Maieusis: Essays in Ancient Philosophy in Honour of Myles Burnyeat 9781281154705, 1281154709

Citation preview

MAIEUSIS

Photo: By Mary Carruthers

Maieusis Essays on Ancient Philosophy in Honour of Myles Burnyeat Edited by

DOMINIC SCOT T

1

1

Great Clarendon Street, Oxford   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  The Several Contributors 2007

The moral rights of the authors have been asserted Database right Oxford University Press (maker) First published 2007 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 978–0–19–928997–4 1 3 5 7 9 10 8 6 4 2

Editor’s Preface It is a great pleasure to present this collection of papers to mark Myles Burnyeat’s retirement. His professional career began in 1964, one year after he graduated in Classics and Philosophy from King’s College Cambridge, with a Lectureship at University College London. After moving back to Cambridge in 1978, he went on to succeed G. E. L. Owen there as Laurence Professor of Ancient Philosophy in 1984. In 1995 he took up a Senior Research Fellowship at All Souls College Oxford, where he remained until his retirement in 2006. Over the years he has also held a large number of visiting positions, many in North America, but also in France, Germany, Hungary, Japan, and Russia. Burnyeat’s name has been closely associated with Plato’s Theaetetus, on which he wrote a series of articles early on in his career before publishing his longawaited (and modestly entitled) ‘Introduction’ to the translation of the dialogue by M. J. Levett, which instantly became the seminal analysis of the dialogue. More recently he has been producing a series of essays on Plato’s Republic, which will prove to be no less influential. His work on Aristotle—evenly distributed between epistemology, ethics, metaphysics, physics, psychology, and rhetoric—has been equally seminal. He has also been at the forefront of research on Hellenistic philosophy, not only through his own papers but also through the edited collections that he has helped bring to press. But his interests spread well beyond the ancient world. Throughout his career, he has deepened our understanding of the relationship between ancient and modern philosophy; and the number of his reviews on purely contemporary philosophy further testifies to the breadth of his interests. What helps to make his work sui generis is the way in which he combines a mastery of historical, philological, literary, and, above all, philosophical perspectives, often within the same essay. Like others of his generation, Burnyeat has maintained and enhanced the profile of the subject around the world. But, more than anyone else, he has enlarged our sense of what a specialist in ancient philosophy should be like. If we were to live up to the standards he has now set, our work would not just combine the best of contemporary philosophy—its insights as well as its rigour—with a deep sensitivity to ancient texts. We would also be able to move effortlessly between different periods in the history of philosophy; to show how literary readings of texts unite, rather than compete, with philosophical analysis; and to draw with ease upon other specialisms, such as the history of mathematics, that many scholars of ancient philosophy still view with phobia. The intellectual authority so evident in his work has always been tempered by a deep kindness towards his students and colleagues, and I have chosen the title

vi

Editor’s Preface

Maieusis partly because it resonates with the Theaetetus, but also to capture the sense that all of the contributors will share of Myles’s role as teacher and colleague. As a reader and commentator, he has the ability to seize hold of the best in one’s work, perhaps an idea one had ventured only tentatively in a footnote, and to make one bring it centre stage; then to nurture it, pit it against criticism, and so make it stronger. His beneficiaries also include his undergraduate students. At Cambridge, he was one of the few lecturers whose audience actually increased in size as the term went on. The reason was his peculiar ability to lecture in ‘layers’, giving initiates an immediate sense of the importance and depth of ancient texts, while at the same time giving more seasoned students (often experts in the field) nuggets to take away and treasure. One famous example is the series of lectures on the Republic that he gave in the mid-1980s to an audience that included second-year undergraduates, Jonathan Lear, Malcolm Schofield, David Sedley, and Gregory Vlastos. As Schofield testifies in his own contribution to this volume, those lectures have continued to cast their spell many years later. This volume is an indication of the gratitude that so many owe to him. Some of the contributors are former students (Hankinson, Harte, Hobbs, Johansen, Lane, Notomi, and myself). Other were colleagues in London (Sorabji) or Cambridge (Denyer, Lloyd, McCabe, Schofield, Sedley, and Wardy). It was with this latter group that Burnyeat helped make Cambridge a mecca for ancient philosophy in the 1980s and 1990s. A third group includes friends and collaborators from other universities (Barnes, Bobzien, Broadie, Cooper, Nehamas). There are very many others in this category who would have liked to contribute; it is only to prevent the book from bulging at the seams that I have had to be so selective. I have not tried to impose a theme upon the essays in this collection, but have instead allowed the contributors to write on whatever is their passion of the moment. The result, as I hope the reader will agree, is an appropriately wide range of styles and approaches, even though many of them converge (often in pairs) as far as their topic is concerned. Cooper and Lane focus on related themes concerning Socrates and the nature of philosophy; Nehamas and I discuss different aspects of er¯os in the Symposium, Republic, and Phaedrus; Denyer and Sedley come to the rescue of two of the Phaedo’s arguments for immortality; Hobbs and Wardy write on war and warriors in Plato (and beyond); Harte and Schofield both discuss different aspects of the Cave allegory in the Republic. My thanks go to all those at OUP who have helped make this possible, especially Peter Momtchiloff and Jean van Altena, to the anonymous readers, and, above all, to the contributors—not least for the alacrity with which they agreed to write, another tribute to the honorand.

Contents 1. Looking Inside Charmides’ Cloak: Seeing Others and Oneself in Plato’s Charmides Mary Margaret McCabe 2. Socrates and Philosophy as a Way of Life John M. Cooper 3. Virtue as the Love of Knowledge in Plato’s Symposium and Republic Melissa Lane 4. Equal Sticks and Stones David Sedley 5. The Phaedo’s Final Argument Nicholas Denyer 6. Beauty of Body, Nobility of Soul: The Pursuit of Love in Plato’s Symposium Alexander Nehamas 7. Er¯os, Philosophy, and Tyranny Dominic Scott 8. Virgil’s Sacred Duo: Phaedrus’ Symposium Speech and Aeneid IX Robert Wardy 9. Plato on War Angela Hobbs 10. Language in the Cave Verity Harte 11. Metaspeleology Malcolm Schofield 12. Why no Platonistic Ideas of Artefacts? Sarah Broadie 13. Plato on What is Not Noburu Notomi 14. The Soul as an Inner Principle of Change: The Basis of Aristotle’s Psychological Naturalism Thomas Johansen 15. Aristotle’s De Interpretatione 8 is about Ambiguity Susanne Bobzien

1 20 44 68 87

97 136 154 176 195 216 232 254

276 301

viii

Contents

16. Sextan Scepticism Jonathan Barnes 17. The Wife of Philinus, or the Doctors’ Dilemma: Medical Signs and Cases and Non-deductive Inference G. E. R. Lloyd 18. Self-Refutation and the Sorites R. J. Hankinson 19. Ideas Leap Barriers: The Value of Historical Studies to Philosophy Richard Sorabji Myles Burnyeat: Publications General Index Name Index Index Locorum

322

335 351 374 391 403 407 412

Notes on Contributors Jnathan B taught ancient philosophy at the universities of Oxford, Geneva, and Paris, and is a Fellow of the British Academy and of the American Academy of Arts and Sciences. He has published enough pages to satisfy the most demanding bureaucrat, his latest book being on Truth, Etc. (Oxford, 2006). He lives in retirement in deepest France. S B is Professor of Philosophy at Yale University, having previously held a position at Oxford University. She is the author of Determinism and Freedom in Stoic Philosophy (Oxford, 1998) and Die stoische Modallogik (Wuerzburg, 1986), and has published on many aspects both of logic and of freedom and determinism, mainly in ancient philosophy. S B has taught at the University of Edinburgh, the University of Texas (Austin), Yale, Rutgers, Princeton, and most recently at the University of St Andrews. She has mainly published on Aristotle’s Physics and Ethics. She is a Fellow of the British Academy, the American Academy of Arts and Science, and the Royal Society of Edinburgh. J C, Stuart Professor of Philosophy at Princeton University, has taught there since 1981, having previously held positions at Harvard and the University of Pittsburgh. He is the author of Reason and Human Good in Aristotle (Cambridge, Mass 1975), Reason and Emotion (Princeton, 1999), and Knowledge, Nature, and the Good (Princeton, 2004), and editor of Seneca: Moral and Political Essays (Cambridge, 1995) and (with J. F. Procope) Plato: Complete Works (Indianapolis, 1997). N D is a Senior Lecturer in Philosophy and Fellow of Trinity College, Cambridge, and a Senior Lecturer in the Faculty of Classics, University of Cambridge, His most recent books are editions of Plato’s Alcibiades (Cambridge, 2001) and Protagoras (forthcoming), in the Cambridge greek and Latin Classics series. R. J. H was educated at Balliol College, Oxford, and King’s College, Cambridge, he is currently Professor of Philosophy and Classics at the University of Texas at Austin. He has written numerous articles on many aspects of Greek philosophy and science, and is the author of several books, including The Sceptics (London, 1995), Cause and Explanation in Ancient Greek Thought (Oxford, 1998) and Galen on Antecedent Causes (Cambridge, 1998). V H is Professor of Philosophy and Classics at Yale University. SHe is the author of Plato on Parts and Wholes (Oxford, 2002) and of various articles in ancient philosophy. A H is Associate Professor in the Philosophy Department at the University of Warwick. She is the author of Plato and the Hero (Cambridge, 2000) and is currently preparing a translation and commentary of Plato’s Symposium for the Oxford Clarendon Series. She has also given numerous broadcasts on philosophy and the history of ideas for the BBC.

x

Notes on Contributors

T K J is a Fellow of Brasenose Colelge, Oxford, and University lecturer in Ancient Philosophy, University of Oxford. He previously taught at Bristol and Edinburgh. His Ph.D. thesis, supervised by Myles Burnyeat, became Aristotle on the Sense-Organs (Cambridge, 1998). Other publications include Plato’s Natural Philosophy (Cambridge, 2004). M L is Senior Lecturer in the History Faculty and Fellow of King’s College at Cambridge University. Work includes Method and Politics in Plato’s Statesman (Cambridge, 1998), arising from her doctorate in philosophy supervised by Myles Burnyeat and Malcolm Schofield, and Plato’s Progeny: How Plato and Socrates still Captivate the Modern Mind (London, 2001). G University retirement Currently, CB3 9AF.

L is Emeritus Professor of Ancient Philosophy and Science at the of Cambridge where he was Master of darwin College from 1989 until his in 2000. He was knighted for services to the history of thought in 1997. he is based at the Needham Research Institute, 8 Sylvester Road, cambridge

M M MC is Professor of Ancient Philosophy at King’s College London. She writes mostly on Plato (most recently, Plato and his Predecessors: The Dramatisation of Reason (Cambridge), and is the general editor of the series Cambridge Studies in the Dialogues of Plato. She currently holds a Leverhulme Trust Major Research Fellowship (2005–8). A N teaches philosophy and comparative literature at Princeton University. He is the author of books and essays on ancient philosophy, Nietzsche, and the philosophy of art. His most recent work is Only a Promise of Happiness: The Place of Beauty in a World of Art (Princeton, 2007). N N is Associate Professor of Philosophy, Keio University, Japan, having received his M.A. in Philosophy from University of Tokyo, and his Ph.D. in Classics from the University of Cambridge. He is the author of The Unity of Plato’s Sophist: Between the Sophist and the Philosopher (Cambridge, 1999; Japanese translation: Nagoya, 2002), and a number of books and articles in Japanese and English. He is currently working on the Sophists and the Socratics. M S met Myles Burnyeat at F. H. Sandbach’s lectures on postAristotelian philosophy in Easter Term 1963. He has taught most of his life at Cambridge, where he is Professor of Ancient Philosophy. With Myles Burnyeat and Jonathan Barnes he edited Doubt and Dogmatism (Oxford, 1980). His most recent book is Plato: Political Philosophy (Oxford, 2006). D S is Professor of Philosophy at the University of Virginia and an Emeritus Fellow of Clare College, Cambridge. Between 1989 and 2007 he taught in the Philosophy Faculty at Cambridge University. He is the author of Recollection and Experience: Plato’s Theory of Learning and its Successors (cambridge, 1995) and Plato’s Meno (Cambridge, 2006). D S is Laurence Professor of Ancient Philosophy at the University of Cambridge. His books include The Hellenistic Philosophers (Cambridge, 1987, with

Notes on Contributors

xi

A. A. Long) and, most recently, Creationism and its Critics in Antiquity (Berkeley, 2007), based on his 2004 Sather Lectures. He edited Oxford Studies in Ancient Philosophy from 1998 to 2007. R S is a Fellow of Wolfson College, Oxford, and of King’s College, London, and was Director of the Institute of Classical studies, London, and Gresham Professor of Rhetoric. Besides editing seventy volumes of Ancient Commentators on Aristotle, he is author of a trilogy of books on cause, time, space and matter, and a trilogy on animals, emotions and individuality, life and death, along with a co-edited book on the ethics of war in different traditions. R W is the author of monographs on ancient Western philosophy and early modern Chinese culture and language. He teaches classics and philosophy at St Catharine’s College, Cambridge.

This page intentionally left blank

1 Looking Inside Charmides’ Cloak: Seeing Others and Oneself in Plato’s Charmides Mary Margaret McCabe

1 . C R I T I A S ’ P RO P O S A L The Charmides offers a series of examples of the untranslatable Greek virtue s¯ophrosun¯e (so I shall refrain from translating it: some offer ‘temperance’ or ‘self-control’), and, among others, an influential definition of it, as knowledge of oneself. One of the reasons, it seems, why self-knowledge comes to grief as an account of a virtue—as it does by the end of the dialogue—is the repeated assimilation of knowledge and self-knowledge to perception and self-perception. This assimilation, I shall claim, is both deliberate and tricky: I hope to offer some explanation of its point. After a good deal of horseplay—some of it apparently pretty smutty—Socrates embarks on a discussion of s¯ophrosun¯e, first with young Charmides, and then with his mentor, Critias.¹ Thereafter almost half of the dialogue is taken up with the refutation of Critias’ passionate declaration that s¯ophrosun¯e is knowing oneself (164c–d, 165b4). That is glossed as the claim that s¯ophrosun¯e is the knowledge of itself ; and that is what comes under critical scrutiny (165c7). (I shall call the first ‘knowledge of the self ’, the second ‘self-knowledge’). Socrates It is an honour, a pleasure, and a matter of considerable trepidation to offer this essay to Myles, an incomparable reader of Plato. I am grateful to various audiences (at King’s College London, at the Moral Sciences Club in Cambridge, and at Rice University) and to various individuals (especially Simon Blackburn, Nick Denyer, Hilary Mackie, Hugh Mellor, Vasilis Politis, Malcolm Schofield, Dominic Scott, Frisbee Sheffield, Nick Smith, Harvey Yunis) for their comments, critical and otherwise. My thanks, especially, to my colleagues Verity Harte and Peter Adamson for a discussion of the Charmides which has gone on for several years. I am also extremely grateful to the Leverhulme Trust, who funded the Major Research Fellowship during the tenure of which I wrote this paper. ¹ The dramatic tone of the dialogue is intensified by the record of the (extreme) political activities of this Critias. Critias was a member of the Thirty in 404/3, against whom Socrates’ civil disobedience is recorded at Ap. 32c; and Charmides was at least implicated in their activities. See Nails 2002: 108–13, 90–4.

