see, not the soul-harmonia opposing and controlling its elements, but rather one part of the soul-harmonia opposing and controlling another. And he might add that of course the controlling part is as determined as the controlled part by some disposition of bodily elements. In effect this would be to replace the soul-body dualism of the Phaedo with an account akin to that of the divided soul in the Republic and Phaedrus, with the addition of a thesis of physicalistic determinism of the functioning of all parts of the soul. Plato’s intention in arguing against the harmonia thesis is no doubt partly to resist this determinism, in support for the insistence on the autonomy of the rational soul that pervades the Phaedo. Yet our discussion of argument B

88

the thesis that the soul is a harmonia

has shown that it is hard to disentangle arguments against the thesis that the soul is a harmonia of physical elements from the thesis that it is a harmonia of elements of any kind. (Gallop’s analysis of the fallacy is independent of the nature of the elements.) Yet the development of Plato’s thought requires that he should at some stage have made that distinction, since there seems to be a perfectly good sense in which the tripartite soul of the Republic and Phaedrus may be called a harmonia, in that it is a composite entity composed of parts whose relations affect its functioning as a whole. Perhaps argument B of the Phaedo discussion was directed specifically against the account of the soul as a physical harmonia, and was not intended to apply to any other sort. (This would count against Gallop’s interpretation.) Alternatively, assuming Gallop to be right, Plato had spotted the fallacy by the time he developed the theory of the tripartite soul and saw that that theory was safe against the arguments of the Phaedo. It is, however, necessary to attribute extraordinary obtuseness to Plato if one accepts, with Hicken and others, that the arguments of the Phaedo are conclusive against the thesis. For if those arguments are sound, they refute the theory of the tripartite soul, in which case the whole political organization of the Republic is based on a psychological theory that Plato had already (assuming the priority of the Phaedo) refuted.

the thesis that the soul is a harmonia 89

Appendix

Analysis of the arguments 92e4–93a2

A1

a4–5

A2

a6–7

A3

a8–9

A4

a11–12

B1

a14–b2

B2

b4–6

B3

b8–c2 c3–10

B4 B5

(understood) B6 d1–5 B7

The properties of an ordering are determined by those of its elements. The activities and passivities of an ordering are determined by those of its elements. An ordering cannot control its elements but must be controlled by them. An ordering cannot be affected or behave in any way opposed to (the behavior of ) its elements. Any ordering that is ordered F-ly is an F ordering. Any ordering that is more ordered is more of an ordering, and any that is less ordered is less of an ordering. No soul is more or less of a soul than any other. There are some good souls and some bad. A good soul is ordered (possesses order); a bad soul is disordered (lacks order). The soul is an ordering. No ordering is more of an ordering than any other That which is neither more nor less an ordering is neither more nor less ordered. That which is neither more nor less ordered has neither more nor less order. No soul is more nor less ordered than any other. No soul has more nor less order than any other. No soul is better or worse than any other.

d6–8

B8

d9–11

B9

d12–e2

B10

e4–5

B11

e7–94a10

B12

a12–b3

Concl. Since B12 contradicts B4, B6 is false.

P(remiss) from A1 from A2 from A2

P from B1

P P P P from B3, B6 (invalid) from B2 P from B6, B7, B8 from B9, B10 from B5, B11 from B4, B6, B12 (RAA)

90

the thesis that the soul is a harmonia

b4–c1

A5

(understood) A6

The soul controls and opposes bodily desires. An ordering opposes and controls its elements. A6 contradicts A3 and A4

c3–e6

A7

e8–95a2

Concl. B6 is false.

P from B6, A5 from A4, A4, A6 From B6, A7 (RAA)

Afterword For further discussion see Barnes [1979], vol. ii, ch. 8 (= 2nd rev. edn. [1982], ch. 22), sects. c-d, Bostock [1986], ch. 6, Scaltsas [1990], Caston [1997], 319–26, and Wagner [2000].

6 Plato and Aristotle on the Criterion of Real Pleasures Both Plato and Aristotle emphasize the diversity of pleasures. Animals of different species, and men of different temperament, physiological condition etc. find pleasure in radically different types of object and activity; the different kinds of men are, therefore, likely to differ radically in their judgements on the pleasantness or unpleasantness of any particular object or activity. One possible response to this situation is to espouse a form of relativism concerning pleasures: every case of pleasure is a case of something’s being pleasant to someone or something, and questions such as ‘Which is the pleasantest sort of life?’ are dismissed as improper. Unless one is prepared to espouse a corresponding form of relativism concerning goodness, the adoption of such a position must involve the abandonment of any connection between pleasure and goodness. Since both Plato and Aristotle maintain some connection between the two, both have an interest in discriminating between pleasures, e.g. in showing that some pleasures are really pleasanter than others, or in showing that some really are pleasures, whereas other purported pleasures are not pleasures at all, or are so only with some qualification. An important part of Plato’s treatment of this topic is his discussion of false pleasures in the Philebus: while that discussion is intended to exclude many kinds of putative pleasures from the good life on the ground that they are unreal or false, it does not appeal to the judgement of any particular sort of person as decisive in discriminating true from false pleasures. In Rep. IX, however (580d–583a), Plato makes such an appeal to establish that the philosophic life is pleasanter than the life of ambition and the life of bodily appetite, while Aristotle makes fairly frequent use

92

the criterion of real pleasures

of a similar criterion, especially in the Nicomachean Ethics. In this paper I shall examine the use of this pattern of argument by Plato and Aristotle, considering, among other things, whose judgement is appealed to, what that judgement is supposed to establish, and whether either philosopher is successful in his employment of this pattern of argument. At 580c9 Plato concludes his first argument in support of the thesis proposed for examination at 544a5–8, viz. that of the types of men whose characters correspond to the various types of political organization, the best man has the most completely satisfactory life and the worst the most completely unsatisfactory (ho aristos eudaimonestatos kai ho kakistos athli¯otatos). The first argument (576b11–580c9) consists of the presentation of the contrasting pictures of the best and worst types of man, the former the ‘Platonically just’ man who enjoys total psychic harmony in that all his lower impulses are harmonized under the direction of the intellect informed by knowledge of the Good, the latter the ‘tyrannical’ man (i.e. the man totally enslaved to the satisfaction of one of his lower appetites) who actually rules as a tyrant. The latter has a life which is, objectively speaking, not worth living (i.e. he is athlios, wretched or unfortunate) because he is, in the modern sense, unhappy, i.e. his life is not at all as he wants it to be; his unhappiness is manifested both in his finding his first-order pursuits predominantly unsatisfying and unpleasant and in the fact that his second-order reflective attitudes are those of anger and shame. This argument commits Plato to the position, surely correct, that it is a necessary condition of having a completely worthwhile life (i.e. of being eudaim¯on) that one’s life should be happy. It does not commit him to the converse, that happiness is a sufficient condition of eudaimonia. That is a point in his favour, since the latter proposition is false; a man who lives and dies happy in the false belief that his beloved wife is faithful to him is not eudaim¯on, a man blessed by heaven. Consequently, this first argument goes only a fairly short way towards establishing Plato’s desired conclusion. He wishes to prove that, of all the types of man considered, the best man has the most worthwhile life and the worst man the most unsatisfactory life. But all that he can fairly claim to have shown by the first argument is that the life of one kind of bad man is an extremely unhappy life; he has further to establish (i) that that kind of man is the worst kind of man, and (ii) that his life is the most unsatisfactory life. (i) is supposed to be established by the comparison of types of man which

the criterion of real pleasures 93 runs from 548c6 to 576b10, but no argument is given for (ii). This is a serious omission; for, since conditions additional to happiness are necessary for eudaimonia, it cannot be assumed that eudaimonia is directly proportional to happiness. Further, this first argument does nothing at all to establish the first conjunct of the conclusion, that the best man has the most worthwhile life. This gap is filled by the second and third arguments, both of which aim to show that the best man has the pleasantest life of all men. It is important to observe that the pleasantness of a life, as conceived in these arguments, is not determined purely by the attitude of the agent to the activities which make up his life or to his life as a whole. Herein lies the crucial difference between Plato’s conception of a really pleasant life and the ordinary conception of happiness. In setting up the second argument Socrates is careful to acknowledge that each of the three types of man will insist that his own life is the pleasantest of all (581c7–10). The argument does not proceed by way of showing that the Platonically just man’s life is pleasanter to him than the life of the ambitious or the appetitive man is to him. Rather, accepting that each man’s life is fully pleasant to him, it tries to show that the just man’s life is pleasanter tout court than the others. The argument thus requires a standard of pleasantness tout court by reference to which the inferior man’s judgement that his life is completely pleasant to him is shown to be, not false indeed, but irrelevant to the question of which life is the most worthwhile. Socrates claims that the judgement of the just man as to which life is pleasantest provides that standard; we have now to consider his grounds for that claim. The argument itself is extremely simple. On any disputed matter the best judge is the man who excels in experience (sc. of the matter in hand), intelligence (phron¯esis) and reason (logos) (582a4–6). But the intellectual excels the two other types in all three respects. First, with respect to experience, he has experience of the two other kinds of pleasure, whereas the ambitious and appetitive men are without experience of his pleasures. He has experienced appetitive pleasures from his childhood onwards, and has experienced ambitious pleasures in his enjoyment of the praise of his fellow intellectuals for his intellectual successes. On the other hand ‘it is impossible for anyone but the philosopher to have tasted how much pleasure there is in the contemplation of reality’ (582c7–9). Then, since intelligence and reason are the attributes which distinguish the philosopher from the other types of man, he naturally excels with

94

the criterion of real pleasures

respect to those also. Hence he is the best judge of pleasures, and consequently his judgement that his own life is the pleasantest is to be accepted as establishing that it is pleasantest tout court. Similarly, his judgement is decisive in awarding second prize in the contest of pleasantness to the life of ambition, since it is more like his own life than is the appetitive life, and consequently in relegating the latter to third place (583a6–11). Our first question, viz. what it is that the philosopher’s judgement establishes to be the case, appears to admit of a straightforward answer. It establishes that his life is the pleasantest of all lives, that is to say, the pleasantest absolutely, not just the pleasantest for him. But while that is stated perfectly plainly in the text at 583a1–3, there is a certain problem as to what status the philosopher’s judgement assigns to the inferior pleasures. Are they simply not as pleasant as his, or are they rather not really pleasant at all? Consequently, is the philosophic life simply the pleasantest of three pleasant lives, or is it the pleasantest life of the three insofar as it alone contains real pleasures, or perhaps insofar as it contains the highest proportion of real pleasures? The question is raised, of course, by the fact that the third argument attempts to establish that only the philosopher’s pleasures are really pleasures. Now we should not automatically read that view of inferior pleasures back into the second argument; Plato might be proceeding cumulatively, arguing firstly that the worst man is the unhappiest of all, secondly that the best man has the pleasantest life of all (even if other kinds of life really are pleasant), and thirdly that in fact only the best man has any real pleasures. There is, however, some textual evidence to suggest that the latter thesis may itself be part of the conclusion of the second argument also. Firstly, at 581d10–e4 the philosopher is said to consider inferior pleasures, in comparison with the pleasures of philosophy, as ‘very far from pleasure’. Secondly, at 582e8–9 Glaucon asserts that necessarily ha ho philosophos ... epainei al¯ethestata einai, which may be rendered either as ‘the things that the philosopher praises are the most real’ or as ‘what the philosopher says in praise (sc. of what he praises) is the truest’. On the former rendering it is immediately implied that the inferior pleasures are less real pleasures than those of the philosopher. That implication is not contained in the latter rendering, but when we recall that part of the philosopher’s praise of his own pleasures consists in describing their rivals as ‘far from pleasure’, it follows that on either

the criterion of real pleasures 95 rendering the philosopher counts the inferior pleasures as less than fully real. It appears, then, that while the main conclusion to be established by appeal to the philosopher’s judgement is that his is the pleasantest life of all, it is also implied that no other life is really pleasant. For, asked whether the appetitive and ambitious lives are pleasant, the philosopher must say ‘No’, since those lives would not be pleasant to him; hence by the principle that the philosopher’s judgement decides what is the case absolutely, those lives are (absolutely) not pleasant. Some inferior pleasures may perhaps be, or approximate to, real pleasures; we may recall the statement that in the case of the philosopher each part of the soul enjoys the best and most real pleasure of which it is capable (586e4–587a1). But a life given over to inferior pleasures is not a real instance of a pleasant life. Since the judgement of the philosopher is set up as a standard for the comparison of lives, it does not, therefore, determine which of a number of really pleasant things is pleasantest; rather, it determines which of a number of putatively pleasant things (i.e. putatively pleasant lives) is really pleasant. To that extent it is correctly described as a criterion of real pleasure. We must now consider the soundness of the argument. Given the truth of the premisses that the best judge of any disputed matter is he who excels all other men in experience, intelligence, and reason and that the philosopher excels all other men in experience of pleasures, intelligence, and reason it plainly follows that the philosopher is the best judge of any dispute about pleasures. Any criticism of the argument must, therefore, be directed at the truth of the premisses. With regard to the truth of the first premiss, while it is true that, other things being equal, A is a better judge of any disputed question than B if A is more experienced, more intelligent, and more rational than B, there are certain areas of dispute, notably those of connoisseurship, where intelligence and rationality are much less relevant than highly developed discriminatory capacities. This point is of some significance, in that it might be claimed with some plausibility that disputes about the relative pleasantness of different ways of life are much more matters of taste than of rational judgement; hence if they are to be settled by appeal to the judgement of experts, the expert in question should be someone more like a wine-taster than a philosopher. The first premiss then may be criticized on the ground that, while purporting to state sufficient conditions for superior expertise in any area, it ignores conditions necessary

96

the criterion of real pleasures

for expertise in some areas where the philosopher’s particular merits are irrelevant. Hence the claims of the philosopher are unduly favoured from the outset. Further difficulties arise about the second premiss, that the philosopher excels the other types of men in experience, intelligence, and reason. Plato offers no arguments for the superiority of the philosopher in intelligence and reason, plainly taking it as obvious. But argument is required, for Plato nowhere explains why someone possessed of the highest intelligence might not prefer to pursue the goals of ambition or appetite. It might be urged that anyone of that level of intelligence must see that those goals are, by comparison with the goals of the intellect, not worth pursuing; but this very argument is presented as part of the proof of that crucial proposition. Plato is, then, faced with a dilemma: either he gives an independent proof of the comparative worthlessness of the goals of the ambitious and appetitive lives, in which case the appeal to the philosopher’s judgement is otiose, or he fails to do so, in which case the appeal itself fails, since one of its premisses is insufficiently supported. Nonetheless, the appeal to the philosopher’s judgement must stand or fall with the claim that he excels the other types of men in experience of pleasures. For the objection of the previous paragraph could show no more than that men of the other types might be as intelligent and as rational as philosophers. If Plato can show that, typically, the philosopher has more experience of pleasures than the others, then, failing some further argument to show that the other types of men are typically more intelligent and rational than philosophers, he will have made out a strong case in favour of his reliance on his appeal to the philosopher’s judgement. This might even go some way towards meeting the objection which we brought against the first premiss, since the notion of experience might perhaps be extended to include discriminatory capacity; it must at least be conceded that the latter is likely to improve as the former is widened. The claim to superiority in experience is therefore crucial. But the claim cannot be sustained. For Plato has to show that the philosopher has sufficient experience of the pleasures of the other types of life to count as rejecting them on the basis of experience, whereas the other types of men lack sufficient experience of his pleasures to reject them on the basis of experience. But all that he urges is that the philosopher must have had some appetitive pleasures and some ambitious ones; as regards his

the criterion of real pleasures 97 negative thesis he says that the money-loving (i.e. appetitive) man need not have had any philosophic pleasures and that it would be difficult for him to do so even if he wanted to, but in contrasting the philosopher with the ambitious man he says that no one but the philosopher can have the pleasure of contemplating reality. It is obviously open to the partisan of appetite or ambition to reply that, in order to have experience of a type of pleasures sufficient for the purposes of this argument, it is necessary to experience those pleasures in their most developed form, and to pursue them with the degree of enthusiasm that the devotee gives them. Hence you can’t claim to know what the pleasures of appetite are really like if all you know is how much you enjoyed afternoon tea in the nursery, or how much you like your refreshing glass of milk after a strenuous bout of dialectic; only the man who can really appreciate a debauch or an orgy knows what the pleasures of appetite are like. A similar point can be made for the case of ambition. Plato then faces a dilemma: either really experiencing the pleasures of a given way of life itself requires that one prefer those pleasures to any others throughout one’s life, or some less stringent condition is adopted. In the former case it is a necessary truth that no one has experience of the pleasures of any way of life other than his own, since anyone who abandons one way of life for another thereby shows that he did not really appreciate it. This position is of course destructive of the claim of the philosopher, or of anyone else, to have experience of the pleasures of more than one way of life; no such criterion is available. If, on the other hand, some less stringent condition is adopted, it will be impossible to establish other than empirically the asymmetry between the experience of the philosopher and that of his rivals which the argument requires. Suppose, for instance, that the following condition were accepted as sufficient; someone has experience of the pleasures of a given way of life if and only if for some part of his adult life he has made those pleasures his predominant pursuit. It cannot now be claimed a priori that the philosopher and he alone has the necessary experience, thus defined, while as an empirical claim it lacks any plausibility whatever. The same result holds if the condition is weakened still further, e.g. if it is sufficient that one should occasionally have experienced full-blown examples of the pleasures of each way of life. It might perhaps be claimed in favour of the asymmetry that, while it is contingently possible to experience occasional full-blown appetitive and ambitious pleasures without living the corresponding way

98

the criterion of real pleasures

of life, it is necessarily impossible to experience full-blown philosophic pleasures without being a philosopher, since an essential element in such pleasures is one’s synoptic understanding of reality. But (a) there seems to be a similar conceptual impossibility in isolating full-blown pleasures of ambition from an ambitious way of life, and (b) it has still not been shown that someone who had achieved such a synoptic view might not thereafter abandon the philosophic life for one of its competitors. If that possibility is excluded by stipulation, then we are back in the first situation, in which there can, by definition, be no experience of more than one way of life.¹ The claim that the philosopher excels the other types of men in his experience of pleasures must therefore be rejected, and with it the claim that the philosopher’s judgement provides the criterion of what is a really pleasant life. The claim to superior experience cannot be maintained as an a priori truth, while as an empirical claim it has neither adequate support in Plato’s text nor any inherent plausibility. On reflection this result ought not to be unwelcome to Plato, since the claim that the philosopher judges by experience what is a really pleasant life and what is not has highly embarrassing consequences. Does the philosopher really pronounce his life more pleasant than a life of perverted or sadistic pleasures from experience of both kinds of pleasures? As a good man the philosopher must surely be without experience of those activities at all, not to speak of experience of the pleasures of those activities. It is a natural response to such examples to maintain that one does not need experience to disqualify such ways of life from counting as really pleasant. They are not really pleasant in themselves, one is inclined to say, but are pleasant only to someone in a corrupted or diseased state, as honey may taste bitter when one is sick. Here the implied criterion of what is really pleasant is not the judgement of the man of experience, but the judgement of the man in a state of physical or psychic health, i.e. the man whose physical and/or psychological components are organized and functioning in the way they ought. In view of the fact that we were led to postulate an appeal to such a criterion by consideration of difficulties which Plato’s criterion was unable to cope with, it is all the more interesting that this, rather than experience, is the criterion which Aristotle regularly relies on in discriminating pleasures. We must now consider his ¹ On the arguments in Rep. IX see also ch. 13 of this volume.

the criterion of real pleasures 99 use of it, with a view to deciding whether it is ultimately more successful than Plato’s appeal to experience. We should first ask why it is important for Aristotle to distinguish between qualified and unqualified pleasures. Firstly we must take account of his doctrine of meaning. Impressed by the relativity of predicates such as ‘good’ and ‘pleasant’, he recognizes that the questions ‘Is x good?’, ‘Is x pleasant?’ etc. are as they stand ambiguous, requiring to be disambiguated by unpacking the predicates into ‘F absolutely’ and ‘F with such and such qualification’ (to this person, in these circumstances etc.). This process of disambiguation requires (reasonably) clear criteria for the application of the specific predicates. Further, in the case of any predicate which is applied both with and without qualifications, the unqualified use is primary; e.g. understanding what it is for something to be good for a sickly body presupposes the understanding of what it is for something to be good for the body simpliciter, i.e. for the healthy body, since things are good for the sickly body insofar as they cause it to approximate to the state of the healthy body. That understanding requires a grasp of the conditions necessary and sufficient for the primary application of the predicate. So the complexity of the concept requires that the investigation of pleasure proceed via the distinction of qualified from unqualified pleasures. The investigation of pleasure is, of course, essential to the main purpose of Aristotle’s ethical enquiry, the identification of the good for man and the conditions under which it can be attained. Aristotle inherits from Plato and Eudoxus a debate on the relation of pleasure to the good for man; according to one extreme position it simply is the good, according to another it is not merely not the good, but not good in any respect. The distinction between qualified and unqualified pleasures affords an elegant resolution. Some at least of the anti-hedonist arguments are effective only against qualified pleasures, such as those of the sick or the wicked, while unqualified pleasure has at least a strong claim to be identified with the good for man. Again, Aristotle consistently acknowledges an obligation to preserve as much as possible of ordinary beliefs; one such belief which his theory must show to be true is that the life of eudaimonia is pleasant (EN 1099a7; cf. 1098b9–12). Given the identification of the life of eudaimonia as the life of excellent activity, the acceptance of the judgement of the physically and psychologically sound man as the criterion of unqualified pleasure, and the requirement that one is not a good man unless one

100

the criterion of real pleasures

enjoys acting well (1099a17–20), it follows that the life of eudaimonia is not merely pleasant, but unqualifiedly pleasant. Since unqualified pleasure is the highest grade of pleasure, the life of eudaimonia wins top marks in respect of pleasure as in other respects. The fact that a depraved man finds depraved activities pleasant is no objection, for their being pleasant to him does not count against their being unpleasant simpliciter, since they are unpleasant to the good man (1173b20–5; 1176a19–24). The results which Aristotle reaches by the use of his criterion are, then, of the highest importance for his whole project; but is the criterion itself sound? If it is to work, two conditions must be satisfied. Firstly, it must be possible to identify the physically and psychologically sound man independently of the fact that he takes pleasure in those things of whose pleasantness his judgement provides the criterion. Secondly, persons thus identified should agree, broadly, in their judgements of what is pleasant and unpleasant. The necessity of these conditions becomes apparent when we consider the perceptual analogies on which Aristotle regularly relies in proposing the judgement of the good man as his criterion of pleasures. Thus if we stipulate that to be sweet just is to taste sweet to the healthy man, we must be able to identify the healthy man other than as the man to whom things which are sweet taste sweet. For if we have no other way of identifying him, our proposed criterion gives the result that to be sweet is to taste sweet to the man to whom those things taste sweet which are such that they taste sweet to the man to whom ...; that is to say, the criterion cannot even be stated. Plainly, we identify the healthy man via our general theory of health and human functioning; the healthy man is he whose organs are functioning as they should do and whose general condition is up to standard, and that is something which can be more or less objectively established, independently of the functioning of the particular sense for whose object the healthy man’s perception provides the criterion. The satisfaction of the second condition allows the possibility of an atypical, but perfectly healthy man whose perception is out of line with that of the majority; there is no incoherence in the suggestion that 5% of healthy tasters might find sugar completely tasteless (such instances in fact occur in the case of certain substances). Yet the acknowledgement of that possibility, which is required by the perceptual analogy, is inconsistent with Aristotle’s application of this criterion to the case of pleasure; he could not admit that

the criterion of real pleasures 101 there might be atypical good men who just happened to enjoy depraved or brutish pleasures. The refusal to admit this possibility is no mere squeamishness on Aristotle’s part; rather it is central to his theory of human virtue. For virtues and vices are states of character, and the best indication of states of character is the pleasure we take in some things, the unpleasantness we find in others; so the man who abstains from bodily pleasures gladly (chair¯on) is temperate, but the man who does so reluctantly is licentious, the man who faces dangers gladly or at least not unwillingly (m¯e lupoumenos) is courageous, the man who does so unwillingly is a coward (1104b3–8). States of character involve both actions and affective states (path¯e ), and hence perfect character requires that the two be harmonized under the direction of the intellect to the pursuit of the good. There could not, therefore, be an ‘atypical good man’ who just happened to like depraved pleasures; for such a man’s desires and consequent pleasures and pains would be out of phase with his judgement of what was best, irrespective of whether he actually indulged his ‘atypical’ tastes. If he did indulge them he would be enjoying bad actions, while if he abstained from indulging them he would be doing good actions (i.e. abstaining from bad pleasures) contrary to his inclinations; in neither event could he be a good man. Aristotle states this emphatically at 1099a17–20: ‘someone who does not enjoy (or rejoice in (chair¯on)) fine actions is not a good man; for no one would call just the man who does not enjoy acting justly, or generous the man who does not enjoy generous actions, and similarly in the other cases’. Aristotle’s criterion thus fails to satisfy the first of our two conditions (p. 100), since it proves impossible to identify the physically and psychologically sound man independently of the fact that he enjoys those very pleasures of whose reality his enjoyment is the criterion. (Since the satisfaction of the second presupposes the satisfaction of the first, Aristotle’s criterion in fact satisfies neither condition.) It might seem that Aristotle could overcome this difficulty fairly easily, in that he could provide an independent identification of the sound man without dislocation of his system. Why not, for instance, identify the sound man as the man who does good actions, and then identify as real pleasures the things that he enjoys? But, as we have just seen, that would distort Aristotle’s thought by locating goodness of character wholly in the sphere of action, in contrast to Aristotle’s own conception of goodness as consisting in the harmonious

102

the criterion of real pleasures

integration of intellect, action, and affective states. Further, even if that point were set aside, it does not seem that good actions can themselves be identified independently of the identification of the good man. For, as is well known, good actions are those which manifest a state of character which is ‘in a mean’, e.g. self-controlled actions manifest a state of character which is neither too given to indulgence in bodily pleasures nor insufficiently appreciative of them. But what counts as being too given to pleasures, or insufficiently appreciative of them, is said to be determined by the judgement of the man of practical wisdom (phronimos; EN 1106b36–1107a2). Hence the identification of good actions presupposes the prior identification of the phronimos, from which it follows that our proposed modification is impossible within the Aristotelian system. It is worth adding that, at least in the Nicomachean Ethics, Aristotle does not attempt the independent identification of the phronimos which his system requires. The same problem recurs in Aristotle’s use of the contrast between things which are pleasant by nature (phusei h¯edea) and things which are only incidentally pleasant (kata sumbeb¯ekos h¯edea), e.g. things which are pleasant to someone who happens to be in a certain abnormal condition. Aristotle uses this contrast at EN 1099a7 ff. as part of his argument for the conclusion that the virtuous life is in itself pleasant (v. supr., pp. 99–100). Most people’s pleasures, he says (a11–15), conflict because they are not of that sort (i.e. pleasant) by nature, but the lover of noble things finds pleasant things which are pleasant by nature; and virtuous actions are of that kind (i.e. pleasant by nature), so that they are both pleasant to those people and pleasant in themselves. This argument fairly clearly presupposes that what is pleasant by nature is an objective matter, and in particular that it is independent of the judgement or preference of the lover of noble things. It is hard to read it otherwise than as recommending the pleasures of the noble life on the ground that the noble man’s preference reflects what is really preferable; it is noteworthy that a13 reads ‘the things which are pleasant by nature (subject) are pleasant to the lover of noble things (predicate)’ and not ‘the things which are pleasant to the lover of noble things (subject) are pleasant by nature (predicate)’. Yet it is the latter reading which would be required if Aristotle were here making the point that things pleasant by nature are constituted as such by the preference of the noble man.

the criterion of real pleasures 103 The argument of this passage, then, requires that there be a state of affairs in virtue of which certain things are pleasant by nature independently of the preference of those who enjoy such things. That immediately threatens the status of the judgement of the good man as a criterion of what is really pleasant. To refer back once again to our perceptual analogy, if sugar is sweet by nature, independently of how it tastes to anyone, then it needs to be explained just how the perception of the normal taster can be relevant to determining whether it really is sweet. Perhaps we might suggest that, while there is some objective fact of the matter in virtue of which sugar actually is sweet by nature, the only way we have of telling that it is is that it tastes sweet to the normal observer. This distinction would be defensible only if it were possible to give some account of the supposed fact of the matter. In fact it is doubtful whether Aristotle even attempts to satisfy that requirement. The concluding paragraphs of the discussion of pleasure in EN VII (1154b9 ff.) might perhaps be interpreted in this sense. Here he is discussing the question why bodily pleasures are commonly held to be the most desirable form of pleasure, and in the course of his discussion draws the familiar distinction between on the one hand pleasures which are experienced when one gets rid of a painful state of bodily deprivation, and on the other pleasures which involve no distress. The latter, he says, belong to the class of things pleasant by nature and not incidentally. Restorative pleasures such as those of convalescence are pleasant incidentally, in that what is strictly speaking enjoyed is the activity of some part of the organism which remains in a healthy condition, and which happens to lead to restoration of the proper state and hence health. The pleasure of satisfying one’s thirst is then, strictly speaking, pleasure in the exercise of the capacity of assimilation of liquid. The fact that the act of assimilation restores the proper balance of bodily elements and so eliminates the distress of thirst is an incidental feature of the situation, just as the fact that a statue is made by a man with knowledge of music is an incidental feature of the situation which is properly to be described as the making of a statue by a sculptor (who is, incidentally, a musician). By contrast (b20) ‘things pleasant by nature are those which bring about the activity of such and such a nature’. Thus the eating of grass in good condition brings about the activities, such as cudchewing and digestion, proper to the nature of the cow, and is therefore pleasant to the cow. Here, it might seem, we have a preference-free test

104

the criterion of real pleasures

for things which are by nature pleasant to man; things which bring about the activities proper to human nature are the things pleasant by nature to man. But the search for things which bring about the activities proper to human nature takes us back to the preference of the phronimos; for the activities proper to human nature are those which constitute the ergon, the function or characteristic activity, of man, i.e. in the first place the intellectual excellences and in the second place the excellences of character. Now we have already seen (pp. 101–2) first that in the Nicomachean Ethics Aristotle defines excellence of character in terms of the dispositions of the phronimos and secondly that excellence consists not merely in doing what the phronimos approves but in doing it in the right spirit, i.e. gladly or with enjoyment. Hence in the practical sphere the description ‘things which bring about the activities proper to human nature’ is coextensive with ‘things which bring about activities which the phronimos enjoys’. It is clear that we have no way of telling which things are by nature enjoyable other than by observing what the phronimos enjoys; what is less clear is whether, as I have suggested, Aristotle may have thought that the enjoyments of the phronimos, rather than constituting certain activities as pleasant by nature, instead give us access to the objective fact that those activities are pleasant by nature. In this connection a crucial, though ultimately indecisive text is EN III. 1113a25 ff., where Aristotle is discussing the question whether the object of wish is what is in fact good, or what each person considers good. Examination of this text (for which time is lacking) indicates that he is concerned to emphasize first that the judgement of the good man is true and secondly that that judgement is our only means of telling what is the case with regard to pleasantness etc. The further question ‘What makes the good man’s judgement true?’ is one to which Aristotle’s text seems to provide no definitive answer; a possible explanation of this silence is that the question itself had not occurred to Aristotle. We see, then, that Aristotle’s attempt to discriminate ‘real’ pleasures (or, equivalently, ‘things pleasant by nature’) from ‘pleasures’ which are so called only with some qualification (or, equivalently, ‘things incidentally pleasant’), though central to his ethical enterprise, leaves two major questions unanswered. These are first the metaphysical question ‘Assuming it to be a fact that certain things are unqualifiedly pleasant, what is the nature of that fact?’ and second the epistemological question ‘How can we discover

the criterion of real pleasures 105 which things are unqualifiedly pleasant?’. The former question is literally unanswered; indeed, as we have seen, it is not even explicitly posed. To the latter Aristotle does supply an answer, viz. ‘By observing what the phronimos enjoys’, but the question yet remains unanswered in the sense that that answer is unsatisfactory, since in the Nicomachean Ethics Aristotle provides no directions on how to identify the phronimos. It is worth remarking in conclusion that he appears to remedy that deficiency in the Eudemian Ethics, since in the last chapter of that work (1249a22 ff.) he does what he nowhere does in the Nicomachean, viz. specify the standard or rule (horos) by which the good man (spoudaios [= phronimos]) determines what is right in action and in the choice of such external goods as wealth and friendship. The standard is ‘whatever choice and use of things good by nature, whether goods of the body or possessions or friends or other good things, will best promote the contemplation of God, that is best and that the finest standard; but if any choice and use impedes the service and contemplation of God, either through deficiency or excess, that is bad’ (1249b17–21). This passage provides the missing identification of the phronimos; good actions are those which best promote the¯oria, and the phronimos is he who does and enjoys good actions. The same point might be put in terms of the ergon of man; it is appropriate to man above all to contemplate reality, and the man in perfect condition is he whose intellect and desires are so harmonized as best to promote the achievement of that end. Yet this identification of the phronimos will be both useless and otiose as a means of identifying real pleasures. It will be useless because of the requirement that the phronimos enjoys good actions; that requirement makes it impossible to find out that e.g. a certain amount of bodily pleasure is really pleasant by observing that that is what the phronimos enjoys. For, given the knowledge that that amount of pleasure best promotes the¯oria, anyone who doesn’t enjoy it is not properly phronimos, but is at best enkrat¯es. And it is precisely that knowledge which makes the proposed criterion of reality for pleasures otiose. For if we know that a certain pleasure best promotes the¯oria, we know that it is best suited to the proper functioning of human nature, i.e. that it is something pleasant by nature, as opposed to something which is pleasant only when nature is in one way or another disorganized. Given independent knowledge of the processes of nature, we have no need to rely on anyone’s judgement as a test of what is naturally, i.e. really or unqualifiedly, pleasant. It is, therefore, probably not accidental that,

106

the criterion of real pleasures

in contrast with the Nicomachean, the Eudemian Ethics are comparatively reticent on the role of the good man as a criterion of real pleasures: he is nowhere described as a standard or measure of what is really pleasant, and only one passage (1235b31 ff.) implies that role. On the other hand, a number of passages (1237a1–10, a25; 1238a27; 1239b30 ff.) indicate that the unqualifiedly good and the unqualifiedly pleasant are determined by the proper functioning of human nature. It is a further question whether this difference of emphasis between the Eudemian and the Nicomachean Ethics is better explained on the hypothesis of the priority of the former or of the latter. Afterword For fuller discussion of EN III. 4, including 1113a25 ff., see Taylor [2006], 159–63. The above interpretation of the final chapter of the Eudemian Ethics is controversial. For an opposed view, which interprets the service and contemplation of God not exclusively as the exercise of the¯oria but as the exercise of both intellectual virtue and virtue of character, see Kenny [1992], ch. 7. I dispute that interpretation in Taylor [1994].

7 Urmson on Aristotle on Pleasure While many of J. O. Urmson’s writings have attained the status of classics in their various fields, his paper ‘Aristotle on Pleasure’ has not attracted more than passing mention from other writers.¹ Yet that essay, brief though it is, raises an issue of major importance for our understanding not only of Aristotle’s views on pleasure but of pleasure itself. That issue concerns the relation between the enjoyment of an activity and the having of pleasant sensations. Contrary to the empiricist tradition, which had tended to equate the two, Urmson both uses Aristotle to urge the distinction between enjoying φ-ing and enjoying sensations caused by φ-ing and finds fault with him for blurring that distinction. While taking issue with Urmson on both points, I count it a great merit of his paper that it focuses attention on a central and unexpectedly complex issue. In that paper Urmson first summarizes Aristotle’s discussion of pleasure in Nicomachean Ethics (hereafter EN) X, which he describes as ‘Aristotle’s most mature and careful account,’ and then discusses some problems arising from it. His main conclusion, with which I broadly agree, is that in that discussion Aristotle offers an analysis of the enjoyment of an activity as something barely distinguishable from the activity, ‘more like the effortless zest with which the activity is performed than a result or concomitant of it’.² This view of enjoyment, which is similar to Gilbert Ryle’s, is opposed to any attempt to represent enjoyment as the obtaining of pleasant experiences accruing from an activity, of which the paradigm case is a pleasant bodily sensation caused by the activity. Urmson quotes a revealing example of the I am grateful to Jonathan Dancy, Justin Gosling, and, especially, Jennifer Hornsby for their comments on earlier drafts. ¹ Urmson [1967]. For comments on Urmson’s paper see Owen [1971–2], Lucash [1974], Hardie [1980], 411, and Gosling and Taylor [1982], 273–6, 297–8. ² Urmson [1967], 326.

108

urmson on aristotle on pleasure

kind of misunderstanding to which the latter picture of enjoyment can lead: at the beginning of his discussion of temperance in EN III. 10–12, Aristotle distinguishes the pleasures of the soul, for example, those springing from the love of honour or the love of learning, from bodily pleasures, pointing out that someone who experiences the former type ‘enjoys what he loves, his body being in no way affected, but rather his intellect’ (1117b29–31). Enjoying intellectual activity, that is, or enjoying acting honourably, is not a matter of having any bodily sensations. Aristotle’s commentator Aspasius, however, failed to see the point, protesting that pleasure and enjoyment are not in the intelligence but in the affective part of the soul.³ It is natural to think in that way, Urmson suggests, if you take it for granted that pleasure and enjoyment are passive experiences connected only causally with any activity. Aristotle, then, is in no danger of construing all pleasure as the occurrence of pleasant sensations. Rather, according to Urmson, he goes to the opposite extreme, attempting to make the experiencing of pleasant sensations itself a case of the enjoyment of an activity. This is seen from his discussion of bodily pleasures in EN III. 10, where bodily pleasures are described as the enjoyment of different kinds of sense perception, enjoyment of pleasant colours, sounds, etc. The bodily pleasures that are the sphere of temperance (Aristotle’s immediate concern in this passage) are those of touch and taste (1118a23–6) and primarily that of touch. The glutton and the drunkard are not interested in tasting what they are eating or drinking, as a wine taster or a cook might be, but in the sensation of the stuff actually sliding down the gullet (a26–31); hence Aristotle’s anecdote of the glutton who wished that he had a throat longer than a crane’s, so as to prolong the sensation of swallowing (literally, ‘since he enjoyed the touch’ (a32–b1)). Similarly, and more plausibly, sexual enjoyment is all a matter of touch (a31–2). Urmson’s account of this passage is as follows. What Aristotle is really doing is describing the intemperate man as one who goes in for excessive eating, drinking, and sex not because he enjoys those activities, but for the sake of the pleasant tactile sensations that they produce; but, lacking a clear grasp of that distinction, Aristotle misdescribes the intemperate man as enjoying the exercise of the sense of touch. This, according to Urmson, is misleading, since enjoying the exercise of the sense of touch ³ Aspasius, Commentaria in Aristotelem Graeca, xix. 88, lines 11–13.

urmson on aristotle on pleasure 109 is quite different from having a pleasant tactile sensation. The latter is a passive experience, while the former is a kind of activity; an example of it would be the enjoyment of the tactile discrimination of different textures. A similar distinction holds in the case of each of the other senses; thus we can distinguish looking at x from seeing x, listening to x from hearing x, tasting (as in wine tasting) from having a taste in the mouth, and smelling (in the sense of sniffing at) from having an odour in the nostrils. In the case of each pair the enjoyment of the first member is a genuine enjoyment of an activity, but in the case of the second what is enjoyed is not an activity but a sensation. If in those cases an activity is involved, it is merely a means of obtaining the object of the enjoyment; it is not that object itself. An example of Urmson’s makes the point clear: ‘We must surely recognize,’ he writes, ‘that one might smell roses not because one enjoyed smelling roses but because one enjoyed the smell—a passive experience to which the activity of smelling is but a means.’ Aristotle blurs this important distinction, and may therefore justly be censured for ‘the uncommon error of assimilating the enjoyment of feelings to the enjoyment of activity’.⁴ I do not intend in this paper to discuss Aristotle’s analysis of the enjoyment of an activity, or Urmson’s account of it. Nor do I wish to take issue with Urmson’s contention that in his discussion of bodily pleasures in EN III Aristotle assumes that the intemperate man overindulges in food, drink, and sex for the sake of pleasant bodily sensations caused by eating, drinking, and sexual activity. Rather I propose to concentrate on Urmson’s doctrine (implied though nowhere explicitly stated) that when A φ’s for the sake of sensation S, which A enjoys, A’s enjoyment is always to be described as the enjoyment of S, and never as the enjoyment of φ-ing. The essence of Urmson’s criticism of Aristotle, as I understand it, is that Aristotle tries to have it both ways: he describes the intemperate man as enjoying the activity of touching, while what he means, and therefore ought to say, is that the intemperate man enjoys tactile sensations, and therefore that his unseemly activities are not what he enjoys, but merely means to what he enjoys. I hope to show that that distinction does not hold, since the intemperate man’s enjoyments are prominent among a range of cases in which it is true both (1) that A enjoys φ-ing and (2) that one of the features ⁴ Urmson [1967], 329, 331.

110

urmson on aristotle on pleasure

(perhaps the principal feature) in virtue of which A enjoys φ-ing is the occurrence of pleasant bodily sensations caused, in the appropriate way, by A’s φ-ing. I do not, of course, deny that sometimes when A’s φ-ing causes a pleasant sensation S, A does not enjoy φ-ing. Such cases occur when the pleasant sensation is a relatively remote effect of φ-ing, for example, the glow of returning circulation produced by vigorous towelling after an agonizingly cold shower. Here the showering might cause the enjoyable glow, without being itself enjoyed; the glow is sufficiently remote from the shower, both temporally and causally, to be disqualified from being counted as a feature of showering in virtue of which showering is enjoyable. But now let us consider the glow in relation, not to the shower, but to the vigorous towelling, which immediately produces the glow. The towelling is no less distinct from the glow than the shower is; plainly, A’s vigorously rubbing himself with a towel is not identical with the pleasant sensation that it causes. Yet when A rubs himself with vigour and enthusiasm, groaning with pleasure at the exquisite sensations of returning circulation, it seems quite artificial to insist that he is not enjoying towelling himself, but instead enjoying the glow. Here, while the towelling undeniably causes the glow, which is the focus of enjoyment, the temporal coincidence and the immediacy of the causal link seem to constitute the towelling as something that is itself enjoyed, in virtue of its causing the glow, rather than as a mere means to the real object of enjoyment, namely, the glow. That is, sometimes the occurrence of sensation S is, rather than a mere effect of φ-ing, a feature of φ-ing in virtue of which φ-ing is enjoyed; the pleasures of taste and touch, including sexual pleasure, provide the clearest examples of this type of situation.⁵ This will become clear from more detailed consideration of an example drawn from this range of pleasures. Suppose that I am a regular glutton for chocolates. I eat chocolates whenever I have the chance, smack my lips over them, can’t keep my mind on anything else when there is a box in the vicinity, and so on. All this ⁵ The principle on which I rely, that in some cases one enjoys φ-ing by enjoying an immediate effect of φ-ing, has analogies elsewhere, e.g. one hears a car approaching by hearing the sound caused by a car approaching. The insistence that one does not enjoy φ-ing but only the effect of φ-ing is parallel to Berkeley’s insistence that, strictly speaking, one does not hear the car, but only the sound (First Dialogue, in A. A. Luce and T. E. Jessop, eds., The Works of George Berkeley [London, 1948–57], ii. 204).

urmson on aristotle on pleasure 111 adds up to the characteristic Aristotelian-Rylean picture of someone who enjoys eating chocolates. Now you ask me why I enjoy eating chocolates, or, to put the same question another way, what it is about eating chocolates that I enjoy. I reply ‘I like the taste of chocolates.’ Having read your Urmson you then say ‘I see. You don’t really enjoy eating chocolates. What you enjoy is the taste of chocolates, which is a sensation. The activity of eating chocolates is merely a means toward that sensation. So you have misdescribed what it is that you enjoy.’ Somewhat baffled, I ask you the following question: ‘You say that I don’t really enjoy eating chocolates, since what I really enjoy is the taste of chocolates. What then would you count as a case of really enjoying eating chocolates, as distinct from enjoying the taste of chocolates?’ Taking a hint from Urmson’s wine taster, you reply ‘Well, for instance, the enjoyment experienced by a connoisseur of flavours, who enjoys identifying different chocolates by their taste, texture, etc., as opposed to someone like yourself who just wolfs down every chocolate in the box regardless of the differences between them.’ My rejoinder is this: ‘What you have done is to describe a chocolate fancier, that is, someone with a specialized interest in chocolates. I might agree at a pinch to describe him as enjoying eating chocolates, though it would be more precise to describe him as enjoying tasting or discriminating them. But there is absolutely no reason, and indeed it would be most misleading to confine the description ‘‘enjoys eating chocolates’’ to him and to refuse to apply it to me. We both enjoy eating chocolates, but for different reasons. I enjoy eating them for the taste, that is (suppose I am an extreme case), all chocolates taste the same to me, and it is a taste that I like. He enjoys eating them for the taste too, but in a different way, in that he enjoys the exercise of his powers of discrimination in respect of taste. It is as absurd to suggest that no one but a chocolate fancier really enjoys eating chocolates as it would be to suggest that no one but a beer connoisseur really enjoys drinking beer.’ The defender of what I shall call the Urmsonian position⁶ need not, however, regard this as a knockdown argument. While he might agree that, at the level of ordinary language, it seems very odd to refuse to describe me as enjoying eating chocolates, nevertheless, he might maintain, ⁶ This abstract person is not to be identified with Urmson, who does not use the argument discussed in the paragraphs that follow.

112

urmson on aristotle on pleasure

there is a logically compelling reason for this refusal. This is that there is merely a contingent connection between the activity of eating chocolates and the experiencing of the taste of chocolates. Thus, if we suppose that a drug could reproduce the taste of chocolates, though no chocolates were in fact eaten, one would get precisely the same enjoyment from taking the drug as one now gets from eating the chocolates. But, at least since Ryle, it has been commonplace to say that the enjoyment of an activity is unobtainable without the activity itself; one cannot get the enjoyment of gardening without actually gardening. So since, as our drug example shows, it is possible to enjoy the taste of chocolates without actually eating any, but logically impossible to have the enjoyment of eating without actually eating, enjoying eating chocolates must be a different enjoyment from enjoying the taste of chocolates, even in the case in which one enjoys eating them because of the taste. In order to show why this argument will not work, we must distinguish two types of case in which one enjoys φ-ing by enjoying sensations caused by one’s φ-ing: in the first type the occurrence of those sensations is non-contingently linked to the belief that one is φ-ing, whereas in the second those sensations can occur in the absence of the belief that one is φ-ing. Sexual enjoyment of mutual caressing is a particularly clear case of the first type. The pleasant sensations of caressing and being caressed are not ‘raw’ sensations occurring in total independence of the context of beliefs in which they are felt. Being caressed feels pleasant in part because of the beliefs the person being caressed has about the person caressing, the nature of the situation, etc.; thus if the caressing were believed to be a prelude to torture it is unlikely that it would feel as pleasant as a ‘normal’ caress, even supposing the physical stimuli to be identical. But this does not show that this kind of sexual enjoyment does not consist in part in the enjoyment of certain sensations; rather it shows that the holding of certain beliefs is a necessary condition for the occurrence of certain sensations. In order to have a soothing sensation, for example, it is necessary to be in a frame of mind conducive to being soothed, which requires that one should have certain beliefs and lack others. It follows that the only way in which sensations of that sort could be reproduced in the absence of the activity giving rise to them would be by the production of a total hallucination of the enjoyment of that activity, including the belief that that activity was being undertaken. But every instance of enjoyment,

urmson on aristotle on pleasure 113 of whatever kind, could (theoretically) be thus reproduced. Hence if the bare logical possibility of the production of a total hallucination of the enjoyment of φ-ing were sufficient to show that what is actually enjoyed is not φ-ing, no one ever enjoys anything. Since that is absurd, the defender of the Urmsonian position must concede that his thesis does not apply to cases of the first type. He must, then, rely on cases of the second type, in which, as in the example of the drug-induced simulation of the taste of chocolates, the pleasant sensation could occur without even the belief that one was φ-ing. His argument is that, since ex hypothesi what one enjoys is precisely the same when one has taken the drug as when one is actually φ-ing, and since what one enjoys is not φ-ing when one has taken the drug, for then one is not φ-ing, what one is enjoying cannot be φ-ing in the case in which one actually is φ-ing. But this argument depends on a crucial ambiguity in the reference of ‘what one enjoys’. It is the same feature that is enjoyed on either occasion, in that either situation is enjoyed because of the occurrence of F. But in the one case F is a feature of φ-ing, in which case one enjoys φ-ing because of F, while in the other F is a feature of a simulacrum of φ-ing, in which case one enjoys that simulacrum because of F.⁷ So what one enjoys is not the same activity in either situation. Hence the argument is invalid, since the reference of ‘what one enjoys’ is the feature F in the first premiss, but the simulacrum of the activity in the second. The principle on which this defence relies, namely, that if one enjoys a situation because of the presence of F, one cannot really or strictly be said to enjoy anything other than F, is mistaken. One might as well argue that, if one admires a woman’s looks because of her lovely hair, it is strictly speaking the hair that one admires, and not her looks. That is plainly false, even though the hair might be present, and admired, in the absence of the woman, when attached to a waxwork model of the woman, for example. I am not, of course, maintaining that one can enjoy φ-ing without actually φ-ing, since I accept the Rylean position on that point. My claim is rather that if one enjoys φ-ing because of some feature of φ-ing, which we may label F, the fact that in certain circumstances one can get F without actually φ-ing does not show that, when one is actually φ-ing, one is not ⁷ Compare one version of the traditional argument from illusion. What one sees cannot be, e.g. a cat, because one could see that very thing, i.e. that irregular black shape, if no cat were present, e.g. if one were looking, not at a cat, but at a stuffed replica. The fallacy is the same.

114

urmson on aristotle on pleasure

really enjoying φ-ing, but merely enjoying the occurrence of F. Someone who denies this claim is committed to the view that if I can get F either by φ-ing or by ψ-ing, I do not really enjoy either activity, but merely enjoy the occurrence of F. So if I enjoy both skiing and tobogganing because they are exhilarating, I do not really enjoy either, but just enjoy being exhilarated (whatever that means). Of course, it is true that I do not enjoy skiing just qua skiing, but qua exhilarating, and similarly for tobogganing. That is to say, there is some description of the feature(s) of the activity in virtue of which it is enjoyable more specific than ‘it is skiing’ or ‘it is tobogganing’. But it is clear that the application of the description ‘enjoys φ-ing’ is not restricted to cases of enjoying φ-ing qua φ-ing, that is, to cases in which the question ‘What is enjoyable about φ-ing?’ elicits no answer other than ‘It is φ-ing.’ It is fairly clear what has gone wrong with the Urmsonian position. One begins by stressing, rightly, the importance of the distinction between the enjoyment of an activity and the experiencing of pleasant sensations consequent upon that activity. Urmson’s example of doing geometry serves as an excellent illustration.⁸ The enjoyment of geometry, as Aristotle pointed out (EN 1175a30–b6), is a matter of effortless concentration on geometry, which tends to make one a better geometer, whereas pleasurable sensations produced by geometry, for instance, a glow of excitement at the prospect of completing an important proof, may distract from one’s geometrizing in the same way that an ‘extraneous’ pleasure would, for example, hearing one’s favourite tune from next door. One then proceeds to generalize this distinction, drawing the conclusion that if one undertakes an activity for the sake of some sensations(s), what one enjoys is not the activity, but the sensation(s). But this is to overlook the fact that the enjoyment of some activities is itself characterizable in part as the enjoyment of sensations. Thus (assuming that tastes are sensations) if one enjoys eating certain foods because one likes their taste, it is mistaken to suggest that what one enjoys is not eating, but some sensation consequent upon eating. That would correctly describe someone who was indifferent to what the food tasted like, but enjoyed certain digestive sensations occurring afterward, for example; it does not describe someone who relishes food for its taste. Again, if one enjoys sexual caressing because it feels pleasant, ⁸ Urmson [1967], 327–8.

urmson on aristotle on pleasure 115 that is enjoying sexual activity, not enjoying sensations consequent upon it; the same is true, obviously, of enjoying intercourse because it feels pleasant. Urmson’s ‘quasi-aesthetic enjoyment of textures’⁹ is not the only kind of tactile enjoyment; there is also the active enjoyment of touching, stroking, etc., and the passive enjoyment of being touched, stroked, etc., whether in a sexual or a non-sexual context. In the former kind of enjoyment the occurrence of tactile sensations is an essential ingredient; necessarily, if one enjoys stroking the cat, it feels nice to stroke the cat. The latter, that is, passive tactile enjoyment, seems to be the enjoyment of (an activity of) having sensations aroused in one.¹⁰ One can apply Urmson’s own test to show that enjoyment of that activity is not to be separated from enjoyment of the sensation(s): it is surely absurd to suppose that enjoyment of the sensation of being stroked might distract one from being stroked, so that one should have to wait until that enjoyment passes off before getting back to concentrating on being stroked once again. Rather, to use Aristotle’s terminology, the enjoyment perfects the activity (1174b23).¹¹ ⁹ Ibid. 329. ¹⁰ Touch is unique among the senses in that it can be exercised both actively and passively. One cannot exercise sight by being seen, hearing by being heard, smell or taste by being smelled or tasted, but one way of exercising the sense of touch is by being touched. Thus if someone strokes the back of my neck, I acquire tactile information, e.g. of the position of the stroker’s hand, by being touched. The explanation seems to be twofold. Firstly, the verb ‘touch’ has two principal uses. According to the first (let us call it ‘A-touching’, i.e. ‘active touching’), touching is a genus of activities of tactile exploration, including probing, stroking, and fondling among its species. This sort of touching requires intentional movement of parts of the body, normally the limbs. That is why it is appropriate to describe the example above as a case of Jane’s touching the back of my neck with her hand, but inappropriate to describe it as a case of my touching her hand with the back of my neck. ‘A-touch’ corresponds to ‘look at’, ‘listen to’, and perhaps ‘sniff ’, rather than to ‘see’, ‘hear’, and ‘smell’. According to the second use (‘P-touching’, i.e. ‘passive touching’), to touch is simply to be in physical contact with; thus in the example the back of my neck P-touches Jane’s hand and vice versa, though I do not A-touch her hand with the back of my neck. Secondly, the primary verb indicating the exercise of the sense of touch, answering to ‘see’, ‘hear’, ‘taste’, and ‘smell’, is not ‘touch’ but ‘feel’. Thus someone anaesthetized might touch something, even intentionally, without feeling it, and hence without exercising the sense of touch. Not all feeling is tactile feeling: e.g. feeling a headache coming on or feeling giddy are not exercises of the sense of touch. Tactile feeling is feeling by P-touching, i.e. gaining information about the shape, size, position, movement, or texture of an object by direct physical contact between that object and one’s body. Since x P-touches y either if x A-touches y or if y A-touches x, one way of exercising the sense of touch is by being A-touched. This does not hold for the other senses; e.g. ‘x is looked at by y’ does not entail ‘x sees y’. Hence there is no suggestion that x normally exercises the sense of sight in being looked at by y. ¹¹ We do not distort Aristotle’s thought by speaking of the activity of having sensations aroused in one. ‘Activity’ renders energeia, which is strictly the actualization of a capacity or faculty; actually having a bodily sensation is the actualization of the capacity to have sensations of that kind.

116

urmson on aristotle on pleasure

I have here spoken of someone who enjoys sex because it feels pleasant to be caressed, etc., and have argued that given that description it is a mistake to separate enjoyment of sensation from enjoyment of activity. There is, however, another description of a sort of sexual activity, which allows that distinction to be made, namely, that according to which activities such as caressing arouse sensations located in the sex organs, which sensations are alone enjoyed. In that case one could properly be said to enjoy, not the activity, but just the sensation to which it gives rise. The difference between the two situations becomes clear when one reflects that in the second case it is not the stroking itself that feels nice, but the sensation that it arouses, which could be located in a quite different part of the body, whereas in the first case it is the stroking itself that feels nice. Clearly, either situation can occur; for what it is worth, I suspect that the former is the more common, but that is immaterial. What is important is that we should not assume that the second description is the only one available, and hence conclude that anyone who enjoys sex because it feels nice must be enjoying sensations and not activities. To sum up, while it is always possible to distinguish enjoying φ-ing from enjoying sensations consequent on φ-ing, it is not always possible even in theory to distinguish enjoying φ-ing from enjoying the occurrence of some sensations, since φ-ing may be an activity the enjoyment of which consists, in part, in the enjoyment of certain sensations. The enjoyments of food, drink, and sex are characteristic examples of that kind of enjoyment, provided that we are prepared to reckon tastes as sensations. Nor, though a distinction exists, is it in every case clear just where the line is to be drawn between those sensations the enjoyment of which is constitutive of the enjoyment of the activity and those that are merely effects of the activity, of which the enjoyment is consequently separable from the enjoyment of the activity itself. Thus enjoying the taste of brandy is clearly not to be distinguished from enjoying drinking brandy. But what about the heart-warming glow that pervades the whole body after a glass or two? Is the enjoyment of that sensation part of the enjoyment of drinking brandy, or merely the enjoyment of a sensation produced by drinking? There neither is nor need be any clear answer to that question. How, then, does this affect Urmson’s criticism of Aristotle? The latter’s doctrine is that the enjoyment of food, drink, and sex, or at least the kind

urmson on aristotle on pleasure 117 of enjoyment of them that is characteristic of the intemperate man, arises wholly from the sense of touch. (This is a paraphrase of 1118a29–32.) Earlier I accepted Urmson’s suggestion regarding what Aristotle means by this: what the intemperate man enjoys about eating, drinking, and sex is the occurrence of certain bodily sensations. I reject, however, the implication that Urmson draws from this, namely, that Aristotle therefore represents the intemperate man as not enjoying those activities themselves. As I have argued, it does not follow from the fact that you like sex for the sake of certain bodily sensations that you do not enjoy sex as such. But it may be true nonetheless that Aristotle in fact represents the intemperate man as not enjoying his intemperate activities themselves. We have to look at what he says. In the case of food and drink the intemperate man gets little or no enjoyment, in Aristotle’s view, from the taste of what is eaten or drunk (nor does he say anything about enjoyment of the alcoholic effects of drink); the enjoyment lies wholly in the sensation of swallowing. Now, that sensation seems to me one of those that, like the warm afterglow of the brandy, are so intimately connected with the activity giving rise to them as not to be clearly separable from it. We can, if we choose, stipulate that someone who likes eating solely because he enjoys the sensation of swallowing is not to be said to enjoy eating, but that would be a stipulation, not an implication of the existing concept. It is worth remarking on the odd reason that Aristotle gives for his assertion that the intemperate man gets little or no enjoyment from taste. The reason is that ‘to taste belongs the discernment of flavours, the kind of thing that wine tasters and cooks do. But they do not enjoy these things at all, or at least intemperate people do not; their enjoyment arises wholly from touch’ (1118a27–31). What Aristotle has done here is to single out one specialized exercise of the sense of taste (one, moreover, that its practitioners do not, in his view, enjoy at all) and treat it as if it were the only exercise of that sense. He appears to think that if you do not enjoy tastes in that way, you cannot enjoy tastes at all. But that is simply false. Aristotle’s view is even more extreme than the one I criticized earlier, according to which only a chocolate fancier really enjoys eating chocolates, which is a different enjoyment from enjoying the taste of chocolates. On Aristotle’s view, only the chocolate fancier’s enjoyment involves taste at all; the person who is greedy over chocolates, but is not a chocolate fancier, enjoys something altogether different, namely, the

118

urmson on aristotle on pleasure

sensation of swallowing the chocolates. It may be sufficient to remark that on that view it seems impossible to account for the fact that people are very often greedy over particular kinds of food and drink while indifferent to others, since presumably it all feels much the same going down.¹² Aristotle’s view perhaps depends on the assumption that the primary function of a sense is discrimination between objects of that sense, for example, that the function of sight is discrimination of colour.¹³ But that assumption does not lead to Aristotle’s conclusion; for someone who is greedy over oysters, but indifferent to rice pudding, is at least discriminating oysters from rice pudding in enjoying the taste of oysters, though it may be no part of his enjoyment to make discriminations between the tastes of different kinds of oysters, as a connoisseur might. We come now to Aristotle’s account of the remaining ‘intemperate’ pleasure, namely, intemperate sexual enjoyment. In saying that that is purely a matter of touch, Aristotle is clearly taking an unduly narrow view of sexual activity, ignoring such psychological factors as the pleasures of intimacy and tenderness, of dominating or being dominated, etc. But he cannot be interpreted on that ground alone as saying that the intemperate do not enjoy sexual activity, but only sexual sensations, since we have seen that in many sexual situations the enjoyment of the former consists, in part, in the enjoyment of the latter. There is, however, one further piece of evidence to be taken into account. After characterizing ‘intemperate’ enjoyment of food and drink as purely a matter of touch, Aristotle distinguishes intemperate pleasure from ‘more refined’ tactile enjoyments, giving as an instance of the latter the enjoyment of a warming massage in the gymnasium. The ground of the distinction is that ‘the touch that the intemperate man enjoys does not involve the whole body, but certain parts only’ (1118b4–8). (The ‘certain parts’ are presumably the gullet, for food and drink, and the sex organs.) This might suggest that Aristotle accepts the second of the two pictures of sexual activity sketched above, according to which sexual activities such as stroking and kissing are valued ¹² Aristotle assigns to the process of swallowing the discrimination of certain properties of food—e.g. it is in swallowing, rather than tasting, that one perceives that what one is eating is warm or oily—and uses this to explain the fact that the same people are not greedy over food as are greedy over drink, which is discriminated by taste (PA 690b29–691a4). This manifestly fails to explain discriminating greed over food; the only plausible explanation of the coexistence of lust for caviare with detestation of sago is the difference in taste. ¹³ e.g. De an. 424a3–10, 432a16.

urmson on aristotle on pleasure 119 only as means to the production of sensations located specifically in the sex organs. That may be what he meant, in which case he does indeed represent the intemperate as enjoying, not those activities themselves, but merely sexual sensations produced by them. That distinction cannot, however, apply to the enjoyment of sexual intercourse itself; the latter is an activity the enjoyment of which consists in part in the enjoyment of sensations occurring in the sex organs. In any case, Aristotle’s words can be taken in another sense, which is perhaps somewhat more plausible. This is that what distinguishes sexual activity from other forms of activity, and consequently sexual pleasure from other forms of pleasure, is the specific parts of the body that it involves. Then to say that sexual enjoyment is all a matter of touch will not be to say that it is enjoyment of sensations and not activities, but merely to say that one enjoys the activities of touching and being touched not with respect to the whole body, but with respect to certain parts only. This, too, is pretty plainly false, especially given the restricted interpretation of ‘certain parts’ assumed above; but the falsehood reflects inadequate observation (or possibly differences between ancient and modern sexual practices) rather than any conceptual error. My conclusion, then, is that while Aristotle believes that the intemperate man goes in for his intemperate activities for the sake of bodily sensations, he does not say or imply that he enjoys only those sensations and not the activities themselves. Intemperate pleasures are important examples of a sort of enjoyment of activity that is itself constituted in part by the enjoyment of sensations. Hence in those cases it is a mistake to infer that someone who enjoys that sort of sensation does not enjoy the activities themselves. But that mistake, if committed by anyone, has been committed by Urmson, not by Aristotle. Where Aristotle goes wrong is in his very queer view of what it is that is enjoyable about ‘intemperate’ activities, and in particular about the intemperate enjoyment of food and drink. He is also wrong, in my view, about sex, though less obviously and grotesquely so. Specifically, he is wrong to classify all these forms of enjoyment as consisting purely in the enjoyment of tactile sensations. But, contra Urmson, he is right to think that the enjoyment of tactile sensations fits happily into his general account of pleasures as the enjoyment of activities, under the heading of enjoyment of the exercise of the senses. For the senses can be employed, and enjoyed, in the having of sensations as much as in active exploration

120

urmson on aristotle on pleasure

and discrimination, and the passive enjoyment of tactile sensations is the paradigm instance of the former kind of enjoyment. Afterword p.108. Aspasius’ commentary is translated by D. Konstan, Aspasius on Aristotle Nicomachean Ethics 1–4, 7–8, London, 2006. The volume belongs to the Ancient Commentators on Aristotle series, under the general editorship of R. Sorabji. The passage cited occurs on p. 89 of the translation.

8 Popular Morality and Unpopular Philosophy It is an especial pleasure for me to contribute to a volume in honour of Kenneth Dover, not merely as an expression of gratitude for the good fortune of having spent ten happy years at Corpus under his benign presidency, but also in recognition of that special combination of candour and rigour which characterizes his work, and which may serve as a paradigm to scholars in his discipline and beyond.

The reader of K. J. Dover’s magisterial Greek Popular Morality in the Time of Plato and Aristotle¹ (hereafter GPM) cannot but be struck by the fact that, while the names of Plato and Aristotle appear in the title as an indication of the temporal scope of the work, reference to their writings in the book itself is comparatively sparse. This comparative neglect stems directly from his conception of the project. In his preface (pp. xi–xii) he distinguishes the subject of the work, viz. ‘the data relevant to popular morality’, from ‘systematic ethical thought’ (i.e. the ethical theories of Plato and Aristotle), and points out that he has not attempted (in GPM) to investigate the relation between the two. The existence of the distinction between popular morality and ethical theory I take as self-evident, and I have no quarrel with Dover’s decision not to undertake the complex task of investigating the relations between the two in the Greece of the fourth century bc; he rightly sees GPM as devoted to the prior task of providing part of the data for such an investigation. But the writings of Plato and Aristotle may themselves be expected to provide, in addition to their authors’ theoretical views, some evidence of current moral attitudes. Any author writing on morality must take some interest in, and therefore provide ¹ Dover [1974].

122

popular morality and unpopular philosophy

evidence for, views current at the time of writing. Moreover, the character of the works of Plato and Aristotle gives special reasons for expecting them to be fruitful sources of such evidence. Plato writes dialogues in which some characters express, not Plato’s own beliefs, but views which Plato presumably believed to be current at the time; Aristotle’s ethical method explicitly counts current moral beliefs as among the φαινόμενα which it is the task of ethical theory to systematize and harmonize (EN 1095a 28–30, EE 1216b 26–35, 1235b13–18). Dover accepts these points (p. 7), and does occasionally use Plato and Aristotle as sources (e.g. pp. 43, 59, 171, 180). But this use is minimal; he is plainly suspicious of the reliability of the philosophers as evidence for current views, as he explicitly acknowledges in the case of statements about what ‘most people’ say and think (p. 7). I doubt the utility of any attempt to discuss a priori the question whether Plato and Aristotle are reliable sources of evidence for the moral views current in their time. (Dover’s own grounds for scepticism are weak; his main claim, viz. that modern philosophers’ assertions about what ‘we’ say, think, and feel are unreliable (ibid.), is both unsupported by any evidence and dubiously relevant to the ancients.) Any view of their reliability or unreliability in general could, in my view, be attained only by the accumulation of particular pieces of evidence, which might enable us finally to judge whether the picture of contemporary morality which emerges from their works is broadly coherent with that presented by such works as GPM. The presence or absence of such broad coherence would, I believe, be the only possible test of their reliability. Needless to say, I shall not undertake such a Herculean task in this essay. I shall restrict myself to a single piece of fieldwork, in which I shall attempt to show that a generalization of Dover’s about Greek moral beliefs can be refuted by evidence provided largely, though not exclusively, by Plato and Aristotle. If successful, this attempt will indicate that, at the least, the student of ancient morality may profitably consult the writings of the philosophers to supplement the invaluable collection of evidence contained in GPM. In the context of a contrast between ancient and modern views on individual freedom and responsibility Dover writes as follows (p. 157): The Greek did not regard himself as having more rights than the laws of the city into which he was born gave him at that time; these rights could be reduced, for the community was sovereign, and no rights were inalienable.

popular morality and unpopular philosophy 123 I shall argue that some well-known passages from Plato and Aristotle, supported by evidence from Sophocles, Thucydides, and elsewhere, provide sufficient ground for the rejection of that assertion. In the passage quoted Dover asserts that (1) ancient Greeks regarded themselves as having certain rights, viz. those ascribed to them by the laws of their native city, and (2) no ancient Greek regarded himself as having any rights other than those specified in (1). One might dispute his assertion by denying (1), on the ground that the ancient Greeks lacked the concept of a right altogether. It is certainly true first that no single Greek expression answers to the English ‘a right’, and secondly that no ancient moral theory treats rights as its basic or even central concept. But these facts give us no ground to deny (1). I shall not offer any general account of rights, but shall accept as sufficient J. S. Mill’s dictum: When we call anything a person’s right, we mean that he has a valid claim on society to protect him in the possession of it, either by the force of law, or by that of education and opinion. (Utilitarianism, ch. 5, Everyman edn. (London, 1957), 49)

Since the ancient Greeks certainly claimed from society the legal protection of such interests as life, property, and civic status, we should therefore accept Dover’s assertion that they regarded themselves as possessed of legal rights. In what follows I shall dispute his assertion (2), arguing that our evidence justifies us in maintaining that certain Greeks claimed certain rights other than those guaranteed by the laws of their native state. In some instances I shall argue that certain Greeks are represented as claiming such rights; if that is accepted we shall have to pursue the further question of what such representations tell us about the actual beliefs of actual Greeks. The first passage which I shall cite is excessively familiar, and much debated. In Plato’s Apology Socrates is represented as claiming that the task of criticizing the moral beliefs of his fellow citizens has been imposed on him by the gods, whose commands must be obeyed, even at the cost of his life (28d–29a). He then imagines a situation in which the jury agrees to discharge him, on condition that he abandons his mission; if he refuses to accept that condition, he is to suffer the death penalty (29c). His imagined reply is ‘Men of Athens, I revere and love you, but I shall obey the god rather than you, and as long as I live and am able, I shall never cease from philosophizing ...’ (29d).