2

Mary Margaret McCabe

wonders first whether self-knowledge is possible; and second whether, even if it is possible, it is any good to us (if not, it can hardly, he suggests, count as a virtue). I shall ignore the question whether it is any good, and consider instead the arguments which are designed to show that self-knowledge is impossible or, if not, thoroughly odd. Oddity seems, at first, to be to its credit: when Socrates objects that selfknowledge is unlike any other knowledge in not having a subject-matter other than itself, Critias responds that this is exactly its distinguishing mark. Critias insists that s¯ophrosun¯e is the only knowledge which is of itself and all the other knowledges² (166c2–3, 166e5–6). For: I. s¯ophrosun¯e is knowledge of itself (it is reflexive) and of other knowledges (it is second-order). II. s¯ophrosun¯e is not of anything else (other than itself and the other knowledges) (it is just second-order). It follows from these two theses that there is a knowledge which is not of anything else but of itself and of the other knowledges, and of ignorance too (167b11–c2). In what follows next the focus of attention is on the relation between knowledge and its objects:³ on the ‘of’ relation. What is it for knowledge to be ‘of ’ its object? And how does that reflect on the possibility of either a reflexive cognitive state or a higher-order one? Socrates offers a triad of arguments against Critias’ proposal: I shall call them, collectively, the Relations Argument: 1. (167c–168a) Reflexivity is apparently impossible in other cases (similar to knowledge): perception, desire, and belief. 2. (168b–d) The ‘of ’ relation expressed by the thought ‘this knowledge has the power (dunamis)⁴ of being of something’ is (in some central cases: larger, double) irreflexive. 3. (168d–169a) Thence it seems implausible that perception should perceive itself (168e9–10); perception may only perceive its own special objects.⁵ ² I translate epist¯em¯e as ‘knowledge’, and tolerate rebarbative expressions like ‘the other knowledges’, in order not to beg the question from the outset about the nature of knowledge (is it, e.g., piecemeal or wholesale? Is it craft-like or virtue-like, or both?). On this, and on whether we should understand knowledge as ‘understanding’, see of course Burnyeat 1981; Nehamas 2004; and also Lyons 1963 and Bailey 2006. ³ I use ‘object’ to refer to the intentional object of knowledge, belief, etc.; what the ontological status of such objects is, I leave open. ⁴ This expression (coupled as it later is with ousia) may seem to anticipate the argument of Resp. 476–480, especially 477c–e. In what follows I argue that the dunamis/ousia characterization of the relation between knowledge and its objects is rendered problematic in the Charmides; I suspend judgement here on how that may affect our reading of Resp. V. ⁵ Aristotle is undoubtedly reading this passage at De an. 425b 12–20: notice the close correlation between Chrm. 168d9–e1 and De an. 425b 17–20.

Looking Inside Charmides’ Cloak

3

The first phase of the argument assimilates perception to some other states⁶ such as emotion and belief—states which we might readily describe as psychological.⁷ The second assimilates it to other, non-psychological relations. Perception occupies a pivotal role, therefore: for it allows Socrates to claim that the features of non-psychological relations also characterize cognition. This turns out to damage the possibility of both reflexive and higher-order cognitive states. I shall wonder just how, in the economy of the dialogue, this conclusion forces us to modify our assumptions: about perception, about self-perception, and about self-knowledge. 2 . T H E R E L AT I O N S A RG U M E N T Consider the Relations Argument in more detail. Critias claims that there is a knowledge which is of itself and the other knowledges, but not of the objects of the other knowledges. In the first phase of the argument, Socrates gets Critias to concede that perceptions must be at least of their special objects: that is, sight must be at least of colour, hearing at least of sound. So, in general, no perception could be only of itself and the other perceptions. The same, Socrates argues, is true for other psychological states: no desire, for example, is of itself and of the other desires, but not of pleasure; and the same for wishes, love, and fear. Cognitive states, too: surely there is no belief which is of itself and the other beliefs, but not of the things the other beliefs believe? (168a3–4).

So it is absurd to suppose that there is a knowledge which is such as to be of no learning but it is of itself and the other knowledges (168a6–8).

Socrates’ argument so far claims that Critias’ original restrictive thesis (that there is some knowledge which is only of itself and the other knowledges) is implausible: all these psychological states must be at least of their special objects (whatever those may be). The next phases of the Relations Argument, however, go further. ⁶ It is unclear whether we should speak here of states or of events. The argument focuses on the relation between psychological subject and object, which suggests that we should be thinking of psychological states; but the resulting view of perception may well be one that sees it in terms of coming to be in that relation, and so as an event rather than a state (compare the theory of perception as an event at Tht. 154 ff.). In what follows I speak of ‘states’ except where the terminology of ‘events’ is obviously more felicitous; since this occurs at 167, where the event is a second-order seeing, I think the dual terminology does not interfere with the point I make. ⁷ In the sequel, it becomes unclear just how we should classify the difference between the psychological and the non-psychological; I use the expression ‘psychological’ in what follows immediately with a health warning.

4

Mary Margaret McCabe For the second phase Critias agrees that:

Knowledge itself ⁸ is of something, and it has some such power (dunamis) as this, to be of something (168b2–3)

Socrates takes this claim (about the ‘power to be of something’) to capture an asymmetric relation. Consequently, he takes licence to construe it on the model of some non-psychological relations, notably larger than, double. Suppose that there is something (Gerald the giraffe, let us say) which is larger than itself and the other larger things, but not larger than the things than which the other larger things are larger. If Gerald is larger than himself, himself will be smaller than him. If, further, he is larger than the other larger things, but not larger than the things they are larger than, he will be both larger and not larger than them.⁹ Or double: suppose something is double itself and the other doubles, but not double the things of which they are double. Then it will be its own half; and its half will be double its own halves, as well.¹⁰ ⁸ This is reading aut¯e, as proposed by Shorey (1907), not haut¯e as given by Burnet. Van der Ben (1985: 55), rightly, I think, glosses ‘in general’; and compares Resp. 438c7, Prm. 134a3. ⁹ Socrates asks: ‘So if we were to find something larger, which is a) larger in relation to itself and the other larger things, but b) not larger in relation to the things in relation to which the other larges are larger, then this would surely happen to it, that c) if it is larger than itself it will be smaller than itself, too. Is that not so?’ (168b10–c2). The argument can be construed the easy way: if Gerald is larger than himself (we should add, as Aristotle would exhort, the qualifications that avoid the logical difficulties: Metaph. 1005b 20–2), then the himself than whom he is larger is smaller than himself: so he is both larger and smaller than himself. This moves directly from (a) to (c). Or it can be construed the hard way: if Gerald is larger than himself and the other larger things, but not larger than the things than which the other larger things are large, then the other larges will be larger than Gerald; so he will be both larger than them and smaller than them, although they are (ex hypothesi) smaller than him. This moves from (a) via (b) to (c), and takes (c) to be demonstrated both by reflexive considerations and by considerations of transitivity and its failure. The easy way leaves (b) out of consideration altogether; why, if not to offer us the hard way, is (b) still in the text? Since it is, contrariwise, two points are being made about these relations at once: about their reflexivity and about their transitivity. In what follows, I shall suggest that both characters of relations are important in the discussion of self-perception. ¹⁰ Socrates’ question (168c4–7) might be construed in two slightly different ways: ‘And so if something is double all the other doubles and itself, then I suppose it will be double in relation to both itself, its half, and the others. Double, after all, is (exactly) of a half.’ Or: ‘And so if something is double all the other doubles and itself, then I suppose it will be double while both itself and the others are halves. Double, after all, is (exactly) of a half.’ Both the double’s relation with itself and its relations with ‘the others’ are expressly included here: so we might expect the resulting puzzle to turn once again on both reflexivity and transitivity. In the case of double, unlike the case of larger, the relation between this double and all the other doubles should indeed be intransitive (if a double has a half, the half of its half will not be half of the original double). But then the puzzle here seems quite different from that of the preceding move about larger: the preceding puzzle works because larger is transitive, this one because double is not. The passage is highly elliptical, however, so that it is unclear just where the puzzle is supposed to arise. Here is an attempt: at least, if something is double itself, itself will be its half, and so both half and double at once. Then also if something is its own half, it, qua half, will also stand in the same relation to its double as its own halves stand in, to the same thing, the double. But any double is exactly double its half; so neither the relation between the double and itself, nor the relation between the double and the other doubles, makes sense.

Looking Inside Charmides’ Cloak

5

Socrates concludes: ‘Whatever has its own power (dunamis) towards itself will also have the being (ousia) towards which the power (dunamis) is’ (168d1–2). This both summarizes the argument so far, generalized across all relations, and provides the springboard for the third phase, on perception. But what does it mean? It manifests, perhaps, two assumptions of the argument. First, it reiterates the supposition that the relations under consideration are asymmetric ones: the power of one relatum is correlated to (‘of’) the being of the other. Second, in such cases there is not only something we can say about the relatum that has the power, there is also something in particular that we can say about the other relatum: that it has the being correlate to the power. Problems then arise when—as here—the relatum that has the power is identical to the relatum that has the being: for then one thing will display opposite properties, the correlate features (the power and the being) of an asymmetric relation. Notice, however, that more has happened here. First of all, Socrates moves easily from claims about perception, emotion, and cognition, in the first phase, to claims about larger and double, in the second, and supposes that we can reasonably generalize across them. One might complain that the generalization is spurious: the relations in question—marked by ‘of’ or ‘than’—are only apparently similar, their differences masked by an over-generous use of the Greek genitive.¹¹ But that complaint might be theory-laden. For it may rest on the supposition that the examples of the first phase are quite different in kind, and so in their relations, from the second—because the first examples are psychological, the second not. Perhaps Socrates, by contrast, makes here no unargued distinction between psychological or mental states and (merely) physical ones; instead, he offers considerations about some relations (e.g. ‘larger than’) as a way of accounting for others (e.g. ‘perceives’). (It is we who give the latter a problematic metaphysical status, by calling them ‘psychological’ or ‘mental’, and then wondering what sort of property those terms describe). Second, this phase of the argument is carefully structured to turn on two features of the (asymmetric) relations under scrutiny: on whether they are reflexive, and on whether they are transitive (where, intuitively, some are, e.g., larger; and some are not, e.g., double). Critias’ proposal was that there is a knowledge which is both of itself and of the other knowledges, but not of the objects of the other knowledges. Translating this into talk of relations, Socrates takes Critias to be committed to saying that self-knowledge is an asymmetric relation which is both reflexive (it applies to itself) and intransitive (if the knowledge is of the other knowledges, it is not of whatever the other knowledges are of ).¹² But this, Socrates suggests, is absurd, on both counts. Both because it ¹¹ The objective genitive in ‘sight is of colour’ and the comparative genitive in ‘the giraffe is larger than the gerbil’. ¹² So, if a is knowledge of b; and b is knowledge of c; still, a is not knowledge of c. If these relations between knowledges are understood as ordered, with a the highest, this means that the

6

Mary Margaret McCabe

insists on reflexivity and because it denies transitivity,¹³ Critias’ original account of self-knowledge fails. In the third phase of the argument, the general principle about relations is applied to perception; but now the focus has narrowed, leaving transitivity out of view to consider the issue of reflexivity alone. Socrates makes a strong claim about perception: Hearing is of nothing but of sound. (168d3) Sight sees nothing that is not coloured. (168d10).

So perceptions have special objects, to which they are exactly correlated: hearing is just of sound, sight just of colour. Given that a perception occurs, on this view, its special object is both necessary and sufficient for its content: the being of its object determines the power (the nature) of the perception itself. But this is a puzzle for self-perception: if sight sees itself, it must see itself coloured; and if sight is not coloured, it cannot see itself. What exactly is the puzzle? Perhaps it turns on the special objects; perhaps Socrates simply needs an empirical claim (as a matter of fact, sight is not coloured¹⁴ ) to generate his conclusion. But that seems just too quick for all that has happened in the argument; this conclusion was available as soon as the first phase said that the special objects are necessary for perception. Instead, I suggest, the model of sight as a relation persists, and the puzzle derives from the principle that the power (in this case, sight) and its object (colour) are asymmetrically related. Then it is a metaphysical mistake to say that sight sees itself, since that would be to conflate distinct relata: the power with its object, hearing with a sound, and sight with colour. This gives Socrates his conclusions, that a. Relations such as size cannot be reflexive (168e5–6). b. Reflexive powers (dunameis) are implausible in cases such as perception and motion¹⁵ (but perhaps not in other cases) (168e9–169a1). Where does this leave us? More significantly, where does this leave self-knowledge? If the argument goes through, it concludes that none of these relations—whether we would call them psychological or not—is reflexive; so Critias’ content of the first-order knowledge is not included in the content of any higher-order one. Socrates later goes to considerable lengths to point out the absurdities of this thought, 170a ff. ¹³ That transitivity, as well as reflexivity, is at issue, is brought out by the detailed treatment of the two relations larger and double, the first of which is (intuitively) transitive, but disallowed as such by Critias’ proposal; the second of which must be intransitive, but whose intransitivity seems not to protect it from absurdity either; see above, n. 9. ¹⁴ Once again this dimension of the argument bears comparison with its Aristotelian commentary in De an. III. 2; it is, of course, a vexed question whether Aristotle thinks that sight is, or is not, literally coloured. See, notably, Burnyeat 1992; Sorabji 1992. ¹⁵ The inclusion of motion at this point is rather a surprise, we might think. It brings out, however, that perception is construed in this argument as a brute causal relation; below I suggest that this is the refutand.

Looking Inside Charmides’ Cloak

7

thesis fails. Why might we care if it does? Critias has proposed that self-knowledge is both reflexive and higher-order, but not transitive: that is, it does not have as its objects both first-order knowledge and the objects of that first-order knowledge, but just the first-order knowledge, simpliciter. In developing his counter-argument, Socrates focuses both on reflexivity and on higher-order thought; but he denies Critias’ thesis on the basis of reflexivity alone. This leaves open two questions, which may well be distinct: i. Does the denial of self-knowledge imply that there can be no knowledge of the self? (Why should we care?) ii. Does the denial of self-knowledge imply that higher-order knowledge (or, in general, higher-order thought) is impossible too? (Why should we care?) I shall suggest that the real cost of the refutation of Critias is that knowledge is denied its higher-order dimension (so the question of just how all these relations might be transitive is important). I shall further suggest that the argument is set out in the elaborate way that it is, and in the context it is, to show how we might avoid paying this price: and why it matters that we do avoid it. En passant, this will deliver a conclusion about how we should read the dialogue: by taking the details of the dramatic context into account when we try to figure out just what the arguments say. 3 . T H E S T RU C T U R E O F A R E BU T TA L Socrates has argued that some ordinary relations (large, double), and perception, emotion, and cognition are asymmetric and therefore irreflexive, fatally to Critias’ account. And he has suggested that there is something problematic about Critias’ claim that these same relations are all intransitive: especially this will mean that the content of higher-order knowledge will be just lower-order knowledge, but not the content of the lower-order knowledge.¹⁶ Is anything about self-knowledge to be salvaged from Critias’ proposal? Three broad responses suggest themselves: 1. Capitulation: the argument works, and so much the worse for self-knowledge and higher-order thought. 2. The argument goes through for relations such as large and half; perception is such a relation, so for perception as well. The argument does not go through for cognition (emotion is another matter, tricky in all sorts of ways; I shall not discuss it in detail here). 3. The argument goes through for relations such as large and half, but not for psychological relations/states. ¹⁶ This is the main focus of attention in later arguments, 169d–171c, which I do not discuss in detail here.

8

Mary Margaret McCabe

Well, perhaps capitulation is just bad news for Critias, whose thesis about s¯ophrosun¯e is made to collapse. But it would not be so good for Socrates, either, to lose the possibility of higher-order thought. When he reflects on how the discussion is to be conducted, Socrates insists on the importance—both argumentative and ethical—of questioning whether we know what we think we know; and he characterizes this as inquiring into oneself (tracking oneself down): How could you think that—however much I test you—I do so for any other reason than I would give for tracking myself down, asking what I am saying, fearing lest I may escape my own notice thinking that I know something, but not knowing it. And so now I claim to be doing just this: looking at the argument most of all for my own sake, and perhaps also for that of my friends. Or don’t you think this is a common good for just about everyone, that each of the things that are should become clear, as it is? (166c–d)

In Socrates’ version of Critias’ thesis, likewise, this inquiry into oneself is connected with self-knowledge, and the ability to test others: but now with the apparent restriction, that self-knowledge is just knowing what one does and does not know. So the person who is s¯ophr¯on alone will know himself and will be able to test what he knows and what he does not, and will have the power to look for what others might know and think that they know, and again what they think they know, but don’t; no-one else can do this. And being s¯ophr¯on and s¯ophrosun¯e and knowing oneself are just this: knowing what one knows and doesn’t know. Is this what you are saying? (167a)

Compare and contrast the two passages. Socrates’ version is heavily ordered—he inquires into whether he knows what he thinks he knows —and unrestricted as to the content of any one of these orders. Critias’ version, on the contrary, insists that higher-order thinking is intransitive (we are allowed to know that we know something, but not its content), and so turns out to be either impossible or absurd. Socrates is committed to showing, therefore, both how we know what we know and why it might matter. Indeed, we ourselves are implicated: if Critias’ version prevails, what are we doing thinking about what Plato thinks Socrates might think? Could we ever know what it is that Socrates might think, if higher-order thought is blocked in Critias’ way? And would we care?

4 . R E T H I N K I N G PE RC E P T I O N The argument needs to be resolved, then, either in favour of (2) or in favour of (3): whatever we may say about reflexivity, we had better save something of higher-order thought, of reflection. To do this, we need to think again about perception (as I suggested, the pivot of the argument) and how it works. On (2), perceptions fail to be reflexive just because the relation between perceiver and perceived conforms to the model of the non-psychological relations. It is, as I shall call it without prejudice, brutish: an exclusive relation between

Looking Inside Charmides’ Cloak

9

perceiver and object with no room for mediation or indirection. On such an account, there will be scope for claiming that perception is quite different in kind from cognition, so that perception falls, where cognition does not, to the Relations Argument. On (3), the divide may come in a different place: between what is psychological and what is not. On this account, perception would have higher-order capacity; it would be, as I shall call it, civilized, since its closest affinity is with cognition and the other psychological states. It remains to be seen whether the civilization of perception could resist the Relations Argument, and rescue any account of self-knowledge, or of self-perception. Should we opt for response (2) to the Relations Argument (and brutish perception) or for response (3) (and civilized perception)? The Relations Argument begins: ‘Well, see how odd is what we are trying to say, my friend; for if you look for the same thing in other cases, you will believe, I think, that this is impossible.’ ‘How is this? What cases?’ ‘Cases like these: for consider whether you believe that there is a sight which is not a sight of the things other sights are sights of, but is sight of itself and the other sights and likewise of the non-sights; and although it is a sight, it sees no colour, but sees itself and the other sights. Do you believe there is any such thing?’ (167c–d; my italics)

Socrates says to Critias: ‘see how odd is what we are trying to say’ (167c4); and follows up his remark with a discussion, in the first instance, of sight.¹⁷ Seeing, here, is both first-order (we are talking about seeing colours) and higher-order (it has as its content something about sight). At the end of the Relations Argument, Socrates says: ‘And sight, I suppose, my excellent friend, if it sees itself, must have some colour; for something colourless, sight could never see.’ ‘It could not.’ ‘So do you see, Critias, that for the things we have gone through, some seem to us to be altogether impossible; and others are hardly credible as having their own power in relation to themselves? For largeness and multitude and such things are altogether impossible, aren’t they?’ (168d–e; my italics)

Here too, the discussion of sight in the argument is juxtaposed to an exhortation to see in the frame dialogue.¹⁸ The sight that is under discussion sees colours; the sight urged in the frame has the conclusion of the argument itself in its scope. Does this double use of sight have any significance? The framing of the discussion of sight with instances of seeing as a second-order psychological event¹⁹ is, first of all, ostentatious, placed strikingly at the opening ¹⁷ id¯e at 167c4 appears emphatically at the beginning of the sentence. ¹⁸ horas at 168e3 runs directly on from id¯ei at e1, where the latter is part of a move within the argument. ¹⁹ See above n. 5; that this is an event is indicated by the imperative at 167c4.

10

Mary Margaret McCabe

and the close of an argument about seeing. It is troubling, too: for the framed argument undermines the possibility of second-order seeing. Are we to say this is a mere literary accident (even were we to have an account to give of what that might be)? Are we to say that the frame is outside the argument, irrelevant to it, merely the gilding around the real picture? The onus probandi, perhaps, is the other way about: without a methodical way of demarcating the gilded frame from the arguments within, and some grounds for supposing it irrelevant, we should prefer to start by supposing that all of the dialogue somehow or other counts towards its point. We should not, therefore, dismiss out of hand the double use of sight here. In that case, it seems that the dialogue itself directs us towards an account of perception as civilized; and wards off the brute. For the simple argument which Socrates seemed to offer (that since sight is not coloured, it cannot see itself ) is undermined by a conception of sight broad enough to include higher-order attitudes. And in that case the brutish view—which denies the possibility of higher-order perception by insisting on the brute relation between perception and its object—is ruled out. Well, an objection might run, perhaps the double use of sight reveals only an ambiguity in ‘sight’, or in ‘perception’ in general: sometimes it is used narrowly of sense perception, when it may be brutish, and sometimes of perceptions of a more general, civilized, second-order kind. This passage, then, is but an instance of that general contrast; if the seeing of the frame is civilized, the seeing attacked in the argument is not. (And in this case the relation between frame and framed has little impact on the arguments themselves.) But this objection may assume too much. For it assumes that Plato must begin with, must indeed already have fixed on, a brutish account of perception (sense perception), and that he extends it to a metaphorical, civilized use without thinking that the structure of the latter has any bearing on the nature of the former. On that assumption this passage (again) is a literary accident (and response (2) to the Relations Argument is the right one). We need not suppose, however, that ‘perception’ here is equivocated; nor does the composition of the passage, and its ostentatious double use of sight, encourage us to do so (there is no literary accident here). Perhaps, instead, Plato starts from a broad conception of perception, whose nature and delineations he is here trying to make clear. His careful composition, then, is part of his examination of the difference between a civilized and a brutish view of perception. Thus the frame invites us to think of perception in general, and sight in particular, as capable of being higher-order, capable of having in its scope both the content of lower-order seeings and those seeings themselves. The structure of the Relations Argument then allows him to suggest that in fact perception may be civilized right through; all perceptions, on such a view, will be broadly cognitive, whatever their order: ‘perception’ is not equivocated at all. If that is what the Relations Argument offers, then the argument must be understood in terms of the third option I

Looking Inside Charmides’ Cloak

11

offered above; and in that case the crucial contrast here is not between the brute and the civilized, but between those states that are psychological and those that are not. Quite right, we of the twenty-first century might think: for the discussion of these higher-order states—we might think—is exactly appropriate for cases—perception, desire, knowledge—where we want to insist that there is something that it is like for me to perceive.²⁰ In just such cases, that is, the higher-order state (perceiving that we perceive) is designed to capture the phenomenon of consciousness, awareness, the state which is peculiarly a part of human psychology.²¹ Is this what Plato has in mind? I think not: for these arguments about self-perception and self-knowledge do not give him an account of consciousness, but instead a very different account of reflection.²² 5 . T WO D R A M AT I C I N C I D E N TS Go back, once again, to the gilded frame: to those parts of the dialogue that are often supposed to be irrelevant to the arguments contained therein, and to the elaborate horseplay of the introduction. Two episodes are striking, once you start to read the dialogue with, as I shall say, foresight (that is, at least twice).²³ They are striking enough, moreover, to ward off the conclusion that the frame drama has no philosophical bearing on the arguments within.

Incident i. Charmides’ introspection: The dialogue is named after the beautiful young man, Charmides, who is apparently s¯ophro¯n in extraordinary measure—and under several different definitions. He is well-born and respectable (157e ff.). He is modest and unassuming (158c–d). And he comes recommended (if such that be) by his mentor Critias, who insists that he is the most s¯ophr¯on of his generation (157d). Socrates seems impressed, and invites Charmides to supply him with an account of s¯ophrosun¯e for discussion. For: ‘This is how I think it best to look at it,’ I said. ‘For it is clear that if s¯ophrosun¯e is present to you, you will have some belief about it. For it is necessary, I suppose, that if it is in ²⁰ Nagel 1974. ²¹ This, e.g., is often taken to be the issue in the parallel passage of De anima; see e.g. Kosman 1973 on ‘perceiving that we perceive’ as a description of consciousness. ²² It would be a mistake, of course, to think that Plato’s lack of interest in the issues which provoke contemporary discussions of consciousness (issues often dominated by the problem of scepticism) render what he does say about higher-order knowledge and perception redundant. It is a broader question than the one that occupies me here exactly what he supposes higher-order knowledge does for us; this is the question that the last part of the Charmides seeks to address. ²³ This is not the place to defend this approach in detail; it requires merely that the dialogues were written carefully, to be read and reread, so that resonances whose significance appears only later in a dialogue should not be disregarded.

12

Mary Margaret McCabe

you, it will present you with some perception²⁴ from which some belief might be yours about it, what s¯ophrosun¯e is and what sort of thing it is. Do you not think so?’ ‘I do think so,’ he said. (158e–159a)

The same motif reappears a bit later: ‘So once again, Charmides,’ I said, ‘turn your mind to the matter even more, and look into yourself, thinking about what s¯ophrosun¯e, by being present to you, makes you be like; and what it is like, that it makes you so; working all these things out, tell me well and bravely what it appears to you to be.’ (160d–e)

With foresight, this is all pretty odd. First of all, is the point a general one (anything that is ‘present to us’ presents us with a perception from which we get an opinion) or particular (when s¯ophrosun¯e is present to us, it presents us with a perception from which we get an opinion), or somewhere in between (some things, virtues for example, when present in us, do this)? If it is general, why would we think it true? We might, of course, think that my experiences are presented to me immediately; would we say the same of any other inner states? Would Socrates say so, the Socrates who later worries lest his own ignorance might escape his notice? Or is his thought that all my moral states are transparent to me? If I am wicked, will my wickedness be ineluctably borne in on me? Or is s¯ophrosun¯e in particular bound to present itself to its possessor? If s¯ophrosun¯e is self-knowledge, this might be plausible—so long as this is not Critias’ model of self-knowledge. For that takes it that our psychological states are each related to their special objects intimately and exclusively: how then could Charmides’ inner inspection form the basis for a belief?

Incident ii. Looking inside Charmides’ cloak Earlier in the dialogue, as Charmides approaches and tries to sit down, everyone who is already seated tries to make a space for him on the bench beside themselves; so one person has to stand up, and another falls off the end of the bench. Socrates seems to rise above the crude slapstick, remarking rather sourly that he will be impressed just so long as Charmides has a beautiful soul (154e). But now when Charmides approaches: Then, my friend, I was straightaway at a loss, and my earlier boldness, in supposing that it would be easy to have a conversation with him, was cut off; for when Critias said that it ²⁴ aisth¯esis: does this mean the faculty to perceive, or what is perceived? opsis, like ‘sight’ in English, can be both subjective and objective; and see Alc. 133a. For aisth¯esis compare De an. 417a 3–6 and Burnyeat’s comment (2002: n. 32). If aisth¯esis is the faculty of perception, and if it is then to be followed by a belief (‘from which some belief ’), does that imply that the faculty of perception is brutish? Not necessarily: the mode of perception’s presentation may be different from the mode of judgement, but still involve more than merely a raw feel: see e.g. the account of judgement given by the soul’s silent dialogue, and also the description of the conversation in the soul at Resp. 523 ff., where perception proffers a report (see McCabe 2006).

Looking Inside Charmides’ Cloak

13

was I who knew the drug [sc. a leaf to cure Charmides’ headache, which Socrates pretends to have²⁵ ], he looked at me with his eyes in the most extraordinary way, and made to ask a question; and everyone in the palaestra flooded right round us. Then, my fine friend, I saw the things inside his cloak; and I was on fire; and I was beside myself,²⁶ and I thought that Cydias was so wise about love, when he advised someone besotted by a beautiful young man, ‘to take care lest he should come as a fawn up to a lion, and become his breakfast.’ (155c–d)

Three things, it is often assumed, happen to Socrates, all to show that he has s¯ophrosun¯e, self-control (after a struggle): first, he sees Charmides close up, and realizes that he is even prettier than he had seemed from afar; second, Charmides ‘looks at him with his eyes’; and thirdly, Socrates sees inside Charmides’ cloak: Critias had promised that Charmides was even more beautiful naked, and Socrates has helped himself to a view.²⁷ But we might pause to notice that what Socrates has a view of is not fully explicit. It is Plato’s readers who suppose that what Socrates sees are Charmides’ genitalia; what Socrates says is that he saw what was inside Charmides’ cloak. But, ‘being inside Charmides’ cloak’ is transitive (as the Relations Argument may later allow us to understand); it includes not only the most exciting parts of Charmides’ anatomy, but his soul too, the part that Socrates says he is interested in. Our assumption that what Socrates sees inside the cloak is physical rests on our further assumption that the objects of sight are just what is physical; and the short passage describing Socrates’ discomfiture is emphatic that what we are dealing with here is perception—in particular, sight. But is that Socrates’ assumption? Compare the two episodes. They are, on scrutiny, pretty striking. Both turn on something about perception (of a sort): Charmides’ perception is somehow ‘inner’; in Socrates’ case it is vague just how far inside he sees. The language of sight is overlaid by some other talk about ‘looking’:²⁸ Charmides is to look inside himself; and Socrates is overthrown by Charmides’ look. Both episodes, ²⁵ The early action of the dialogue turns on a subterfuge suggested by Critias and agreed by Socrates: that Socrates should pretend to have a cure for the headache from which Charmides is suffering. The leaf is a fiction; but Socrates elaborates the subterfuge by saying that the leaf cannot be effective without some incantations: and those just turn out to be dialectical discussion. The fact that Socrates seems to be lying about the leaf is strikingly contrasted with what one might suppose to be the good faith of his emphasis on philosophical discussion, and its bona fide connection with the priest-king Zalmoxis (see below). His pretence is later alluded to when he suggests that the person with the skill of s¯ophrosun¯e should be able to tell pretenders from those who are sincere (173a ff.). The leaf episode has the ambivalence which also seems to characterize Socrates’ looking inside Charmides’ cloak. ²⁶ en emautou: literally, ‘I was no longer inside my own [something or other]’. LSJ suggest ‘in my own house’; the vagueness of the expression is telling, on the interpretation I offer here. ²⁷ Cf. e.g. Schmid 1998: 8, 91. ²⁸ Alc. 132c ff. (and see Denyer 2001: ad loc. and Brunschwig 1973) offers a different account of the connection between sight, looking, and self-perception.

14

Mary Margaret McCabe

moreover, are somehow about being (or failing to be) s¯ophro¯n (when what that means is still undecided). Again with foresight, they reflect on whether either Socrates or Charmides has the virtue of self-knowledge, and on what it would be for them to have it. But the perception which they have (or are invited to have) should not be explained in terms of the brutish claim, which isolates perceptions and their objects from other reflective states. For both perceptions, on one account, turn out false: since Charmides is not in fact an exemplar of s¯ophrosun¯e. He does not know himself; he has no understanding of whatever his own virtue might be; and the other descriptors of the virtue (nobility, good manners) are left to one side: after all, they perform no explanatory role in the psychology of virtue (this is emphasized by the injunction to Charmides to look inside). So if either case of ‘looking inside’—Charmides’ or Socrates’—is a perception of s¯ophrosun¯e (rather than, e.g., just an attempt at perceiving, a looking-but-not-seeing), it is a mistaken one. If the perception is mistaken, it cannot easily be understood as a direct relation between perceiver and some object; since what is perceived is somehow or other false, or not there. So the relation between perception and what is perceived is a complicated one, and so in some measure indirect. Neither case can be, straightforwardly, brutish.²⁹

6 . T H E C I V I L I Z I N G O F PE RC E P T I O N The double use of sight at 167–9 together with these mistaken perceptions at the very outset of the dialogue invite us to reject the brutish view of perception and, with it, the bracketing of perception with the non-psychological relations like large and double. Instead, we might press the connection made between perception and the other psychological states: perhaps, that is, perception is here conceived as analogous to belief and desire. After all, the Relations Argument itself is bracketed by two examples of perception with higher-order content. Perception, then, like belief, is somehow civilized. What does it mean, then, for the object to be presented to perception, as Socrates puts it to Charmides? Presentations are themselves cognitive (notice that at 169a what is presented is unconvincingness); when the presentation is made to perception, it is made by an appearance. That is why Charmides is encouraged to tell Socrates how—on introspection—s¯ophrosun¯e appears to him to be (160e1); and how what he sees at the end of the argument (168e4) is something that appears impossible. Perception, then, is closely connected to appearance: to the way in which things appear for the perceiver (the language of objects determining ²⁹ In part this point is about intentionality. If perception is intentional, it can be mistaken; but it cannot then be brutish.