124

popular morality and unpopular philosophy

Is this a counter-example to Dover’s generalization, i.e. a case in which a Greek claims some right not assigned to him by the laws of his native state? We must be cautious here. First, the situation is doubly imaginary; Socrates is represented by Plato as representing a situation which does not obtain in fact. In fact, the activity of philosophizing was not forbidden by the law of Athens, and was therefore one which any citizen had the legal right to undertake, in that minimal sense of ‘a right’ in which one has a legal right to do whatever one is not legally forbidden to do. Secondly, the imagined situation is not one in which, strictly speaking, Socrates’ disobedience would be a breach of law, but merely a breach of a condition imposed by a court. Finally, Socrates’ justification for his imagined disobedience is not that the prohibition on his philosophizing would violate his individual right to philosophize, but that he is under a divinely imposed obligation to philosophize, whose divine authority overrides his obligation to obey the authorities of Athens. Hence, it might be maintained, claims to rights are simply not at issue, and the case is irrelevant to Dover’s assertion. It is the third point which is crucial. The essence of the position ascribed to Socrates in the Apology is that divine authority transcends human authority; hence any human injunction, obedience to which requires the violation of obligations imposed by divine authority, must be disobeyed. Hence the distinction between a law properly so called and a mere court order is irrelevant;² Socrates could not be imagined to say ‘Gentlemen, I shall obey the god rather than you; but of course if the Assembly should ² Pace A. D. Woozley, for whom the distinction between the two is central to his attempt, in Woozley [1979], 40–6, to reconcile the positions attributed to Socrates in Ap. and Crt. Woozley’s position is anticipated by Hirzel [1900], 67 n. 2. Brickhouse and Smith [1989] argue (pp. 149–53) that Socrates need not be construed as claiming a right to philosophize which overrides an imaginary law forbidding that activity. Socrates, they claim, would count such a law, even if promulgated by due legal process, as legally invalid, on the ground that, in requiring him to violate a divine command, it conflicted with the existing law against impiety. The two authors and I are in agreement that Plato represents Socrates as committed to refusing to comply with a law forbidding him to philosophize. The difference between us turns on the grounds on which Socrates would justify that refusal; on my interpretation he would do so because the divine command overrode the legal prohibition (assumed to be legally valid), whereas on the Brickhouse–Smith view he would do so on the ground that the prohibition was legally invalid because in conflict with a prior legal prohibition on impiety. On that particular question Plato’s text is, strictly speaking, indeterminate; he does not spell out the grounds on which Socrates would refuse to comply in the imagined situation. I think, however, that Socrates’ words to the jury ‘I shall obey the god rather than you’ suggest that Plato has in mind that Socrates’ ultimate appeal would be rather to authority transcending the legal system than to grounds internal to that system. He does not, after all, object that the imaginary court order would be legally incompetent, though, as Brickhouse and Smith point out (pp. 143–7), it almost certainly would have been.

popular morality and unpopular philosophy 125 duly enact a law forbidding anyone to philosophize, then I should have to disobey the god.’ (It was in any case the common practice of forensic orators to identify the jury with the people, and hence with the state; see GPM 292.) For the same reason the facts of Athenian law are irrelevant; Socrates appeals to an authority which would override that of the law, if the two were to be in conflict. But where does Socrates claim a right to disobey the law in the imaginary situation? The claim is implicit in the claim that he is under an obligation to do so. What one is obliged (by an authority) to do, one is ipso facto authorized (by that authority) to do, and the stronger claim has to be understood as implying the weaker. So Socrates is represented as implicitly claiming, in the imagined situation, the right to continue to philosophize in violation of the law, in virtue of his stronger, explicit claim to be under the obligation, in that situation, to continue to philosophize. We do not, of course, know whether the historical Socrates, either at his trial or on any other occasion, actually claimed that right, in virtue of that obligation. (It is quite plausible that he did so at the trial; it would have been a striking example of the μεγαληγoρία which was held to have contributed to his condemnation (Xen. Ap. 1).) What Plato’s text shows is that he thought it appropriate to present it as central to his defence of Socrates, as following from the fundamental Socratic theses that one must never under any circumstances act unjustly (Crt. 49b8) and that disobedience to divine commands is an instance of acting unjustly (Ap. 29b6 f.). It was not, then, a view which Plato expected his reader to dismiss out of hand; rather, it was a position which he could expect to be taken seriously, and perhaps even, on reflection, to carry conviction. On the other hand, the belief that divine law, with the obligations and rights which that law creates, overrides the law of the state in a case where the two conflict was not a personal idiosyncrasy of Socrates’. This is clear from the famous scene in Sophocles’ Antigone where Antigone defies Creon’s edict that her brother Polynices shall be left unburied, on the ground that that edict is a violation of divine law. Having admitted (447 f.) that she knew of the edict, she responds to Creon’s challenge ‘And yet you dared to contravene these laws?’ (449) with the words Yes, for it was not Zeus who proclaimed them, nor did Justice, the companion of the gods of the underworld, determine such laws among men. Nor did I think

126

popular morality and unpopular philosophy

that your decrees had such force that you, a mortal, could outrun the unwritten and unshakeable laws of the gods. For they live, not yesterday and now, but for ever, and none knows whence they came. (450–7)

Here, as in the Apology, someone is represented as claiming that obedience to divine ordinances requires disobedience to the law of the state. (The fact that Creon’s edict is promulgated via a proclamation, κήρυγμα, rather than a written law is insignificant; breach of it is consistently described (see 59, 381 f., 481, 663, 847) as breach of νόμoς or νόμoι.) Antigone conceives herself as under an obligation, imposed by divine law, to bury her brother, and therefore (as in the case of Socrates) as having the right to do so. (It is tempting to construe the passage as expressing the belief that the dead have a right to burial, but that temptation should be resisted, for there is no suggestion that the dead have any kind of claim on the living; rather it is the gods who make demands on the living, demands which in their turn ground the claim of entitlement to do what the gods require one to do.) The context is, of course, fictional, and we cannot assume that the quoted lines express the beliefs of the poet himself. Rather, they express an opinion on the relation of divine to human law which is not merely a serious option but which is treated with approval in the drama as a whole. (While Antigone is not presented as an entirely sympathetic figure, the message of the play is that on the crucial issue, the superiority of divine authority to that of the state, she is in the right and Creon in the wrong, and it may be assumed that the audience was expected to agree.) The thesis, then, that the obligation to obey divine commands (and the consequent right to do so) may in certain circumstances override the law of the state is expressed with some approval by Sophocles in the fifth century and enthusiastically endorsed by Plato in the fourth. Plato’s endorsement may reflect an actual claim on the part of the historical Socrates to the right to disobedience in a hypothetical situation, though we are not justified in asserting categorically that it does. Since Sophocles and Plato may be assumed to have expected some measure of acceptance of this thesis on the part of their audience, we have evidence that they believed that some Greeks regarded themselves as having certain rights other than those guaranteed by the laws of the state. And since they had the advantage over us of being members of the society whose attitudes are in question, we

popular morality and unpopular philosophy 127 should accept that that belief of theirs is likely to have been true, at least until we have definitive grounds for holding it false. Strictly speaking, then, we have discovered counter-examples to Dover’s generalization. But it might justly be objected that they contradict the letter rather than the spirit of his assertion. Our examples show that it is likely that some Greeks believed that they had obligations other than those imposed by the law of the state, and it is only the belief in those obligations which justifies the attribution to them of the belief that they had rights other than those assigned to them by law. But what Dover, on the most plausible construal, is concerned to deny is that Greeks ascribed to themselves nonlegal rights not grounded in prior obligations, but possessed in virtue of some intrinsic characteristic of the possessor, i.e. something closer to the traditional conception of Natural Rights or the Rights of Man. I shall now try to show that, even if his assertion is interpreted in that way, the writings of Plato and Aristotle provide some contrary evidence. In Plato’s Gorgias Callicles argues (483c–d) that while aggression and self-aggrandizement on the part of the strong against the weak are conventionally (νόμωι) regarded as wrong (ἄδικoν) and disgraceful (αἰσχρ´oν), ‘I think that nature itself shows this, that it is right (δίκαιoν) that the superior should have more than the inferior and the more capable more than the less capable.’ This is shown, he asserts, not only by the behaviour of animals, but also by that of men in every age and country. Such behaviour is in accordance with the nature of right (κατὰ φύσιν τὴν τoῦ δικαίoυ), and indeed with the law of nature (κατὰ νόμoν γε τὸν τῆς φύσεως ), though not in accordance with the law we lay down (483c–d). So as far as the real nature of what is right, and the law of nature, are concerned, a stronger party who attacks a weaker (e.g. Hitler invading Poland in 1939) does what that nature and that law warrant or entitle him to do. That is to say, in the minimal sense of ‘a right’ (see above) he does what he has a natural right to do. But Callicles goes further; he maintains, not merely that the stronger has a right to attack the weaker in the minimal sense that it is not wrong for him to do so, but also in the stronger sense that the attempt to restrain him is itself a wrong against him.³ This further claim is implied by his highly pejorative ³ In making this distinction I apply one of the distinctions between various senses of ‘a right’ first formulated by Hohfeld [1920] and now accepted as standard. My ‘minimal sense’ corresponds to a Hohfeldian ‘bare liberty’, my ‘stronger sense’ to his ‘claim-right’. For details of Hohfeld’s distinctions see Waldron [1984], 5–7.

128

popular morality and unpopular philosophy

description of conventional moral education at 483e–4846: we take the best and strongest, he says, and tame them like lions, fashioning them (sc. in the way we want), enslaving them with spells and incantations to the effect that one must be fair, and that that is fine and good. But a man of sufficient spirit will break free from this slavery, trample all that mumbo-jumbo and those unnatural laws underfoot, and emerge as master instead of slave; ‘and there’, he concludes, ‘what is naturally right flashes forth’ (484a6–b1). It is a clear implication of this description that in being induced to accept conventional moral norms the superior individual is being wronged; he is being tricked, his nature is being distorted by irrationality and force, and in breaking free from this slavery he is attaining a freedom to which he is entitled. Callicles, then, believes in at least one natural right, the natural right of the strong to tyrannize the weak.⁴ They have that right (1) in the minimal sense that it is not (naturally, i.e. really) wrong for them to do so, and (2) in the stronger sense that it is naturally (i.e. really) wrong to try to restrain them from doing so by means of conventional morality. In this portrayal of Callicles Plato therefore represents a counter-instance to Dover’s generalization, on the restricted interpretation adopted above. But what does this representation tell us about the views of actual Greeks? I assume that the vexed question of whether Callicles was a historical person is undecidable; even if he was, the attribution of a thesis to him in a Platonic dialogue is not sufficient evidence that he actually maintained that thesis (cf. the remarks on Socrates above). Further, in this case we cannot make the assumption for which I argued with reference to the Apology, viz. that Plato, in presenting his thesis with approval, must be assumed to count on similar approval in his readers, since he clearly disapproves of Callicles’ thesis, and does his best to refute it. Rather, the dialogue shows that Plato regarded that thesis as representing a serious challenge to conventional morality, which had to be met by the resources of Socratic dialectic. But ⁴ The ‘Calliclean’ thesis that might is right has frequently been attributed also to Thrasymachus in Rep. 1 (e.g. by Kerferd [1947–8]). I have not used this passage, since I believe that while the position of Thrasymachus has close affinities with that of Callicles, the former is not, taking his position as a whole, concerned to claim that the stronger has any kind of right to exploit the weaker, but rather that his doing so, while admittedly unjust, is sensible, admirable, and productive of a good life for himself (343b–344c, 348b–e). At 341a he comes close to the assertion that the ruler who is correct in his assessment of his own interest is entitled to the obedience of his subjects, but that suggestion drops out of sight in the rest of the discussion. On this complex question the judicious discussion by Annas [1981], ch. 1, is very helpful (esp. pp. 48–50).

popular morality and unpopular philosophy 129 that implies that Plato believed that at least some people actually did accept Callicles’ thesis, or at least were seriously tempted by it. It is hardly plausible that he should have exerted such dramatic power in the presentation, and such dialectical ingenuity in the rebuttal, of a thesis which he conceived of as a mere abstract possibility, which would not have entered the head of any actual individual. Rather, the impression given by the Gorgias is that Callicles’ view was in the air, that it was attractive to a certain kind of talented, self-confident, and unscrupulous person, of whom the literary figure of Callicles is representative. Plato says as much at Leg. 890a: All of these views (sc. that different kinds of things are fine by nature and by convention, that nothing at all is naturally right etc.) are maintained by men who are thought clever by young people, poets and others, who say that the height of justice is to conquer by force. As a result young people fall into impiety, thinking that the gods whom the law requires one to take account of do not exist, and we find mutually hostile groups of people dragging them by these means towards the naturally correct life, which is really a life of domination over others, not of servitude to others according to law.

Confirmation of the existence of such views in the late fifth century comes from a famous passage in Thucydides’ Melian dialogue, in which the Athenians reply to the Melians’ assertion that they are confident of the support of the gods ‘since we are righteously taking a stand against wrongdoers’ (ὅτι ὅσιoι πρὸς oὐ δικαίoυς ἱστάμεθα, V. 104. 4). The Athenians reply (105. 1 f.) that they have just as good ground to rely on divine favour, since neither in their conduct nor in their demands are they departing from received opinion about the gods or received norms of human behaviour. For as far as the gods are concerned we believe, and as far as men are concerned it is apparent, that by a universal natural necessity they exercise mastery over what they have in their power. We neither laid down this law nor were the first to make use of it once it had been laid down, but we found it in existence and make use of it in the knowledge that it will continue for ever ...

Here the Athenians are not setting aside questions of justification, and confining themselves to the neutral, factual claim that this is how everyone necessarily behaves. Rather, in response to the Melians’ claim to have justice, and hence the gods, the traditional guardians of justice (see GPM 255), on their side, they present a counter-claim to justification and its consequent divine favour, viz. the inevitable law of nature that the victor

130

popular morality and unpopular philosophy

should be in control.⁵ Of course, the context is that of international relations, not personal morality. But the Athenians put their case in abstract terms which bring the political instance under a universal law, applicable both to individuals (divine as well as human) and to communities. Again, Thucydides is no more reporting an actual debate than Plato is reporting actual conversations.⁶ Rather, he, like Plato, provides evidence for the currency of such views at the time of writing.⁷ At Rhet. I. 13 Aristotle distinguishes two kinds of law (νόμoς ): (1) that which is specific to each community, subdivided into written and unwritten law; (2) that which is common and according to nature. ‘For there is something which everyone has an inkling of, a common, natural right and wrong, even when there is no community to which the parties belong or no agreement between them’ (1373b6–9). He cites three instances in confirmation of his claim that there is universal recognition of the categories of natural right and wrong: the first is the passage from the Antigone discussed above, the second some lines of Empedocles on the universal prohibition against taking life (31 B 135 DK), and the third a ⁵ Here I venture to disagree with the analysis of the Athenian position given by A. Andrewes (Gomme, Andrewes, and Dover, [1970], 172). In his view the Athenians, consistently with their earlier refusal (V.89) to argue about justice or injustice, do not dispute the Melians’ claim to be in the right, appealing instead to a universal rule that force overrules justice. It is true that both sides say at the start of the debate (80 f.) that they will leave considerations of right and wrong out of account, but the Melians explicitly reintroduce the appeal to right in 104, claiming thereby to deserve divine favour. In reply the Athenians say that they believe they will enjoy that favour no less than the Melians, since what they are doing and demanding is sanctioned by beliefs about the gods and normal human behaviour. The verb translated ‘demand’ is δικαι´oω, whose formation from δίκαιoν carries the connotation ‘claim as a right’; the Athenians do not then dispute the Melians’ claim, which was endorsed by normal Greek sentiment, that the gods care about right and wrong. They accept it, claiming that in virtue of the law of nature, which applies to gods and men alike, right is on their side. (In another passage from Thucydides, IV. 86. 6, Brasidas is represented as referring to ‘the justification of force, which fortune confers’. For further Thucydidean passages in a similar vein see GPM 312.) In their claim that the νόμoς that one controls what one has in one’s power applies to gods as well as men, the Athenians may have in mind such myths as that of the castration of Uranus by Kronos, and the subsequent dethronement of Kronos by Zeus. If I am right, they accept that such violent acts are justified. That view of mythical violence is defended by Callicles at Gorg. 484b–c, where he cites Pindar’s lines ‘Law the king of all, Mortals and immortals, Acts with supreme hand To justify the most violent deeds’, as a statement of the law of nature (as he conceives it: see above). At Euthphr. 5e–6a Euthyphro cites the action of Zeus, the ‘best and justest of gods’, in dethroning his father as justification for his own prosecution of his father. The contrary view of Zeus’ act, viz. that it was a violation of justice, is given by the Unjust Argument at Aristoph. Nub. 904–6. ⁶ See the careful discussion by Andrewes, op. cit. 182–7. ⁷ So does Nub. 1405 ff., where Pheidippides claims that it is δίκαιoν for him to beat his father, on the grounds, among others, that that is the kind of thing that cockerels and other creatures do (1427–9). Making due allowance for comic exaggeration, the joke requires that that is the sort of thing that some people were saying; there is no point in making fun of views which no one holds.

popular morality and unpopular philosophy 131 speech by Alcidamas, a pupil of Gorgias. A scholiast supplies Alcidamas’ words: ‘God has set everyone free; nature has made no one a slave.’ In citing this as illustrative of a belief in natural right and wrong Aristotle understands Alcidamas as claiming that anyone’s being a slave is, in the nature of things, an instance of wrong or injustice (ἄδικoν answers to both terms), which amounts to the recognition of a universal natural right not to be enslaved. In another passage (Pol. 1253b20–3) Aristotle confirms that that claim was made by some of his contemporaries: ‘Others say that the rule of master over slaves is contrary to nature, and that it is by convention (νόμωι) that one man is a slave and another free, but by nature there is no difference. So it is not right (δίκαιoν); for it is a violation of nature.’ Aristotle thus confirms Plato’s testimony that some Greeks believed in principles of natural right and wrong, which were capable of grounding claims to rights or entitlements, even in cases where those rights were not conceded by the positive law of the state. Of course, such principles need not conflict with positive law, nor would they even normally be expected to do so. Ordinarily, they were seen as presupposed in, and as guaranteeing, the moral status of the legal code. This idea goes back at least to Heraclitus, who asserts that ‘All human laws are nourished by the one divine (sc. law)’ (22 B 114 DK), and is expressed e.g. by Pericles in the Funeral Speech (Thuc. II. 37. 3): In the public sphere it is chiefly fear which makes us law-abiding; we obey whoever is in office and the laws, particularly those which are laid down for the redress of injustice and such unwritten laws as bring acknowledged disgrace (sc. on those who break them).

There is therefore a considerable degree of continuity between written and unwritten law, whether the latter was conceived (despite Aristotle’s distinction cited above) as a set of universal moral norms, as in the Antigone passage, Xen. Mem. IV. 4. 19 f., and Ar. Rhet. 1368b7–9, or, as in Pl. Leg. 793a–c and 841b, as ancestral tradition, which was itself not very clearly distinguished from moral universals with respect to content, and certainly was possessed of no less normative force.⁸ Given this continuity, we ⁸ In Isocr. 12. 169 the requirement that the dead must be buried is described as ‘an ancient custom and ancestral law, which all men regularly use, not laid down by human nature, but imposed by divine power’, and a few lines later (170) as ‘the common law of all the Greeks’. The same requirement, counted at Eur. Supp. 19 among the νόμιμα of the gods, and described at 563 as ‘the ancient law of the

132

popular morality and unpopular philosophy

find occasional appeal to unwritten law in forensic speeches, e.g. Lys. 6. 10, where the jury trying the case of Andocides’ alleged profanation of the mysteries is urged ‘to use not only the written laws, but also the unwritten ones, which are expounded by the Eumolpidae, which no one has authority to annul and no one dares contradict, nor does anyone know who promulgated them.’ Such appeals are infrequent, and when they do occur usually stress the continuity of written and unwritten law, taking the form ‘such-and-such conduct is in accordance with (or contrary to) not just written, but also unwritten law’ (the latter also designated by expressions such as ‘the law common to all men’ and ‘that which belongs to nature’); instances are Dem. 18. 275; 23. 61, 70, 85; 43. 22; 45. 53; Isa. 2. 24 (for further instances see GPM 268). I have not come across any passage in which an orator assumes the existence of a conflict between the two, saying e.g. ‘According to the written law my opponent is indeed in the right, but according to the unwritten, which has superior force, right is on my side.’ In Rhet. I. 15 Aristotle advises an advocate who has the written law against him to argue in just that way, adducing such considerations as the contrast between on the one hand the permanent principles of equity and the universal norms which constitute the unwritten law and on the other the variability of the written (1375a27–b15). So this type of argument was part of the advocate’s equipment, which it is appropriate to include in an instructional manual. Why, then, do we not find actual instances of it? Presumably because it is an argument of last resort, to be employed only by someone who acknowledges that his case is weak, and only too readily countered by the arguments for sticking to the letter of the law which Aristotle adduces in the same chapter (1375b16–25). Appeal to the unwritten law, or to what is naturally right, to override claims validated by the written law is thus a high-risk strategy for a pleader who aims to persuade a jury, and it is hardly surprising that we find no actual examples of its use.⁹ gods’, is elsewhere in the play called the law of all the Greeks (311, 526, 671) and one of the laws of mortals (378). Divine origin, ancient tradition, and universal acceptance are closely interwoven aspects of the popular conception of unwritten law. ⁹ Dover observes (GPM 306) that a claim to defeat man-made law by recourse to divine law might itself be characterized by an opponent as impious, as requiring a breach of publicly defined religious obligations (presumably, in the first instance, the juror’s oath, interpreted by the opponent as an oath to uphold the written law). Aristotle is aware that the oath might be represented, to suit the requirements of the case on either side, as binding the juror to uphold either written or unwritten law (Rhet. 1375a29–31 [unwritten], 1375b16–18 [written]).

popular morality and unpopular philosophy 133 We have, then, discovered some counter-examples to Dover’s generalization, not in the speeches of forensic advocates, but in the work of abstract writers, philosophers, historians, and dramatists. Nor is that surprising. For the claim to possess rights other than those recognized by the law is an abstract claim (though it may have considerable practical consequences), belonging to the area of discourse in which actual institutions are evaluated. That discourse, which is in modern times conducted via the intellectual press and similar media, was in the ancient world largely the province of the types of writer just enumerated. Dover’s decision to concentrate on the forensic orators to the comparative neglect of evidence from those areas of literature, especially philosophy, has, in my view, proved detrimental in this case. Wittgenstein identifies a main cause of philosophical disease as a onesided diet: ‘one nourishes one’s thinking with only one kind of example’ (Philosophical Investigations, i. 593). The point has as much application to other disciplines as to philosophy. Afterword For a more sceptical view of the attribution to the Greeks of a theory of human rights see Burnyeat [1994].

9 Socratic Ethics1 First a word in elucidation of my title. By ‘Socratic Ethics’ I intend, not ‘the ethical theories of the historical Socrates’ but ‘the ethical theories of Plato’s Socratic dialogues’, and by ‘Socratic dialogues’ I refer to those dialogues, generally reckoned to be early, in which (a) there is no mention of the Theory of Forms, and (b) Socratic elenchus is a central or at least prominent feature of the discussion.² I do not, therefore, propose to grapple directly with the classical ‘Socratic question’, viz. the question of how much of the historical Socrates may be recovered from our literary sources. Rather, I shall be exploring the internal connection between certain items of Plato’s presentation of Socratic ethics, on the one hand the attempt to reach definitions and on the other the two theses which have come to be labelled as ‘the Socratic paradoxes’, (i) that virtue is knowledge and (ii) that no one does wrong willingly or intentionally. Yet this enquiry, though not directly historical in intention, is not totally without historical interest, in that all those items are ascribed to Socrates by Aristotle. The two theses are attributed to him, directly or by implication, in the Nicomachean Ethics. At 1144b19–20 Aristotle says that Socrates thought that all the virtues were sorts (or perhaps instances) of phron¯esis, and a few lines later (29–30) that they were kinds (or instances) of knowledge, which I take to be in the context equivalent statements. Earlier (1116b4–5) he has attributed to Socrates an instance of this generalization, viz. the thesis that courage is a kind (or instance) of knowledge. As regards the second thesis, Aristotle famously sets up his discussion of akrasia in part via the claim, on which Socrates used to insist, that ‘no one acts contrary to what is best in the belief that he is doing so, but through error’ (1145b26–7), ¹ Some material from this chapter also appears in chapter 10 of this volume. ² The dialogues discussed in this chapter are Charmides, Crito, Euthyphro, Gorgias, Laches, Meno, and Protagoras.

socratic ethics 135 which Gerasimos Santas³ has shown to be the foundation of the doctrine that no one does wrong willingly. In a well-known passage of Meta. A 6 Aristotle singles out the search for ethical definitions as central to Socrates’ contribution to the development of philosophy: ‘Socrates ... was occupying himself with ethical matters, neglecting the world of nature as a whole but seeking the universal in ethical matters; he was the first to concentrate attention on definitions’ (987b1–4). He repeats this information in Meta. 1078b17 ff. with some important additions (b27–32): ‘there are two things which may justly be ascribed to Socrates, inductive arguments and general definitions, for both are concerned with the starting point of knowledge; Socrates did not, however, separate the universal or the definitions, but they [viz. Plato and his followers] did, calling them the Forms of things.’ Of course, we cannot simply take it for granted that Aristotle’s evidence is independent of Plato. That the evidence for the Socratic denial of akrasia is derived from the dialogues is suggested by Aristotle’s opening statement (1145b23–4) that ‘it would be astonishing, as Socrates thought, if though knowledge were present something else overcame it and dragged it about like a slave’, where the picturesque metaphor is a verbal echo of Prot. 352c1–2. Nothing in the contexts of the other Nicomachean Ethics passages requires that they should be independent of Platonic texts. On the other hand, in the Metaphysics Aristotle is attempting to identify the contribution of the historical Socrates to the development of philosophy, and in particular to the development of Platonism, and he distinguishes Socrates from Plato by a criterion which is not derived from the Platonic dialogues, viz. that Socrates did not separate the Forms, whereas Plato did.⁴ We are therefore justified in believing that not merely the Platonic but also the historical Socrates did seek ethical definitions, and that in his search for them he made use of inductive arguments (and also elenchus of his interlocutors).⁵ Beyond that I do not think we can go far in reconstructing the views of the historical Socrates. It seems to me quite ³ Santas [1964]; Santas [1979]. ⁴ As was argued long ago by Ross in his [1924]: vol. i, xxxiii–xlv. ⁵ In his Memorabilia Xenophon ascribes all three to Socrates. For definitions see I. 1. 16, III. 9. 4–9, IV. 6, esp. sect. 1; inductive arguments and elenchus are pervasive features of Socrates’ argumentative technique, as reported by Xenophon, e.g. II. 3 (induction), III. 6 (elenchus). I do not assume that Xenophon’s portrait is historically more accurate than Plato’s, but merely cite these passages as further evidence, possibly independent of Plato, for Socrates’ practice of argument.

136

socratic ethics

likely that he did maintain the Socratic paradoxes in some form or other.⁶ On the question of what kind of definition he sought, we have virtually no information from any source other than Plato,⁷ and are therefore systematically debarred from being able to determine how much is Plato and how much Socrates. From now on, therefore, I shall have no more to say about the historical Socrates. I shall instead consider what Plato’s Socrates is looking for when he looks for an ethical definition, and how the search for definitions connects with the Socratic paradoxes. I shall suggest that while Plato falls short of presenting Socrates as possessing a single clear conception of definition, one paradigm of definition predominates. Further, that paradigm has an intimate connection with the paradoxes; more precisely, the demand for definition is satisfied via the development of a psychological theory in which the paradoxes have a central position. I shall be focusing primarily on the Meno, which, though probably one of the latest of the Socratic dialogues, and in its introduction of the theory of recollection transitional to the metaphysical dialogues of the middle period, is nonetheless Socratic by my criteria of the prominence of elenchus and the absence of Forms, and is of all the Socratic dialogues the richest in evidence for the Platonic/Socratic theory and practice of definition. What is a Socratic definition a definition of? Not, in the first instance, of any of the items that might spring to the mind of the modern reader, such as a term, the meaning of a word, or a concept. Ideally, a Socratic definition answers the question ‘What is ...?’, where the blank is filled in by a word designating some quality or feature of agents, such as courage or excellence. So literally, what are to be defined are those qualities or features themselves, not anything standing for them, as words or perhaps concepts might be thought to do. But, of course, we cannot in general ⁶ At Mem III. 9. 4, Xenophon’s Socrates maintains that everyone does what seems to him to be most advantageous to him, a claim which leads to the first paradox, that no one does wrong willingly. (The claim is repeated at IV. 6. 6.) The second paradox, that virtue is knowledge, is maintained at III. 9. 5; ‘he said that justice and all the rest of excellence is wisdom’. (For the explicit identification of wisdom with knowledge see IV. 6. 7.) ⁷ The passages from Xenophon cited in n. 5 above contain instances of both types of definition distinguished in the chapter, viz. conceptual elucidation and substantive account, without any sign of a distinction between the two. Nor do these passages provide any clear evidence of Xenophon’s view of what the theoretical role of Socratic definition was. In general, Xenophon’s examples of Socratic definition are so few, and his treatment of them so cursory, that it is unsafe to base any conclusions on them.

socratic ethics 137 draw a sharp distinction between specifying what something is and defining or elucidating the concept of that thing. The concept of F (where ‘F’ is some general term) is what we understand or possess when we use the term ‘F’ with understanding, and in some cases saying what F is just is defining or elucidating the concept of F. Thus if I answer the question ‘What is justice?’ by saying that justice is giving everyone their due, I have thereby attempted (however inadequately) to elucidate the concept of justice, in that the answer is intended to make explicit what is standardly conveyed by our talk of justice. The ultimate authority for the correctness of that sort of definition is the competent speaker of the language in which the elucidation is expressed, and the ultimate test which that authority applies is conformity with his or her linguistic intuitions. In other cases, however, the question ‘What ‘is F?’ is aimed to elicit, not an elucidation of the ordinary concept of F, but an account of the phenomenon couched in terms of the best available scientific theory. For example, ‘light is a stream of photons’ is not an elucidation of the ordinary concept of light, i.e. of what the standard speaker of English understands by the word ‘light’. It is an account of what light is, i.e. of what science has discovered light to be, and that account presupposes, but is not exhausted by, the grasp of the concept which is available to the competent but pre-theoretical speaker of the language. Hence the ultimate test of its adequacy is not its fit with that speaker’s linguistic intuitions, but its explanatory power, empirical testability, or whatever else constitutes the test of a good scientific theory. The form of words ‘What is ... ?’ may express the search for a definition of either kind (and in any case the distinction between the two is less sharp than the foregoing oversimplification has suggested⁸). Precisely what kind of search is afoot has to be determined by the context, which cannot be guaranteed to provide an unambiguous result. We learn from the Euthyphro that a Socratic definition of a given quality should (a) specify what is common to all and only those things to which the name of the quality applies (5d), (b) specify that in virtue of which the name applies to them (i.e. give the nature of the quality, not merely a distinguishing mark of its presence) (6d, 11a), and (c) provide a criterion by reference to which disputed cases may be determined (6e). Requirements (a) and (b) are explicitly endorsed at Meno 72c, where Meno is invited to ⁸ See Putnam [1975].

138

socratic ethics

specify ‘the single nature they [i.e. the various types of human excellence] all have in virtue of which (di’ ho) they are excellences’, and though the Meno has nothing to say about disputed cases of aret¯e, we have no reason to suppose that the third requirement has been abandoned. These three requirements are satisfied alike by conceptual elucidations and by that kind of account which we have contrasted with those, and which we might call, traditionally, ‘real definitions’ or, perhaps more informatively, ‘substantive scientific accounts’. But the specific context of the Meno in which Socrates is prompted to ask ‘What is excellence?’ suggests that elucidation of the ordinary concept is unlikely to be sufficient for the purpose of the dialogue. The context is provided by Meno’s opening question (70a): ‘Is excellence something which can be taught, or is it rather acquired by practice, or is it a natural endowment, or is it acquired in some other way?’ That is to say, Meno wants to know how one may come to acquire excellence, i.e. how one can come to be an excellent or outstanding person, to which Socrates correctly replies (71b) that one cannot answer that question unless one first understands what excellence is. Now, what kind of understanding of the nature of a quality is required if one is to know how to acquire that quality? Socrates generalizes Meno’s question, maintaining (cf. Rep. I. 354b–c) that in general one cannot answer the question ‘Is x F?’ unless one can first answer the question ‘What is x?’ and glossing that requirement as analogous to the requirement that one should know who Meno is in order to know whether he is, for instance, handsome. The gloss is unhelpful: in order to know whether some predicate holds of an individual subject such as Meno one must indeed be able to individuate that subject, but the ability to individuate Meno does not require that one possess any uniquely individuating specification of Meno; one might, for instance, individuate Meno ostensively, just as ‘that man over there’. The analogue in the case of a universal such as excellence seems to be no more than the minimal requirement to know what we are talking about when we use the word, a requirement which again does not presuppose the ability to give a verbal specification, but which might be satisfied by the ability reliably to recognize typical instances. But that minimal knowledge of what excellence is (which Socrates and Meno obviously have already, otherwise the conversation could not even start) is clearly no help at all when it comes to the question of how to acquire excellence.

socratic ethics 139 An analogy with another evaluative predication may help to make the point clearer. Asking how to acquire excellence at tennis, I am told by Socrates that I must first know what excellence at tennis is, just as in order to know whether Meno is handsome I must first know who Meno is. But in that sense, i.e. that I must be able to identify Meno, of course I know what excellence at tennis is. Unless I could recognize it when I came across it, how could I ever raise the question of how to acquire it? So I need more than the bare ability to know what I am talking about. Perhaps, then, I need an elucidation of the concept. But now what counts as such an elucidation? A bare specification of the meaning of the expression ‘excellence at tennis’, for example ‘ability to play tennis to a high standard’, is plainly no help in guiding me towards the acquisition of excellence at tennis. For that I need some substantive knowledge of what excellence at tennis consists in, i.e. of what qualities go to make a first-class tennis player. There are at least two kinds of qualities which excellence at tennis may be said to consist in: first of all, since the game is composed of a variety of kinds of stroke, one might analyse excellence at tennis into the excellent performance of each of those kinds, having a good service, a good forehand, etc. That kind of analysis would be of some use to the novice, insofar as it articulates the undifferentiated goal of acquiring excellence at tennis into a number of subordinate goals. But it provides no guidance as to how each of the subordinate goals is to be pursued. For that, the general notion of excellence at tennis, and thereby the specific excellences constitutive of that excellence, excellent service, etc., have to be analysed via a specification of that complex of qualities which are causally necessary and sufficient for success at the game: for instance excellence in tennis is achieved via a complex of attributes such as hand–eye coordination, speed of reaction, stamina, tactical insight, and motivation. That is to say, I need to have the rudiments of a theory of tennis, which explains excellence by identifying its causes and thereby indicating appropriate methods of acquiring it. The practical orientation of the discussion in the Meno thus suggests that the account of excellence which is being sought ought to be an explanatory account of the type exemplified immediately above. Such an account would not only fulfil the requirements stated in the Euthyphro, but would give Meno the practical guidance he is looking for, as an elucidation of the ordinary concept of excellence would not. The text of the dialogue does, I

140

socratic ethics

believe, confirm that that type of account is predominant in Plato’s mind, but the evidence is by no means unambiguous: indeed the complexities are such as to force us to acknowledge that the distinction between conceptual elucidation and explanatory theory emerges with less than total clarity. Invited to say what excellence is, Meno first responds by the suggestion that there are various forms of human excellence, corresponding to certain primary social roles held by various kinds of person, the adult male, adult (married) woman, child, etc. (71e–72a). Socrates then poses his usual challenge to say what all these kinds of excellence have in common in virtue of which they all count as types of excellence (72a–d), and himself takes what is presumably the first step towards meeting this challenge by pointing out that excellence in each of Meno’s primary social roles requires the possession of justice and self-control (73a–c). This looks like a move towards the first of our two patterns of analysis of the conception of excellence at tennis, in which a complex characteristic is decomposed into its constituents; completed to Socrates’ satisfaction it would exhibit human excellence of each of Meno’s primary types as consisting in the possession of the same set of specific excellences or virtues, justice, self-control, courage, wisdom and perhaps some others (cf. 74a, 88a; Prot. 349b). But, as in our tennis example, the practical question of how to acquire excellence would still require a further account of each of those specific virtues, which would, it seems, have to be of the causally explanatory type. Socrates, however, is dissatisfied with this account for a quite different reason, viz. that it, like Meno’s original suggestion, fails to meet the requirement to say what all the instances (in this case all the specific virtues) have in common (74a). Prima facie the application of that requirement to this case is extremely puzzling. Surely what the specific virtues have in common is precisely that they are all constituents of overall human excellence: what more could Socrates possibly expect by way of answer? It is as if, being told that hand–eye coordination and motivation are constituents of excellence at tennis, one were to ask what they have in common. Plainly, they need have nothing in common other than that both are required in order to excel at tennis. If (implausibly) it were to turn out that coordination and motivation are themselves manifestations of some further, common power or property, that would have to be discovered empirically, not guaranteed a priori by the fact of our counting both these attributes as constituents of excellence at tennis.

socratic ethics 141 Socrates nevertheless presses his demand for a specification of what the specific virtues have in common, elucidating this demand by providing a model of the kind of specification he is looking for, a definition of shape (74b). This procedure is flawed, since the model is appropriate only if justice and self-control stand to overall human excellence as circularity and squareness stand to shape, but in fact they do not. Whereas specific shapes are all determinates of a common determinable, justice and self-control are not determinates of the determinable excellence, but constituents of excellence.⁹ There is no determinable of which they are determinates. There is indeed a general concept of which they are instances, viz. ‘constituent of overall human excellence’, but Socrates and Meno have already agreed that that concept applies to justice and self-control, so giving that definition would not advance the discussion. But although Socrates has taken a wrong turning in looking for a general account of excellence analogous to his definition of shape, there is some interest in seeing what kind of definition the latter is. In fact Socrates gives two definitions of shape. The first, ‘the only thing which always accompanies colour’ (75b10–11), does not appear to satisfy the requirement that a definition should specify what the thing to be defined is, and not merely give a distinguishing mark. One does not say what a property is by pointing to another property of which it is the invariable accompaniment, any more than one explains what night is by saying that it is the only thing which invariably follows day. (In fact, since a number of properties, for example visibility, extension, size, luminosity, always accompany colour, the so-called ‘definition’ is true of nothing.) However, nothing in the text indicates that this account is recognized as failing to meet the standard requirements for a definition; indeed Socrates says (75b11–c1) that he would be content if Meno could define excellence in that way. Meno demurs indeed, describing the definition as ‘silly’, but for an altogether different reason, viz. that it would not be informative to someone who did not understand what colour is (75c). He states a requirement of the informativeness of any definition, viz. that it should not contain any term which is not understood by the person to whom the definition is given, but that requirement is not an objection to the correctness of this or any definition. In reply Socrates gives another definition of shape, ⁹ Determinates of a single determinable, such as specific shapes and specific colours, are mutually incompatible; constituents of a complex property are (a) mutually compatible, (b) such that every instance of the complex property in fact instantiates some subset of the constituents. See Searle [1967].

142

socratic ethics

having first made sure that Meno understands all the component terms, viz. ‘shape is the limit of a solid’ (76a7). This account of shape, as the external boundary or limit of a body, does appear to be an elucidation of the concept. And since that is given by Socrates as a model of the kind of definition of excellence he is looking for, we have to recognize a divergence between the kind of account suggested by the practical direction of the enquiry, viz. a causally explanatory account, and the preferred Socratic model. A similar situation occurs in the Laches, where the practically motivated search for a definition of courage is illustrated by a model definition of quickness as ‘the power of doing many things in a short time’ (192b1–2), which is also a conceptual elucidation. The situation is complicated even further by Socrates’ offer of yet another definition, this time (in deference to Meno’s insistence) of colour. This is not a conceptual elucidation, but an explanatory account of colour in terms of a scientific theory, viz. the physiology of Empedocles: ‘Colour is a flowing out [sc. from the coloured object] of shapes [i.e. physical particles of various shapes] symmetrical to vision [i.e. of such shapes and sizes as enable them to penetrate the channels in the eye] and perceptible’ (76d4–5).¹⁰ (The modern analogue would be an account of colour in terms of light waves of different lengths.) Here we have an example of just the kind of explanatory causal theory suggested by the practical motivation of the dialogue; were Socrates to say that this is the most satisfactory of the definitions he has offered, we should have a nice fit between the shape of the dialogue and the explicit witness of the text. In fact he says that it is inferior to the definition of shape (76e; presumably it is the second of the two definitions of shape that he refers to, but the text does not make that explicit). But he does not say why it is inferior. Perhaps Plato is aware of the distinction between conceptual elucidation and causally explanatory account, and regards the former as in general the preferred sort of definition. But Socrates does not say so, and the text suggests other possibilities. Socrates offers the definition of colour as one in the manner of Gorgias, which will therefore be familiar to Meno (76c),¹¹ and remarks on its ‘high-flown’ character, which he thinks pleases Meno (76e3–4). This suggests that Socrates may regard this definition as inferior to the ¹⁰ Cf. DK 31 A 86 (Theophrastus) and 92 (Aetius). ¹¹ Gorgias had reputedly been a student of Empedocles (DL VIII. 58: DK 82 A 3), and Meno had studied with Gorgias (71c).

socratic ethics 143 other, not because causally explanatory accounts are as such inferior to conceptual elucidations, but because this particular account is couched in overelaborate technical terminology. Perhaps, too, he thinks that it is simply false, but has merely picked on it to gratify Meno by citing something familiar. Given these hints, it would be rash to conclude that Socrates’ preference for the definition of shape indicates a preference for one type of definition, conceptual elucidation, over another, a causally explanatory account. Indeed, the text thus far gives no indication that Plato is even aware of the distinction between the two types of definition. Let me sum up the position so far. Socrates responds to Meno’s practical question of how excellence is to be acquired by stating a minimal requirement for any enquiry, viz. that the object of the enquiry should be identified. But that requirement is already fulfilled from the outset. Socrates is in fact demanding the fulfilment of a stronger requirement, viz. to provide an account of excellence which will meet the standards set out in the Euthyphro. Those standards do not differentiate between a conceptual elucidation and a causally explanatory account, but the practical ends of the enquiry would be better served by the latter. In the course of the discussion Socrates gives samples of accounts: the first, which gives a distinguishing mark of the object ‘defined’ rather than specifying what the object is, fails to meet the Euthyphro standard, the second is a conceptual elucidation, the third a causally explanatory account. Socrates gives no sign of regarding the three as belonging to different kinds, but merely states that, for reasons unspecified, he prefers the second to the third. Thus far, then, the question ‘What kind of definition is Socrates looking for in the Meno?’ has no unambiguous answer. If such an answer is to be had, we must take into consideration the rest of the dialogue. That may seem unpromising, since in that portion of the dialogue nothing more is said about the methodology of definition, nor are any more examples given. The dialogue does, however, reach an answer, or rather answers, to the question ‘What is excellence?’ Excellence is first (87c–89a) argued to be knowledge, then another argument leads to the revision of that account in favour of the answer ‘Excellence is true opinion’ (99b–c), an account which is further qualified by a strong hint at 100a that the former answer gives the true account of genuine excellence, while the latter gives an account of what passes for excellence by ordinary standards. What kind of definition is represented by these answers?

144

socratic ethics

What these answers share is a conception of excellence as a cognitive state, or more portentously, a grasp of truth. We are not concerned with the details of the distinction between knowledge and true opinion, but its essence is the firmness or reliability of the grasp; knowledge is a reliable (because systematic) grasp of the truth, true opinion is an unreliable (because unsystematic) grasp of truth. What sort of account of excellence is provided by the conception of it as a cognitive state? Note that Socrates and Meno are not discussing excellence in some theoretical sphere, such as excellence at mathematics; the paradigms of excellence in this final stage of the discussion are individuals such as Pericles who embody the same ideal of success in public and private affairs as Meno had assumed from the outset. That is to say, cognitive states are presented as giving an account of all-embracing social and political merit, of the state of the totally well-rounded, successful and admirable person (= man, by this stage of the discussion). In what sense is that person’s admirable state a cognitive one? The sort of excellence in question is above all practical, manifested in action and a whole style of behaviour; it does not seem that specification of a cognitive state could give an account of the manifestation of excellence. That is to say, ‘knowledge or true opinion’ does not offer the same kind of account of excellence as ‘having a good service, ground strokes and volley’ does of excellence at tennis, since the latter account does precisely specify the kind of actions which manifest that excellence, whereas the former does not. Rather, it gives an account of what is manifest in excellent performance, as ‘coordination, stamina, courtcraft, etc.’ does of what is manifested in excellent tennis-playing. And as that account expresses, not an elucidation of the concept of excellent tennis-playing, but an empirical theory of the causes of the type of play which we count as excellent, so the cognitive account of overall human excellence expresses, not an elucidation of that concept, but a causally explanatory theory, what we may call the Cognitive Theory of Virtue. We find, then, that despite Socrates’ expressed preference for the conceptual definition of shape over the causally explanatory account of colour, the account of excellence which he endorses in the concluding section of the dialogue is of a type represented, not by the former, but by the latter. The aim of our enquiry into the nature of definition in the Socratic dialogues was the elucidation of the connection between definition and the Socratic paradoxes. That aim has now been achieved in part, since

socratic ethics 145 the first paradox, the thesis that virtue is knowledge, has now turned out to be the Cognitive Theory. But that elucidation lacks content without further exploration of just what the Cognitive Theory claims. That exploration will also, I hope, indicate the connection between the Cognitive Theory and the second paradox, the thesis that no one does wrong willingly. The Cognitive Theory is a theory to the effect that overall success in human life is guaranteed by the possession of certain cognitive states. This theory in turn rests on a theory of the explanation of intentional action which combines to a remarkable degree a staggering audacity and simplicity with a high degree of plausibility. It states that provided that the agent has a conception of what is overall best for the agent, or (equivalently) what is maximally productive of eudaimonia (for the agent), that conception is sufficient to motivate action with a view to its own realization. That is emphatically not to say that motivation does not require desire as well as belief. On the contrary, Socrates makes clear his view (77c1–2; 78b4–6) that everyone desires good things, which in context has to be interpreted as the strong thesis that the desire for good is an invariable motive. That desire is then conceived as a standing motive, which requires to be focused in one direction or another via a conception of the overall good. Given that focus, desire is as it were locked on to the target which is picked out by the conception, without the possibility of interference by conflicting desires. Hence, given the standing desire, all that is required for correct conduct, i.e. for the manifestations of excellence, is the correct focus. And that focus has to be a correct conception of the good for the agent, i.e. a correct conception of eudaimonia. I have just stated the theory underlying the Cognitive Theory in a conditional form, viz. provided that the agent has a conception of what is overall good, that conception provides uniform motivation. That form allows the possibility that an agent might lack the conception of such a good altogether, and simply be motivated in a haphazard way by considerations of particular goods on each occasion of action. A stronger version of the theory would be the thesis that on every occasion every agent is motivated by his or her conception of what is best overall. That version has the disadvantages (a) of going beyond the letter of the relevant texts (Meno 77–8; Prot. 358), and (b) of extreme implausibility. On the other hand, the weaker thesis has the disadvantage that it leaves the uniformity of

146

socratic ethics

motivation, given the conception of overall good, virtually unexplained. If agents are capable, lacking the conception of overall good, of being motivated towards different, and therefore potentially conflicting goods, what guarantees that the conception of overall good, once in place, will be sufficient to silence conflicting motivations? Prot. 358c–d states that it is impossible (literally, ‘not in human nature’) to choose what you think bad (sc. for you) instead of what is good, but gives no argument or explanation. Meno 77e–78a gives an argument, but an unsound one: to want things which you think bad for you is to want to be harmed, and to want to be harmed is to want to be wretched and unfortunate, but no one wants to be wretched and unfortunate. (The first premiss is false, because wanting does not cross envisaged causal connections: from the fact that I want x, and believe that x causes y, it does not follow that I want y.) Plato might perhaps rely on that argument, faced with the objection which I have just raised. I suspect, however, that the objection would not have occurred to him, because he may not have made the distinction between the desire for particular goods and desire for the good overall on which the objection depends. The argument of Prot. 355c–d, leading to the conclusion that it would be absurd to describe someone as knowingly doing bad things (i.e. things bad for him) because he is overcome by the desire for goods, depends on failure to make that distinction. It is absurd to describe someone as abandoning the conception of the overall good because he is seduced by the attractions of the overall good, but not absurd to describe one as doing so because he is seduced by particular goods which are incompatible with the overall good (for example short-term pleasures which are believed to be harmful in the long run). It is therefore quite likely that Plato was not sensitive to the distinction; and with the distinction goes the distinction between the weaker and the stronger forms of the theory of motivation on which the Cognitive Theory rests. On the weaker form, motivation is (a) always towards some envisaged good, and (b) towards the envisaged overall good when the conception of that good is present. On the stronger form, motivation is always towards the envisaged overall good. On either form, excellence in human performance is guaranteed by possession, with greater or less reliability, of the correct conception of the good. In turning to the second Socratic paradox I shall assume Santas’ distinction between the self-interested paradox, ‘No one acts intentionally against his or her overall interest’, and the moral form, ‘No one does intentionally

socratic ethics 147 what is morally wrong.’ The former falls directly out of the theory of motivation which we have seen to underly the Cognitive Theory. The latter, as Santas demonstrates, follows from the former, together with the thesis, argued for in the Gorgias, that it is always in the agent’s interest to do what is morally right. The agent must, of course, believe that it is in his or her interest to do what is morally right, but the correct conception of the good for the agent will guarantee that he or she has that belief. For the correct conception of the agent’s good (see Gorg. 504a–d) is that it is a state of the personality organized in accordance with the requirements of virtue, analogous with bodily health (cf. Crt. 47e). Given that conception, together with true beliefs as to what actions will realize it, the prudential thesis guarantees that the agent will be motivated to do those actions. Hence, by contraposition, if the agent is not thus motivated, he or she either lacks that conception or lacks those true beliefs. In either case, the agent’s wrongdoing is due to error, and is therefore unintentional, which is what the moral paradox states. The practical orientation of the initial enquiry in the Meno, then, suggests that the kind of account of excellence which is required is a causally explanatory account. That hypothesis is neither confirmed nor refuted by Socrates’ model definitions, which are of diverse kinds, including both a causal account and a conceptual elucidation, without any clear recognition of the distinction between those kinds. On the other hand the alternative accounts of excellence which are finally arrived at are in fact causal accounts. Cognitive states explain how practical excellence is achieved, the explanation being embedded in a theory of motivation which also yields the Socratic paradoxes. To guard against possible misunderstanding, I do not of course deny that the Socratic conception of excellence, as mentioned in the Crito and spelt out in the Gorgias, is revisionary. The ‘Periclean’ paradigm of excellence is assumed in the Meno only for the sake of argument, and in fact represents at best an approximation to real excellence. But that does not affect my central contention, that the relation of the cognitive grasp of the good to excellence itself is in the Socratic dialogues a causal one, the causal mechanism being that described in the theory of motivation which I have just sketched. In the authentic Socratic conception the good of which the excellent person has knowledge is itself a state of the personality, analogous to bodily health, in which the various impulses to action are

148

socratic ethics

harmonized under the direction of the agent’s conception of the supreme good. But now the theory collapses into inconsistency. By the Cognitive Theory excellence is knowledge of how to achieve the good, while by the analogy with health excellence is the health of the soul and therefore the good itself which is to be achieved via knowledge. On the former account excellence should, like medicine, be valuable only instrumentally, whereas by the latter it is valuable intrinsically. Further, the theory of the Gorgias requires that the personality should be organized in such a way as to realize a prior conception of the good, whereas the only conception of the good envisaged in the Cognitive Theory is the conception of the state of the personality itself. The conception of the supreme good as the Form of the Good, developed in the middle dialogues, gives Plato at least the sketch-plan of an escape from this dilemma. In the foregoing I have dealt largely, though not exclusively, with the Meno. The causal character of Socratic accounts of the virtues is also discernible in the Protagoras, where Socrates argues for the thesis that all the virtues are forms of knowledge, and is suggested by the Laches and the Charmides, where the respective discussions indicate, without definitively enunciating, accounts of courage as ‘knowledge of what is fearful’ (an account adopted at Prot. 360d) and s¯ophrosun¯e as ‘knowledge of the good’. In conclusion, all that I hope to have shown is that in the most extended treatment of Socratic definition, the definition sought and arrived at is not a conceptual elucidation, but a causally explanatory account, that that sort of definition is prominent in other Socratic dialogues, and that, given the theory of motivation presupposed in the particular definition which is given in the Meno, the connection between that definition and the Socratic paradoxes is both intimate and reasonably perspicuous. To repeat myself, I do not think that there is any indication that Plato made any explicit distinction between conceptual elucidation and causal account; the distinction which I have emphasized (and whose sharpness I may have exaggerated) emerges, not in Plato’s theory, but in his dialectical practice. Afterword p. 141 ‘a definition of shape’. I take for granted the standard interpretation of sch¯ema in 74b–76d as ‘shape’. Scott [2006] argues (pp. 37–42) that sch¯ema means ‘surface’ rather than ‘shape’, and hence

socratic ethics 149 that in the definition of colour at 76d4–5 the expression which I render ‘a flowing out of shapes’ (aporro¯e sch¯emat¯on) should be understood as ‘a flowing out from (or possibly ‘of’) surfaces’. That specific issue does not affect the general question of the type of definition which is being proposed. On either rendering the definition of sch¯ema amounts to an elucidation of the concept, and that of colour to a causal explanatory account.

10 Platonic Ethics The fundamental question of Platonic ethics is ‘How should one live?’ (Rep. I.352d, Gorg. 500c). That question is not to be understood as ‘What is the morally best way to live?’, as is shown by the fact that in Rep. I an appropriate, though in Plato’s view false, answer to it is that given by Thrasymachus, namely that one should live by emancipating oneself to the best of one’s ability from the restraints of morality with a view to the furtherance of one’s own interest. Rather it is to be understood as ‘How may one achieve the life which is, objectively, but from the point of view of one’s own interest, the most worth living?’ (Rep. I.344e). The Greek term for the achievement of such a life is eudaimonia (literally ‘having a favourable guardian spirit’ (daim¯on) ), conventionally translated ‘happiness’, but in view of its objective character better rendered ‘blessedness’ or ‘well-being’. According to Aristotle (EN 1095a18–20) it was universally acknowledged (a) that eudaimonia was the supreme good and (b) that the term meant ‘living well’ and ‘doing well’; nothing in the texts of Plato suggests that his use of the term conflicts with these claims. In the same passage Aristotle tells us that there were substantive disputes about what living well amounted to, some holding, for example, that it consists in acquiring wealth, others that it consists in a life of honour or of intellectual achievement; Plato depicts such substantive disputes in Socrates’ confrontations with Thrasymachus in Rep. I and Callicles in the Gorgias. The agreement on ‘How is one to live well?’ as the basic question of ethics forecloses certain ethical disputes while leaving others open. Most fundamentally, ethical questions are approached from the standpoint of the individual’s interest, the promotion of which is assumed to be the primary function of the individual’s practical rationality. On that assumption one has adequate reason to undertake any action if and only if so doing

platonic ethics 151 will contribute to one’s living well, i.e. to one’s having an objectively worthwhile life. The conception of an objectively worthwhile life should not be construed in a narrowly egoistic way, since it may be part of an objectively worthwhile life that one cares for the good of others, not merely instrumentally, with a view to the benefits one may expect to gain from such benevolence, but for its own sake. Nevertheless, it is broadly egoistic,¹ in that it is assumed that the value to the carer of that selfless care lies primarily in its contribution to the life of the carer, and only secondarily in its contribution to the life of the person cared for. Since this broadly egoistic conception of the role of practical reason is assumed from the outset, there is no room in Platonic thought for theories of a Kantian type, which seek to identify moral principles as imperatives binding unconditionally on any rational agent in total independence of any considerations of the interest of that or any other agent.² It is therefore unsurprising that our texts provide no hint that the very conception of that type of theory had so much as occurred to Plato. The broadly egoistic starting point of Plato’s ethical enquiries is not, then, open to question. By contrast, the status of morality is an eagerly debated question in some of his major dialogues. By morality I understand a socially regulated system of norms imposing restraints on the pursuit of self-interest with the general aim of furthering social cooperation, for which the nearest Greek terms are to dikaion and dikaiosun¯e, conventionally ‘the just’ and ‘justice’. The questions ‘What is the morally right thing to do?’ and ‘Which is the morally best way to live?’ were certainly not unintelligible to Plato, or to Greeks of his time generally. They were, however, both distinct from and posterior to the fundamental question which we have already identified, ‘How should one live?’, i.e. ‘How should one achieve the best life for oneself?’ The former questions were posterior in that, whereas ¹ For a useful discussion of these two varieties of egoism, labelled respectively ‘moral solipsism’ and ‘moral egocentrism’, see Irwin [1977], 255. ² This contrast between Platonic and Kantian theory leaves open the question whether, in the former, the agent’s interest may itself be seen as consisting in the acquisition of states of the personality (i.e. virtues) which have value independently of their contribution to the agent’s eudaimonia. There are theoretically at least three possible views on the relationship of virtue to eudaimonia: (i) virtue is valuable purely instrumentally, as a means to eudaimonia, (ii) virtue is at least partly constitutive of eudaimonia, and is intrinsically valuable qua constitutive of eudaimonia, (iii) virtue is valuable both in its own right and as either a means to or as a constituent of eudaimonia. I argue in note 21 below that (ii) is closer than (i) to giving an account of Plato’s view in the early dialogues. I know of no evidence to suggest that Plato was aware of (iii) as an alternative to (i) and (ii). (EN 1097b2–5 indicates that Aristotle may have been.)

152

platonic ethics

the rationality of the pursuit of the best life for oneself was unquestioned (indeed one could go so far as to say that that pursuit constituted practical rationality for the Greeks), the rationality of the individual’s observance of the dictates of morality required to be established by showing that their acceptance was necessary for the achievement of the individual good life. Plato attempts to meet the challenge to provide this justification of morality in the Gorgias and, in a much more elaborate and extended form, in the Republic. He accepts, then, that morality requires justification, in the form of a defence of its rationality, that this justification must be in terms of a broad conception of individual interest, and that such a justification can be provided. The requirement of justification seems an inevitable response to the fact that morality is essentially cooperative, requiring sacrifices from the individual for the common good. By Plato’s time there had been developed theories of the social nature of morality, which attempted to ground morality in self-interest (in a similar fashion to the theories of Hobbes and Hume³) by showing how norms of self-restraint and social cooperation would naturally develop in primitive societies as a device for mutual protection against the onslaughts of wild animals or (more plausibly) of unsocialized individuals.⁴ But the success of such theories seems limited. They show convincingly why self-interested individuals have reasons to prefer the existence of such institutions to a Hobbesian war of all against all, ‘since we benefit from one another’s justice and goodness’ (Prot. 327b). But they are unable to show that, given that the institutions exist, each individual benefits more from the sacrifices which he or she is required by the norms of the institution to make than he would do by taking advantage of the sacrifices of others to promote his own interest. The sacrifices might be regarded as one’s subscription to the mutual security club, which it is in one’s interest to pay, since the other members would not accept that one should enjoy the benefits of membership without paying one’s dues. But if one can get away without paying, as one fairly clearly can now and again (though not, doubtless, always), why pay in those circumstances? Of ³ Hobbes, Leviathan, chs. 13, 17; Hume, Treatise III. ii. 1–2. ⁴ Plato provides examples in the myth in the Protagoras (320c–322d) and in the theory proposed by Glaucon in Rep. II (358e–359b); an example independent of Plato is the so-called ‘Anonymus Iamblichi’, on which see Guthrie [1969], 71–4 and 314–15. For a valuable discussion of the historical sources of this tradition see Kahn [1981].

platonic ethics 153 course, it is unfair not to pay, which is a perfectly good reason for someone who is already committed to being fair, but the theory was supposed to generate a purely self-interested reason for undertaking that commitment, which it clearly fails to do. As Glaucon points out (Rep. 362a), the most that the theory can provide is a self-interested reason for making other people believe that one always deals fairly with them, which falls short of a reason for always actually doing so; but the latter, not the former, is what is required for the justification of morality. These problems reflect a crisis in traditional Greek morality, to which Plato’s ethics presents a sustained response. Traditional morality recognized certain states of character, principally courage, self-control, justice or fairness and piety, as the principal qualities which made their possessor an outstanding and admirable person.⁵ The young were brought up to regard possession of these qualities as fine and admirable (kalon) and the lack of them as disgraceful (aischron), and their inculcation was the principal aim of education.⁶ These were the most important among the excellences (aretai), i.e. those qualities stable possession of which, together with such external goods as wealth, position in society and physical health, constituted success in life (to eu z¯en or eudaimonia), which, we have seen, was universally acknowledged as the supreme good. But the arguments which we have just glanced at show that the claims of certain aretai, notably justice, to be constitutive of the agent’s good are at odds with the other-regarding character of those qualities. A further difficulty arises from the fact that some of these qualities are no less indeterminate in character than eudaimonia itself. Thus even if it is granted that success in life requires piety, i.e. a proper attitude to the gods, including respect for those obligations which the gods impose on us, there can be apparently irresoluble disputes about what kind of conduct really is required by the gods, as in the famous ⁵ Pindar, Isthmian 8.24–7; Aesch. Septem 610; Euripides, fr. 282.23–7; Xenophon, Mem III. ix. 1–5; Plato, Prot. 329c, Meno 73e–74a, Rep. 427e, etc. (See Irwin [1977], p. 287, n. 1.) ‘Self-control’ renders s¯ophrosun¯e, a term which lacks a precise English equivalent. It connotes primarily a proper sense of oneself and one’s limitations in relation to others, and derivatively various applications of that sense, especially control of the bodily appetites. Hence in many contexts, including those provided by Plato’s tripartite psychology, it is appropriately rendered ‘self-control’. Where in this chapter I use the terms ‘self-control’ and ‘self-controlled’ they correspond to the Greek s¯ophrosun¯e and its cognate adjective s¯ophr¯on, but I have not attempted total uniformity of usage, preferring sometimes to use the Greek term. (In my commentary on the Protagoras ( Taylor [1976 and 1991]), I prefer the rendering ‘soundness of mind’, for reasons explained there on pp. 122–4.) ⁶ Prot. 324d–326e.

154

platonic ethics

example in Herodotus (III. 38) of the diverse customs of different nations in the disposal of the dead, a central case of religious obligation. Both difficulties may be seen to have prompted an emphasis on the distinction between the nature or reality of things (phusis) and convention (nomos).⁷ The lack of coincidence between the agent’s interest and the demands of morality leads to the claim that while nature prompts us to seek our own interest, the demands of morality spring from nothing but convention (with the implication that the latter, unlike the promptings of nature, lack any authority).⁸ The indeterminacy of some of the conventional excellences is similarly attributed to the facts that (a) their requirements arise from nothing more than convention, and (b) different societies have different conventions.⁹ These criticisms of conventional morality should not be assimilated to modern attacks on the objectivity of values. Critics who, like Glaucon or the sophist Antiphon,¹⁰ denigrate conventional morality by contrast with the promptings of nature assume that the latter is a locus of objective value. The extreme form of this position is upheld by Callicles, who maintains that unrestrained self-assertion is naturally right (phusei dikaion), and that the conventional morality (nom¯oi dikaion) which opposes it is, therefore, naturally wrong, i.e. really or objectively immoral (Gorg. 483c–484c).¹¹ While Thrasymachus is not prepared to go that far, he asserts that injustice is a form of wisdom and an excellence, in that it gives the agent a worthwhile life, while justice is weakness and folly, in that it harms the agent and promotes the good of others (Rep. 343b–344c, 348c–e). We have no reason to interpret these claims otherwise than as statements of fact. It is significant that the earliest application of the nature–convention contrast to morality, attributed to Archelaus, who is said to have been a pupil of Anaxagoras and a teacher of Socrates, states simply that ‘the just and the disgraceful are by convention, not by nature’ (DL II. 16; DK 60 A 1). Here, what is purely a matter of convention is what is morally right and wrong, not what is good or bad; for example there is no suggestion that it is purely (or at all) a matter of convention that health is a good ⁷ Principal discussions of the nomos–phusis distinction include Heinimann [1945]; Guthrie [1969], ch. 4; Kerferd [1981], ch. 10. ⁸ Rep. 359c. Cf. Prot. 337c–d. ⁹ Herodotus III. 38. ¹⁰ DK 87 B 44. See Guthrie [1969], chs. 4 (a) (ii) and II (5); Saunders [1977–8]. ¹¹ I discuss the point more fully in chapter 8 of this volume.

platonic ethics 155 state and illness a bad one. Later, when the sceptics applied their universal strategy of suspension of judgement to the special case of claims about value, they did not confine their critique to moral value, but applied it to value generally;¹² but that seems to have been a post-Platonic development. At Tht. 172a–b Socrates asserts that ‘in matters of what is just and unjust and holy and unholy [people] are willing to maintain that none of these things is so in reality (phusei) or has its own nature (ousian), but what is agreed on [sc. by each community] is the truth for as long as it is agreed’, but contrasts that conventionalism about morals with the position about what is advantageous, where ‘no one would dare to say that what a community lays down as advantageous for itself is so in fact’. That assertion is confirmed by our other evidence: both sides of the dispute about nature and convention accepted that genuine values were part of nature, the critics of conventional morality attacking its values as spurious because they are merely conventional and therefore not part of nature, its defenders urging that on the contrary moral values are natural and therefore genuine. We have already noticed one defensive move, the theory (outlined in Protagoras’ myth) that moral conventions are themselves natural, in that they are strategies for cooperation developed by human beings struggling for survival in a hostile environment. But that defence was insufficient, since it failed to show that moral value passed the primary test for being natural, namely that of promoting the individual well-being of the agent. If his defence of traditional morality was to pass that test, Plato had to develop a better theory of the nature of morality and of human nature, in the hope of demonstrating the objective goodness of the traditional virtues via their contribution to the perfection of that nature, and therefore to the objectively worthwhile life for the agent. In what follows I shall set out what I take to be the main lines of Plato’s attempt to develop such a theory. I shall distinguish three stages in this process: i the theory of the early dialogues¹³ ii the theory of the Republic iii developments subsequent to the Republic. ¹² See Annas [1998]. ¹³ For the purpose of this chapter I count the following as early dialogues: Apology, Crito, Euthyphro, Charmides, Laches, Lysis, Euthydemus, Protagoras, Meno, Gorgias.

156

platonic ethics

i The early dialogues A central preoccupation of the early dialogues is the search for definitions, whether of individual excellences (courage in Laches, s¯ophrosun¯e in Charmides, piety in Euthyphro), of excellence in general (Meno) or of friendship (Lysis), an aspect of life intimately related to excellence and the good. In order to understand the prominence of definition in these dialogues, and its connection with the theory of the nature of the virtues which emerges in them, it is necessary to consider what Plato’s Socrates is looking for when he looks for an ethical definition. In outline, the project of the early dialogues is to give accounts of the traditional virtues which will exhibit them as natural goods; to investigate how that project was carried out would require close examination of the relevant texts, for which space is lacking here.¹⁴ I must confine myself in this chapter to a bald statement of the results of that investigation, focusing primarily on the Meno, which, though probably one of the latest of the dialogues which I here count as early, and in its introduction of the theory of recollection transitional to the metaphysical dialogues of the middle period, is of all the early dialogues the richest in evidence for the Platonic/Socratic theory and practice of definition. A Socratic definition is not, in the first instance, a definition of any of the items that might spring to the mind of the modern reader, such as a term, the meaning of a word, or a concept. Ideally, a Socratic definition answers the question ‘What is ... ?’, where the blank is filled in by a word designating some quality or feature of agents, such as courage or excellence. So, literally, what are to be defined are those qualities or features themselves, not anything standing for them, as words or perhaps concepts might be thought to do. But, of course, we cannot in general draw a sharp distinction between specifying what something is and defining or elucidating the concept of that thing. The concept of F (where ‘F’ is some general term) is what we understand or possess when we use the term ‘F’ with understanding, and in some cases saying what F is precisely is defining or elucidating the concept of F. Thus if I answer the question ‘What is justice?’ by saying that justice is giving everyone their due, I have thereby attempted (however inadequately) to elucidate ¹⁴ For fuller discussion see chapter 9 of this volume.

platonic ethics 157 the concept of justice, in that the answer is intended to make explicit what is standardly conveyed by our talk of justice. The ultimate authority for the correctness of that sort of definition is the competent speaker of the language in which the elucidation is expressed, and the ultimate test which that authority applies is conformity with his or her linguistic intuitions. In other cases, however, the question ‘What is F?’ is aimed to elicit, not an elucidation of the ordinary concept of F, but an account of the phenomenon couched in terms of the best available scientific theory. For example, ‘light is a stream of photons’ is not an elucidation of the ordinary concept of light, i.e. of what the standard speaker of English understands by the word ‘light’. It is an account of what light is, i.e. of what science has discovered light to be, and that account presupposes, but is not exhausted by, the grasp of the concept which is available to the competent but pre-theoretical speaker of the language.¹⁵ Hence the ultimate test of its adequacy is not its fit with that speaker’s linguistic intuitions, but its explanatory power, empirical testability, or whatever else constitutes the test of a good scientific theory. The form of words ‘What is ... ?’ may express the search for a definition of either kind (and in any case the distinction between the two is less sharp than the foregoing oversimplification has suggested¹⁶). Precisely what kind of search is afoot has to be determined by the context, which cannot be guaranteed to provide an unambiguous result. We learn from the Euthyphro that a Socratic definition of a given quality should (a) specify what is common to all and only those things to which the name of the quality applies (5d), (b) specify that in virtue of which the name applies to them (i.e. give the nature of the quality, not a mere distinguishing mark of its presence) (6d, 11a), and (c) provide a criterion by reference to which disputed cases may be determined (6e). Requirements (a) and (b) are explicitly endorsed at Meno 72c, where Meno is invited to specify ‘the single nature they [i.e. the various types of human excellence] all have in virtue of which (di’ ho) they are excellences’, and though the Meno has nothing to say about disputed cases of aret¯e, we have no reason to suppose that the third ¹⁵ In certain cases the scientific account may even demand revision of the pre-existing concept, as in cases where the latter itself carries connotations of a superseded scientific theory, for example hysteria. ¹⁶ See Putnam [1975].