Looking Inside Charmides’ Cloak

15

the cognitive state is displaced).³⁰ And appearance, we should notice, is both somehow partial (perception is not the same thing as reflection or belief) and also somehow compelling: this is how it really seems. That perception of this kind can be wrong, is part of its seeming; but that it seems so with such compulsion is why it seems so real. How, then, is perception distinct from belief? The difference lies in part in the power it has (we may call this its attitude) to its objects. Belief does believing towards its objects: I believe that the moon is made of blue cheese. Desire, equally, does desiring towards its objects: I wish that I had finished this paper, I wish he were mine—I wish that the moon were made of blue cheese. By parity, perception does perceiving towards its objects. What does that mean? Well, on the evidence of what Socrates says to Charmides, it reports the objects as they are presented. We could describe this in terms of how the object is seen (I see an argument as valid; I see Charmides as having an intemperate soul), or in terms of the content of the report (I see that the argument is valid, I see that Charmides has an intemperate soul). The report, therefore, is a complex one, taking in the whole of the presentation; it forms the basis for a belief; and the belief in turn can come under the scrutiny of reason or thought.³¹ These attitudes are civilized throughout. On such an account, furthermore, the gap we might postulate between sense perception and introspection will narrow: they are both, somehow, of how things potently seem (it is not the case that inner seeing is somehow metaphorical, or that the canonical case is what we would call physical³²). But this narrowing of the gap, legitimated by a civilized account of perception, will further resist the thought that either perception or introspection is somehow a matter of the perceiver being purely affected by the object perceived (this is the model of the Relations Argument; and it is rejected, I am suggesting, by the frame dialogue). When Charmides looks inside himself, and when Socrates looks inside Charmides’ cloak, what is in there provides them with the faculty to perceive; but it does not constitute the seeing. The soul is no more a theatre than is the outside world. In that case, then, there is both continuity and discontinuity between the attitude that is perception and the attitude that is belief. They are continuous by being complex in content, civilized not brutish, susceptible to higher-order ³⁰ Plato is not worried here, I think, about the subjective nature of appearances (and so, e.g., about the problem of explaining the subjective in terms of objective facts of the matter). Instead, he is interested in the perspective imposed by appearance; and thence by the partialness of it: on this see the argument of Phd. 74 ff., where the problem of appearance is not that things appear thus and so (but aren’t, really), but rather that they appear thus and so, and are; and also appear differently, and are. ³¹ That perception reports to thought is a theme of the discussion of thought at Resp. 523–5. ³² See below on the holism of Zalmoxis; I am grateful to Malcolm Schofield for making me see that holism would encourage the view that there is no standardly dualist view of the relation between the physical and the psychological in play here.

16

Mary Margaret McCabe

attitudes of the same type. They are discontinuous, because the attitude of perceiving and reporting one’s appearance is quite different from the belief based upon it. This ensures that the two cognitive states neither collapse into each other, nor are completely detached from each other: there may be, just as Plato describes, a complex set of relations between belief and perception, inquiry and reflection, even desire: just as we find instantiated in the frame dialogue. And it is the frame dialogue that allows us to see that this is so. 7 . S E L F - K N OW L E D G E A N D K N OW L E D G E O F T H E S E L F Return now to the dialectical structure of the dialogue. The Relations Argument can be resisted if perception is not brutish; and the civilized model of perception is to be found in the assumptions of the frame dialogue. What is more, if perception is civilized, there is no reason why it could not be higher-order: no reason why I should not both perceive something and perceive that I perceive it, without risk that the higher-order attitude turns out either vacuous or absurd. For the content of the first-order attitude may be included in the content of the higher-order one; there is nothing about complex content that demands the intransitivity of relations like double or large. We might, however, find there to be something rather unexpected about this account of perceiving that we perceive. If Socrates does indeed see something about the inner Charmides, what he sees is wrong. When Charmides offers an opinion derived from his inner looking, that opinion too is false. So neither seeing is privileged; and inner seeing is not private, either, since Socrates sees inside Charmides’ cloak (rather than his own). If—on this account—we perceive that we perceive, the higher-order perception is not going to generate anything at all like our own idea of consciousness.³³ Higher-order attitudes, then, must be doing some other work here than expressing the incontrovertible view of our own inner states, or an awareness that we are feeling, thinking, perceiving thus or so. Consciousness, I think, has never been the issue here. Instead, the Relations Argument was interested in reflexivity, and in how knowledge could be selfknowledge. But that interest, of course, was the result of a muddle: for the original project to explain knowledge of the self was replaced by a discussion of self-knowledge, which was revealed in turn to be impossibly reflexive.³⁴ The ³³ Or, better, our ideas of consciousness: see e.g. Tye 1995: ch. 1 on the contrast between phenomenal consciousness (the what it is like to feel pain, etc.) and the sort of consciousness that occurs via introspection of my own inner (mental) states—some kind of self -consciousness, perhaps, as in Locke, Essay, II. 1. 19, or as discussed by Ryle (1949: ch. 1). Neither phenomenal consciousness nor self-consciousness can be Plato’s targets here; the first because it is private, the second because it is exclusively mine. There is a great deal more to be said about just how Plato treats these issues, but not here. ³⁴ The reference of ‘self ’ in reflexive expressions is always tricky: a sign on the tables in the University Library in Cambridge announces them to be ‘self-clearing’.

Looking Inside Charmides’ Cloak

17

muddle is explicit in the text. For Socrates expressly (169e) puts aside any objection to treating knowledge of the self and self-knowledge as the same; his recusatio serves to emphasize the possibility that they are not. Now perhaps the reason for this crabwise approach to knowledge of the self may turn out to be just what is left over from the Relations Argument: the possibility of higher-order thought. For, I suggest, the target all along has been the connection not between self-knowledge and reflexivity (for this is rejected by the Relations Argument), but between knowledge of the self and higher-order thought.³⁵ That connection has been left untouched: but its importance is emphasized in the frame dialogue. The account we must give of the argument about self-knowledge, therefore, will be incomplete unless we read it within its frame: for it is the frame that supplies its interpretation. If the dialogue offers an account of perception as civilized, and also suggests that there are significant analogies between perception and knowledge, then knowledge, too, may be construed in a civilized manner. If knowledge is civilized, it is to be understood not as a brute relation between a power and its objects, but as an attitude with complex content; and this allows for the possibility of a higher-order knowledge, which can embed the content of the first-order attitude in its content (content is not intransitive). What, then, are we to say of knowledge of the self? It should be no longer brutishly reflexive, for sure, and so no longer paradoxical in the manner the Relations Argument urged. Instead, because knowledge may be both higher-order and reflective, we may know what we know and do not know without this kind of knowledge being empty or absurd. This reflective attitude may still be construed as knowledge of knowledge (by virtue of its higher-order component): how exactly is it knowledge of the self? That knowledge of the self may be construed as knowledge of knowledge is congenial to an intellectualist account of Socratic inquiry. Especially, this reflective knowledge will include other knowledges within its scope: it will, therefore, be a suitable condition for a complex investigation into what one does and does not know—this kind of knowledge, therefore, will have a holistic cast. That thought, in turn, would be welcome to one of the heroes of the dialogue: the Thracian priest-king Zalmoxis, who has a holistic approach to medicine (you can’t fix someone’s headache without fixing their whole soul), and who supposes, further, that your soul is the source of value for the rest of you. Socrates reports him as having said: As you should not try to cure the eyes without the head, nor the head without the body, likewise you should not try to cure the body without the soul. But this is the reason why the Greek doctors fail to cure many diseases, because they care nothing for the whole, which should be cared for, since while the whole is not well, it is impossible for the part ³⁵ This is why transitivity is one of the features of relations in which Socrates is interested in the Relations Argument.

18

Mary Margaret McCabe

to be so. For everything arises from the soul: all good and evil for the body and the whole person; and it flows out thence as if from the head to the eyes. (156e)

But the combination of Socratic higher-order thought and Zalmoxis’ holism have a rather uneasy consequence. For notice the Heraclitean tone of Socrates’ account of inquiry at 166d1: he is searching for himself, but there is no guarantee he will find him.³⁶ Fair enough: for if knowledge is holistic, it may seem that it needs to be acquired as a whole; and that, perhaps, is an extraordinarily hard task, if not an impossible one. Further, if knowledge of the self is knowledge of knowledge, the self that is sought may be so complex that the search will never end: the self as a whole is elusive, just because knowledge itself is (well-nigh) impossible to complete (and Socrates is dead). Plato would reject as spurious any puzzle that makes the self elusive on the basis of reflexivity;³⁷ he is right, I think, to observe that the real problem of the elusive self is that reflection on its limits never ends. REFERENCES Bailey, D. (2006), ‘Knowledge in Plato’s Charmides’ (unpublished). Brunschwig, J. (1973), ‘Sur quelques emplois d’opsis’, in Zetesis, Festschrift for E. J. De Strycker (Antwerp), 24–39. Burnyeat, M. F. (1981), ‘Aristotle on Understanding Knowledge’, in E. Berti (ed.), Aristotle on Science (Padua), 97–139. (1992), ‘Is an Aristotelian Philosophy of Mind Still Credible? A Draft’, in Nussbaum and Rorty (1992), 15–26. (2002), ‘De Anima II. 5’, Phronesis, 47: 28–90. Denyer, N. C. (2001), Plato: Alcibiades. Cambridge. Hume, D. (1978), A Treatise on Human Nature, 2nd edn., ed. L. A. Selby-Bigge. Oxford. Kosman, L. A. (1973), ‘Perceiving that we Perceive: On the Soul III.2’, Philosophical Review, 84: 499–519. Locke, J. (1975), An Essay concerning Human Understanding, ed. P. H. Nidditch. Oxford. Lyons, J. J. (1963), Structural Semantics. Cambridge. McCabe, M. M. (2006), ‘Is Dialectic as Dialectic Does? The Virtue of Philosophical Conversation’, in B. Reis, (ed.), The Virtuous Life in Greek Ethics (Cambridge), 70–98. Nagel, T. (1974), ‘What is it Like to be a Bat?’, Philosophical Review, 83: 435–50. Nails, D. (2002), The People of Plato. Indianapolis. Nehamas, A. (2004), ‘Epist¯em¯e and Logos in Plato’s Later Thought’, in Nehamas, Virtues of Authenticity, (Princeton), 224–48. Nussbaum, M. C., and Rorty, A., (1992) (eds.), Essays on Aristotle’s De Anima. Oxford. Ryle, G. (1949), The Concept of Mind. London. ³⁶ Compare the journeying metaphor in Heraclitus: ‘I sought out myself ’ (DK22B101), or ‘Going to the limits of the soul, you would not find them, even if you travelled every road; it has such a deep account’ (DK22B45). ³⁷ See e.g. Hume, 1978: 252.

Looking Inside Charmides’ Cloak

19

Schmid, W. T. (1998), Plato’s Charmides and the Socratic Ideal of Rationality. Albany, NY. Shorey, P. (1907), ‘Emendation of Plato Charmides 168b’, Classical Philology, 2: 340. Sorabji, R. R. K. (1992), ‘Intentionality and Physiological Processes: Aristotle’s Theory of Sense-Perception’, in Nussbaum and Rorty (1992), 195–226. Tye, M. (1995), Ten Problems of Consciousness. Cambridge, Mass. Van der Ben, N. (1985), The Charmides of Plato: Problems and Interpretations. Amsterdam.

2 Socrates and Philosophy as a Way of Life John M. Cooper

I One aspect of philosophy in antiquity, probably more than any other, sets it apart from philosophy nowadays, or indeed at any time since the end of antiquity. This is the idea of philosophy as a way of life. This conception was not prominent, or even perhaps present, at all periods or in all authors. But for many ancient philosophers, and in its public image, philosophy was not merely and purely a subject of study. Philosophy in antiquity was, as it has been ever since, a range of intellectually challenging topics and questions to be investigated, whether orally or in writing or both: conflicting arguments to be sorted out; conclusions to be proposed and defended via philosophical arguments that one would state and explain; alternative and competing theories to one’s own to be canvassed and rejected for reasons one would retail, defend, and recommend to others. But ancient philosophers were not always just lecturers in an intellectual ‘‘specialty’’, nor was their profession merely that of a teacher of doctrines and methods of argument, and the like. Ancient philosophers differed quite particularly from other people with regard to their way of life, even from others engaged in careers of teaching or other intellectual work—for example, in medicine or mathematics. I don’t have in mind here trivial externals such as wearing a beard or frequenting the city’s public places in torn clothing; I mean that philosophers lived their philosophy—and not just in that they spent their working hours engaged in philosophical pursuits. Doctors or mathematicians also spent most of their days I thank an audience at Georgia State University, Atlanta, where I delivered a preliminary version of this paper in April 2006 for sharp and helpful discussion which spurred me to make a number of clarifications and other changes. Later versions were delivered at the University of Oslo and Middlebury College, Vermont, and I thank my hosts and those in the audience for interesting and helpful discussions, which enabled me to make further adjustments. Alexander Nehamas gave me two sets of extraordinarily perceptive comments and criticisms on successive versions of the paper. Any merit of this final version is due in very considerable measure to his instructive objections and suggestions.

Socrates and Philosophy as a Way of Life

21

in their own professional pursuits. A philosopher made philosophy the basis of his whole life. In fact, more even than it meant investigating, discussing, and teaching the subject, being a philosopher, for many ancient philosophers, meant living one’s whole life a certain way—philosophically—and encouraging others to live that way, too. Our heightened awareness nowadays of this aspect of ancient philosophy is due in considerable measure to the efforts, going back to the 1960s, of the French scholar of Plotinus and Neoplatonism, Pierre Hadot, who thought it an essential characteristic of all ancient philosophy, from its pre-Socratic beginnings, without interruption or exception, down to its end with the last pagan Platonists of the sixth century .¹ That is an exaggeration. Our best evidence suggests that in fact Socrates, or perhaps rather his image in the writings of Plato, Xenophon, and the other writers of Socratic dialogues who had personal knowledge of him, is responsible for initiating and setting the contours of the whole subsequent development of philosophy in antiquity conceived as a way of life.² There seems, in any event, no doubt that the model of Socrates as a philosopher leading a philosophical life, especially as refracted through the writings of the Socratics, played a decisive role in giving this conception the prominence that it enjoyed in Hellenistic and later ancient philosophy. Recently, Alexander Nehamas has illuminatingly treated Socrates as such a model philosopher, in his book The ¹ Besides the articles collected in Hadot 1995, see also Hadot 2002 and 1998. ² Before Socrates there were the Pythagoreans in southern Italy and later in mainland Greece, at Thebes and Elis. They plainly constituted some sort of cult or ‘‘brotherhood’’ with some sort of common life together. This combined what we can recognize as philosophical ideas with religious dogmas, ritual practices, and taboos, whether in a political community as at Croton and Metapontum in Pythagoras’ own day (last half of sixth century ) or in private organizations (‘‘schools’’). Murky as the whole history of early Pythagoreanism is, however, I do not think the role of philosophical ideas in this mix (say, ideas about the immortality or transmigration of souls or anything to do with the importance of numbers) was at all comparable to the role of philosophy in Socratic and later conceptions of philosophy as a way of life. For these later thinkers, logical, philosophical argument and analysis, and the philosophical and moral ideas that reason so employed was found to lead us to, was the essence and foundation of philosophy as a way of life. In that sense, the Pythagorean way of life was not an instance of the later Greek conception. Nor do I see any warrant for Hadot’s offhand suggestion (1995: 71) that similar ‘‘schools’’ existed generally ‘‘among the Presocratics’’, in which ‘‘the thought, life-style, and writings of a master were religiously preserved’’. There is no reason to think that Parmenides or his Eleatic followers (e.g. Zeno) had any special way of life, or thought of their philosophical inquiries or arguments as helping to define or support any such life (Plato makes no such suggestion in his own Parmenides or Sophist-Statesman). Nor do we have evidence of any identifiable ‘‘followers’’ at all of Empedocles, or Anaxagoras, or Heraclitus (despite Plato’s conceit at Tht. 179e–180d), or the earlier Milesian philosophers Thales, Anaximander, and Anaximenes—such as might have constituted such a ‘‘school’’ with a way of life. We do know of followers of Democritus, who taught his atomist philosophy through the fourth century, and eventually to Epicurus; but we know no reason to think that they led any special way of life as part of their devotion to Democritus’ philosophical ideas. In fact, it is only what came to be called philosophy, under the influence of Socrates and Plato (see n. 4 below), that made it possible even to conceive of philosophy as a way of life. As for Pythagorean communities in Plato’s own lifetime, there is little reason to think that they bore any resemblance to life in Plato’s Academy, and a good deal for thinking that they did not. On this see Lynch 1972: 61–2.

22

John M. Cooper

Art of Living. Nehamas links Socrates to Montaigne, Nietzsche, and Foucault as practitioners of an ‘‘art of living’’ through philosophical writing (or discussion) in which one establishes oneself as an unusual and distinctive person, a person of ‘‘an uncommon and idiosyncratic character, a set of features and a mode of life that set oneself apart from the rest of the world and make one memorable not only for what one did or said, but also for who one was’’ (Nehamas 1998: 5). In this paper I do not pursue the question of origins for the ancient idea of philosophy as a way of life, or of Socrates’ influence in the establishment of this idea in subsequent ancient philosophy. Nor do I follow Nehamas in treating Socrates as one who practiced in his philosophizing an art of self-construction as an unique individual. I attempt simply to examine Socrates’ conception of philosophy itself as a way of life, of philosophy as something to be lived, not merely thought and talked about. In speaking of ‘‘Socrates’’, unless I specify the character of that name in one or another specific work, I will mean the historical figure as he was taken by the later ancient tradition to be presented through these Socratic writings (which is the only or chief way they knew him)—and particularly, of course, through the Socratic dialogues of Plato.³ I will argue that, ³ This means that as I use the name, ‘‘Socrates’’ refers to both a literary figure and, in a way, a historical person. My Socrates is the historical Socrates as presumed to have been presented in certain dialogues of Plato, the dialogues of Xenophon, and those of the so-called minor Socratics. I make no attempt to get behind the literary figure(s) to the genuine historical personage. All these authors, of course, knew Socrates personally and had participated in or attended his philosophical discussions and inquiries (Xenophon perhaps less intently than the others). But the ‘‘Socratic dialogue’’ (Σωκρατικὸς λόγος) was a recognized genre of writing (see Aristotle’s comments, Poet. 1, 1447b 11), in which, it seems, though it was certainly not a form of out-and-out fiction, authors were not expected to attempt anything like an ‘‘accurate’’ presentation of actual Socratic conversations, or to attempt to convey with any specificity the ideas or philosophical concerns of the historical person. (See C. H. Kahn’s somewhat extreme account of the authors’ freedoms (Kahn, 1996)). Nehamas’s more complex account is close to my own guiding assumptions here see his 1998: 6–8, 93–6, and 101–21 with nn.) Since Socrates wrote nothing, any influence of ‘‘Socrates’’ on the history of philosophy after the time of these personal friends could hardly be much else than an influence of this literary character. Often already with Aristotle, as when he discusses the views of Socrates on lack of moral self-control (Eth. Nic. VII 2–3; see also Eth. Eud. III 1, 1229a 15 on courage), Socrates is taken to hold the same views as the character of Plato’s dialogues (in this instance, the Protagoras). (For a similar reliance on a Socratic author for Socrates’ view, see Rh. II 20, 1389b 4–8, reporting, it seems, Xen. Mem. I 2. 9—unless in this instance Aristotle goes back to Polycrates himself, Socrates’ accuser presumably referred to by Xenophon.) From Aristotle’s remarks elsewhere (Metaph. A 6, 987b 1–4, and M 4, 1078b 17–19, 27–31), we can tell that by his time there was uncertainty and dispute about what the historical person had in fact believed and what characterized his own approach to philosophy. I should add that so far as concerns Socrates in Plato’s works, I distinguish between some of Plato’s dialogues as Socratic and others not (see Cooper 1997: pp. xii–xviii). Already in antiquity, as we can see from Diogenes Laertius in his Life of Socrates (II 45), it was recognized that although Socrates is the main speaker in most of Plato’s dialogues, in some of them Plato puts into Socrates’ mouth philosophical views of Plato’s own devising, which go beyond and even contradict some of what he himself attributes to Socrates in the dialogues that establish for posterity and define Socrates’ philosophical and moral personality and his philosophical interests—esp., the Apology and Phaedo with its autobiographical story. As Socratic dialogues of Plato I have in mind Euthyphro, Apology, Crito, Alcibiades, Charmides, Laches, Lysis, Euthydemus, Protagoras, Gorgias, Meno, Greater and Lesser Hippias, Ion, Menexenus, Clitophon, and Minos (on

Socrates and Philosophy as a Way of Life

23

for Socrates, to live a philosophical life meant to live with the idea that reason, conceived as the capacity for argument and analysis in pursuit of the truth about things, is our highest and most essential capacity. To live a philosophical life is therefore to live consistently on the basis of reason so conceived—philosophical reason, we can say—in everything that we do. Socrates had a quite particular conception, not followed by all his successors in this tradition, of what living on the basis of reason entailed. It is my main purpose to explain and explore those particularities. But, in a nutshell, both for Socrates and all his successors, this is what being a philosopher and living a philosophical life meant: living according to reason, conceived as a capacity for argument and analysis in pursuit of the truth.

II We need only recall Plato’s Apology of Socrates to see how the conception of Socrates as a philosopher who made a point of living his own philosophy might have gained currency. The Apology can also give us insight into what, for Socrates, philosophy as a way of life actually meant. Socrates is represented as having devoted himself over many years to what seems to have been full-time engagement in discussions in the public places of Athens with various people. Some of them were visitors, some young men who flocked round to listen to him, some adult persons with settled positions and reputations in Athenian society. These discussions were philosophical in character (as philosophy came to be understood, and as Plato himself advocated it should be understood⁴). They the assumption that they are all genuine works of Plato). To these one might add Resp. I. In this last case, as in several of the others (notably Protagoras, Meno, Gorgias, and Euthydemus) one must be prepared to find suggestions of ideas and concerns of Plato’s own, going beyond those of the character Socrates as presented in Apology and the bulk of the Socratic dialogues—ones that we can see further developed and pursued in non-Socratic works of Plato. So the division between Socratic and non-Socratic dialogues is not absolute, and has to be treated with circumspection. ⁴ As Hadot (2002: 15 ff.) argues, Plato was responsible for fixing the meaning of ‘‘philosophy’’ that has lasted ever since, at least in academic circles. It involves (on my own characterization) a commitment to logical reasoning as the fundamental method for the formation of philosophical beliefs, and to a broad scope for philosophical beliefs, including beliefs resulting from inquiries into physical science and metaphysical issues, not just on the questions of practical life and politics that Socrates devoted himself to. That Plato was responsible for establishing this understanding of philosophy was the view of Jaeger (1928). Following Burkert (1960), Hadot suggests, not implausibly, as others have done too, that the old tradition that Pythagoras was the first to introduce the term ‘‘philosopher’’ (which he applied to himself ), is an invention—deriving from an anecdote, set out most fully for us in Cic. Tusc. V 8–9, of Heraclides Ponticus in the second half of the fourth century, written under the influence of Plato’s own conception of philosophy. The term ‘‘philosopher’’, once it was introduced at some point in the mid to late fifth century, at first sometimes meant only, rather indiscriminately, a person devoted to ‘‘intellectual and general culture’’ (see Herodotus I 30, concerning Solon, and Thuc. II 40. 1, in the Funeral Oration). But as a passage of Gorgias’ Helen (B11 DK, sect. 13) shows, there was already, in the latter part of the

24

John M. Cooper

consisted of questions that Socrates would ask about some matter of importance for human life, to get the discussion started, and then further questions he would direct at the respondent’s answers. He would emphasize in his questions (and ancillary comments) logical relations of implication, inconsistency, and the like, in trying to work out between himself and the respondent what the respondent actually thought, insofar as the respondent was to be taken seriously as someone with opinions about what the truth was on the subjects under exploration. To hold opinions responsibly about what is true, Socrates assumed, is to be prepared to explain and defend them, by appealing to reasons that in fact do support them as being true, and to be committed to accepting as further opinions (or parts of the initial ones) any logical or other consequences plausibly to be drawn from them. His discussions, moreover, were always confined to issues about human life, how to lead it correctly and well—in particular, how to conceive and appreciate the value of the various highly rated traditional virtues (justice, courage, temperance, etc.), in comparison with other things apparently also of value (such as bodily health, physical strength, wealth, bodily or other pleasures, and so on).⁵ fifth century, a special usage in which it specifically indicated someone engaged especially in logical argumentation. (See also Dissoi Logoi, DK 90, 1. 1 and 9. 1, in the latter passage accepting the MSS. reading reported in their apparatus, instead of DK’s text.) Relatedly, the chorus in Aristophanes’ Ecclesiazusae (produced c. 392) talks (571) of the need for a deep ‘‘philosophical thought’’ from Praxagora if she is to go to the defense of her female friends; the thought turns out to be Praxagora’s plan for common ownership of all property and equal shares of everything. Her and the others’ comments on this show that a ‘‘philosopher’’ is one who can think up new things, contrary to all tradition, by relying on ideas of their own while dismissing custom and ordinary experience if they conflict with what their reasoning has, they think, revealed to be true or right. Furthermore, the author of the Hippocratic treatise On Ancient Medicine (probably dating from about 410 ) once refers disparagingly (in his text as our MSS present it) to the investigations of people like Empedocles ‘‘into nature’’ as ‘‘philosophy’’ (ch. 20, li. 5, in Jones 1923). (See Cooper 2004, 3–42.) Thus already, before any effect that Plato’s writings might have had, there was in play a well developed notion of a ‘‘philosopher’’ as someone engaged in logical argument and trusting to reason in pursuit of the truth about how things actually are, while, if that pursuit of the truth requires it, disregarding experience and convention. Thus Plato drew upon a preexisting tendency to use the term specifically for those devoted to logical argumentation and prepared to follow reason, in pursuit of the truth, wherever it led, even if it conflicted with custom and experience. He was able to establish that usage and to defeat Isocrates’ rearguard counter-efforts during the same period to reserve the word for a person, like himself, devoted to general culture, including a close knowledge of Greek history and, especially, the cultivation of language and style in speaking—abilities that were of some actual use. Isocrates disdained technicalities of all kinds, including fussy logical argumentation whether or not (according to the practitioners) in pursuit of the truth—Isocrates didn’t notice that some uses of logical argument might not be just for their brilliant effect but for discovering or establishing the (perhaps surprising) truth about something. Isocrates wanted to call the other, more technical, people ‘‘sophists’’, by contrast: see e.g. Helen 1 and 6, and Antid. 261 ff. Even Isocrates, however, knows the special use of the word ‘‘philosophy’’ that I have referred to. In his exhibition piece, usually thought to date from c. 385, on Busiris as an ancient king of Egypt who introduced or encouraged the discovery of all kinds of good things that the Greeks then borrowed for their own cultural enrichment, he includes among these ‘‘philosophy’’, which he describes as including both the establishment of political laws and ‘‘the investigation of the nature of things’’; he links this to the study of astronomy, arithmetic, and geometry (§§22–3). ⁵ See Pl., Ap. 19d, 30a–b; Xen., Mem. I 1. 16.

Socrates and Philosophy as a Way of Life

25

Socrates famously maintains in Plato’s Apology that one’s soul and its condition, whether good or bad, is the most important thing for anyone (30b). It is vastly more important than any of those other values. For him, the good condition of the soul is entirely a matter of one’s ability, unfalteringly and with an inexhaustible thoroughness, to understand, explain, and defend by philosophical argument and analysis one’s own values and commitments. One’s understanding of one’s values and commitments is the basis, for him, of all a person’s decisions and actions. Thus, the good and the bad condition of the soul is determined solely by whether or not one makes the use of one’s own powers of reasoning, in pursuit of the truth, the fundamental and constant guide for the living of one’s life, and by how fully and well one does so. How well does one’s use of this power measure up to the standards for reasoning implicit in the very capacity to think things out and decide what to think, and to do? Socrates also famously maintains that the unexamined life is not worth living for a human being (I return to this below, sect. III). This is a life like his own, a life devoted to examining others and himself at the same time (Ap. 38a), in discussions on just these questions about the rational defensibility of one’s own values and commitments. But it is more than that, as well. The viewpoint that the state of one’s soul is the most important thing determining the quality for better or worse of one’s life implies the decision to follow in one’s daily actions—other actions, I mean, than those involved in carrying on such discussions—the values and commitments that are most completely supported by the results of one’s examination of others and self-examination, so far as one has gotten at any given time in that process. One could not really hold Socrates’ view about the importance of the state of one’s soul if, instead of following one’s own reasoned views about what is true about values, however provisionally arrived at, one opted to follow some other authority. We learn from Platonic dialogues such as Protagoras and Meno that Socrates thought (so to speak as a matter of human psychology, the psychology of human action) that human beings always and necessarily act in whatever way, at the moment of action, they think it best to act. It would be possible, nonetheless, for someone to follow other authorities than that of their own reason. They could still think that that was better for them. Of course, they might well, and presumably would, be unable to rationally defend that view with much success—but, disregarding the value of their own soul as they would be doing, they would not care about this. But someone who held the Socratic view about the importance of the soul could not live in any other way than by following the guidance of their own values and commitments. Thus, for Socrates, philosophy as a way of life—this life that he himself led and was recommending to the young men and others with whom he conversed—was, essentially, a commitment to as rigorous, constant, and complete as possible a use of one’s own capacity for reasoning, in pursuit of the truth, in all the decisions and actions that make up one’s life. It was only by making this commitment, he thought, that one could care for one’s soul in the way that its overriding importance demanded.

26

John M. Cooper

What I have said so far is no more than a sketch of how, for Socrates, philosophy was a way of life. Many questions arise about the several elements in this view, and in what follows I will be pursuing some of these. I have suggested that Socrates was the first ancient philosopher to conceive of philosophy as a way of life, and that his conception laid the groundwork for the whole subsequent development of this aspect of ancient philosophy. We can already see, then, that at the root of this whole development is the idea that to be a philosopher is to be fundamentally committed to the use of one’s own capacity for reasoning in living one’s life. The philosophical life is essentially simply a life led on that basis. This is the basic commitment that every true and full philosopher made in adopting philosophy, in choosing to be a philosopher, whatever ancient school they belonged to. Pierre Hadot speaks of an ‘‘existential option’’ as needed when anyone aligns himself with the doctrines of any specific school.⁶ But this is incorrect. Any specific philosophical views and orientations that might characterize an ancient philosopher in the subsequent historical development (as a Platonist or Aristotelian, or Stoic or Epicurean or Skeptic) do not result from anything ‘‘existential’’. They result simply from his or her coming to accept specific ideas, all of them supported by philosophical reasoning in pursuit of the truth, that these philosophical schools might put forward about what, if one does use one’s powers of reasoning fully and correctly, one must hold about values and actions. One’s ‘‘option’’ for any one of these philosophies in particular, far-reaching as the consequences might be for one’s way of life, does not deserve to be called an ‘‘existential’’ one. The only ‘‘existential’’ option involved is the basic commitment to being a philosopher, to living on the basis of philosophical reason. The choice to be an Epicurean, or a Stoic, for example, depends—certainly, by the standards of these philosophical movements themselves, it ought to depend—on rational arguments in favor of the fundamental principles of the philosophical school in question. It is crucial for a correct understanding of what ancient philosophy was, or is, that one sees the central force of the fundamental commitment to living a life on the basis of philosophical reason. It is this that set philosophers off as a single group from the rest of the population. ⁶ On ‘‘existential choices’’ or ‘‘options’’ see Hadot 2002: 102, 103, 129, 132, 176, etc. Hadot begins to use this terminology only with his chapter on ‘‘The Hellenistic Schools’’, but he makes it clear from the first occasion (p. 102) that he thinks that Socratic philosophizing, and Plato’s and Aristotle’s schools too, were characterized by such a fundamental option or choice. For him, a different ‘‘existential choice’’ is needed for entering any particular ‘‘school’’. However, there is no reason to think that any ancient philosopher made a choice first to be a Stoic and live a Stoic life, or any other specific philosophical life. For one thing, many of them studied at more than one of the Athenian schools, simultaneously or in sequence, before finally settling in one philosophical milieu or another. First came the decision to live a philosophical life (perhaps, of course, under the influence of the attractions of some particular version of it)—to live according to reason. Even if at the same time one decided to live as a Stoic or an Epicurean or a Platonist, that specific choice was logically subsequent.

Socrates and Philosophy as a Way of Life

27

III The central feature of philosophy as a way of life for Socrates, which distinguishes it from all later such conceptions, is that it is a life of constant and continued examination of the moral opinions of others, and of self-examination in the process.⁷ So highly does Socrates rank the good of his soul over all other values, and so closely does he tie it to philosophical investigations and self-examination, that he tells the jury at Plato’s version of his trial that should they decide to acquit him on condition that he give these practices up, he would refuse (29d). Why would he refuse? At first he gives as his reason that, in preference to them, he must obey the god Apollo, whose oracle’s reply to Chaerephon’s inquiry (21a) constituted a command to him to devote himself to examining others. He adds to this that the jury, as the representatives of the Athenian people, should find their own interest better served by having him continue rather than give it up. I do not doubt that Socrates is sincere in giving these as two of his reasons (however unlikely they might be to persuade the jury).⁸ But later, after being convicted, he adds what must have been, for him, the deepest and most fundamental reason why he would under all circumstances continue to practice philosophy in his own way (37e–38a). When considering exile as a possible penalty to propose, he first repeats, on the assumption that exile would be an ineffective escape for him unless he coupled it with cessation, that to cease would be to disobey Apollo. He has already said that he will not do that. But the jury must not have believed him before: they voted to condemn him. So ⁷ The life of a Pyrrhonian skeptic might not appear very different. But Socrates’ ideas about what philosophical analysis and argument might lead humans to know excluded from the outset anything to do with ‘‘things in the heavens or below the earth’’, i.e. the study of nature other than the specifically human. Pyrrhonian skeptics lived no less skeptically about that than about the knowledge of human values and right action, and were open and indeed committed to continued examination of arguments in every area of philosophical study. On the other hand, they certainly lacked Socrates’ faith in and existential commitment to living according to reason, however much of their time they may have devoted to examining arguments that profess to tell us what reasoning in pursuit of the truth has to teach us. In fact, it is much more Academic skeptics whose way of life really does resemble Socrates’ own. See Cooper 2004a. ⁸ In evaluating Socrates’ choice to give his jurors these reasons, one ought to bear in mind, as for the first, that the charge he is facing is one of impiety. By drawing special attention to Apollo’s reply to Chaerephon, he is able to claim that the very activities which he thinks have led the Athenians to regard him as impious are precisely the ones that the great god Apollo directed him to engage in. These activities were in fact expressions of his own deep-seated piety, he can claim, and thus evidence not of impiety but its very opposite. However purely retrospective the specific role he assigns, now at the end of his life, to the oracle at Delphi may have been, making this claim is not just a clever rhetorical move. It is something Socrates thinks he has strong reasons to believe: see my discussion below, in the last two paragraphs of this section. As regards the second reason, Socrates seems to be imitating fourth-century orators who in arguing cases sometimes try to convince the jury that the Athenians’ own interests require a vote in their favor. (See e.g. Lys. 1. 34–6 and Dem. 21. 219–24, and Yunis 2005: 204–8. I have profited from discussion with Yunis about this oratorical practice.) But this too is something that Socrates thinks he has good reason to assert.