158

platonic ethics

requirement has been abandoned. These three requirements are satisfied alike by conceptual elucidations and by that kind of account which we have contrasted with those, and which we might call, traditionally, ‘real definitions’ or, perhaps more informatively, ‘substantive or scientific accounts’. In the course of the discussion Socrates gives two model definitions, of shape and colour respectively. The first of these, ‘Shape is the limit of a solid’ (76a7), is a conceptual elucidation, whereas the second (76d4–5) is a substantive account of colour in terms of a scientific theory, namely the physiology of Empedocles.¹⁷ He says that the latter is inferior to the definition of shape, but he does not say why it is inferior, and it would therefore be rash to conclude that Socrates’ preference for the definition of shape indicates a preference for one type of definition, conceptual elucidation, over another, a causally explanatory account. Indeed, the text thus far gives no indication that Plato is even aware of the distinction between those types of definition. So if we are to answer the question ‘What kind of definition is Socrates looking for in the Meno?’, we must take into consideration the rest of the dialogue, where, though nothing more is said about the methodology of definition, the question ‘What is excellence?’ is answered. Excellence is first (87c–89a) argued to be knowledge, then another argument leads to the revision of that account in favour of the answer ‘Excellence is true opinion’ (99b–c), an outcome which is further qualified by a strong hint at 100a that the former answer gives the true account of genuine excellence, while the latter gives an account of what passes for excellence by ordinary standards. What these answers share is a conception of excellence as a cognitive state, or more portentously, a grasp of truth. We are not concerned with the details of the distinction between knowledge and true opinion, but its essence is the firmness or reliability of the grasp; knowledge is a reliable (because systematic) grasp of truth, while true opinion is an unreliable (because unsystematic) grasp of truth. What sort of account of excellence is provided by the conception of it as a cognitive state? Note ¹⁷ ‘Colour is a flowing out [sc. from the coloured object] of shapes [i.e. physical particles of various shapes] symmetrical to vision [i.e. of such shapes and sizes as enable them to penetrate the channels in the eye] and perceptible’ (76d4–5). ( The modern analogue would be an account of colour in terms of light waves of different lengths.)

platonic ethics 159 that Socrates and Meno are not discussing excellence in some theoretical sphere, such as excellence at mathematics; the paradigms of excellence in this final stage of the discussion are individuals such as Pericles who embody the same ideal of success in public and private affairs as Meno had assumed from the outset. That is to say, cognitive states are presented as giving an account of all-embracing social and political merit, of the state of the totally well-rounded, successful and admirable person (= man, by this stage of the discussion). In what sense is that person’s admirable state a cognitive one? The sort of excellence in question is above all practical, manifested in action and a whole style of behaviour; it does not seem that specification of a cognitive state could give an account of the manifestation of excellence. In other words, ‘knowledge or true opinion’ does not offer the same kind of account of excellence as ‘having a good service, ground strokes and volley’ does of excellence at tennis, since the latter account does precisely specify the kind of actions which manifest that excellence, whereas the former does not. Rather, it gives an account of what is manifested in excellent performance, as ‘coordination, stamina, courtcraft, etc.’ does of what is manifested in excellent tennisplaying. And as that account expresses not an elucidation of the concept of excellent tennis-playing but an empirical theory of the causes of the type of play which we count as excellent, so that cognitive account of overall human excellence expresses, not an elucidation of that concept, but a causally explanatory theory, what we may call the Cognitive Theory of Excellence. We find, then, that despite Socrates’ expressed preference for the conceptual definition of shape over the causally explanatory account of colour, the account of excellence which he endorses in the concluding section of the dialogue is of a type represented, not by the former, but by the latter. The aim of our enquiry into the nature of definition in the early dialogues was the elucidation of the nature of the theory of virtue to which that practice of definition is preliminary. That aim has now been achieved in part, with the identification of that theory as the Cognitive Theory of Excellence (more familiar as the first of the so-called ‘Socratic paradoxes’, the thesis that virtue is knowledge). But that elucidation lacks content without further exploration of just what the Cognitive Theory claims. That exploration will also, I hope, indicate the connection between the Cognitive Theory and the second paradox, the thesis that no one does

160

platonic ethics

wrong willingly,¹⁸ and also with the much-discussed question of the unity of virtue. The Cognitive Theory is a theory to the effect that overall success in human life is guaranteed by the possession of certain cognitive states.¹⁹ This theory in turn rests on a theory of the explanation of intentional action, which combines to a remarkable degree a staggering audacity and simplicity with a high degree of plausibility. It states that provided that the agent has a conception of what is overall best for the agent, or (equivalently) what is maximally productive of eudaimonia (for the agent), that conception is sufficient to motivate action with a view to its own realization. That is emphatically not to say that motivation does not require desire as well as belief. On the contrary, Socrates makes clear his view (77c1–2, 78b4–6) that everyone desires good things, which in context has to be interpreted as the strong thesis that the desire for good is an invariable motive. That desire is then conceived as a standing motive, which requires to be focused in one direction or another via a conception of the overall good. Given that focus, desire is as it were locked on to the target which is picked out by the conception, without the possibility of interference by conflicting desires. Hence, given the standing desire, all that is required for correct conduct, i.e. for the manifestations of excellence, is the correct focus. And ¹⁸ On the paradoxes see Santas [1964]. ¹⁹ There has been a lively debate among recent commentators on whether the Socratic claim is the extreme claim that virtue (i.e. knowledge) is sufficient for success in life (i.e. eudaimonia) by itself, or the less ambitious claim that it is sufficient given a (modest) sufficiency of those goods which it is beyond the power of the agent to procure, for instance good health. The former view is maintained by Irwin [1977], 100 and [1986], the latter by Vlastos [1984], and by Brickhouse and Smith [1987]. The issue seems to me not to be addressed very clearly in the dialogues. While the Socratic claim that the good man cannot be harmed (Ap. 30c, 41c–d), if taken literally, implies the former view, that view is clearly inconsistent with the assertion at Crt. 47e that it is not worth living with a sickly body, unless we attribute to Socrates the implausible view that virtue guarantees good health. He might indeed have held that virtue will minimize the causes of ill-health by eliminating those which spring from lack of self-control, but it seems very implausible that he should have thought that such factors as climate or accident could have no effect on one’s health. As Irwin points out ([1986], 92–4), Socrates does argue at Euthyd. 279–80 that wisdom always makes those who possess it eutuchein, so that the wise do not need good luck, but in the context that seems to amount just to the claim that knowledgeable practitioners of any skill (doctors, etc.) are more successful than the unskilled, and that they do not need to rely on luck, as the unskilled do. In general, Socrates’ insistence on the paramount importance of the state of one’s soul (assumed to be within one’s control) and the comparative worthlessness of other reputed goods (Ap. 28b, Prot. 313a) has the outcome that the virtuous agent will need very little from fortune to give him or her a totally worthwhile life, but it is pressing the literal reading of the texts too far to take him to claim that he needs literally nothing.

platonic ethics 161 that focus has to be a correct conception of the good for the agent, i.e. a correct conception of eudaimonia. From this theory it follows immediately, given the conception of excellence accepted in the Meno as what is manifested in excellent conduct, that the traditional virtues are in fact one and the same state of the agent. Courage is that stable state which is manifested in the proper handling of fearful situations, justice the stable state which is manifested in one’s proper dealings with others, piety the stable state which is manifested in one’s proper dealings with the gods. And the stable state in question is the same state in every case, namely the agent’s grasp of the correct conception of the good. Indeed, that same state is designated by the non-synonymous names of the traditional virtues, whose distinct connotations pick out the distinct manifestations of that single state: for example the connotation of the name ‘justice’, i.e. ‘what is manifested in one’s proper dealings with others’, is distinct from the connotation of the name ‘piety’, i.e. ‘what is manifested in one’s proper dealings with the gods’. The point of retaining these various names is to do justice to the fact that for the achievement of overall success in life the same cognitive state has to be manifested in various ways, whether in various types of conduct which overlap only partially (for example if all courageous actions are self-controlled but not vice versa) or in different aspects of coextensive action-types (for example if all and only just actions are pious, their being pious, i.e. appropriate to dealings with the gods, is a specific modification of that attribute of the action-type which is their being just, i.e. appropriate to dealings with others). But the retention of the names of the traditional virtues should not disguise the essence of the theory, that what is manifested in all these different ways is identical, namely the agent’s grasp of his overall good. The ‘parts’ of total excellence are not distinct motive-forces or tendencies to action, as on the traditional conception, which allowed that they might be separable from one another. Nor are they distinct cognitive states, for example kinds of knowledge, as knowledge of history is a distinct type of knowledge from knowledge of geometry; knowledge of how to treat others and knowledge of how to control one’s passions are not distinct types of knowledge, but rather different aspects of a comprehensive knowledge of how to live, which is what controls one’s activity in all areas. That is precisely the account of the virtues which the cognitive theory of the Meno would lead us to expect,

162

platonic ethics

and it is the theory which we find explicitly argued for in the Protagoras and implicit in the accounts of individual virtues in the Euthyphro, Laches, and Charmides.²⁰ The central claim of the Cognitive Theory, as so far elucidated, is not one from which Thrasymachus or Callicles need dissent; that theory claims that a reliable grasp of the good is sufficient for overall success in life, but is of itself neutral between incompatible conceptions of the good. In order to realize Socrates’ aim of vindicating the traditional conception of virtue it has to be supplemented by an account of the good, which will show either that traditional virtue (or a sufficiently close approximation to it) is instrumentally necessary for success in life, or that it is at least partially constitutive of it.²¹ In advance of that account the Cognitive Theory entails the self-interested version of the second Socratic paradox ‘No one acts intentionally against his overall interest.’ The supplementation of the Cognitive Theory with that account yields the moral version of the paradox ‘No one does intentionally what is morally wrong.’ But when we look for arguments in favour of that account, the dialogues provide us with very little. Crt. 47e states, but does not argue for, the analogy between health of the body and justice in the soul, asserting that injustice ²⁰ I discuss the unity of virtue more fully in Taylor [1976 and 1991] and in chapter 4 of this volume. Vlastos has vigorously contested this account of the doctrine in his [1981]. For a reply to his main points see Taylor [1991], 221–4. ²¹ Irwin argues in his [1977] (for example on pp. 84–5) that the craft analogy commits the Socrates of the early (= Socratic) dialogues to an exclusively instrumental view of the relation of virtue to eudaimonia, and treats acceptance or rejection of this instrumental view as the criterion for distinguishing between the views of ‘Socrates’ and those of ‘Plato’. But first, not all crafts (technai) are exclusively instrumental in character. A techn¯e is any skilled activity which can be systematically taught, a description which embraces the performing arts as well as the productive: see, for instance, Symp. 187b (music), Leg. 816a (dancing). The craft analogy itself, therefore, need not prevent Socrates from claiming both that virtue is a techn¯e and that it is at least partly constitutive of eudaimonia, while Crt. 47e shows him firmly committed to the latter claim. So unless Irwin is to attribute the view of the Crito to ‘Plato’ rather than to ‘Socrates’, he must abandon the view that the latter’s conception of the relation of virtue to eudaimonia is exclusively that of an instrumental means. Irwin’s view is criticized by Zeyl in his [1982]. The above is not to say that the position of the early dialogues on the relation of virtue to eudaimonia is consistent. As pointed out in the text (p. 165 below), in different passages Socrates is represented as maintaining both that virtue is knowledge (sc. of the good) and that it is itself the health of the soul, and therefore the good itself. Those theses are inconsistent, since knowledge of the good requires that the content of that knowledge should be independently specifiable, which is impossible if the good of which one has knowledge is the very state of having knowledge of the good. The fault, however, lies not with the craft analogy, which can be stated without inconsistency as the thesis that success in life is a skilled activity analogous to skill in the performing arts, but with the conjunction of the claims that excellence is a state of knowledge of how to achieve a goal and that it is that goal itself.

platonic ethics 163 damages the soul as sickness does the body, and that as it is not worth living with a sickly and diseased body, even less is it worth living with a corrupted soul. On the strength of that analogy it is agreed (48b) that living well (the universally acknowledged good) is the same thing as living creditably (kal¯os) and justly (i.e. that the good life is identical with the moral life), but Thrasymachus could properly point out that no argument has been given for the crucial claims that injustice harms the soul and that justice benefits it. Gorg. 504–5 merely gives a more extended version of the analogy: all types of craftsmen aim to produce a good product, and in every case the goodness of the product consists in its order and arrangement. So a well-made boat has all its parts properly fitted together, and in a healthy individual all the bodily constituents are properly ordered. As health is the name for the proper ordering of bodily components, the name for the proper ordering of psychic components is justice and s¯ophrosun¯e, which expresses itself in proper conduct towards gods and men (piety and justice) and in the proper control of pleasure and pain (courage) (507a–c). Once again, the crucial identification of the proper order of the soul with conventional justice and s¯ophrosun¯e is unargued. It is an appropriate ad hominem rejoinder to Callicles that his ideal of the unrestrained satisfaction of the pleasure of the moment does not provide an adequate rule for the long-term planning of one’s life, and that without some such rule one’s life will collapse into a chaos of conflicting desires.²² To be satisfying one’s life must be coherent, and to be coherent it must contain some discrimination of pleasures into the more and less significant, and some circumscription of the pursuit of the latter with a view to the greater enjoyment of the former. But that element of rational planning and self-control might be exercised in the pursuit of a life of injustice and self-indulgence; Don Juan could satisfy that requirement by making the seduction of as many women as possible his paramount aim, and by refusing to be distracted from it by the momentary attractions of a life of scholarship or of quiet domesticity. ²² Of course, one’s only rule for planning one’s life might be ‘Don’t plan; take every pleasure as it comes.’ Such a rule is not formally self-contradictory, nor is the description of someone as attempting to live by it. It is, however, practically self-defeating, in the sense that anyone who tried to put it into practice would find that the attempt to adhere to it consistently required him to break it. The reason is that some pleasures, as they come, require that one should plan, for example the pleasure of getting started on a career.

164

platonic ethics

Earlier in the Gorgias (473–5) Socrates argues against Polus that injustice is against the agent’s interest, and hence (by the second paradox) that no one acts unjustly intentionally (509e). The argument does not contain any positive account of the agent’s good, but proceeds directly to the conclusion that injustice is bad for the agent, relying on Polus’ admission that injustice, while more advantageous to the agent than justice, is more disgraceful than justice; Socrates gets Polus to agree that if x is more disgraceful than y, then x is either more unpleasant than y or worse (i.e. more disadvantageous) than y, and then concludes that injustice is worse than justice via the claim (agreed by Polus) that injustice is not more unpleasant than justice. The weakness of the argument is obvious; as Callicles points out (482d–e), a tougher-minded opponent (such as Thrasymachus) would not have made the initial concession that injustice is more disgraceful than justice. Further, even given that concession, Socrates’ argument requires the general principle that if action-type a is more disgraceful to the agent than action-type b, it must be either more unpleasant to the agent or worse for the agent, but there is no ground to think that that principle is true; an action-type might be disgraceful to an agent (i.e. such as to bring him into justified disrepute) which was neither unpleasant to that agent nor bad for him, but either unpleasant to or bad for others. The only other early dialogue which contains any account of the good is the Protagoras, where Socrates is represented as arguing for the conclusion that courage is a kind of knowledge from the premiss that the good is pleasure. Without renewing the controversy as to whether this premiss is (as I believe) presented in the dialogue as Socrates’ own view or merely that of the majority of ordinary people,²³ it has once again to be observed ²³ Commentators are divided on the question of whether Socrates is represented as seriously espousing hedonism. I append a list of some writings on either side of the dispute. For the thesis that he is serious see: Grote [1865], vol. 2, pp. 87–9 J. and A. M. Adam [1905], xxix–xxxiii Hackforth [1928] Vlastos [1956], xl (note) Dodds [1959], 21–2 Crombie [1962], vol. 1, pp. 232–45 (with reservations) Irwin [1977], ch. 4 Taylor [1991], 208–9 Gosling and Taylor [1982], ch. 3 Nussbaum [1986], ch. 4 Cronquist [1980]

platonic ethics 165 that it is (a) not itself supported by any argument and (b) insufficient to provide the desired vindication of conventional morality. Taken as a whole, then, the early dialogues fail to realize Plato’s project of providing that vindication. The theory of motivation which is outlined in them requires to be complemented by an account of eudaimonia, but no such account is provided. Instead, the whole theory rests on the analogy between conventional virtue and bodily health, which begs the crucial question of the value of conventional virtue to its possessor. Moreover, that analogy threatens the cognitive account of virtue itself; by that account virtue is knowledge, i.e. knowledge of how to achieve eudaimonia, whereas on the analogy virtue is eudaimonia itself. If the theory were to fit the analogy, knowledge of how to achieve eudaimonia should be analogous to knowledge of how to achieve health, i.e. to medicine rather than to health itself. But then virtue would be of purely instrumental value, whereas the analogy with health represents it as having intrinsic value. A further difficulty is this, that the cognitive account of virtue depends on the thesis that a cognitive grasp of the good is sufficient to motivate the agent to achieve it, a thesis which, though notoriously beset by the counter-evidence of ordinary experience, is defended by nothing more than a single unsound argument (Meno 77–8). When we turn to the Republic we find Plato developing a more elaborate psychology which enables him at the least to make a serious effort to remedy these deficiencies, in that it not only provides the material for the necessary account of eudaimonia but also allows him to abandon the counter-intuitive claims that virtue is knowledge and that it is impossible to act contrary to one’s conception of one’s overall good. Against: Taylor [1926], 260–1 Sullivan [1961] Raven [1965], 44–9 Gulley [1968], 110–18 Vlastos [1969] Manuwald [1975] Kahn [1976] Dyson [1976] Duncan [1978] Zeyl [1980] Stokes [1986], 358–439 Kahn [1988a] and [1988b]. For a reconsideration of this issue see ch. 15 of this volume.

166

platonic ethics

ii The Republic Among the many ways in which the Republic is innovative is its attempt at a comprehensive integration of individual psychology with political theory. That there was some connection was not, of course, a novel idea; naturalistic theories such as those of the Protagoras myth had shown how social institutions including morality would naturally have developed in response to the needs of individuals for protection and cooperation. According to these theories morality was therefore both social, i.e. a set of social norms, and natural, i.e. grounded in individual human needs. We saw, however, that these theories failed to show that the observance of morality by any particular individual is an intrinsic good to that individual. Plato’s innovation in the Republic may be seen as an attempt to bridge the gap between the individual’s good and that of the community by internalizing the social nature of morality, in that the individual personality is itself organized on a social model, and its best state, which is the supreme good for the individual, consists in a certain social organization. However counter-intuitive, the social conception of individual morality is not an arbitrary construction, designed to fill a gap left by earlier theory, but has a firm theoretical base. This may be set out as follows. 1 Key evaluative predicates such as ‘good’, ‘just’, ‘courageous’, and ‘self-controlled’ are applicable to communities as well as to individuals (368e–369a). 2 Any predicate which applies both to an individual and to a community applies to the one in virtue of the same feature or features as those in virtue of which it applies to the other (435a–b). 3 Since the perfectly good individual is wise, self-controlled, courageous and just, by 2 the perfectly good community is wise, self-controlled, courageous, and just (427e). 4 The perfectly good community is just in virtue of the fact that the members of the three functionally defined classes into which it is divided (rulers, military auxiliaries and economic producers) stick to the social function which defines their respective class, and to which they are fitted by their natural abilities, developed by appropriate education (433a–434c).

platonic ethics 167 5 The psychology of every individual comprises a tripartite structure of intellect, self-assertive motivation, and bodily appetite corresponding to the political structure of the perfectly good community (435e–441c). 6 Therefore, by 2, 4, and 5, the perfectly good individual is just in virtue of a relation between the three elements of his or her personality corresponding to that between the classes in the perfectly good community which constitutes the justice of that community (see 4) (443c–444a). In this derivation premiss 2 has a pivotal role, mediating the inferences from the character of the individual to that of the community (step 3) and conversely (step 6). It is an a priori thesis, which applies to the case of the community and the individual²⁴ the Socratic thesis (see above, p. 157) that all the things to which a single predicate ‘F’ applies share a single common nature in virtue of which they are all Fs. Unfortunately for Plato the thesis is false; even leaving aside cases of simple equivocation like ‘pen’ or ‘cape’, a predicate may apply to things of different kinds, not in virtue of the fact that all the things to which it applies share a common nature, but in virtue of the fact that each kind of application relates differently to a central notion.²⁵ Ironically, ‘just’ provides a clear example. A political community is just if either it is internally organized according to just principles, or in its relations with other communities it acts according to just principles. An individual, however, is just only in the latter way, not in the former, since the notion of just principles has no application to the psychological organization of an individual. That is because just principles assign rights and obligations to the individuals composing a community, whereas the elements in an individual’s psychological organization are not themselves individuals, and are therefore not subjects of rights and obligations. Even this very crude sketch shows that the justice of a principle is the primary application of the predicate ‘just’, and that communities and individuals are derivatively just in virtue of different relations to the primary application. Plato’s error is twofold, first in assuming the univocity of the predicate, and ²⁴ The crucial sentence (435a5–7) is ‘So, if things larger and smaller are called the same, are they alike in the respect in which they are called the same, or unlike?’ It is assumed without argument that the community and the individual are examples of ‘things larger and smaller’. ²⁵ As is well known, Aristotle discovered this kind of application. See Meta. 1003a33–b12, EE 1236a16–23: for discussion see Owen [1960].

168

platonic ethics

secondly in applying to the individual member of a community an internal, structural model of justice appropriate to the complex social entity, not to its individual components. The identification in step 4 of adherence to one’s generic social role as the organizational principle of justice in a community may also seem very contrived, but this too has some theoretical backing. First, it depends on step 3, and therefore on premiss 2, on which 3 relies to identify the traditional list of individual virtues (see above, p. 153) as the virtues of the perfect community. It therefore succumbs to the refutation of 2 in the preceding paragraph. But waiving that objection for the sake of argument and granting that the excellence of the perfect community consists in wisdom, courage, self-control, and justice, what grounds the identification of justice with adherence to one’s generic social role? Here Plato appeals to the traditional conception of social justice as each one’s having his own and doing his own, i.e. that each individual should be secure in the possession of what he or she is entitled to and should not encroach on the entitlement of another (433e–434a). Traditionally this expresses an individualistic principle of ownership, but Plato transforms it into a collectivist principle of service to the community; what belongs to one is above all the contribution one makes to the common good, and to be treated unjustly is to be deprived of that contribution, and thereby of the good itself, which can be realized (for all) only if each makes (and a fortiori is allowed to make) his or her specific contribution to it (434a–c). Social justice is thus redefined, via the ‘doing one’s own and having one’s own’ principle, as adherence to optimum social organization. I shall here assume the conclusion for which I have argued elsewhere,²⁶ that the criterion of the optimum social organization is the maximization of eudaimonia, understood as the provision for every member of the community of the conditions either for the realization of eudaimonia or for as close an approximation to it as the limitations of the individual’s psychological capacities allow. Given that account of social justice, it follows by premiss 2 that individual justice is optimal psychological organization, which is that very state of eudaimonia which it is the function of social organization to make as widely available as possible. But the claim that the good for the individual, which it is the aim of social organization to realize, is optimum ²⁶ In Taylor [1986].

platonic ethics 169 psychological organization advances us little beyond the truism that the good is living well and doing well. It makes some advance, locating doing well in the possession of a certain psychological state, rather than in the possession of external goods, but Heraclitus (DK 22 B 119) and Democritus (DK 68 B 170–1) had already said as much. If the theory is to say more, and in particular if it is to provide the vindication of morality which eluded the early dialogues, it must (a) give an account of optimum psychological organization which is both informative and acceptable and (b) justify the claim that that account is an account of individual justice. Since space is lacking for even the most summary account of the psychology of the Republic,²⁷ I must be content with dogmatic statement. There are three principal²⁸ elements in the personality, the intellect, the bodily appetites, above all those for food, drink, and sex, and a loosely defined cluster of motivations which Plato calls ‘spirit’ (thumos or to thumoeides), including anger, shame, ambition, and a sense of honour or self-respect, all of which may be understood as aspects of a fundamental impulse of self-regard and self-assertiveness.²⁹ The intellect is not a purely ratiocinative faculty, but has its own motivations; hence the tripartition is at least in part a distinction between three kinds of motivation, towards intellectual activity, self-realization, and bodily satisfaction respectively. But in addition to providing its own specific motivations the intellect has the function of directing and coordinating the activity of all three kinds of motivation with a view to the realization of the agent’s overall good, since only the intellect is capable of the grasp of the good presupposed by that direction and coordination. There are therefore two ways in which the intellect is supreme in the state of optimal psychological organization. First, all the agent’s specific desires are directed by the intellect with a view to the agent’s overall good, and secondly, that good consists in a life in which the satisfaction of the specific desires of the intellect (i.e. desires for intellectual activity) takes priority over the satisfaction of the other kinds of desire. The satisfaction of intellectual desire should be understood as neither purely theoretical nor (because purely theoretical) exclusively egoistic; according to the metaphysical system of the Republic, the supreme object of understanding is the Form of the good, and someone who grasps ²⁷ See, for instance, Woods [1987]. ²⁸ Rep. 443d7 explicitly leaves open the possibility that there may be others. ²⁹ See Gosling [1973], ch. 3. Cf. Rawls [1963].

170

platonic ethics

what goodness itself is is thereby motivated to realize it not merely in his own life but (by the theory of love of the Symposium and Phaedrus) in the lives of those he loves and in the community of which he is a member. A central feature of this psychology is Plato’s abandonment of the theory of uniform motivation which was presupposed in the early dialogues, and with it the Cognitive Theory of Excellence and the strong version of the unity of virtue which that theory implied. Plato no longer accepts either the strong thesis that every intentional action is aimed at the realization of the agent’s conception of his or her overall good nor the weaker thesis that whenever that conception is present it motivates, since the bodily and ‘spirited’ appetites motivate independently of, and even in opposition to, the conception of the overall good. Hence what makes the difference between the virtuous and the non-virtuous agent is not simply the possession by the former of a cognitive state which the latter lacks, but the possession by the former of a psychological structure lacking in the latter, in which the specific desires are appropriately responsive to the direction of the intellect. And that responsiveness is not guaranteed by the content of the intellect’s direction, but requires that the specific desires should have been conditioned by the process of education described in books II–III to respond instinctively to the guidance of the intellect by loving what the intellect reveals as good and hating what it reveals as bad. Since what is manifested in the various types of virtuous conduct is no longer a single cognitive state, the former version of the unity thesis has to go. But since by the new theory it is the same psychological structure which is manifested in those types of conduct, it might seem that the strong theory could survive the shift from cognitive state to total structure, with ‘courage’, ‘wisdom’, etc. functioning as non-synonymous names of that total structure. In fact the shift is more substantial, for two reasons. First, it seems that those names apply, not to the same total structure under different aspects, but to different aspects of the structure; thus self-control is ‘a certain ordering and mastery of pleasures and appetites’ (430e) while justice is the state in which each element plays its proper part in the optimal structure, wisdom is the care of the intellect for the whole (441d–e), and courage is the retention by the spirit, despite pleasures and pains, of the instructions of the intellect about what is to be feared and what is not (442c). While justice and self-control are hard to distinguish on this account (except that the terms are non-synonymous), courage and wisdom are not the total structure

platonic ethics 171 itself but aspects of it, ascribed in the first instance to particular elements of that structure. Secondly, the possibility of disorderly appetite indicates that while self-control and justice are not only mutually necessitating but severally impossible without wisdom, the possession of wisdom does not guarantee self-control. It appears, then, that the psychology of the Republic requires the abandonment of the unity of virtue doctrine, in that even the weakest form of that doctrine, the thesis that anyone who possesses any of the virtues necessarily possesses them all, has to be replaced by a still weaker thesis, namely that courage, self-control, and justice all require wisdom (and perhaps require one another also), but wisdom does not guarantee the presence of the other virtues. Does Plato give us adequate reason to accept that this psychological organization is in fact optimal? To do so he has to show that each of the elements in the psychological structure functions at its best when coordinated by the intellect so as to make the appropriate contribution to a life where the highest priority is given to the pursuit of intellectual satisfactions. We may accept on the basis of the argument against Callicles (see above) that a satisfactory life is possible for the individual only if his or her potentially conflicting motivations are intentionally coordinated, which is a rational process requiring the identification of priorities and long-term goals. But we need further argument to show that the supreme long-term goal of the optimal life must be the theoretical understanding of reality (of which the primary object of understanding is goodness), and the realization of that understanding in practical, including political, life. Plato attempts to meet this challenge by arguing in book IX that the life devoted to those goals is the pleasantest life possible, but his arguments are unsatisfactory. He gives two main arguments: the first (famously recalled by Mill in Utilitarianism) is that the devotee of intellectual pleasures is the appropriate judge of which life is the pleasantest, since he has experience of the pleasures of appetite and ambition (the dominant goals of the rival lives), whereas the adherents of those pleasures lack experience of intellectual pleasures. This argument fails because experience of the pleasures of a life requires commitment to the activities and the values constituting that life, but the intellectual is as remote from immersion in the rival lives as the rivals are from the intellectual life. The second argument depends on the conception of pleasure as the making good of a deficiency in the organism (for example hunger is a state of bodily depletion and ignorance a state of intellectual

172

platonic ethics

depletion, and the pleasures of satisfying hunger and of discovery are the processes of making good the respective depletions). Plato uses this model to make two basic points, whose relation to one another is obscure; the first is that since the state of bodily depletion is painful, what we think of as bodily pleasures are in fact mostly episodes of getting rid of pain, not genuine pleasure. The second is that whereas bodily deficiencies cannot be properly or genuinely made good, intellectual deficiencies can be. Leaving aside the question of the adequacy of the depletion model in general, its application in this argument is highly obscure, since it is unclear whether the inferiority of bodily pleasures is supposed to lie in the fact that they require to be repeated (since, for instance, one gets hungry again a few hours after having eaten) or in the alleged insatiability of the desires which they presuppose (so that one can never get enough food) or simply in the alleged confusion between bodily pleasure and the getting rid of bodily distress. If the point is the latter, Plato’s diagnosis of the alleged confusion is itself confused, since the depletion model yields the result that the process of making good the depletion is pleasant, irrespective of whether the depletion is painful. Hence whenever the depletion is painful, getting rid of that painful lack will be genuinely pleasant, and there will be no confusion of genuine pleasure with something else, namely the getting rid of pain. If, on the other hand, the target is the insatiability of bodily desires, the alleged fact should be denied; normal bodily desires are not insatiable, unless ‘insatiable’ is reinterpreted as ‘recurrent’, in which case the point is after all the first, i.e. that bodily desires are recurrent whereas intellectual desires are not. But with respect to that point it is not clear that the need for recurrent satisfaction differentiates a life devoted to bodily satisfactions from one devoted to intellectual; no doubt a truth once discovered does not have to be rediscovered, but a meal once eaten does not have to be eaten over again, and an intellectual life will require repeated acts of thought (whether new discoveries or the recapitulation of truths already known) no less than a life of bodily satisfactions will require repeated episodes of bodily pleasure. (The point also applies to the pleasures of the life of ambition, which is for the purposes of the argument required to share the defects of the life of bodily pleasure, but which is in fact barely mentioned.)³⁰ ³⁰ For fuller discussion of the arguments of Rep. IX see Annas [1981], ch. 12; Gosling and Taylor [1982], chs. 6 and 17.2; Stokes [1990].

platonic ethics 173 Plato does not, then, succeed in establishing the optimality of his preferred psychological structure. Does he fare any better in showing that that structure captures the nature of justice as a virtue of the agent? Since his account of justice is avowedly revisionary, he cannot be held to the requirement to show that the presence of that structure in an agent is necessary and sufficient for that person’s being just by ordinary standards. Someone who conforms consistently to ordinary morality as it is depicted by Glaucon³¹ is just by ordinary standards but not by Plato’s, since his commitment to justice is conditioned, not by acceptance of the value of justice for itself, but by the belief that he could not succeed in doing what he would really like to do, viz. to promote his own interests at the expense of others. The most that can be required is that Plato should show that the presence of the structure is sufficient for the performance of a sufficient range of central cases of just dealing and the corresponding avoidance of unjust dealing. At 442e–443a Socrates asserts that the Platonically just agent will never commit any major crime such as theft, sacrilege, treason, or adultery, and while he gives no argument we may concede that the control of appetite which characterizes such an agent will make him proof against the standard temptations to such wickedness. Moreover, as we saw, his love for the good will make him concerned for the good of others and for that of the community. How, then, could he fail to be just by conventional standards?³² The flaw in the theory is that the structure itself defines the good for the agent; hence concern for the good of others and for that of the community is concern to maintain that structure in others and the corresponding structure in the community. Consequently just actions are redefined as whatever actions serve to create and preserve that structure, and unjust as whatever destroy it (443e). That redefinition clearly licenses substantial interference with the autonomy of others, with a view to the promotion of their own good (as redefined) or the good of the community; indeed, if the setting up of the ideal Platonic state required extermination and enslavement of whole populations, by this account such acts would be just ³¹ This assumes that ordinary morality is not so confused as to make consistent adherence to it impossible. It is not clear that Plato would accept that charitable assumption; the arguments against the accounts of justice proposed in Republic I by Cephalus and Polemarchus suggest that he thinks that ordinary moral beliefs are thoroughly confused. ³² See Sachs [1963]; Demos [1964]; Vlastos [1971b]; Irwin [1977], ch. 7, sects. 10–11 and 19.

174

platonic ethics

(not merely permitted, but required). This is explicitly acknowledged in the text: at 540e–541a Socrates recommends the foundation of the ideal state by the process of expelling the whole population of an existing city over the age of ten, taking the children from their parents, and bringing them up in the educational system which he has just described. Plainly the forced evacuation of an entire city and the enforced separation of a complete generation are acts of extreme violence, which could not in practice be perpetrated without considerable loss of life. As regards loss of autonomy, the subjection of the lowest class in the ideal state is complete enough for their state to be described as one of slavery (590c–d); it is, of course, paternalistic slavery, since it is better for the lower classes to be enslaved to those who have their good at heart and who know what that good is than to be enslaved to their own lower nature and to mistaken conceptions of what is good for them. Yet it is slavery nonetheless, since in the last resort the direction of their lives rests not with their own intellect and will, but with those of the rulers. The Platonically just agent will not, therefore, be unjust from the vulgar motives of private gain or personal lust, but he will not be just for all that. The form of injustice to which he will be prone is something much more terrible, the enforcement of an ideology which, in virtue of its comprehensiveness and its redefinition of benevolence, admits no limitation in the name of individual liberty, and is therefore liable to press its claims to the extremes of tyranny.³³

iii Developments subsequent to the Republic The ethical theory of the Republic represents Plato’s most sustained attempt to vindicate the claims of morality. Subsequent developments in his ethical ³³ Plato is quite explicit in drawing the implication at Pol. 293c–d: [T]he only constitution worthy of the name ... must be the one in which the rulers are ... men really possessed of the scientific understanding of the art of government. Then we must not take into consideration whether their rule be by laws or without them over willing or unwilling subjects or whether they themselves be rich or poor men. No. They may purge the city for its better health by putting some of the citizens to death or banishing others. (tr. Skemp from [1952]) A similar view is expressed at Leg. 735b–736c.

platonic ethics 175 thought narrowly conceived (i.e. as distinct from political theory, which requires separate treatment) amount to modifications of detail, not to any radical shift of view. They may therefore be dealt with briefly. The second of the two stages which we have distinguished in the development of Plato’s ethical thought was marked off from the earlier by the abandonment of the ‘Socratic’ theory of uniform motivation and hence by the rejection of the Cognitive Theory of Excellence and the thesis of the unity of virtue. In the dialogues subsequent to the Republic the nonuniform character of motivation is even more strongly emphasized, while in the Statesman he insists on the disunity of virtue (see below, pp. 177–8). These developments may be attributed in part to an increasing sense of the dichotomy, already present but not dominant in the Republic (611a–d), between the rational element in the personality, motivated by the good and only contingently and temporarily embodied, and the non-rational spirit and appetites, which spring from the body and are motivated independently of the good.³⁴ To the extent that the reason is identified with the real self, spirit and appetite come to be seen rather as alien forces requiring to be kept in subjection by reason than as manifestations of a uniformly rational agency. This allows Plato, despite having abandoned the thesis that knowledge of what is best is sufficient for doing what is best, to continue to maintain the second of the Socratic paradoxes, that no one voluntarily acts wrongly. Contrary to the Socratic position, action against one’s better judgement is possible; but such action is not voluntary, since the agent (= the rational self) is overwhelmed by external forces, i.e. the non-rational passions (Tim. 86e, Leg. 734b, 860–3). On the Socratic model, all purported cases of action against one’s better judgement had to be explained as cases of intellectual error; Plato now recognizes that some cases have to be explained by a mismatch between intellectual judgement and passion, but saves the doctrine by counting such mismatches as sources of involuntary ³⁴ In the Phaedo the appetitive and spirited motivations are attributed to the body (66b–c, 68b–c), which is sharply distinguished from the (rational) soul, whose task is to master and control the body (80a). The rational soul is implicitly identified with the self: the survival of the self is the survival of the soul, and the task of the philosopher is to prepare for what is ordinarily thought of as death, but which is in fact the fullness of life, free from the distractions of the body (64c–69e). While it might seem obvious that on this model also the motivations of the body are independent of the good, it is not clear that that is so, since the body is also described as a source of illusions, i.e. false beliefs (81b, 83d). Hence the spurious morality of the non-philosopher, guided by a calculation of bodily pleasures (68b–69c), is, although undoubtedly a state of enslavement of the soul to the body, nevertheless not incompatible with the Socratic theory of motivation.

176

platonic ethics

action.³⁵ That saving move has its cost, in that it breaks up the unity of the agent, but Plato’s dichotomy of rational and non-rational elements in the personality, itself a reflection of the more fundamental dichotomy of rationally apprehended reality and the imperfectly rational material world, already encourages that split. The split is, however, fatal to the project of vindicating morality by showing it to be constitutive of the best life for the agent. For morality concerns the embodied agent, whereas the best life for the agent is the discarnate life of pure thought (Tim. 90). The most that Plato could hope to provide by way of vindication of morality would therefore be the claim that immorality as ordinarily conceived hinders the achievement of that life (Tim. 90). This fundamental dichotomy tends to drive Plato further from the doctrine of the unity of virtue: if wisdom is in the last analysis the activity of the immortal ‘real self’ and courage, self-control and justice different aspects of the subjection of the mortal and non-rational by that self, then the connection between the virtues has been loosened to the extent that, so far from being in any sense the same virtue, they are no longer even virtues of literally the same subject. Rather, the perfection of the real self requires, contingently and temporarily, the cooperation of the mortal self. This cooperation consists in the exercise of the four cardinal virtues, wisdom, s¯ophrosun¯e, justice, and courage (Leg. 631c), as is brought out in a later passage (653a–c) describing the general aim of education. This is essentially that of the primary education of the Republic, so to train the motivational impulses by means of the basic stimuli of pleasure and pain that the child comes to like what reason dictates and to hate what reason forbids. The affective responses are formed before the rational judgements, but when the child is mature enough to form those judgements they agree with the content of the affective responses ‘and this agreement as a whole is aret¯e’ (653b6). This formula may plausibly suggest an account of what courage, s¯ophrosun¯e, and justice have in common; each is an agreement between a specific motivational impulse and a rational judgement,³⁶ or (perhaps closer to the theory of the Republic) each is an aspect of a state of agreement ³⁵ R. M. Hare makes the same move in [1963], ch. 5 (modified in [1981], 23–4, 58–60). ³⁶ Compare Aristotle’s account of the conditions of correct choice in EN 1139a22–6: ‘since excellence of character is a disposition concerned with choice, and choice is deliberative desire, for this reason the judgement (logos) must be true and the desire right, if the choice is to be good, and the one must say the same thing as the other pursues’.

platonic ethics 177 between the agent’s motivations taken globally and the deliverances of his or her reason. There is, however, still the problem of what wisdom itself has in common with those virtues whose essence consists in agreement between the non-rational and wisdom. It may be a sense of this difficulty which prompts Plato to say at the end of the work (963) that while it is not hard to see how the virtues differ from one another, it is a real problem to explain how states as different as courage and wisdom are one, i.e. to determine what they have in common; this problem he leaves unanswered. I suggest that this problem reflects Plato’s difficulties in fitting his account of virtue to his sense of the dichotomy between the mortal and immortal elements in the soul. That same difficulty is also, I think, reflected in another feature of the treatment of the virtues in the later dialogues, namely that Plato sometimes reverts to the earlier ‘Presocratic’ tradition of treating them as separable components of excellence, as when he asserts that the state in which justice, wisdom (phron¯esis), and s¯ophrosun¯e are ‘unified together with courage’ is better than courage alone (630a–b), citing as an example of the latter, as Protagoras had done in opposition to Socrates (Prot. 349d), the courage of wicked and licentious soldiers (cf. 696b–e). The conception of courage as a non-rational impulse combating fear, to be found even in animals (cf. Lach. 197a–b), recurs at Leg. 963e and at Statesman 306. In the latter passage the Eleatic Stranger goes out of his way to emphasize the unhomogeneity of the specific virtues, by first stating the conventional view of courage and s¯ophrosun¯e as parts of total excellence, and then urging the ‘unfamiliar’ thesis (306b13) that they are hostile and opposed to each other, in the sense that courage, understood as an aggressive impulse, is opposed to s¯ophrosun¯e, understood as an impulse to quiet and unassertive behaviour. It is the task of the statesman to devise forms of education and political institutions which will harmonize these opposed impulses for the benefit of the individual and the community. Here once again we see Plato apparently reverting to conceptions of the specific virtues which were rejected in the early dialogues, courage as aggressiveness in Laches (197a–b), s¯ophrosun¯e as quietness in Charmides (159b–160b).³⁷ In so doing ³⁷ There is even a trace of these conceptions in Meno 88a–c, where Socrates, arguing that knowledge is the only unconditional good, includes the virtues of s¯ophrosun¯e, courage, and justice (as well as learning, good memory, personal splendour, ‘and everything of that kind’) among the things which are good only on condition that they are directed by knowledge (i.e. directed aright), but which are harmful if misdirected. But in fact this passage seems to draw the very distinction between motivational

178

platonic ethics

he ignores the crucial distinction between motivational drives and the proper organization of those in the integrated personality which was one of the main achievements of his mature theory, and thereby generates a spurious paradox. Aggressiveness may indeed be opposed to quietness, but courage is not opposed to s¯ophrosun¯e, since both are aspects of a structure of motivations organized under the direction of the intellect. I confess to being puzzled as to why Plato should have blurred this distinction (which is central to Aristotle’s theory of virtue as well as to Plato’s own) in his later writings; all that I can suggest is that his sense of the discontinuity between the rational and immortal elements in the personality on the one hand and the non-rational and mortal on the other may have made him uncomfortable with accounts of the specific virtues as modifications of the latter by the directive activity of the former. According to the theory of the Republic, the specific virtues, while no longer a single virtue as in the early dialogues, are still virtues of a single subject; yet the argument of Rep. X. 611b–d that since the soul is immortal and what is immortal cannot be composite the tripartite soul is not the true soul leads to the conclusion that the tripartite soul is not a genuine unity, but rather an adventitious agglomerate of disparate elements, like the sea-god Glaucus overgrown with shells and weed. If Plato takes that conclusion seriously, then there is no single subject for all the specific virtues; wisdom is an attribute of the immortal soul, and the other virtues attributes of the mortal elements, which may have made it easier to revert sometimes to the traditional view of them as non-rational impulses, requiring the direction of reason to attain the status of true virtues. Of course, one need not conceive drive and virtue proper which is blurred in the later dialogues. For Socrates asks Meno whether, if it is not the case that the virtues are sometimes beneficial and sometimes harmful, they can be anything other than knowledge. For instance, if courage is not intelligence (phron¯esis) but a sort of boldness, is it not the case that when someone is bold without thought (nous) he is harmed, but when he is bold with thought he is benefited? After citing the examples of s¯ophrosun¯e and learning Socrates concludes as follows (c1–5): So to sum up, all the undertakings and endurances of the soul result in eudaimonia if they are directed by intelligence, but in the opposite if they are directed by folly. So it seems. If excellence, then, is one of the things in the soul and it is necessary that it should be beneficial, it must be intelligence ... ‘Undertakings and endurances of the soul’, i.e. the acts which one is prompted to by motivational drives, are clearly distinguished from excellence, which is identified as knowledge. It follows that courage, for example, which is acknowledged to be a part of excellence (see above), is not itself a motivational drive such as boldness, but is itself knowledge.

platonic ethics 179 of them in that way, since the dualistic conception of the soul still allows the alternative (closer to the Aristotelian view) that, for example, ‘courage’ is not the name of a non-rational impulse requiring to be modified by reason, but the name of the state of having that impulse properly modified by reason, which seems to be the view which predominates in the Laws. There seems, then, to be evidence of some vacillation in the conception of the virtues in the later dialogues, which may perhaps be explained by the increased influence in this period of the dichotomy between the rational and non-rational in Plato’s view of the soul. The above account of the ‘moral’ virtues as consisting in agreement between affective responses and rational judgement, produced by the prerational training of the non-rational elements in the personality, has close and obvious affinities with Aristotle’s theory, on which it was doubtless an influence. Another similarity between Plato’s later theory and Aristotle’s is found in the section of the Laws (660e–663d) where the Athenian Stranger discusses the requirement in an adequate code of legislation to show that the good man will be eudaim¯on and the wicked wretched. Since the point of this provision in a code of legislation is to motivate people to obey the law, it is assumed, in line with the generally hedonistic account of motivation which is taken for granted in the Laws (see, for example, 636d–e) that the appropriate way to show that the good agent will be eudaim¯on is to show that his or her life will be pleasant, and the rival lives unpleasant. The strategy is that of Rep. IX, but the arguments have none of the metaphysical elaboration of those employed there. Instead, the Stranger argues simply that everyone sees his or her own preferred life as pleasantest from its own perspective, and that the correct perspective from which the assessment should be made is that of the virtuous agent (663b–d). No attempt is made to support this principle; Plato does not, for instance, employ the analogy, to which Aristotle sometimes appeals (see, for instance, EN 1173b22–5, 1176a8–22), with the perception of the healthy as the criterion of correctness in judgements of sensible qualities such as colour or taste.³⁸ The central theme of this chapter has been Plato’s attempt to anchor morality on the ‘natural’ side of the nature–convention dichotomy by ³⁸ For fuller discussion see ch. 6 of this volume.

180

platonic ethics

grounding it in an adequate theory of human nature. That theory develops from the optimistic oversimplification of the early dialogues to the more complex psychology of the middle and later periods, leading to a picture of the virtuous agent as one who achieves through intellectual and emotional training the right fit between intellectual judgement and affective response. That picture of the ideally developed agent was perhaps Plato’s most important legacy to his successors, from Aristotle through the Stoics to Christian theorists and their post-Christian followers. But the theory underlying the picture faced the problems (a) of giving an account of the element of intellectual judgement which would justify its claim to truth and (b) of defending the integrity of the personality against the threat of dualism, in which the intellect arrogates value to itself, and the affective elements in the personality are correspondingly devalued along with the body. Plato himself solved neither problem: both are visible in Aristotle, Stoicism, and Christianity, while at least the former remains to trouble their ‘realist’ successors. Afterword For valuable discussions of Plato’s ethics and moral psychology see Cooper [1999], especially chs. 3–5. pp. 152–5. For a more optimistic assessment of the capacity of the myth in the Protagoras to provide a justification of morality see Taylor [2007].

11 The Atomists Atomism was the creation of two thinkers of the fifth century bc, Leucippus and Democritus. The former, attested by Aristotle, our primary source, as the founder of the theory, was a shadowy figure even in antiquity, being eclipsed by his more celebrated successor Democritus to such an extent that the theory came to be generally regarded as the work of the latter. Epicurus, who developed and popularized atomism in the late fourth and early third centuries bc (following in the tradition of various figures such as Nausiphanes and Anaxarchus, now little more than names), went so far as to deny that Leucippus ever existed. Only a little more is known about Democritus. The precise relation between Leucippus and Democritus is unclear. Plato never mentions either by name. Aristotle and his followers treat Leucippus as the founder of the theory, but also assign its basic principles to both Leucippus and Democritus; later sources tend to treat the theory as the work of Democritus alone. While it is clear that the theory originated with Leucippus, it is possible that the two collaborated to some extent and almost certain that Democritus developed the theory in a number of areas, for example, extending it to include a materialistic psychology, a sophisticated epistemology, and an account of the development of human society that laid particular stress on the human capacity to learn from chance experience.¹

Physical Principles According to Aristotle (GC 324a35–325a31), the atomists attempted to reconcile the observable data of plurality, motion, and change with the A version of this chapter had already appeared as part of the chapter ‘Anaxagoras and the Atomists’ in Taylor [1997], and material from it also appears in Taylor [1999]. Permission from the respective publishers to reprint that material is gratefully acknowledged. ¹ For Democritus’ poetics, which falls outside the scope of this chapter, see Most [1999].

182

the atomists

Eleatic denial of the possibility of coming to be or ceasing to be. Like Anaxagoras and Empedocles, they postulated unchangeable primary things, and explained apparent generation and corruption by the coming together and separation of those things. But their conceptions of the primary things and processes differed radically from those of Anaxagoras and Empedocles. For Anaxagoras the primary things were observable stuffs and properties, and for Empedocles they were the elements, earth, air, fire, and water: for both, the primary processes were mixing and separation of those primary things. By contrast, for the atomists the primary things were not properties or stuffs but physical individuals, and the primary processes were not mixing and separation but the formation and dissolution of aggregates of those individuals. Again, the basic individuals were unobservable, in contrast with the observable stuffs of Anaxagoras and the observable elements of Empedocles. Consequently, their properties could not be observed but had to be assigned to those individuals by theory. Since the theory had to account for an assumed infinity of phenomena, it assumed an infinite number of basic individuals, while postulating as few explanatory properties as possible, specifically shape, size, spatial ordering, and orientation within a given ordering.² All observable bodies are aggregates of basic individuals, which must therefore be too small to be perceived.³ These basic corpuscles are physically indivisible (atomon, literally uncuttable), not merely in fact but in principle; Aristotle reports (GC 316a14–b7) an (unsound) atomistic argument, which has some affinities with one of Zeno’s arguments against plurality (DK 29 B 2), that if (as e.g. Anaxagoras maintained) it were theoretically possible to divide a material thing ad infinitum, the division must reduce the thing to nothing. This argument was supported by another for the same conclusion; atoms are theoretically indivisible because they contain no void. On this conception bodies can split only along their interstices; hence, where there are no interstices, as in an atom, no splitting is possible. (The same principle probably accounted for the immunity of the atoms to other kinds of ² To adapt Aristotle’s example (Meta. 985b18–19) AN differs from NA in ordering, and AN from AZ in orientation within a given ordering. ³ While most of the ancient sources agree that atoms are too small to be perceptible, some late sources indicate that some atoms are very large (even on one account ‘as big as a world’). It seems to me most likely that the atomists held that, while there are atoms of all possible sizes (for the same reason that there are atoms of all possible shapes), all the atoms in our world are too small to be perceived. See Barnes [1979], ch. 17 (b).

the atomists 183 change, such as reshaping, compression, and expansion. All were probably assumed to require displacement of matter within an atom, which is impossible without any gaps to receive the displaced matter.) It is tempting to connect the assumption that bodies can split only along their interstices with the Principle of Sufficient Reason, to which the atomists appealed as a fundamental principle of explanation—arguing, for instance that the number of atomic shapes must be infinite, because there is no more reason for an atom to have one shape than another (Simplicius, In Phys. 28.9–10).⁴ Given the total homogeneity of an atom, they may have thought, there could be no reason why it should split at any point, or in any direction, rather than any other. Hence by the Principle of Sufficient Reason, it could not split at all. The programme of reconciling the data of perception with the demands of Eleatic theory led the atomists to posit a void or empty space (a) as that which separates atoms from one another and (b) as that in which they move. Parmenides had argued (DK 28 B 8.22–5) that there could not be many things if there were no void to separate them, and Melissus had argued (DK 30 B 7) that there could be no motion without a void into which the moving object moves; Aristotle attests that the atomists accepted both theses (Phys. 213a32–4, GC 325a27–8). To the question what it is that separates atoms from one another, and into which they move, their answer was simply ‘nothing’, ‘what is not’, or ‘the empty’, which they appear to have treated as interchangeable terms. They did not, then, shrink from the conclusion that what is no more is than what is not (Aristotle, Meta. 985b8; Plutarch, Col. 1108f).⁵ But the assertion that what separates distinct objects is nothing leads straight to incoherence; either there is nothing which separates those objects, in which case they are not separate from one another, or there is something which separates them, in which case ‘nothing’ is the name of something. We have no idea whether this challenge was actually put to the atomists, or if it were, how they might have met it. The most we can offer is the following suggestion of an appropriate defence. There is indeed something ⁴ For a full discussion of the atomists’ use of this principle, see Makin [1993]. ⁵ Plutarch states this maxim in what is presumably the atomists’ own terminology: ‘The thing is no more than the no-thing’, where ‘thing’ represents the word den, an artificial formation specifically coined to contrast with m¯eden ‘nothing’, itself etymologically equivalent to m¯ed’ hen ‘not one [sc. thing]’.

184

the atomists

which separates any two non-adjacent atoms, namely an interval. But an interval is not any kind of thing: it is merely a gap, an absence of anything. So there are indeed gaps between atoms, but gaps are nothings, and when an atom moves, it moves into a gap. But that can hardly be the whole story. For the notion of an interval or gap between objects presupposes a continuous dimension in which the objects and the interval between them are alike situated. That is to say, the atomists’ conception of the void cannot have been merely that of the non-being of a physical object; it was at least that of a gap in space, where space is conceived, however inchoately, as a continuous dimension. The atomists also claimed that the void is infinite in extent and used the term ‘the infinite’ as another designation of it; this is most naturally interpreted as the claim that empty space is infinite in extent. They believed, then, that the universe consists of an infinitely large collection of indivisible physical objects (atoms) moving in infinite space, where space is a three-dimensional continuum of which any part may be either occupied or unoccupied.⁶ In this empty space the atoms are in a state of eternal motion. This motion is not the product of design, but is determined by an infinite series of prior atomic interactions⁷ (whence two of Aristotle’s principle criticisms of Democritus, that he eliminated final causation (GA 789b2–3) and made all atomic motion ‘unnatural’ (De caelo 300b8–16).⁸ The theoretical role of the void in accounting for the separation of atoms from one another has an interesting implication that is recorded by Philoponus (In Phys. 494.19–25, In GC 158.26–159.7). Since atoms are separated from one another by the void, they can never strictly speaking come into contact with one another. For if they did, even momentarily, there would be nothing separating them from one another. But then they would be as inseparable from one another as the inseparable parts of a single atom, whose indivisibility is attributed to the lack of void in it (see above); indeed, the two former atoms would now be parts of a single larger atom. But, the atomists ⁶ For a fuller discussion, see Sedley [1982]. ⁷ On the nature of these, see below, pp. 188–9. ⁸ In Aristotle’s system natural motion is motion that is intrinsic to the nature of a thing of a certain kind; for example, it is natural for a stone to move downwards, that is, to fall to the earth when unsupported. Things may also be caused, by the exercise of external force, to move in ways contrary to their natural motion; for example, a stone may be thrown upwards. The atomists’ thesis that all atomic motion is the product of precedent atomic interaction is thus in Aristotle’s terms equivalent to the thesis that all atomic motion is unnatural, a claim that he held to be incoherent (since the concept of unnatural motion presupposes that of natural motion).

the atomists 185 held, it is impossible that two things should become one. Holding atomic fusion to be theoretically impossible, and taking it that any case of contact between atoms would be a case of fusion (since only the intervening void prevents fusion), they perhaps drew the conclusion that contact itself is theoretically impossible.⁹ Hence what appears to be impact is in fact action at an extremely short distance. Rather than actually banging into one another, atoms have to be conceived as repelling one another by some sort of force transmitted through the void. Again, though no source directly attests this, the interlocking of atoms, which is the fundamental principle of the formation of aggregates, is not strictly speaking interlocking, since the principle of no contact between atoms forbids interlocking as much as impact. Just as impact has to be reconstrued as something similar to magnetic repulsion, so interlocking has to be reconstrued as quasi-magnetic attraction. If this suggestion is correct (and it is fair to point out that no ancient source other than Philoponus supports it) it is a striking fact that, whereas the post-Renaissance corpuscular philosophy that developed from Greek atomism tended to take the impossibility of action at a distance as an axiom, the original form of the theory contained the a priori thesis that all action is action at a distance. Consequently that impact, so far from giving us our most fundamental conception of physical interaction, is itself a mere appearance that disappears from the world when the description of reality is pursued with full rigour.¹⁰

Chance and Necessity While the broad outlines of the views of the atomists on these topics can be fairly readily reconstructed, there is much obscurity about the details. The atomists’ universe is purposeless, mechanistic, and deterministic; every ⁹ See Kline and Matheson [1987] and Godfrey [1990]. Bodn´ar [1998] argues (at 49–53) that, rather than providing evidence for the actual views of the atomists, the texts of Philoponus are mere guesses prompted by his interpretation of the Aristotelian texts on which he is commenting. ¹⁰ Restrictions of space preclude discussion of various questions about the nature of atoms that have been the subject of much scholarly dispute. The vexed question of whether atoms have weight is discussed by numerous writers, most fully by O’Brien [1981], with cogent criticism by Furley [1983]. On the question of whether, and in what sense, atoms may be said to have parts, see for example Barnes [1979], ch. 17 (c), and Furley [1967], ch. 6 and [1987], ch. 9, sections 3–4. I discuss these matters in Taylor [1999], 179–84 (weight) and 164–71 (parts).

186

the atomists

event has a cause, and causes necessitate their effects.¹¹ Broadly speaking the process is mechanical; ultimately, everything in the world happens as a result of atomic interaction. The process of atomic interaction has neither beginning nor end, and any particular stage of that process is causally necessitated by a preceding stage. But exactly how the atomists saw the process as operating is obscure. This obscurity is largely attributable to the fragmentary nature of the evidence that we possess, but perhaps the statement of the theory itself was not altogether free from obscurity. The fundamental text is the single fragment of Leucippus (DK 67 B 1) ‘Nothing happens at random, but everything from reason and by necessity.’ The denial that anything happens ‘at random’ (mat¯en) might well be taken in isolation to amount to an assertion that all natural events are purposive, since the adverb and its cognates frequently have the sense ‘in vain’ (i.e. not in accordance with one’s purpose) or ‘pointlessly’. If that were the sense of mat¯en then ‘from reason’ (ek logou) would most naturally be understood as ‘for a purpose’. These renderings are, however, very unlikely. The majority of the sources follow Aristotle (GA 789b2–3) in asserting that Democritus denied purposiveness in the natural world, explaining everything by mechanistic ‘necessity’. A reading of Leucippus which has him assert, not merely (contra Democritus) that some, but that all natural events are purposive, posits a dislocation between the fundamental world-views of the two of such magnitude that we should expect it to have left some trace in the tradition. Moreover, the attribution of all events to necessity, a central feature of the mechanistic Democritean world-view, is itself attested in the fragment of Leucippus. We ought, then, to look for an interpretation of the fragment that allows it to be consistent with Democritus’ denial of final causation. Such an interpretation is available without forcing the texts. Sometimes (e.g. Herodotus VII. 103. 2; Plato Tht. 189d) mat¯en is to be rendered not as ‘without purpose’ but as ‘without reason’ (‘in vain’ and ‘empty’ have similar ranges of application). Given that construal of mat¯en ‘from reason’ is to be construed as ‘for a reason’, where the conception of reason is linked to that of rational explanation. The first part of the fragment (‘Nothing happens at random, but everything from reason’) thus asserts, ¹¹ On the absence of explicit evidence for the early Greek philosophers’ reflection on causal explanation, see Vegetti [1999].

the atomists 187 not universal purposiveness in nature, but a principle that we have already seen to be pervasive in atomism, the Principle of Sufficient Reason. Instead of a radical discontinuity between Leucippus and Democritus, the fragment, thus construed, attests commitment to a principle basic to atomism. The second half (‘and by necessity’) makes a stronger claim, which links the notion of rational explanation to the notions of necessity and of cause. The stronger claim is that whatever happens has to happen, cannot but happen. This amounts to a specification of the reason whose existence is asserted in the first half of the sentence; nothing happens without a reason, and, in the case of everything that happens, the reason for which it happens is that it has to happen.¹² There are, therefore, no chance events, that is, no events which simply happen. On the other hand, we have evidence that the atomists assigned some role to chance in the causation of events, though precisely what role is not easy to determine. Aristotle (Phys. 196a24–8), Simplicius (In Phys. 327.24–6, 330.14–20), and Themistius (In Phys. 49. 13–16) all say that Democritus attributed the formation of every primal cosmic swirl¹³ to chance (indeed Aristotle finds a special absurdity in the theory that while events in a cosmos occur in regular causal sequences, the cosmos itself comes into being purely by chance). That might be thought to be confirmed by the statement in Diogenes Laertius’ summary of Democritus’ cosmology that he identified the cosmic swirl itself with necessity (IX. 45). On this interpretation, the statement that everything happens by necessity is confined to events within a cosmos and states that all such events are determined by the atomic motions constituting the swirl. The swirl itself, however, is not determined by anything; it just happens. On this view necessity governs, but is local to, a world order, which itself arises by chance from a precosmic state where there is no necessity. The recognition of pure chance is, however, inconsistent with the Principle of Sufficient Reason, which we know the atomists accepted. A reconciliation is suggested by a passage of Aetius (I. 29. 7) ‘Democritus ¹² The best discussion of the fragment is Barnes [1984], who, while finally opting for an agnostic stance, is more sympathetic to the view that Leucippus may have accepted universal teleology. The non-teleological interpretation that I propose is also maintained by McKirahan [1994], 321–2. ¹³ On the atomists’ theory a world order begins to form when some of the infinite mass of randomly jostling atoms form a circular eddy or swirl.

188

the atomists

and the Stoics say that it [i.e. chance] is a cause which is unclear to human reason,’ which may be read as asserting that the ascription of events to chance is a confession of ignorance of their causes, not a denial that they have causes. Some other pieces of evidence support this suggestion. Diogenes’ summary of the cosmology of Leucippus (IX. 30–3) concludes with the sentence ‘Just like the coming into being of worlds, so do their growth, decay, and destruction occur according to a certain necessity, the nature of which he does not explain.’ In line with his famous dictum, then, Leucippus held that all events, including the formation of worlds, happen according to necessity but was unable to say what it is that necessitates cosmic events. It is then plausible that either he himself or Democritus said that such events may be said to occur by chance, in the sense that we are (whether merely in fact or in principle is indeterminate) ignorant of their causes. Explanations of specific kinds of events and of particular events were governed by the principle that there are no chance events, but no attempt was made to offer explanations of the fundamental cosmic processes themselves. That need not imply that they are literally uncaused, but that they might as well be treated as such, since their actual causes are of a degree of complexity outstripping the powers of the human mind to discover. For the atomists, then, everything happens of necessity; the identification of necessity with the mechanical forces of impact and motion may have been due to Democritus. What exactly was his view on this? Aetius reports him as identifying necessity with ‘impact and motion and a blow of matter’ (I. 26. 2). Are impact and motion given equal status in this identification, or is it taken for granted that motion is always caused by prior impact? On the former construal some motion may be either uncaused or attributable to a cause other than impact. In favour of the first alternative is Aristotle’s evidence (Phys. 252a32–b2) that Democritus held that one should not ask for a cause of what is always the case. He might then have said that the atoms are simply always in motion. But while that principle allows him to exclude the question ‘What causes the atoms to be in motion?’ the Principle of Sufficient Reason requires that the question ‘Why is any particular atom moving with any particular motion?’ should have an answer, and it might appear inevitable that that answer should refer to a prior atomic collision, as is attested by various sources (e.g. Simplicius, In Phys. 42. 10–11; Alexander, In Meta. 36. 21–5).

the atomists 189 We have, however, to recall the evidence from Philoponus that atoms never actually collide or come into contact, with its implication that the basic physical forces are attraction and repulsion. On that view, most atomic motion is explained by the analogue of impact, namely repulsion, while the immobility of atoms relative to one another is explained by attraction, since the relative stability of atoms in an aggregate has to be explained, not by their literal interlocking but by their being held together as if interlocked by an attractive force operating over the tiny gaps between the atoms in the aggregate. But in addition, some form of attraction may also have explained some atomic motions; Sextus cites Democritus (M VII. 116–18) as holding that things of the same kind tend to congregate together, and as illustrating that by examples of the behaviour of animate (birds flocking together) and inanimate things (grains of different sorts being separated out by the action of a sieve, pebbles of different shapes being sorted together by the action of waves on a beach). That this principle was applied to the atoms appears from Diogenes’ account of the cosmogony of Leucippus where atoms of all shapes form a swirling mass from which they are then separated out ‘like to like’. The separation out of atoms of different sizes could adequately be accounted for by the stronger centripetal tendency of the larger, itself a function of their greater mass. But the context in Diogenes, where the atoms have just been described as being of all shapes, with no mention so far of size, suggests that ‘like to like’ is here to be understood as ‘like to like in shape’. Aetius’ report of Democritus’ account of sound (IV. 19. 3) asserts that atoms of like shape congregate together, and it contains the same illustrative examples as the Sextus passage. It is plausible, though not explicitly asserted, that this same principle accounts for the formation of aggregates of spherical atoms, for example, flames. We have, then, some evidence that Democritus’ dynamics postulated three fundamental forces, a repulsive force that plays the role of impact in a conventional corpuscular theory and two kinds of attractive force, one that draws together atoms of the same shape and another that holds together atoms of different shapes in an atomic aggregate. It is plausible that he applied the term ‘necessity’ to all three, regarding them alike as irresistible. It must, however, be acknowledged first that the evidence for this theory is extremely fragmentary and secondly that even if it is accepted we have no idea whether or how Democritus attempted to unify these

190

the atomists

forces into a unified theory. Stated thus baldly, the theory has obvious difficulties, for example, if two atoms of the same shape collide, do they rebound or stick together? If all atoms have both an attractive and a repulsive force, there must be some yet more basic principles determining what force or combination of forces determines their motion. Our sources give no hint of whether Democritus had so much as considered such questions.

Epistemology While we have no evidence to suggest that Leucippus was concerned with epistemological questions, there is abundant evidence of their importance for Democritus. It is quite likely that the latter’s epistemological interests were stimulated at least in part by his fellow-citizen and elder contemporary Protagoras. Our evidence is highly problematic, in that it provides support for the attribution to Democritus of two diametrically opposed positions on the reliability of the senses. On the one hand, we have a number of passages, including some direct quotations, in which he appears to reject the senses as totally unreliable; on the other, a number of passages ascribe to him the doctrine that all appearances are true, which aligns him with Protagorean subjectivism, a position that he is reported as having explicitly rejected (Plutarch, Col. 1108f). The former interpretation is supported mainly by evidence from Sextus, and the latter mainly by evidence from Aristotle and his commentators, but we cannot resolve the question by simply setting aside one body of evidence in favour of the other, since (a) in the course of a few lines (Meta. 1009b7–17) Aristotle reports both that Democritus says that either nothing is true, or it is unclear to us, and that he asserts that what appears in perception is necessarily true, and (b) Sextus (M VII. 136) ascribes some of Democritus’ condemnation of the senses to a work in which ‘he had undertaken to give the senses control over belief’. Prima facie, then, the evidence suggests that both interpretations reflect aspects of Democritus’ thought. Was that thought, then, totally inconsistent? Or can the appearance of systematic contradiction be eliminated or at least mitigated? The former interpretation is based on the atomists’ account of the secondary qualities, whose observer-dependence Democritus seems to

the atomists 191 have been the first philosopher to recognize. Our senses present the world to us as consisting of things characterized by colour, sound, taste, smell, and so forth, but in reality the world consists of atoms moving in the void, and neither atoms nor the void are characterized by any secondary quality. We thus have a dichotomy between how things seem to us and how they are in reality, expressed in the celebrated slogan (DK 68 B 9): ‘By convention sweet and by convention bitter, by convention hot, by convention cold, by convention colour, but in reality atoms and the void.’ Further, the distinction between the reality of things and the appearances which that reality presents has to be supplemented by an account of the causal processes via which we receive those appearances. Atomic aggregates affect us by emitting from their surfaces continuous streams of films of atoms which impinge on our sense organs, and the resulting perceptual states are a function of the interaction between those films and the atomic structure of the organs. For instance, for an object to be red is for it constantly to emit films of atoms of such a nature that, when those films collide with an appropriately situated perceiver, the object will look red to that perceiver. Hence we are doubly distanced from reality not only phenomenologically, in that things appear differently from how they are, but also causally, in that we perceive atomic aggregates via the physical intervention of other aggregates (viz. the atomic films) and the action of those latter on our sense organs. A number of fragments stress the cognitive gulf that separates us from reality: (B 6) ‘By this principle man must know that he is removed from reality’; (B 8) ‘Yet it will be clear that to know how each thing is in reality is impossible’; (B 10) ‘That in reality we do not know how each thing is or is not has been shown many times’; and (B 117) ‘In reality we know nothing, for truth is in the depths.’ This evidence immediately presents a major problem of interpretation. On the one hand, B 9 and associated reports stress the gulf between appearance and reality, claiming that the senses are unreliable in that they misrepresent reality. That dogmatic claim presupposes that we have some form of access to reality, which enables us to find the sensory picture unfaithful to how things are in fact. On the other hand, B 6, 8, 10, and 117 make the much more radical claim that reality is totally inaccessible, thereby undercutting the thesis that there is a gulf between appearance and reality. B 7, ‘This argument too shows that in reality we know nothing

192

the atomists

about anything, but each person’s opinion is something which flows in,’¹⁴ and the second half of B 9, ‘In fact we know nothing firm, but what changes according to the condition of our body and of the things that enter it and come up against it,’ attempt uneasily to straddle the two positions, since they draw the radically sceptical conclusion from a premiss about the mechanism of perception that presupposes access to the truth about that mechanism. We might conclude that Democritus simply failed to distinguish the dogmatic claim that the senses misrepresent reality from the sceptical claim that we can know nothing whatever about reality. An alternative strategy is to look for a way of interpreting the evidence that will tend to bring the two claims nearer to consonance with one another. We can bring the two claims closer to one another if the ‘sceptical’ fragments are interpreted as referring, not to cognitive states generally but specifically to states of sensory cognition. These fragments will then simply reiterate the thesis that we know nothing about the nature of reality through the senses, a thesis that is consistent with the slogan stated in the first half of B 9 and that dissolves the apparent tension internal to B 7 and the second half of B 9. Support for that suggestion comes from consideration of the context in which Sextus quotes B 6–10, namely that of Democritus’ critique of the senses, of which Sextus observes: ‘In these passages he more or less abolishes every kind of apprehension, even if the senses are the only ones which he attacks specifically.’ It thus appears that Sextus understands Democritus as referring in these fragments to the senses only, though in his (i.e. Sextus’) view the critique there directed against the senses in fact ¹⁴ The Greek of the last clause is epirysmi¯e hekastoisin h¯e doxis. I translate epirysmi¯e as an adjective, qualifying doxis (opinion), having the sense ‘flowing in’, from the verb epirre¯o. That is the sense of the word (which is found only in this passage (quoted by Sextus M VII. 137)) attested in the fifth-century ad lexicon of Hesychius. On the other hand, rysmos (an Ionic form of rythmos) was an atomistic technical term for ‘shape’ (Ar. Meta. 985b15–16), and one of the titles preserved in Diogenes Laertius’ list of the works of Democritus (IX. 47) is Peri Ameipsirysmi¯on (On Changes of Shape), where ameipsirysmi¯e is a noun. Further, though the noun epirysmi¯e is not itself found, the verb epirrythmizein does occur (very rarely) in the sense ‘alter’. Some scholars (including Guthrie [1965] and Barnes [1979]) therefore interpret the word here as a noun, a variant for ameipsirysmi¯e, giving the sense ‘opinion is a reshaping’. (de Ley [1969] actually proposes emending Sextus’ text to read ameipsirysmi¯e.) The point of the fragment is the same on either interpretation, namely, that our opinions about the world are determined by the impact on our receptive mechanisms of the flow of atoms from objects around us. That impact, produced by the constant influx of atoms, produces constant alteration (reshaping) of those mechanisms. The alternative interpretations pick out different stages in the causal process; since the whole process is required for an account of opinion and its relation to the reality of things, nothing substantial hinges on the choice of interpretation.

the atomists 193 applies to all forms of apprehension. This is confirmed by the distinction that Sextus immediately attributes to Democritus between the ‘bastard’ knowledge provided by the senses and the ‘genuine’ knowledge provided by the intellect (B 11). The latter is specifically said to be concerned with things that fall below the limits of sensory discrimination, and we must therefore suppose that the atomic theory itself is to be ascribed to this form of knowledge. This is supported by those passages (M VIII. 6–7, 56) in which Sextus associates the position of Democritus with that of Plato; both reject the senses as sources of knowledge and maintain that only intelligible things are real. For Plato, of course, the intelligible things are the Forms, whereas for Democritus they are the atoms, which are inaccessible to perception and, consequently, such that their properties are determinable only by theory. On this interpretation the position expressed in the fragments cited by Sextus is not general scepticism, but what we might term theoretical realism. The character of the physical world is neither revealed by perception nor inaccessible to us; it is revealed by a theory which, starting from perceptual data, explains those data as appearances generated by the interaction between a world of imperceptible physical atoms and sensory mechanisms also composed of atoms. But now, as Sextus points out (M VIII. 56) and Democritus himself recognized (in the famous ‘Complaint of the Senses’ (B 125)), scepticism threatens once again because the theory has to take perceptual data as its starting point. As a result, if the senses are altogether unreliable, there are no reliable data on which to base the theory, so, as the senses say to the mind in B 125, ‘Our overthrow is a fall for you.’ Commentators who read B 125 as expressing commitment to scepticism on the part of Democritus¹⁵ naturally reject the foregoing unitary interpretation. On this view B 117 and B 6–10 are not restricted to sensory cognition but express a full-blooded rejection of any form of knowledge, which must be seen as superseding the distinction between appearance and reality drawn in B 9 (first part) and B 11 and the claim to ‘genuine knowledge’ in the latter. Yet Sextus presents B 6–11 in a single context (M VII. 135–40) without any suggestion of a conflict within the collection. Moreover, in PH I. 213–14 he points out that, though the sceptics resemble Democritus in appealing to phenomena of conflicting appearances, such as the honey ¹⁵ For instance, Barnes [1979], ch. 24.