28

John M. Cooper

now, he says, they will think he is speaking ‘‘ironically’’ in repeating that claim.⁹ By that, he means they will think he is saying something he doesn’t really mean, and perhaps that he is preening himself at the same time. But, he goes on to say, they will believe him even less if he says—this must be his deepest reason—that in fact ‘‘it is the greatest good for a person to discuss virtue every day and those other things about which you hear me conversing and testing myself and others, for the unexamined life is not worth living for human beings’’.¹⁰ But why does he think that it is the greatest good to converse every day, testing himself and others—such a good that he would never willingly give it up? The answer must be (I’ll come back to this below) that it is only by doing so that, he thinks, one can keep on improving one’s grasp on, and one’s rationally justified confidence in, one’s own rationally examined system of values—or improve it in other ways, by refinement and clarification, for example. One is constantly testing it against the possibly different views of others, learning new reasons, or deepening one’s understanding of the reasons one had already come to appreciate that indicate that one’s value-commitments are correct. Still, why should the philosophical life, the life of reason, require continued conversation and testing on these topics, right up to the time of one’s death? Why does Socrates think that the high ranking given to the good of one’s soul requires this? One might think that, at some stage in one’s self-examinations and examinations of others’ opinions, one might reach a point where one had done enough of that. One could then set aside the constant practice of philosophical argument and examination, and begin to live with the confidence that one’s own practical outlook, formed and confirmed by that process, was rationally well founded and endorsed by reason itself. The basic good of one’s soul would now be established. One would have achieved the perfection of one’s understanding of everything that is of value in a human life and the ability thereby to work out what to do in every situation of life, and to give an adequate rational defense of any of one’s actions if required. Continued attention to the good of one’s soul as the first priority would from then on consist in acting steadily and continuously on that understanding in a daily life not spent in philosophical argument and discussion. One could assume (or resume) the life of a politician, say, or a physician, or sportsman, or farmer and estate manager, or whatever. We might expect that such a person’s way of leading their life would be altered in ⁹ I put this adverb in quotes, because I am not sure just which of the many subtleties of the Greek verb εἰρωνεύεσθαι, which Socrates applies to himself here, are in play. Nehamas (1998: 49) thinks it is boastfulness by claiming to be a favorite of the god that Socrates is wary about if he should tell the jury again about his ‘‘divine mission’’. It is clear, as Nehamas emphasizes in his discussion of Socratic ‘‘irony’’, that εἰρωνεία always involves an implication that one is superior to those one is exercising it on. So it always risks being found offensive. But I am not sure that Socrates is alluding to offensiveness by boasting here. He is certainly saying that the jury will think he is saying something he doesn’t really believe, and that could be offensive enough. It is, in any event, adequate reason for the jury not to take what he says seriously. ¹⁰ Translations from Plato are from Cooper 1997, sometimes, as here, with alterations.

Socrates and Philosophy as a Way of Life

29

various ways, deepened perhaps, and widened in its perspectives, certainly. Their philosophically arrived at and supported commitments would be evident in many ways in their life, and it would not be unreasonable to call it a philosophical life. Nonetheless, the life would henceforward be that of an ordinary citizen, showing the effects of philosophy, but not itself a life of philosophy, of constant engagement in philosophical argument and analysis. This is how, to put a more favorable spin on it than certainly he intended, the character Callicles in Plato’s Gorgias thinks of philosophy: you need to do philosophy when young, but you reach a point where enough is enough, and you are free to move on to some career or other, something really important and worthwhile, leaving behind all but occasional or ancillary engagement in philosophical study and discussion.¹¹ It is also the view implied by Seneca when, challenged to name a Stoic wise man—one who had reached the goal to which philosophical study was ultimately directed—Seneca instances the younger Cato (‘‘of Utica’’).¹² Cato was a full-time Roman statesman and administrator of the first century , noted for his probity and, even more, for Stoically committing suicide rather than accept a pardon and reconciliation with Julius Caesar upon Caesar’s final victory over the Republicans, one of whose leaders Cato was. It is worth noting, however, that for Seneca Cato was a wise man (σοφός, sapiens), not a philosopher (φιλόσοφος). Perhaps not fully coherently (if a philosopher is thought of, literally, as a seeker after wisdom, who therefore cannot yet have attained it), he treats others of his Stoic heroes (including Socrates) as philosophers as well as wise men. For Seneca, Cato’s life was that of a wise man, because it was led with full commitment and consistency with the Stoic theory as its basis, but not that of a philosopher (even though part of Cato’s legend has him reading philosophical books in his tent when on campaign). A Stoic philosopher might also be a wise man, but his life, the way he lived his philosophy, would involve constant engagement in philosophical discussion. For Socrates, as we have seen, philosophy could be a way of life only in the second ¹¹ This is also how Isocrates thought: see Antid. 261–9. ¹² Seneca, Constant. 2. 1–3, 7. 1; Prov. 2. 7. Hadot refers to Seneca’s use of Cato as an example in these passages, but he fails to attend to the important distinction, observed by Seneca, between a wise man and a philosopher. He says that ‘‘Posterity admired [Cato] as a philosopher’’ (Hadot 2002: 173; see also 1995: 272). Hadot cites no evidence for this beyond these passages from Seneca. But Seneca never calls Cato a philosopher in either of the works cited, but only a wise man (sapiens), in De Constantia, or a brave and virtuous one, in De Providentia. Nor does he do so elsewhere: the closest he comes is at Moral Letters 14. 12, where he speaks of Cato as ‘‘philosophizing’’. But the context shows that this merely refers to Cato’s commitment to Stoic philosophical principles. Socrates, by contrast, he describes with both terms: for Socrates as philosopher see Letters 44. 3 and 71. 7, and other passages. Seneca reserves the term ‘‘philosopher’’ for teachers and writers of philosophy. Cicero once (Fin. 4. 61) speaks of Cato as ‘‘studiosissimus philosophiae’’, which means only that he was a devoted student of philosophy. In extant Latin literature what appears to be the sole reference to Cato as a philosophus comes in Apuleius (mid-2nd century ), Apol. 42, where he even calls Varro one, too! (I thank Robert Kaster for his help on this.) So Hadot seems quite unjustified in saying that ‘‘posterity’’ admired Cato as a philosopher.

30

John M. Cooper

way, through daily devotion to philosophical argument and discussion, not in Cato’s. But why? Surely, Socrates would not altogether disapprove of someone who, like Cato, set such a limit on their devotion to philosophical inquiry and discussion. However, it appears, he would think that any one who did that, and was satisfied simply to live on the basis of their previously examined set of values, was failing to take seriously enough their commitment to follow reason as their guide in living their lives. They would be following reason, up to a point, in that they were living according to their own thought-through, examined ideas, rather than by accepting instead some other authority (tradition, convention, inherited ideas, poetic portrayals of models for living—in short, the weight and influence of the surrounding culture). But, as I said above, to live on the basis of reason requires that one’s use of this power measure up fully to the standards for reasoning implicit in the very capacity to think things out and decide what to think, and to do. Apparently, Socrates thinks that the standards for its proper use implicit in this capacity tell against ever thinking that one has successfully completed the task of examination and testing of others’ and one’s own ideas about human values. For this reason, anyone who attempted to live a philosophical life without continuous philosophical discussions and examinations, instead of Socrates’ own way, would thereby repudiate their own Socratic claim that the good of the soul is the most important thing for human life. They would in fact be neglecting it, by ranking some other value or values ahead of the soul’s good condition: whatever the other values are that they go off, leaving philosophical argument aside, to pursue. There are two related sides to Socrates’ reasons for thinking this, as I will explain below. Both of them begin from the thought that reason, by its very nature and essential dynamic, aims at and is unsatisfied with anything less than final and complete knowledge—wisdom—about what truly is valuable in human life. Socrates thinks that human beings, as it appears, are not capable of achieving this knowledge. On the other hand, as it must have seemed to Socrates himself from his personal experience, one certainly can, by devoted application to philosophical discussion, extend and improve one’s grasp of the reasons that do support a certain set of values and commitments—the very ones, as it seems, that full wisdom would establish once and for all as true. Thus, since the final good condition of the soul—wisdom or knowledge—is apparently unattainable by human beings, but progress and improvement is always possible, proper attention to the good of the soul as the most important good to aim at in human life requires nothing less than precisely the life of philosophy to which Socrates devoted himself. Hence, for Socrates, philosophy as a way of life not only did not include a way of living that makes philosophizing for the most part a preliminary, and then sets it to the periphery. It positively ruled it out. For that reason, if one is to be true to the original, Socratic outlook on philosophy, one should exclude from consideration as illustrating philosophy as a way of life any

Socrates and Philosophy as a Way of Life

31

way of life, however much based on philosophical principles (Stoic or Epicurean ones, for example), that does not involve constant engagement in the activities of argument and analysis proper to philosophy itself. Such lives of ‘‘applied philosophy’’, including Cato’s, deserve separate classification and consideration. Let me turn now to consider the two sides that I said there are to Socrates’ reasons for thinking that proper attention to the good of one’s soul requires constant philosophizing. I said that for Socrates the good condition of the soul is entirely a matter of one’s ability, unfalteringly and with an inexhaustible thoroughness, to understand, explain, and defend by argument and analysis, to others and to oneself, one’s own values and commitments. This good consists, therefore, in the perfection of the ability to understand and argue well about values. It is the condition in which that ability is made absolutely secure and permanent. That is what Socrates understands by wisdom. But, as I mentioned, it seemed to Socrates, and surely not unreasonably so, that human beings could never achieve this condition. One must be prepared to face up to challenges that anyone might pose against any of one’s ideas that one had carefully examined and felt quite persuaded of. But how can a human being, with all the limitations there manifestly are on human intelligence and individual human experience, ever be confident that they had already fully considered and rejected every possible challenge, or that any remaining ones, if ever raised, could immediately be turned aside on the basis of adequate reasons already found among those on the basis of which one has become confident of the idea in question? In fact, how could anyone be absolutely sure (given their current state of belief on some question) that no unheralded future experience of their own or others could possibly provide what they ought rationally to consider, perhaps only for a moment, when presented with it, to be an acceptable basis for doubting its truth? No human being can already know about all possible relevant experiences people might have, and all grounds of possible objection to any view on these matters that they might have arrived at. Knowledge or wisdom must be unshakable; but it would be extremely rash of anyone—and so, it would go counter to reason itself—to claim to know anything about human good and human bad in that unshakable way. However, the perfection of one’s ability to understand, explain, and defend by argument one’s own values and commitments requires nothing less than that knowledge. Plainly, it is this perfected ability that Socrates is denying when he denies having any knowledge (wisdom) himself, and when he doubts whether anyone else does either.¹³ ¹³ One must take Socrates’ denial as intended with the utmost seriousness. It is deeply mistaken to say, as many commentators (including Hadot (2002: 25)) continue to do, that Socrates is ‘‘ironical’’ or feigning in any way when he says repeatedly in the Apology and other dialogues of Plato that he has no wisdom. When he denies that he is wise in the way in which the prominent Athenians were held to be whom he reports in the Apology having questioned about their alleged wisdom (21b–23b), Socrates is registering his considered view that he is not in a position to withstand any and all such requests for explanation and defense of any of his own values and commitments. This

32

John M. Cooper

These pragmatic considerations give Socrates adequate grounds for holding that, as it seems, human beings cannot in fact achieve the soul’s final good, so that a full concern for that good as the most important of all things of value to us requires us to live the life of self-examination that he himself led. But there is another side to the matter as well, one involving religious considerations. As part of his strategy in rebutting the charges of impiety brought against him, Plato has the Socrates of his Apology piously link these limitations of human beings to a positive attribution to the gods, and to Apollo in particular, of this knowledge—wisdom—about human beings and human life (23a–b). Only god, or the divine, he says, is wise (sophos).¹⁴ Socrates, and human beings at their best, are mere philo-sophers, lovers of wisdom: they are seekers after wisdom or knowledge, who will never possess it. Thus Socrates is able to align his pragmatic recognition of the limitations on the human capacity for knowledge with a pious and appropriately modest deference, and consequent due reverence, towards the greater powers of the gods. This gives him a second reason for insisting on the human incapacity for full knowledge: pious deference to and reverence for the gods requires it. There is a positive side as well to the piety to which Socrates lays claim in his defense. On his account, philosophers are people who, by unrelentingly applying their minds and developing their powers of thought and argument, can come closer and closer to possessing the wisdom that belongs to divinity by its very nature—without, however, ever reaching it. This they do day by day and as much of the day as they can manage, by discussing, debating, examining themselves and others, on the all-important Socratic questions about human good and human life. All that intense, devoted activity of philosophizing would be carried out with a view to piously assimilating oneself, as a rational being, to the higher, perfect rationality (wisdom) of god. These pious links forged by Socrates between philosophy—philosophy as a way of life—and the active (but duly modest and reverent) pursuit of the wisdom that is reserved to god alone are, of course, just the first stage in a long historical development. In their distinctive (and widely divergent) ways, each quite different from Socrates’ emphasis on continuous discussion and reevaluation of one’s own ideas, Plato and late Platonists, Aristotle, the Stoics, even Epicurus—all the main philosophical schools and movements of antiquity except, of course, the Skeptics, is what he thought any such allegation of wisdom would attribute to its possessor. He was aware that he might well, if questioned extensively enough, contradict himself or fall into implausible and unreasonable-looking assertions as to which he could not always immediately dispel or explain away that appearance. It was the inability of these alleged wise people to sustain that test that he thought entitled him to declare them not wise at all. So, likewise, himself. ¹⁴ See, to similar effect, Socrates’ remarks in Prt. 344b ff., in attributing approvingly to Simonides the thought that god alone can be good, while even the best human being can only at times become good, by thinking and acting in the knowledgeable way on some occasion or for some period. Human beings can at best approximate to god’s condition of permanent knowledge by sometimes seeing and understanding a situation aright—but then, later, in a similar or a different situation, they will fail to do that. See also Phdr. 248a, 278d.

Socrates and Philosophy as a Way of Life

33

Academic and Pyrrhonian—all make the philosophical life the best for human beings by also making it the most divine or godlike.