194

the atomists

that tastes sweet to the healthy and bitter to the sick, Democritus in fact uses those phenomena to support, not the sceptical position that it is impossible to tell how the honey is in fact, but the dogmatic position that the honey is itself neither sweet nor bitter. (I interpret the latter as the assertion that sweetness and bitterness are not intrinsic attributes of the structure of atoms which is the honey (see p. 191).) Sextus, in short, sees Democritus not as a sceptic, but as a dogmatist. Indeed, Sextus does not cite B 125, and it is possible that he did not know the text from which it comes; M VIII. 56 shows that he was aware of the problem that is dramatized in the fragment, but he clearly saw it as a difficulty for Democritus, rather than as signalling Democritus’ rejection of the basis of his own theory. At this point we should consider in what sense the theory of atomism takes the data of the senses as its starting point, and whether that role is in fact threatened by the appearance-reality gap insisted on in B 9. According to Aristotle (GC 315b6–15, 325a24–6), the theory started from sensory data in the sense that its role was to save the appearances, that is, to explain all sensory data as appearances of an objective world. Both Aristotle and Philoponus (In GC 23. 1–16) mention conflicting appearances as among the data to be saved; the theory has to explain both the honey’s tasting sweet to the healthy and its tasting bitter to the sick, and neither appearance has any pretensions to represent more faithfully than the other how things are in reality. All appearances make an equal contribution to the theory. That is a position which atomism shares with Protagoras, but the latter assures the equal status of appearances by abandoning objectivity; in the Protagorean world there is nothing more to reality than the totality of equipollent appearances. For Democritus, by contrast, the reconciliation of the equipollence of appearances with the objectivity of the physical world requires the gap between appearance and reality. Without the gap, a world of equipollent appearances is inconsistent, and hence not objective. But there is no ground for denying equipollence; qua appearance, every appearance is as good as every other. Hence the task of theory is to arrive at the best description of an objective world that will satisfy the requirement of showing how all the conflicting appearances come about.¹⁶ So far from threatening the foundations of the theory, then, the appearance-reality gap is essential to it. In that case, what is the point ¹⁶ For a similar view see McKim [1984].

the atomists 195 of the complaint of the senses in B 125? Does not that text provide conclusive evidence that Democritus believed that the gap threatened the theory, and hence (assuming that he understood his own theory) conclusive evidence against the interpretation that I am advancing? I do not think so, for the simple reason that we lack the context from which the quotation comes. The point of the complaint need not (and given the nature of Democritus’ theory certainly should not) be the admission that the theory is self-refuting. It is at least as likely to be a warning against misunderstanding the account of the appearance-reality gap as requiring the abandonment of sensory evidence. We may imagine an anti-empiricist opponent (Plato, say) appealing to the gap to support the claim that the senses are altogether unreliable, and should therefore be abandoned. In reply Democritus points out that the attack on the senses itself relies on sensory evidence. Sextus does indeed align Democritus with Plato in this regard (M VIII. 56). It is my contention, however, that when we put the Aristotelian evidence of the atomists’ acceptance of the appearances as the starting point of their theory together with all the other evidence, including the fragments, we have to conclude that the picture of Democritus as a failed Platonist is a misunderstanding. The atomists’ distinction between appearance and reality does not involve ‘doing away with sensible things’; on the contrary, appearances are fundamental to the theory, first as providing the data that the theory has to explain and secondly as providing the primary application for the observationally based terminology that is used to describe the nature and behaviour of the entities posited by the theory.¹⁷ A final objection, however, comes from Aristotle himself, who describes Democritus as concluding from conflicting appearances ‘that either nothing is true, or it is unclear to us’ (Meta. 1009b11–12). This is a very puzzling passage, for a number of reasons. Aristotle is explaining why some people go along with Protagoras in believing that whatever seems to be the case is so, and in the immediate context (1009a38ff.) he cites the phenomena of conflicting appearances and the lack of a decisive criterion for choosing between them as conducing to that belief. But at b9 he shifts from the thought that conflicting appearances lead to the view that all appearances are true to the sceptical account of those phenomena, namely that it is unclear which of the appearances is true or false, ‘for this is no more true ¹⁷ See chapter 1 of this volume.

196

the atomists

than that, but they are alike.’ This, Aristotle says (i.e. the belief that none of the appearances is truer than any other), is why Democritus said that either nothing is true, or it is unclear to us. So Democritus is represented as posing a choice of adopting either the dogmatic stance that none of the appearances is true, or the sceptical stance that it is unclear (which is true). Yet, in the next sentence Aristotle says that because Democritus and others assimilate thought to perception, they hold that what appears in perception is necessarily true (cf. GC 315b9; they (i.e. Leucippus and Democritus) thought that the truth was in appearance). So unless Aristotle is radically confused, the disjunction ‘either none of the appearances is true, or it is unclear to us’ must be consistent with the thesis that all perceptions are true. If ‘it is unclear to us’ is read as ‘it is unclear to us which is true,’ then the claims are inconsistent. I suggest, however, that what Democritus said was to the effect that either nothing is true, or it (i.e. the truth) is unclear. The first alternative he plainly rejected, so he maintained the second. And that is precisely what he maintains in B 117: the truth (about the atoms and the void) is in the depths, that is, it is not apparent in perception—it is unclear (ad¯elon) in the sense that it is not plain to see. That he used the term ad¯elon to apply to atoms and the void is attested by Sextus (M VII. 140), who cites Diotimus as evidence for Democritus’ holding that the appearances are the criterion for the things that are unclear and approving Anaxagoras’ slogan ‘the appearances are the sight of the things that are unclear’. The truth, then, that is, the real nature of things, is unclear (i.e. non-evident), but all perceptions are true in that all are equipollent and indispensable to theory. If that is what Democritus held, then it may reasonably be said that ‘true’ is the wrong word to characterize the role of appearances in his theory. ‘All appearances are equipollent’ is equally compatible with ‘All appearances are false,’ and in view of his insistence on the non-evident character of the truth, it would surely have been less misleading for him to say the latter. Though there are some difficult issues here, I shall not argue the point, since I am not concerned with defending Democritus’ thesis that all appearances are true. I do, however, accept that he actually maintained that thesis and have sought to explain why he did and how he held it together with (a) his rejection of Protagorean subjectivism and (b) the views expressed in the fragments cited by Sextus.

the atomists 197 The atomists’ account of appearances depends on the whole theory of perception of which it is part, and that in turn on their theory of human nature, and ultimately of the natural world as a whole. The theory is entirely speculative, since it posits as explanatory entities microscopic structures of whose existence and nature there could be no experimental confirmation. Developments in sciences such as neurophysiology have revised our conceptions of the structures underlying perceptual phenomena to such an extent that modern accounts would have been unrecognizable to Leucippus or Democritus; but the basic intuitions of ancient atomism, that appearances are to be explained at the level of the internal structure of the perceiver and of the perceived object, and that the ideal of science is to incorporate the description of those structures within the scope of a unified theory of the nature of matter, have stood the test of time.

Psychology Democritus’ uncompromising materialism extended to his psychology. Though there is some conflict in the sources, the best evidence is that he drew no distinction between the rational soul or mind and the non-rational soul or life principle, giving a single account of both as a physical structure of spherical atoms permeating the entire body. This theory of the identity of soul and mind extended beyond identity of physical structure to identity of function, in that Democritus explained thought, the activity of the rational soul, by the same process as that by which he explained perception, one of the activities of the sensitive or non-rational soul. Both are produced by the impact on the soul of extremely fine, fast-moving films of atoms (eid¯ola) constantly emitted in continuous streams by the surfaces of everything around us. This theory combines a causal account of both perception and thought with a crude pictorial view of thought. The paradigm case of perception is vision; seeing something and thinking of something both consist in picturing the thing seen or thought of, and picturing consists in having a series of actual physical pictures of the thing impinge on one’s soul. While this assimilation of thought to experience has some affinities with classical empiricism, it differs in the crucial respect that whereas the basic doctrine of empiricism is that thought derives from experience, for Democritus thought is a form of experience, or, more precisely,

198

the atomists

the categories of thought and experience are insufficiently differentiated to allow one to be characterized as more fundamental than the other. Among other difficulties, this theory faces the problem of accounting for the distinction, central to Democritus’ epistemology, between perception of the observable properties of atomic aggregates and thought of the unobservable structure of those aggregates. We have no knowledge of how, if at all, Democritus attempted to deal with this problem.¹⁸

Ethics and Politics The evidence for Democritus’ ethical views differs radically from that for the areas just discussed, since while the ethical doxography is meagre, our sources preserve a large body of purported quotations on ethical topics: the great majority from two collections, that of Stobaeus (fifth century ad) and a collection entitled The Sayings of Democrates. While the bulk of this material is probably Democritean in origin, the existing quotations represent a long process of excerpting and paraphrase, making it difficult to determine how close any particular saying is to Democritus’ own words. Various features of style and content suggest that Stobaeus’ collection of maxims contains a greater proportion of authentically Democritean material than does the collection which passes under the name of ‘Democrates’.¹⁹ Subject to the limitations imposed by the nature of this material, we can draw some tentative conclusions about Democritus’ ethical views. He was engaged with the wide-ranging contemporary debates on individual and social ethics of which we have evidence from Plato and other sources. On what Socrates presents as the fundamental question in ethics, ‘How should one live?’ (Plato, Gorg. 500c, Rep. I. 352d), Democritus is the earliest thinker reported as having explicitly posited a supreme good or goal, which he called ‘cheerfulness’ or ‘well-being’ and which he appears to have identified with the untroubled enjoyment of life. It is reasonable to suppose that he shared the presumption of the primacy of self-interest which is common both to the Platonic Socrates and to his immoralist opponents, Callicles and Thrasymachus. Having identified the ultimate ¹⁸ For further discussion of Democritus’ psychology, see Laks [1999] and Taylor [1999], 200–11. ¹⁹ For details see Taylor [1999], 222–7.

the atomists 199 human interest with cheerfulness, the evidence of the testimonia and the fragments is that he thought that it was to be achieved by moderation, including moderation in the pursuit of pleasures, by discrimination of useful from harmful pleasures, and by conformity to conventional morality. The upshot is a recommendation to a life of moderate, enlightened hedonism, which has some affinities with the life recommended by Socrates (whether in his own person or as representing ordinary enlightened views is disputed) in Plato’s Protagoras, and, more obviously, with the Epicurean ideal of which it was the forerunner.²⁰ An interesting feature of the fragments is the frequent stress on individual conscience, or sense of shame.²¹ Some fragments stress the pleasures of a good conscience and the torments of a bad one (B 174, B 215) while others recommend that one should be motivated by one’s internal sense of shame rather than by concern for the opinion of others (B 244, B 264, B 84). This theme may well reflect the interest, discernible in contemporary debates, in what later came to be known as the question of the sanctions of morality. A recurrent theme in criticisms of conventional morality was that, since the enforcement of morality rests on conventions, someone who can escape conventional sanctions, for example, by doing wrong in secret, has no reason to comply with moral demands.²² A defender of conventional morality who, like Democritus and Plato, accepts the primacy of selfinterest therefore faces the challenge of showing, in one way or another, that self-interest is best promoted by the observance of conventional moral precepts. Democritus seems to have attempted this both by appeal to divine sanctions (not post mortem, since for the atomists the soul-atoms were scattered on the death of the body, but in the form of misfortunes occurring during life, B 175), and by appeal to the ‘internal sanction’ of conscience. Democritus seems to have been the earliest thinker to make the latter central to his attempt to derive morality from self-interest, thus opening up a path followed by others including Butler and J. S. Mill. ²⁰ For a fuller discussion, see Kahn [1985]. This valuable study identifies a number of areas, such as the conflict between reason and desire, in which Democritus’ thought shows significant similarities to, and contrasts with, the early views of Plato. ²¹ While the relation between the concepts of conscience and of shame raises some intricate philosophical issues, I am not concerned to differentiate them, since the basic concept of self-reproach, which we find in the fragments, is common to the two. ²² See Antiphon DK 87 B 44, Critias DK 88 B 25; Glaucon’s tale of Gyges’ ring in Plato, Rep. 359b–360d; and Decleva Caizzi [1999].

200

the atomists

The attempt, however pursued, to ground morality in self-interest involves the rejection of the antithesis between law or convention (nomos) and nature (physis) that underlies much criticism of morality in the fifth and fourth centuries. For Antiphon, Callicles, Thrasymachus, and Glaucon, nature prompts one to seek one’s own interest while law and convention seek, more or less successfully, to inhibit one from doing so. But if one’s long-term interest is the attainment of a pleasant life, and if the natural consequences of wrongdoing, including ill-health, insecurity, and the pangs of conscience, give one an unpleasant life, while the natural consequences of right-doing give one a contrastingly pleasant life, then nature and convention point in the same direction, not in opposite directions as the critics of morality had alleged. (We have no evidence whether Democritus had considered the objections that conscience is a product of convention, and that exhorting people to develop their conscience assumes that it must be.) Though the texts contain no express mention of the nomosphysis contrast itself, several of them refer to law in such a way as to suggest rejection of the antithesis. B 248 asserts that the aim of law is to benefit people, thus contradicting Glaucon’s claim (Plato, Rep. II. 359c) that law constrains people contrary to their natural bent. B 248 is supplemented and explained by B 245; laws interfere with people’s living as they please only to stop them from harming one another, to which they are prompted by envy. So law frees people from the aggression of others, thus benefiting them by giving them the opportunity to follow the promptings of nature towards their own advantage. The strongest expression of the integration of nomos and physis is found in B 252: the city’s being well run is the greatest good, and if it is preserved everything is preserved, while if it is destroyed everything is destroyed. That is to say, a stable community is necessary for the attainment of that well-being which is nature’s goal for us. This quotation encapsulates the central point in the defence of nomos (emphasized in Protagoras’ myth (Plato, Prot. 322a–323a)) that law and civilization are not contrary to nature but required for human nature to flourish; that point is also central to the Epicurean account of the development of civilization (see especially Lucretius V).²³ ²³ For a fuller discussion see Procop´e [1989/90], and for Democritean theology see Broadie [1999], 220, and Taylor [1999], 211–16.

the atomists 201

Conclusion Atomism can thus be seen as a multifaceted phenomenon, linked in a variety of ways to various doctrines, both preceding, contemporary, and subsequent. Atomistic physics is one of a number of attempts to accommodate the Ionian tradition of comprehensive natural philosophy to the demands of Eleatic logic. Atomistic epistemology takes up the challenge of Protagorean subjectivism, breaks new ground in its treatment of the relation of appearance to reality, and constitutes a pioneering attempt to grapple with the challenge of scepticism. Atomistic ethics moves us into the world of the sophists and of early Plato in its treatment of the themes of the goal of life, and of the relations between self-interest and morality and between nomos and physis. The atomism of Leucippus and Democritus exercised a continuing influence throughout subsequent centuries, whether as a challenge to be faced, most notably by Aristotle, or as a forerunner to Epicureanism in all its aspects, and thereby to the revival of atomistic physics in the Corpuscular Philosophy of the seventeenth century.

202 the atomists

Appendix I conclude with a brief discussion of the vexed question of the connections (or lack of them) between Democritus’ ethics and his physical theory. In an earlier discussion (chapter 1 of this volume, endorsed without further argument in Gosling and Taylor [1982]) I argued against Vlastos’ claim (Vlastos [1945/6]) to find significant connections between the content of the two areas of Democritus’ thought. Vlastos’ position has found some recent defenders (and my views some critics), notably Sassi [1978] and Farrar [1988]; these discussions seem to me to call for some re-examination of the question. It is, I take it, common ground that in composing his ethical writings Democritus had not abandoned his physical theory, and therefore that, at the very least, he would have sought to include nothing in the former that was inconsistent with the latter. I shall make the stronger assumption that he took for granted in the ethical writings the atomistic view of the soul as a physical substance pervading the body. However, I remain unconvinced of any closer connection between physics and ethics. In particular, I see no indication that any ethical conclusions (e.g. that the good is ‘cheerfulness’) were supposed to be derived from the physical theory, or that the physical theory provided any characterizations of the nature of any ethically significant psychological state. In other words, I see no evidence that Democritus believed in type-type identities between ethical states such as cheerfulness and physical states such as having one’s soul-atoms in ‘dynamic equilibrium’ (Vlastos [1945/6], 584, Farrar [1988], 229). My earlier criticisms of this kind of view still stand. There is, however, one particular point on which I now think that I took scepticism too far. This was in my rejection of Vlastos’ interpretation of B 33, that teaching creates a new nature by altering the configuration of the soulatoms. My reason was that rythmos was an atomistic technical term for the shape of an individual atom, not for the configuration of an atomic aggregate, for which their term was diathig¯e. Hence metarythmizei (or metarysmoi) in the fragment could not mean ‘reshape’ in the sense of ‘produce a new configuration’. But, as Vlastos had already pointed out, the catalogue of Democritean titles includes Peri ameipsirysmi¯on, On changes of shape (DL IX. 47), which cannot refer to changes in the shapes of individual atoms (since they are unchangeable in respect of shape), and must therefore refer to changes in the shape of atomic aggregates. Further, Hesychius glosses ameipsirysmein as ‘change the constitution (synkrisin) or be transformed,’ and though he does not attribute the word to any author it is at least likely to have been used in that sense by Democritus, since neither the verb nor its cognates are attested to anyone else. It therefore now seems to me that

the atomists 203 Vlastos’ reading of the fragment is probably correct. For Democritus, teaching, like thought and perception is a physical process involving the impact of eid¯ola on the soul, with consequent rearrangement of the soul-aggregate. (Cf. B 197: ‘The unwise are shaped (rysmountai) by the gifts of fortune ...,’ and n. 14 above (p. 192.).) Acceptance of that causal picture does not, of course, commit one to endorsing type-type psychological identities. Psycho-physical identity having been set aside, some looser connections between Democritus’ ethics and other areas of his thought may perhaps be discerned. In chapter 1 of this volume I argue for a structural parallel between ethics and epistemology, a suggestion that still seems plausible to me. Another vague connection is with cosmology. It is not unreasonable to suppose that Democritus saw at least an analogy between the formation of worlds (kosmoi) from the primitive atomic chaos by the aggregation of atoms under the force of necessity and the formation of communities (also termed kosmoi, B 258, 259) by individuals driven by necessity to combine in order to survive. It may also be (as suggested by, for example, M¨uller [1965]) that the aggregation of like individuals to like, which is attested as operating in the formation of worlds (DK 67 A 1.31), had some counterpart in the social sphere.

Afterword For a sustained critique of the interpretation of the atomists contained in this chapter see Konstan [2000]. My suggestion that Philoponus’ testimony justifies the belief that the atomists denied atomic contact is demolished by Mansfeld [2007]. For fuller discussion of Democritus’ treatment of the nomos-phusis contrast and the relation of that treatment to some Platonic discussions see Taylor [2007].

12 Aristotle on the Practical Intellect1 In this paper I shall consider Aristotle’s conception in the Nicomachean Ethics of the role of the practical intellect, and specifically the question of how the intellect is related to the agent’s long-term goals. Does the intellect itself lay down the goals that are to be pursued, or is it restricted to working out means to the achievement of goals which it takes as given by some non-intellectual faculty, e.g. by the passions as in Hume’s theory? Discussion of this question will prove to have some far-reaching implications for Aristotle’s conception of the role of thought in the life of the moral agent.

1. The task of the practical intellect in general The first point to note is that there is not just one activity or kind of activity which is the practical exercise of the intellect, but many: just as there are many theoretical sciences and kinds of sciences, so there are many practical enquiries and kinds thereof. We have first the distinction (VI. 4) between productive and practical crafts, the former (e.g. housebuilding and medicine) directed to the production of things separate from their own exercise, the latter (e.g. musical performance) directed only towards their own excellent performance. The product of a productive craft need not be a substance, such as a house or pot, but may be a state (e.g. the healthy state of a patient), but in either case the craftsman aims to produce something beyond his own excellent performance (indeed, excellent performance of a productive craft presupposes such an external object, being defined as the kind of performance which generally produces the desired object). In ¹ Versions of this paper have been read in Edinburgh, Oxford, Boulder, Cornell, Pittsburgh, Munich, and Trier. I am grateful to the audiences at those universities for their helpful comments.

aristotle on the practical intellect 205 the case of a practical craft, however, ‘the excellent practice is itself the goal’ (1140b7). Even more central to our enquiry is a distinction relating specifically to practical crafts. In the opening chapter of the Nicomachean Ethics Aristotle points out that different crafts, both practical and productive, may be subordinated to one another in a hierarchy, the goals of the more specific crafts being subsumed under the goals of the more general, e.g. various ancillary crafts such as those of bridle-making are subsumed under the craft of horsemanship, which is in turn subsumed under, for example, the art of warfare. The supreme practical craft, he argues in chapter 2, is that of politics, which subsumes the ends of all other crafts under the highest end, that of promoting the good life for all the members of the community. Excellent practice, then (eupraxia), which is the end of praxis, may be confined to a special area of life, or may be non-specific, i.e. identical with the excellent organization of one’s life as a whole; the latter is the mark of the phronimos, who ‘is able to deliberate well about the things which are good and beneficial to him, not in a restricted sphere, e.g. with a view to health or strength, but such as promote living well as a whole’ (1140a25–8). Here we see Aristotle implying, rather than stating in so many words, the distinction between the first-order pursuit of intellectual goals and the second-order activity of organizing the pursuit of those goals into a satisfactory life. The clearest indication that Aristotle has seen this distinction is the fact that he assigns a particular name, phron¯esis, to the specific excellence of that second-order activity, distinguished alike from theoretical excellence (sophia) and from first-order practical excellence, i.e. skill at any particular craft. We have, then, narrowed down our enquiry to focus on that activity of the intellect whose object is the securing for the individual of the overall good life (to eu z¯en hol¯os 1140a28 = eudaimonia 1095a19–20), and whose specific excellence is phron¯esis. In investigating that activity we shall, of course, have to take into account both those features which it shares with other manifestations of practical intelligence (the various sorts of poi¯esis and praxis) and those which are peculiar to it. In general, the task of the practical intellect, whether exercised in poi¯esis or in praxis, is the initiation of change. Change is brought about by deliberation, whose field of operation is things which are (a) capable of change, (b) within the agent’s power to affect. Yet at the same time the ergon, the specific function of the intellect as such, is to attain to truth

206

aristotle on the practical intellect

(1139a28). Hence, in its action-initiating function, the practical intellect must still be seen as aiming at the truth; specifically, its object is practical truth, which Aristotle defines as ‘truth in agreement with right desire’ (1139a26–31). This might well seem to represent a mistaken assimilation of the practical to the theoretical; would not Aristotle have done better to say that the function of the theoretical intellect is to reach the truth, that of the practical to reach the good? But that itself would be a mistake. Aristotle shows a firm grasp of the insight that things are wanted under some description; hence desire can itself be represented as having a propositional content, e.g. ‘Such and such would be a good thing to have’, or ‘I must drink’. If the agent is to reach a correct decision on any practical matter, two things must be the case: first his practical beliefs must be true, and second he must be motivated to act in accordance with those beliefs, i.e. in Aristotle’s terms his desires must say the same thing as his intelligence (VI. 2). Suppose my aim is to be healthy; then if I am to reach correct decisions on how to act, say in deciding what, how much etc., to eat, I must have true beliefs, e.g. that too much animal fat is bad for you, and that such and such foods are high (low) in animal fat, and I must want to abstain from those foods I truly believe to be high in fat and to eat appropriate quantities of those I truly believe to be low in fat. The aim of the practical intellect, then, is to reach decisions to act; and decisions are reasoned judgements (prohairesis presupposes bouleusis 1112a15–16), expressive of desire. Aristotle’s formula ‘desiderative intellect or intellectual desire’ seems happily to express that complex situation, which is also reflected by his insistence that the practical intellect both issues in action and has truth as its object.² This direction towards practical truth is common both to the departmental, first-order function of the practical intellect and to its general, second-order search for the good life; this is shown by the parallelism between the definition of techn¯e, productive excellence, as ‘a productive disposition with a true logos’ (i.e. a true conception of what is to be produced, 1140a9–10) and of phron¯esis, overall practical excellence, as ‘a true practical disposition, with a logos, concerned with things good and bad for man’ (1140b5–6).³ In its search for practical truth, i.e. in deliberation, ² On practical truth see M¨uller [1982], ch. 6, esp. 231–6 and 258–9. ³ Aristotle does not discuss departmental praxis, e.g. musical performance, or its specific excellence; he distinguishes poi¯esis from praxis in VI. 4 on the ground that the former aims at the production of an object separate from the performance, whereas the latter does not, but when he discusses praxis and its

aristotle on the practical intellect 207 the practical intellect operates under an important restriction, viz. that deliberation is not concerned with ends but with the choice of actions conducive to the attainment of ends already given. So a doctor doesn’t deliberate about whether to cure someone, or a statesman about whether to create eunomia; they take these as the aims of their respective undertakings, as already given, and busy themselves with the enquiry into how those aims are to be realized (1112b11–16). We desire things, e.g. health or eudaimonia, as ends, and select the means to them, but it is inappropriate to speak of selecting those ends themselves (1111b26–9). We have to be cautious here. First, Aristotle does not say that the desire for a given end lacks propositional content; the theory of desire expressed in VI. 2, MA 7 and De an. III. 9–10 requires that the desire for e.g. health is expressed in a thought such as ‘Health is something good’. Secondly, it is unspecified in EN III how determinate is the conception of the end which enters into the desire for it; the example of health suggests a fairly full conception of what health is, which can be a starting point for practical medical studies, whereas the example of eudaimonia is compatible with a thin conception, e.g. I want a worthwhile life, but have no clear conception of what kind of life that would be. Thirdly, the account of deliberation is silent on the question of how the conception of the end and the belief that the end is something good are acquired: that account does not, therefore, exclude the possibility that that conception and/or belief are acquired as the result of some process of intellectual enquiry. All that Aristotle is committed to is the negative thesis that that conception and that belief are not the result of deliberation. And that negative thesis has the air of a tautology; for ‘deliberation’ seems just to be the name for that species of practical thinking which is concerned with the selection of actions conducive to the achievement of a given end. For any particular process of deliberation, therefore, it is trivially true that the end assumed in that process was not arrived at via that process. But is there any reason why the end assumed in one piece of deliberation should not have been arrived at via another? For example, might not strategic deliberation lead excellence, phron¯esis, in ch. 5 he confines himself to the search for the good life as a whole, apparently forgetting about departmental praxis. He gives no example of such praxis; even musical performance, which one would expect to be such an example (as it is in MM 1179a8–10), is described at 1103a32–b2 as poi¯esis, in apparent contrast to praxis. (I am grateful to Prof. A. W. M¨uller for this information.) See also M¨uller [1982], 218–21.

208

aristotle on the practical intellect

to the result that the cavalry ought to be equipped with a certain sort of bridle, and the smith take that conception as the starting point for his deliberation as to how to make it? To be complete, Aristotle’s theory must allow for such chains of deliberation. But any deliberation or chain of deliberations must start from the conception of some end which is not arrived at by deliberation; otherwise the chain of deliberation could have had no beginning.

2. Deliberation towards given ends and grasp of the right ends Deliberation, then, cannot be autonomous; necessarily, it takes its starting points from elsewhere. This restriction on deliberation is applied by Hume to reason in general, in his famous insistence that reason is, and ought only to be, the slave of the passions (Treatise II. iii. 3). The function of reason in Hume’s theory is the discerning of the relations between ideas, which in the practical sphere amounts to deliberation; reason can determine that if a certain thing is done, a certain outcome will result, or that if such and such is to be achieved, such and such must be done, but reason itself is incapable of determining that a given end is to be pursued. Ends are dictated by desires, i.e. by the passions; reason is restricted to working out the best means to what we happen to want, i.e. it is the slave of the passions. An influential tradition of exegesis, deriving from Julius Walter’s Die Lehre der praktischen Vernunft in der griechischen Philosophie (Jena, 1874),⁴ has attempted to fit Aristotle also into this mould. On this view the functioning of the practical intellect alike in its second-order activity and in its departmental role is identified with deliberation. The excellence of the practical intellect, phron¯esis, is therefore identified with skill in deliberation to the attainment of ends determined not by the intellect but by e¯thik¯e aret¯e, the excellence of the non-intellectual, appetitive side of the personality. This interpretation fits Aristotle’s insistence on the necessity of proper upbringing for the good man (e.g. 1095b2–5), and is supported above all by two famous passages from VI. 12–13, 1144a7–9 and 1145a4–6, both of which say that eudaimonia requires phron¯esis and excellence of character, the latter ⁴ For details see Allan [1953].

aristotle on the practical intellect 209 making one’s aim right, the former enabling one to do what promotes the achievement of that aim (ta pros ton skopon, ta pros to telos).⁵ Some of the defects of what we might call the ‘Humean’ interpretation of Aristotle were well brought out by Allan,⁶ and his work has been taken further by others, notably Richard Sorabji, David Wiggins, and John McDowell.⁷ Allan’s principal contribution is to cite the abundant textual evidence to show that Aristotle’s conception of phron¯esis is not restricted to skill in deliberation towards a predetermined end, but at least includes a correct conception of that end itself. A single instance will suffice here: at 1140b11–21, Aristotle argues that the conception of the good which is the starting point of the practical reasoning of the phronimos will not be available to someone who has been corrupted by excessive desire for pleasure or aversion from distress; e.g. someone who can’t endure any distress will simply not appreciate that courage is a necessary component of the good life. It is in virtue of preserving this conception that s¯ophrosun¯e is so called, h¯os s¯oizousan t¯en phron¯esin. This conclusion, firmly enough established by Allan, is strengthened by consideration of the structural parallelism between the respective excellences of the theoretical intellect (sophia) and the practical (phron¯esis). Both consist of two parts, (a) a correct grasp of first principles, (b) the ability to reason correctly from principles thus grasped. Aristotle’s term for the former, certainly in the theoretical sphere and perhaps also in the practical,⁸ is nous (1141a7–8, 1142a25–6, 1143a35–b5); the latter he calls epist¯em¯e in the theoretical sphere (1139b31–2), euboulia in the practical (1142b16, 31–3). Sophia is nous + epist¯em¯e (1141a18–20); phron¯esis is nous + euboulia. Aristotle does not, indeed, give the latter formula in so many words; but he comes very close to it in 1142b31–3, and in fact deviates from it in the direction of an even stronger emphasis on the grasp of the end, which he not only distinguishes from euboulia, but appears actually to identify with phron¯esis itself; euboulia is ‘correctness concerning what conduces to the end, of which phron¯esis is the true conception’, where ‘the end’ is a more natural antecedent for ‘which’ than ‘what conduces to the end’.⁹ ⁵ See also 1178a16–19. ⁶ See note 4. ⁷ Sorabji [1973–4],Wiggins [1975–6], McDowell [1980]. ⁸ It is disputed whether (as maintained by Kenny [1978], 171–2) the word nous designates the correct grasp of practical, as well as theoretical principles. On practical nous see below. ⁹ So Gauthier and Jolif [1958/9], ii. 518–19; Kenny [1979], 106–7; M¨uller [1982], 267–8; and Bostock [2000], 85. The opposite view is defended by Burnet [1900].

210

aristotle on the practical intellect

One component, then, of the excellence of the practical intellect is having a correct conception of what the good for man is. Aristotle’s language here is uncompromisingly cognitivist; phron¯esis is a true conception of the end (1142b33), it is a true practical disposition concerned with things good and bad for man (1140b4–6, 20–1), excellence of character requires ‘excellence either natural or acquired in having correct beliefs (orthodoxein) about the first principle’ (1151a18–19). Just as someone who possesses scientific knowledge must deduce what he knows from first principles which are themselves known (An. post. I. 1, 71a1–2), so the phronimos must derive his reasoned selection of action from first principles which he grasps as true. How does the phronimos come to grasp those true principles?

3. Who or what grasps ethical principles? Three interpretations Here we find ourselves facing a problem which has taxed the ingenuity of some of Aristotle’s ablest commentators. Basically, the problem is posed by the fact that Aristotle suggests at various points three different accounts of how the possessor of practical wisdom acquires a reliable grasp of principles;¹⁰ these accounts are, moreover, (a) prima facie incompatible with one another, (b) such that two of them appear to threaten the possibility of knowledge of principles.¹¹ On the first, ‘Humean’ account one comes to a grasp of principles, not via the intellect at all, but through the habituation (ethismos) of the appetites, resulting, where successful, in excellence of character.¹² Having a reliable grasp of the correct principles of conduct is simply a matter of having been brought up to like and enjoy good kinds of action and to dislike bad ones and find them unpleasant (EN ii. 1–3, especially 1104b3–13); and the good and bad are simply those kinds of action which the person of practical wisdom likes and dislikes. On the second account, the principles of conduct are ‘reputable’ (endoxa) ¹⁰ Though I initially set these out separately for the sake of clarity in exposition, an adequate theory (as I later suggest, in common with most commentators) must combine elements of these different proposals. ¹¹ Cf. Irwin [1978], 258: ‘Aristotle seems to recognize three incompatible ways to first principles; habituated virtue without wisdom [= my first alternative], virtue including wisdom [= the third alternative], and dialectic [= the second alternative].’ ¹² Bostock [2000], 88–96 maintains this position tentatively. Cf. M¨uller [1982], 304.

aristotle on the practical intellect 211 or well-grounded opinions, identified as those accepted by everyone, or by most people or by the wise, i.e. theorists of repute.¹³ A reliable grasp of such principles is achieved by a method of critical enquiry, exemplified by Aristotle’s own procedure, which seeks to identify the most plausible principles and to establish their credentials by, as far as possible, removing objections and eliminating apparent inconsistencies between them. On the third, the account of the grasp of practical principles is identical with that of theoretical: like the latter, the former are grasped by nous, which is the name for a grasp of general principles arrived at by induction from sensible particulars.¹⁴ The first two accounts are prima facie inimical to claims that the rational agent possesses knowledge of principles; on the first the basic attitudes are not cognitive at all, but rather affective, while on the second, it might appear, principles are not known, but accepted on the ground that they, or propositions supporting them, are believed by everyone, or by most people, or by some eminent theorist or other. But on that account, surely, nothing more can be claimed for them than plausibility. Let us take the first account, according to which the first principles are grasped by habituation of character. This is supported by 1098b3–4, where Aristotle lists various ways in which principles are grasped. ‘Some’, he says ‘by induction, some by a sort of habituation and some in other ways’, and given his insistence (1095b4–7) on proper upbringing for a grasp of the starting points of moral reasoning, it is an ¹³ Irwin [1988] defends a version of this position. In his view the first principles of ethics are discovered, not by ‘pure dialectic’, but by ‘strong dialectic’. In strong dialectic principles based on well-founded opinions are themselves justified by fundamental truths of metaphysics. Irwin writes: ‘Aristotle needs strong dialectic, to make up for the weakness of pure dialectic. He thinks he can evaluate the common beliefs from a sufficiently independent view, because he has metaphysical arguments to justify some of his leading principles, and especially those principles that support his view of substance, matter, nature and essence. His metaphysical arguments are not non-dialectical, since they do not stray outside common beliefs altogether. But they are not merely dialectical, since they rest on premisses that we must believe if we are to believe in an objective world at all, or if we are to understand or to explain nature. The arguments of strong dialectic resting on those principles underlie Aristotle’s conception of the soul and of human nature, which in turn underlies the main claims of the Ethics’ (358). ¹⁴ So Reeve [1992], ch. 1. On Aristotle’s doctrine of the grasp of theoretical principles see Barnes [1994]. Cooper [1975] agrees with the thesis that the first principles of ethics are grasped by nous, but rejects the inductive account of nous. In his view the principles which are grasped by nous are established by dialectical thinking: ‘Aristotle holds both that the principles of the sciences are known intuitively, by nous, and that they can be established by discursive dialectical argument’ (67). ‘A similar condition obtains in his moral philosophy’ (69). Since Reeve too insists on the indispensable role of dialectic in the clarification of principles which have already been justified by experience (40), his position does not differ widely from that of Cooper.

212

aristotle on the practical intellect

extremely plausible inference that he regards habituation as the process by which moral principles are grasped. But how is that compatible with the requirement that that process should be cognitively stable? For Aristotle habituation is the training of the appetitive element of the personality, which is not itself adapted for the discovery of truth, but is rational only derivatively, to the extent that it is capable of modification by rational considerations; it can ‘listen to reason as to its father’ (1103a1–3). The exercise of such a capacity cannot therefore itself be a reliable method of acquiring truth. Rather we must conceive of the process of habituation, not as a conditioning of ‘blind’ appetites, but as a ‘twin-track’ process in which the appetitive responses are progressively refined under the guidance of the intellect which is itself undergoing a parallel process of refinement or rather enlightenment; the clearer the insight the intellect has of ethical principles, the more precise the instructions it can issue to the desires.¹⁵ But the desiderative component is essential for the grasp of principles to do its motivational work. This interpretation has the advantage of eliminating the prima facie incompatibility between the thesis that practical principles are acquired by habituation and the other two. If habituation itself presupposes a method by which the intellectual element is developed so as to grasp the truth, then, consistently with the thesis that principles are grasped by habituation, either of the other theses might be a correct account of that method. The thesis that the principles of practical thought are ‘reputable opinions’ (endoxa) raises a number of complex problems. First, what is the scientific status of endoxa and hence of arguments from them? Second, does Aristotle claim that all or only some moral principles are endoxa? At Top. I. 1–2 he distinguished scientific reasoning, whose principles are necessary truths, from dialectical reasoning, whose principles are endoxa. The latter is not itself scientific, but is a critical method among whose uses Aristotle particularly mentions the role of dialectical reasoning in examining the principles of the sciences; the principles of any science cannot be discussed within that science, since they are primitive with respect to it, but dialectic ‘being a technique of examination provides a way towards the principles of all the sciences’ (101b3–4). It is not totally clear what Aristotle means by ‘a way towards’ the principles. Does he mean that dialectic proves the principles? It ¹⁵ So Woods [1986], 148; Reeve [1992], 60–1. For fuller discussion see Sherman [1989], ch. 5.

aristotle on the practical intellect 213 is hard to see how arguments from received opinion are supposed to issue in knowledge of necessary truths. A possible response to these difficulties is to suggest that the role of dialectic in general is not to prove principles, but (a) to find supporting arguments for principles already grasped inductively¹⁶ and (b) to provide arguments against putative principles. If that general account were applied to the special case of practical principles, the second and third accounts listed above cease to be rivals, not merely to the ‘habituation’ thesis, but also to one another. The phronimos grasps first principles inductively, thereby exercising nous, and is able to support them by arguments from endoxa. Do the texts provide any support for this appealingly eirenic suggestion? Disappointingly, the texts are indecisive. Some may indeed be read as supporting this suggestion: thus EE 1216b26–8 says that the aim of ethical enquiry must be ‘to seek conviction through the arguments, using ta phainomena (i.e. ta endoxa) as pieces of evidence and examples’, which suggests the supportive role described above. EN 1098b9–12 is also naturally read in the same way. Aristotle has argued for his identification of human good as excellent actualization of soul from general principles about the function of a thing, viz. that in the case of things which have a function their good or bad state is determined by the performance of their function, and that the function of a human being is rational activity. He then supports this conclusion by arguments from common opinion, designed to show that excellent rational activity fits generally accepted beliefs about human good, prefacing these by the statement ‘We must examine it not only from the conclusion and the premisses of the argument but also from what is said about it; for everything which is the case is consistent with what is true, but what is true soon disagrees with what is false’. Elsewhere, however, the phainomena seem to provide the principles of proofs themselves, though in these passages it is not clear that the conception of proof is uniform. EE 1215a7–8 says that refutations of opinions are demonstrations of the theories (logoi) opposed to them, which suggests a formal proof by reductio ad absurdum. But at EN 1145b2–7 Aristotle introduces his discussion of akrasia by saying that, as in the other cases, we must begin by setting out the phainomena and asking questions, and so prove (deiknunai) all the reputable ¹⁶ So Reeve [1992], 40: ‘Dialectical arguments are not proofs of first principles ... Instead, dialectic takes principles already justified by experience and made true by reality and ‘‘clarifies’’ them. And this is as true in ethics, where the object of enquiry is the good, as it is in science ...’

214

aristotle on the practical intellect

opinions, or if not all, as many as possible and the most trustworthy ‘for if the difficulties are resolved and the reputable opinions are left in place, a sufficient proof will have been given’ (cf. Phys. 211a7–11). If we may take the phrase ‘as in the other cases’ as indicating that this method is to be employed in every case (as is perhaps also suggested by EN 1095a28–b7), then the notion of proof in ethics is radically different from that suggested by the axiomatic model. According to the latter, proving a proposition is showing that it has to be true, since it follows from one or more premisses which, independently, have to be true. On the ethical model proving a proposition consists in showing that it is (or follows from) a reputable opinion which has survived the process of setting out the phainomena and raising difficulties about them. Instead of a system of proof Aristotle appears to be offering a coherentist scheme of justification which issues at best in defeasible judgements of the form ‘Since p forms part of the best available scheme of ethical beliefs, it may be held as true until good reasons are discovered for revising the scheme’.

4. The role of phron¯esis in the choice of ends This assumes, however, that all ethical principles either are or are derived from endoxa, and it is not clear that that is true. In Aristotle’s ethics, general philosophical principles, especially those from his metaphysics and psychology, play a fundamental role. We have already cited his use of principles about the function of things in identifying the good for man; other instances are the metaphysical arguments in EN X for the thesis that theoretical excellence is the highest good, and his frequent employment of the principle that, for any subject matter where things appear to different observers, the way things really are is the way they appear to the observer in good or proper condition (e.g. the way things taste to the healthy person is how they really taste).¹⁷ He gives no sign of thinking of these as merely reputable opinions, or as derived from such opinions. Rather they are principles fundamental to his whole scheme of thought, and as such they seem to be regarded as unassailably true. But how are they known? He ¹⁷ For fuller discussion of this principle see Gosling and Taylor [1982], ch. 17.2, and ch. 6 of this volume.

aristotle on the practical intellect 215 gives us no explicit guidance. It remains a possibility that they are supposed to be grasped inductively, by nous. The crucial text is EN 1143a35–b14.¹⁸ There Aristotle distinguishes two objects of nous, viz. undemonstrated principles on the one hand and particular instances falling under them on the other; these are the ‘extremes in either direction’, i.e. the starting points and finishing points of reasoning, and they share the property of being undemonstrated (since there is no demonstration of singular propositions in Aristotle’s syllogistic). This distinction is then complicated by the introduction of the distinction between theoretical and practical reasoning, giving a distinction between nous of undemonstrated theoretical principles and nous of particular practical instances (b1–3). Is practical nous here then just a faculty of moral perception, by the exercise of which we see instances of conduct as falling under moral characteristics, without any reference to principles? That appears to contradict 1142a23–30, where that very faculty is called a special sort of perception and is contrasted with nous, precisely on the ground that the latter is concerned with undemonstrated principles, which must, given the context, be practical principles. A resolution of this apparent contradiction is suggested by 1143b4–5, where Aristotle says that particular instances of conduct are ‘principles of that for the sake of which , for universals come from particulars: so we must have perception of the latter, which we call nous’; this appears to describe the inductive procedure familiar from An. post. II. 19.¹⁹ Aristotle would then have a tidy and unified doctrine were he to hold that there is nous of moral as well as of theoretical principles, that both kinds of nous are acquired inductively, and that perception of morally significant instances is (another kind of) nous, which provides the perceptual data from which moral principles are inductively derived.²⁰ He does not, however, set things out thus neatly. Rather, the emphasis in this discussion of nous is on its perceptual role; it is a natural endowment which develops with experience (b6–9), which is why we must pay attention to the undemonstrated pronouncements of people of age and practical sense ‘for, having an eye from experience, they see aright’ ¹⁸ For a full discussion of this difficult passage see Dahl [1984], 227–36. ¹⁹ Cf. EN 1142a19, where Aristotle suggests tentatively that the principles come from experience. ²⁰ So Kenny [1978], 170–1; [1979], 151–2; Engberg-Pedersen [1983], 211–19; Dahl [1984], 41–5.

216

aristotle on the practical intellect

(b11–14). The sort of eye one has from experience would naturally seem to be the trained perception of particular instances as falling under some significant characteristic, exemplified by a general’s eye for terrain; what the untrained observer sees purely as a pleasant rural scene, the trained military eye simply sees as the place for the enemy to concentrate his armoured reserve. That type of perception is indeed trained by experience, but is nonetheless exercised in particular instances, not in grasping any general proposition, as in standard cases of induction. Yet a little later (1144a29–36) Aristotle attributes the grasp of first principles of conduct (represented schematically as ‘Since such and such (let it be whatever you like) is the best thing and the good ...’) to ‘the eye of the soul’, pointing out that that eye does not see properly without excellence of character. It therefore appears that Aristotle may have thought that the possession of the trained eye enables its possessor to formulate universal principles. But he also emphasizes that correct conduct cannot be formulated in exceptionless generalizations, but has to be determined by particular circumstances discerned by perception (1104a6–10, 1109b20–3, 1126b2–4; cf. IX. 2). On that conception of phron¯esis general principles specify the end only indeterminately, e.g. one has a general conception of a good life which embraces virtues of character such as courage, but what it is to be courageous cannot be exhaustively specified in any formula, but has to be recognized by the trained judgement (i.e. perception) of the courageous person.²¹ That is certainly Aristotle’s dominant account of excellence of character. It therefore seems that the view of practical nous which that account requires fits awkwardly with the texts which suggest that nous ought to be a grasp of universal principles, reached by induction from particular instances. His actual doctrine of goodness of character tends rather to assimilate it to the quasi-perceptual grasp of instances of moral concepts. A solution of this problem is possible, if we assume (a) that it is the task of induction to lead from sensible particulars to the mastery of a general concept, and (b) that this mastery consists precisely in the ability to recognize an indefinite number of particular instances falling under the concept, without the necessity of subsuming these particular instances under a single general formula.²² According to this account, habituation ²¹ See Engberg-Pedersen [1983], 198–9. ²² Dahl [1984], 44, expresses this idea exactly: ‘What nous grasps are particular actions that are good or to be done. Its grasp of these is based on experience and hence on a kind of induction. Implicit in

aristotle on the practical intellect 217 itself is understood as the acquisition of this perceptual capacity, and furthermore the acquisition of that capacity may itself be fostered by the reputable opinions of people of practical sense on whose unproved sayings and opinions we are to rely no less than on proofs, since they have the eye of experience and so see aright (1143b11–14). Reputable opinions and perception are thus mutually reinforcing elements in moral education.²³ Aristotle’s account, though obscure in detail, is unambiguous in its central contention that cognitive capacities are necessary for the correct orientation of the appetitive responses. This result presents two problems. First, what are we to make of those passages which seem to support the Humean interpretation? The second, more substantial problem is this: does the exercise of moral nous allow any genuine role to deliberation? In fact, the passages from VI. 12 and 13 fit our interpretation quite well. At 1144a6 ff. Aristotle is making the point that the achievement of eudaimonia requires excellence both of the practical intellect (i.e. phron¯esis) and of motivation (i.e. e¯thik¯e aret¯e). The latter, he says, makes one’s aim right, the former makes right the things one does to achieve that aim. On the interpretation which I have proposed, both these things are true, though admittedly they do not amount to a complete account of phron¯esis. On the Humean interpretation, e¯thik¯e aret¯e selects the mark to aim at, phron¯esis the means to it. On my interpretation it is phron¯esis which selects the mark, though Aristotle does not say so here. What e¯thik¯e aret¯e does is to make one’s aim right, i.e. it motivates one actually to aim at what phron¯esis has selected, and phron¯esis has then the further job of finding the means to that aim.²⁴ 1145a4–6 makes the same point: for prohairesis to be right we need both phron¯esis and e¯thik¯e aret¯e; h¯e men gar to telos h¯e de ta pros to telos poiei prattein. The verb to be supplied after telos is presumably this grasp are universal ends that one should be aiming to promote. One sees the universal ends that one should be aiming at by seeing them in the actions that one recognizes one should perform. One sees them in these actions because, in recognizing that one should perform them, one recognizes that actions like them should be performed. Whether or not one can articulate the content of what one has grasped, one has inductively acquired a universal end.’ Cf. Engberg-Pedersen [1983], 218–21. For a fuller exposition of this conception of the mastery of a concept and its connection with Wittgenstein’s remarks (in Philosophical Investigations I, paras. 143 ff.) on following a rule, see McDowell [1979]. ²³ So Sherman [1989], 43–4. ²⁴ Here I follow Allan [1953], who himself acknowledges his debt to Loening [1903], and Gauthier and Jolif [1958/59], ii. 577. Cf. Sherman [1989], 80–1.

218

aristotle on the practical intellect

poiei, and the sentence is to be read as ‘the one (i.e. aret¯e) makes the telos (i.e. makes something the telos), the other (i.e. phron¯esis) makes one do the things conducive to the telos’, where ‘makes the telos’ is understood in the same sense as ‘makes the aim right’ in the previous passage. Finally, it is worth mentioning two passages which might appear to support the Humean interpretation, though in fact they tell firmly against it, in favour of the interpretation I have defended. The first is the assertion at 1139a35–6 that ‘thought by itself moves nothing’, which defenders of the Humean interpretation have interpreted in the sense of ‘reason is the slave of the passions’. But the sentence runs ‘thought by itself moves nothing, but practical thought, i.e. thought directed with a view to something, does; for this (i.e. practical thought) is also the first principle of productive thinking’. Aristotle agrees with Hume that the initiation of action requires desire as well as thought: he rejects Hume’s central thesis that the function of reason is simply to devise ways of carrying out the dictates of desire. The second passage is one mentioned earlier, 1151a15–20, where Aristotle is reiterating his point that goodness of character preserves one’s grasp of the first principles of morals, while wickedness destroys it (cf. 1140b11–21). In practical reasoning the starting point is the goal, in mathematics it is the axioms; ‘for neither in one nor the other is reason demonstrative of the first principle, but excellence, whether natural or acquired, of having correct beliefs about the first principle’. Plainly this cannot mean that our grasp of first principles in general is not given by the intellect, but somehow emanates from a non-rational element, since that is obviously inconsistent with Aristotle’s general doctrine of how we come to grasp first principles.²⁵ ²⁵ Most commentators (e.g. Engberg-Pedersen [1983], 185; Reeve [1992], 49; Irwin [1999], 111; Bostock [2000], 90 and 224) interpret the passage in precisely the sense which I reject. Bostock is typical; he translates a17–19 as follows: ‘In neither case is it reasoning (logos) which teaches us the first principles, but [in the case of actions] it is virtue, either natural or habitual (ethist¯e), that teaches us right opinion about the first principle.’ The supplementation in square brackets is essential for this interpretation; but this supplementation seems to me unnecessary, and therefore unjustified. Without supplementation it is quite clear that the phrase all’ aret¯e ¯e phusik¯e ¯e ethist¯e tou orthodoxein peri t¯en arch¯en applies both to theoretical and to practical thought, not merely to the latter, as Bostock et al. suppose. Aristotle is saying here that in neither kind of thought are the principles proved by reasoning, but are acquired by another kind of excellence. The theme of the surrounding context is certainly the role of ethical virtue and vice in the preservation or corruption of the grasp of ethical principles, but it does not follow that this particular sentence refers to the specifically ethical sphere rather than to both the ethical and the theoretical. Ross’s original translation (Ross [1925]) agrees with Bostock et al.: ‘in actions the final cause is the first principle, as the hypotheses are in mathematics; neither in that case is it argument that teaches the

aristotle on the practical intellect 219 Rather the point is that in neither form of reasoning does reason demonstrate those principles, which have therefore to be known non-demonstratively.²⁶ It is, then, clearly Aristotle’s doctrine that excellence in deliberation (euboulia) is only one element of phron¯esis, the other being the grasp of first principles. But that schematic account can be realized in either of two ways. In the first, set out in the doctrine of the practical syllogism, it is nous (understood as the grasp of general principles) which formulates the major premiss, thereby providing a complete specification of the end to be achieved, while the premiss or premisses setting out the steps by which this end is to be achieved are worked out in a quasi-deductive way by euboulia. Thus suppose that shelter is necessary for survival, and hence a good. It is possible to specify completely what shelter is, and given our knowledge of the world, specifically of the nature and availability of building materials, and our knowledge of techniques of building, we can work out that if shelter is to be obtained then such and such materials must be organized in such and such ways.

5. Two tasks of the practical intellect This model certainly applies readily to technical deliberation, from which Aristotle normally takes his examples of practical reasoning. But on the other hand Aristotle’s doctrine of excellence of character, and the moral nous which that excellence presupposes, does not allow the end to be specified in advance of the process of ‘deliberation’. Hence that doctrine requires the second way in which the schematic account is realized. In this way, the end is specified indeterminately, e.g. ‘the good life first principles, nor is it so here—virtue either natural or produced by habituation is what teaches right opinion about the first principle.’ So Gauthier and Jolif [1958/9], i. 208, and Dirlmeier [1969], 108. But the revision of Ross by Urmson (Urmson [1984]) favours the reading which I propose: ‘neither in that case is it reason that teaches the first principles, nor is it so here—excellence either natural or produced by habituation is what teaches right opinion about the first principle.’ Mu¨ ller [1982], 271 also favours that reading: his paraphrase runs ‘In mathematics no less than in practical thought [Praxis] a natural or acquired proficiency [Qualifiziertheit] brings about correct assessment in the area of starting-points.’ ²⁶ ‘Acquired’ renders ethist¯e a 19, literally ‘acquired by habituation’. Here is another passage in which Aristotle says that ethismos is a method of acquiring correct beliefs; but since it is cited as a possible source of such beliefs in the theoretical as well as in the practical sphere it is likely that it has the more general connotation of ‘training’ than the specific connotation ‘training of the appetitive soul’ which it has in EN II.

220

aristotle on the practical intellect

requires that one act courageously’, and that indeterminate conception is made determinate by the exercise of moral insight in concrete situations. Excellence of character is a disposition to select actions, a disposition which is in a mean, i.e. which consists in being neither too given to a certain motivation nor insufficiently sensitive to it. And what counts as being neither excessively nor insufficiently motivated cannot be exhaustively specified by any formula, but is determined by the judgement of the phronimos. For example, being courageous is being neither excessively timorous nor being insufficiently motivated by fear, and similarly being neither excessively nor insufficiently bold (thrasus). But what counts as being courageous in any actual situation has to be discerned by the phronimos via his informed judgement of the appropriate action in those circumstances; there is no way in which he can read off the requirements of courage from his antecedent grasp of the nature of courage, ‘for it does not fall under any craft or manual of instruction (parangelia), but those who act must always look to the particular occasion, as is also the case in medicine and steersmanship’ (1104a6–10). That is why the phronimos needs nous; he has to have an eye for the requirements of courage, a sense of what is appropriate for the courageous man, which eludes capture in any formula. Lacking that, his grasp of the end is indeterminate; the inexperienced individual who believes that courage is a good and wants to be courageous, but lacks that insight into what courage requires in this or that situation, does not yet have a determinate conception of what courage is. The acquisition of that determinate conception does not consist in learning rules or formulae; it consists in the development of a style of behaviour, of the ability to respond flexibly and appropriately to an indefinite variety of situations, like the expertise of a master helmsman.²⁷ But now the role of deliberation in phron¯esis appears to be reduced to vanishing point; for the phronimos does not deliberate how to achieve a predetermined goal, any more than the helmsman deliberates how to steer the ship. Deliberation would seem to have at best a subsidiary, partly technical role; e.g. having decided that courage requires me to stand and fight now, I may have to set about improvising a weapon, a task calling for technical deliberation. It seems to me, then, that so far from tending to identify phron¯esis with excellence in deliberation, Aristotle’s theory has in fact the opposite ²⁷ See Sherman [1989], esp. ch. 2.

aristotle on the practical intellect 221 tendency; phron¯esis tends towards identification with moral nous, understood as the capacity for moral perception, leaving deliberation with a role of uncertain and at best marginal scope. Perhaps that is Aristotle’s final position; we conceive the ends of action only indeterminately, e.g. as living temperately, courageously, etc., and the task of practical thinking is to make those conceptions determinate. Perhaps that is what deliberation is, and we are misled by those passages which suggest that deliberation requires a determinately specified end.²⁸ We must, however, consider another possibility, which also has some textual support. The practical skills by which Aristotle illustrates the activity of the phronimos, medicine and navigation, have independently specifiable goals, respectively the maintenance or restoration of health and a safe and speedy voyage. What counts as doing well in these activities may not be formulable in any finite set of directions, but it is determined by the formulae ‘doing what best promotes health’ and ‘doing what best assures a safe and speedy voyage’. The concluding sentences of EN VI suggest that an independent goal for phron¯esis, answering to health in the case of medicine, is provided by the promotion of theoretical activity. That is fairly clearly the case in the final chapter of the Eudemian Ethics, and the position would fit well with the supreme value assigned to theoretical activity in EN X.²⁹ On that view the criterion of right action is always that this action best promotes the¯oria, whether directly, by furthering the agent’s own theorizing, or that of others, or indirectly, e.g. by promoting conditions, such as a stable and prosperous society, in which the¯oria can flourish. It may, however, be, as in the case of medicine, that it is impossible to give a finite formulation of all the kinds of action which have that character, and that the insight of the phronimos will not necessarily refer to the promotion of the¯oria, but will simply be the immediate insight that this or that is the right thing to do here and now. On this model the grasp of the end is simply the grasp of the truth that the¯oria is the supreme good, while the task of phron¯esis is to make determinate in action the indeterminate conception of ‘character and conduct such as best promotes the¯oria’. On the other model the grasp of the end is the grasp of the truth that the good is living excellently, while the task of phron¯esis is simply to make that indeterminate conception determinate. ²⁸ So Woods [1986], 159, and Broadie [1991], ch. 4. ²⁹ See Gauthier and Jolif [1958/9], i. 29*, ii. 2, 560–3.

222

aristotle on the practical intellect

Either model allows that moral nous has the twofold task of first grasping the end and then making determinate this (more or less) indeterminate grasp. Either model thus allows Aristotle to hold both that phron¯esis is the supreme virtue of the practical intellect and that it is concerned with ta pros to telos. The fact that the text of the Nicomachean Ethics contains indications of support for both these models may suggest that Aristotle may have found it difficult to choose between them, or even that he had not clearly distinguished them from one another. Either would be symptomatic of a tension which seems to me to pervade the Nicomachean Ethics, between the demands on the one hand to see morality as autonomous and, on the other, to elevate the theoretical intellect as the bearer of supreme value.³⁰ I do not believe that Aristotle ever satisfactorily resolved that tension, but that is another and much longer story. Afterword

For further discussion see Striker [2006]. ³⁰ Cf. Monan [1968], chs. 5 (esp. 111–15) and 8.

13 Plato on Rationality and Happiness The inclusion in the programme of this conference¹ of papers on both Plato and Socrates immediately poses the problem of which of the views on rationality and happiness attributed by Plato to the dramatic character named ‘Socrates’ are to be attributed to Plato and which to Socrates. To attribute them all to Socrates would restrict the writer on Plato to pronouncements on these topics of characters in the dialogues other than Socrates, thereby giving an artificially restricted, and hence distorted, picture of Plato’s views. The attribution of all to Plato would have the rationale that it was Plato, and not Socrates, who wrote the works in which those views are expressed; but that policy would restrict the writer on Socrates to authors other than Plato, which would in effect eliminate Socrates from the programme of the conference. Inevitably, some principle must be agreed on for attributing some theses and arguments of the Platonic character ‘Socrates’ to Socrates and some to Plato. For my own part, in this paper I assume the broadly developmental account of Plato’s writings set out in my recent Socrates.² There I distinguish a group of twelve dialogues (plus the Apology) plausibly thought of as early, and devoted largely to lively portrayal of a Socrates whose primary interests are elenctic argument and the quest for definitions of ethically significant concepts, from a ‘middle’ group which assumes theories, notably the theory of Forms, which are plausibly ascribed to Plato, but not to Socrates, and a later group of more technical and systematic dialogues where the role of expositor and leader of discussion is, for the most part, played by figures other than Socrates. While remaining broadly agnostic about which specific views can be attributed to the historical Socrates I regard the first group as Socratic in the sense ¹ Held in Buffalo in 2000.

² Taylor [1998].

224

plato on rationality and happiness

that in writing them Plato saw himself as portraying the actual individual Socrates as the ideal philosopher, and his interests and methods of argument as paradigmatic of philosophy. For the purposes of this paper, and purely as a matter of expository convenience, I restrict myself to the treatment of happiness and rationality in the ‘middle’ and later dialogues, where Socrates either speaks more for Plato than in the ‘Socratic’ group, or does not speak at all.³ Turning to our central topics, rationality and happiness, we must begin with the trite point that if happiness is to be recognized as a concept of central importance in ancient ethical theory, it has to be construed not in the predominant modern sense of content with one’s life, but as eudaimonia, the achievement of an objectively worthwhile life, the sort of life the gods give one when they are favourably disposed. Being content with one’s life is a necessary, but not (as we shall see in detail in connection with the arguments of Republic IX) a sufficient condition for being happy on that conception. Indeed our modern conception has some of these connotations; so when one wishes a couple happiness in their marriage, one wishes them not merely contentment with their life, but well-grounded contentment, i.e. that they should be pleased with their lot because things are really going well for them. But if things don’t in fact go well for them, yet they remain content with their lot, perhaps because they adapt to unfavourable circumstances, or remain incurably optimistic that things will improve, they may not have the sort of life one wished for them, but it seems unreasonable to insist that they are not happy. The objective connotations of such an expression as ‘I wish you every happiness’ (= every blessing, every good fortune) make it reasonable to translate eudaimonia as ‘happiness’; but we have to discount the dominant subjective connotations of the modern concept if we are to avoid misunderstanding of the ancient. Rational beings are beings which use reasons in forming their beliefs and in shaping their actions, and rationality I understand as the virtue specific to these aspects of life. Rationality thus conceived contrasts with irrationality; a rational agent is one in whom reasons play their proper role in the formation of belief and in the guidance of action, as opposed to an irrational agent, in whom reasons do not play their proper role. There is of ³ Donald Morrison discusses the ‘Socratic’ dialogues in Morrison [2003].

plato on rationality and happiness 225 course another use of ‘rational’, exemplified by my use of the expression ‘rational beings’, in which ‘rational’ contrasts not with ‘irrational’ but with ‘non-rational’. Most (perhaps all) non-human animals are not guided by reasons at all, but they are not thereby irrational, but rather non-rational. The rational/non-rational contrast also applies to individual psychology, since it is distinctive of Plato in the ‘middle dialogues’ to identify nonrational elements within the personality, and to present a picture of proper human development, and therefore of rationality itself, as constituted by the proper interrelation of the non-rational and the rational elements of the personality. There is also discernible in the dialogues of this ‘period’ a tendency to identify the rational soul with the real self, and thus to identify the virtue of rationality with the perfect functioning of the rational soul, independently of its relation to non-rational elements in the personality. In this paper I shall explore the tension between these two conceptions of rationality, which arise from two conceptions of the self, and consider the extent to which Plato succeeded in resolving the tension as it affects either conception. The complex psychology of the ‘middle’ dialogues, with its concomitantly complex picture of rationality, contrasts with a simpler psychology, found in at least some of the ‘Socratic’ group of dialogues, where there is a single uniform motivation, towards the achievement of the agent’s overall good, leaving no room for conflict of motivation between rational and non-rational elements in the personality.⁴ Ordinary thought, as represented by Meno and by ‘the many’ in the Protagoras,⁵ rejects this account, insisting on the commonsense position that it is possible to pursue what one knows or believes to be bad for oneself, because one is overcome by such non-rational impulses as lust or fear, but Socrates argues that all alleged cases of giving in to what one knows or believes to be bad for one are in fact attributable to misconception or miscalculation.⁶ Everyone really wants what is best for him or herself overall, but may have a wrong conception of what their good consists in, or, while conceiving it ⁴ Morrison [2003] challenges the claim that Socrates is portrayed as an egoist in the ‘Socratic’ dialogues. It is not germane to my present purpose to discuss this issue in general. For my purpose here, all that I maintain is that in the Meno and Protagoras Socrates rejects the possibility of akrasia on the strength of commitment to an egoistic psychology, and that in the Gorgias he maintains against Polus that whatever we do we do in the belief that it is best for us (see n. 7 below). ⁵ Meno 77c, Prot. 352d–e. ⁶ Meno 77c–78b, Prot. 353c–357e, 358b–d.

226

plato on rationality and happiness

correctly, may simply fail to estimate correctly how it is to be achieved, through carelessness, or haste, or overconcentration on the immediate at the expense of more distant goals. These various kinds of misconception or miscalculation are forms of irrationality, in that they are the ways in which the intellect fails to perform its proper function; we can then sum up this ‘Socratic’ view in the slogan that irrationality consists in cognitive failure. Consideration of the kinds of cognitive failure discussed in the argument with the many in the Protagoras indicates certain links (on a developmental view possibilities of transition) between this ‘Socratic’ psychology and the more complex psychology of the Republic. Suppose, for instance, that one judges that because of the risk of sexually transmitted disease one should avoid an attractive sexual encounter, but is overcome by lust and gives in to temptation against one’s better judgement. According to Socrates one simply misestimates the relative values of the pleasure in prospect from the encounter and the pain in prospect from the disease, because one tends to exaggerate immediate pleasures and pains and underestimate distant ones (Prot. 356a–c). But is one more attracted by the immediate pleasure because one overestimates it, or is it rather the case that one overestimates it because one is more attracted to it? Socrates appears to assume the former, but he gives no argument to rebut the latter alternative. But if there can be motivational sources of cognitive failure, such as failure to weigh up various alternatives adequately because one is, prior to the weighing up, already attracted to one alternative more than to any of the others, or jumping to a conclusion in advance of proper deliberation because of the attractiveness of that envisaged conclusion, then rationality will no longer reduce to correctness in practical judgement. It will indeed be true that every instance of practical irrationality is a case of cognitive failure, in that the agent will have failed to form a true practical judgement which, had it been formed, would have led to the right action. But that failure may itself be caused by an excessive attachment e.g. to some kind of pleasure, and hence the systematic avoidance of cognitive failure may require the elimination of its causes, such as excessive attachment to that kind of pleasure. A parallel situation can arise in the purely cognitive sphere. I may make mistakes in arithmetic through failure in concentration, and I may fail to concentrate because I am distracted every time a pretty girl walks past when I am working in the library. In that situation the

plato on rationality and happiness 227 best way to improve my performance in arithmetic tests may be not to spend yet more hours on arithmetical problems, but instead to cultivate an attitude of indifference to the presence of pretty girls in the library. Here reliable arithmetical competence requires that calculative skill be supported by the appropriate kind of motivation. That appears to be the situation in the Gorgias, where we find Socrates arguing against Polus (467–8) that whatever anyone does they do in the belief that it is best for them,⁷ and against Callicles (504–8) that the best state of the personality is that in which the desire for pleasure is under control, so that only desires for good pleasures are allowed to be satisfied, while desires for bad pleasures, which are compared to the desires of sick people for things which are bad for them, are restrained. While Socrates does not discuss the issue explicitly, we achieve the best fit between these passages on the assumption that the effect of bad desires is to cause the people who have them to have the false belief that satisfying those desires will be good for them. There seem, then, to be stronger and weaker versions of what is conventionally known as ‘Socratic intellectualism’. On both, an agent chooses to ϕ iff that agent believes that ϕing is, of the alternatives available, the best for the agent. On the stronger version, the desire to ϕ is always aroused by the antecedent belief that ϕing is, of the alternatives available, the best for the agent, whereas on the weaker the causal priority may be reversed, allowing it to be the case that the agent believes that ϕing is the best for him/her of the available alternatives because he/she antecedently wants to ϕ. While I know of no evidence that this distinction is explicitly drawn in the Socratic dialogues, it seems to me that the treatment of the desire for pleasure in the Gorgias is best explained on the assumption that Socrates implicitly maintains the weaker version in that dialogue. The Gorgias is thus intermediate between strong Socratic intellectualism and ⁷ I am not persuaded by Morrison’s non-egoistic reading of this passage (Morrison [2003]). The goods for the sake of which people are said to undertake unpleasant or neutral actions, wisdom, health, wealth, etc., are most naturally understood as good for the person who has them. At 468b1–4 Socrates makes the claim that whenever we walk or stand still it is because we think it better to do so (sc. than not), and therefore for the sake of the good. While indeed this sentence does not contain the explicit specification that we do so because it is better for us, Socrates immediately continues ‘So (oukoun) if we kill or exile anyone or deprive them of their property, we do so thinking that it is better for us to do those things than not’ (b4–6). Here the explicitly egoistic claim is expressed as following from the immediately preceding sentence, which should therefore be understood (as is in any case most natural) as implicitly containing that claim.

228

plato on rationality and happiness

the psychology of the Republic. In strong Socratic intellectualism desire is focused on its object by the belief that that object is the best for the agent; in the Gorgias desire can cause the belief that the object is best for the agent, and thereby motivate the agent to choose that object, whereas in the Republic desire motivates independently of evaluative belief. I leave aside the question whether that intermediate position has any chronological implications. In the Republic the optimal state of the agent is that of psychic harmony, in which each of the three parts of the soul performs the task for which it is naturally fitted. Rationality consists in the intellect’s performing its proper task, namely the direction of the life of the agent towards the agent’s good, including the direction of the functioning of the other two parts. It thus requires not merely the cognitive and ratiocinative excellence of the intellect itself, but the integrated functioning of all three parts, so that each performs its function excellently iff the others do. In examining the relation between rationality and happiness in the Republic, then, we have to consider the contribution to rationality, not merely of the intellect, but also of the other parts of the soul. Considering the intellect first, we see that its contribution to rationality is twofold. First, it has the task of caring for the welfare of the soul (i.e. person) as a whole (441e5), in contrast with the other parts, which are confined to the satisfaction of their specific interests. Second, it has, like the other parts of the soul, its own specific desires and pleasures (580d6–7), which, in the case of the intellect, are the desires for, and pleasure in, the exercise of the intellect itself. These two functions give rise to a distinction between two different activities of the intellect. The second points to the first-order role of the intellect in the search for knowledge and understanding, hence to the theoretical intellect. The first points to the second-order role of the intellect in coordinating the agent’s various firstorder desires and pursuits in order to attain a maximally good life, hence to the practical intellect. Clearly, the first-order desires of the intellect will be among those which the intellect has to coordinate in its second-order role, and the pursuit of that role may require on occasion the subordination of some intellectual desire to a desire of some other kind, as when an aspiring mathematician is so gripped by the fascination of his study that he neglects to eat or sleep, thus endangering his health; in that case it would be the function of the practical intellect to curb his mathematical

plato on rationality and happiness 229 fervour in favour of the satisfaction of those bodily desires necessary for health. That does not, of course, assume that health is unconditionally a greater good than mathematical knowledge. All that it assumes is the contrast between the long-term good of health and the short-term good of the achievement of mathematical knowledge here and now. It is left open whether the long-term good of health is itself an ultimate good, or whether its goodness is derivative from the necessity of health for the achievement of yet further long-term goods, which might, of course, include the achievement of mathematical knowledge or some other object of the theoretical intellect. It would therefore be a serious misunderstanding to construe Plato as subordinating the theoretical role of the intellect to the practical, where the latter is conceived as the attainment of the best life for the agent. For we have not yet considered Plato’s account of what determines the best life. The coordinating role of the practical intellect, its choice of which desires are to be satisfied, and which pleasures pursued, presupposes a conception of the best life, such that that life is better attained by subordinating desire D1 to D2 than vice versa. It is clear that the kinds of conception familiar from certain modern, broadly ‘subjectivist’ theories of value, which seek to identify the best life by some criterion of maximal desire-satisfaction, or of agreeable states of consciousness, are not adequate to represent Plato’s thought. For him the best life is that in which the best desires are satisfied, and in which the agent’s satisfaction with his/her life is the best kind of satisfaction, and the best desires and satisfaction are identified by metaphysical considerations, which yield the result that the first-order goals of the theoretical intellect are the goals whose achievement constitutes the best life. These considerations are spelled out most fully in Republic IX. 580c–588a, where, as part of his demonstration that the best person is the happiest (eudaimonestatos) and the worst the most wretched (544a5–8) Socrates argues that the life of the best person is the best, and that of the worst person the worst, not in respect of orderliness and fineness and excellence (in which it is unquestioned that the best person excels the worst (588a7–10)), but specifically in respect of pleasure. In the first argument (580c–583b), famously taken over by Mill in Utilitarianism to demonstrate the superiority of higher over lower pleasures, Socrates argues that the best person (identified with the philosopher) is the best judge of pleasures.

230

plato on rationality and happiness

The reason is that he is superior to adherents of the pleasures of the other parts of the soul not only in reasoning (logos) and intelligence (phron¯esis), but also in experience; he has had experience of the pleasures specific to the other parts of the soul, whereas the adherents of those pleasures have no experiences of the pleasures specific to the intellect. The main interest of this argument is to make clear the distinction between the subjective and objective conceptions of happiness mentioned at the outset of this paper; Socrates is not attempting to argue that the philosopher enjoys his/her life more than the adherents of the ‘lower’ pleasures do theirs, or that the former is better pleased with his/her life than the latter are with theirs. He states explicitly that each kind of person thinks that they have the pleasantest life (581a–582a), with the implication that all of them are content with their specific kind of life. Rather, he is trying to show that the philosopher’s preference for his/her way of life over the others is a better-grounded preference than that of the other for their ways of life, in that it is preference from an epistemically privileged standpoint. This standpoint is that of the observer who is best qualified to judge between the value of kinds of pleasures in virtue of his reason, intelligence, and experience. It is highly plausible that the judgement of the observer who is thus qualified is not intended to be ultimate, that is to say that the superiority of the preferred kind of pleasures does not consist just in its being preferred by the qualified observer. Rather, the preference of the qualified observer is grounded in that observer’s reason, intelligence, and experience. He prefers that kind of pleasure because reason, intelligence, and experience reveal to him the way in which that kind of pleasure is superior. Socrates’ concluding argument for the superiority of the pleasures of the intellect (583b–588a) may then be taken as showing the respect in which reason, intelligence, and experience recognize those pleasures as superior. This argument depends on the conception of pleasure in general as the making good of a deficiency in the organism; thus hunger is a state of bodily depletion and ignorance a state of intellectual depletion, and the pleasures of satisfying hunger and of discovery are the processes of making good the respective depletions. Socrates uses this general model to make two points, whose relation to one another is obscure. The first is that since the state of bodily depletion is painful, what we think of as bodily pleasures are for the most part not instances of genuine pleasure at all, but merely

plato on rationality and happiness 231 episodes of getting rid of pain, which are mistaken for real pleasures. The second is that whereas bodily deficiencies cannot be geniunely made good, intellectual deficiencies can be. While there are many problematic features of this argument, it is clear that the second point is supposed to result from the fact that the intellect and its objects possess a higher degree of reality than bodily-based desires and their objects. Knowledge, true belief, intelligence, and virtue possess ‘pure being’ to a greater extent than food, drink, etc., since what is stable and real and immortal ‘more truly is’ than what is unstable and mortal (585b12–c5). Since the same contrast holds between soul and body (d5), it follows that ‘What itself more truly is and is filled with things which more truly are is more truly filled than what itself less truly is and is filled with things which less truly are’ (d7–9). The soul more truly is than the body in that (a) it is immortal, whereas the body is mortal, (b) when it is freed from association with the body, it attains a changeless state of understanding of reality (Phaedo 79d), whereas the body is always subject to change; since what is immortal is the intellect, not the complex soul including the bodily-based elements of desire and spirit (Rep. 611b–612a; for discussion see below), these features coincide in assigning the highest level of reality to the intellect. The objects of the intellect are the changeless Forms and the eternal truths about them which constitute knowledge, which contrast with the impermanent objects of perception, which, in Plato’s view, manifest truth only in a qualified and hence imperfect way (see especially Rep. V. 475e–480a). The basic thought then seems to be that knowledge and understanding, and consequently the pleasure in those states, are permanent possessions of the intellect, whereas all bodily pleasures are evanescent, leading to recurrent desires for more episodes of the same kind of pleasure, e.g. another meal, another sexual encounter. If that is the thought, then the validity of the contrast is highly dubious, since it presupposes a totally static view of the intellectual life which eliminates any element of activity, such as discovery or even the recapitulation of thoughts which one has previously had. Nonetheless, it seems that this is Plato’s most explicitly stated ground for his fundamental conviction that the goals of the intellect are the highest goals, and hence that the optimal state of the embodied person is that in which the bodily-based impulses of desire and spirit find their proper function in subordination to, and thereby promotion of, those goals. And in exercising their proper function these impulses themselves achieve ‘their

232

plato on rationality and happiness

best pleasures, and the truest of which they are capable’ (586e6–587a1). Presumably the thought here is that the bodily-based pleasures contributing to the life focused on the goals of the intellect are free of the tendencies to insatiability and obsessionality which, in Plato’s view, characterize those pleasures when they are allowed to dominate the agent’s life. Hence the intellect ‘cares for the whole soul’ (441e6) by so organizing these impulses that they not only promote the proper functioning of the intellect, and thereby help to achieve the good life for the agent, but also achieve their own optimal satisfaction. In the embodied agent, then, rationality consists in the intellect’s performance of its directive function, which has two aspects, first that of harmoniously coordinating the short-term desires specific to the three parts of the soul in order best to achieve the agent’s long-term goals, and second that of securing the achievement of knowledge and understanding of reality (i.e. the Forms) as the supreme long-term goal. As we have seen, rationality requires the proper functioning of the non-rational parts of the soul. We have now to consider those parts, and specifically the questions (a) in what sense they are non-rational and (b) how the harmonious coordination of the three parts is envisaged. In Republic IV the non-rationality of the bodily appetites is a crucial premiss in the argument leading to the identification of distinct parts of the soul. Appetite and reason can be opposed, namely in the situation where appetite urges the enjoyment of food, drink, or sex, but reason urges restraint; and conflict of motivation requires distinct sources of motivation (439a–c). These distinct sources of motivation cannot be thought of simply as conflicting reasons, for and against a given course of action, for Socrates proceeds to describe appetite as in itself blind or impervious to reason. Appetite is simply focused on its specific, internal, object, e.g. thirst is in itself nothing more than desire for drink, and any considerations about whether drink is good or bad must be supplied by a separate principle, namely the intellect (ibid.). This account is presented in explicit (438a) opposition to the ‘Socratic’ account mentioned earlier in this paper, according to which all desire is for things conceived as good, i.e. as promoting the agent’s overall good. On the new account the conception of things as good belongs to the intellect, desire being limited to the conception of its own internal object, e.g. to the object of thirst as drink. Bodily appetites are thus conceived as evaluatively blind animal urges,

plato on rationality and happiness 233 which, in the situation of motivational conflict, arise from pathological and diseased conditions and drag the agent about in opposition to the commands of reason (439c9–d2). The kernel of truth in this account is the abandonment of the ‘Socratic’ thesis of uniform motivation; some things are wanted without the belief that they promote the agent’s overall good, and some in the belief, or even knowledge, that they are detrimental to the achievement of that good. But that is a feature of all wants, whatever their object, not a specific feature of bodily appetites, as our earlier example of the obsessive mathematician is sufficient to demonstrate. All desires have to be modified and coordinated in the light of the agent’s long-term goals, and all have the capacity to be in some degree recalcitrant to those processes of modification and coordination. That recalcitrance, and the conflicts of motivation which evince it, thus provide no ground for the strong characterization of the bodily appetites as non-rational which the passage expresses. The difference between the treatments of appetite and spirit (thumos) is instructive. Spirit is a complex form of motivation whose primitive form is anger, and whose conceptual core is a sense of self-worth which manifests itself in such emotions as pride, shame, and indignation. Like bodily appetites, these emotions too can be recalcitrant to reason, as in the case of Odysseus, whose anger at the outrageous conduct of Penelope’s maidservants with the suitors threatens to carry him away and has to be checked by his better judgement (441b–c). Spirit is thus distinct from the intellect, yet so closely allied to it as to be liable to be taken for an aspect of it, since when appetite conflicts with intellect, spirit always sides with the latter, as in the case of Leontius (439e–440e). The point is that the emotions which manifest spirit are themselves permeated by evaluations of what is right and wrong, fitting and unfitting for the agent. Thus Odysseus is angry because the suitors and their women are wronging him and his household, and Leontius is ashamed of himself for giving in to his morbid desire to stare at mutilated corpses. These emotions are not only conceptually highly complex, but also express rationally grounded judgements of value; hence only rational beings are capable of these emotions in their developed form, though the primitive basis of anger and aggression is common to rational and non-rational animals. Spirit, then, is non-rational only in the strictly limited sense of being capable of resistance to rational judgement; in that sense all first-order

234

plato on rationality and happiness

motivation whatever is non-rational. Appetite, on the other hand, is conceived by Plato as non-rational in a much fuller sense, in that not merely is it capable of being resistant to reason, but it is incapable of incorporating rational judgement into its own content as spirit does, and as all desire does on the ‘Socratic’ model. The primitive thumos (i.e. rage) of infants and non-human animals is value blind, but developed human thumos is permeated with evaluation.⁸ Appetite, however, seems to remain primitive on Plato’s new model, so that it can only be restrained, not permeated and thereby transformed, by reason. That picture is strongly suggested by the image of the soul-chariot in the Phaedrus, with its good and bad horses. The good horse (i.e. spirit) is ‘a lover of honour together with self-control and a sense of shame, ... needing no whip, but controlled by the word of command alone’ (253d6–e1), while the bad horse (i.e. appetite) is ‘deaf [sc. to the charioteer’s commands], barely yielding to the whip and goad’ (e4–5). That is, while spirit is sensitive to reason,⁹ appetite is insensitive to it, and can only be suppressed forcibly. This picture is amplified in the description which follows of the response of the two to erotic stimulation (254a–255a), the good horse obedient to reason and responsive to a sense of shame, the bad violent and disobedient, until it is finally tamed (tapein¯otheis 254e7) by the repeated experience of forcible restraint. This gives a profoundly pessimistic picture of the bodily appetites as not merely conceptually primitive but also as intrinsically disorderly, in that they are not simply capable of conflict with rational judgement on occasion, as spirit is, but that they tend continually to that conflict. They are, it appears, incapable of being refined or educated, but can merely be restrained until they are ‘tamed’, and even then they have a tendency to break out of control if given the slightest opportunity (256c). The ideal embodied life is that of those who are ‘in control of themselves and orderly, having enslaved that through which wickedness came to be in the soul and liberated that through which goodness came to be there’ (256b1–3). This model does not seem to allow for any genuine harmonization of the appetites with the other parts of the soul; since the bad horse is given ⁸ That is why Aristotle maintains that giving in to one’s thumos against one’s better judgement is less disgraceful than the corresponding failure in respect of bodily desires (EN 1149a24–b2). ⁹ The Greek κελεύσματι ... καὶ λόγῳ, translated ‘word of command’, has the implication of rational command.

plato on rationality and happiness 235 no positive contribution to the functioning of the soul, it can hardly be said to be coordinated with the other parts. At best it ‘ceases from its wickedness and follows the foresight of the charioteer’ (254e6–7), i.e. it stops being disruptive. The appetites are not so much non-rational as irrational, since they are essentially opposed to the promptings of reason. And as rationality requires the elimination of irrationality, on this model the ideal of rationality is to approach as closely as possible to the elimination of the bodily appetites. The ideal of the harmonious integration of the parts of the soul, in which each performs its proper function and achieves its truest pleasures (see above), has altogether disappeared. Bodily appetites can be totally eliminated only if the body itself is eliminated; hence the pessimism of the soul-chariot image in the Phaedrus points towards the more thorough-going dualism of the Phaedo. In the latter dialogue bodily appetites and pleasures belong not to the soul but to the body, the soul being identified with the rational soul, whose relations to the body are described via eloquent metaphors of hindering and imprisonment. The body ‘fills us with lusts and desires and fears and fantasies of all kinds and a great deal of nonsense’, so that it is impossible to think straight (66c); the desires, pleasures, and pains of the body nail the soul to the body, making it accept as true whatever the body says (83d), whereas the enlightened person regards the senses as a source of deception and takes as real only what is grasped by pure thought (83a–c). The real self is the intellect, which can achieve its full potential only when it is freed from the shackles of the body; hence the true philosopher ‘practises nothing else than dying and being dead’ (64a4–5). Rationality and hence happiness are impossible in the embodied state, but only, if at all, after the complete separation of soul from body when ‘God himself shall release us, and being thus pure, through separation from the body’s folly, we shall probably be in like company, and shall know through our own selves all that is unsullied ... because never will it be permissible for impure to touch pure’ (67a–b, trans. Gallop [1975]). The message that true rationality and happiness are impossible for the embodied agent cannot be dismissed as an aberration peculiar to the special dramatic situation in the Phaedo, where Socrates’ departure from embodied life is imminent. It recurs in the digression in the Theaetetus, where Socrates connects it with the doctrine that the material world, in contrast to the divine nature, is intrinsically contaminated with evil: ‘It is impossible for

236

plato on rationality and happiness

evils to be eliminated ... since there must always be something opposed to good; yet it cannot be situated among the gods, but it necessarily pervades mortal nature and this world. For this reason one must try to escape as quickly as possible from this world to that. Escape consists in becoming as like god as possible, and becoming like god is becoming just and holy together with wisdom’ (176a–b). This might be read in a spirit of bland reductionism, according to which escaping from the material world and becoming like god are mere metaphors for becoming virtuous, leaving us with the familiar idea that since the divine is the paradigm of goodness, the person who most resembles the gods is the virtuous person. But while Socrates is here certainly maintaining that the way to become like god is to establish the rule of reason, the most godlike part, in one’s soul and thereby become completely virtuous (which is essentially the Republic’s account of the nature of rationality and virtue), the purely reductionist reading loses the force of the strongly evaluative dualism which contrasts the perfection of the divine with the inevitable contamination of the material world. Since the material is as such a source of disorder and chaos (as in the Timaeus and Statesman¹⁰), involvement with the material world cannot but hinder the rational soul in its task of seeking union with the perfect changeless rational order of the world of Forms. The embodied soul, like the material world as a whole, is capable of receiving rational order, however imperfectly and temporarily, and rationality has the task of imposing that order, whether at the cosmic level, as in the activity of the Demiourgos in the Timaeus, or at the level of the individual agent, but the imposition of that order, which is embodied rationality, is merely an approximation to the perfect rationality of the disembodied soul. Hence ‘escape from this world to that’ is no mere metaphor for becoming virtuous; rather, the achievement of virtue by the embodied agent is a stage on a journey which will be completed only when the rational soul, freed from its bodily shackles, attains its true perfection in union with the Forms (Tim. 90a–d).¹¹ Prima facie, then, we have an opposition between the relatively optimistic picture of embodied rationality in the Republic, where the non-rational elements are harmoniously integrated with the intellect to the maximal satisfaction of all, and the pessimism of the Phaedo, Phaedrus, and Theaetetus, ¹⁰ Tim. 53a–b, Pol. 273a–d. ¹¹ I am here indebted to Annas [1999], ch. 3. See also Sedley [1999].

plato on rationality and happiness 237 where rationality requires escape from the body, which can be fully achieved only after death. Yet that appearance is misleading to the extent that even the optimism of the Republic seems to give at best a provisional and approximate account of rationality. We should first recall the point made above, that the tripartition of the soul in Republic IV itself depends on the picture of the appetites as reason-blind instinctual drives which is the core of the pessimism of the ‘dualistic’ dialogues. It is hard to reconcile that picture with the Republic’s own account of individual s¯ophrosun¯e as the harmonious agreement between the rational and non-rational parts of the soul that the former should be in control (442c10–d1), and with the result that that rational control guarantees the other parts of the soul their best and truest pleasures (586e4–587a1). There is thus internal to the Republic itself a tension between an optimistic and a pessimistic picture of human rationality, a tension which, as Bernard Williams points out,¹² has its counterpart at the political level. Secondly, Rep. 611b–612a distinguishes the true, i.e. the rational soul, which alone is immortal, from the tripartite soul, which is an unnatural amalgam of the true soul and the accretions, namely spirit and the bodily appetites, which have grown on to it through its physical embodiment, like the shells, stones, and seaweed which encrust the sea-god Glaucus and conceal his true form. On this analogy, not only is the tripartite soul not identical with the real self, it is not strictly speaking a unitary substance at all. Insofar as the soul is the self, the tripartite soul is not the human soul; the non-rational parts are not parts of a human soul, any more than the barnacles and weed which cover the sea-god are parts of him. What they are parts of is an adventitious composite which conceals his true nature; just so the tripartite soul is not the real self, but a composite which conceals the real nature of that self. Hence the optimistic picture of rationality as the health of the soul (444d–e), i.e. the harmonious coordination of rational and non-rational elements, is at best an approximation to true rationality and true health. It is the best possible state of the unnatural amalgam; but the true self can only be healthy when it is freed from the amalgam and thereby enabled to exercise its rationality unfettered. Socratic psychology provides no room for a dichotomy between the real self, identified with the rational self, and the embodied self. The embodied ¹² Williams [1973].

238

plato on rationality and happiness

agent has a uniform motivation toward his or her overall good, rationality consists in the achievement of a correct conception of that good, and rationality thus conceived is, assuming a minimum level of goods, such as health, which are outside the agent’s control, sufficient for eudaimonia.¹³ In the dialogues which I have discussed in this paper Plato abandons the assumption of uniform motivation in favour of a complex psychology which yields a correspondingly complex conception of rationality, requiring not merely the correct conception of goods to be pursued, but also the integration of non-rational elements in the personality with the directing function of the intellect. This new conception of rationality, which is identified with the health of the complex soul, is, subject to the same qualification, also held to be sufficient for eudaimonia. But the heightened psychological realism of the new conception is purchased at a high price, that of the integration of the agent. For in that conception the embodied agent is split off from the real self, which is identified with the intellect. The supreme good of the real self is rationality itself, i.e. its perfect functioning in theoretical understanding of reality; but that good, which is true eudaimonia, it can achieve fully only when freed from the entanglements of the body. In its embodied state the best it can do is so to modify the non-rational elements that the latter play their proper role in an embodied life which is itself directed to a goal, theoretical understanding, the complete achievement of which is bound to be frustrated by the embodied state itself. Plato is torn between the view that the non-rational elements in the soul should as far as possible be suppressed, as in the Phaedo and Phaedrus, and a more positive view, which has some affinities with Aristotle’s theory of ethismos, in which they can be trained to provide a pre-rational appreciation of the good and the beautiful.¹⁴ On either view true rationality and hence true eudaimonia are beyond the reach of the embodied agent. Nor should that surprise us. For eudaimonia is living well, and for Plato embodiment is in the last resort an obstacle to the rational agent’s (i.e. the rational soul’s) living well. The approximation to living well which is the harmony of the tripartite soul ¹³ In allowing the qualification expressed in this sentence I side with Vlastos [1984] and with Brickhouse and Smith [1987] against Irwin [1986] and [1995], ch. 4, esp. 58–60. Socrates’ insistence on the paramount importance of the state of one’s soul, which he assumes to be within one’s control, and on the comparative worthlessness of other reputed goods (Ap. 28b, 29d, Prot. 313a) has the effect of considerably narrowing the gap between these opposed positions. See ch. 10 of this volume, p. 160 n. 19. ¹⁴ See Rep. 401a–403c, and Diotima’s account of erotic education in Symp. 209e–212a.

plato on rationality and happiness 239 is not in fact the life of a unified being, but the best that can be achieved by a rational being forcibly yoked to beings of a nature alien to it. Once Plato had shattered the unified Socratic conception of an agent at once rational and embodied, it was a task for Aristotle and later philosophers to put Humpty Dumpty together again.¹⁵ ¹⁵ A symptom of the difficulty which Aristotle found in this task is his hesitation in EN X. 7 on whether the human being (anthr¯opos) is identical with the theoretical intellect (nous) or with the embodied complex (suntheton) which includes the intellect. For a fuller discussion see Yu [2003]. Inwood [2003] and McPherran [2003] discuss similar issues in later philosophers.

14 Pleasure: Aristotle’s Response to Plato Aristotle discusses pleasure in the context of lively debate both about its nature and about its value, of which we have evidence in his own writings and those of others, above all of Plato.¹ For him the question of value predominates. His treatment of the topic belongs to the ethical treatises, not to his discussion of the soul and its faculties, and while both principal discussions include accounts of the nature of pleasure those accounts are subordinated to his evaluative interests; his primary concern is to give pleasure its proper place in his account of the best form of human life, and it is because that concern requires a proper understanding of what pleasure is that the account of its nature engages his attention. Aristotle’s survey of current views on the value of pleasure reveals a wide range of conflicting opinions, from Eudoxus’ identification of pleasure with the good at one extreme to, at the other, the denial that any pleasure is good, either in itself or incidentally. This diversity does not lend itself to the eirenic project mentioned in EE I. 6 of showing that every opinion possesses some truth. While some apparent conflicts can be reconciled some theses have simply to be rejected, the best that can be done for them being an explanation of how people have come to hold them (EN 1154a21–b20). Aristotle himself is firmly committed from the outset to the view that pleasure is an inseparable attribute of the best life. The EE begins from the unargued claims (a) that the good is eudaimonia and (b) that eudaimonia is the finest, best, and pleasantest of all things (1214a1–8). In EN thesis (a) is first declared to be established by universal consent (1095a17–20) and later established by an argument to the effect that eudaimonia alone fulfils the formal criteria of the good, viz. those of being sought for its own ¹ For discussion see Gosling and Taylor [1982] (hereafter referred to as G&T).

pleasure: aristotle’s response to plato 241 sake alone and of being self-sufficient (1097a15–b21). A further argument leads to the substantive account of human good as ‘activity of the soul in accordance with excellence’, i.e. the excellent realization of specifically human capacities (1097b22–1098a20). This is confirmed by a number of arguments to the effect that that kind of activity possesses some agreed marks of the good life; one such mark is that the life of excellent activity is intrinsically pleasant (1099a7–28), leading to the position which is the starting point of EE, that the good life is finest, best, and pleasantest (1099a23–31). The thesis that the life of excellent activity is intrinsically pleasant seems partly to rest on the basic intuition that a wholly satisfactory life must be pleasant (otherwise it would lack a feature which counts significantly towards its being worthwhile), but is also supported by a thesis which is less obviously part of common evaluative consciousness, viz. that ‘the person who does not enjoy fine actions is not good’ (1099a17–18). Though Aristotle does say that no one would call a person just unless that person enjoyed acting justly (a18–19), it is not obvious that that is an unbiased report of actual contemporary Greek usage, independent of his own substantive view, derived from his account of the psychology of virtue and the process of habituation necessary to inculcate virtue, that it would not be correct to call anyone virtuous who did not enjoy acting virtuously. Given this prior commitment to the intrinsic pleasantness of the good life, Aristotle’s discussion of pleasure has two main functions. First he has to rebut arguments which purport to establish that pleasure cannot contribute to the good life, either because pleasure is bad, or because it is not good. This defensive strategy will lead him to give an account of pleasure insofar as he seeks to show that those who expel pleasure from the good life do so because they have mistaken views of what pleasure is. Further, his own account of pleasure should provide further positive arguments in support of the thesis that pleasure is at least inseparable from the good life. The contents of the discussions of pleasure in EN bear out these suggestions. As is well known, the work contains two treatments of the topic, VII. 11–14 and X. 1–5, with a certain degree of overlap in subject matter, but no cross-references, either explicit or implicit, in either direction. They are clearly two independent discussions which owe their position in the text of EN to the hand of an editor, possibly Aristotle himself but more plausibly a later redactor, and since EN VII is one of

242

pleasure: aristotle’s response to plato

the books common to both versions of the Ethics, it is generally assumed that its discussion of pleasure belongs originally to EE, and further that that discussion is earlier than the one in EN X. (I shall return to the question of the temporal relation of the two discussions at the end of this paper.) The discussion of book VII is very largely devoted to the examination, leading to the rebuttal, of arguments hostile to pleasure, to which is appended a brief statement of Aristotle’s positive view. In book X, by contrast, the positive view is set out much more elaborately, while the anti-hedonist arguments appear as arguments against a thesis which is absent from VII, namely Eudoxus’ thesis that pleasure is the good. Most of these arguments are rebutted, but Eudoxus’ thesis is not explicitly endorsed, and the ensuing discussion appears to point rather to pleasure’s being an inseparable aspect (or perhaps accompaniment) of the good than to its being the good itself. I shall not attempt to deal with all the issues raised in these complex passages, confining myself instead to a single central issue. Both passages contain criticism of a certain view of the nature of pleasure, seen as foundational to many of the arguments hostile to pleasure which Aristotle is attempting to rebut. This is the view of pleasure as a perceived process of replenishment of a natural lack, and thereby a return from a state of deficiency, where something necessary to the proper functioning of the organism is lacking, to a state of equilibrium and thus of normal function. The terminology in which this view is stated is not always so specific as that which I have just used. At 1152b12–14 the first argument for the position that no pleasure is good is stated as follows: ‘every pleasure is a perceived coming into being (genesis) of a natural state, but no coming into being is of the same kind as its completion, e.g. no process of building is the same kind of thing as a house’. While ‘perceived coming into being of a natural state’ is less specific than ‘replenishment of a natural lack’, the discussion of 1152b25–1153a7 strongly suggests that replenishments of lacks are at least paradigm cases of the pleasures described by the theory; the examples mentioned include cases in which nature is ‘replenished’ (1153a2–6), which are contrasted with cases like pleasure in thinking, where there is neither distress nor desire, ‘because one’s nature is not deficient’ (1152b36–1153a2). (The genesis terminology and its application to cases of physiological deficiency recall Plato, Phil. 53c–54e, where the view of certain clever people that pleasure is always a process of coming

pleasure: aristotle’s response to plato 243 to be, never a state of being, is applied to the pleasures of eating and drinking to reduce to absurdity the claim that the good life is the one devoted to those pleasures.) Similarly in book X we have at 1173a29–b20 a series of arguments against the theory that pleasure is a process of change (kin¯esis) and a coming into being, arguments which have some overlap with those from book VII just mentioned. Here too the only kind of process mentioned is that of replenishment, the theory is said (b13–15) to be based on consideration of the pleasure and distress associated with food and drink (troph¯e), and the concluding argument is that the pleasures of thought etc. which involve no distress cannot be processes of coming to be ‘since there has been no lack of which there could come to be replenishment’ (b15–20). The evidence therefore suggests that the account of pleasure as the perceived coming into being of a natural state is not an alternative theory to that of pleasure as the perceived replenishment of a natural lack, but merely a less specific designation of that very theory. The theory is familiar from well-known passages of Plato, notably Gorg. 494–7, Rep. 585d–e, and Phil. 31–2. The paradigm cases are those of pleasures in the satisfaction of bodily-based appetites, especially those for food, drink, and sex. The Gorgias passage illustrates the simplest stage of the theory. Bodily appetite is either identified with or seen as arising from bodily deficiency, which is experienced as unpleasant. This unpleasant consciousness (lup¯e) prompts the agent to make good the deficiency, and the process of filling up the deficiency (anapl¯er¯osis) is experienced as pleasant. There are a number of unclarities even at this stage. First, it is unclear whether distress and pleasure are literally identified with physical deficiency and physical replenishment respectively, or are thought of as effects of those physical conditions. It is implied, though not explicitly stated, that pleasure and distress involve awareness, but it is not clear whether that is awareness of physical deprivation and replenishment, or awareness of pleasure and distress themselves. If the former, is the thought that (a) pleasure and distress are the awareness of those physical conditions, or (b) that those conditions, given that the agent is aware of them, are pleasure and distress?² A further set of problems arises from the assimilation of sexual desire to the model ² If b is the case, then awareness of physical deprivation and of replenishment is awareness of (respectively) distress and of pleasure.

244

pleasure: aristotle’s response to plato

of hunger and thirst. In the latter two cases bodily desire can plausibly be seen as a response to physiological deficiency, dehydration, lack of protein, carbohydrate, etc., and physiological deficiency can be identified as such by its reference to bodily functioning; without appropriate solid and liquid nourishment bodily function is impaired, and pleasure is (or is a response to) the process of restoring proper functioning. But sexual desire cannot plausibly be seen as a response to a physiological deficiency which impairs bodily functioning; either it is simply a response to a lack of sexual pleasure, in which case the physiological model loses its explanatory force, or else it is a response to a lack of sexual activity, which is itself conceived of either as a precondition of proper bodily functioning, or (perhaps more plausibly) as part of that functioning.³ The application of the theory to sexual pleasure can thus be seen as an extension (in one way or another) from the cases where it has the clearest application. Given one kind of extension, desire is seen in both kinds of case as a response to a deficiency, and pleasure as bound up with the making good of that deficiency, though it is only in the primary cases that the deficient items can be identified as constituents of the organism, whose mutual adjustment is a precondition of correct functioning. Given the other kind, desire is a response to deficiency in some cases and excess in others, and pleasure is a response to the restoration of equilibrium in both. A further extension is exhibited by the application to mental pleasures in Rep. IX; as hunger and thirst are states of bodily deprivation, and bodily pleasures are the making good of those lacks, so ignorance or lack of understanding can be seen as states of mental deprivation, and the making good of those lacks in learning as mental pleasures. Here too the question arises of what it is that is lacking; is it a precondition of proper mental functioning, which would assimilate the mental case to those of hunger and thirst, or is it that functioning itself, which would rather assimilate it to the sexual case? If the acquisition of knowledge or understanding is thought of as a prerequisite of the exercise of those faculties, then the model of hunger and thirst is appropriate. But if the soul is seen as lacking understanding of ³ A more plausible theory of sexual pleasure would result if the specific notion of deficiency were replaced by the more general notion of imbalance. Sexual desire could then be seen, not implausibly, as a response to an excess of some physiological component (in modern terms a hormonal imbalance, e.g. an excessive level of testosterone), and pleasure a response to the restoration of the balance via the discharge of the excess in the process of copulation. In this case too we have an extension of the original model, though in a different direction from that suggested in the preceding paragraph.

pleasure: aristotle’s response to plato 245 some subject matter, then making good that lack would appear to consist in coming to have that understanding, which is not a process identifiable as completed prior to the exercise of understanding. One might think of the dissatisfaction of someone trying to make sense of a complex pattern, say a visual or musical pattern; that dissatisfaction is alleviated when and only when one has come to see the pattern, and one has come to see it when and only when one has seen it. So having come to see it is not a prerequisite of seeing it. There are, then, some unexpected complexities in the replenishment model of pleasure, not only in its extension to mental pleasures, but also in its application to those very cases of bodily-based appetites which originally suggest it. But even setting these complexities aside, the model has a general feature which makes it particularly problematic for someone who, like Aristotle, seeks to assure the place of pleasure in the good life. For according to the model pleasure is seen as something essentially remedial, as bound up with (to use a deliberately vague expression) the process of getting rid of an imperfect and undesirable state. It seems at best an alleviation of the troubles of the human condition; consequently it is hard to see how pleasure thus conceived could have any role in the ideally good life, much less be a necessary feature of it. Hence it is not surprising to find Aristotle in EN VII citing arguments which rely on this model in support of the theses that no pleasure is good (1152b12–15) and that pleasure is not the good (1152b22–3). A possible response to this objection would be to accept that the constant fluctuation of deficiency/desire and replenishment/pleasure is a necessary feature of human life. The ideally good life, envisaged as free from deficiency and its associated distress, is not a possible human life, though perhaps it might be possible for some other creature, such as a god (provided that the god is conceived non-anthropomorphically). Yet traditionally the life of the gods was regarded as blessed (makarios) in the highest degree, and the blessed life as supremely pleasant (EN 1152b5–7). If the replenishment model is accepted, either those traditional beliefs would have to be abandoned, or the model would have to be construed as a model of human pleasure only, and divine pleasure conceived as something altogether different. On either account the defence of pleasure which the model allows is comparatively weak; pleasure is not something to be hoped for or aspired to for its own sake, but is at best something to be welcomed as an amelioration of our

246

pleasure: aristotle’s response to plato

imperfect condition.⁴ And even that welcome, it would seem, should be qualified; for if pleasure is essentially remedial, arising when a deficiency is remedied, would we not do better to avoid those deficiencies in the first place than to seek to remedy them? As Socrates argues against Callicles, it would surely be the height of irrationality to seek to have an itch in order to have the pleasure of scratching it (Gorg. 494c–d), yet on the replenishment model that ought to be a paradigm of a pleasure. Even if one distinguishes necessary pleasures, i.e. those arising from deficiencies whose satisfaction is necessary for human life, from unnecessary, the tendency of the model will be to favour asceticism. For the strategy recommended by the model will be to remedy only those deficiencies which cannot be avoided, and any deficiency whose satisfaction is not necessary for survival can, it seems, be avoided by eliminating the desire which generates the deficiency. Thus to someone subject to sexual desire lack of sexual activity is perceived as a deficiency; but if one ceases to want sex one no longer feels the lack of it. The replenishment model is not, then, well adapted to assure the place of pleasure in the good life. It also has some independent defects. First, as Plato had pointed out (Rep. 584b, Phil. 51b–52b), many kinds of pleasures are not preceded by episodes of desire. So I can enjoy e.g. the smell of a rose, or a beautiful view, or memories of childhood holidays, without previously having desired to have, or felt, any lack of those experiences. The scent is wafted through the open window, the view is disclosed at the crest of the hill, the pleasant memories simply occur to me, all without any antecedent longings. Here the phenomenology gives no support to the replenishment theory. Of course that does not refute the theory, since not all deficiencies make themselves apparent in desire; I may suffer from vitamin C deficiency without any desire to take vitamin C (or indeed without any awareness of the deficiency). But if the replenishment theory is not supported by the phenomenology in these cases, the onus must be on the proponent of the theory to show why the cases are best described in terms of the theory. In the vitamin C example physiological theory enables us to identify the deficiency independently of phenomenology; proper bodily functioning requires a certain level of vitamin C, and failure to maintain that level reveals itself in various symptoms, which may have nothing to do with ⁴ Cf. Glaucon’s account of justice in Rep. II.

pleasure: aristotle’s response to plato 247 desire for substances containing vitamin C. But in the cases of pleasure cited above nothing analogous allows us to identify any unfelt lack; why should we suppose that my delight in the scent of the rose is prompted by the making good of a deficiency of which I was unaware, however that deficiency is to be identified (on the problem of identification see below)? One might perhaps adopt that theory as the best explanation, or, at the extreme, in default of any other possible explanation; both strategies require examination of possible alternatives, and a more careful examination of the replenishment model itself. Earlier we saw that the most plausible application of that model to sexual pleasure was the following; the model postulates that sexual activity is necessary to a worthwhile life, and consequently explains the experience of the lack of it as unpleasant, and the experience of the making good of that lack as pleasant.⁵ In this case the notion of lack or deficiency is derivative from that of worthwhile activity, indeed the lack is precisely the lack of a worthwhile activity. Now while we should not expect any single answer to the question ‘What makes an activity worthwhile?’, it seems undeniable that one feature which makes at least some activities (to some extent) worthwhile is that those activities are enjoyable, and equally undeniable that sex is worthwhile at least partly because it is enjoyable. It follows that we do not give a complete account of sexual pleasure by describing it as the making good (or as arising from the making good) of a lack of sexual activity. For the perception of the absence of sex as a lack presupposes that sex is seen as worthwhile, and it is seen as worthwhile at least partly insofar as it is seen as enjoyable. So even in such a central case as that of sexual pleasure, the deficiency/replenishment model presupposes a prior account of what it is that makes sex enjoyable. This is so even in the case of a sort of pleasure which is typically preceded by an episode of felt desire. The necessity of such an account is even more clearly apparent in the kinds of case considered above, where there is no preceding desire. And given such an account, the positing of an unfelt lack appears quite otiose. For ⁵ Here the postulation is that of the theorist who is seeking to explain why humans and other animals find sex pleasant. The subject, whether human or non-human, which experiences pleasure need not (and in the non-human case presumably does not) itself endorse or even entertain that postulation. Instead the theory has to include the further postulation of a natural nisus towards a life satisfactory or worthwhile for creatures of that kind, such that the lack of conditions necessary for that life is experienced as unpleasant.

248

pleasure: aristotle’s response to plato

what is that supposed lack a lack of? Either it is just a lack of pleasure, in which case the ‘theory’ reduces to the tautology that pleasure is the making good of a lack of pleasure, or else it is a lack of whatever it is that makes e.g. smelling the rose pleasant (alternatively, a lack of whatever that pleasure consists in). But in that case the explanatory work is being done by that account, whatever it is, not by the posited lack. Aristotle’s own response to the deficiency/replenishment model can be seen as making this point. That model now appears to have got things the wrong way round. It seeks to give a general account of pleasure via the notion of making good a deficiency, but in most cases the deficiency can be specified only as a deficiency of pleasure, or as a deficiency of whatever features it is in virtue of which things are pleasant. In a few cases indeed, but only a few, the deficiency can be specified independently of pleasure, e.g. hunger and thirst, to which we might add what might be classified as addictive pleasures, e.g. the pleasures of nicotine, alcohol, and other drugs. In the case of the first two effective bodily functioning demands an appropriate level of nutrition; in the addictive cases an acquired habit leads to a situation in which a given level of the drug is demanded. In either type of case deficiency is experienced as a craving, and the making good of the deficiency is experienced as the pleasure of satisfying the craving. It is important to distinguish the peculiar nature of addictive pleasures from non-addictive pleasures in the same kind of object; the pleasure of satisfying a craving for alcohol is distinct from the simple enjoyment of alcoholic drinks, as is made clear by the fact that either kind of pleasure may be experienced without the other. Addictions, and consequently addictive pleasures, are pathological, a manifestation of the malfunctioning of the organism. In fact there appear to be comparatively few non-pathological pleasures which the deficiency/replenishment model actually fits; in addition to the pleasures of satisfying hunger and thirst the pleasure of getting warm when one has been cold is the one which springs most readily to mind. This suggests that the theorist who wishes to defend the status of pleasure as a necessary constituent of the good life has two options. Such a theorist, we assume, will not seek to defend pathological pleasures. They have no role in the good life, and the explanation of the way in which they are pleasant, and why they are, will presumably reveal them to be a special case, perhaps counted as pleasures because of some

pleasure: aristotle’s response to plato 249 resemblances to the normal or standard case. For the standard case itself one alternative open to the theorist is that of recognizing two irreducibly different kinds of pleasure, one of which fits the deficiency/replenishment model, and has comparatively few members, the other, which does not fit that model, containing all the rest. The other alternative is that of devising a single account which applies to all. Aristotle chooses the second alternative. On this account the factor common to all pleasures is the exercise of natural capacities in appropriate conditions. The basic idea is that pleasures are appropriate to the different species of animals; every species has capacities for activities which constitute its specific life, and when those capacities are exercised (i.e. when the corresponding activities are undertaken) in the appropriate conditions their exercise is pleasant to the individual member of the species (EN 1176a3–8). So since it is constitutive of the life of some kinds of dogs to chase hares, when the conditions for chasing are appropriate (e.g. both dog and hare are healthy, the ground is not too rough, the scent is good) then the dog will enjoy chasing the hare.⁶ Human life is constituted by the capacities shared with animals, growth, reproduction, nutrition, perception, and locomotion (the first three also shared with vegetables), with the addition of the specifically human capacity for thought; hence some human pleasures, notably those in food, drink, and sex, are kinds of pleasures which animals also enjoy, while intellectual pleasures are specific to humans. Pleasures are common only at the generic level; thus animals of every kind enjoy food and sex, but each kind of animal enjoys its specific kind(s) of food, while rejecting the kinds which other species enjoy, and each kind enjoys sex only with members of its own species. Again, non-human animals enjoy the exercise of perceptual capacities only instrumentally (e.g. the lion enjoys the scent of the deer as a sign of its approaching meal (EN 1118a16–23)), whereas humans enjoy them intrinsically; i.e. aesthetic enjoyment is a specifically human kind of pleasure. This analysis has the advantage of applying as well to cases which fit the deficiency/replenishment account as to those which do not. Taking our previous cases as examples of the latter, the exercise of the sensory capacities (smell and sight) and of memory are constitutive of specifically ⁶ It is not clear whether this analysis would commit Aristotle to agreeing with the hunting lobby that hares etc. enjoy being hunted; if not, it must presumably be because fear inhibits the pleasure which the prey would otherwise feel in exercising its specific capacity for flight.

250

pleasure: aristotle’s response to plato

human life, and will therefore be in appropriate circumstances pleasant. (On the problems of identifying appropriate circumstances see below.) The principal problem in the application of the deficiency/replenishment analysis to intellectual pleasures was that that analysis applied straightforwardly to the pleasure of acquiring knowledge or understanding, but not so straightforwardly to the pleasure of exercising those capacities. That problem now disappears, since both kinds of intellectual activity are characteristically human. Turning to the cases of the pleasures of satisfying hunger and thirst, the crucial point is that those drives belong to a natural pattern of animal activity, which, since humans are a species of animals, is ipso facto part of the natural pattern of human activity. Unlike addictions, hunger and thirst are not pathological conditions; on the contrary, they are essential elements in the proper functioning of the animal’s capacities to seek and acquire nourishment. There is something wrong with an animal which is not hungry when it is short of food or not thirsty when dehydrated (if the cat refuses to eat or drink for days on end you take it to the vet), and also with an animal which wants food when it is not hungry or drink when it is not thirsty. Satisfying one’s hunger and slaking one’s thirst are perfectly genuine cases of pleasure; the crucial point for Aristotle is that they count as such insofar as they fall under the general classification of the appropriate exercise of natural capacities. The point that on the Aristotelian analysis the pleasures of satisfying hunger and thirst are genuine pleasures is worth emphasizing, since it brings out a divergence between Aristotle and Plato. At Rep. 583c–585a Socrates argues that the great majority of bodily ‘pleasures’, viz. those which involve the replenishment of some deficiency, are not in fact pleasures at all; they are rather processes of escape from distress, which people mistake for genuine pleasures through lack of experience of the latter. Again at Phil. 44a–b a similar view is attributed to ‘people who are said to be very expert about nature’, and is given qualified approval by Socrates. Plato appears to accept this argument as showing that genuine pleasure must be free of any element of distress, since later in the Philebus (51e–52a) Socrates admits the pleasures of learning among genuine pleasures subject to the proviso that ‘they do not involve any actual hunger for learning, and that there is no distress from the start through hunger for knowledge.’⁷ The claim that ⁷ Trans. Gosling [1975].

pleasure: aristotle’s response to plato 251 most bodily ‘pleasures’ are not in fact instances of pleasure, since in those cases the process of getting rid of the distress arising from bodily deficiency is mistaken for genuine pleasure, appears to embody a confusion. Even if it is granted (i) that the state of bodily deficiency (e.g. hunger) is unpleasant and (ii) that the state of having got rid of that deficiency (e.g. having satisfied one’s hunger) is neither pleasant nor unpleasant, it does not follow that the process of transition from the state of deficiency to the state of repletion is not really pleasant.⁸ Aristotle’s analysis allows him to escape this error; the process of transition from deficiency to repletion is the process in which the nutritional capacity is appropriately exercised, and is therefore standardly pleasant. Equally, the analysis points away from the mistaken belief that states of repletion (as distinct from processes of replenishment) are neutral between pleasure and pain. For those states are as much part of the pattern of the exercise of natural capacities as are the processes which give rise to them, and are therefore, like the latter, naturally experienced as pleasant. The attractions of this analysis, given Aristotle’s project of vindicating the claims of pleasure to a place in the best life, are obvious. Since the best human life consists in the excellent exercise of specifically human, i.e. rational capacities, it follows immediately from the analysis that that life must be, not merely pleasant, but intrinsically pleasant, i.e. pleasant just in virtue of being the kind of life that it is; ‘their life (i.e. the life of those who exercise rational capacities excellently) has no need of pleasure as a sort of adornment, but it has pleasure in itself’ (EN 1099a15–16). At the same time the wide diversity of human capacities and activities, answering to a corresponding diversity of human interests, gives a ready explanation of the diversity of kinds of pleasure, and of the observation that what is pleasant to one person may be unpleasant or neutral to another. Capacities are developed to different degrees in different individuals; so someone with a gift for mathematics will naturally enjoy doing mathematics, whereas for someone whose mathematical capacity is undeveloped the activity will be burdensome. ⁸ In fact both (i) and (ii) are highly questionable. Not all cases of bodily deficiency are unpleasant, first because in some cases the person who has the deficiency is not aware of it, and secondly because even when one is conscious of the deficiency, moderate degrees of hunger and thirst, particularly when one expects to satisfy them within a fairly short time, need not be experienced as unpleasant. And states of having satisfied desires grounded in bodily deficiencies, such as being replete having been hungry, or being warm having been cold, are paradigms of pleasant states.

252

pleasure: aristotle’s response to plato

Though appealing, the analysis faces some major problems. Perhaps the most pressing is this. Pleasure is or arises from the exercise of natural capacities in good or appropriate conditions, but it is problematic whether it is possible to identify appropriate conditions without including among them the condition that the activity is pleasant to the agent. If so, the analysis is viciously circular. Some of Aristotle’s formulations do indeed expose him to this criticism. Thus he asserts that ‘To each person that kind of thing is pleasant which he is called ‘‘a lover of ’’, e.g. a horse to the horse-lover, and a spectacle to the lover of spectacles’ (EN 1099a8–10). Of course it is a truism that people enjoy the kind of things they are keen on, but it is a truism because it is at least a necessary condition for being keen on something that one should enjoy the activities appropriate to that thing. And if being keen on a certain activity is one of the conditions necessary for undertaking that activity in good or appropriate conditions, and thereby for taking pleasure in that activity, then it follows that pleasure itself is one of the necessary conditions for pleasure. Aristotle might reply that it is possible to identify good or appropriate conditions independently of pleasure, by appeal to the notion of the natural functioning of the species. Thus if a certain species is naturally adapted to eat a certain kind of food, he might claim that a healthy individual of that species will enjoy nutritious samples of that food, provided that there are no interfering factors such as cold, fear, or contamination of the food; this view is suggested, though not explicitly spelled out, at EN 1176a3–9. But that claim is open to counter-examples; a diet might be perfectly nutritious but unpalatable, perhaps because the ingredients are lacking in flavour, or because it palls through lack of variety. Aristotle might attempt to deal with these counter-examples by appeal to his principle that, in cases of pleasure (as generally in cases of perceptual appearances) the criterion of truth is how things seem to the spoudaios, the person in good condition (EN 1099a21–4, 1113a29–33, 1166a12–13, 1173b22–5, 1176a16–19). So, just as we count honey as really sweet because it tastes sweet to the healthy person, even though it may taste bitter to someone who is ill, we judge the pleasantness of the nutritious diet by how it tastes to the person whose appetite itself is healthy, not pampered or jaded. But this move makes it clear that the analysis can be protected against the possibility of counter-examples only at the price of circularity. For if you do not count as having a healthy appetite unless you enjoy healthy food, then the claim that healthy food is

pleasure: aristotle’s response to plato 253 really pleasant because it tastes so to the person with a healthy appetite is self-guaranteeing, because vacuous. The variety of human tastes and interests reinforces this difficulty. Aristotle counts pleasure in the exercise of the senses as a paradigm of human pleasure. Since for him sense-perception is the realization of a sensory capacity by its appropriate object, sensory pleasure requires that both the capacity and the object should be in good condition. On the side of the capacity the requirement is that the sensory apparatus should be functioning well, while on the side of the object the requirement is variously expressed: ‘every kind of perception is exercised on a perceptual object, and the perfect exercise is that of perception in good condition exercised on the finest (kalliston) of the objects falling under the perception’ (EN 1174b14–16); ‘in the case of each kind of perception the best exercise is that of the one in best condition exercised on the best (kratiston) of the objects falling under it’ (b18–19); ... ‘there is pleasure of every kind of perception, and also of thought and speculation, and the pleasantest is the most perfect, and the most perfect is the pleasure of the faculty in good condition exercised on the best (spoudaiotaton) of the objects falling under it’ (b20–3). The problem is to understand what Aristotle means by the quoted adjectives, ‘finest’ and ‘best’. The requirement that the sensory apparatus be working to perfection seems naturally matched by the requirement that the object be such as to stimulate perfect exercise, e.g. that perfect sight be stimulated by maximally visible objects (in terms of Aristotle’s theory, colours). But that condition is manifestly insufficient to guarantee that the perception is pleasant; as Anthony Kenny memorably points out, ‘the most sensitive nose in the world put in front of the most powerfully smelling manure in the world will not necessarily find the experience pleasant.’⁹ Nor need the absence of pleasure be attributed, as Kenny’s example might perhaps suggest, to the sense’s being overpowered by the object, as the eyes can be dazzled by too bright light; an object might be maximally visible or smellable, i.e. most clearly detected by the sense over the widest range of conditions, without the perception’s being pleasant. But if the finest and best objects are not the maximally perceptible objects, what are they? It is tempting to suggest that they are the most beautiful objects, but ‘beautiful’, unlike the Greek kalon, has natural application only to the ⁹ Kenny [1963], 149.

254

pleasure: aristotle’s response to plato

objects of sight and hearing. It is hard to know how to understand ‘beautiful smell’, ‘beautiful taste’, and ‘beautiful tactile sensation’, unless ‘beautiful’ is understood as ‘delightful’, in which case the object’s being fine and good is not independent of its being pleasant. Aristotle’s thesis is clearly intended to apply to all the sense modalities; so the object’s being kalon or spoudaion is its being fine-looking, fine-tasting, etc.; and the difficulty that those attributes are not applicable independently of the percipient’s pleasure in the appearance becomes general. Clearly, it will be hopeless to appeal to the judgement of the spoudaios to determine which sensory object is spoudaion. For the analysis of sensory pleasure requires that both faculty and object should be in good condition; but if the object’s being in good condition just is its seeming pleasant to the person whose faculties are in good condition then its being in good condition is identical with its being pleasant, and so cannot be part of the analysis or explanatory account of its being pleasant. Further, either the good condition of the sense-faculty includes its judging the right things pleasant, or it does not. If the former, the account is doubly vacuous, since one can neither identify good objects without appeal to a faculty in good condition, nor vice versa. But if the latter, then the account breaks down. For it is clearly possible that two people with perfect hearing, as measured by auditory testing, might disagree on which sounds are pleasant and which unpleasant, one adoring the bagpipes and detesting church bells, the other hating the former and loving the latter. Such diversity can be accommodated within a general account of pleasure as (arising from) the exercise, in good conditions, of the activities characteristic of the species only by the acknowledgement that it is characteristic of humans, unlike members of other species, to have different interests and preferences from one another, with the consequence that the description of good conditions for the exercise of the activity must include the condition that that exercise satisfies the agent’s preferences, interests, etc. That is not to revert to the deficiency/replenishment account of pleasure, since a preference or an interest is not a deficiency. Preferences etc. can indeed give rise to deficiencies. Thus a keen sailor obliged to live far from suitable water is likely to find life frustrating. But in that case the deficiency presupposes the preference, and is identified in terms of it; one lacks what one needs (viz. accessible water) in order to satisfy one’s preference (viz. for sailing), and in consequence of that lack one’s life is lacking something (viz. sailing) required to make it satisfactory. It is not the case, as the

pleasure: aristotle’s response to plato 255 deficiency/replenishment account maintains, that the preference is either the deficiency itself, or the awareness of the deficiency. The moral is that neither the notions of deficiency and replenishment nor those of activities proper to members of a species offer a reductive account of pleasure. Any account of pleasure must make room for notions of wanting, preference, interest, etc., but those notions do not offer the prospect of reduction, since pleasure itself figures in any account of them. (Being keen on sailing involves enjoying sailing, being pleased at the prospect of a sailing trip etc., subject to all the usual qualifications.) At EN 1153a12–15 Aristotle sums up his rejection of the deficiency/replenishment account of pleasure in these words: ‘Therefore it is not correct to say that pleasure is a perceived process of coming into being; rather one should say that it is the actualization of the natural state, and instead of ‘‘perceived’’ one should say ‘‘unimpeded’’.’ I take the requirement that the actualization should be ‘unimpeded’ to sum up the absence of obstacles, both internal and external, to the exercise of the capacity in appropriate conditions. That exercise could be ‘impeded’ by the inappropriate condition of the object, e.g. unpalatable food, or by the inappropriate condition of the subject, e.g. anxiety, loss of appetite, or both. Understanding ‘unimpeded’ in this broad sense, I take Aristotle to be offering as an improvement on the replenishment account the account of pleasure as the unimpeded actualization of a natural capacity. Thus understood his account raises two interrelated questions. First, is it an account of what we enjoy, or take pleasure in, or is it an account of what enjoyment is? That is to say, is Aristotle saying that what we enjoy is always the unimpeded exercise of a natural capacity, or that enjoyment (= pleasure) is the unimpeded exercise of a natural capacity? Secondly, what unimpeded exercise is he talking about? Taking the enjoyment of food as an example, is he talking about the unimpeded exercise of the capacity to take in nourishment (the nutritive capacity) or of the unimpeded exercise of the capacity to be aware of taking in nourishment, a perceptual capacity, perhaps to be identified with the sense of taste or of touch (see below)? In advance of answers to these questions we have four possible interpretations of Aristotle’s account: 1. What we enjoy when we enjoy food is unimpededly taking in nourishment;

256

pleasure: aristotle’s response to plato

2. Enjoying food is unimpededly taking in nourishment; 3. What we enjoy when we enjoy food is unimpededly perceiving our taking in nourishment; 4. Enjoying food is unimpededly perceiving our taking in nourishment. We must, of course, allow for the possibility that the account is undifferentiated between some (or conceivably all) of these alternatives. Thus if Aristotle does not distinguish between an account of what is enjoyed and an account of what enjoyment is, then 1 and 2 would collapse into one another, as would 3 and 4. That would apparently present a straight choice between an undifferentiated account of the pleasure of food as the unimpeded exercise of the nutritive capacity and an undifferentiated account of it as the unimpeded exercise of a perceptual capacity. But if the former presupposes awareness of the exercise of the nutritive capacity as a necessary condition of pleasure (i.e. enjoying food/what we enjoy when we enjoy food is unimpededly taking in nourishment, provided that one is aware of doing so) the gap between the two rival accounts is narrowed.¹⁰ Another possibility is that Aristotle does distinguish between an account of what is enjoyed and an account of what enjoyment is, and offers both. Thus 1 and 4 above are not only consistent with one another, but together offer a reasonably plausible comprehensive account of enjoyment and its objects; generalized it would claim that what we enjoy is the unimpeded exercise of natural capacities, and that enjoyment is the awareness of that exercise. There is a well-known difficulty confronting the attribution of theses 1 and 2 to Aristotle. This is that absorbing nourishment is a process (kin¯esis) which goes from a beginning to an end through a series of stages, takes time to complete, is not complete till it is over, can be interrupted and takes place quickly or slowly. But in EN X. 3–4 Aristotle argues that none of those marks of processes is true of pleasure, which is something whole and complete, like sight; the point is that pleasure, like sight, is complete as soon as it has occurred, unlike processes such as building which approach completion in a series of stages and achieve it only when the process is over. Hence at least by the time of writing EN X Aristotle appears firmly committed to the thesis that no pleasure is a kin¯esis, and hence to rejecting theses 1 and 2. As regards book VII, though the arguments for that conclusion are lacking, Aristotle appears to accept the conclusion itself; for ¹⁰ Cf n. 2 above.

pleasure: aristotle’s response to plato 257 he asserts (1153a9–12) that pleasures are not processes of coming to be (nor do all involve any such process) but activities and an end-state (sc. the state in which a process of coming to be is completed), and that they occur not when something is coming to be but when something (sc. a capacity) is utilized. A few lines later (a15–17) he says that the reason that people think that pleasure is a coming to be is that they think an activity is a coming to be, whereas they are different. So here, as in book X, it appears that no pleasure is a kin¯esis. This result in turn faces an equally well-known difficulty on the other side, viz. that at various places Aristotle speaks either explicitly or implicitly of processes such as building, writing, and calculating, which are plainly kin¯eseis by the criteria of EN X. 3–4, as things which are enjoyed (1173a15, b30; 1174a6; 1175a12–17, a34–5, b18). A possible way out of this difficulty relies on the distinction between theses 1 and 2 above. In these cases what is enjoyed is the process carried out in the appropriate conditions; so the devotee of building enjoys unimpeded building. What is denied in both discussions of pleasure is not that claim about what is enjoyed, but the corresponding claim about what enjoyment is, in this case the claim that enjoying building just is building unimpededly. Given that denial, and the positive claim that pleasure is not (a) kin¯esis or genesis but (an) energeia, the enjoyment of building would have to be some energeia supervening on the building itself, perhaps perception or awareness of the building, as suggested in thesis 4. An obvious objection to that way out of the difficulty is that Aristotle nowhere explicitly distinguishes the two questions ‘What is enjoyment?’ and ‘What kinds of things do people enjoy?’ A fortiori, he never points out that the insistence that pleasure is never a kin¯esis, but always an energeia, applies to the first question only, and that in at least some cases one correctly answers the second question by citing some kin¯eseis such as building or assimilating nourishment. Throughout he presents the discussion as if there were a single question ‘What is pleasure?’, to which ‘Pleasure is (a) perceived genesis/kin¯esis’ and ‘Pleasure is (an) unimpeded energeia’ are conflicting answers. The distinction of question 1 from question 2 (as of 3 from 4) is not, then, grounded directly in the text; it emerges from a process of sympathetic interpretation, as a distinction which offers Aristotle a way out of a difficulty. Hence if that difficulty can be resolved by an interpretation which remains closer to the text, by avoiding the appeal to that distinction, that interpretation is to be preferred.

258

pleasure: aristotle’s response to plato

In G&T we propose an alternative solution, which has at least the advantage of not requiring the distinction of questions 1 and 2. There is a single undifferentiated question, ‘What is pleasure?’, to which the unitary answer is ‘Pleasure is the unimpeded exercise of capacity’. Thus the pleasure of thinking is the unimpeded exercise of the capacity to think, and of building the unimpeded exercise of the capacity to build. Some capacities, such as the capacity to see, are exercised in acts which are themselves energeiai by the criteria of EN X. 4 (and also by the grammatical criteria of Meta.  6). Others, such as the capacity to build, are exercised in acts which are kin¯eseis by those criteria, since every act of building proceeds by stages, is not complete till it is over etc. But every stage in the process of building is also an exercise of building capacity (i.e. an energeia), and it is under the latter description that it is enjoyed. Hence the answer to ‘What is enjoyed in building?’ is ‘The unimpeded exercise of building capacity’, and the answer to ‘What is enjoying building?’ is ‘Enjoying building is exercising building capacity unimpededly’. David Bostock dismisses this suggestion, saying that ‘it is completely obvious that this is not Aristotle’s view of the matter’.¹¹ Aristotle, he maintains, makes it quite clear both in the Nicomachean Ethics and in the Metaphysics that walking and building simply are processes (= kin¯eseis) ‘no matter what one’s motive might be for undertaking’ them. (The question of motive is obviously irrelevant; an act of walking is an exercise of the capacity to walk irrespective of one’s motive for walking.) Bostock apparently takes the fact that these simply are processes as sufficient to establish that it cannot be the case that they are enjoyed qua exercises of their respective capacities. But he gives no argument for this conclusion, nor is it clear how it is supposed to follow. It would, of course, follow if ‘simply are processes’ implies ‘are processes and nothing else, hence not exercises of capacities’; but it is abundantly clear from Aristotle’s general account of capacities and their exercise that that is simply false. Walking, building etc., and in general things that people do, are all exercises of capacities. We still require an argument to establish that it cannot be Aristotle’s view that it is qua exercises of capacities, not qua processes, that they are enjoyed.¹² ¹¹ Bostock [1988]. The quotations are from 262–3. See also Bostock [2000], 160–5. ¹² One might seek to defend Bostock by arguing that Aristotle holds that, while one indeed exercises the capacity to build by building, the building and the exercise of the capacity to build are two distinct (though presumably spatio-temporally concurrent) entities, in modern terminology distinct events, not,

pleasure: aristotle’s response to plato 259 Bostock’s dismissal of this view is therefore too swift. Stronger reasons for scepticism are provided by considerations similar to those applied above to the distinction between theses 1 and 2, viz. that the thesis that processes are enjoyed not qua processes but qua exercises of capacity is also extraneous to the texts, and is simply imported to resolve a difficulty. After all, the use of the ‘qua’ terminology (in its Greek original h¯ei) is a standard piece of Aristotelian technicality. If what he means is that the process of building e.g. a house is enjoyed not qua building a house but qua the exercise of the capacity of building, why should he not say precisely that? As above, sympathetic interpretation should be subordinated to fidelity to the text. Bostock’s own solution of the difficulty can justly claim superiority in that respect, since it is closely based on the text. His claim is that Aristotle thinks that there are just two kinds of specifically human pleasure, viz. pleasure in thought and pleasure in the exercise of the senses. These are energeiai by the various criteria of the Nicomachean Ethics and of Meta. , and are listed among the examples of energeiai in the latter (1048b23–4, 33–4). What are loosely described as pleasures in or of activities such as eating or building are in fact pleasures in or of the associated thoughts and perceptions. The following quotation expresses the central point: We do, of course, speak of enjoying eating and drinking, just as we also speak of enjoying building, or writing, or hosts of other things which Aristotle will say are processes. But in all cases, as I interpret him, his view is that the place where the pleasure is to be found is in the associated thoughts and perceptions. Thus the builder may enjoy seeing his wall go up so straightly and so cleanly, as he may also enjoy the feel of the trowel in his hand, and the bodily sensations produced by the effortless exercise of his muscles. He may also enjoy first anticipating and then contemplating the completed building. In these thoughts and perceptions there may be pleasure, but not in the actual process of building. And Aristotle’s fundamental thought here is that pleasure takes place in the mind, but one can hardly say this of building, any more than of eating and drinking. (271)

In support of this claim Bostock points to the fact that the account of pleasure in EN X. 4, which is said to make clear what sort of thing it is (1174a13–14), is in fact an account of the pleasures of perception and thought (1174b14–26, b33–1175a1). As he points out, in that chapter as I have assumed, one entity with two non-equivalent descriptions. I know of no evidence justifying the attribution to Aristotle of such a theory, but cannot pursue the question here.

260

pleasure: aristotle’s response to plato

Aristotle ties pleasure closely to life (1175a10–21); it is significant that the examples of the activities which constitute life are thought and hearing (sc. hearing music), since this section thereby illustrates Aristotle’s claim at 1170a16–19 (repeated more emphatically at a33–b1) that human life consists primarily in perception and thought. One might add the fact that the discussion in X. 5 of the different kinds of human pleasures is introduced by the remark that the pleasures of thought are different from those of perception, and the different kinds of pleasure (sc. of thought, of perception, or perhaps of both¹³) from one another (1175a26–8), which suggests that no other kinds of pleasure need to be considered. Another passage which supports Bostock’s analysis is the account of s¯ophrosun¯e at EN III. 10. The virtue is a proper disposition towards the pleasures of food, drink, and sex. Its sphere is delimited via a classification of pleasures first into ‘psychic’ (exemplified by the pleasures of ambition, learning, and telling and listening to stories) and bodily, and then via a division of the latter. The principle of classification of bodily pleasures is according to the various senses. The pleasures appropriate to s¯ophrosun¯e are distinguished from those of sight (the examples cited are pleasures in colours, shapes, and drawing), hearing (pleasure in music and acting) and smell, allowing them to be identified as pleasures of touch and taste; Aristotle actually says ‘these (sc. pleasures) are touch and taste’ (1118a26). In the case of intemperate people taste is of little or no importance, the pleasures not only of sex but also of food and drink being ascribed to the sense of touch. Taste is important for the discrimination of flavours, but the intemperate are not at all interested in flavour, but merely in the tactile sensation of swallowing; hence the greedy man who wished that his gullet were longer than a crane’s ‘since he enjoyed the touch’ (1118a26–b1). It is possible that in this passage Aristotle characterizes not merely intemperate enjoyment, but all enjoyment of food, drink, and sex as pleasure in bodily sensations, specifically sexual sensations and the sensation of swallowing. He thinks that these are particularly discreditable forms of enjoyment because ¹³ All the manuscripts agree in reading ‘the pleasures of thought differ in kind from those of perception’. Following these words they vary between ‘and themselves from one another’, ‘and these from one another’ and ‘and these themselves from one another’. On the first reading pleasures of thought are referred to, on the second and third pleasures of perception. The best sense would be ‘and pleasures of either kind from one another’, which may be what Aristotle meant, but which is not confirmed by any manuscript.

pleasure: aristotle’s response to plato 261 they are common to humans and other animals (1118b2–3); this point, which applies to every case, not merely to that of the intemperate, may suggest that the account of these pleasures as pleasures in bodily sensations is also general. Whatever is the truth on that particular point, the main interest of the chapter as a whole for our present discussion is in its confirmation of the thesis that Aristotle regards the fundamental classification of human pleasures as that of pleasure in or of thought on the one hand and in or of the exercise of the senses on the other. The application of this analysis to the four possible accounts of the enjoyment of food listed above yields the result that Aristotle definitely rejects 1 and 2, and accepts either 3 or 4, or an undifferentiated thesis covering both. In the particular case of food, the tactile sensation account suggested in III. 10 favours thesis 3; Aristotle looks to be saying that what we enjoy is having certain tactile sensations, the having of which is itself an exercise of the sense of touch. Yet it does not follow that what we enjoy is not eating, but just a sort of perception. Rather, what is enjoyable about eating, on this (bizarre) view, is the sensation of swallowing; to put it another way, the way we enjoy eating, according to Aristotle, is by enjoying the sensation of swallowing.¹⁴ If we generalize this result to the problematic cases of enjoyment of activities like building, we arrive at a reductive account, not of enjoyment itself, but of its object; enjoyment of building is just enjoyment of perception of and thought about building, as sketched in the quotation from Bostock given above. But Bostock takes the upshot to be that pleasure is in these thoughts and perceptions, not in the actual process of building, on the ground that Aristotle thinks that pleasure takes place in the mind, whereas building does not. But he produces no evidence that Aristotle thinks that pleasure is ‘in the mind’ in a sense which is inconsistent with one’s literally enjoying building. Of course pleasure is not in the body, as Aristotle points out (1173b9–11), but then neither is building in the body. Building is something which an embodied agent does, and sensory pleasure is also an attribute of an embodied agent. Rather than accept Bostock’s contention that pleasure is in the builder’s thoughts and sensations and not in the process of building itself, we should say that the pleasure which is in the builder’s thoughts and sensations is the builder’s pleasure in the process ¹⁴ For fuller discussion see ch. 7 of this volume.

262

pleasure: aristotle’s response to plato

of building itself. After all, what else could pleasure in the actual process of building be, than the pleasure that is in the builder’s thoughts and perceptions? But now we are confronted by the crucial ambiguity of the expression ‘pleasure in the builder’s thoughts and perceptions’. This may be construed as ‘enjoyment of the builder’s thoughts and perceptions’. On that construal, the notion of enjoyment remains primitive, and the account of pleasure consists in the reductive account of its object; enjoyment of building just is enjoyment of the builder’s thoughts and perceptions. But the expression can also be understood as ‘enjoyment consisting in the builder’s thoughts and perceptions’. On this construal, unlike the other, we are given an account of what enjoyment itself is, viz. certain kinds of thought and perception. For the builder to enjoy building is for him to see the wall going up straight and cleanly, to feel his muscles moving effortlessly, to think of all this as something worth doing etc. We may recall how Aristotle moves in EN III. 10 from speaking of enjoying the objects of sight and hearing (chairontes tois dia t¯es opse¯os, tois peri t¯en ako¯en) to the statement that the pleasures with which s¯ophrosun¯e is concerned are touch and taste. That might of course be understood as the claim that they are the pleasures of touch and taste, but equally it can be understood literally as the claim that those enjoyments are exercises of those senses. And we should also recall that the discussion of X. 4, which is to make it clearer what pleasure is, starts by explaining that seeing, unlike processes, is complete at every moment of its existence, and goes on to show that pleasure shares that characteristic. It is certainly possible, and perhaps even natural, to take this as making a point, not about the objects of pleasure, but about pleasure itself, namely that in a crucial point it is like seeing. The texts which favour Bostock’s account, then, at least leave it open that Aristotle is attempting to provide an account of pleasure as consisting in thought and perception, or that his theory is undifferentiated between that and an account of what is enjoyed as thought and perception.¹⁵ This issue affects the crucial question of how Aristotle’s account is supposed to apply to the virtuous agent’s pleasure in his or her virtuous activity, which is crucial, as we have seen, to virtue and the good life. Does Aristotle think that what the virtuous agent enjoys is being aware, in thought and/or ¹⁵ Bostock himself explicitly leaves these questions open at the conclusion of his paper (272).

pleasure: aristotle’s response to plato 263 perception, of acting virtuously, or that his or her enjoyment of virtuous action just is his or her awareness, in thought and/or perception, of acting virtuously, or is his view undifferentiated between the two? The nearest I can come to answering this question is to offer the following tentative suggestion. Notoriously, the discussion of book VII appears to identify pleasure with unimpeded activity, whereas that of book X avoids that identification, preferring to describe pleasure as something which perfects activity in a special way ‘as a sort of supervening perfection, like the charm of those in their prime’ (1174b32–3). My suggestion is that this change may reflect an increased awareness on Aristotle’s part of the distinction between an account of what is enjoyed and an account of what enjoyment is. The ‘unimpeded activity’ formula straddles the two, whereas if Aristotle had come to a clearer conception of pleasure in an activity as a sort of awareness of that activity, he would be reluctant to identify it with the activity itself, while yet seeking for a way of characterizing the inseparability of the awareness from the activity. An additional attraction of the idea of pleasure as awareness, divided into thought and perception, is that it is applicable to all pleasures, including cases where the object of pleasure is nothing but thought or perception itself. For Aristotle, thought and perception are self-intimating; we are aware of thought by or in thinking, and of perception by or in perceiving (De an. 425b12–25, EN 1170a29–33), while in the case of other activities thought and perception are the means by which we are aware of doing them.¹⁶ I suggest, then, that the virtuous agent is aware in thought of what the content of his or her good prohairesis is, and in perception that the description of the action fits ¹⁶ According to G&T the distinction between Books VII and X is terminological only. Both share the same substantive view of pleasure as the unimpeded, i.e. perfect, actualization of capacity, but whereas that view is expressed in VII as ‘pleasure is unimpeded activity’ the thought in X is that pleasure is the perfection in virtue of which the unimpeded activity is perfect, or in other words the formal cause of its perfect actualization (249). That suggestion now seems to me less plausible. If Aristotle’s point is one which requires to be expressed by means of his own terminology of kinds of cause, and specifically via the notion of a formal cause, it is mysterious why he does not employ that terminology. The ‘charm of those in their prime’ appears to be a simile intended to elucidate a relation (between pleasure and the activities enjoyed) which eludes literal exposition in standard terminology. In rendering Aristotle’s tois akmaiois h¯e h¯ora as ‘the charm of those in their prime’ I revert to the traditional understanding of this expression as referring to the visible aspect of the perfection of those in the prime of life, normally translated ‘the bloom on the cheek of youth’. G&T, pointing out that that sense of h¯ora is secondary to its primary sense of ‘season’, which is then extended to that of the right season, the springtime of life, render, in conformity with their interpretation of pleasure as the formal cause of perfect activity, ‘the springtime of youth for those in their prime’ (212). As the Greek phrase may bear either sense, the rendering must be determined by one’s overall interpretation of the context.

264

pleasure: aristotle’s response to plato

the content of that prohairesis,¹⁷ and that those thoughts and perceptions are inseparable from the agent’s enjoyment of virtuous activity. But whether they are that enjoyment, or its object, or undifferentiated between the two I am unable to determine. This suggestion has an obvious affinity with G. E. L. Owen’s celebrated thesis,¹⁸ but is not a mere restatement of it. Owen sees the discussions of books VII and X as simply directed to different questions, the former to the question of what is really enjoyed or enjoyable, the latter to the question of what enjoyment is. Moreover, the methods of the two discussions are different; the former proceeds by looking for some feature common to everything which is enjoyed, the latter by reviewing the logical characteristics of pleasure-verbs. On the alleged difference of method, I adhere to the criticism of Owen’s view in G&T.¹⁹ On the content of the two discussions I agree with Owen in detecting a shift, but identify the shift differently. For Owen the two discussions are directed to quite different questions, and it is then puzzling why those questions are expressed in the same words ‘What is pleasure?’ I see the two discussions as stages in the articulation of a single enquiry. In each case Aristotle is addressing the question ‘What is pleasure?’, because the answer to that question is a precondition of the correct evaluation of pleasure. But Aristotle’s question is itself ambiguous between ‘What do we enjoy?’ and ‘What is enjoyment?’, and my suggestion is that the discussion of book X shows some indication, absent from book VII, that Aristotle had moved towards separating those questions. If that is correct, it favours the prevalent (though not universal) view that the discussion in book VII is the earlier.²⁰ ¹⁷ For instance, the agent is aware in thought that his/her prohairesis is to eat a portion of chicken qua healthy food, and via perception that this food on the plate is a portion of chicken (EN 1147a3–7, b9–17). The agent’s knowledge of what he or she is doing thus combines direct awareness of his or her intentions with perceptual knowledge of whether and how that intention is realized. ¹⁸ Owen [1971–2]. ¹⁹ See ch. 11.3. ²⁰ For further discussion see Broadie [2003].

15 The Hedonism of the Protagoras Reconsidered The long-standing debate on ‘the hedonism of the Protagoras’, i.e. on the question of whether in the dialogue Socrates is represented as maintaining in his own person some version of the thesis that pleasure is the good, or merely arguing ad homines that that thesis must be accepted by the ordinary person and/or the sophists, shows little sign of going away. From Grote [1865] onwards scholarly opinion has been divided, with the majority inclining now one way and now the other. In the last thirty or so years the latter view has commanded majority support, its most redoubtable champions being Gregory Vlastos [1969, 1991] and Charles Kahn [1988a, 1988b, 1996], but there have been dissidents, including Terence Irwin [1977, 1995], Martha Nussbaum [1986], myself [1976, 1991], and most recently George Rudebusch [1999]. Writers on either side of the dispute are, of course, able to point to indications in the text which support their particular view; were that not so the debate could hardly have continued so long without consensus. In this re-examination of the question I shall argue that the inconclusive nature of the dispute is no accident, but an involuntary tribute on the part of modern scholars to Plato’s skill as a writer of philosophical dialogue. Plato, it now seems to me, deliberately leaves undetermined the question ‘Who is it who maintains hedonism?’, precisely because in writing the dialogue his interest is not in attributing to this person or that the thesis that pleasure is the good, but in the thesis itself as an abstract proposition. His interest in that proposition, I suggest, stemmed from the facts (a) that the sketch of an ethical/psychological theory which he attached to the figure of Socrates in the early dialogues required an account of the good and (b) that the abstract thesis ‘Pleasure is the good’ filled that gap with an account which he felt deserved to be taken seriously.