IV There is a further consequence of Socrates’ denial of human wisdom. In Plato’s Socratic dialogues (with the partial exception of the Apology, which is not really a dialogue at all) Socrates always takes the role of questioner of others’ opinions. He never presents, argues for, or explains directly any of his own philosophical views.¹⁵ He never allows himself to be questioned by someone else about his own opinions on what is good or bad—or his philosophical reasons for holding them.¹⁶ There is an interlude in his examination of the sophist Protagoras in the Protagoras that one might regard as an exception, but there Socrates agrees to be questioned on the interpretation of a poem of Simonides, not on his own substantive views on moral topics.¹⁷ In that case, as everywhere else where Socrates ¹⁵ One of the main differences between the Platonic and the Xenophontic Socrates is that in Xenophon, while in philosophical dialogues Socrates does most often appear as questioner (and cf. Mem. I 2. 36 where an enemy notes his ‘‘habit’’ of asking questions when he already knows the answer), sometimes he is the answerer instead. See Mem. III 8, where Aristippus takes his turn putting Socrates’ opinions to the test (ἐλέγχειν), as Socrates had subjected Aristippus previously to such scrutiny; and in IV 4, a conversation between Socrates and Hippias, after Hippias objects to Socrates’ practice of questioning others without stating opinions of his own (§9), Socrates launches into an account of justice, and Hippias then performs the Socratic role of critical examiner of this account. (Socrates is also depicted frequently just explaining to his young men various things about life and about the virtues, and encouraging them outright toward specific behaviors: see Mem. I 5, I 7, II 4, III 9.) Xenophon does not make any part of his Socrates’ persona a principled avoidance of taking a position in argument and opening oneself to explication and defense of the position taken. ¹⁶ Thrasymachus in Resp. I refers with contempt and outrage to this feature of Socratic discussions, 336b–337a: Socrates does nothing but ask questions and then refute the answers, just to show his own superiority, but he won’t answer questions himself—something much harder than to ask them—and will use any ploy (such as ironical praise of the greater competence of the proposed questioner, who therefore ought to speak instead) so as to avoid having to do that. ¹⁷ Socrates is also made to answer questions at several points in the free-for-all eristic-refutatory display in the Euthydemus (283a–e, 287d–288a, 293b–e, 295b–298a, 300e–303a), but these do not concern his moral opinions and are not aimed at raising serious, substantive difficulties that anyone with those views would have to face: the whole point of the sophists’ effort is just to force the answerer into a verbal contradiction. In addition, Socrates begins his discussion with Polus in the Gorgias by offering Polus the opportunity to be the questioner (462a1–b3). But this is just a device allowing Socrates the opportunity to explain a view of his own as to what rhetoric is—it is a form of ‘‘flattery’’, aiming to give its audiences pleasure, no actual craft (τέχνη) aiming to aid or benefit or improve them in any way (463a–466a). Having expounded this view in response to Polus’ questions, he again resumes his accustomed role of questioner (466e). But in fact, just because Socrates in the Gorgias is glad to put forward in this positive way something he holds and is prepared to defend in dialectical argument as answerer, one can doubt whether the Socrates of the Gorgias is a pure instance of the Socrates we are discussing—the character in such Socratic works of Plato’s as Apology, Laches, Protagoras, and so on, the one whose practices and underlying views played a role in the later development of philosophy. In the first century  Quintilian says roundly that in putting forward just this theory in the Gorgias ‘‘Socrates’’ is in fact a mere mouthpiece for Plato (cuius persona videtur Plato significare quid sentiat (Inst. II 15. 26)).

34

John M. Cooper

discusses such questions, readers have a considerable philosophical-interpretative inquiry of their own to make before they can hazard any claim about Socrates’ own views and, especially, about his philosophical reasons for holding them. This is true even if, as I think, such inquiry can reveal a fairly extensive, quite plausible, underlying Socratic theory about why the soul and its good condition is the most important of all human values, and why the virtues, including justice, are elements in that good condition. Socrates never explains and argues for that theory, however. Moreover, in Plato’s Socratic dialogues he never clearly explains his reasons for this abstention. But they are not far to seek, and they seem to be reasons of principle. The Theaetetus is not a Socratic dialogue (as I am understanding that term), but in one passage the Socrates-character does refer to the ‘‘common reproach against’’ him—that ‘‘I am always asking questions of other people but never myself express any views¹⁸ about anything’’ (150c4–7). This is a reproach naturally addressed to the Socrates of Plato’s Socratic dialogues. The Socrates of the Theaetetus says that it is true. And he goes on to offer an elaborate explanation and defense of this behavior, by describing himself as a professional, or semi-professional, midwife for the mind or soul, helping young men give birth to their own ideas, but barred from having any ideas of his own to put forward. The god, he says, ‘‘compels me to act as a midwife, but has blocked me from giving birth’’. We should, however, leave aside the theory of Socratic midwifery in considering the reasons that Socrates (the Socrates of Plato’s Socratic dialogues) has for his practices. As Myles Burnyeat (1977) has illuminatingly emphasized, Plato gives a broad hint that Socratic midwifery is a late addition of his own to the Socratic image. He has Theaetetus say roundly that he has never before heard of it (though he has heard lots else about Socrates).¹⁹ Plato reinforces this when he has Socrates ¹⁸ I translate Burnet’s text here (the revised OCT of Winifred Hicken agrees); but in fact only one of the three families of manuscripts recognized by current scholarship has the reading ἀποφαίνομαι (‘‘express … views’’), and then (if I am interpreting correctly the editorial practice followed in the revised OCT text, as explained on p. xix) only in two of the members of that family. The other two families (and, apparently, the remaining members of the one just mentioned) have ἀποκρίνομαι (‘‘answer questions’’) instead. Burnet and Hicken apparently chose to follow  W and P because the Anonymous Commentary on the Theaetetus, which dates from the second century  or earlier, perhaps as early as the first century  (see Bastianini and Sedley 1995), quotes Plato’s text with ἀποφαίνομαι, and because Plutarch, not quoting, however, but simply describing Socrates’ practices in Quaest. Plat. 999 f., uses this term. One should also note that, if either of these readings arose by ‘‘correction’’ from the other, ἀποκρίνομαι is by far the more likely to be the interloper—the reference to questioning just before invites a reader to expect a following reference to answering. The text I have translated therefore seems on balance more likely to be what Plato wrote. ¹⁹ He has heard, as Socrates goes on immediately to suggest, that Socrates is well known for reducing the people he discusses things with to pure puzzlement (ἀπορειν), inability any longer to know what to say on the topic. This is a clear reminiscence (and verbal echo) of what Meno says to Socrates in the famous passage of the Meno where he compares Socrates to a torpedo fish that numbs anyone who comes into contact with it (80a). Theaetetus says that he has heard a lot about Socrates’ conversations and the questions he asks, and has been much impressed with those reports (148e).

Socrates and Philosophy as a Way of Life

35

respond that up to the time of the Theaetetus conversation (at the very end of his life) no one has had any idea that midwifery was his trade (149a5–10). And, as Socrates explains and exemplifies his practice of midwifery in the Theaetetus, it centrally involves very extensive, positive philosophical theory construction of his own—indeed, on topics in metaphysics and physics and theory of knowledge which we know Socrates (the historical person, refracted through the Socratic writings) never engaged in.²⁰ But in fact, before launching into this account of his midwifery, Socrates has already offered a preliminary reason why he (the Socrates of the Socratic dialogues) has limited himself to asking questions of others and has not proposed, explained, and argued for opinions of his own, about which others might then question him. Having said, ‘‘I am always asking questions of other people but never myself express any views about anything,’’ he adds, ‘‘because I have nothing wise in me’’ (διὰ τὸ μηδὲν ἔχειν σοφόν). This is the reason we need to consider if we wish to understand that Socrates’ forbearance from taking the answerer’s position in a philosophical discussion. Evidently, Socrates thought that those who did set forth opinions, thereby opening themselves to having to answer any of the sorts of questions that Socrates would go on to ask, thereby implied one or the other of two things. Either they thought they already knew the answer (were wise on the subject)—such were the people he went to so as to test the oracle (see Ap. 21b ff.) as well as some of his interlocutors in Plato’s Socratic dialogues: Euthyphro, perhaps Critias in the Charmides, the generals in Laches, Protagoras, the rhetoricians in Gorgias, Meno, Hippias in his two dialogues, Ion. Or else they were at least ready to present an answer as something they were prepared to assert with an advance commitment to its rational acceptability and defensibility. They imply that they have thought about (maybe then for the first time), are convinced, and so are prepared to defend and explore the consequences of their answer. This would be the situation of his young men when he engaged any of them in a discussion, asking for their opinion on some matter to do with human life, and exploring with them the philosophical difficulties that then arise.²¹ We see this exemplified by others ²⁰ In dealing with Theaetetus’ first proposal about the nature of knowledge (152a–186e), Socrates specially emphasizes two functions of his midwife’s art. First, he is able to delay or speed up the process of delivery of the offspring into the final light of day. It is not born yet when Theaetetus announces the initial proposal at 151e (‘‘a person who knows something perceives what he knows’’, i.e. is conscious, is aware, of it), but requires the midwife’s delaying skills before it can be declared ‘‘brought to birth’’ (at 160e). In the meantime it undergoes midwife-supplied development through discussions of Protagorean and Heraclitean theories and an elaborate account of what ‘‘perception’’ itself should be construed to mean. Only then, secondly, does Socrates begin the elaborate series of intricate arguments that are aimed at assessing the philosophical value of the proposal so fleshed out—the process of discovering whether it is a ‘‘genuine’’ and true birth, or a phantom, a wind-egg, and falsehood (161a–186e). Socrates’ discussions in the Socratic dialogues include no analogue of either of these processes. ²¹ It is noteworthy that in his defense in the Apology Socrates suppresses mention of this part of his philosophical activities. About the young men who flocked around him, he only says that they witnessed his examinations and refutations of the alleged or self-professed wise, enjoyed them

36

John M. Cooper

of his interlocutors in Plato’s dialogues, not all of them young men: Crito and Charmides in their dialogues, Alcibiades, Lysis, Clinias in the Euthydemus, Polemarchus in Republic I. Socrates did not think himself in either of these positions. Certainly he had thought hard and long about all the matters he questioned others about (in part precisely through prior such examinations). Certainly he had views on these subjects, well formed and sometimes quite unusual or even provocative ones—as we see from the Apology. (I return to this below, sect. V.) But, as we have already seen, he was not convinced of any opinions he held, however much he was committed to living on their basis, so long as he found no good reason to doubt them. His attitude was one of openness to the need for further thought before one could declare any of them with final certainty. For one with such an attitude, it could very reasonably seem that the right stance to take in philosophical argument was the one that Socrates did assume: that of questioning others and continuing to think and rethink one’s own opinions in the course, and as a result, of doing so. Since he is not actually convinced about anything, shifting to the position of answerer would carry intellectual (and, given the subject of the discussion, moral) discomfort. By doing so he would either present himself as wise, or as convinced enough of something to be willing to assert it with advance commitment to its being true, and with the assurance that he would be able to defend it with complete success. There are two further, less important, considerations. First, Socrates reports in the Apology that those who were witnesses to his examinations of the wise regularly inferred that he must be (positively) wise himself on all the matters on which he refuted the views of others (Ap. 23a). If, then, with this reputation, he put forward views of his own, and undertook to explain and defend them before others, there would be a great danger that others, especially the young men, would just take his word for the truth. They would not adopt the attitude of openness and need for further thought that he regarded as appropriate, and would not think the matter through for themselves. Yet that, if anything, is greatly, and then went off on their own to try their hand at similar refutations of others (23c, 33c)—including perhaps their own parents, though Socrates understandably leaves that unsaid. The impression he leaves in the Apology is that his philosophical activities consisted of just two things: refutations of false claims to wisdom and exhortations to all and sundry to care more about their souls, and about virtue, than about wealth, glory, political power, and position, and the like. It is understandable that, in defending himself against Meletus’ charge of impiety and corruption, Socrates would suppress mention of discussions on moral questions with the young men. Readers of the Apology need to correct for this omission by taking account of the evidence of many dialogues of Plato, as specified in the next sentence of my main text, about the discussions that Socrates held with older friends of his and his young men. In those discussions his interlocutor was by no means held either by others or by himself to be a wise man. It is natural to think that the daily discussions about virtue that Socrates refers to when he makes his comments about the unexamined life (Ap. 38a, discussed above) were primarily such discussions with friends, young and old (even if they also included refutations of puffed-up, self-satisfied allegedly wise people). So one finds oblique mention of such discussions even in the Apology, if one reads carefully.

Socrates and Philosophy as a Way of Life

37

what he wanted to teach them. In addition, no questioner who might put him and his views to the test would be likely to raise sufficiently the difficulties that (he thinks) would need to be dealt with, or to bring out at all the real complexities of the position—things that Socrates was so expert himself at doing in the examination of others’ opinions. So any one hearing the exchange might not appreciate adequately the complexities and the difficulties to be surmounted before one could become convinced of Socrates’ own views. In this way, too, he would run the risk that others would wrongly take his word for the truth. Since the crux of the matter for Socrates was never to think one was in final possession of the answers, it was therefore all the more important that he should avoid the answerer’s role in a discussion and examination, and adopt always only that of the questioner—the seeker after the truth. Readers of the dialogues sometimes (like Thrasymachus in Republic I) think Socrates’ refusal to present moral views of his own for scrutiny and examination by another person is nothing more than wily avoidance of the much more difficult answerer’s position, and a desire to relish his own superiority when he defeats others who do assume it. The considerations I have just advanced show, however, that Socrates’ stance in this respect, and his behavior, are not so merely self-serving as they may appear. They have substantial backing in his own philosophical conceptions and ambitions. Strictly speaking, then, in the kind of question-and-answer, ‘‘dialectical’’ discussion that Socrates devotes himself to in living the life of philosophy, all the opinions expressed—all the commitments undertaken—are the answerer’s, none the questioner’s. It is answerers who initially advance a thesis. It is answerers who, having presumptively agreed to hold a discussion with Socrates on that basis, have to respond to subsequent questions by either accepting or rejecting some further premise offered to them for their consideration by the questioner. And if some consequence is deduced from these assembled answers that is incompatible or at any rate not in good harmony with the initial thesis, it is the answerer’s position that suffers. His initial thesis is shown not to consort well with his additional views as elicited by the questioner. The answerer himself, as a personal matter, also suffers. He shows himself not wise, and this can be an embarrassment if he has claimed moral and intellectual expertise, whether by advertising himself as a sophist or in some other way. Even if he has made no such claim (as is the case with many of Socrates’ interlocutors in both Xenophon’s and Plato’s Socratic dialogues), there can be embarrassment. People professing to be particularly devoted to friendship or moderation or some other important good, and convinced enough about it and its value to declare their views on the subject as part of such a discussion, ought to have paid sufficient attention to the object of their alleged devotion to have something satisfactory to say about it. They ought not to fall into contradiction or unalleviated gross implausibility in trying to discuss it. In either case, on Socrates’ views about what is the most important thing of value in any human life, the interlocutor will have shown himself not

38

John M. Cooper

to care sufficiently for the thing of the greatest value, his own soul. Instead of answering questions when pressed to do so by Socrates, he ought to make it his first task in life to engage, like Socrates, as questioner in such discussions, in an unrelenting effort to come to a better understanding, as comprehensive and clear as possible, of all of the vast realm of human values. Socrates is persuaded, as we have seen, that proper care of the soul demands nothing less. In any such discussion one would normally expect the answerer to say what he really thinks, not only in stating his initial conviction, which Socrates then goes on to examine, but also in responding to subsequent questioning. One would, in general, expect the answerer to consider seriously what the best, most reasonable, thing to say is in response to any question that arises for someone with their convinced belief in the thesis they have set up. At any rate, that is the most obvious way for Socrates’ questioning to help others, as he says he wants to do, to test themselves and improve their ideas, and so their soul. This normal presumption applies in Socrates’ discussions in Xenophon and Plato. Indeed, Socrates sometimes insists that the answerer’s answers to subsequent questions represent the answerer’s (current) actual beliefs. This might be particularly important when he is conducting a discussion with some puffed-up claimant to wisdom, whom he is trying to humiliate. But strict insistence on ‘‘saying what you believe’’ in answering these subsequent questions is not, in general, necessary for his purposes. This is particularly clear when his fellow discussant is an earnest young man who might not have definite beliefs on all the questions Socrates asks. He could nonetheless be encouraged to think through the pros and cons of one or another attractive-looking initial response, when confronted for the first time with the question. He could learn a lot from that. But it applies also, most of the time, when it is a reputed wise person. Someone is shown not to be wise, and helpfully humiliated, even when (without ‘‘saying what he believes’’) he shows himself not sufficiently knowledgeable about matters of values (and logic) to avoid agreeing to things in a discussion that either contradict his original thesis or lead to obviously implausible consequences without having the intellectual resources to defend them nonetheless. Hence, Socrates’ expectation, most of the time, that his interlocutor will answer all subsequent questions, after setting up a thesis, by saying what they believe establishes only a normal presumption, not a rule of any sort for Socratic dialectic. Here it is important to realize that, despite what Gregory Vlastos claims, Socrates never insists on ‘‘saying what you believe’’ as a general rule governing the conduct of the answerer in such a ‘‘dialectical’’ discussion. In the rare cases when he does insist, it concerns only such subsequent questions, and always under some very special dialectical circumstance.²² Every Socratic discussion, as we have ²² See Vlastos (1994), who claims (see pp. 7 ff.) that for success in his inquiries Socrates requires the interlocutor to ‘‘say what you believe’’. I do not have room to argue the point here, but examination of the contexts of all the passages to which Vlastos refers to illustrate Socrates’