266

the hedonism of the protagoras

It will be helpful to remind ourselves of the wider context within which the issue of hedonism arises in the dialogue. Protagoras had begun by maintaining that the specific ‘parts of virtue’, wisdom, self-control or soundness of mind (σωϕρoσύνη), courage, justice, and holiness or piety (ὁσιότης ) are all as distinct from one other as the different parts of the face are from one another (329d–e), but by 349d has been forced by Socrates’ arguments to retreat to the lesser claim that while wisdom, σωϕρoσύνη, justice and ὁσιότης are ‘pretty similar to one another’ courage is ‘totally different from all the others’, the evidence being that some people are exceptionally courageous while altogether lacking the other virtues (indeed being characterized to an extreme degree by the opposite vices). Socrates gives an argument whose aim is to force Protagoras to agree that courage is identical with wisdom (349e–350c), but Protagoras identifies a fallacy in that argument, which is immediately dropped (350c–351b). At 351b3 Socrates abruptly begins a new argument designed to force Protagoras to accept the identity of courage with wisdom, which he finally does at 360d–e. It is this final argument which relies on the assumption that pleasure is the good. It should be noted from the outset that Socrates is unambiguously represented as not merely forcing Protagoras to admit the identity of courage with wisdom but as maintaining that identity himself. This is stated explicitly at 361a–b, where Socrates imagines the argument as mocking both Protagoras and himself for having reversed their original positions: ‘How absurd you both are, Socrates and Protagoras! You (i.e. Socrates) previously said that virtue can’t be taught, but now you are ~ ἀπoδε ~ ιξαι) that urging (σπεύδεις ) the opposite, trying to show (ἐπιχειρων everything is knowledge, justice and self-control and courage, in which case virtue most certainly could be taught. For if virtue were something different from knowledge, as Protagoras was trying to say, it would plainly be something that could not be taught; but now if it turns out as a whole to be knowledge, as you are urging (σπεύδεις ), Socrates, it would be amazing if it could not be taught.’ Socrates, then, is maintaining the identity of virtue as a whole, and thereby courage specifically, with knowledge or wisdom,¹ and by ‘maintaining’ is to be understood not merely asserting, ¹ I use these terms interchangeably, since the Greek terms σoϕία and ἐπιστήμη appear to be used interchangeably in this section of the dialogue. Thus at 352d1–2 Protagoras says that σoϕία and

the hedonism of the protagoras 267 but arguing, as is indicated by ‘trying to show’ at 361b1. There is therefore an expectation that Socrates will be presented as maintaining in his own person, not merely his conclusion, but also the premisses from which he derives that conclusion, including the identification of pleasure with the good. We must now examine the text to see whether it bears out that expectation. Socrates begins by securing Protagoras’ agreement to the theses that someone who has lived a life of pain and misery has not lived well, and that someone who has completed a pleasant life has lived well (351b). This agreement is not without qualification, for when Socrates sums up Protagoras’ admissions as ‘So a pleasant life is good and an unpleasant one bad’ Protagoras adds the qualification ‘Provided the things you take pleasure in are praiseworthy’ (εἴπερ τo ~ις καλo ~ις ... ζῴη ἡδόμενoς c1–2). So Protagoras rejects evaluative hedonism; he thinks that a life’s being pleasant is not sufficient for its being good, since it may be disqualified from counting as a good life if the pleasures which characterize it fail to be καλά, i.e. if they are in one way or other (the specification is not specifically moral) shameful, disgraceful, or inferior. Socrates responds, ‘Surely you don’t go along with most people in calling some pleasant things bad and some unpleasant things good’ (c2–3). Here Socrates makes it clear that he thinks that most people agree with Protagoras in rejecting evaluative hedonism, while the force of his question ‘surely you don’t ... (μὴ καὶ σύ ...)’ gives a strong conversational implication that Socrates thinks that Protagoras ought not to agree with the majority on that point. But is that because Socrates thinks that, because evaluative hedonism is true, both Protagoras and the majority would be wrong to reject it, or because, independently of the truth of evaluative hedonism, Protagoras is to be expected not to reject it? In Taylor [1976] and [1991] (p. 166) I assert the former, but the issue now seems to me not so clear. Socrates’ next utterance (c4–6) is the question ‘in so far as things are pleasant, are they not to that extent good, leaving their other consequences out of account, and the same with unpleasant things; in so far as they are unpleasant, are they not to that extent bad?’ In the passage cited I take it that this is a rhetorical question having the force ἐπιστήμη are the most powerful (κράτιστoν (sing.) ) force in human affairs, i.e. he is evidently using the two words as names for the same thing. And at 360d Socrates obliges Protagoras to agree to the ~ δεινων ~ καὶ μὴ δεινων ~ and immediately goes on to say that he has definition of courage as σoϕία των been seeking to show that all the virtues, including courage, are ἐπιστήμη (361b1–2).

268

the hedonism of the protagoras

of an assertion of Socrates’ own view. It certainly could be, but equally it could be a genuine question designed to elicit Protagoras’ response, having the conversational force ‘isn’t this rather the way you see it?’. Socrates’ question is prefaced by the words ἐγὼ γὰρ λέγω (c4), and we might hope that the interpretation of that phrase would settle the interpretation of the whole sentence, but unfortunately the phrase itself is ambiguous between ‘what I say is’ and ‘what I mean is’. The former rendering demands the reading of the whole sentence as a rhetorical question asserting Socrates’ own view; the latter rendering allows that reading but does not demand it, and equally allows the other reading. Protagoras’ answer (c7–d2) ‘I don’t know whether I should give such a simple answer to your question and say that all pleasant things are good and all unpleasant things bad’ is indeterminate between the two readings, as is the succeeding interchange down to e3.² At e3 Protagoras agrees to investigate the question ‘whether pleasure itself is not good’ (e2–3),³ adding ‘if it apppears that pleasure and good are the same, we shall agree; but if not, we shall then argue further’ (e5–7). Protagoras is thus represented as understanding Socrates as maintaining ‘that pleasure and good are the same,’⁴ and seeing the investigation as supposed to lead either to agreement or to disagreement between Socrates and himself on that point. In Gosling and Taylor [1982] we maintain (p. 50) that this shows that the identity of pleasure and good (however understood) is Socrates’ own thesis, since ‘Plato might perhaps make Protagoras misidentify Socrates’ thesis; he would hardly represent him as mistakenly believing that Socrates has a thesis’. That assertion now seems to me simplistic. Why should Plato not invite the reader to ask precisely the question whether Socrates is asserting the thesis, as Protagoras is represented as supposing, or merely raising it for discussion? And once the reader has asked that question he may be inclined to see the ambiguous phraseology which has preceded as intended to avoid a definitive answer. So far the dialectical situation is that Protagoras is represented as agreeing with the majority in rejecting evaluative hedonism. He holds (as they do) ² Note the repetition of the ambiguous λέγω at e2. ³ I discuss the ambiguity of this question in Taylor [1976] and [1991], 168–70, concluding that it is possible, but not certain, that what is at issue is a thesis undifferentiated between the alternatives ‘Pleasure is the only underivatively good thing’ and ‘ ‘‘Pleasant’’ and ‘‘good’’ are names for the same characteristic’. For what it is worth, Protagoras appears to take the question in the latter sense. ⁴ See previous note.

the hedonism of the protagoras 269 that there are some bad pleasures, and it has been implied that he thinks that such pleasures are bad insofar as they are disgraceful or discreditable, i.e. because they are deficient on a measure of value which is not itself determined by pleasure and its opposite, distress. Socrates, on the other hand, has at least proposed and perhaps himself maintained a view according to which (i) a life’s being pleasant is both necessary and sufficient for its being good, (ii) there are no bad pleasures, (iii) anything pleasant is good to the extent to which it is pleasant, leaving other consequences out of account, (iv) pleasure itself is good (or perhaps (iv a)) pleasure itself is the good.⁵ While this position (or cluster of positions) is not precisely delineated, it is clear that its overall thrust is opposed to the commonsense anti-hedonism of Protagoras, who has explicitly rejected i and ii, and who expects to find himself disputing iv. The next step is to examine Protagoras’ view on the relation between the good and the pleasant (352a6–8), which amounts to examining his reasons for rejecting the hedonist thesis (or theses). The examination proceeds indirectly. Protagoras agrees enthusiastically that someone who knows what is good and bad will never be overcome by any motivation so as to do what he knows to be bad (352c2–d3), and Socrates explicitly states his agreement (d5). The majority of ordinary people, however, do not agree, but maintain that people do what they know to be bad because they are overcome by pleasure, distress, or by specific motives such as lust or fear (d4–e2, referring back to b3–c2). Socrates and Protagoras agree that what these people say is false (353a3), and Socrates invites Protagoras to join him in persuading them and teaching them the real nature of the phenomenon which they (mistakenly) call ‘being overcome by pleasure’ (352e5–353a2). While it is as yet unclear (as Protagoras remarks, 353a7–8) how the examination of common views is germane to the dispute between Protagoras and the ‘Socratic’ hedonistic thesis, the dialectical set-up certainly leads us to expect that Socrates and Protagoras will attempt to convince the many by arguing from premisses which they (i.e. Socrates and Protagoras themselves) believe to be true. Socrates now imagines a dialogue between the many on the one hand and Protagoras and himself on the other which certainly starts out ⁵ See n. 3.

270

the hedonism of the protagoras

in conformity with that expectation. The many challenge Socrates and Protagoras to say what they think ‘being overcome by pleasure’ really is, and Socrates undertakes to tell them (353c1–4). But immediately Socrates starts from the description of the phenomenon given by the many, viz. that people do things they know to be bad (πoνηρά) because they are overcome by the pleasures of food, drink, and sex, and asks why they (i.e. the many) call the things they do bad (c4–9). The answer which he suggests to them, and which Protagoras immediately agrees that the many would give, is that the only reason they consider those things bad is that they produce consequences such as illness and poverty which are unpleasant in the long run (c–e). Similarly, Protagoras and Socrates agree that the only reason the many have for counting immediately unpleasant things such as medical treatment good is that they lead in the long run to pleasures and the avoidance of distress (354a–c). The many are then shown that, given this basis for evaluation, they cannot defend their claim that people do things they know to be bad because they are overcome by pleasure; since they identify pleasure with goodness, that would be to say that people do what they know to be bad because they are overcome by the good (354e–355d). Since they have a single standard of evaluation, the predominance of pleasure over distress, the only explanation that they can give of failure to do what is best is failure to apply that standard correctly. Socrates concludes the imagined dialogue by reminding the many of their challenge to Protagoras and Socrates to tell them what ‘being overcome by pleasure’ really is. Had Protagoras and Socrates given straightaway the correct answer ‘Error’ (ἀμαϑία) the many would have laughed at them; but now they have to laugh at themselves, since they have themselves to acknowledge that, given the only standard of evaluation that they have, that is the correct answer (357c–d). Leaving aside the question of what detailed flaws affect the argument,⁶ the overall strategic position is that Socrates has argued against the many that evaluative hedonism commits them to holding that the only possible explanation of bad conduct is intellectual error. Of course the many started by denying evaluative hedonism, since they agreed that there are bad pleasures. But that denial was superficial. Since the only standard of evaluation which they would acknowledge is a hedonistic one, the only ⁶ See Taylor [1976] and [1991], 179–86.

the hedonism of the protagoras 271 sense in which they can agree that pleasures are bad is that they lead to an unfavourable balance of pleasure over distress overall. Their value scheme is hedonistic, and the recognition of bad pleasures can be accommodated to that scheme of evaluation. And given that scheme, they have no option but to accept the Socratic claim that failure to do what is best (i.e. pleasantest) overall must be attributed to mistake or error, specifically error in estimating or calculating the overall consequences, measured in terms of pleasure and distress. So they have no option but to accept as true the account of ‘being overcome by pleasure’ as ‘error’ which Socrates had proposed to them on his and Protagoras’ behalf. And clearly Socrates and Protagoras are represented as themselves taking that account to be true. But why? We have not been clearly told. If Socrates is represented as himself maintaining, and not merely proposing for consideration the hedonistic thesis enunciated at 351b–d he should himself consider the argument which convinced the many to be not merely effective against them, but sound. But we saw that the presentation of that argument was sufficiently nuanced to make it unclear whether Socrates was represented as maintaining it in his own person. At 358a, having concluded his imaginary argument with the many Socrates turns to Protagoras and the other sophists, and begins by asking ~ them ‘whether what I am saying seems to you true or false’ (πότερoν δoκω ~ ~ ὑμ ιν ἀληϑη λέγειν ἢ ψεύδεσϑαι). The first-person form of the question explicitly invites assent to or dissent from whatever Socrates has so far said in his own person, including, of course, his assertions about what the many will say, but does not invite assent to or dissent from the content of those assertions themselves. But the form of the reply blurs that distinction; ‘they ~ ἐδόκει all agreed that what had been said was absolutely true’ (‘ϒπερϕυως ~ ~ ἅπασιν ἀληϑη ε ἰναι τὰ εἰρημένα). Strictly speaking that should refer only to what had been said by Socrates, but in the context the question arises whether the reference of τὰ εἰρημένα might not be to what had been said in the whole of the imagined dialogue with the many. That possibility is made increasingly lively by Socrates’ next utterance, ‘So you agree that the pleasant is good and the unpleasant bad’. The ‘so’ makes it clear that the sophists are committed to this as something which has been said, but said by whom? If the reference is back to 351b–e, then this has been said by Socrates, and that assumption lends support to the reading of λέγω at 351c4 as ‘say’, rather than ‘mean’ (see above). But in the conversational context

272

the hedonism of the protagoras

the more natural reference is to the discussion with the many which has just concluded, and in that discussion it was the many, not Socrates, who said that the pleasant is good and the unpleasant bad. (They ‘said’ it, of course, in the sense that they accepted Socrates’ insistence that they had no other standard of good and bad than pleasure and distress.) In that case the sophists draw no distinction between what Socrates has himself said and what he has induced the many to agree to, since they answer a question about the truth or falsity of the former by reference to the latter. Plato might, of course, have written this passage with the intention of representing the sophists as inept in argument, since they are unable to distinguish Socrates’ own assertions, which they are committed to accepting, from those which he elicited from others, to which they are not so committed. But the upshot of that is that Socrates is represented as equally inept, if he does not notice the sophists’ confusion, or as unscrupulous, if he is represented as intentionally taking advantage of that confusion. I discuss this issue in Taylor [1976] and [1991], 208–9. I do not there consider the hypothesis that Plato represents both the sophists and Socrates as inept arguers, and shall not consider it here either. In that discussion I argue that if Plato (a) represents Socrates as not himself maintaining the conclusions which he elicits from the many and (b) represents the sophists as failing to distinguish those conclusions from Socrates’ own view, then he represents Socrates as arguing dishonestly. Holding that to be implausible I conclude that ‘Socrates is represented by Plato as sharing the assumptions of the common man and the conclusions which he (Socrates) derives from those assumptions, rather than merely forcing the common man to accept the implications of assumptions which he (Socrates) rejects’ (p. 209). This argument is supported by a subsidiary argument, viz. that Socrates’ conclusion that cowardly action arises from error as to the best course of action is derived, independently from any hedonistic assumptions, from the thesis of 358c–d that no one goes for what he considers a worse alternative when he thinks a better alternative available. Since the hedonistic assumption is therefore not necessary for Socrates’ argument it is especially implausible ‘that he should be represented as arguing with conscious dishonesty on such a subsidiary point’ (ibid.). I no longer find this argument persuasive. Taking the subsidiary argument first, I am no longer clear that the thesis of 358c6–d4 is intended to be taken as independent of the hedonistic assumption in which the whole

the hedonism of the protagoras 273 discussion is embedded, and which is stated at b6–7. Again, the hypothesis of dishonesty on Socrates’ part no longer seems so incredible, especially in the light of the depiction of Socrates as an eristic combatant which has been so prominent earlier in the dialogue, particularly in the whole Simonides episode (and see esp. 339e). But more crucially I am now dubious of a tacit assumption of this argument, viz. that the sophists’ slide from their agreement to what Socrates has said to their acceptance of the hedonistic assumption is intended by Plato as a mark that that assumption is to be attributed specifically to Socrates. It now seems to me at least as plausible that that slide is intended by Plato as a mark precisely of indifference as to whether the thesis is to be attributed to Socrates or only to the many. The reader, that is to say, is not supposed to ask ‘But is that Socrates’ thesis, or merely an assumption of the many?’ Rather the reader is supposed to ask ‘Suppose we take the overall pleasure of a life as our account of the good, what follows?’ and to discover, by following out the argument, that what follows is that courage, in common with the other virtues, is identical with knowledge of the good. To recapitulate, at 351b Socrates abandons his first unsuccessful attempt to show that courage is knowledge or wisdom and begins what proves to be a second argument for the same conclusion by introducing, without any preamble, a thesis of evaluative hedonism. Not only is the precise content of that thesis unclear (see above), but it is indeterminate whether Socrates is represented as maintaining it himself or merely as introducing it for Protagoras’ consideration. Protagoras initially rejects the thesis, maintaining that if one’s life is to be good it is not sufficient that it be pleasant, since it must satisfy the further condition that the pleasures which make it pleasant should be fine. He is, however, willing to examine it further, and is reluctantly persuaded to consider, as part of that examination, the opinion of most ordinary people that it is possible to do what one knows to be bad because one is overcome by pleasure (an opinion which Socrates and Protagoras agree in rejecting). Those people acknowledge that they have no other standard of good and bad than (respectively) the promotion of a pleasant and an unpleasant life, and that given that scheme of evaluation it is inconsistent for them to hold that one can do things one knows to be bad because one is overcome by pleasure. Socrates then secures the agreement of Protagoras and the other sophists that what he is saying is true, and on the strength of that that the pleasant is good and that what

274

the hedonism of the protagoras

promotes a pleasant life is fine; i.e. Protagoras withdraws his objection to Socrates’ initial thesis, conceding that pleasure is itself the only determinant of what is fine. Since nothing has occurred to motivate this change of mind other than the argument with the many, it is plausible that the sophists take the acknowledgement of the many that they have no non-hedonistic standard of value as committing them to the same acknowledgement, i.e. that they make no distinction between what Socrates has maintained and what has been maintained by any of the parties to the argument. The sophists then agree that no one acts contrary to his knowledge or belief of what it is best to do, and hence that ‘giving in to oneself’ is error or mistake (358b6–c3), and that it is not in human nature to go for what you think to be bad in preference for what you take to be good (c6–d2). The former admission is explicitly subject to the condition that ‘the pleasant is good’; that condition is not explicit in the second admission, but it is possible that it is intended to be understood from the earlier sentence. Socrates now applies these admissions to the special case of courage and eventually secures Protagoras’ reluctant agreement that courage is ‘wisdom concerning what is to be feared and not’ (360d1–2) and hence that it is impossible to be both ignorant and courageous (e1–5), as he had maintained at 349d. In his concluding summary of the argument Socrates describes himself as seeking to show and as urging that all the virtues, including courage, are knowledge. While that indeed creates a retrospective indication that Socrates has himself maintained the hedonistic premiss from which that conclusion is derived, that has to be set against the ambiguous presentation of the thesis at 351b–e and the central role of the dialogue with the many, where it is emphasized that it is they who have no non-hedonistic evaluative standard. Once their commitment to hedonism has been secured, the immediate assent of the sophists to that position suggests that alignment with the many is as central to their depiction as is agreement with Socrates. I suggest, then, that we should at least take seriously the idea that the tentativeness and indirectness of the presentation of hedonism in the Protagoras is a device of Plato’s, intentionally adopted to leave it indeterminate whether Socrates is himself espousing that thesis or not. We have in the dialogue an account of human good as a life with a favourable overall balance of pleasure over distress, and of the goodness of an action as a positive contribution to such a life. That account is

the hedonism of the protagoras 275 presented as one to which the ordinary person has no alternative; perhaps the point is that it is implicit in our ordinary pre-theoretical assumptions, or more ambitiously that, failing a metaphysically more demanding theory, of which the ordinary person is ex hypothesi incapable, it is what the ordinary person should accept as an adequate ground for his or her moral beliefs. Since the sophists are champions of ordinary moral beliefs and practices (as Protagoras makes clear at 328a–c) it should therefore be congenial to them too.⁷ Given that account Socrates can derive the theses that virtue is knowledge (and hence that all the specific virtues are in fact one) and that no one acts contrary to what they believe best. It is thus an account which he can use both to support theses which he maintains or at least adumbrates in other dialogues, such as Meno and Laches, and to refute Protagoras’s conception of the virtues as independent motivating forces. Why, then, should Plato shrink from presenting him as unambiguously espousing it? That question invites us to plunge into the deep waters of Plato’s authorial intentions. One possibility is that in writing the Protagoras Plato was concerned to reflect faithfully the views of the historical Socrates, and that he was unwilling to present Socrates as unambiguously espousing hedonism because he knew that Socrates had not in fact done so. I do not myself propose that explanation of the phenomenon, since I do not believe that the dialogues give us any evidence that Plato was ever concerned with historical accuracy in the attribution of views to Socrates. From the outset, in my view, Socrates figures in Plato’s presentation as an exemplar of philosophy, and it was as germane to that presentation to attribute to the figure of Socrates positions of Plato’s own as to record his own views. I believe that Plato’s portrayal of Socrates contains many aspects of the personality and argumentative methods of the historical Socrates, and it is quite likely that some of the views which ⁷ Weiss [1990] emphasizes this aspect of the dialogue. On her reading Socrates’ main aim is to provide for Protagoras the defence of his claims as a teacher which he fails to provide for himself, by presenting expertise in the application of the hedonistic calculus as a model of sophistic expertise. While agreeing with her that Plato may intend to suggest the μετρητικὴ τέχνη as an appropriate subject for sophistic teaching, I disagree with her final claim that the account of courage as knowledge of things to be feared and not is intended by Plato as a reductio ad absurdum of the proposed model. I believe that that was Plato’s preferred account of courage prior to the Republic, and that in his view the only questionable element in it was the hedonistic determination of what is and is not to be feared.

276

the hedonism of the protagoras

Plato attributes to him were in fact held by that individual, but historical authenticity was never a criterion for the attribution of a view to him.⁸ Another suggestion, which has the merit of not relying on the implausible hypothesis of biographical authenticity, is that Plato avoided a straightforward portrayal of a hedonist Socrates because that would have been inconsistent with a portrayal of Socrates as an anti-hedonist which was already established in earlier Platonic dialogues. This too raises large questions, both about the relative chronology of the dialogues and about the degree of consistency in the portrayal of Socrates which the reader expected. While the personality of Socrates is recognizably the same across a range of dialogues, and while there are indeed certain beliefs, e.g. that one must never act unjustly, and that virtue is knowledge, which he maintains in more than one dialogue, there are also discrepancies between dialogues. Thus the agnosticism of the Apology on the survival of the soul after death is strictly inconsistent with the confident assertions of immortality in the Phaedo. It will not do to claim that the Socrates of the Apology is the real Socrates, whereas the Socrates of the Phaedo is Plato’s Socrates, since, for the reasons given in the previous paragraph, there is no sharp line to be drawn between the two. Rather, the reader accepts that Apology and Phaedo have different aims, the former a justification of Socrates’ philosophical activity, the latter a philosophical meditation on the nature of the soul and of death, to which Socrates’ divergent stances on the afterlife are appropriate. Hence, even if it were the case that Plato had already portrayed an anti-hedonist Socrates before writing the Protagoras, it would not be obvious that the reader would have had to conclude that a hedonist portrayal in the latter could not be taken seriously. Had such a situation obtained, a possible reader’s reaction might have been ‘Two very different Socrateses; so one of them must be speaking for Plato, while the other is closer to the historical. Which is which?’ But the crucial point is that, lacking any reliable data on relative chronology, we have no good evidence for the existence of an anti-hedonist portrayal antedating the Protagoras. The only plausible ⁸ It is therefore irrelevant to my purpose to consider whether any non-Platonic evidence, such as that of the Cologne papyrus PK¨oln 205 (discussed briefly in Taylor [1998], 23), should be taken to suggest that the historical Socrates may actually have been a hedonist. In fact we have no more reason to suppose the author of the dialogue from which that fragment comes to have been more concerned with historical accuracy than Plato (or, for that matter, any writer of S¯okratikoi logoi).

the hedonism of the protagoras 277 candidate is the Gorgias; and the case for an early dating of that dialogue, though ably defended by Kahn,⁹ remains at best conjectural and in my opinion unconvincing.¹⁰ I propose an alternative hypothesis, namely that Plato deliberately avoided an unequivocal attribution of the hedonistic thesis to Socrates because he wished to indicate to the reader that his own attitude to the thesis was not one of unqualified commitment. Anticipating the reader’s instinctive acceptance of Socrates as speaking in the dialogues as the representative of philosophy, and therefore as expressing the authority of the author, he is careful to indicate that this is not a thesis which that authoritative voice proclaims as its own. Not that it is unequivocally assigned to anyone else. As we have seen, some indications point to the ascription of the thesis to the authoritative figure of Socrates, while others suggest that it is a thesis appropriate to the majority of ordinary people, and thereby to the sophists. What we are to make of this, I suggest, is that Plato is here exploring, rather than maintaining, an account of the good, an exploration prompted by the investigation of the topic of virtue and knowledge which we find undertaken in the Socratic dialogues. In the Meno Socrates argues positively, but inconclusively, in favour of the thesis that virtue is knowledge, but does not address the question ‘Knowledge of what?’ In the Laches he adumbrates the suggestion that courage is knowledge of good and evil, but abandons that suggestion on the ground that in that case courage would be identical with virtue as a whole, which contradicts the original hypothesis that courage is one of a number of virtues. The Charmides ends with another unsuccessful attempt to define a specific virtue as some kind ⁹ His fullest argument for the early dating of the Gorgias is in Kahn [1988a] (discussed by Taylor [1991], pp. xviii–xx). In Kahn [1996] the commitment to an early date is more tentative. There Kahn stresses the importance of the Gorgias in paving the way for the defence of Socratic ethics (as expressed in the Apology and Crito) against the radical challenge posed by immoralism, a defence which, begun in the Gorgias, is finally achieved in the Republic. While still inclining to the view that the Gorgias preceded the other ‘threshold’ dialogues in which (in his view) the central themes of the Republic are adumbrated, he concedes that ‘[s]uch biographical speculation is not essential to the interpretation, and readers may prefer to think of all these dialogues as composed concurrently’, since ‘[t]he essential point ... is not chronological but thematic’ (p. 128). ¹⁰ I leave out of account the Apology and Crito, for in those works Socrates does not address the question of the nature of goodness, and therefore does not take up any stance for or against evaluative hedonism. In those works he does indeed treat virtue, especially justice, as an intrinsic good of the highest value, but it is naive to maintain, as Zeyl [1980] does, that that stance is so obviously incompatible with hedonism as to deprive a subsequent picture of a hedonistic Socrates of any credibility. For a fuller defence of this view see Gosling and Taylor [1982], 62–5.

278

the hedonism of the protagoras

of knowledge, in this case the suggestion that σωϕρoσύνη is knowledge of knowledge. The Euthydemus too ends in aporia; all goods other than wisdom are good only if directed by wisdom (the same conclusion as in the Meno). So wisdom is excellence in the kingly art which properly directs our life. But now what is the goal of this art? Not to provide those conditional goods which it directs. So its goal can only be wisdom. But what sort of wisdom? Not departmental wisdom such as shoemaking, for that is subordinated to the kingly art itself. So the goal of the kingly art can be nothing other than skill in the kingly art itself. But that answer is totally uninformative. Clearly, the informative answer that is required is that the kingly art aims at the achievement of human good, which in turn requires some informative account of what that good is. Similarly the Meno’s question invites the answer ‘knowledge of the good’, in turn inviting the further question ‘What is the good?’; the same question is prompted by the abortive investigations of the Laches and Charmides. My suggestion is that Plato provides an answer to this question in the Protagoras, namely that the good is an overall pleasant life, but that this answer is presented tentatively, as a hypothesis for examination, not wholeheartedly endorsed.¹¹ The merits of the thesis are its simplicity and its capacity to provide a grounding for ordinary evaluative beliefs. Its defects, as Plato saw them, were to be made apparent in those later dialogues, notably the Republic, in which the relation of pleasure to the good is examined in the light of developments in his metaphysics and psychology. If, as I have argued, Plato deliberately avoids portraying Socrates in the Protagoras as unambiguously endorsing hedonism, the traditional problem of reconciling the hedonistic Socrates of that dialogue with the antihedonistic Socrates of the Gorgias and Phaedo simply lapses. That said, the hypothesis which I have advanced leaves the picture of the development of Plato’s views on pleasure defended in Gosling and Taylor [1982]¹² ¹¹ It should be emphasized that this suggestion does not involve the claim that the Laches, Charmides, Meno, and Euthydemus were all written before the Protagoras. My hypothesis is that those dialogues provide evidence that at a certain period Plato was engaged with a number of interconnected questions, in which the thesis ‘The good is an overall pleasant life’ played a central role. It has nothing to say about the sequence in which he committed those questions to paper, or the sequence (which may have been different) in which he made those results available to the public. ¹² See especially pp. 97–8.

the hedonism of the protagoras 279 largely untouched. Assuming the Protagoras to be his earliest treatment of the topic, it will still be the case that his starting point is a theory which identifies pleasure with goodness, from which he retreats under the influence of a growing commitment to body–soul dualism, which obliges him to reject the view that all pleasures are good qua pleasures and that pleasure provides a single standard of value. From that perspective the value of pleasures is not ultimate, but dependent on the value of the subject and object of pleasure. Hence, since the soul and the objects with which the soul is concerned, i.e. the Forms, are more valuable than the body and its concerns, the pleasures of the soul are intrinsically more valuable than those of the body (Rep. 583b–586a), and indeed, since the pleasures of the body immerse the soul in illusion and distract it from reality, most bodily pleasures have negative value (Phd. 64c–69e, 80c–84b). The principal qualification to that view which I now propose is that whereas Gosling and Taylor take it for granted that the early hedonism is Plato’s own position, I now suggest that it is one which he presents for consideration as filling a crucial gap in his theory, without positively endorsing it.¹³ Afterword Both Rowe [2003] and Kahn [2003] maintain positions not dissimilar to that proposed in this chapter. The former writes (p. 142): ‘My own thesis is ... that Plato is prepared, in the Protagoras, at least to experiment with the thesis that the pleasant is the same as the good, in order to introduce the intellectualist position ... and that, at the very least, he is using the pleasant as a stalking-horse for the good ....’ He does, however, conclude that Socrates is to be seen not merely as taking hedonism seriously, but ‘in some sense’ endorsing some version of it (p. 146). Allowing for the indeterminacy of ‘in some sense’ his final view is thus closer to the position which I previously maintained, and from which I withdraw in this chapter. ¹³ Subsequently to the Prague conference I reread the brief appendix (pp. 167–71) on ‘Hedonism in the Protagoras’ in Annas [1999]. (I blush to admit that, despite my admiration for this splendid book, that particular section had completely slipped from my mind when I wrote the paper.) Annas is in agreement, for slightly different reasons, with my main contention that Plato introduces hedonism in the dialogue as a thesis ‘worth formulating and discussing’ (p. 170), without attributing it unambiguously to any character in the dialogue. She does not offer any suggestion as to why Plato thought it worth discussing, and does not, therefore, endorse the suggestion which I propose in the latter part of the paper.

280

the hedonism of the protagoras

Kahn characterizes the treatment of hedonism in the dialogue as ‘Plato’s quasi-hedonism’ and sums up its role as follows: ‘Meeting the many halfway, this dialogue offers a persuasive simplification which the more complex theory of motivation in the Republic is designed to replace’ (p. 174). That comes close to what I suggest in the penultimate paragraph of the present chapter.

16 Wisdom and Courage in the Protagoras and the Nicomachean Ethics At Prot. 359a Socrates begins his final attack on the thesis about courage and the other virtues which Protagoras had enunciated at 349d, viz. that while the other four principal virtues, wisdom, justice, soundness of mind (s¯ophrosun¯e), and piety or holiness (hosiot¯es) are fairly similar to one another, courage is totally different from all the rest. Protagoras supports this thesis by alleging that we find many people who are altogether lacking in the other virtues, but are remarkably courageous. The immediate aim of Socrates’ attack is to show that, since courage turns out to be itself a kind of wisdom, it is impossible for people totally lacking in wisdom (amathestatoi) to be courageous (360e); his further aim is to establish that, in consequence of that account of courage, all the specific virtues are kinds of knowledge, and hence that virtue in general is knowledge (361b). The final attack relies on the agreement of the sophists at 358a–e to the unitary account of motivation extracted from the imaginary dialogue with ‘the many’ (352a–357e). According to that account, no one ever chooses to do anything other than the action which he judges will produce the best available outcome (358c6–d4). Fear is agreed to be the expectation of evil (358d6–7), i.e. expectation of a bad outcome. Hence, Socrates claims, no one ever intentionally does something which he fears, when he has the option of doing something which he does not fear; since to do so would be intentionally to bring about an outcome which the agent judges bad, in preference to trying to bring about an outcome which he does not judge bad. But according to the general motivational principle just stated, that is impossible (358e).

282

wisdom and courage

Notice that what is agreed on is not the simple thesis that no one ever does what he fears, but the qualified thesis that no one ever does what he fears when he has the option of doing something which he does not fear. The general motivational principle on which it depends is that no one ever intentionally seeks to bring about any outcome other than the best available; but the best available outcome might still be bad, when one has to choose between evils, a situation which Socrates explicitly envisages (358d2–4). So one might have to seek to bring about an outcome which one expected to be bad (i.e. feared, by the definition), provided that one expected all available alternatives to be worse. In the attack on Protagoras which follows immediately, the situation of a choice between evils is not explicitly mentioned; it appears to be assumed that there is always open to the agent a choice between outcomes which are expected to be bad (i.e. feared) and outcomes which are not expected to be bad (i.e. not feared). The attack itself proceeds as follows. Socrates elicits from Protagoras a repetition of his earlier assertions that the courageous are bold (tharraleoi) and that they ‘go for’ what they want. Socrates then asks whether courageous and cowardly people go for the same things. Protagoras says ‘No’; cowards go for ta tharralea, courageous people for ta deina. That, at any rate, is the ordinary view, which Protagoras accepts. Socrates, however, dissents, pointing out that according to what has just been agreed it is impossible for anyone to go for something he takes to be deinon, since to take something to be deinon is to take it to be bad, and no one, it has been agreed, goes for anything he takes to be bad. So cowards and courageous alike go for the same things, viz. the things ha ge tharrousi, i.e. the things they feel tharros about, or, equivalently, the things which they find tharralea (on the translation of these expressions see below). Protagoras protests that the courageous and the cowardly go for opposite things; courageous men go into battle, cowards run away. Socrates replies that from the principles already agreed it follows that their opposed conduct expresses opposite beliefs about which outcome is worse, i.e. more to be feared. The courageous man stands firm because he makes the correct judgement that running away is worse, i.e. more to be feared, than facing death in battle; the coward flees because he makes the incorrect judgement that facing death is worse, i.e. more to be feared, than flight. Cowardice is thus error or ignorance (amathia) about what is and is not to be feared,

wisdom and courage 283 and its opposite, courage, wisdom, or knowledge (sophia) about the same things. It will be useful to say a word about the terminology used in the above summary. Protagoras’ assertion that the courageous are tharraleoi looks back to the argument of 349e–351b, where he had maintained that the courageous are tharraleoi but had denied the converse. The adjective is formed from tharros, whose basic sense is a positive attitude to danger, prompting the agent to ready engagement in dangerous situations. In the course of that discussion it emerges that there are various types of tharros, ranging from sheer animal ferocity and aggressiveness through delight in danger, especially in war, to confidence that one can escape unscathed, often as a result of one’s technical skill. The claim that courageous people are tharraleoi is then just the claim that they face danger readily; the source of their tharros is left unspecified. But the assertion that cowards go for ta tharralea is slightly less straightforward; it has to mean that cowards go for things that inspire them with tharros, but that can hardly mean ‘things that inspire them with a positive attitude to danger’, since ex hypothesi cowards lack that attitude. In the case of the coward the relevant sense of tharros is ‘confidence’; the only kind of thing the coward is ready to go for (i.e. to face) is something which he is confident is not really dangerous. So the sense is that cowards go for things that fill them with confidence, specifically the confidence that they will come out unscathed, whereas the courageous go for things that are to be feared, i.e. things where they believe that there is real danger. And now Socrates argues that courageous and coward alike go for ta tharralea, since it is impossible for anyone to go for what he thinks is bad, and only bad things are to be feared. In each case the expression ta tharralea picks out the way things appear to the agent in question, not the way they are in fact. We have just seen that for the coward ta tharralea are situations in which he is confident of escaping unscathed; what sort of situations are tharralea for the courageous person? Here too the crucial notion is that of confidence, but not confidence in escape from e.g. death or injury as in the case of the coward. The courageous person is willing to face those things because he is confident that facing them is a better thing to do than running away. What is common to both the coward and the courageous person is then their confidence that their action will promote the best outcome. The difference lies in their divergent conceptions of what the best outcome is; for the coward it is escaping injury or death,

284

wisdom and courage

for the courageous person it is promoting the agent’s eudaimonia by acting virtuously. In this picture fear and confidence are simply the opposite poles of the agent’s evaluative system; fear just is negative evaluation, i.e. evaluation as something overall bad, and confidence is positive evaluation, i.e. evaluation as something overall good. ‘Overall’ is important. I might start off by thinking of wounds and death as something bad, and therefore to be feared, but when in the light of my evaluation of the courses open to me it becomes apparent that by facing wounds and death I shall be acting courageously, and thereby promoting my eudaimonia, it is no longer open to me to look on facing wounds and death as other than good. I no longer see them as evils, and hence I no longer fear them (since fear is just the expectation of evil). Hence Plato’s psychology has no room for conflict between the desire to act courageously and the fear of wounds or death, and therefore no room for the description of an agent who acts courageously despite that fear. The nearest that this system can come to admitting conflict is in the contrast between incomplete and complete evaluation; incomplete evaluation of the situation can issue in evaluation based on partial evidence, e.g. in fear of wounds or death, but evaluation in the light of the agent’s complete value system eliminates that fear. Since these evaluations are (a) opposed and (b) made in the light of the agent’s complete evaluation of the situation, it is impossible for the agent to experience both simultaneously, for precisely the reason that it is impossible for the agent to have inconsistent all-inclusive evaluative beliefs. Finally, fear and confidence do not provide considerations which the agent can take into account in reaching his final evaluation, or promptings which might either reinforce or threaten the stability or effectiveness of that evaluation, since they are nothing other than those evaluations themselves. The fundamental difference between the account of courage in the concluding section of the Protagoras and Aristotle’s treatment of the virtue in the Nicomachean Ethics is found in the different views of fear and tharros which they embody. In the former, as we have just seen, they are identified with negative and positive evaluation respectively, and are hence the outcome of the agent’s deliberation, whereas for Aristotle they are motivational impulses to which the agent must respond in arriving at his decision. Hence they are not the outcome of deliberation, but part of the process leading to that outcome, as it were materials which the

wisdom and courage 285 agent shapes in deliberation. We must examine Aristotle’s account in more detail in order to determine whether his account of fear and tharros gives a better account of courage than that found in the final section of the Protagoras.¹ As is well known, Aristotle’s general theory of virtue of character (¯ethik¯e aret¯e) is that each specific virtue is a disposition with respect to a specific motivational impulse (or impulses), a disposition which is ‘in a mean’, i.e. which consists in being neither excessively given to that impulse nor insufficiently responsive to it. Courage has the special feature of being a correct disposition with respect not to a single impulse, but to two, fear and tharros (1107a33–b4). I shall consider below in more detail what Aristotle means by tharros, but at this preliminary stage it suffices to take it as standing for a range of positive attitudes to danger or harm, and thereby a set of motivational impulses prompting the agent towards situations of danger or harm, whereas fear prompts the agent to avoid such situations. The fact that courage has to do with these two opposed motivations complicates the standard picture according to which each specific virtue is opposed to two specific vices (kakiai), the one an excess of the motivation, the other a deficiency of it (as e.g. temperance is a mean between excessive indulgence in bodily pleasures (akolasia) and insufficient responsiveness to them (anaisth¯esia)). But since Aristotle is inexplicit on how far fear and tharros vary independently of one another, the details of the more complicated picture are obscure. One may be excessively or insufficiently fearful, and excessively or insufficiently ‘bold’ (using that adjective to render the notion of being responsive to tharros). Aristotle treats being insufficiently fearful as having ‘an excess of lack of fear’ (1107b1)² and distinguishes that from having an excess of tharros. He is correct to distinguish the two, since being insufficiently averse to danger is not the same disposition as being too keen on it; someone who is literally indifferent to danger satisfies the former description but not the latter. Yet while the two dispositions are not identical with one another, being too keen on danger implies being insufficiently averse to it. At the other extreme the person who is excessively fearful and the person who is insufficiently bold are characterized by the single term ‘cowardly’ (deilos), which implies that Aristotle thinks that those ¹ The following section of the paper incorporates some material from Taylor [2006]. ² The person characterized by this excess is described as t¯ei aphobiai (sc. huperball¯on). I take that description to be logically equivalent to ‘deficient with respect to fear’, i.e. t¯oi phobeisthai elleip¯on.

286

wisdom and courage

defects are at least inseparable (or perhaps even that there is a single defect describable in two ways). But here too the possibility of indifference shows that while being excessively fearful implies being insufficiently bold the converse is not true. Aristotle should therefore distinguish excess of fear from insufficiency of boldness, as he does insufficiency of fear from excess of boldness. Tharros is not, then, simply absence of fear; if it were, courage would be the correct response to a single motivation, viz. fear, and would be intermediate between the excess of that motivation (which would also be describable as deficiency of tharros) and the deficiency of it (= excess of tharros); i.e. the complexities which we have just pointed out would not occur. Rather, tharros is a positive motivation, which involves finding something attractive in situations of danger or harm; while its motivational tendency is thus opposed to that of fear, which is aversion to danger or harm, it can coexist with fear. Aristotle has thus hit upon an important insight into the nature of courage, that it is a matter of correct response to opposed impulses arising in situations of danger. But in order to appreciate that insight fully we need to arrive at a clearer conception of what range of attitudes he includes under the heading of tharros. He is, however, regrettably inexplicit. In III. 7 he describes various types of excessive or deficient state opposed to courage. The first type is aggressive fury, typical of uncivilized peoples such as the Celts, who fear nothing, not even waves or earthquakes (1115b26–8; the parallel passage from Eudemian Ethics (1229b28–30) describes the Celts as drawing their swords and marching against the waves, adding that they do so from spirit (thumos) though they are aware how great the danger is). As this is one of the kinds of motivation described as tharros at Prot. 351a–b (cf. Lach. 197b), we might expect Aristotle to ascribe it to excess of tharros, but instead he ascribes it to excessive absence of fear, apparently distinguishing it from conduct prompted by excessive tharros. A description of the latter follows immediately (1115b28–32); the person who is excessive in tharros is a boaster who pretends to be courageous but is really cowardly, putting on a show of eagerness to face danger in advance, but shirking it when the time comes. Taken literally, that surely describes someone who is not actually excessive in boldness, but someone who merely pretends to be. We must, therefore, take it that Aristotle here expresses himself somewhat loosely; he intends to describe, not someone who merely pretends to be

wisdom and courage 287 eager to face danger though he is really cowardly all the time, but someone who is genuinely eager for it in advance, but finds his courage ebbing away as the moment of danger approaches. But if boldness is just eagerness to face danger, why are the Celts who are eager to fight the stormy sea not a paradigm of excess in that respect? Perhaps the answer is that Aristotle is assuming the conception of tharros expressed in Rhet. 1383a17–19 ‘hope, accompanied by imaginative representation, that safety is near at hand and fearful things non-existent or far off’. The Celts are filled, not with hope that they will escape the waves, but with mad fury against them, while the ‘bold coward’ (thrasudeilos, Aristotle’s term for the person with excess of tharros described above) starts off by hoping to escape but loses hope as danger nears. While that gives a good explanation of the different treatment of the two types, some problems remain. (a) In what sense is the bold coward’s hope excessive? Is it unreasonable hope? But the courageous person normally hopes to escape (in the sense of surviving) the danger, and that hope is ex hypothesi not unreasonable. I suggest that the bold coward is seen as starting off by making light of the dangers and exaggerating positive aspects of the coming danger, e.g. that he will win great glory by undergoing it, but as tending towards reversing those assessments the closer danger looms. If that is correct, fear and boldness are interdependent aspects of the agent’s attitude to danger. There are positive and negative aspects of envisaged danger, and the proper attitude to it consists in getting the balance between the two right. One can go wrong by tipping the balance either way; but either way the defect has to be described in the relation between the ways of representation. One can give too much weight to the fearful aspect (and thereby insufficient weight to the positive, e.g. that by facing danger in battle we stand to win a glorious victory) or vice versa; but there is no room for the description of someone who merely gives too much weight to the fearful without giving insufficient weight to the positive, or too much weight to the positive without giving insufficient weight to the fearful.³ ³ That is why it is impossible to be too averse to danger without being insufficiently keen on it, and too keen without being insufficiently averse (see above). Being too averse to it is giving too much

288

wisdom and courage

(b) Elsewhere Aristotle applies terms cognate with tharros to people who are unruffled in situations where they have no hope of escape (suffering financial loss (1115a21–2), facing a flogging (1115a23–4)). They face these situations with a calm demeanour which expresses a serene and untroubled frame of mind. But is the effect of that not merely to reduce tharros to fearlessness, which we have seen to be contrary to Aristotle’s express view? That need not be so. Here too we may appeal to the idea of positive and negative assessments of the situation of danger or harm; the unruffled calm can be seen as the outward expression, not merely of lack of fear, but of assessments of the situation under such descriptions as ‘demanding dignity in the face of adversity’ or ‘a chance for me to show what I’m made of’. Aristotle’s conception of tharros is therefore primarily cognitive; it is motivation prompted by positive assessment of envisaged danger or harm. Sheer animal fury, which in ordinary idiom counts as tharros (see the passages of Protagoras and Laches cited above), lacks sufficient cognitive content to count, and it may be because he sees the Celts as driven by an aggressive spirit resembling that that he does not find it appropriate to describe them as prompted by an excess of tharros. But on the other hand hope or confidence of escape from danger are not the only forms of positive assessment; other forms might include assessment as exciting, glorious, noble, worthy of a Spartan, etc., i.e. any description under which the dangerous situation can be seen by the agent as attractive or worthwhile. This cognitive account of tharros has its counterpart in Aristotle’s treatment of fear, which is motivation prompted by negative assessment of envisaged danger or harm. There is this difference, that envisaged danger or harm are positively assessed not as such (i.e. they are not seen as attractive or worthwhile because they are dangerous or harmful), but for other reasons, such as those listed above. Danger and harm, on the other hand, are negatively assessed under those very descriptions, since harm is the loss of good, an intrinsic evil, which is accompanied by pain and distress, further weight to the negative factors relative to the positive, which implies giving too little weight to the positive factors relative to the negative, and being too keen is giving too much weight to the positive factors relative to the negative, which implies giving too little weight to the negative factors relative to the positive. The indifferent person, who gives no weight at all to either type of factor, gives insufficient weight to both without giving too much weight to either.

wisdom and courage 289 intrinsic evils, and danger is the prospect of those evils. Aversion to danger and harm, i.e. fear, is part of a properly functioning human psychology. Wounds, pain, and death are among the things which Aristotle counts as ‘fearful for a human being’ (ta ... kat’ anthr¯opon (sc. phobera) (1115b9)), death being the most fearful thing of all (1115a26). The attitude of the courageous person to these things is to be ‘undisturbed as appropriate to a human being’ (anekpl¯ektos h¯os anthr¯opos (1115b11), but being undisturbed in that way is not a matter of feeling no fear; rather it is not being overwhelmed by the fear which one does feel. The courageous person ‘will fear such things, but will endure them as he should and as reason prescribes for the sake of the fine’ (1115b11–13). Courage, then, consists in facing up to the prospect of what we find unpleasant and naturally wish to avoid; it is therefore itself something unpleasant (epilupon (1117a33–4)), and paradoxically the better a person one is the more distressing it is to face up to the prospect of loss of life and the things that make life worth living, for the better one is the more worthwhile one’s life is, and hence the more one has to lose by death (1117b9–13). Hence the courageous person will undergo dangers reluctantly (akonti (b8)), though ‘he will endure them because it is fine to do so and disgraceful not to’ (b9). Earlier (II. 3, 1104b3–8) Aristotle has said that pleasure and distress are differentiating marks of different types of character, explaining (via examples) that ‘the person who abstains from bodily pleasures and takes pleasure in so doing is temperate, while the person who finds it disagreeable to do so is intemperate, and the person who endures frightening things with pleasure, or at least without distress (m¯e lupoumenos) is courageous, while the person who feels distress at doing so is cowardly’. There is an obvious tension between this passage and those cited in the preceding paragraph. Aristotle appears to be aware of the difficulty, which he attempts to meet at 1117a35–b6 by comparing the courageous person facing danger to a boxer who puts up with the unpleasantness of arduous training and the pain of being knocked about in the ring for the sake of the crown and the other rewards of victory. These will indeed be very pleasant when they arrive, but when one is actually in the ring the prospect of them is overshadowed by the pain one is enduring. This does not provide an adequate answer. First, it presents the value of courage as something purely instrumental, like that of a painful medical procedure; in itself courage is something irksome, which is worthwhile only for the sake of subsequent rewards. That cannot

290

wisdom and courage

easily be reconciled with Aristotle’s requirement that fully virtuous actions are chosen for their own sake (di’ auta (1105a32)). Nor can it meet the demand of 1104b7–8 for some kind of positive response at the time of action to the actual situation of facing danger. Secondly, the example of the boxer represents the rewards of courage as bestowed by others, which tends to assimilate true courage to civic courage; but Aristotle has just distinguished true courage from civic courage in the preceding chapter of Book III (1116a17–b3). The ground of that distinction is precisely that whereas the truly courageous person is motivated by his own internalized values, the person of civic courage is motivated by external factors such as rewards and regard for the opinion of others. A better solution of the difficulty is provided by Aristotle’s account of courage as the appropriate response to both fear and tharros. The courageous agent does indeed regard imminent death as involving a loss which he shrinks from and would prefer not to have to undergo. But at the same time he regards the very act of undergoing that loss as something admirable, and as such undertakes it either gladly or at least without reluctance. Every sane person, in Aristotle’s view, faces death reluctantly. The difference between the courageous person and the coward is that reluctance is the latter’s overall attitude (since the value of the courageous act is either outweighed or is not present to his mind at all) whereas for the courageous person reluctance is the response to the situation qua loss of goods, which is outweighed by the value of the situation qua courageous act. That is to say, the courageous person experiences both fear (of pain and death etc.) and tharros, i.e. enthusiasm (or something similar) at the prospect of doing something noble. The right response to these feelings, which manifests the agent’s courage, is that in which tharros predominates, so that the fearful situation is accepted gladly or at least readily. Anthony Duff makes a somewhat similar suggestion.⁴ He describes a martyr who ‘goes willingly (even gladly) to her death, because she sees that this is required of her: she is of good hope and good cheer; she exhibits, I suggest, an appropriate kind of tharsos [an alternative spelling of tharros]. Her tharsos does not consist in any expectation of survival—she has none—nor in a love of danger for its own sake—she finds neither value nor pleasure in mere danger. It consists rather in her hopeful confidence [author’s italics] in ⁴ Duff [1987], 10–11.

wisdom and courage 291 the worth of her action and her sacrifice: she is willing and happy to give up her life in this cause. ... Her tharsos provides an appropriate pathos which partly constitutes her courage: she takes pleasure in facing the right dangers for the right reasons; her passion speaks with the same voice as reason, since it is informed by her understanding of what is good and noble.’ Duff goes on to suggest (11) that the martyr’s natural fear of death is not simply outweighed by tharros thus described, but transformed, so that the painful expectation of death, though still present, no longer disturbs her nor prompts her to flight. Similarly, Michael Stocker writes:⁵ ‘[C]ourageous warriors maintain their fears and master them. But the sort of mastering important here does not involve a continuing struggle with the fears nor does it involve getting rid of them. Rather, it involves integrating the dangers and the victory, the fear and the confidence, into one coherent and settled emotional appreciation of the situation. ... [T]he courageous warrior with these fears will have a different sort of eagerness. Rather than a light and easy eagerness it will be a more solemn, studied or regretting eagerness ... attaining victory is pleasurable, but acting courageously need not be pleasurable and may indeed be painful ... Such solemnity, regret and lack of pleasure need not involve indecision, vacillation, ambivalence or other lack of whole-heartedness. The fear had by courageous warriors does not involve conflict. Rather there is the regret or sadness that the goal involves such a risk ... the fear and the confidence have interpenetrated each other, losing their separate identities.’ These penetrating studies, which undoubtedly provide subtle and illuminating analyses of various kinds of tharros, and thereby much valuable insight into certain special types of courage, share the view that Aristotle’s courageous person must be completely free from the inclination to avoid pain or death, and thereby, in respect of motivation, totally unconflicted. Such conflict as exists in that person is a conflict, not of motivation, but of values, in that the action which the courageous person gladly undertakes is seen as involving both the positive values of nobility etc. and the negative values of pain and death. Given that someone in the exalted state of Duff’s martyr is totally unconflicted with respect to motivation, one might well query whether she feels any fear at all. If her negative assessment of the situation is no more than sadness at the loss of what she has valued in life, ⁵ Stocker [1990], 144–5.

292

wisdom and courage

that seems insufficient to amount to fear, and in any case she need not feel even that, if her sorrow at parting from life is turned into the joy of witnessing for her faith. Unlike the martyr, Stocker’s clear-sighted warriors do feel regret at having to be in a situation demanding self-sacrifice, and would presumably be glad if, in the event, that sacrifice was not required. But such regret, lacking motivational force, does not amount to fear. These descriptions read more like accounts of superhuman courage than of what we recognize as the virtue of the person of ordinary courage. Fear, as we ordinarily conceive of it, has motivational force, and if courage is, as Duff and Stocker suggest, incompatible with motivational conflict, then we are faced with the choice of revising our concept of fear or of conceding that the person of Aristotelian courage does not, after all, feel fear at the time of action. We may feel forced to that unpalatable choice by the thought that only totally unconflicted courage can be distinguished from self-control. That seems to me a mistake: the self-controlled person has bad desires which he or she controls through reason (1154b12–13), but the desire to avoid pain and death, even in circumstances when undergoing them is fine, is not a bad desire, but an aspect of properly functioning human nature. In the case of fear, the self-controlled person would be someone who was by instinct cowardly, i.e. someone who instinctively sought to avoid any and every distress, but mastered that bad impulse through reason. A normally courageous person, on the other hand, is not someone cowardly by nature, but someone who, faced with imminent and serious harm, feels the appropriate amounts of fear and tharros. Such a person can still feel some degree of genuine motivational conflict, provided that his or her overall attitude is as I have described. This conflict is more than a conflict of values. It has genuine motivational content, in that it takes effort on the part of the courageous person to overcome his or her fear; on the other hand, that effort is exerted gladly, without hesitation or vacillation. My emphasis on the cognitive character of fear and tharros in Aristotle may seem to bring them close to Socrates’ account of those attitudes in Prot. 359–60, where, I argued above, fear just is assessment as something bad and tharros assessment as something good. While both accounts are cognitive in character, they diverge in this crucial respect, that whereas the Socratic account equates the attitudes with total assessments resulting from the agent’s practical thought, Aristotle sees them as directed to specific

wisdom and courage 293 aspects of the total situation, and thus as providing input to the agent’s practical thought. This enables him to describe courage not simply as a display of tharros, but as the correct response to both fear and tharros, and thereby to do justice to the fact that courage, unlike other virtues of character such as temperance or generosity, essentially involves a conflict of inclinations within the agent, since the prospect of noble action in the face of death (Aristotle’s paradigm case of courage) requires the agent to see the situation as one in which evils have to be weighed against goods. While Aristotle’s general ethical ideal is that of effortless harmony between the agent’s impulses and practical judgement, he has to recognize that that is impossible in the case of courage, since it could be attained only by someone insensible of the losses involved in injury and death, i.e. someone ‘far from being human’ (1119a9–10).⁶ Aristotle is therefore better equipped than Socrates (in Prot. 359–60) to do justice to the special character of courage as the proper response to conflicting impulses. He is also able to preserve more of our ordinary ways of describing courage. He is not obliged to reject, as Socrates is, the commonsense description of the opposite motivations of the courageous person and the coward. He has no problem in accepting that the courageous person goes for what he sees as fearful, whereas the coward flees from it. Socrates has to deny that because he identifies seeing something as fearful with assessing it as bad overall (359d); Aristotle can accept it because he has a psychologically more realistic account of fear as a response to something seen as bad in a certain specific way, viz. as conflicting with our natural instinct for selfpreservation and the avoidance of pain. Socrates’ view of fear and tharros yields the paradox that the motivations of the courageous person and the coward are identical; both go for ta tharralea, i.e. the things they judge best overall. The difference lies wholly in their divergent assessments of what is better and what worse overall. Aristotle is not committed to that paradox. His courageous person certainly does what he thinks best overall, since his disposition is to give the appropriate weight to the conflicting considerations presented by fear and tharros. But his coward need not do ⁶ While the actual context of the quotation refers to someone totally insensitive to bodily pleasures, it applies equally to fear. As suggested above, the person totally unconflicted in face of imminent death might be far from human in the direction of being divine. Such a person is not insensible (as a beast is) of the losses involved in injury or death; rather those losses have no significance when viewed from the special perspective of superhuman courage.

294

wisdom and courage

what he thinks best overall, i.e. best in the light of all available reasons. Because he has not developed appropriate responses to fear, he may be simply overwhelmed by the prospect of pain and death, to the extent that he is unable to give proper consideration to other reasons for acting, or at the extreme to consider them at all. Or he may weigh up the various considerations, and decide in the light of them that he ought to stay at his post, but still give way to fear and run away. The second of those descriptions characterizes the coward as one of Aristotle’s two types of akrat¯es, the weak akratic, who reaches a decision by deliberation but does not stick to it (1150b19–22). The first description suggests the other type of akrat¯es, the impetuous (propet¯es) person, who does not complete his deliberation because he is overwhelmed by passion (1150b19–28). But that assumes that the coward is capable of complete deliberation about dangerous situations, but is prevented from completing it by the onset of fear, whereas the truth may be that he is in general incapable of deliberating properly about such situations, because fear effectively blots out any other considerations. Such a person is not even an akratic, much less someone who decides, as the courageous person does, what it is best to do, and who differs from the courageous person only in his judgements of what is better and what worse. Socrates’ coward is simply someone who reaches the wrong evaluative judgement through a failure in deliberation. Aristotle’s description of cowardice covers a range of different cases, in some of which the effect of fear is not to cause deliberation to yield the wrong result, but to disable deliberation either in part or altogether. In that respect it is closer to the truth.

17 Socrates1 Socrates is the first Western philosopher to have left to posterity any sense of his individual personality, and is a central figure in the subsequent development of philosophy. Both of these aspects are due primarily to Plato. It is via his portrayal by Plato’s literary genius that Socrates is a living figure for subsequent generations, and thereby an exemplar of the ideals of philosophy, above all dedication to truth and intellectual integrity. It was under the influence of Socrates that Plato applied systematic techniques of argument pioneered by Socrates and his contemporaries, the sophists, to the fundamental questions of human nature and conduct which primarily interested Socrates, thereby placing ethics and psychology at the centre of the philosophical agenda. But while Plato brings Socrates to centre stage he also hides him; since Socrates wrote nothing himself we depend on others for our knowledge of him, and it is above all Plato’s representation of Socrates which constitutes the figure of perennial philosophical significance. But that representation was itself the expression of Plato’s understanding of an actual historical individual and the events of his life. We must, therefore, begin with a brief account of the little we know of that individual and those events.

Life Socrates was born in Athens around 470 bc, and lived in the city all his life, apart from military service abroad. We know very little of the circumstances of his life. His father, Sophroniscus, is said by some ancient sources to have been a stonemason, and in Plato’s Theaetetus (149a) Socrates ¹ A version of this chapter was given as the 2004 A. E. Taylor Lecture at the University of Edinburgh.

296

socrates

says that his mother, Phainarete, was a midwife. That may indeed be true, though the fact that the name literally means ‘revealing excellence’ suggests the possibility that Plato has invented the story in allusion to Socrates’ role as midwife to the ideas of others (Tht. 149–51). Since Socrates served in the infantry, who had to provide their own arms and equipment, his circumstances, at least initially, must have been reasonably prosperous, but Plato and other writers emphasize his poverty in later life, which they attribute to his spending all his time in philosophical discussion. The same sources stress that, unlike the Sophists, he never took payment for his philosophical activity, and he may have depended largely on support from wealthier friends. During his lifetime Athens became the principal centre of intellectual and cultural life in Greece, attracting from all over the Greek world intellectuals who developed and popularized the tradition of natural philosophy begun by the Ionian philosophers of the previous century, together with exciting new argumentative techniques and radical questioning of traditional beliefs about theology, morals, and society. Socrates was actively interested in most of these areas. Plato and others attest to his interest at one stage in questions of cosmology and physiology, though the sources agree that his interests subsequently shifted to fundamental questions of conduct. He never engaged in formal philosophical instruction, or set up any school; his philosophical activity consisted in informal conversation, partly with a circle of mainly younger associates whom he attracted by the force of his intellect and personality, but also with others, including sophists and prominent citizens. Some of his associates, including Plato and some of his relations, were opposed to the Athenian democratic system, and it may be that Socrates shared that attitude to some extent. He married relatively late in life; at the time of his death at about the age of 70 his eldest son was an adolescent, and he had two more small sons, the younger probably a baby. His wife (who must have been at least thirty years younger than he) was Xanthippe. Her bad temper (attested by Xenophon and others, but not by Plato) became legendary; stories of her abuse of Socrates, and his equanimity in putting up with it, were a stock comic theme from antiquity to modern times. Thus Chaucer’s Wife of Bath describes in the Prologue to her tale (727–32) how Socrates sat quietly while Xanthippe ‘caste pisse upon his heed,’ merely remarking mildly ‘Before the thunder stops it comes on to rain.’ (The story goes back to Diogenes Laertius’ life of Socrates,

socrates 297 Lives of the Philosophers II. 36.) One element in this comic tradition is the story that Socrates had another wife, or possibly a concubine, while married to Xanthippe; stories of how the two women switched from quarrelling with one another to concerted assaults on Socrates afforded rich material. Ancient sources attribute the origin of this tale to Aristotle, but the supposed original source is lost, and the historical basis extremely dubious. We know of no specific events in Socrates’ life till after the outbreak of the Peloponnesian War with Sparta in 432. He served with distinction in various campaigns, most notably the battle of Delium in 424, where it was said (Plato, Lach. 181b) that if everyone had behaved like Socrates the battle would not have been lost. By the 420s he had become sufficiently well known to be caricatured in several comic dramas. In the single example to survive complete, the Clouds of Aristophanes, first produced in 423, he appears as a representative of subversive contemporary tendencies, the head of a disreputable academy whose curriculum combines training in argumentative trickery with atheistic natural philosophy. Later, in his Apology (Defence of Socrates), Plato represents this portrayal as the origin of prejudice against Socrates which culminated in his condemnation on charges of impiety and corruption of the young (18a–19d); we have no reason to discount that evidence. The only occasion on which Socrates is known to have intervened in public life took place in 406. After a naval engagement the Athenian commanders had failed to pick up survivors, and the popular assembly voted to try them collectively, instead of individually as required by law. At that period most civic offices were assigned by lot, and Socrates happened to be a member of the executive committee whose function was to prepare business for the assembly. In that capacity he was the only one to oppose the illegal proposal. A few years later when, after final defeat in the war, the democracy was temporarily overthrown by a junta known as the Thirty Tyrants, he showed the same adherence to legality and morality by refusing, at the risk of his own life, to obey an order from the tyrants to take part in the arrest of an innocent man. It is likely that he remained neutral during the civil war in which the tyranny was overthrown, since he had friends in both camps; in particular, two of the most prominent among the tyrants, Critias and Charmides, both relatives of Plato, were among his close associates.

298

socrates

It is probable that that was at least a contributory factor in the accusation brought against him under the restored democracy. The explicit charges were failure to recognize (or perhaps ‘to believe in’) the gods of the state religion and the introduction of new divinities, coupled with corruption of the young. The case was tried early in 399, and the prosecution demanded the death penalty. We have no evidence of the detail of the prosecution’s case. On the religious aspect the prosecutors may have sought to represent Socrates as the leader of an illegal private cult, and may have used his claim, amply attested by Plato, to be guided by a private divine sign or voice in support of that charge. It is highly likely that the charge of corruption centred on his associations with notorious enemies of the state, particularly the tyrants mentioned above and Alcibiades, an intimate of Socrates who had instigated a disastrous invasion of Sicily in 415 and had later defected to Sparta. Our knowledge of the trial is based on two versions of Socrates’ defence, by Plato and Xenophon, each of whom, while preserving a core of fact, presents the defence in the light of his own agenda; Xenophon relies wholly on Socrates’ adherence to conventional piety and morality, while Plato gives a radically unconventional picture of Socrates’ philosophical activity as the fulfilment of a divine mission to perfect the souls of his fellow-citizens by subjecting their basic beliefs and values to philosophical criticism. Socrates was condemned to death. Plato’s Phaedo gives a moving picture of his last hours, spent among his followers in discussion of the immortality of the soul and the task of philosophy to free it from the trammels of the body, followed by his tranquil death from self-administration of hemlock. While there is dispute about the relative degrees of realism and idealization in the description of the effects of the poison, there is little doubt that the primary aim of the whole work is less historical accuracy than depiction of the ideal philosophical death.

Socratic Literature Besides Plato and Xenophon no fewer than nine associates of Socrates are reported by various ancient sources as having written imaginative accounts of Socrates’ conversations, creating a body of literature collectively known as ‘Socratic conversations’ (or ‘discourses’) (S¯okratikoi logoi). For the most part only the titles of these works survive, indicating that Socrates’ relations

socrates 299 with certain individuals, especially Alcibiades (who figures prominently in some Platonic dialogues, notably the Alcibiades and the Symposium), were a theme common to Plato and the other Socratic writers. Apart from Plato and Xenophon, the only Socratic writer of whose works any significant fragments survive is Aeschines of Sphettus; the fragments of his Alcibiades show Socrates using his characteristic critical method (see below) to convince Alcibiades of the vanity of his political ambitions. They thus provide evidence that the programme of defending Socrates against the slanders occasioned by his associations with political undesirables was not confined to Plato and Xenophon, but they provide no evidence for Socrates’ thought to complement those sources. For information specifically about the thought of Socrates we are in fact almost wholly dependent on Plato, since our other principal source, Xenophon, focuses on the practical and moral import of Socrates’ conversations, with comparatively little theoretical content, in keeping with his overall purpose (see above) of portraying Socrates as a good man and sound citizen. There is a systematic difficulty in determining which of the views attributed to Socrates in Plato’s dialogues were actually held by the historical person, and scholarly opinion has embraced all possible positions. In the nineteenth century the dominant consensus (primarily on the part of German scholars) divided the Platonic writings into three broad groups, distinguished both chronologically and doctrinally. The first ‘early’ group, including Laches, Charmides, Protagoras and those dialogues dealing directly with the trial of Socrates (Euthyphro, Apology, and Crito) was generally held to give a veridical account of the personality, views, and philosophical activity of the historical Socrates. Thereafter Plato’s philosophy developed in directions independent of Socrates, and the importance of the dramatic figure of Socrates in the dialogues correspondingly declined, until its virtual disappearance in works such as the Sophist and Statesman, which were taken to be late, and its total disappearance from the Laws, unfinished on Plato’s death and generally regarded as his last work. This ‘developmental’ model was supported by the stylometric studies of the later nineteenth century, in which a number of scholars, working largely independently of one another, converged on the identification of six dialogues, Sophist, Statesman, Timaeus, Critias, Philebus, and Laws, as a group distinct in various features of style and vocabulary from the rest of the Platonic corpus, and fixed as late by the presence of the Laws. Parmenides, Phaedrus, Republic, and

300

socrates

Theaetetus, which are by the same criteria closer in style to the late group than the rest of the dialogues, were identified as ‘middle’ dialogues, and the remainder as ‘early’. While this developmental model, with its assumption that the early dialogues veridically represent the historical Socrates, is still highly significant today, notably in the influential work of Gregory Vlastos and others, it has undergone challenge from two opposite extremes, on the one side the thesis maintained by John Burnet and A. E. Taylor in the early twentieth century that all the doctrines attributed by Plato to Socrates in the dialogues were actually maintained by the historical Socrates, and on the other the views of those who, stressing the fact that all our information about Socrates derives from sources with their own literary and philosophical agenda, urge that the historical Socrates is inaccessible to us and should therefore disappear from the history of philosophy. The Burnet/Taylor thesis has few if any adherents today; not only does it present an implausible picture of a Plato who devoted the great part of his literary career to recounting the views of someone else, but it rests on an assumption about the nature of Plato’s attitude to Socrates, viz. that it would have been disrespectful to Socrates for Plato to do other than represent his views with historical accuracy, which seems totally foreign to the character of the dialogues themselves. It is clear from the dialogues that Plato’s attitude to Socrates was that the latter’s life and activity represented the paradigm of philosophy, and it is totally in keeping with that attitude that Plato should ascribe to Socrates what he (Plato) regards as the philosophical truth, whether or not Socrates himself had maintained it. What we may call the sceptical view of Socrates, on the other hand, is widely accepted today, and while its extreme versions are exaggerated and oversimplified, it is based on an important insight into the nature of our sources. The insight is simply that all our knowledge of Socrates is based on sources in which historical veridicality is at best one among the author’s concerns, and generally not the principal concern. Oversimplification consists in the characterization of these sources as fiction, as opposed to factual biography, and exaggeration in the conclusion that the historical Socrates is inaccessible. The dichotomy between biography and fiction seems to me inapplicable to the Socratic literature, including Plato’s Socratic dialogues (and indeed I doubt its appropriateness to most ancient biographical writing); Socratic conversation is a form of biography, but biography whose factual constraints

socrates 301 are looser than is standardly the case in the modern world. That is not to say that there are no factual constraints; Plato’s dialogues do present an actual historical individual, some of the events in whose life are known, and are no doubt faithful to the spirit and nature of the philosophical conversation which was that individual’s principal activity. But when it comes to specific doctrines, while there are some doctrines maintained by Plato’s Socrates which it is virtually certain that the historical Socrates did not maintain, there is none that it is certain that he did. In the first class the paradigm case is the theory of separate Forms (i.e. intelligible universal natures existing separately from their sensible instances) which we find maintained by Socrates in several dialogues, but which Aristotle (whose evidence I regard as independent of the dialogues on this point) explicitly says Socrates did not hold (Meta. 1078b27–b32). On the other hand, theses characteristically regarded as ‘Socratic’, e.g. that Virtue is Knowledge (see below), are not ascribed to Socrates by sources which are clearly independent of their appearance in the Platonic dialogues. They may in fact have been maintained by Socrates, or they may have been suggested to Plato, in the form in which they appear in the dialogues, by things which Socrates said. We cannot be sure, and in any case it is not of the first importance, since the philosophical significance of these doctrines consists in the role which they play, and the arguments by which they are supported, in the dialogues in which they appear. The brief account of Socrates’ thought which follows is to be understood as based on that assumption. It identifies some central themes in the portrayal of Socrates in those dialogues, generally considered comparatively early compositions, in which the personality and argumentative style of Socrates are more prominent than in dialogues devoted to the more systematic exposition of Plato’s own thought (see above). The attribution of any specific doctrine to the historical Socrates must be correspondingly tentative.

Thought i Disavowal of Wisdom In these dialogues Socrates is presented for the most part not as a systematic or authoritative teacher, but as a questioner and enquirer. His enquiries are all focused on questions of conduct, broadly understood, and frequently

302

socrates

consist of attempts to reach an agreed definition of some fundamental value, such as courage, or goodness in general. Typically Socrates is depicted as engaged with one or more people in conversation on some specific, often practical topic, which leads on to the more general issues just mentioned. Socrates elicits the views of his interlocutors on these issues and subjects them to critical examination, conducted with a minimum of philosophical technicality, and utilizing other assumptions, usually of a commonsense kind, which the parties to the discussion agree on. Usually this procedure reveals inconsistency among the set of beliefs (including the general thesis or proposed definition) which the person examined holds, which is taken as requiring the abandonment of the thesis or definition. Frequently the dialogue ends with the acknowledgement by Socrates and the others that, having failed to settle the general issue raised, they are unable to proceed further; they thus end up in a state of aporia, i.e. a state with no way out. This procedure of enquiry, rather than instruction, and its frequent aporetic outcome are in keeping with Socrates’ denial (Ap. 21b) that he possesses any wisdom (i.e. expertise). It is the mark of an expert to be able to define the concepts in the area of his expertise and to expound that area systematically, neither of which Socrates can do. In later antiquity Socrates was regularly reported as having said that he knew nothing, or, paradoxically, that he knew nothing except that he knew nothing. Either formulation goes beyond anything we find in Plato. Though Socrates frequently says in the dialogues that he does not know the answer to this or that particular question, he never says that he knows nothing, and occasionally makes emphatic claims to knowledge, most notably in the Apology, where he twice claims to know that abandoning his divine mission to philosophize would be bad and disgraceful (29b, 37b). What he does disavow is having any wisdom. He seems to apply the notion of wisdom firstly to divine wisdom, a complete and perspicuous understanding of everything, which belongs to the gods alone, and is consequently unavailable to humans, and then to human expertise of the sort possessed by craftsmen such as builders and shoemakers, a systematic mastery of a technique which enables its possessor to apply it successfully and to expound and pass it on to others. The sophists claimed to possess, and to teach to others, a practical expertise applying not to any specialized area of human activity but to human life as such, mastery of which guaranteed overall success in personal and political life; this was ‘the political craft’ (Ap.

socrates 303 19d–20c, Prot. 319a). Socrates rejects that claim, not on the ground that such expertise is not available to humans; but because the sophists’ activity fails to meet the ordinary criteria for human expertise, particularly that of being systematically learned and taught (Prot. 319d–320b, Meno 89c–94e). He denies that he possesses this expertise himself (Ap. 20c), but does not say that it is impossible that he, or any human being, should possess it. This disavowal of expertise is not incompatible with the claim to know particular things. The non-expert can know some particular things, but not in the way the expert knows them; specifically the non-expert is not able, as the expert is, to relate his particular items of knowledge to a comprehensive system which provides explanations of their truth by relating them to other items of knowledge and to the system as a whole. But that raises the problem of the source of Socrates’ non-expert knowledge of moral truths. Usually a non-expert knows some particular things because he has been told by an expert, or because he has picked them up from some intermediate source whose authority is ultimately derived from that of the expert. But Socrates does not recognize any moral experts, among human beings at any rate. So what is the source of his non-expert knowledge? The dialogues provide no clear or uniform answer to this question. Sometimes he suggests that the application of his critical method is sufficient, not merely to reveal inconsistency in his interlocutor’s beliefs, but to prove that some are false, and hence that their negations are true. Thus at the end of the argument with Callicles in the Gorgias he claims (508e–509a) that the conclusion that it is always better to suffer wrong than to do it has been established by ‘arguments of iron and adamant’ (i.e. of irresistible force), while conjoining that claim with a disavowal of knowledge: ‘I do not know how these things are, but no one I have ever met, as in the present case, has been able to deny them without making himself ridiculous.’ This presents a contrast between expert knowledge, which Socrates disclaims, and a favourable epistemic position produced by repeated application of Socrates’ critical method of argument. There are some propositions which repeated experiment shows no one capable of denying without selfcontradiction. While it is always theoretically possible that someone might come up with a way of escape from this position, realistically the arguments establishing those propositions are so firmly entrenched as to be irresistible. While it is an attractive suggestion that Socrates considers the moral truths

304

socrates

which he non-expertly knows to be of this kind, it receives no clear confirmation from the dialogues. There is, for instance, no indication in the Crito that Socrates’ unshakeable commitment to the fundamental principle that one must never act unjustly (49a) is based on critical examination of his and Crito’s moral beliefs. It has to be acknowledged that while Socrates indicates that critical examination is sometimes capable of establishing truth beyond at least the practical possibility of rebuttal, and sometimes suggests that he knows some moral truths on the strength of good arguments for them, he gives no general account of the grounds of his non-expert moral knowledge. ii Religion One might perhaps speculate that the source of Socrates’ non-expert moral knowledge is supposed to be divine revelation, but though Socrates’ attitudes to the divine are an important element in his portrayal by both Plato and Xenophon, neither in fact suggests that Socrates believed that his moral beliefs were divinely inspired. What he did believe, according to both writers, is that throughout his life he was guided by a private sign or voice which he accepted, apparently without question, as being of divine origin, but the content of that guidance appears to have been, not moral principles, but day-to-day practical affairs, and it had the peculiar feature that its guidance was always negative, warning Socrates against some course of action that he might otherwise have undertaken (Plato, Ap. 31c–d). Thus Xenophon reports him (Ap. 4) as explaining his failure to prepare his defence by the fact that the divine sign had told him not to, while in Plato’s Apology (40a–b) he says that he is confident that his conduct at his trial has been correct because the divine sign has not opposed it. Such a claim to continuous private divine guidance (as opposed to occasional private revelations, e.g. in dreams) was certainly unusual, and, as suggested above, it is quite likely that it at least contributed to the charge of religious unorthodoxy which was one of the grounds of his condemnation. The actual stance of the historical Socrates towards conventional religion is not altogether easy to reconstruct from the sources. Xenophon, as pointed out above, stresses his conventional piety, as measured by public observance and private conversation, e.g. his demonstration to an irreligious acquaintance of the providential ordering of the world, down to such details as the design of the eyelashes to shield the eyes from the wind (Mem. 1.4); on

socrates 305 that account it is difficult to see how the charge of impiety could have been brought at all. Plato’s presentation is more complex. He does indeed represent Socrates as concerned on occasion with prophetic dreams (Crt. 44a–b, Phd. 60e–61b) and with ritual, most famously in his report of Socrates’ last words ‘Crito, we owe a cock to Asclepius; pay it and don’t forget’ (Phd. 118a). But it is notable that all of these instances arise in the context of Socrates’ imminent death. When Plato represents Socrates as praying on various occasions throughout his life he almost always makes him pray for nothing but wisdom and virtue, while in his most extensive discussion of piety, in the Euthyphro, he suggests that Socrates thinks that what the gods require from humans is nothing other than moral virtue.² That fits well with his Apology, where Socrates’ rebuttal of the charge of impiety has nothing at all to say about ritual, consisting wholly in the claim that Socrates’ life has been the fulfilment of his divine mission to promote the welfare, identified with the moral virtue, of his fellow-citizens. Plato’s view of Socratic religion seems then to be that the essence of service to the gods is moral virtue, and that ritual fills its proper role, as in Socrates’ life and death, as a complement to the fulfilment of that primary task. If that reflects Socrates’ own view, then it is possible that it was seen by conservatively-minded contemporaries as presenting a radical challenge to traditional ideas of the relations between gods and humans, which were founded on the belief that divine favour and protection for individuals and the community were secured by performance of the appropriate prayers and rituals, and thereby as justifying his condemnation for neglecting the state religion in favour of a new religion of his own.³ iii Definitions In the procedure of enquiry sketched in (i) above the search for general definitions is central. This arises naturally from Socrates’ search for expertise; the expert knows about his or her subject, and according to Socrates the primary knowledge concerning any subject is precisely knowledge of what that subject is. The general pattern of argument in the dialogues is that some specific question about a subject, e.g. how is one to acquire goodness, is problematic in the absence of an agreed conception of what that subject is. Hence before the problematic question can be pursued, the definition ² See ch. 4 of this volume.

³ See Burnyeat [1997].

306

socrates

of the subject must first be sought. The problematic question may be of various kinds; it may be, as in the example above (from the Meno) how goodness as such is to be acquired, or how a specific virtue is to be acquired (courage in the Laches), or whether a virtue is advantageous to its possessor (justice in the Republic). The Euthyphro exemplifies another pattern; it is disputed whether a particular action, Euthyphro’s prosecution of his father for homicide, is an instance of piety or holiness, and Socrates maintains that the question will be settled when, and only when, the definition of piety is arrived at. This pattern has given rise to the accusation that Socrates is guilty of the ‘Socratic fallacy’ of maintaining that in general it is impossible to tell whether anything is an instance of a property unless one already possesses a general definition of that property. That general position would be methodologically disastrous for Socrates, since his approved strategy for reaching a definition is to consider what instances of the kind or property in question have in common, and it is impossible to do that if you have to know the definition before you can even identify the instances from which the definition is to be derived. In fact the argument of the Euthyphro does not involve that fallacy; even if it is granted that there are some disputed cases where the question ‘Is this an instance of F?’ cannot be settled without answering the prior question ‘What is F?’ it does not follow that there are no undisputed cases where instances of F can be recognized without a definition. In the Hippias Major, however, Socrates does argue (286c–e) that you cannot tell whether anything is fine or beautiful (kalon) unless you know, i.e. can give a definition of, what fineness or beauty is; so though the Socratic fallacy is not a pervasive defect of Socrates’ argumentative method, we do appear to find at least one instance of it in the dialogues.⁴ The question ‘What is F?’ can itself be understood in various ways; it may be a request for an elucidation of the linguistic meaning of the term ‘F’, or a request for a substantive account of what the property of F-ness consists in, including, where appropriate, the decomposition of a complex property into its components (e.g. goodness consists of justice, self-control etc.) and explanatory accounts of properties (e.g. self-control consists of the control of the bodily appetites by reason). The practical nature of the questions which often give rise to the search for definitions suggests that ⁴ For full discussion see Benson [2000], ch. 6.

socrates 307 the latter kind of definition is what is sought. Someone who wants to know how virtue is to be acquired will not be helped by a specification of the meaning of ‘virtue’ as ‘a property contributing to overall success in life’; what they are looking for is precisely an account of what it is that constitutes or guarantees success in life. That is confirmed by the fact that the Laches, Meno, and Protagoras, all of which start from the practical question of how either a specific virtue or goodness in general is to be acquired, converge on the suggestion that courage (in the Laches) and goodness (in the Meno and Protagoras) are identical with knowledge, which is itself part of a substantive theory of the nature of goodness (see next section). It must, however, be acknowledged that Plato shows no awareness of the theoretical distinction between a purely conceptual definition and the kind of substantive account which is favoured by the structure of the dialogues just mentioned. Even in the Meno, the dialogue in which definition is treated in the greatest detail, he gives model definitions of either kind without any explicit differentiation. Substantive accounts are favoured over conceptual definitions by his practice, not in the light of any theoretical discrimination between the two.⁵ iv Ethics The picture of Socrates as a non-expert enquirer outlined above needs to be qualified to this extent, that in some dialogues, specifically the Protagoras, Gorgias, and Meno, he is represented as arguing positively, though not conclusively, in favour of certain propositions which amount to at least the outline of a theory of human nature and of human good. The basic theses of this theory are: 1. Every agent has a single overall aim, the achievement of a completely satisfactory life for him or herself. 2. Knowledge of what constitutes such a life is both necessary and sufficient for the achievement of it. 3. Such a life consists in the practice of the virtues of justice, selfcontrol, courage, and holiness, which are identical with one another in that they are the application to different kinds of situation of the fundamental virtue of knowledge (of what the good for humans is and how it is to be achieved). ⁵ See ch. 9 of this volume.

308

socrates

Thesis 2 is the famous thesis that ‘Virtue is Knowledge’, from which together with thesis 1 follows the still more famous thesis that ‘No one does wrong willingly’ (the latter two often referred to as ‘The Socratic Paradoxes’). The idea expressed in the second paradox is that, since everyone necessarily has the single aim of achieving the best life for him or herself, any action which does not in fact promote that aim must be explained by the agent’s mistaken belief that it does promote it. Socrates is thus the first of a succession of philosophers throughout the ages to deny the possibility of acting against one’s better judgement (often ascribed to weakness of will); that position remains as controversial today as it was in antiquity. The identification of the conventional moral and social virtues as applications of the fundamental knowledge of what the human good is (with the implication that the virtues are identical with one another, conventionally labelled ‘The Unity of the Virtues’), though central to the proto-theory, is never adequately argued for. It is supported at Crt. 47e by an analogy between virtue of soul and health of body; justice and injustice are respectively the health and sickness of the soul. So, just as it is not worth living with a diseased and corrupted body, it is not worth living with a diseased and corrupted soul. But that is not an argument. Even granted that health is an intrinsically desirable and disease an intrinsically undesirable state, the crucial claims that justice is the health of the soul and injustice its disease require defence, not mere assertion. Plato supplies some arguments in the Gorgias, but they are weak. Socrates first argues that successful tyrants, who manifest the extreme of injustice, do not get what they really want, i.e. the best life for themselves, because their injustice is bad for them. The crucial argument for that conclusion (473a–475c) starts from the premiss, conceded by Socrates’ opponent Polus, that acting unjustly, while good (i.e. advantageous) for the agent, is disgraceful. It is next agreed that whatever is disgraceful is so either because it is unpleasant or because it is harmful. Since acting unjustly is clearly not unpleasant, it must therefore be harmful. Hence the life of injustice is harmful to the unjust agent. This argument fails because it ignores the relativity of the concepts of unpleasantness and harmfulness. To be acceptable the first premiss must be read as ‘Whatever is disgraceful to anyone is so either because it is unpleasant to someone or because it is harmful to someone.’ From that premiss it clearly does not follow that because injustice is not unpleasant to the unjust person it must be harmful

socrates 309 to that person. It could be harmful to someone else, and its being so could be the ground of its being disgraceful to the unjust person (as indeed we ordinarily think). Later in the dialogue (503e–504d) Socrates argues against Callicles that since the goodness of anything, such as a boat or a house, depends on the proper proportion and order of its components, the goodness of body and soul alike depend on the proper proportion and order of their components, respectively health in the case of the body and justice and self-control in the case of the soul. The analogy of health and virtue, simply asserted in the Crito, is here supported by the general principle that goodness depends on the organization of components, but that principle is insufficient to ground the analogy, since the proper organization of components is determined by the function, point, or aim of the thing which those components make up. So in order to know which organization of psychic components is the appropriate one for humans we need a prior conception of what our aims in life should be. One conception of these aims may indeed identify the optimum organization as that defined by the conventional moral virtues, but another may identify as optimum a quite different organization, say one which affords the maximum scope to certain kinds of self-expression, as exemplified by a figure such as the Nietzschean Superman. Socrates provides no argument to exclude that possibility. In addition to the failure to establish that virtue is always in the agent’s interest, the proto-theory is more deeply flawed, in that it proves to be incoherent. This emerges when we consider Proposition 2 ‘Virtue is Knowledge’, and ask what virtue is knowledge of. The answer suggested by the Meno and the Protagoras is that virtue is knowledge of the best life for the agent; given the standing motivation to achieve that life, knowledge of what it consists in will be necessary if one is to pursue it reliably, and sufficient to guarantee success in that pursuit. But that requires that the best life for the agent is something distinct from the knowledge which guarantees that one will achieve that life. ‘Virtue is knowledge of the best life for the agent’ will be parallel to ‘Medicine is knowledge of health,’ and the value of that knowledge will be purely instrumental and derivative from the intrinsic value of the success in life which it guarantees. But Socrates, as we have seen, treats virtue as analogous, not to medicine, but to health itself, and hence as intrinsically, not merely instrumentally valuable. Virtue is not, then, a means to some independently specifiable

310

socrates

condition of life which we can identify as the best life, well-being, or happiness (in Greek, eudaimonia); rather it is a constituent of such a life, and one of the most difficult questions about Socratic ethics is whether Socrates recognizes any other constituents. That is to say, for Socrates a life is worth living either solely or at least primarily in virtue of the fact that it is a life of virtue. The incoherence of the proto-theory thus consists in the fact that Socrates maintains both that virtue is knowledge of what the agent’s good is and that it is that good itself, whereas these two theses are inconsistent with one another. It could indeed be the case both that virtue is knowledge of what the agent’s good is and that the agent’s good is knowledge, but in that case the knowledge which is the agent’s good has to be a distinct item or body of knowledge from the knowledge of what the agent’s good is. So if Socrates is to maintain that virtue is knowledge he must either specify that knowledge as knowledge of something other than what the agent’s good is, or he must abandon the thesis that virtue is the agent’s good. There are indications in the dialogues that Plato was conscious of this difficulty. In the Euthydemus he represents Socrates as grappling inconclusively with the problem, and in the Republic he offers a solution in a conception of human good as consisting in a state of the personality in which the non-rational impulses are directed by the intellect, informed indeed by knowledge, but by knowledge not of human good, but of goodness itself, a universal principle of rationality. This conception retains from the proto-theory the thesis that human good is virtue, but abandons the claim that knowledge is virtue, since virtue is not identical with knowledge but directed by it, the knowledge in question being knowledge of the universal good. The Protagoras may plausibly be seen as exploring another solution to this puzzle, since in that dialogue Socrates sets out an account of goodness whose central theses are (i) virtue is knowledge of human good, (ii) human good is a life in which pleasure predominates over distress. Whether Socrates is represented as adopting this solution in his own person, or merely as proposing it as a theory which ordinary people and sophists such as Protagoras ought to accept (a question on which there has been much dispute),⁶ it represents a way out of the impasse which blocks the proto-theory, though not a way which Plato was himself to adopt. Having ⁶ For discussion see ch. 15 of this volume.

socrates 311 experimented with this solution, which retains the identity of virtue with knowledge while abandoning the identity of virtue with human good, he settled instead for the Republic’s solution, which maintains the latter identity while abandoning the former. The proto-theory is not strictly inconsistent with Socrates’ disavowal of wisdom or expertise, since it is presented in outline only, not established by conclusive argument as expertise requires. But the presentation of Socrates as even a proto-theorist has at least a very different emphasis from the depiction of him simply as a questioner and generator of aporiai. I believe that it is impossible to tell how much of this theory is Plato’s own, and how much was actually held by Socrates. That it was at least suggested to Plato by certain ideas which had emerged in Socrates’ conversations seems highly likely, but we are not in my view justified in asserting more than that.

Later influence The proto-theory just sketched was an important element in the development of Plato’s own ethical theory, and via Plato on those of Aristotle and the post-Aristotelian philosophical schools. With the exception of the Epicureans, each of the main schools adopted Socrates as in effect a patron saint, stressing aspects of his thought and personality congenial to its particular philosophical standpoint; the sceptics, especially those in the Platonic Academy, which was converted to scepticism by Arcesilaus just over a century after its foundation and remained sceptical for two centuries, stressed Socrates’ disavowal of wisdom and the undogmatic character of his questioning technique. The Cynics, whose doctrines and way of life derived from Antisthenes, one of Socrates’ associates, claimed to emulate the austerity of his lifestyle and to accept his doctrine that virtue is sufficient for happiness. Via the Cynics Socrates became a major influence on Stoicism, which combined the Cynic doctrine that happiness consists in living according to nature with the doctrine that for rational beings the life according to nature is the life in accordance with rationality. Accepting the essentials of the proto-theory outlined above they drew the conclusion that moral virtue is the only good, everything else being indifferent, i.e. neither good nor bad. A particularly significant figure in the Stoics’ canonization of Socrates is Epictetus, who adopted Socrates as the exemplar of the

312

socrates

philosophical life and reproduced in his protreptic discourses features of Socratic method such as elenctic and inductive arguments. The influence of Socrates was not confined to the ancient philosophical schools. The second-century Christian apologist Justin claimed him as a forerunner of Christianity, a characterization which was revived by Renaissance Neoplatonists such as Marsilio Ficino. In medieval Islam he was revered, though not well understood, as a sage and a defender of (and martyr for) monotheism against idolatry. In the Enlightenment era he was appropriated by rationalists such as Voltaire as an exemplar of natural virtue and a martyr in the struggle of rationality against superstition. In the nineteenth century Hegel, Kierkegaard, and Nietzsche identified him as a central figure in developments in the history of philosophy to which their own respective theories responded, and in the last quarter of the twentieth century he was a major influence on the later thought of Foucault. The perennial fascination of Socrates owes less, however, to any specific doctrines than to Plato’s portrayal of him as the exemplar of a philosophical life, i.e. a life dedicated to following the argument wherever it might lead, even when it in fact led to hardship, poverty, judicial condemnation, and consequent death. Plato’s depiction of how Socrates lived for philosophy would in any case have made him immortal; his presentation of how he died for it has given him a unique status in its history.

bibliography 313

Bibliography

Ancient Sources Diogenes Laertius, ed. R. D. Hicks, Loeb Classical Library (Greek with facing English Translation), Cambridge, Mass., 1925. Ferguson, John, Socrates: A Source Book, London, Macmillan for the Open University Press, 1970. This contains a comprehensive collection of passages of ancient works (in English translation) referring to Socrates. Plato. Oxford Classical Text, ed. John Burnet, 5 vols., Clarendon Press, Oxford, 1902–6. New edition of vol. i, ed. E. A. Duke et al., 1995. New edition of Republic, ed. S. Slings, 2003. All the dialogues are available in numerous English translations. Xenophon, Memorabilia, ed. E. C. Marchant, Loeb, London, 1923. Apology and Symposium ed. O. J. Todd, Loeb, London, 1961. English translation, Xenophon, Conversations of Socrates, trans. Hugh Tredennick and Robin Waterfield, Penguin Classics, Harmondsworth, 1990. Aristophanes. Clouds, ed. K. J. Dover, Clarendon Press, Oxford, 1968. Translated by B. B. Rogers, Loeb, London, 1924. Minor Socratic writers. Socratis et Socraticorum Reliquiae, ed. Gabriele Giannantoni, 4 vols., Bibliopolis, Naples, 1991. English translation of the principal fragments of Aeschines in G. C. Field, Plato and His Contemporaries, Methuen, London, 1930, ch. 11. Modern Works Comprehensive Survey Guthrie, W. K. C., A History of Greek Philosophy, vol. iii, part 2, Cambridge University Press, Cambridge, 1969. Published separately 1972 under title Socrates. Socratic Literature and the Problem of the Historical Socrates Burnet, John, Greek Philosophy, Thales to Plato, Macmillan, London, 1914. ‘The Socratic Doctrine of the Soul’, Proceedings of the British Academy 7 (1915–16), 235–60.

314 bibliography D¨oring, Klaus, ‘‘Sokrates, die Sokratiker und die von ihnen begr¨undeten Traditionen’’, in Helmut Flashar, ed., Grundriss der Geschichte der Philosophie: Die Philosophie der Antike 2/1. Sophistik, Sokrates, Sokratik, Mathematik, Medizin, Schwabe AG Verlag, Basel, 1998, 139–364. Kahn, Charles H., Plato and the Socratic Dialogue, Cambridge University Press, Cambridge, 1996. Patzer, Andreas, ed., Der historische Sokrates, Wissenschaftliche Buchgesellschaft, Darmstadt, 1987. Rutherford, R. B., The Art of Plato, Duckworth, London, 1995. Taylor, A. E., ‘‘Plato’s Biography of Socrates’’, Proceedings of the British Academy 8 (1917–18), 93–132. Socrates, Peter Davies, London, 1932. Varia Socratica, James Parker and Co., Oxford, 1911. Taylor, C. C. W., Socrates, Oxford University Press, Oxford, 1998. Vander Waerdt, Paul A., ed., The Socratic Movement, Cornell University Press, Ithaca, NY, 1994. Critical and Analytical Studies, Primarily on Plato’s Presentation of Socrates Benson, Hugh H., Socratic Wisdom, Oxford University Press, Oxford, 2000. Brickhouse, Thomas C., and Smith, Nicholas D., Plato’s Socrates, Oxford University Press, Oxford, 1994. The Philosophy of Socrates, Westview Press, Boulder, Colo., 2000. Irwin, Terence, Plato’s Ethics, Oxford University Press, Oxford, 1995. McPherran, Mark L., The Religion of Socrates, Pennsylvania State University Press, University Park, 1996. Santas, Gerasimos Xenophon, Socrates, Routledge and Kegan Paul, London, 1979. Vlastos, Gregory, Socrates, Ironist and Moral Philosopher, Cambridge University Press, Cambridge, 1991. Socratic Studies, ed. Myles Burnyeat, Cambridge University Press, Cambridge, 1994. Works on the Trial of Socrates Brickhouse, Thomas C., and Smith, Nicholas D., Socrates on Trial, Clarendon Press, Oxford, 1989. This contains a useful guide to modern literature on Socrates. eds., The Trial and Execution of Socrates: Sources and Controversies, Oxford University Press, Oxford, 2002. Hansen, Mogens Herman, The Trial of Sokrates: from the Athenian Point of View, The Royal Danish Academy of Sciences and Letters, Copenhagen, 1995. Stone, I. F., The Trial of Socrates, Jonathan Cape, London, 1988.

bibliography 315 Collections of Articles Ahbel-Rappe, Sara, and Kamtekar, Rachana, eds., A Companion to Socrates, Blackwell Publishing, Malden, Mass., 2006. Benson, Hugh H., ed., Essays on the Philosophy of Socrates, Oxford University Press, Oxford, 1992. Gower, Barry S., and Stokes, Michael C., eds., Socratic Questions, Routledge, London, 1992. Karasmanis, Vassilis, ed., Socrates 2400 Years Since his Death, Hellenic Ministry of Culture and European Cultural Centre of Delphi, Athens, 2004. (Selected articles from this volume are reprinted in Lindsay Judson and Vassilis Karasmanis, eds., Remembering Socrates: Philosophical Essays, Clarendon Press, Oxford, 2006.) Prior, William T., ed., Socrates, 4 vols., Routledge, London, 1996. Smith, Nicholas D., and Woodruff, Paul B., eds., Reason and Religion in Socratic Philosophy, Oxford University Press, Oxford 2000. Vlastos, Gregory, ed., The Philosophy of Socrates, Doubleday and Co., Garden City, NY, 1971. Later Influence Fitzpatrick, P. J., ‘The Legacy of Socrates’, in Gower and Stokes, eds., Socratic Questions (see above), 153–208. Long, A. A., ‘Socrates in Hellenistic Philosophy’, Classical Quarterly ns 38 (1988), 150–71. Montuori, Mario, Socrates: Physiology of a Myth, trans. J. M. P. Langdale and M. Langdale, Gieben, Amsterdam, 1981. Original Italian edition G. C. Sanzoni, Florence, 1974. Nehamas, Alexander, The Art of Living: Socratic Reflections from Plato to Foucault, University of California Press, Berkeley and Los Angeles, 1998.

18 Democritus and Lucretius on Death and Dying In this paper I shall compare the views of death and of the process of dying attributed to Democritus by various ancient sources with those found in Lucretius, especially in book III of De rerum natura. Though this topic has recently been the subject of an excellent study by James Warren,¹ his sagacious treatment still seems to me to leave some issues open for discussion. In particular, I shall suggest that the differences between our two authors on death and dying arise naturally from their differing conceptions of the nature of the soul. Democritus conceives of the soul as a simple psycho-physical structure, a web of spherical and hence highly mobile atoms permeating the entire structure of the body. The most explicit evidence is that of Lucretius himself, who reports (in order to reject) Democritus’ view that there are exactly as many soul-atoms as body-atoms, arranged alternately (III. 370–4)² (i.e. presumably that between any two body-atoms there is a soul-atom and vice versa). Democritus appears to have ascribed the principal functions of the soul, viz. thought, perception or sensation (aisth¯esis) and the initiation of motion to this structure as a whole (Ar. De an. 404a27–31, 405a8–13 (DK 68 A 101)). The weight of the evidence points to his having made no distinction of parts of the soul, and specifically to his not having distinguished the mind or rational (part of the) soul from the non-rational (part or parts of the) soul. There is indeed a conflict of evidence on this point, since two passages of ps.-Plutarch’s Epitome (Aet. IV. 4. 6, IV. 5. 1 (DK ¹ Warren [2002]. For a fuller treatment of Epicurean attitudes to death see Warren [2004]. ² Subsequent references to Lucretius are to book III unless otherwise specified.

death and dying 317 68 A 105)) report that he distinguished the intellect from the non-rational soul and located it in a particular part of the body. But the two passages are mutually inconsistent, the former locating the intellect in the chest (i.e. the Epicurean view, see below) the latter (followed by Theod. Cure V. 22) in the head or more specifically the brain (the Platonic view, Tim. 70a–b). It is clear that whoever compiled the epitome was not working from an original text, but merely assumed that Democritus’ view must have been identical with one or other of the currently most influential views. On the other side we have the explicit testimony of Philoponus (In De an. 35.12 (DK 68 A 105)) that Democritus made no distinction between the intellect and the rest of the soul, of Sextus (supported by Lucretius, loc. cit.) that he held that thought occurs throughout the whole body (M VII. 349 (DK 68 A 107)), and of Aristotle (locc. citt.) that he identified mind and soul. The weight of the evidence is clearly in favour of the latter view. The whole animal is then a psycho-physical mechanism, in which the psychically active element, the soul, permeates the whole, in a way similar to that in which an electric charge permeates the charged body. There is no indication of a central locus or nucleus of psychic energy controlling or integrating psychic activity, in such a way that peripheral activity depends on central functioning. Soul-atoms (or more strictly clusters of soul-atoms) are adapted to cause bodily motion and to respond to external activity (i.e. to perceive or to think) wherever they are in the body, and the animal is psychically active, i.e. alive, so long as it contains enough soul-atoms to preserve psychic function throughout the whole body or substantial portions of it. Since death is the cessation of psychic functioning, consequent on the loss of a number of soul-atoms sufficient to sustain that functioning, this view of the soul naturally leads to a view of the distinction between life and death as a continuum, rather than as a sharp cut-off. If a person has lost so many soul-atoms that thought, sight, hearing, smell, taste, and bodily functions such as respiration and voluntary movement have all ceased, yet a few soul-atoms remain at the periphery, e.g. in the fingers and toes, there is no theoretical reason for Democritus to deny that sensation might occur in those parts. Sensation does not, as we have seen, require the transmission of information to any central sensorium. And we have doxographical evidence that Democritus actually held exactly that view.

318

death and dying

The most explicit evidence is from ps.-Plutarch (Aet. IV. 4. 7 (DK 68 A 117)): Democritus says that everything contains a sort of soul, including corpses, which is why it is always apparent that they have a sort of warmth and perception, though most of it is breathed out.

This is indeed the source whose evidence we judged unsatisfactory on the distinction of the intellect from the non-rational soul, and were it unsupported it would be rash to rely on it. But it is supported by Cicero, who reports (Tusc. I. 34. 82 (DK 68 A 160)) that Epicurus criticized Democritus for maintaining that sensation can persist after death, by Alexander (In Top. 21. 21 (DK 68 A 117)) and by Proclus (In Rep. II. 113. 6 (DK 68 B 1)) and Tertullian (De an. 51. 2 (DK 68 A 160)), who indicate that Democritus cited instances of apparently dead people coming back to life (such as Er in Plato, Rep. X), explaining these by the theory that a ‘spark of life’ (i.e. some soul-atoms) remains in the corpse, and (according to Tertullian) supporting this by appeal to the (supposed) phenomenon of the continued growth of hair and nails after death.³ As Warren points out,⁴ the theory that even minimal perception might persist after death was anathema to orthodox Epicureans, since their central doctrine that ‘Death is nothing to us’ rests on the premisses that death is the dissolution of the soul-body complex, that sensation ceases when the complex is dissolved, and that what is without sensation is of no concern to us (KD 2). One strategy for reconciling Epicurean orthodoxy with the phenomena adduced by Democritus is that of defining death as the total and irrevocable cessation of all vital functioning,⁵ from which it would follow that in cases such as that of Er where a ‘spark of life’ is assumed to remain in the body, the person is not actually dead. The passage of Cicero cited above suggests that the ‘followers of Democritus’ (‘Democritii’, whoever ³ Warren [2002] cites physiological evidence pointing out that this supposed phenomenon is a misinterpretation of the effect of the shrinking and desiccation of the skin after death (p. 198, n. 19). ⁴ [2002], 199–200. ⁵ Modern authorities unanimously define death as the irreversible cessation of vital functioning. For an example see the model code proposed by the US President’s Commission for Study of Ethical Problems in Medicine and Biomedical and Behavioural Research (Washington, DC, 1981). This contains the following statement of criteria of death: ‘An individual who has sustained either (1) irreversible cessation of circulatory and respiratory function or (2) irreversible cessation of all functions of the entire brain, including the brainstem, is dead. A determination of death must be made in accordance with accepted medical standards.’ (Cited in Zaner [1988], p. vii.)

death and dying

319

they were) may have made that move in response to Epicurus’ criticism; Democritus does not say (they may be supposed to have replied) that sensation persists after death, but rather that in some cases where death is supposed to have occurred it has not actually occurred. But that strategy threatens the consolatory power of the mantra ‘Death is nothing to us’, which assumes a fairly clear criterion for determining when death has occurred. If, for all we know, the objects we commonly regard as corpses have some minimal sensation, then even the staunchest Epicurean has grounds for being concerned about what happens to them, thus reopening the way for the fears about the treatment of one’s corpse which Lucretius is at such pains to dispel (870–93).⁶ An alternative strategy would simply be to deny the occurrence of the phenomena which Democritus appealed to. That, however, looks suspiciously like dogmatism, especially in view of the fact that the Epicureans themselves had grounds to make use of similar phenomena (see the evidence from Lucretius cited below). A sounder strategy would be to adduce theoretical grounds for the impossibility of the occurrence of sensation in isolation from the activity of the integrated psycho-physical complex; if that strategy were successful the phenomena apparently pointing to the occurrence of sensation in that situation would have to be interpreted in some other way. I shall now suggest that in book III Lucretius provides evidence that the Epicureans in fact took that latter route. A central element in this strategy is the distinction of the mind, located in the chest, from the rest of the soul, scattered throughout the body. Though this distinction is not explicit in what we possess of Epicurus’ own texts, it is confirmed by a scholiast on Hdt. 67 (Usener 311, Long/Sedley 14i), by Aetius IV. 4. 6 (cited above), and most fully by Lucretius’ distinction of the animus, located in the chest, from the anima pervading the body. The animus is the mind (94–5), which Lucretius identifies with the seat of the emotions (141–2). He stresses the fact that together with the anima it forms a unified system in which the animus has a controlling function (94–5, 136–9); the rest of the soul ‘obeys and is moved according to the direction and impulse of the mind’ (numen mentis momenque, 143–4). The example given in this passage is the effect of emotion on the body as a whole, e.g. sweating, pallor, and fainting brought on by fear; the fearful stimulus is registered ⁶ See Warren [2004], 20–3.

320

death and dying

in the animus and its effects are transmitted throughout the organism by the anima (153–60). Again, the animus is the source of bodily movement, its directions being transmitted to the limbs via the anima (161–4). Here the anima is functioning as the nervous system, transmitting information and impulses from the central animus to the periphery of the embodied organism. The crucial question for our enquiry is whether sensation too requires the activity of the unified animus-anima system, or whether it is a function of the anima alone. Here Lucretius’ account of the micro-structure of the soul is relevant. Again in opposition to Democritus, who held that all the soul-atoms were of the same kind, viz. very small and spherical, and hence fiery (Ar. De an. 403b31–404a9 (DK 67 A 28)), Lucretius constructs the soul of atoms of different kinds, atoms of heat, air, wind (231–6) and a fourth type, distinguished from the others by being smaller and yet more mobile (241–4), which is said to be the source, not only of motion for the other soul-atoms (247–8), but of sensation throughout the body (245, 272). This type of atom lies deep within the body (273–4), and is described as ‘the soul of the soul’ (anima animae, 275, 280–1) which ‘controls the whole body’ (dominatur corpore toto, 281). It is, then, the vital principle of the soul itself, the source of activity, including sensation, throughout the whole psychic structure. It does not appear to be identical with the animus, which also includes component atoms of the other three types (288–93). Rather it seems to be seen as the active principle of the mind itself, as it were the motor of all psychic activity, including thought and sensation. It appears from this that while the activity of the anima is necessary for sensation (e.g. feeling a pain in one’s foot requires the stimulation of anima-atoms situated in the foot) it is not sufficient by itself, since the activity of the anima itself presupposes the activity of this vital central element of the mind and thereby of the whole psychic system. The mind, or more strictly this elusive vital element within the mind, may be seen as a centre of consciousness, unifying and giving mental reality to the activity of the anima throughout the body. Lucretius emphasizes the centrality of the mind at 396–400: for without the mind (mens) and the animus no part of the soul (anima) can remain in the limbs for a tiny moment of time, but it accompanies it readily and scatters into the air.

death and dying

321

This passage picks up another key theme, that of the instability of the soul. Since the soul is composed as a whole of highly mobile atoms and since the atoms of the vital centre, the ‘soul of the soul’, are the most mobile of all, that element of the soul is conceived as something like a highly volatile gas, instantly dissipated as soon as its bodily container is shattered (425–44). Given that sensation requires the central organizing activity of the animus, and that the latter is dissipated instantly as soon is its bodily container suffers sufficiently significant damage, the Epicureans then have good ground for denying that sensation can continue after such damage, and hence that it can continue after death. But at the same time they emphasize certain phenomena which point to activity of the anima independent of the animus. In principle, such phenomena ought to be able to occur after the dissipation of the animus into the air. And that possibility might appear once again to open the way for the Democritean suggestion that sensation might occur in the ‘inanimate’ (i.e. animus-less) corpse. One of Lucretius’ main premisses in his arguments for the mortality of the soul is the thesis that the soul is not a unitary substance, but a whole composed of parts. Since anything composed of parts is liable to destruction through the progressive loss of parts, the soul, being such a composite, is destructible in that way. Hence his emphasis on cases where psychic activity ceases in parts of the body while continuing in others, e.g. when paralysis creeps up (as in Plato’s description of the death of Socrates in the Phaedo) from the extremities to the centre (526–32). Lucretius’ explanation is that the anima-atoms in the limbs are dispersed, while those in the rest of the body remain active. Another phenomenon supporting the particulate theory of the anima, which is more immediately relevant to our topic, is the continued activity of severed bodily parts. Lucretius gives gruesome descriptions of the twitching of severed limbs, of a severed head preserving its living expression and open eyes ‘until it has yielded up all the traces of the soul’ and of the discrete parts of a chopped-up snake writhing and attacking each other (634–63). Here again, the point is to demonstrate that the ‘force of the soul’ (vis animai) is physically dispersed by the process of physical dissection, and hence that the soul must be mortal (638–41). Here the anima continues to perform at least one vital function, the causation of movement, for some time after the bodily parts concerned are physically

322

death and dying

cut off from contact with the organizing and directing animus.⁷ Should we not then expect that the other main activity of the anima, sensation, would similarly continue for some time in separation from the animus? Yet this graphic passage mentions only the motion of the severed limbs and the lifelike expression of the severed head. The only mention of sensation is the statement that, in the heat of battle, the man who loses a limb does not feel any pain, because the blow is so sudden and his mind is concentrating on the fight (645–7). Since the whole living person does not feel pain, it seems plausible that Lucretius’ view is that the severed limb or head does not itself feel pain either, and this is confirmed by his explicit statement at II. 910–14 that sensation is a function of the whole organism, which cannot occur in a separated part. The front part of the severed snake is indeed described as ‘struck by the burning pain of the wound’, but that presumably is the part which houses whatever centre of consciousness corresponds in the snake to the human animus. Lucretius thus differentiates sensation from the initiation of motion, in that whereas the latter function of the anima can persist after the severing of communication with the animus, the former cannot. It remains to investigate the grounds for that distinction. As another proof of the particulateness, and therefore of the mortality of the soul, Lucretius cites the generation of maggots in decomposing corpses as evidence that ‘seeds of the soul’ (semina animai) remain scattered throughout the body (713–21). Since, like ancient theorists generally, Lucretius was ignorant of the true explanation of this phenomenon, his account is that some of the atoms of the dead body are recycled to form the bodies of the maggots. But since the latter are alive they must contain soul-atoms too, which must also have been recycled from those composing the corpse. Hence the corpse must contain soul-atoms as well as body-atoms. Now this is precisely the foundation of the Democritean view that the corpse retains some degree of sensation, and the phrase ‘seeds of the soul’ is reminiscent of the ‘spark of life’ (empureuma t¯es z¯oe¯s) in the Proclus passage cited above (DK 68 B 1). Yet the treatment of this residue of the soul by Lucretius is altogether different from that by Democritus. For the latter the residue is the physical basis of the continuation of some ⁷ Note the contrast with 398–9: ‘nam sine mente animoque nequit residere per artus | temporis exiguam partem pars ulla animai’.

death and dying

323

vital functioning of the ‘dead’ person, and it provides the potential in at least some cases, such as that of Plato’s Er, for the complete recovery of function. For Lucretius the vital functioning of the person ceases with death; hence the subsequent activity of residual soul-atoms scattered throughout the body cannot continue the vital functioning of the person, in however attenuated a form. Rather it constitutes the vital functioning of those different creatures in whose bodies the soul-atoms have become incorporated, viz. the maggots. Once again, we have to ask what explains the difference between the two accounts. What I take to be central to the Epicurean position (represented here by Lucretius) is the insight that what we may call the higher psychic functions, thought, perception/sensation, and the initiation and control of bodily motion, require the integration of incoming information and outgoing direction into a single, centrally controlled system. This idea is not original to Epicureanism; we find the central idea in Plato’s Theaetetus, in the denial that the separate senses are housed in the body like the warriors hidden inside the Trojan horse (184d), and it is foreshadowed in Empedocles’ theory of the blood round the heart as the organ of thought (DK 31 B 105). It seems, however, to be absent from the atomism of Democritus. On the question I have discussed in this paper, whether sensation can persist after the cessation of other aspects of life, it is the application of this picture to consciousness which is crucial. In the doctrine that it is the central animus, and more especially the elusive ‘fourth nature’ of the animus which accounts for the ‘sensiferos motus’ of the soul I suggest that we find an anticipation of Descartes’s doctrine that, unlike the body, the soul, whose only essential attribute is consciousness, is necessarily unitary (Meditations 6, Adam-Tannery VII. 85–6). Descartes’s point is that we are necessarily aware of all the different exercises of mental functioning, thought, will, perception etc. as functions of one and the same self. Though various abnormal phenomena, such as schizophrenia and split-brain phenomena,⁸ may cast doubt on the necessity of Descartes’s thesis, it is clearly correct as an account of normal functioning, including that of consciousness. Though one feels bodily sensations in different parts of the body, not in the brain, it is the integrated person who feels them. I can indeed conceive of feeling ⁸ See Nagel [1971], repr. in Glover [1976], 111–26 and in Nagel [1979], 147–64; Glover [1988], chs. 1–3.

324

death and dying

pain in a severed hand, but were that to happen it would be I, this embodied person here, who felt the pain in the hand lying over there. Can we imagine the hand itself feeling pain? But then whose pain would the hand feel? Suppose it is my hand that has been cut off, but I feel no pain in the hand (though perhaps I feel it in the arm from which the hand has been severed).⁹ Then, if the hand itself feels pain, it is not my pain; nor, presumably, is it anyone else’s pain. Then it would seem that the pain that the hand feels is simply the hand’s pain, no one else’s. But now the hand is being treated as a centre of consciousness in its own right, as if, by being cut off, it has somehow become itself a sort of primitive person, or at least an animal. It is surely clear that that is pure superstition; since an attached hand is not itself an animal, but part of an animal, how could severing it confer on it the extra capacities necessary for being an animal? While I am not suggesting that the Epicurean position somehow relies on this argument, I think it not unreasonable to claim for the Epicureans the fundamental insight which underlies it. We can now see why the Epicureans might reasonably allow motion to persist in separated bodily parts, while denying that consciousness persists in them. While voluntary motion, like consciousness, is a function of the integrated animal, the twitching of a severed limb is not an instance of voluntary motion. Rather it is a purely mechanical effect of prior stimuli, persisting temporarily after the connection with the source of those stimuli has been cut, rather as a residual electric charge may persist briefly in a charged body (e.g. a radio receiver) after the current has been disconnected. So when the limb twitches the question ‘Whose twitching is occurring?’ does not arise; the twitching is not anyone’s twitching, but merely the twitching of the limb. That is exactly the situation we saw to be impossible in the case of pain; since any pain must be felt by some sentient being, there can be no pain which is no one’s pain, but merely pain in a hand or leg. To repeat, the supposition that there is pain in a hand or leg detached from a functioning centre of consciousness is the supposition that the hand or leg has itself become just such a centre of consciousness in its own right, which is impossible. The Epicurean account of the soul thus represents a major advance over that of Democritus, in its recognition that consciousness is the activity of ⁹ Or perhaps I feel pain in the ‘phantom hand’, i.e. it feels to me (a) that my hand is still attached to my body, (b) that I have pain in that hand. That is simply another way in which I feel pain; it is not a case of the hand’s feeling pain.

death and dying

325

an integrated psychical system, involving the integration of diverse sensory input and control of voluntary movement, which it is plausible to locate in some functionally central bodily part. Hence they were better able than Democritus to recognize that continued post-mortem activity of soul-atoms scattered throughout the body, which they and Democritus alike accepted, is not of itself a sufficient ground to postulate the occurrence of post-mortem sensation. In insisting on the central controlling function of the animus, and in particular on the necessity of that function for the functioning of the anima as a sensory system, they may reasonably be seen (allowing for their ignorance of anatomy and neurophysiology) as foreshadowing the modern scientific picture in which those functions are fulfilled by the brain. And in taking the dispersal of the animus as constituting death they can therefore be seen as foreshadowing the dominant modern conception of death as brain death, the irreversible cessation of the functioning of the brain. On that conception, whatever other vital functions, such as breathing or circulation, continue (normally with artificial help), a person whose brain has ceased permanently to function (e.g. because of brain damage) is dead, and conversely someone whose brain continues to function in the absence of other vital functions is alive.¹⁰ On these crucial points the Epicurean picture is in essential agreement with the modern. Where it differs (apart from the obvious anatomical disagreement on the location of the vital centre) is that the Epicureans assume that the dispersal of the animus is (virtually) instantaneous and irreversible, whereas modern science recognizes that brain function may itself cease gradually rather than instantaneously, and that in certain circumstances, such as drug intoxication and hypothermia, where the brain itself has not undergone major trauma, even total cessation of brain function is not irreversible.¹¹ Modern science ¹⁰ The lively contemporary controversy on whether the crucial concept is the functioning of the whole brain, the ‘higher brain’ (including the cerebrum and cortex), or the brainstem is not relevant to the present discussion. See Zaner [1988]; Lamb [1996], especially ch. 5. ¹¹ See e.g. Lamb [1996], 54: ‘Certain drugs and low body temperature can place the neurons in ‘‘suspended animation’’. Under these conditions they may survive deprivation of oxygen or glucose for some time without sustaining irreversible damage.’ Lamb cites a report from The Lancet of 6 March 1976 (p. 535), describing a drug-intoxicated patient who, 12 hours after losing consciousness, showed no spontaneous respiration and a total absence of all sensory responses and brainstem reflexes, but who had almost totally recovered within 72 hours. For these reasons the standard guidelines for the identification of death in the UK and USA require the absence of drug intoxication, hypothermia, and other factors which ‘mimic’ clinical death. See Plum [1999], 34–65.

326

death and dying

has therefore largely abandoned the idea of a moment of death,¹² which is central to the Epicurean picture, underlying its confidence that ‘death is nothing to us’. And, ironically, though Democritus’ failure to identify a controlling centre of psychic function makes his picture of the soul and its activity in general less satisfactory than the Epicurean, it made it easier for him to recognize a possibility which the theoretically more advanced Epicureans denied, but which modern science has shown actually to occur, viz. the return to life of someone who satisfies all the criteria, save that of irreversibility, of clinical death.¹³ ¹² See e.g. Brody [1999], 79, ‘the death of the organism is a process rather than an event’; Plum [1999], 53, ‘Natural death is not an event. It is a process in which different organs or parts of organs permanently lose their life-supporting properties at widely varying times and rates.’ For a defence of the traditional concept of a moment of death see Lamb [1996], ch. 7. ¹³ I am indebted to my colleague Dr D. S. Fairweather for advice on medical questions relating to this paper.

19 Socrates under the Severans1 In this paper I examine the representation of Socrates by pagan and Christian writers of the latter half of the second and first half of the third centuries ad. The pagans to be considered are Athenaeus and Aelian, and the Christians Clement, Origen, Tertullian, and Minucius Felix. The contrast between Christian and pagan approaches is striking. Unsurprisingly, the Christian views of Socrates are doctrinally determined. On the other hand, so far from expressing doctrinal uniformity, they manifest a fundamental dichotomy between radically opposed attitudes to the pagan world, a dichotomy which focuses on the figure of Socrates and on one particular aspect of that figure, his daimonion or divine sign. This concentration on the daimonion is absent from the pagan writers, as indeed is any specific ideologically based focus. They write not as dogmatists, even in the sense of adherents of any particular philosophical school, but as anthologists and encyclopedists, for whom the primary interest of Socrates is less as a philosopher, whether dogmatic or critical, than as a source of entertaining and morally improving anecdote. The opposed views of Socrates expressed by the Christian writers whom we shall be considering reflect a fundamental tension running through Christian attitudes to the pagan world, and in particular to its intellectual and cultural achievements. Granted that the Christian revelation had brought the promise of salvation which was absent from paganism, was the thought of the pagan world to be seen in a way analogous to the Jewish Law, as some sort of anticipation of the revelation which had superseded it, or was the whole of that world, including the thought of its great figures, to ¹ I am delighted to offer this paper to a collection in honour of Ewen Bowie, valued as a friend and colleague for over forty years. The paper was presented to the Arizona Colloquium in Ancient Philosophy in Tucson in 2005. I am indebted to writings: Edwards [2007] and Frede [2004]. I am also most grateful to Michael Trapp advice.

whom I have Tenth Annual the following for his helpful

328

socrates under the severans

be seen as belonging to the realm of darkness which Christ had overthrown? Jerome’s famous and terrifying vision of judgement, in which he is punished for his love of pagan literature with the condemnation ‘Ciceronianus es, non Christianus’ (Ep. XXII. 30), may be taken as representative of the latter, darker view (though his writings as a whole show an appreciation of pagan literature suggestive of the other side of the dichotomy). But from the earliest times the more positive view had its adherents. One of the earliest Christian apologists, Justin (second century ad), represented Socrates as a partial forerunner of Christianity in that he rejected traditional Greek religion in favour of the worship of the one true God, and justified this appropriation of Socrates by the slogan that ‘whatever has been well said by them [i.e. the pagans] belongs to us Christians’ (II Apol. 13).² The treatment of the history of Greek philosophy by Christian writers shows both sides of the dichotomy. While the main aim of Hippolytus’ Refutation of All Heresies (third century ad) is to show how the errors of Christian heretics derive from pagan Greek philosophy, Clement’s treatment of pagan philosophy in the Protrepticus and Stromateis is more eirenic; thus Plato, Pythagoras, and Socrates are described as following Moses (who says in Genesis 1: 3 ‘He spoke, and it came to be’) in saying that they hear the voice of God when they look at the divinely sustained world order (Strom. V. 14. 99. 3). In what follows I shall examine the specific application of this dichotomy to the presentation of Socrates and his daimonion s¯emeion. For Clement, Socrates is a central figure in Greek thought, in that the different philosophical sects took their principal doctrines from him (Strom. VI. 2. 5. 1).³ He is therefore a conduit for the transmission of the knowledge of the true God to the pagans. Clement counts him among those who had the correct conception of God, citing Xenophon Mem. IV. 3. 13 ff. where he tells Euthydemus that though God is invisible he is revealed through his providential ordering of nature, and quoting Antisthenes (fr. 40B Decleva Caizzi (= fr. 24 Mullach)) to the effect that no one can learn the nature of God from any image, because God is not like anything else, which, according to Clement, Antisthenes held because he was an associate of Socrates (Protr. VI. 71. 1–3). Clement cites various Platonic texts to show ² For fuller discussion of Justin’s treatment of Socrates see Edwards and Frede opp. citt. ³ This was the view of the schools themselves, each of which, apart from the Epicureans, claimed to derive their doctrines from Socrates. The pioneering study of this topic is Long [1988]. Subsequent work includes several of the contributions to Vander Waerdt [1994].

socrates under the severans 329 Socrates as an exemplar of the proper worship of God. The essence of that worship is willing separation from the body and its passions, which Clement finds perfectly exemplified in Socrates’ account of the philosophic life in the Phaedo: The acceptable sacrifice to God is willing separation from the body and its passions. This is the true worship of God, which is why philosophy was reasonably said by Socrates to be a preparation for death. For it is the person who does not make any use of sight as evidence in his thinking nor drag in any of the other senses, but approaches things by pure intelligence alone who undertakes true philosophy. (Strom. V. 11. 67. 1)⁴

Socrates’ concluding prayer at Phaedr. 279b ‘O Pan and you other gods, grant that I may be good within, and such external things as I have, may they be in harmony with what is within’ is cited by a somewhat flexible interpretation as fulfilling Christ’s two great commandments ‘Love God’ and ‘Love thy neighbour’ (Strom. V. 14. 97. 1), and in the same passage Clement seeks to find the same message in Tht. 185e ho gar kal¯os leg¯on kalos te kai agathos and Prot. 309c–d to soph¯otaton kalliston estin. Another instance of creative interpretation in a Christian direction occurs at Strom. V. 2. 14. 1; discussing Crt. 48b Clement cites Socrates as maintaining that living and dying well is more important than simply living, and takes that as evidence of Socrates’ hope for the afterlife. In fact the passage does not mention dying; Socrates’ point is simply that living well is more important than simply living, and while that does imply that living well requires facing death rather than acting unjustly, that evaluation is independent of hope for the afterlife. Socrates’ distance from traditional religion is referred to obliquely in a passage (Protr. VII. 76. 3–6) where Euripides’ criticisms of traditional tales of divine immorality are praised as ‘worthy of Socratic discussion’ (diatrib¯e). Socrates anticipates Christianity, then, not merely in his conception of God but in his living the Christian life in advance of its formal institution. It is not, therefore, surprising to find that his divine sign is given a Christian interpretation as a guardian angel (Strom. I. 17. 83. 4; I. 21. 153. 5; V. 14. 91. 4; VI. 6. 53. 2–3). While the conception of a tutelary spirit fits the role of the Platonic daimonion s¯emeion in warning Socrates against ⁴ The final sentence is a close paraphrase of Phd. 65e–66a, and the words in italics translate a direct quotation.

330

socrates under the severans

doing things which would be bad for him, and thereby protecting him from harm (Apol. 31d, Theages 128d), it misses a central aspect of the Platonic account, viz. that Socrates’ sign was something unique to him, which thereby picked him out as an extraordinary, and indeed somewhat uncanny, individual. By contrast, in the Christian conception every soul is assigned a guardian angel whose function is to preserve it against temptation (for a partial Platonic anticipation see Rep. 621d–e). In that respect, if in no other, hostile accounts of the sign (see below) are nearer the spirit of Plato’s own. Clement’s pupil Origen discusses Socrates in his defence of Christianity against the attack by the pagan philosopher Celsus. A consistent theme in his defence is that various accusations which Celsus brings against Christ and his followers could equally be alleged against Socrates, whom everyone, including Celsus, accepts as a paradigm of goodness and wisdom (Cels. III. 66, IV. 97). Thus Celsus alleges (Cels. II. 41) that Jesus had not been able to show that he was invulnerable to evils, meaning that he had not been able to escape the ignominious death of a criminal. Origen replies: If he considers as evils poverty and a cross and the conspiracy of wicked men, obviously he would say that evil also befell Socrates, who would not have been able to prove that he was pure from all evils.

But in any case Origen is able to appeal to the example of Socrates himself to show that these alleged evils are not evils at all. In another context (Cels. III. 8) he objects to Celsus’ counting external harms such as death and injury as evils, saying that he ‘has failed to pay attention to the fine saying of Socrates ‘‘Anytus and Meletus can kill me, but not harm me; for it is not allowable for that which is better to be harmed by that which is worse’’ ’.⁵ In reply to Celsus’ assertion that the miracles attributed to Jesus are incredible, Origen replies (Cels. VI. 8) that similar scepticism would cast doubt on aspects of Socrates which show him as enjoying divine favour, citing his widely reported prophetic dream of Plato as a swan,⁶ and his ⁵ The ‘fine saying’ paraphrases Plato, Ap. 30c6–d1. ⁶ Olympiodorus Vita Plat. 4: cf. Apuleius De Plat. 1. 1, DL III. 5, Pausanias I. 30. 3, Suidas s.v. Plato, Tertullian De an. 46. 9. In the standard version Socrates dreams that a cygnet sits in his lap, then grows its feathers and flies away with a beautiful song. Meeting Plato the next day, he recognizes him as the swan. Athenaeus XI. 507c–d gives an anti-Platonic version, in which he figures not as a swan but as a crow which pecks Socrates’ bald head, prompting Socrates to say that he thinks that Plato will tell many lies against him.

socrates under the severans 331 divine sign. Again, the implication is at least that Celsus would not wish to discredit such stories, although it is not wholly clear whether Origen himself accepts them. For him, Socrates is not free of the taint of idolatry; he was praised by the oracle not for the truth of his opinions, but probably because of ‘the sacrifices and burnt-offerings which he brought to him and to the other demons’ (Cels. VII. 6). But it does seem that in Origen’s view Socrates shares some of the aura of the Christian martyrs. In response to Celsus’ objections to honours paid to the tombs of martyrs he responds (Cels. IV. 59) ‘After the death of Anytus and Socrates would a wise man pay similar attention to the tomb of Anytus’ body and that of Socrates, and would he build a similar tomb or sepulchre for both?’

For Clement, then, Socrates is an opponent of idolatry and a precursor of the true religion, whose divine sign is a genuine mark of divine approval in the form of protection by a good angel. Origen’s portrayal is more ambiguous; while Socrates can be held up to the pagans as a paradigm of goodness he is not free of the taint of idolatry.⁷ When we come to Tertullian, however, Socrates belongs firmly to the unregenerate pagan world which is totally subject to the powers of evil (De an. 1), and hence his alleged goodness cannot approach true Christian virtue. For Tertullian the pagan gods are in fact evil demons, and so far from rejecting them Socrates manifested his adherence to them, in his appeal to the Delphic oracle’s endorsement of his philosophical mission and in his dying injunction to Crito to fulfil a vow to sacrifice a cock to Asclepius (ibid.). The ground of the belief that Socrates rejected the gods, which led to his condemnation, was simply that instead of swearing by the traditional gods he preferred to swear ‘by the dog’ etc. (Apol. XIV. 7). On this point Tertullian is inconsistent; he elsewhere cites the oracle’s endorsement of Socrates as the wisest of men as in effect an abdication of divinity, since Socrates denied the existence of the gods (Ad nat. I. 4. 7). Specifically, since pagan divinities as a whole are evil demons, the same is true of Socrates’ daimonion. Since the Christian revelation had not yet shattered the demonic powers, there was no source of truth available to guide Socrates, who was instead subject to the direction of a ‘pessimum ⁷ For fuller discussion see Edwards [2007].

332

socrates under the severans

revera paedagogum (De an. 1) ... dehortatorium plane a bono’ (Apol. XXII. 1).⁸ Hence it is no surprise that Socrates was far from being the paragon of virtue which pagan tradition alleges. His untroubled acceptance of condemnation and death was motivated, not by hope of blessed immortality, as true Christian constancy is, but by the malicious desire to discomfit his accusers (De an. 1), and in any case his easy death by hemlock did not stand comparison with the torments which the Christian martyrs had to endure (ibid.). When it comes to sexual matters, Tertullian lets his rhetoric, and perhaps his imagination, run riot. Like other pagans, including Cato the Censor, Socrates cared so little for the chastity of women, including his wife, that he was willing to advocate community of wives: ‘O sapientiae Atticae, o Romanae gravitatis exemplum: leno est philosophus et censor!’ (Apol. XXXIX. 12.)⁹ As for corrupting the young men, Tertullian not only accepts that the charge was well founded, but interprets it as referring to sexual seduction, and takes the opportunity to compare Socrates’ laxity in that respect with Christian chastity (Apol. XLVI. 10). All of these themes recur in the Octavius of Minucius Felix. Socrates’ daimonion was an evil spirit (26. 9, 38. 5); Socrates is listed among pagan philosophers ‘quos corruptores et adulteros novimus et tyrannos’ (38. 5). In the latter passage the author repeats the Epicurean Zeno’s insulting description of Socrates as an ‘Attic buffoon’ (scurra Atticus),¹⁰ on the ground that he confessed that he knew nothing while ‘glorying in the testimony of a totally deceptive demon’ (‘nihil se scire confessus testimonio licet fallacissimi daemonis gloriosus’; the deceptive demon is the Delphic oracle). But together with these passages in which the influence of Tertullian’s hostility to Socrates is apparent, there are traces in the text of the favourable picture of Socrates as a monotheist and therefore a precursor of Christianity. At 19. 13 Socrates is included in a list of philosophical monotheists; Minucius cites the same passage of Xenophon’s Memorabilia as Clement (IV. 3. 13 ff.; v. supr.) in attributing to him the claim that since God is imperceptible the question of his form should not be investigated. These philosophers ⁸ This became the regular view of Christian apologists writing in Latin. In addition to Munucius Felix (cited below), cf. Lactantius Div. inst. II. 14. 9, Chalcidius Comm. 168, Rufinus Hist. eccles. II. 14. 5. Presumably the authority of Tertullian accounts for the wide spread of this view among Western writers. ⁹ Perhaps Tertullian assumed that the community of women advocated for the guardians in Plato’s Republic reflected not merely the theoretical opinion but also the practice of the historical Socrates. ¹⁰ Recorded by Cicero ND I. 34. 93

socrates under the severans 333 had anticipated the Christian view of God ‘ut quivis arbitretur aut nunc Christianos philosophos esse aut philosophos fuisse iam tunc Christianos’. Christian responses to Socrates in the writers we have considered focus primarily on Socrates’ relation to the divine (of which his daimonion s¯emeion is a central element) and tend towards opposite poles, one stressing the role of Socrates as a forerunner of Christian monotheism (involving the construal of the divine sign as a guardian angel), the other stressing his commitment to idolatrous polytheism, in which the sign is seen as the manifestation of a demonic familiar. Though different writers incline to one pole or the other, we see from the case of Minucius that both views of Socrates may be found at different points in the work of a single writer (indeed, in that case, in the course of a single comparatively short work).¹¹ While that may be explained partly by lack of concern for strict consistency on the part of the writer we should also acknowledge that the evidence from the primary sources on Socrates’ religious attitudes is itself ambivalent to the extent of affording scope for widely divergent interpretations, even on the part of a single writer. On the one hand he accepts the authority of the Delphic oracle and prophetic dreams, and is concerned with matters of ritual, most famously in his last words enjoining the sacrifice to Asclepius; in Xenophon’s portrayal his conventional piety is a central theme. On the other hand Plato’s account of his service to the god in his Apology says nothing whatever about ritual. That service consists wholly in the fulfilment of his divine mission to lead his fellow-citizens to care for their souls by practising virtue, which fits well with a reading of the Euthyphro which takes that dialogue to conclude with the suggestion that piety is nothing other than virtue as a whole.¹² Those aspects of Plato’s portrayal tend towards a view of Socrates as the proponent of a moralistic alternative to conventional civic religion,¹³ which can readily be incorporated into a monotheistic theology. Of course Christian writers interpret Socrates in the light of their dominant theological preconceptions (it would be very surprising if they did not). We have simply to acknowledge that the primary ¹¹ Even Tertullian is not wholly consistent in his hostility to Socrates, who figures at Apol. XI. 15 in a list of outstanding individuals, and is specifically said to excel in wisdom. ¹² I argue for that reading in chapter 4 of this volume. For other views see McPherran [1999], ch. 2 and the exchange between McPherran and Calef in OSAP [1995] (see bibliography). ¹³ See Burnyeat [1997].

334

socrates under the severans

evidence is such as to make possible the application of widely divergent preconceptions. In the pagan writers to be considered the focus on Socrates shifts from the doctrinal to the biographical and anecdotal. In the tradition of encyclopedic miscellany (containing a strong element of moralizing biography) to which their work belongs Socrates figures less as a source of philosophical doctrine or as a paradigm of philosophical method than as an exemplar of significant traits of character and a fount of practical wisdom, frequently expressed in pointed epigrams. Within that broad framework there is room for anecdotes illustrating different viewpoints on Socrates, including some less adulatory than in the tradition deriving from Plato and Xenophon; Athenaeus’ format of table-talk ascribed to a variety of speakers lends itself readily to that diversity of viewpoint. A number of stories, each ending in a pithy punchline, concern the simplicity of Socrates’ mode of life, particularly with respect to food and drink. He often used to be seen walking up and down in front of his house in the evening, and when asked what he was doing used to say ‘I’m collecting a relish (opson) for dinner’ (Deipn. IV. 157e). He used to say that he differed from others in that they lived in order to eat, whereas he ate in order to live (IV. 158 f.). Seeing someone helping himself greedily to the relish at dinner he asked ‘Fellow-guests, which of you is it who treats the bread as relish and the relish as bread?’ (V. 186d) A textually incomplete anecdote records Socrates’ saying about a type of cup called a bombylios, which apparently released the liquid drop by drop, that ‘Those who drink as much as they want from a cup quit soonest, while those who drink from a bombylios which drips little by little ...’ (XI. 784d). Perhaps the point was that the person who drinks only a little at a time carries on longer and ends up drinking more than the person who takes his fill straight away (i.e. too much), or perhaps (as suggested by Xen. Symp. II. 26, cited by Athenaeus at XI. 504d–e) the sense is the opposite one, that quick drinking encourages excess while drinking small quantities at a time promotes (decorous) jocularity. Either way the message favours moderation. A passage which tempers this picture of Socratic austerity is I. 20 f., where Socrates is described (following Xen. Symp. II. 16–20) as enjoying dancing and thinking it good exercise A stock theme of the favourable biographical tradition and a source of memorable Socratic ‘one-liners’ is his equanimity in putting up with

socrates under the severans 335 the ill-temper of his wife Xanthippe. Athenaeus gives us only a single instance (XIV. 644), also retailed by Aelian (VH XI. 12). Alcibiades had sent Socrates a cake, which Xanthippe trampled underfoot (presumably in a jealous rage), prompting Socrates to laugh and say ‘Well, you won’t be able to have a share of it either.’ Socrates’ relations with Alcibiades could, and certainly did, provide ammunition for hostile critics. We find an instance of this at V. 219b, in the context of an anti-philosophical diatribe by a speaker named Masurius, who makes Socrates one of his principal targets. He quotes verses purporting to be by Aspasia, encouraging Socrates in his pursuit of Alcibiades, and offering consolation when Socrates wept for lack of success. A little later (221e) he describes Socrates as ‘passing the time with Aspasia’s flutegirls in their workshops’ (ergast¯eri¯on: the term is sometimes used as a euphemism for ‘brothel’ (v. LSJ, s.v. 1), and clearly has that application here). These aspersions follow an extended passage (215e–216e) in which he denies the truth of the accounts, mostly from Plato, of Socrates’ courage on the battlefield. These accounts must be false, he says, because neither Thucydides nor any other historian or poet mentions them! Another anti-philosophical speaker, Myrtilus, picks up the Aspasia theme. In the course of extensive quotation of a catalogue of accounts of love affairs by the elegiac poet Hermesianax of Colophon (4th–3rd century bc) he describes Socrates’ passion for Aspasia (XIII. 599a–b): And with what fiery power did Cypris, in her wrath, heat Socrates, whom Apollo had declared to be supreme among men in wisdom! Yea, though his soul was deep, yet he laboured with lighter pains when he visited the house of Aspasia; nor could he find any remedy, though he had discovered the many cross-paths of logic. (tr. C. B. Gulick (Loeb))

The same speaker strikes a darker note with a hostile description of Socrates’ conduct at his trial (XIII. 611a): Socrates died because he used ‘inappropriate arguments’ and ‘discoursed to the jury with arguments of the most knavish sort on the theme of justice’ (... tous akairous elenchous di’ hous Sokrates men apethanen ho pros tous eis ta dikast¯eria diakl¯eroumenous dialegomenos peri tou dikaiou kleptistatous ontas). The next sentence adds that for the same reason the notorious atheists Theodorus and Diagoras were condemned respectively to death and exile; the men ... de construction associates them with Socrates in a single group.

336

socrates under the severans

He goes on to attack Socrates indirectly via his association with Aeschines of Sphettus, quoting a speech of Lysias in prosecution of Aeschines for defaulting on a debt. Lysias was well known as an associate of Socrates (Phaedr. 227a, Rep. 328b), and was reputed to have a written a speech for his defence at his trial which Socrates rejected as out of character (Cic. De or. I. 231, DL II. 40–1), but that does not prevent the speaker from giving a hostile slant to points in the speech which are either neutral or clearly intended to favour Socrates. The first point is that Aeschines had originally borrowed money to set up in the perfumery business (611e–612a). This is described as ‘a consequence of Socrates’ philosophy’, and the fact that Xenophon has Socrates disapprove of the use of perfume (Symp. II. 3, also cited by Athenaeus at 686d–e) is then taken to imply that Socrates’ disapproval was hypocritical. The speaker then quotes Lysias as saying that he would never have expected such dishonesty from a pupil (math¯et¯es) of Socrates, especially one who had said so many solemn things (pollous kai semnous legonta logous) about justice and virtue (612b). The implication of the use of the word ‘pupil’ is that Aeschines had learned these ‘many solemn things’ from Socrates, and Lysias is clearly seeking to stress the special depravity of someone who had turned to dishonesty despite such excellent instruction. But in the mouth of Athenaeus’ speaker the implication is once again that Socrates’ solemn instruction is so much hypocrisy. As in the more celebrated cases of Alcibiades, Critias, and Charmides, Socrates is indirectly answerable for the misdeeds of his associates. Athenaeus thus reflects both hagiographical and anti-Socratic traditions, though the context of the hostile passages is mostly that of diatribes attacking philosophy and philosophers in general rather than Socrates in particular. Aelian’s presentation, by contrast, is almost wholly adulatory. Socrates figures in a list of the ‘best of the Greeks’ who were very poor, along with heroes such as Phocion and Epaminondas and the paragon of justice, Aristides (VH II. 43); at IV. 16 Socrates is the paradigm wise man, as Aristides is the paradigm just man and Epaminondas the paradigm general. It is symbolic of his heroic status that his birthday, 6 Thargelion, was also the date of the victories of Artemisium and Plataea (II. 25).¹⁴ He figures in ¹⁴ Diogenes Laertius gives that as Socrates’ birthday, citing Apollodorus’ Chronicles and Demetrius of Phaleron (II. 44). Stephen A. White argues convincingly in White [2000] that that was in fact the day of his death, which took place thirty days after the sacred embassy set out for Delos on the sixth day of the preceding month, Mounichion.

socrates under the severans 337 a list of wise benefactors, beginning with Chiron, who instructed Achilles, and Nestor, who counselled Agamemnon; if the Athenians had listened to Socrates, they would have been completely happy and would have practised philosophy (XII. 25). An instance of his wise advice is given in a fragment preserved by Stobaeus (fr. 4, Stob. II. 31. 38), in which Socrates rebukes fathers who do not educate their sons properly and then prosecute them for not maintaining them in old age; they are demanding the impossible, he said, because it is impossible for people who have not learned what is right to do it.¹⁵ His exceptional temperance made him so fit that he alone in Athens was not affected by the plague (XIII. 27). He figures in a list of philosophers who were good soldiers (VII. 14). But his courage is not restricted to the battlefield; when some young men try to frighten him by dressing up as Furies and waylaying him after dark he is not in the least frightened, but starts asking them questions as if they were in the Lyceum or the Academy (IX. 29). He is also immune to ridicule; some people are driven to suicide by ridicule, but Socrates merely laughs (V. 8). An example of this indifference is the story that at a performance of the Clouds, when some foreigners asked who this Socrates was who was being lampooned he stood up and remained standing throughout the performance (II. 13). The story that he, and other great men, used sometimes to play with children (XII. 15) presumably indicates indifference to general opinion, as well as the childlike simplicity which is typically the mark of the great.¹⁶ There are the usual Xanthippe stories; besides the story of Alcibiades’ cake (see above), we have an instance where Xanthippe refused to go to see a procession dressed in a coat of Socrates’ (presumably they were so poor that she had nothing else to put on): ‘So you are not going to see, but to be seen’ comments Socrates (VII. 10). But one anecdote gives the relationship an unusual slant; instead of the usual nagging we find Xanthippe testifying to Socrates’ imperturbability (see above), saying that he always keeps the same expression whatever happens, and that he is always serene in mind, superior to any distress and fear (IX. 7). He is a good influence on Alcibiades: he refuses to accept his extravagant presents (IX. 29),¹⁷ tells him ironically ¹⁵ There is presumably an echo here of the theme, prominent in Protagoras (319e–320b, 324d–328a) and Meno (93a–94e), of the failure of prominent citizens to educate their children properly. ¹⁶ Cf. the story that Heraclitus, found playing dice with children, said that that was better than engaging in politics with adults (DL IX.3 (=DK 22 A 1 (3))). ¹⁷ Despite Xanthippe’s urging him to do so. Contrast her reaction to the cake; it is difficult for Xanthippe to win in these anecdotes.

338

socrates under the severans

that since he despises individual members of the d¯emos he should have the same attitude to the d¯emos as a whole (II. 1),¹⁸ and cuts him down to size by showing him that the estates of which he is so proud are too small to figure in a map of the world (III. 28). Others too suffer the edge of his tongue; when Antisthenes (a proto-Cynic) ostentatiously displays his torn cloak he says ‘Won’t you stop showing off?’ (IX. 35), and when Archelaus of Macedon has his house decorated at vast expense by the painter Zeuxis he says that people travel a long way to see the house, but that no one would go to see Archelaus unless they were paid to (XIV. 17).¹⁹ Some anecdotes retail matter familiar from Plato and Xenophon, his divine sign (VIII. 1), his refusal to act illegally in the cases of the generals after Arginusae and of Leon of Salamis (VIII. 10) and his indifference to the mode of his burial (fr. 3, Stob. IV. 55. 10). Others elaborate Platonic or Xenophontine material; Socrates’ insistence that the corpse left after his death will not be him (Phd. 115c–e) gets a further twist in the story that that was why he refused Apollodorus’ offer of elegant clothing to put on to drink the hemlock (I. 16), and Xenophon’s entertaining account of his conversation with the high-class prostitute Theodote (Mem. III. 11) is complemented by an exchange with another lady, Callisto (XIII. 32). In reply to her boast that while she can seduce Socrates’ associates he can’t draw away any of her clients Socrates replies that that is because she leads people down the easy path to vice, while he leads them on the difficult uphill path to virtue.²⁰ The only hint of hostility in Aelian’s anecdotes is an accusation of luxury attributed to the Cynic Diogenes (IV. 11), but even that is turned rather against the speaker than against Socrates: the exaggerated austerity of the Cynic finds ostentatious luxury in Socrates’ modest house and camp-bed (Prot. 310b–c) and the slippers which he wore on special occasions (Symp. 174a). While the Christian and pagan writers I have discussed show knowledge of the primary sources, especially Plato and Xenophon, neither class of writer is concerned with the detail of Socrates’ philosophical activity, taking that to include his argumentative practice and any doctrines which he may have maintained. For both types of writer he is primarily an emblematic ¹⁸ Cf. Xen. Mem. III. 7. 6–7 ¹⁹ The reference is to Archelaus’ invitations to poets, including Antiphon and Euripides, to stay at his court. ²⁰ This exchange also has echoes of Prodicus’ fable of the choice of Herakles (Xen. Mem II. 1. 21–34).

socrates under the severans 339 figure. For Athenaeus and Aelian he is the emblematic philosopher, conceived in a favourable sense as a sage and source of morally uplifting tales, or in an unfavourable sense as a pretentious hypocrite. For the Christians he is equally an emblematic figure, and equally open to favourable or unfavourable interpretation. For his Christian admirers he is emblematic of the anticipation by the Greeks of the Christian revelation, for his opponents he is emblematic of the enslavement of the pagan world to demonic powers. The writers I have considered interpret Socrates in the light of their particular preconceptions and interests; in that respect they resemble Socratic writers of every age.

General Bibliography Adam, J. [1890], Platonis Euthyphro, Cambridge. and Adam, A. M. [1905], Platonis Protagoras, 2nd edn., Cambridge. Allan, D. J. [1953], ‘Aristotle on the Origin of Moral Principles’, Actes du XI e Congr`es Internationale de Philosophie, vol. xii, 120–7, repr. in J. Barnes, M. Schofield, and R. Sorabji, eds., Articles on Aristotle 2: Ethics and Politics, London, 1977. Annas, J. [1981], An Introduction to Plato’s Republic, Oxford. [1998], ‘Doing without Objective Values: Ancient and Modern Strategies’, in S. Everson, ed., Companions to Ancient Thought 4: Ethics, Cambridge, 193–220. [1999], Platonic Ethics, Old and New, Ithaca, NY. Archer-Hind, R. D. [1883], The Phaedo of Plato, London. Bailey, C. [1928], The Greek Atomists and Epicurus, Oxford. Barnes, J. [1979], The Presocratic Philosophers, 2 vols., London. 2nd rev. edn. (1 vol.), 1982. [1984], ‘Reason and Necessity in Leucippus’, in Benakis [1984], i. 141–58. [1994], Aristotle, Posterior Analytics, 2nd rev. edn., Oxford. Benakis, L. [1984], ed., Proceedings of the First International Conference on Democritus, 2 vols., Xanthi. Benson, H. H. [1992], ed., Essays on the Philosophy of Socrates, Oxford. [2000] Socratic Wisdom: The Model of Knowledge in Plato’s Early Dialogues, Oxford. Bluck, R. S. [1955], Plato’s Phaedo, London. Bodn´ar, I. M. [1998], ‘Atomic Independence and Indivisibility’, OSAP 16: 35–61. Bostock, D. [1986], Plato’s Phaedo, Oxford. [1988], ‘Pleasure and Activity in Aristotle’s Ethics’, Phronesis 33: 251–72. [2000], Aristotle’s Ethics, Oxford. Brickhouse, T. C., and Smith, N. D. [1987], ‘Socrates on Goods, Virtue and Happiness’, OSAP 5: 1–28. [1989], Socrates on Trial, Oxford. Broadie, S. [1991], Ethics with Aristotle, New York. [1999], ‘Rational Theology’, in Long [1999], 205–24. [2003], ‘Reply to C. C. W. Taylor’, in R. Heinaman, ed., Plato and Aristotle’s Ethics, Aldershot, 21–7.

general bibliography 341 Brody, B. A. [1999], ‘How Much of the Brain Must Be Dead?’, in S. J. Youngner, R. M. Arnold, and R. Shapiro, eds., The Definition of Death: Contemporary Controversies, Baltimore, 71–82. Burnet, J. [1900], The Ethics of Aristotle, London. [1924], Plato’s Euthyphro, Apology of Socrates and Crito, Oxford. [1925], Plato’s Phaedo, Oxford. Burnyeat, M. F. [1994], ‘Did the Ancient Greeks have the Concept of Human Rights?’, Polis 13: 1–11. [1997], ‘The Impiety of Socrates’, Ancient Philosophy 17: 1–12, repr. in T. C. Brickhouse and N. D. Smith, eds., The Trial and Execution of Socrates: Sources and Controversies, New York, 2002. Calef, S. C. [1995a], ‘Piety and the Unity of Virtue in Euthyphro 11 E-14 C’, OSAP 13: 1–26. [1995b], ‘Further Reflections on Socratic Piety: A Reply to Mark McPherran’, OSAP 13: 37–43. Caston, V. [1997], ‘Epiphenomenalisms, Ancient and Modern’, PR 106: 309–63. Cooper, J. [1975], Reason and Human Good in Aristotle, Cambridge, Mass. [1999], Reason and Emotion: Essays on Ancient Moral Psychology and Ethical Theory, Princeton. Croiset, M. [1946], Platon (Bud´e edn.), vol. i, Paris. Crombie, I. M. [1962/3], An Examination of Plato’s Doctrines, 2 vols. (vol. i 1962, vol. ii 1963), London. Cronquist, J. [1980], ‘The Point of the Hedonism in Plato’s Protagoras’, Prudentia 12: 63–81. Dahl, N. O. [1984], Practical Reason, Aristotle, and Weakness of Will, Minneapolis. Davidson, D. [1969], ‘How is Weakness of the Will Possible?’, in J. Feinberg, ed., Moral Concepts, Oxford, 93–113, repr. in Davidson, Essays on Actions and Events, Oxford, 1980. Decleva Caizzi, F. [1999], ‘Protagoras and Antiphon: Sophistic Debates on Justice’, in Long [1999], 311–31. de Ley, H. [1969], ‘όξις ἐπιρυσμίη: A Critical Note on Democritus fr. 7’, Hermes 97: 497–8. Demos, R. [1964], ‘A Fallacy in Plato’s Republic?’, PR 73: 395–8, repr. in Vlastos [1971a], vol. ii. Dirlmeier, F. [1969], Aristoteles, Nikomachische Ethik (Aristoteles, Werke in ¨ Deutscher Ubersetzung, vol. vii), Berlin. Dodds, E. R. [1959], Plato, Gorgias. A Revised Text with Introduction and Commentary, Oxford. Dover, K. J. [1974], Greek Popular Morality in the Time of Plato and Aristotle, Oxford.

342 general bibliography Duff, A. [1987], ‘Aristotelian Courage’, Ratio 29: 2–15. Duncan, R. [1978], ‘Courage in Plato’s Protagoras’, Phronesis 23: 216–28. Dyroff, A. [1899], Demokritstudien, Leipzig. Dyson, M. [1976], ‘Knowledge and Hedonism in Plato’s Protagoras’, JHS 96: 32–45. Edwards, M. [2007], ‘Socrates and the Early Church’, in M. B. Trapp, ed., Socrates from Antiquity to the Enlightenment, Aldershot, 127–41. Engberg-Pedersen, T. [1983], Aristotle’s Theory of Moral Insight, Oxford. Farrar, C. [1988], The Origins of Democratic Thinking: The Invention of Politics in Classical Athens, Cambridge. Fine, G. [1999], ed., Plato, 2 vols., Oxford: vol. i Metaphysics and Epistemology, vol. ii Ethics, Politics, Religion, and the Soul. Frede, M. [2004], ‘The Early Christian Reception of Socrates’, in V. Karasmanis, ed., Socrates 2400 Years Since his Death, Athens, 481–90, repr. in L. Judson and V. Karasmanis, eds., Remembering Socrates: Philosophical Essays, Oxford, 2006. Friedl¨ander, P. [1964], Plato, trans. H. Meyerhoff, 3 vols., New York. Furley, D. J. [1967], Two Studies in the Greek Atomists, Princeton. [1983], ‘Weight and Motion in Democritus’ Theory’, OSAP 1: 193–209, repr. in Furley, Cosmic Problems, Cambridge, 1989. [1987], The Greek Cosmologists, vol. i: The Formation of the Atomic Theory and its Earliest Critics, Cambridge. Gallop, D. [1975], Plato, Phaedo, Oxford. ´ Gauthier, R. A., and Jolif, J. Y. [1958/9], Aristote, L’Ethique a` Nicomaque: Introduction, Traduction et Commentaire, 2 vols. in 3 parts (vol. i 1958, vol. ii 1959), Louvain. Glover, J. [1976], ed., The Philosophy of Mind, Oxford. [1988], I: The Philosophy and Psychology of Personal Identity, Harmondsworth. Godfrey, R. [1990], ‘Democritus and the Impossibility of Collision’, Philosophy 65: 212–17. Gomme, A. W., Andrewes, A., and Dover, K. J. [1970], A Historical Commentary on Thucydides, vol. iv, Oxford. Gomperz, T. [1905], Greek Thinkers, tr. G. G. Barry, vol. ii, London. Gosling, J. C. B. [1973], Plato, London. [1975], Plato, Philebus, Oxford. and Taylor, C. C. W. [1982], The Greeks on Pleasure, Oxford. Gottschalk, H. B. [1971], ‘Soul as Harmonia’, Phronesis 16: 179–98. Grote, G. [1865], Plato and the Other Companions of Sokrates, 3 vols., London. Gulley, N. [1968], The Philosophy of Socrates, London.

general bibliography 343 Guthrie, W. K. C. [1965], A History of Greek Philosophy, ii: The Presocratic Tradition from Parmenides to Democritus, Cambridge. [1969], A History of Greek Philosophy, iii: The Fifth-Century Enlightenment, Cambridge. Hackforth, R. [1928], ‘Hedonism in Plato’s Protagoras’, CQ 22: 39–42. [1955], Plato’s Phaedo, Cambridge. Hardie, W. F. R. [1980], Aristotle’s Ethical Theory, 2nd edn., Oxford. Hare, R. M., [1963], Freedom and Reason, Oxford. [1971a], ‘Wanting: Some Pitfalls’, in R. Binkley, R. Bronaugh, and A. Marras, eds., Agent, Action and Reason, Oxford, 81–127, repr. in Hare [1971b]. [1971b], Practical Inferences, London. [1981], Moral Thinking, Oxford. Havliˇcek, A., and Karf´ık, F. [2003], eds., Plato’s Protagoras. Proceedings of the Third Symposium Platonicum Pragense, Prague. Heidel, W. A. [1900], ‘On Plato’s Euthyphro’, Transactions of the American Philological Association, 31: 163–81. Heinimann, F. [1945], Nomos und Phusis, Basel. Hicken, W. F. [1954], ‘Phaedo 93a11-94b3’, CQ ns 4: 16–22. Hirzel, R. [1900], Agraphos Nomos (Abhandlungen der Phil.-Hist. Klasse der K¨oniglichen S¨achsischen Akademie der Wissenschaften 20.1), Leipzig. Hohfeld, W. [1920], Fundamental Legal Conceptions, New Haven. Inwood, B. [2003], ‘Reason, Rationalization, and Happiness in Seneca’, in Yu and Gracia [2003], 91–107. Irwin, T. [1978], ‘First Principles in Aristotle’s Ethics’, in P. French, T. Uehling, and H. Wettstein, eds., Midwest Studies in Philosophy III: Studies in Ethical Theory, 252–72. [1977], Plato’s Moral Theory, Oxford. Rev. and expanded as Plato’s Ethics, Oxford, 1995. [1980], ‘The Metaphysical and Psychological Basis of Aristotle’s Ethics’, in Rorty [1980], 35–53. [1986], ‘Socrates the Epicurean?’, Illinois Classical Studies 11, ch. 6, repr. in Benson [1992]. [1988], Aristotle’s First Principles, Oxford. [1999], Aristotle, Nicomachean Ethics, 2nd edn., Indianapolis. Jackson, B. D. [1971], ‘The Prayers of Socrates’, Phronesis 16: 14–37. Jowett, B. [1871], The Dialogues of Plato Translated into English, 5 vols., Oxford. Kahn, C. H. [1976], ‘Plato on the Unity of the Virtues’, in W. H. Werkmeister, ed., Facets of Plato’s Philosophy, Phronesis suppl. vol. ii, Assen, 21–39.

344 general bibliography Kahn, C. H. [1981], ‘The Origins of Social Contract Theory in the Fifth Century B.C.’, in G. B. Kerferd, ed., The Sophists and their Legacy, Hermes Einzelschriften 44, Wiesbaden, 92–108. [1985], ‘Democritus and the Origins of Moral Psychology’, American Journal of Philology 106: 1–31. (Shorter version, ‘Democritus on Moral Psychology’, in Benakis [1984], i. 307–15.) [1988a], ‘On the Relative Date of the Gorgias and the Protagoras’, OSAP 6: 69–102. [1988b], ‘Plato and Socrates in the Protagoras’, Methexis 1: 33–52. [1996], Plato and the Socratic Dialogue, Cambridge. [2003], ‘Socrates and Hedonism’, in Havl´ıcˇ ek and Karf ´ık [2003], 165–74. (Also in V. Karasmanis, ed., Socrates 2400 Years Since his Death, Athens, 2004, 111–15.) Kenny, A. [1963], Action, Emotion and Will, London. [1978], The Aristotelian Ethics, Oxford. [1979], Aristotle’s Theory of the Will, London. [1992], Aristotle on the Perfect Life, Oxford. Kerferd, G. B. [1947–8], ‘The Doctrine of Thrasymachus in Plato’s ‘‘Republic’’ ’, Durham University Journal 9: 19–27 [1981], The Sophistic Movement, Cambridge. Kline, A. D., and Matheson, C. A. [1987], ‘The Logical Impossibility of Collision’, Philosophy 62: 509–15. Konstan, D. [2000], ‘Democritus the Physicist’ (review of Taylor [1999]), Apeiron 33: 125–44. Kraut, R. [1997], ed., Plato’s Republic: Critical Essays, Lanham, Md. Laks, A. [1999], ‘Soul, Sensation, and Thought’, in Long [1999], 250–70. Lamb, D. [1996], Death, Brain Death and Ethics, Avebury. Loening, R. [1903], Die Zurechnungslehre des Aristoteles, Jena, repr. Hildesheim 1967. Long, A. A. [1988], ‘Socrates in Hellenistic Philosophy’, CQ ns 38: 150–71. [1999], ed., The Cambridge Companion to Early Greek Philosophy, Cambridge. Lucash, F. [1974], ‘More Pleasure in Aristotle’, Rivista di Filosofia Neo-scolastica 66: 126–30. McDowell, J. [1979], ‘Virtue and Reason’, The Monist 62: 331–50, repr. in R. Crisp and M. Slote, eds., Virtue Ethics, Oxford, 1997 and in Sherman [1989]. [1980], ‘The Role of Eudaimonia in Aristotle’s Ethics’, Proceedings of the African Classical Association 15, repr. in Rorty [1980], 359–76. McGibbon, D. [1960], ‘Pleasure as the ‘‘Criterion’’ in Democritus’, Phronesis 5: 75–7.

general bibliography 345 McKim, R. [1984], ‘Democritus against Scepticism: All Sense-Impressions are True’, in Benakis [1984], i. 281–90. McKirahan, R. D., Jr. [1994], Philosophy before Socrates, Indianapolis. McPherran, M. [1995], ‘Socratic Piety: In Response to Scott Calef’, OSAP 13: 27–35. [1999], The Religion of Socrates, University Park, Pa. [2003], ‘Reason’s Ascent: Happiness and the Disunity of Virtue in Plato and Plotinus’, in Yu and Gracia [2003], 135–58. Makin, S. [1993], Indifference Arguments, Oxford. Mansfeld, J. [2007], ‘Out of Touch: Philoponus as Source for Democritus’, in A. Brancacci and P.-M. Morel, eds., Democritus: Science, the Arts, and the Care of the Soul, Leiden, 277–92. Manuwald, B. [1975], ‘Lust und Tapferkeit: Zum gedanklichen Verh¨altnis zweier Abschnitte in Platons Protagoras’, Phronesis 20: 22–50. Monan, J. D. [1968], Moral Knowledge and its Methodology in Aristotle, Oxford. Morrison, D. [2003], ‘Happiness, Rationality, and Egoism in Plato’s Socrates’, in Yu and Gracia [2003], 17–34. Most, G. W. [1999], ‘The Poetics of Early Greek Philosophy’, in Long [1999], 332–62. Mourelatos, A. P. D. [2005], ‘Intrinsic and Relational Properties of Atoms in the Democritean Ontology’, in R. Salles, ed., Metaphysics, Soul, and Ethics in Ancient Thought: Themes from the Work of Richard Sorabji, Oxford, 39–63. Muller, ¨ A. W. [1982], Praktisches Folgern und Selbstgestaltung nach Aristoteles, Freiburg. Muller, ¨ C. W. [1965], Gleiches zu Gleichem. Ein Prinzip fr¨uhgriechischen Denkens, Wiesbaden. Nagel, T. [1971], ‘Brain Bisection and the Unity of Consciousness’, Synthese 22: 396–413, repr. in Glover [1976] and in Nagel [1979]. [1979] Mortal Questions, Cambridge. Natorp, P. [1893], Die Ethika des Demokritos: Text und Untersuchungen, Marburg. Nussbaum, M. C. [1986], The Fragility of Goodness, Cambridge. O’Brien, D. [1981], Theories of Weight in the Ancient World, vol. i: Democritus, Weight and Size, Paris. Owen, G. E. L. [1960], ‘Logic and Metaphysics in Some Earlier Works of Aristotle’, in I. D¨uring and G. E. L. Owen, eds., Plato and Aristotle in the Mid-fourth Century, G¨oteborg, 163–90, repr. in J. Barnes, M. Schofield, and R. Sorabji, eds., Articles on Aristotle 3: Metaphysics, London, 1979, in Owen, Logic, Science and Dialectic: Collected Papers in Greek Philosophy, ed. M. C. Nussbaum, London, 1986, and in

346 general bibliography German translation in F.-P. Hager, ed., Metaphysik und Theologie des Aristoteles, Darmstadt, 1969. [1971–2], ‘Aristotelian Pleasures’, PAS 72: 135–52, repr. in J. Barnes, M. Schofield, and R. Sorabji, eds., Articles on Aristotle 2: Ethics and Politics, London, 1977, and in Owen, Logic, Science and Dialectic. Penner, T. [1973], ‘The Unity of Virtue’, PR 82: 35–68, repr. in Benson [1992] and in Fine [1999], vol. ii. and Rowe, C. J. [1994], ‘The Desire for Good: Is the Meno Inconsistent with the Gorgias?’, Phronesis 39: 1–25. Plum, F. [1999], ‘Clinical Standards and Technological Confirmatory Tests in Diagnosing Brain Death’, in S. J. Youngner, R. M. Arnold, and R. Shapiro, eds., The Definition of Death: Contemporary Controversies, Baltimore, 34–65. Procop´e, J. [1989/90], ‘Democritus on Politics and the Care of the Soul’, CQ ns 39: 307–31 and 40: 21–45. Putnam, H. [1975], ‘The Meaning of ‘‘Meaning’’ ’, in Putnam, Mind, Language and Reality: Philosophical Papers, ii. 215–71, Cambridge. Rabinowitz, W. G. [1958], ‘Platonic Piety: An Essay toward the Solution of an Enigma’, Phronesis 3: 108–20. Raven, J. E. [1965], Plato’s Thought in the Making, Cambridge. Rawls, J. [1963], ‘The Sense of Justice’, PR 72: 281–305, repr. in J. Feinberg, ed., Moral Concepts, Oxford, 1969, and in Rawls, Collected Papers, ed. S. Freeman, Cambridge, Mass., 1999. Reeve, C. D. C. [1992], Practices of Reason, 1992. Rorty, A. O. [1980], ed., Essays on Aristotle’s Ethics, Berkeley and Los Angeles. Ross, W. D. [1924], Aristotle’s Metaphysics, 2 vols., Oxford. [1925], Ethica Nicomachea, in The Works of Aristotle Translated into English, vol. vii, Oxford. Rowe, C. J. [2003], ‘Hedonism in the Protagoras again: Protagoras 351b ff.’, in Havl´ıcˇ ek and Karf ´ık [2003], 133–47. Rudebusch, G. [1999], Socrates, Pleasure, and Value, New York. Sachs, D. [1963], ‘A Fallacy in Plato’s Republic’, PR 72: 141–58, repr. in Vlastos, [1971a], vol. ii and in Kraut [1997]. Santas, G. [1964], ‘The Socratic Paradoxes’, PR 73: 147–64, repr. in A. Sensonske and N. Fleming, eds., Plato’s Meno: Text and Criticism, Belmont, Calif., 1965 and as ch. 6 of Santas, Socrates: Philosophy in Plato’s Early Dialogues, London, 1979. Sassi, M. M. [1978], Le teorie della percezione in Democrito, Florence. Saunders, T. J. [1977–8], ‘Antiphon the Sophist on Natural Laws (B44DK)’, PAS 78: 215–36.

general bibliography 347 Scaltsas, T. [1990], ‘Soul as Attunement: An Analogy or a Model?’, in P. Nicalopoulos, ed., Greek Studies in the Philosophy and History of Science (Boston Studies in the Philosophy of Science 121), 109–19. Schueler, G. F. [1983], ‘Akrasia Revisited’, Mind 92, 580–4. Scott, D. [2006], Plato’s Meno, Cambridge. Searle, J. [1967], ‘Determinables and Determinates’, in P. Edwards, ed., The Encyclopedia of Philosophy, London, i. 357–9. Sedley, D. N. [1982], ‘Two Conceptions of Vacuum’, Phronesis 27: 175–93. [1999], ‘The Idea of Godlikeness’, in Fine [1999], ii. 309–28. Segvic, H. [2000], ‘No One Errs Willingly: The Meaning of Socratic Intellectualism’, OSAP 19: 1–45. Sherman, N. [1989], The Fabric of Character: Aristotle’s Theory of Virtue, Oxford. Skemp, J. B. [1952], Plato’s Statesman, London. Sorabji, R. [1973–4], ‘Aristotle on the Role of the Intellect in Virtue’, PAS 74: 107–29, repr. in Rorty [1980]. Stocker, M. [1990], ‘Courage, the Doctrine of the Mean and the Possibility of Evaluative and Emotional Coherence’, in Stocker, Plural and Conflicting Values, Oxford, 129–64. Stokes, M. C. [1986], Plato’s Socratic Conversations, London. [1990], ‘Some Pleasures of Plato, Republic IX’, Polis 9: 2–51. Striker, G. [1974], Κριτήριoν τῆς ἀληθείαζ. Nachrichten der Akademie der Wissenschaften in G¨ottingen, Phil.-Hist Klasse, No. 2, 47–110. [1977], ‘Epicurus on the Truth of Sense-Impressions’, Archiv f¨ur Geschichte der Philosophie 59: 125–42. [2006], ‘Aristotle’s Ethics as Political Science’, in B. Reis, ed., The Virtuous Life in Greek Ethics, Cambridge, 127–41. Sullivan, J. P. [1961], ‘The Hedonism in Plato’s Protagoras’, Phronesis 6: 9–28. Taylor, A. E., [1926], Plato: The Man and His Work, London. [1934], The Laws of Plato Translated into English, London. Taylor, C. C. W. [1976], Plato, Protagoras, Oxford. 2nd rev. edn. 1991. [1984], ‘Reply to Schueler on Akrasia’, Mind 93, 584–6. [1986], ‘Plato’s Totalitarianism’, Polis 5.2: 4–29, repr. in Kraut [1997] and in Fine [1999], vol. ii. [1994], review of Kenny [1992], International Philosophical Quarterly 34: 115–17. [1997], ed., Routledge History of Philosophy vol. i: From the Beginning to Plato, London. [1998], Socrates, Oxford.

348 general bibliography Taylor, C. C. W. [1999], The Atomists: Leucippus and Democritus, Fragments. A Text and Translation with a Commentary, Toronto. [2006], Aristotle, Nicomachean Ethics, Books II–IV, Translated with a Commentary, Oxford. [2007], ‘Nomos and Phusis in Democritus and Plato’, Social Philosophy and Policy 24.2: 1–20. Also in D. Keyt and F. D. Miller, Jr., eds., Freedom, Reason and the Polis: Essays in Ancient Greek Political Philosophy, Cambridge, 2007, 1–20. Urmson, J. O. [1967], ‘Aristotle on Pleasure’, in J. M. E. Moravcsik, ed., Aristotle: A Collection of Critical Essays, Garden City, NY, 323–33. [1984], Nicomachean Ethics (Ross’s translation (see Ross [1925]) rev. J. O. Urmson), in J. Barnes, ed., The Complete Works of Aristotle. The Revised Oxford Translation, 2 vols., Princeton. Vander Waerdt, P. A. [1994], ed., The Socratic Movement, Ithaca, NY. Vegetti, M. [1999], ‘Culpability, Responsibility, Cause: Philosophy, Historiography, and Medicine in the Fifth Century’, in Long [1999], 271–89. Vlastos, G. [1945/6], ‘Ethics and Physics in Democritus’, PR 54: 578–92 and 55: 53–64, repr. in Vlastos, Studies in Greek Philosophy, ed. D. W. Graham, 2 vols., Princeton, 1995, vol. i. [1956], Plato, Protagoras, Indianapolis. [1969], ‘Socrates on Acrasia’, Phoenix 23: 71–88, repr. in Vlastos, Studies in Greek Philosophy, vol. ii. [1971a], ed., Plato, 2 vols., Garden City, NY: vol. i Metaphysics and Epistemology, vol. ii Ethics, Politics, and Philosophy of Art and Religion. [1971b], ‘Justice and Happiness in the Republic’, in Vlastos [1971a], vol. ii, 66–95, repr. in Vlastos, Platonic Studies (2nd edn.). [1973] Platonic Studies, Princeton. Second edition 1981. [1981] ‘The Unity of the Virtues in the Protagoras’, in Vlastos, Platonic Studies (2nd edn.), 221–69. Revised version of paper originally published in Review of Metaphysics 25 (1971–2). [1984], ‘Happiness and Virtue in Socrates’ Moral Theory’, Proceedings of the Cambridge Philological Society 30: 182–213. [1991], Socrates, Ironist and Moral Philosopher, Cambridge. von Fritz, K. [1938], Philosophie und sprachlicher Ausdruck bei Demokrit, Plato und Aristoteles, New York. Wagner, E. [2000], ‘Supervenience and the Thesis that the Soul is a Harmonia’, Southwest Philosophy Review 16: 1–20, repr. in E. Wagner, ed., Essays on Plato’s Psychology, Lanham, Md., 2001. Waldron, J. [1984], Theories of Rights, Oxford.

general bibliography 349 Warren, J. [2002], ‘Democritus, the Epicureans, Death and Dying’, CQ ns 52: 193–206. [2004], Facing Death: Epicurus and his Critics, Oxford. Watson, G. [1977], ‘Skepticism about Weakness of Will’, PR 86: 316–39. Weiss, R. [1990], ‘Hedonism in the Protagoras and the Sophist’s Guarantee’, Ancient Philosophy 10: 17–39. White, S. A. [2000], ‘Socrates at Colonus’, in N. D. Smith and P. B. Woodruff, eds., Reason and Religion in Socratic Philosophy, Oxford, 151–75. Wiggins, D. [1975–6], ‘Deliberation and Practical Reason’, PAS 76: 29–51, repr. in Rorty [1980]. Williams, B. [1973], ‘The Analogy of City and Soul in Plato’s Republic’, in E. N. Lee, A. P. D. Mourelatos, and R. M. Rorty, eds., Exegesis and Argument: Studies in Greek Philosophy Presented to Gregory Vlastos, Phronesis suppl. vol. 1, Assen, 196–206, repr. in Kraut [1997], in Fine [1999], vol. ii, in Wagner [2001], and in Williams, The Sense of the Past: Essays in the History of Philosophy, ed. M. Burnyeat, Princeton, 2006. Wolfsdorf, D. [2006], ‘Desire for Good in Meno 77B2-78B6’, CQ ns 56: 77–92. Woods, M. J. [1986], ‘Intuition and Perception in Aristotle’s Ethics’, OSAP 4: 145–66, repr. in Woods, Four Prague Lectures and Other Texts, Prague, 2001. [1987], ‘Plato’s Division of the Soul’, Proceedings of the British Academy 73: 23–47. Woozley, A. D. [1979], Law and Obedience, London. Yu, J. [2003], ‘Will Aristotle Count Socrates Happy? Socrates and the Nicomachean Ethics’, in Yu and Gracia [2003], 51–73. and Gracia, J. J. E. [2003], eds., Rationality and Happiness: From the Ancients to the Early Medievals, Rochester, NY. Zaner R. M. [1988], ed., Death: Beyond Whole Brain Criteria, Dordrecht. Zeyl, D. [1980], ‘Socrates and Hedonism’, Phronesis 25: 250–69, repr. in J. P. Anton and A. Preus, eds., Essays in Ancient Greek Philosophy, vol. iii, Albany, NY, 1989. [1982], ‘Socratic Virtue and Happiness’, Archiv f¨ur Geschichte der Philosophie 64: 225–38.

Publications of C. C. W. Taylor (including critical notices, but excluding reviews and other occasional pieces)

Items reprinted in the present volume are indicated by an asterisk. 1. ‘Pleasure’, Analysis 23, suppl. vol. (1963), 2–19. 2. Critical Notice of R. M. Hare, Freedom and Reason, Mind 74 (1965), 280–98. Repr. in G. Wallace and A. D. M. Walker, eds, The Definition of Morality, Methuen, London, 1970. 3. ‘States, Activities and Performances’ (symposium with T. C. Potts), Proceedings of the Aristotelian Society, suppl. vol. 39 (1965), 85–102. ∗ 4. ‘Pleasure, Knowledge and Sensation in Democritus’, Phronesis 12 (1967), 6–27. Repr. in T. Irwin, ed., Articles on Greek and Roman Philosophy, Garland Publishing Inc., New York, 1994. 5. ‘Plato and the Mathematicians’, Philosophical Quarterly 17 (1967), 193–203. 6. ‘Understanding a Want’, in J. J. MacIntosh and S. C. Coval, eds., The Business of Reason, Routledge and Kegan Paul, London, 1969, 244–61. 7. ‘Forms as Causes in the Phaedo’, Mind 78 (1969), 45–59. Repr. in Irwin op. cit. no. 4 above and in N. D. Smith, ed., Plato: Critical Assessments, Routledge, London, 1998, vol. ii. 8. Plato, Protagoras. Translated with notes. Clarendon Press, Oxford, 1976. Rev. edn. 1991. 9. ‘Berkeley’s Theory of Abstract Ideas’, Philosophical Quarterly 28 (1978), 97–115. Repr. in G. Pitcher, ed., The Philosophy of George Berkeley, Garland Publishing Inc., New York, 1988. ∗ 10. ‘ ‘‘All Perceptions Are True’’ ’, in M. Schofield, M. Burnyeat, and J. Barnes, eds., Doubt and Dogmatism: Studies in Hellenistic Epistemology, Clarendon Press, Oxford, 1980, 105–24. ∗ 11. ‘Plato, Hare and Davidson on Akrasia’, Mind 89 (1980), 499–518. ∗ 12. ‘The End of the Euthyphro’, Phronesis 27 (1982), 109–18. Repr. in Irwin op. cit. no. 4 above. 13. (with J. C. B. Gosling) The Greeks on Pleasure, Clarendon Press, Oxford, 1982.

publications of c. c. w. taylor 351 ∗

14. ‘The Arguments in the Phaedo Concerning the Thesis that the Soul is a Harmonia’, in J. P. Anton and A. Preus, eds., Essays in Ancient Greek Philosophy, vol. ii, SUNY Press, Albany, NY, 1983, 217–31. Repr. in E. Wagner, ed., Essays on Plato’s Psychology, Lexington Books, Lanham, Md., 2001. ∗ 15. ‘Plato and Aristotle on the Criterion of Real Pleasures’, in J. Harmatta, ed., Actes du viie Congr`es de la F.I.E.C., vol. ii, Akad. Kiad´o, Budapest, 1983, 345–56. 16. ‘Reply to Schueler on Akrasia’, Mind 92 (1984), 584–6. 17. ‘Berkeley on Archetypes’, Archiv f¨ur Geschichte der Philosophie 67 (1985), 65–79. 18. ‘Action and Inaction in Berkeley’, in John Foster and Howard Robinson, eds., Essays on Berkeley, Clarendon Press, Oxford, 1985, 211–25. 19. ‘Emotions and Wants’, in J. Marks, ed., The Ways of Desire, Precedent Publishing, Chicago, 1986, 217–31. 20. ‘Plato’s Totalitarianism’, Polis 5.2 (1986), 4–29. Repr. in R. Kraut, ed., Plato’s Republic: Critical Essays, Rowman and Littlefield Publishers Inc., Lanham, Md., 1997, and in G. Fine, ed., Plato (Oxford Readings in Philosophy), vol. ii, Oxford University Press, Oxford, 1999. 21. Critical Notice of Martha C. Nussbaum, The Fragility of Goodness, Mind 96 (1987), 407–14. 22. ‘Hellenistic Ethics: A Discussion of Malcolm Schofield and Gisela Striker, eds., The Norms of Nature, Studies in Hellenistic Ethics’, Oxford Studies in Ancient Philosophy 5 (1987), 235–45. 23. (with J. P. Dancy and J. M. E. Moravcsik, eds.) Human Agency: Language, Duty, and Value. Philosophical Essays in Honor of J. O. Urmson, Stanford University Press, Stanford, Calif., 1988. ∗ 24. ‘Urmson on Aristotle on Pleasure’, in no. 23 above, 120–32. 25. ‘Aristotle’s Epistemology’, in S. Everson, ed., Companions to Ancient Thought 1, Epistemology, Cambridge University Press, Cambridge, 1990, 116–42. 26. (with J. C. B. Gosling) ‘The Hedonic Calculus in the Protagoras and the Phaedo: A Reply’, Journal of the History of Philosophy 28 (1990), 115–16. ∗ 27. ‘Popular Morality and Unpopular Philosophy’, in E. M. Craik, ed., ‘Owls to Athens’: Essays on Classical Subjects Presented to Sir Kenneth Dover, Clarendon Press, Oxford, 1990, 233–43. ∗ 28. ‘Socratic Ethics’, in B. S. Gower and M. C. Stokes, eds., Socratic Questions, Routledge, London, 1992, 137–52. 29. Critical Notice of Gregory Vlastos, Socrates: Ironist and Moral Philosopher, Philosophical Quarterly 42 (1992), 228–34.

352 publications of c. c. w. taylor 30. Critical Notice of C. Gill, ed., The Person and the Human Mind, Polis 11 (1992), 62–71. 31. Ethics and the Environment, ed. C. C. W. Taylor, Corpus Christi College, Oxford, 1992 (proceedings of conference held at CCC in 1991). 32. ‘Politics’, in J. Barnes, ed., The Cambridge Companion to Aristotle, Cambridge University Press, Cambridge, 1995, 233–58. 33. Various articles on ancient philosophy in T. Honderich, ed., The Oxford Companion to Philosophy, Clarendon Press, Oxford, 1995. Rev. versions of some entries in 2nd rev. edn., 2005. 34. Various articles on ancient philosophy in S. Hornblower and A. Spawforth, eds., The Oxford Classical Dictionary, 3rd edn., Oxford University Press, Oxford, 1996. Article on Sophists reprinted in Hornblower and Spawforth, eds., The Oxford Companion to Classical Civilization, Oxford University Press, Oxford, 1998. 35. Plato, Protagoras, trans. with introd. and notes, Oxford University Press, Oxford (World’s Classics), 1996. 36. Routledge History of Philosophy, i. From the Beginning to Plato, ed. C. C. W. Taylor, Routledge, London, 1997. (Introduction and chapter on Anaxagoras and the Atomists by CCWT.) ∗ 37. ‘Platonic Ethics’, in S. Everson, ed., Companions to Ancient Thought 4, Ethics, Cambridge University Press, Cambridge, 1998, 49–76. 38. Articles on Democritus, Eudaimonia, and Leucippus in E. Craig, ed., The Routledge Encyclopedia of Philosophy, Routledge, London, 1998. 39. Socrates (Past Masters Series), Oxford University Press, Oxford, 1998. German translation 1999, Dutch and Korean translations 2001, Turkish translation 2002, Chinese translation 2007. Reissued (with R. M. Hare, Plato, and J. Barnes, Aristotle) in Greek Philosophers: Socrates, Plato and Aristotle, Oxford University Press, Oxford, 1999. Reissued in Very Short Introductions series, Oxford University Press, Oxford, 2000. 40. The Atomists: Leucippus and Democritus, Fragments. A Text and Translation with a Commentary (Phoenix Presocratics series), Toronto University Press, Toronto, 1999. ∗ 41. ‘The Atomists’, in A. A. Long, ed., The Cambridge Companion to Early Greek Philosophy, Cambridge University Press, Cambridge, 1999, 181–204. 42. ‘Democritus’, in C. Rowe and M. Schofield, eds., The Cambridge History of Greek and Roman Political Thought, Cambridge University Press, Cambridge, 2000, 122–9. 43. ‘Ethics and Politics in Aristotle. A Discussion of Richard Kraut, Aristotle, Political Philosophy’, Oxford Studies in Ancient Philosophy 23 (2002), 265–77.

publications of c. c. w. taylor 353 44. ‘The Origins of our Present Paradigms’, in J. Annas and C. Rowe, eds., New Perspectives on Plato, Modern and Ancient, Harvard University Press, Cambridge, Mass. and Center for Hellenic Studies, Washington, DC, 2002, 73–84. ∗ 45. ‘Aristoteles u¨ ber den praktischen Intellekt’, in T. Buchheim, H. Flashar, and R. A. H King, eds., Kann man heute noch etwas anfangen mit Aristoteles?, Meiner Verlag, Hamburg, 2003, 142–62. ∗ 46. ‘Plato on Rationality and Happiness’, in J. Yu and J. J. E. Gracia, eds., Rationality and Happiness: From the Ancients to the Early Medievals, University of Rochester Press, Rochester, NY, 2003, 35–49. ∗ 47. ‘Pleasure: Aristotle’s Response to Plato’, in R. Heinaman, ed., Plato and Aristotle’s Ethics (Proceedings of the Fourth Keeling Colloquium), Ashgate, Aldershot, 2003, 1–20. ∗ 48. ‘The Hedonism of the Protagoras Reconsidered’, in A. Havl´ıcˇ ek and F. Karf´ık, eds., Plato’s Protagoras, Proceedings of the Third Symposium Platonicum Pragense, OIKOUMENH, Prague, 2003, 148–64. ∗ 49. ‘Sagezza e Coraggio nel Protagora e nell’ Etica Nicomachea’, in G. Casertano, ed., Il Protagora di Platone: struttura e problematiche, Loffredo Editore, Naples, 2004, 716–28. 50. ‘Socrates the Sophist’, in V. Karasmanis, ed., Socrates 2400 Years Since his Death, European Cultural Centre of Delphi and Hellenic Ministry of Culture, Athens/Delphi, 2004, 211–19. Repr. (with some changes) in L. Judson and V. Karasmanis, eds., Remembering Socrates: Philosophical Essays, Clarendon Press, Oxford, 2006. 51. ‘Gosling, Justin Cyril Bertrand, (1930– )’ and ‘Urmson, James Opie (1915– )’, in S. Brown, ed., The Dictionary of Twentieth-Century British Philosophers, Thoemmes Continuum, Bristol, 2005, 338–9, 1065–6. ∗ 52. ‘Socrates’, in D. Borchert, ed., The Encyclopedia of Philosophy, 2nd edn., Macmillan Reference USA, Detroit, 2005, vol. ix, 105–14. 53. ‘Ancient Philosophy, Impact of’, in A. C. Grayling, A. Pyle, and N. Goulder, eds., The Continuum Encyclopedia of British Philosophy, Thoemmes Continuum, London, 2006, i. 93–9. 54. Aristotle, Nicomachean Ethics, Books II–IV. Translated with a Commentary. Clarendon Press, Oxford, 2006. ∗ 55. ‘Democritus and Lucretius on Death and Dying’, in A. Brancacci and P.-M. Morel, eds., Democritus: Science, the Arts, and the Care of the Soul, Proceedings of the International Colloquium on Democritus (Paris, 18–20 September 2003) (Philosophia Antiqua 102), Brill, Leiden, 2007, 77–86. ∗ 56. ‘Socrates under the Severans’, in S. Swain, S. Harrison, and J. Elsner, eds., Severan Culture, Cambridge University Press, Cambridge, 2007, 496–507.

354

publications of c. c. w. taylor

57. ‘Nomos and Phusis in Democritus and Plato’, Social Philosophy and Policy, 24.2 (2007), 1–20. Also in D. Keyt and F. D. Miller, Jr., eds., Freedom, Reason and the Polis: Essays in Ancient Greek Political Philosophy, Cambridge University, Press, Cambridge, 2007, 1–20. 58. Pleasure, Mind, and Soul: Selected Papers in Ancient Philosophy. Clarendon Press, Oxford, 2008. Forthcoming 59. ‘Plato’s Epistemology’, in G. Fine, ed., The Oxford Handbook to Plato, Clarendon Press, Oxford, 2008, 288–325.

Index of passages cited Principal discussions are indicated by bold type. Aelian VH I.16 338 II.1 338 II.13 337 II.25 336 II.43 336 III.28 338 IV.11 338 V.8 337 VII.10 337 VII.14 337 VIII.1 338 VIII.10 338 IX.7 337 IX.29 337 IX.35 338 XI.12 335 XII.25 337 XIII.27 337 XIII.32 338 XIV.17 338 Aesch. Septem 597–612 64 610 153 Aet. I.26.2 188 I.29.7 187–8 IV.4.6 316, 319 IV.4.7 318 IV.5.1 316 IV.9.5 24 IV.19.3 189 Alexander In Meta. 36.21–5 188 In Top. 21.21 318 Antiphon DK 87 B 22 6–7 B 44 154, 199 Antisthenes Fr. 40B Decleva Caizzi 328 Apuleius De Plat. 1.1 330

Ar. An. post. I.1, 71a1–2 210 II.19 215 De an. 403b31–404a9 320 404a5 ff. 4 404a27 ff. 18 404a27–31 316–17 405a8–13 316–17 405a11–13 8 407b27–8 78 I.4 74, 77–8 408a30–31 77 424a3–10 118 425b12–25 263 432a16 118 III.9–10 207 De caelo 295a10–12 15 300b8–10 8 300b8–16 184 EE 1214a1–8 240 1215a7–8 213 I.6 240 1216b26–8 213 1216b26–35 122 1229b28–30 286 1235b3–18 122 1235b31 ff. 106 1236a16–23 167 1237a1–10 106 1237a25 106 1238a27 106 1239b30 ff. 106 VIII.3 221 1249a22 ff. 105–6 1249b17–21 105 EN 1095a17–20 240 1095a18–20 150 1095a19 205 1095a28–b7 214 1095a28–30 122 1095b2–5 208 1095b4–9 211

356 index of passages cited Ar. (continued) I.6 240 1097a7–28 241 1097a15–b21 241 1097a17–18 241 1097a18–19 241 1097a23–31 241 1097b2–5 151 1097b22–1098a20 241 1098b3–4 211 1098b9–12 99, 213 1099a7 ff. 99, 102 1099a8–10 252 1099a11–15 102 1099a13 102 1099a15–16 251 1099a17–20 101 1099a21–4 252 1103a1–3 212 1103a32–b2 207 II 219 II.1–3 210 (esp. 1104b3–13) 1104a6–10 216, 220 1104b3–8 101, 289 1104b7–8 290 1105a32 290 1107a33–b4 285 1107b1 285 1109b20–23 216 III 207 1112a15–16 206 1112b11–16 207 1112b26–9 207 1113a25 ff. 104, 106 1113a29–33 252 1115a21–2 288 1115a23–4 288 1115a26 289 III.7 286 1115b9 289 1115b11 289 1115b11–13 289 1115b26–8 286 1115b28–32 286 1116a17–b3 290 1116b4–5 134 1117a33–4 289 1117b8–9 289 1117b9–13 289 III.10 260, 262

III.10–12 108–9 1117b29–31 108 1118a16–23 249 1118a23–6 108 1118a26 260 1118a26–31 108 1118a26–b1 260 1118a27–31 117 1118a29–32 117 1118a32–b1 108 1118a31–2 108 1118b2–3 261 1118b4–8 118 1119a9–10 293 1126b2–4 216 1130a10–13 67 VI 221 VI.2 206–7 1139a22–6 176 1139a26–31 206 1139a28 206 1139a35–6 218 1139b31–2 209 1140a1 204, 206–7 1140a9–10 206 VI.5 207 1140a25–8 205 1140b4–6 210 1140b5–6 206 1140b7 205 1140b11–12 209 1140b11–21 218 1140b20–21 210 1141a7–8 209 1141a18–20 209 1142a9 215 1142a23–30 215 1142a25–6 209 1142b16 209 1142b31–3 209 1142b33 210 1143a35–b5 209 1143a35–b14 215 1143b4–5 215 1143b6–9 215 1143b11–14 216–17 1144a7–9 208, 217 1144a24–b2 234 1144a29–36 216 1144b19–20 134 1145a4–6 208, 217–18

index of passages cited 357 1145b2–7 213 1145b23–4 135 1145b26–7 134 1150b19–28 294 1151a15–20 218 1151a18–19 210 VII.11–14 241 1152b5–7 245 1152b12–14 242 1152b12–15 245 1152b22–3 245 1152b25–1153a7 242 1152b36–1153a2 242 1153a2–6 242 1153a9–12 257 1153a12–15 255 1153a15–17 257 1154a21–b20 240 1154b9 ff. 103 1154b12–13 292 1154b20 103 IX.2 216 1166a12–13 252 1170a29–33 263 X 214, 221 X.1–5 241 X.3–4 256–60, 262 1173a15 257 1173a29–b20 243 1173b9–11 261 1173b13–15 243 1173b15–20 243 1173b20–5 100 1173b22–5 179–252 1173b30 257 1174a3–7 264 1174a6 257 1174a13–14 259 1174b9–17 264 1174b14–16 253 1174b14–26 259 1174b18–19 253 1174b20–23 253 1174b23 115 1174b32–3 263 1174b33–1175a1 259 1175a10–21 260 1175a12–17 257 1175a16–19 260 1175a26–8 260 1175a30–b6 114

1175a33–b1 260 1175a34–5 257 1175b18 257 1176a3–8 249 1176a3–9 252 1176a8–22 179 1176a16–19 252 1176a19–24 100 X.7 239 GA 789b2–3 184, 186 GC 314a21 ff. 19 315b6–15 194 315b9 196 316a14–b7 182 324a35–325a31 181–2 325a27–8 183 325a23 ff. 15, 19 325a24–6 194 MA 7 207 Meta. 978b1–4 135 985b15–16 192 985b18 183 985b18–19 182 1003a33–b12 167 1009a38 ff. 195 1009b7 ff. 4, 5–6, 190 1009b9 195–6 1009b11–12 195 1009b12 ff. 18 1011b26 30  6 135 1048b23–4 259 1048b33–4 259 1078b17 ff. 135 1078b27–32 135, 301 MM 1179a8–10 207 PA 690b29–691a4 118 Phys. 196a24–8 187 211a7–11 214 213a2–4 183 252a32–b2 188 Pol. 1253b20–3 131 Rhet. 1368b7–9 131 I.13 130 1373b6–9 130 I.15 132 1375a27–b15 132 1375a29–31 132 1383a17–19 287 ps.-Ar. De mundo 396b7 ff. 75

358 index of passages cited Aristocles See Eus. Aristoph. Nub. general 297, 337 904–6 130 1405ff. 130 Aspasius In EN 88.11–13 108 Ath. Deipn. I.20f 334 IV.157e 334 IV.158f 334 V.186d 334 V. 215e–216e 335 V.219b 335 V.221e 335 XI.504d–e 334 XI.507c–d 330 XI.784d 334 XIII.599a–b 335 XIII.611a 335 XIII.611e–612a 336 XIV.644 335 XV.686d–e 336 Chalcidius Comm. 168 332 Cic. De fin. V.8.23 2, 13 De or. I.231 336 ND I.34.93 332 Tusc. I.34.82 318 Clem. Protr. VI.71.1–3 328 VII.76.3–6 329 Strom. I.17.83.4 329 I.21.153.5 329 III.3.17.1 79 V.2.5.1 328 V.2.14.1 329 V.11.67.1 329 V.14.91.4 329 V.14.97.1 329 V.14.99.3 328 VI.2.5.1 328 VI.6.53.2–3 329 Cologne papyrus (PK¨oln 205) 276 Critias DK 88 B 25 199

Dem. 18.275 132 23. 61, 70, 85 132 43.22 132 45.53 132 Democritus DK 68 B 3 11, 13 B 6 191–2 B 6–10 15, 193 B 7 15, 191–2 B 8 191–2 B 9 17, 191–2, 193–4 B 10 191–2 B 11 16, 193 B 33 4, 9–11, 202–3 B 37 1–2, 3 B 65 3 B 69 4, 9 B 71 13 B 72 13 B 74 11–12, 13, 20 B 84 199 B 98–9 3 B 112 3 B 117 15, 191, 193, 196 B 125 4, 9, 14, 40, 193, 194–5 B 164 15 B 170–1 169 B 174 3, 13, 199 B 175 199 B 187 2, 4–5, 9 B 188 12 B 189 2 B 191 4, 7–9, 11, 13 B 194 3 B 197 203 B 200 13 B 204 13 B 215 13, 199 B 230 13 B 244 199 B 248 200 B 252 200 B 257 6 B 258 203 B 259 203 B 264 199 DL II.16 154 II.36 297

index of passages cited 359 II.40–1 336 II.44 336 III.5 330 VIII.46 78 VIII.58 142 IX.3 337 IX.30–3 188 IX.31 15, 23, 203 IX.45 4, 6, 13, 187 IX.45–9 16 IX.46 6 IX.47 192, 202 X.31 26, 29, 37 X.32 29, 31–3, 37, 40 X.38 27 X.50–1 24, 27 X.50–2 23, 25 X.54 38 X. 87 27 X. 97 27 X.104 27 X.129 26 X.146 27 X.146–7 28, 39 X.147 23 Empedocles DK 31 B 23 74 B 105 323 B 117 80 B 135 130 Ep. KD 2 318 Eur. Suppl. 19, 311, 526, 563, 671 131–2 Fr. 282.23–7 153 Eus. PE XIV.20.9 24–5, 37 Heraclitus DK 22 B 40 3 B 51 74 B114 131 B 119 169 Herodotus III.38 154 VII.103.2 186 Isa. 2.24 132 Isocr. 12.169–70 131

Jerome Ep. XXII.30 328 Justin II. Apol. 13 328 Lactantius Div. inst. II.14.9 332 Leucippus DK 67 B 1 186–7 Lucr. II.910–14 322 III.94–5 319 III.98–135 78 III.136–9 319 III.141–2 319 III.143–4 319 III.153–60 320 III.231–6 320 III.241–5 320 III.247–8 320 III.272–5 320 III.280–1 320 III.288–93 320 III.396–400 320 III.398–9 322 III.425–44 321 III.526–32 321 III.634–63 321 III.638–41 321 III.645–7 322 III.713–21 322 III.870–93 319 IV.473–9 39–40 IV.522–721 39 IV.622–4 39 V 200 Lys. 6.10 132 Macrobius DK 44 A 73 78–9 Melissus DK 30 B 7 183 Minucius Felix Octavius 19.13 332 26.9 332 38.5 332 Olympiodorus Vita Plat. 4 330

360 index of passages cited Origen Cels. II.41 330 III.8 330 III.66 330 IV.59 331 IV.97 330 VI.8 330–1 VII.6 331 Parmenides DK 28 B 8.22–5 183 Pausanias I.30.3 330 Philolaus DK 44 B 6 6 Philoponus In De an. 35.12 317 In De an. 71.19 ff. 19 In De an. I.4 81 In GC 23.1–16 194 In GC 158–.26–159.7 184 In Phys. 494.19–25 184 Pindar Isthmian 8.24–7 153 Pl. Ap. general 124, 297, 305 20c 303 21b 302 28b 160 28d–29d 123–5 29b 302 29b6 ff. 125 30c 160 30c6–d1 330 31c–d 304 31d 330 37b 302 40a–b 304 41c–d 160 Charm. general 72, 148, 156, 162, 277–8 159b–160b 177 174d–175a 66 Crt. general 124 44a–b 305 47e 147, 160, 162–3, 308–9 48b 163, 329 49b8 125 Euthphr. general 156, 162, 305–7 5d 137, 157 5d6–9 63 5e–6a 130

5e9–d1 63 6d 137, 157 6e 137, 157 10a–11b 62 11a 137, 157 11b–16a 62–72 11e2–4 62 11e4–5 62 11e6–12a2 62 12a3–c9 62 12c10–d3 63 12d1–3 71 12e5–8 63, 64 12e7–8 64 12e9–13d3 65 13d4–8 65 13d9–e5 65 13e6–14a10 65 14a11–b7 65 14b2–5 68 14b7–c6 66 14c5–6 66, 68 14d1–2 66, 68 14d6–7 65 14d9–e7 66 14e2–4 68 15a5–6 66 15a9–10 66 15b1–c10 66 Euthyd. general 278, 310 278–80 160 Gorg. general 278 467–8 227 473–5 164 473a–475c 308 474e–475e 49–50 476e–477e 50 482d–e 164 483c–d 127 483c–484c 154 483e–484b 128 484a6–b1 128 484b–c 130 494–7 243 494a–c 246 500c 150, 198 503e–504d 309 504–5 163 504–8 227 504a–d 147 507a–c 163

index of passages cited 361 507b 64 508e–509a 303 509e 164 Hp. Ma. 286c–e 306 Lach. general 68–9, 72, 148, 156, 162, 275, 277–8, 307 181b 297 192b1–2 142 197a–b 177 197b 286 Lysis general 156 Leg. general 179 630a–b 177 631c 176 635a–c 176 636d–e 179 653b6 176 660e–663d 179 663b–d 179 696b–e 177 734b 175 735b–736c 174 793a–c 131 816a 162 841b 131 860–3 175 890a 129 903b–904b 67 963 177 Meno general 68–9, 156, 161, 275, 277, 307 70a 138 71b 138 71c 142 71e–72a 140 72a–d 140 72c 137–8, 156 73a–c 140 73e–74a 153 74a 140 74b 141, 149 75b10–11 141 75b11–c1 141 75c 141 76a7 142, 158 76c 142 76d4–5 142, 149, 158 76e3–4 142 77–8 145 77b–78b 50–4, 61 77c 225

77c–78a 146 77c–78b 225 77c1–2 145 78–9 70 78b4–6 145 87c–89a 143, 158 88a 140 88a–c 177–8 89c–94e 303 93e–94a 337 99b–c 143, 158 100a 143, 158 Phd. general 278, 298 60e–61b 305 61d–e 78 64a4–5 235 64c–69e 175, 279 65e–66a 329 66b–c 175 66c 235 67a–b 235 68b–c 175 68b–69c 175 79d 231 80a 175 80c–84b 279 81b 175 81b–83d 20–1 83a–c 235 83d 175, 235 85b5–c1 75 85e–86d 73 85e3–86b5 73 85e5–6 75 86b6–c2 73 86b9 77 86c6–7 76–7 88b6–7 78 88c 73 91e–92e 79 92d2 76 92e4–93a2 81 92e4–93a10 81 93a4–5 81 93a6–7 81 93a8–9 81 93a11–12 81 93a14–b2 82 93b4–6 82 93b8–c2 82 93c3–10 82

362 index of passages cited Pl. (continued) 93d1–5 82, 83 93d4 83 93d6–8 83 93d9–11 84 93d12–e2 84 93e4–5 84 93e7–94a10 84 94b4 81 94b4–c1 87 94c3–e6 87 94e8–95a2 87 95a2 81 115c–e 338 118a 305 Phaedr. general 87–8, 170 237b7–c1 21 253d6–e1 234 253e4–5 234 254a–255a 234 254e6–7 235 254e7 234 256b1–3 234 256c 234 277a 336 279b 329 Phil. 31–2 243 44a–b 250 51b–52b 246 51e–52a 250 53c–54e 242 Pol. 273a–d 236 293c–d 174 306 177 Prot. general 68–9, 72, 148, 164–5, 199, 307, 310–11 309c–d 329 310b–c 338 313a 160 319d–320b 337 320c–322d 152 322a–323a 200 324d–326e 153 324d–328a 337 327b 152 328a–c 275 329c 153 329d–e 266, 274 330–1 71 337c–d 154 339e 273

349b 140 349d 177, 266, 274, 281 349e–350c 266 349e–351b 283 350c–351b 266 351a–b 286 351b 267, 271, 273 351b–e 274 351b3 266 351c1–2 267 351c2–3 267 351c4 268, 271 351c4–6 267 351c7–d2 268 351e2–3 268 351e5–7 268 352a–357e 281 352a6–8 269 352b3–c2 269 352c1–2 135 352c2–d3 269 352d–e 225 352d1–2 266–7 352d4–e2 269 352d5 269 352e5–353a2 269 353a7–8 269 353c–e 270 353c–357e 225 353c1–4 270 353c4–9 270 354a–c 270 354e–355d 270 355c–d 146 356a–c 49, 226 357c–d 270 358 145 358a 271 358a–e 281 358b–d 225 358b–e 48–9 358b6–7 273 358b6–c3 274 358c–d 146, 272 358c6–d2 274 358c6–d4 272–3, 281 358d2–4 282 358d6–7 281 358e 281 359–60 292–3 359a 281

index of passages cited 363 359d 293 360d 148, 267 360d–e 266 360d1–2 274 360e 281 360e1–5 274 361a–b 266 361b 69, 281 361b1–2 267 Rep. general 87–8, 278 328b 236 341a 128 343b–344c 128, 154 344e 150 348b–e 128, 154 352d 150, 198 354b–c 138 358e–359b 152 359b–360d 199 359c 154, 200 362a 153 368e–369a 166 401a–403c 238 427e 153, 166 433a–434c 166 433e–434a 168 434a–c 168 435a–b 166 435a5–7 167 435e–441c 167 438a 232 439a–c 232 439c–440e 233 439c9–d2 233 441b–c 233 441e5 228 441e6 232 442c10–d1 237 442e–443a 173 443c–444a 167 443d7 169 443e 173 444d–e 237 475e–480a 231 544a5–8 92, 229 548c6–576b10 93 576b11–580c9 92 580c–d 92 580c–583b 229 580c–588a 229 580d–583a 91

580d6–7 228 581a–582a 230 581c7–10 91 581d10–e4 94 582a4–6 93 582c7–9 93 582e8–9 94 583a1–3 94 583a6–11 94 583b–586a 279 583b–588a 230 583c–585a 246 584b 246 585b12–c5 231 585d–e 243 585d5 231 585d7–9 231 586e4–587a1 95, 237 586e6–587a1 232 611a–d 175 611b–d 178 612d–e 330 Symp. general 170 174a 338 187b 162 209e–212a 238 Theages 128d 330 Tht. 149–51 296 149a 295 172a–b 155 176a–b 236 184d 323 185e 329 189d 186 Tim. 53a–b 236 70a–b 317 90 176 90a–d 236 Plut. Col. 1108f 19, 183, 190 1109a–b 24 1109a–e 25–6, 34 1120c ff. 35–6 1121a–b 34, 36 1123b–c 32 Proclus In Rep. II.113.6 318, 322 Rufinus Hist. eccles. II.14.5 332

364 index of passages cited Scholiast on Ep. Hdt. 67 319 Sext. M VII.116–18 189 VII.135–40 15–17, 193 VII.136 190 VII.140 196 VII.190–8 35 VII.203–4 10, 24 VII.210 33, 37 VII.210–16 27 VII.349 317 VII.389 19 VIII.6–7 193 VIII.9 23, 24, 28, 31, 33–4, 37 VIII.63 23 VIII.63–4 24, 32–3 VIII.185 24 VIII.355 23 PH I.10 35 I.213–14 193–4 II.63 17 Simplicius In Phys. 28.9–10 183 42.10–11 188 327.24–6 187 330.14–20 187 Sophocles Antigone 59, 381 ff., 447–57, 481, 663, 847 125–6 Stob. I.41.2 6 II.7.3i 6, 13–14 II.31.38 337 IV.55.10 338 Strabo I, p. 61 13 Suidas s.v. Plato 330 Tertullian Ad nat. I.4.7 331 Apol. XI.15 333 XIV.7 331 XXII.1 332 XXXIX.12 332

XLVI.10 332 De an. 1 331–2 46.9 330 51.2 318 Themistius In Phys. 49.13–16 187 Theod. Cure V.22 317 Theophr. Sens. 16 20 63 17 69 19 Thuc II.37.3 131 IV.86.6 130 V.80 ff. 130 V.89 130 V.104.4 129–30 V.105.1 ff. 129 Xen. Ap. 1 125 4 304 Mem. I.1.16 135 I.4 304 II.1.21–34 338 II.3 135 III.6 135 III.7.6–7 338 III.9.1–5 153 III.9.4 136 III.9.4–9 135 III.9.5 136 III.11 338 IV.3.13 ff. 328, 332 IV.4.19 ff. 131 IV.6 (esp. 1) 135 IV.6.6 136 IV.6.7 136 Symp. II.3 336 II.16–20 334 II.26 334 Zeno of Elea DK 29 B 2 182

General Index Principal discussions are indicated by bold type. Academy 311 Aeschines of Sphettus 299, 336 akrasia 42–61 Aristotle on 134–5, 294 Plato on 48–55, 175–6, 225–8, 269–70 Alcibiades 298–9, 335–8 Alcidamas 131 Allan, D. 209 Anaxagoras 14, 20, 22, 154, 182, 196 Anaxarchus 181 Andrewes, A. 130 Annas, J. 279 Anonymus Iamblichi 152 Antiphon 6, 200, 338 Antisthenes 311, 338 Arcesilaus 311 Archelaus (teacher of Socrates) 154 Archelaus (king of Macedon) 338 Archytas 6 Aristides 336 Aristotle on akrasia 134–5, 294 on deliberation 205–8, 217–22 on dialectic 211–13 on endoxa 210–14, 217 on eudaimonia 205 on habituation 210–12, 216–17, 219 on induction 211, 216 on kin¯esis and energeia 256–8 on pleasure 99–106, 107–20, 240–64 on practical intellect 204–22 on prohairesis 206, 217, 263–4 on the¯oria 221 on virtue of character general 217–22 courage 220, 284–94 s¯ophrosun¯e 209, 260–1 of intellect epist¯em¯e 209 euboulia 209 nous 209, 211, 215–16, 220–2

phron¯esis 205–6, 208–10, 214–22 sophia 209 (See also list of passages cited.) Aristoxenus 78 Aspasia 335 Aspasius 108, 120 atomists on chance and necessity 185–90 epistemology 190–7 ethics and politics 198–203 physical principles 181–5 principle of sufficient reason 183, 187–8 (See also Democritus, Leucippus.) Bailey, C. 1, 21 Barnes, J. 187 Berkeley, G. 110 Bodn´ar, I. 185 Bostock, D. 218, 258–63 Brasidas 130 Brickhouse, T. and Smith, N. 124 Brody, B. 326 Burnet, J. 300 Burnyeat, M. 135 Butler, J. 199 Callicles 127–9, 130, 150, 162–4, 171, 198, 200, 227, 309 Callisto 338 Cephalus and Polemarchus 173 Charmides 297, 236 Chaucer, G. 296 Cicero 24 (See also index of passages cited.) Cooper, J. 211 Critias 297, 336 Cynics 311 Dahl, N. 216–17 Davidson, D. 42–8, 54, 60

366 general index Democrates 198 Democritus on death 316–26 on knowledge 1–22 gn¯om¯e, gn¯esi¯e and skoti¯e 16–18, 193 on pleasure 1–22 on soul 316–17 on well-being 2, 4–5, 6–9, 11, 198–9 (euest¯o and euthumia) (See also atomists, and index of passages cited.) Descartes, R. 323 Diagoras of Melos 335 Diogenes the Cynic 338 Diotimus 21 Dover, K. J. 121–33 Duff, A. 290–2 Dyroff, A. 1, 21 Echecrates 78 Eddington, A. S. 16 Eleatics 14–15, 17, 183 Empedocles 20, 142, 158, 182 Epaminondas 336 Epictetus 311–12 Epicureans 199, 311 on soul and death 323–6 Epicurus on knowledge and perception 23–41 on hallucinations 24, 29–33 Euripides 329, 338 Euthydemus 328 Euthyphro 62–72 passim Farrar, C. 202 Ficino, M. 312 Foucault, M. 312 Galen 14 Gallop, D. 86–8 Glaucon 94, 173, 199–200 Gorgias 131, 142 Gosling, J. and Taylor, C. 258, 263–4, 268, 278–9 Hare, R. M. 55–8, 176 Hegel, G. W. F. 312 Hermesianax of Colophon 335 Hesychius 192 Hippolytus of Rome 328

Hobbes, T. 152 Hohfeld, W. 127 Hume, D. 36, 152, 204, 208, 218 Irwin, T. 61, 151, 162, 210, 211 Islamic view of Socrates 312 Justin 312, 328 Kahn, C. 199, 277, 279–80 Kant, I. 67 Kenny, A. 106, 253 Kierkegaard, S. 312 Konstan, D. 203 Lamb, D. 325–6 law divine 125–6 of nature 127 positive 122–6, 131–2 unwritten 126, 131–2 Leon of Salamis 338 Leontius 233 Leucippus 15, 19, 181, 189, 190 (See also atomists.) Locke, J. 36–7 Lucretius on death 316–26 on anima and animus 319–21 on mortality of soul 321–3 Lysias 336 Mansfeld, J. 203 Meno 137–43 passim Mill, J. S. 123, 171, 199, 229 Morrison D. 225, 227 Moses 328 M¨uller, A. 207, 219 M¨uller, C. 203 Natorp, P. 1–3, 11 nature (phusis) and convention (nomos) 153–5, 179–80, 200, 203 Nausiphanes 181 Nietzsche, F. 312 Odysseus 233 Owen, G. E. L. 264

general index 367 Pericles 131, 159 Phainarete 296 Philolaus 6–7, 76, 78–9 Philoponus 184–5, 189, 203 (See also index of passages cited.) Phocion 336 Photius 6 Plato on akrasia 48–55, 175–6, 225–8, 269–70 on eudaimonia 224–39 on intellect, theoretical and practical 228–9 on pleasure 91–8, 171–2, 243–8, 265–80 on virtue unity of 68–72, 161–2, 175–8, 266–7 cognitive theory of 144–8, 158–61, 165, 307–11 courage 69, 163, 170–1, 177, 179, 281–9 holiness 62–72, 163 justice 62–5, 69–70, 163, 167–9, 170–1, 173–4, 177 s¯ophrosun¯e 69, 163, 170–1, 177, 237 wisdom 170–1, 281–4 comparison of lives, (Rep. IX) 91–8, 171–2, 229–32 (See also Socrates, and index of passages cited.) Plum, F. 326 Polus 164, 227, 308 Protagoras 19, 190, 194 (as character in Prot.) 265–80 passim Pythagoras 328

later influence and reception 311–12, 327–39 Plato’s portrayal of 223–4, 275–7, 295 Socratic literature 298–301 thought disavowal of wisdom 301–4 ethical proto-theory 307–11 religion 304–5, 333–4 divine sign 304, 327–34 Socratic definition 136–44, 148, 156–9, 305–7 ‘Socratic fallacy’ 306 Socratic intellectualism 227–8 Socratic paradoxes 134, 136, 144–8, 162, 308 (as character in Plato’s dialogues 48–55, 62–72, 73–88, 91–8, 123–5, 134–49, 156–74, 224–39, 243–51, 265–80, 281–4, 292–4 passim) (See also Plato, Xenophon.) Sophroniscus 295 soul as harmonia 73–90 Democritus on 316–7 Plato on tripartite soul 87–8, 169–71, 232–5, 237 rational and non-rational soul 175–9, 235–9 soul-body dichotomy 175, 235 Stocker, M. 291–2 Stoics 311 Striker, G. 23 Stobaeus 198

Reeve, C. 211, 213 rights 123, 127 legal and non-legal 121–33 natural 127–31 Ross, D. 218–9 Rowe, C. 279 Russell, B. 14 Ryle, G. 107

Taylor, A. E. 300 Taylor, C. 202–3, 267–8, 272–3 Theodorus 335 Theodote 338 Thrasymachus 128, 150, 154, 162–4, 198, 200

Sassi, M. 202 Simmias 73, 76, 78–9, 80 Socrates biography 295–8 death 298, 321

Vlastos, G. 1, 4–11, 20–2, 162, 202–3, 300 Voltaire 312

Urmson, J. O. 107–20, 219

Walter, J. 208 Warren, J. 317–8

368 general index Watson, G. 58 Weiss, R. 275 Williams, B. 237 Wittgenstein, L. 16, 133, 217 Woozley, A. D. 124 Xanthippe 296–7, 335, 337

Xenophon portrayal of Socrates 296–9, 304–5 (See also index of passages cited.) Zeno of Elea 182 Zeno of Sidon (Epicurean) 332 Zeuxis 338