Socrates and Philosophy as a Way of Life

39

seen, begins from a thesis concerning moral values and how best to lead a human life stated by the interlocutor as something he is convinced of. If one is aiming at coming as close as one can to achieving wisdom oneself, as Socrates is, one ought not to limit oneself in examining such theses to engaging with further opinions about these or other human values of specific persons who are committed to their truth. One needs to cast one’s net wider. One needs to allow the exploratory consideration of all kinds of ideas, whether held decisively by someone you can find to engage with in discussion or not. We do not know, and Socrates does not make a point of insisting, whether Meno firmly believed any of the things he says in attempting to explain and defend his Gorgianic thesis about virtue. The same applies to all Socrates’ other interlocutors (except in the special situations I have referred to). It does not matter for Socrates’ purposes whether they believed what they said or not in answering his subsequent questions after they have stated their thesis. In principle, the same goes also for stating their initial theses. It does seem, however, that Socrates’ interlocutors are always represented as holding convictions about, for example, virtue or some particular virtue—and, as I have explained, Socrates does think that anyone who puts forward a thesis for serious dialectical discussion implies that that represents a conviction of his. Accordingly, the main thesis, or succession of theses, they advance on their topic is presented as a full belief of theirs. However that may be, in all his discussions Socrates was testing, while expanding, the range and adequacy of his own grasp of the moral issues on both sides of difficult and debated questions about human nature, human virtue, and the particular virtues of courage, temperance, justice, adherence to this principle shows that in fact they occur in special situations, particular features of which explain Socrates’ (reasonably) insisting on it in those situations, but only in them. Take, e.g., the Crito, where Socrates and Crito are testing the value of the latter’s advice that he should be allowed to bribe Socrates’ way out of prison. Socrates wants this to be a common decision, after common evaluation of the reasons on both sides (46b3–4, c2–3; see also 48d8–e1). So it is quite in order and perfectly natural that, when Socrates reaches what is for him the crucial point in the argument against Crito’s plan (49c–d), he takes good care to make sure that Crito really agrees with him and does not just say he does—especially since, as Socrates emphasizes, this is a point that only a few people actually do accept. Though the special features vary from case to case, the same applies to the passages of the Protagoras (331c4–d1), Gorgias (500b6–d4), and Republic I (346a4) to which Vlastos refers. In the Republic it is in order to make sure that Thrasymachus cannot accuse him again of verbal trickery in his argument that Socrates takes care to get him to declare outright that what he is agreeing to about ‘‘ruling’’ in other contexts besides politics is what he really believes, i.e what he really means to assert. This, he says, is needed if they are ‘‘to come to some definite conclusion’’. In the Gorgias it is to avoid a repeat of Callicles’ belated claim that something he had earlier said that got him into trouble was said in jest. As for Protagoras, Socrates actually goes out of his way only two pages after Vlastos’s passage, at 333c, to say that he does not need Protagoras to say what he believes. It is enough that Protagoras makes a serious effort to follow out the consequences of a view that he chooses to attribute not to himself but to ‘‘the many’’, even though Socrates may suspect that it is in fact Protagoras’ view too, and to see what its consequences are for his own original thesis. Vlastos has to call this a case of ‘‘waiving’’ the requirement (pp. 10–11). In fact, there was no such requirement to be waived. Earlier Socrates required only that Protagoras be serious about his role in the discussion: namely, he must speak for himself in explaining and defending and following out the consequences of the view he has previously espoused.

40

John M. Cooper

piety, wisdom, and the rest, which his daily discussions and debates concerned. He was also assisting others to advance their own understanding through the same process. For these purposes, there was no need for any rule to ‘‘say what you believe’’, and Socrates did not think there was.

V In my preliminary sketch in section II I mentioned one aspect of philosophy as a way of life for Socrates that I have left aside till now. So far I have been concerned primarily with the philosophical activities of argument and discussion which, in one way, lie at the center of the Socratic way of life. But in another way, the center is occupied by a second consequence of his fundamental view about the overriding value to any human being of their soul and its good condition. This is Socrates’ commitment to follow in all the actions of daily life, as well as in his performance of citizenly duties, and indeed in any circumstance calling for difficult or stressful decision, whatever definite ideas about human values he may have arrived at (so far) in his philosophical reflections and investigations. The first of these ideas, of course, is his conception of the overriding value of the human soul and its good condition—wisdom or total knowledge about the whole range of things of value for human beings, and their relative rankings or importance. But he also thinks that wisdom is the same thing, somehow, as a philosophically explained and articulated version of the traditional Greek virtues of justice, piety, temperance, and courage. Accordingly, so far as concerns the actions of daily life and all one’s practical decisions, for Socrates—as in fact for the whole subsequent Greek tradition of philosophy as a way of life—living one’s philosophy meant living on the basis of a philosophically grounded conception of justice, piety, temperance, and courage as fundamental goods for a human being. In the passage of the Apology I referred to above (beginning of sect. III), where Socrates announces his refusal to give up philosophizing in public even if offered acquittal in return, he first identifies ‘‘wisdom and truth’’ as the best possible state of the soul (29e1–3).²³ But he goes on to describe the (hypothetical) person whose soul was in this state as having acquired ‘‘virtue’’ (ἀρετή). Thus Socrates can repeat what he has just said about the greater importance of wisdom and truth than wealth, glory, and power, but this time with virtue taking their place: I go around doing nothing but persuading both young and old among you not to concern yourselves with your body or your money in preference to or as strongly as with the best possible state of your soul, saying to you: Virtue does not come from having money, ²³ The repeated καί here seems clearly epexegetic: wisdom, i.e. truth.

Socrates and Philosophy as a Way of Life

41

but from having virtue money and everything else become good for human beings, both individually and collectively. (30a8–b4)²⁴

‘‘Virtue’’ here, as we see from Socrates’ own emphasis just before (29b6–7) on the badness of unjust action, refers to justice and the other recognized virtues. So, for Socrates, the commitment to the value of wisdom, which expresses itself in a devotion to philosophical examination and self-examination, will also express itself in the resolute avoidance of any vicious act, whether it is one of injustice, or cowardice, or weak self-indulgence. Any and all such behavior will expose one as accepting a wrong system of values—a system that ranks the good state of the soul (wisdom, and these other virtues) lower in value than its true ranking, by expressing a preference for whatever alleged other good one acts to obtain or achieve in one’s unjust or cowardly or self-indulgent action. I will not discuss here how and why Socrates does think that wisdom is also, or somehow implies, justice and courage and the other virtues. His arguments in the Protagoras to show the identity of wisdom and the other recognized virtues rely on ideas about the particular virtues that are remote from anything that could plausibly be thought a philosophical basis for maintaining this connection. And with the sole exception of Republic I—a very special case—his discussions in other dialogues seem simply to assume that justice and courage are, like wisdom, virtues, and as such part of the good condition of the soul. In his discussions he explores other aspects of them. The Apology certainly suggests that he possesses some thought-out philosophical rationale to support these commitments. But since, as questioner in his discussions, we never see him bringing it clearly to bear in his arguments, we are left with little basis for attempting to work out on our own what it was. A good deal of subsequent Greek moral theory is devoted to developing philosophical analyses of the human soul and of human nature that are intended to ground these Socratic ideas. This effort begins already in Republic I, which is why I said just now that this is a very special case of a Socratic dialogue. ²⁴ Translation based on Grube–Reeve in Cooper 1997. The translation of the last clause follows Burnet (1924) in construing the Greek (n. ad 30b3). On another construction it would be translated: ‘‘Virtue does not come from having money, but from virtue come money and all other public and private goods for human beings.’’ This second construction might seem the easier and more natural (though I agree with Burnyeat (2003), that the other construction has parallels). Burnet’s reason for not accepting it was a legitimate wish not to have Socrates say or imply that if you are virtuous (just or courageous, e.g.), then you can be sure of making a lot of money, being healthy, etc. Socrates cannot have meant to say he went around telling that to the Athenians. His main point is manifestly that the fundamental value of wisdom and virtue is not in any way derivative from the values of these other things. If it were, they would rank equally or ahead of it in the true system of values; but Socrates’ claim is that wisdom and virtue are in sole first place in the rank order of values. In fact, however, even on the second construction Socrates can, with strain, be interpreted as saying much the same thing as on the other. On the second construction, by calling money, etc., ‘‘goods’’ in saying that they come from virtue, he might be taken to mean, not that these things reliably do come from virtue, but that when, and only when, they do come from virtue, they become actually good for the individual or city. Without virtue they are not goods at all. That is, insofar as these are good, they must come from virtue. Burnet’s construal is preferable because it does not require such strained interpretation to make acceptable sense of the Greek.

42

John M. Cooper

It continues in the remainder of the Republic and in other dialogues of Plato. Both Aristotle and the Stoics add their own distinctive thoughts in this same effort of philosophical theory construction. For our purposes in offering an account of Socratic philosophy as a way of life, it suffices to point out that on Socrates’ view, however we attempt to articulate and explain it, the life of reason, of philosophy, will include a fundamental devotion to the social virtues recognized and prized in Athens and other Greek cities (and, of course, among ourselves too): in particular justice, but also physical and moral courage, honesty, self-control, respect for others, friendly and supportive attitudes to all with whom one might come into contact, and so on. A successful life of reason and philosophy will therefore also be a life of moral virtue at its highest. The truest philosopher will also be the most morally, socially virtuous person—precisely because only a philosopher can have achieved the reasoned, argued understanding of just why those (or rather, some philosophically improved version of them) really are part of the good condition of the soul. VI If, then, we seek the origins of the traditional Greek conception of philosophy as a way of life, it appears that it is to Socrates that we must look. At this first moment, the philosophical life—philosophy itself as a way of life—simply was the life that Socrates is presented as living in Plato’s (and the other Socratics’) dialogues. It was a life devoted to the care of one’s soul. It consisted in constant and continued philosophizing: that is to say, in discussion and critical examination of one’s own and others’ opinions on questions of human life, and how best to lead it. It was a life aimed at making reason itself, not tradition or habit or what ‘‘feels good’’ or ‘‘feels right’’ to you, one’s guide in all that one does, and in one’s manner and methods of living. It was also a life deeply committed to living virtuously, or as nearly so as a mere human being can manage, by upholding philosophically examined and defended standards of all the traditional virtues of justice, piety, temperance, and courage in all one’s actions. REFERENCES Bastianini, G., and D. N. Sedley (1995) (trans.), Corpus dei papiri filosofici greci e latini, part 3, 227–562. Florence. Burkert, W. (1960), ‘‘Platon oder Pythagoras? Zum Ursprung des Wortes ‘Philosophie’ ’’, Hermes, 88: 159–77. Burnet, J. (1924) (ed.), Plato’s Euthyphro, Apology of Socrates and Crito. Oxford. Burnyeat, M. F. (1977), ‘‘Socratic Midwifery, Platonic Inspiration’’, Bulletin of the Institute of Classical Studies, 24: 7–16. (2003), ‘‘Apology 30b 2–4: Socrates, Money, and the Grammar of γίγνεσθαι’’, Journal of Hellenic Studies, 123: 1–25.

Socrates and Philosophy as a Way of Life

43

Cooper, John M. (1997) (ed.), Plato: Complete Works. Indianapolis. (2004a), ‘‘Arcesilaus: Socratic and Skeptic’’, in Cooper (2004b), 81–103. (2004b), Knowledge, Nature, and the Good. Princeton. (2004c), ‘‘Method and Science in On Ancient Medicine’’, in Cooper (2004b), 3–42. Hadot, Pierre (1995), Philosophy as a Way of Life, trans. Michael Chase, ed. Arnold I. Davidson. Oxford. 1st published in French, 1981, 2nd edn. 1987. (1998), The Inner Citadel: The Meditations of Marcus Aurelius, trans. Michael Chase. Cambridge, Mass. 1st published in French, 1992. (2002), What is Ancient Philosophy?, trans. Michael Chase. Cambridge, Mass. 1st published in French, 1995. Jaeger, W. (1948), ‘‘On the Origin and Cycle of the Philosophic Ideal of Life’’ (1928), trans. Richard Robinson, in Jaeger, Aristotle: Fundamentals of the History of His Development (Oxford), Appendix II. Jones, W. H. S. (1923) (ed. and trans.), Hippocrates. Cambridge, Mass. Kahn, C. H. (1996), Plato and the Socratic Dialogue. Cambridge. Lynch, John P. (1972), Aristotle’s School. Berkeley. Nehamas, Alexander (1998), The Art of Living: Socratic Reflections from Plato to Foucault. Berkeley. Quintilian (1920), Institutio Oratoria, trans. H. E. Butler, Loeb Classical Library, 4 vols. Cambridge, Mass. Seneca (1928), De Constantia Sapientis, ed. and trans. J. W. Basore, Loeb Classical Library. Cambridge, Mass. Vlastos, G. (1994), ‘‘The Socratic Elenchus: Method is All’’, in Vlastos, Socratic Studies, ed. Myles Burnyeat (Cambridge), 1–37. Yunis, Harvey (2005), ‘‘The Rhetoric of Law in Fourth-Century Athens’’, in M. Gagarin and D. Cohen (eds.), The Cambridge Companion to Ancient Greek Law (Cambridge), 191–208.

3 Virtue as the Love of Knowledge in Plato’s Symposium and Republic Melissa Lane

I . I N T RO D U C T I O N Socrates confronted Plato with a paradox. Convinced that Socrates was the best man of his generation (Phd. 118a15–17), perhaps the best man who had ever lived so far, Plato had to face the fact that, on his own admission, Socrates did not have what he himself considered necessary in order to be what he was. If knowledge of aret¯e is required for having aret¯e and so for living well, then Socrates, who lacked that knowledge, could not have been virtuous and could not have lived well. Yet he was, and he did.

So Alexander Nehamas states a central problem about Socrates (Nehamas 1998: 68). That this poses a problem is evident, not only to later commentators,¹ but even within Plato’s Symposium. There, Alcibiades is sufficiently troubled by the problem to propound a solution at some length: that Socrates does possess knowledge, though he conceals it and feigns ignorance. Nehamas, for his part, tacitly rejects Alcibiades’ diagnosis in holding that no adequate solution is offered I am grateful to Dominic Scott for his acute editorial eye and pen, and to Amber Carpenter, Stephen Halliwell, Antony Hatzistavrou, Shimon Malin, Malcolm Schofield, Frisbee Sheffield, and participants in seminar discussions at Colgate University and the University of St Andrews, for discussion of earlier versions of this chapter. I owe a special debt to Malcolm for directing my attention to Republic VI, though he does not necessarily agree with what I have made of it. I am also grateful to Amy Price and the Centre for History and Economics, King’s College Cambridge, for investment in proofreading; to the Cambridge University Faculty of Classics for affording me special access to its Library, and to participants in its May Week/Laurence Seminar 2006 for discussion of the De Officiis passage quoted herein. In this context I should also like to express a deep general gratitude to Myles Burnyeat, whose teaching (together with that of Malcolm Schofield) informs this chapter as it does everything that I have written about Plato. ¹ As observed in Lane 2001: 13–14, the problem was especially widely noticed in the late Renaissance and the subsequent Enlightenment debates about reason, virtue, and scepticism. For example, Montaigne argued that Socrates may have simply been naturally wise, and Rousseau that he was virtuous out of ignorance; while an opposing line taken by Diderot was that he actually did possess knowledge, perhaps even to the point of obnoxious pedantry.

Virtue as the Love of Knowledge

45

in Plato, because Plato was in fact unable to solve the problem: at bottom, he could not explain Socrates. Nehamas is right to reject Alcibiades’ proposed solution, which is flawed. However, it is argued here that Plato does offer a viable solution, one which he actually proposes in the voice of Socrates in Republic VI (485a4–487a8). In this section of the Republic, we find a discussion of the qualities possessed by a nature most suited to philosophy (a ‘natural philosopher’): justice, temperance, courage, magnanimity, and many others. Such qualities are not virtues in the full sense: they are preconditions for acquiring philosophical knowledge and hence true virtue. Instead, I shall call them ‘natural virtues’, as they are the virtues of the philosopher’s nature before it is perfected by education.² On this account, natural virtues originate with a primitive version of temperance, which arises from a psychological, and even physiological, drive to learn. This drive is in fact a form of love, and it is so powerful as to exert what I call (following an image in Republic VI) a hydraulic effect: psychic energy flows into this love, depriving other desires of the energy to oppose or distract one from the desire to learn.³ Originating with this natural and dispositional form of temperance, the love of learning then develops in a complex interplay between disposition and evaluation, which we will trace below. Here, it suffices to observe that the natural virtues do not presuppose or require that the natural philosopher has already gained the knowledge that she or he seeks. On the contrary: natural virtues are preconditions for the acquisition of knowledge-based virtues because only someone who is naturally abstemious as a result of their love for philosophy could go on to develop the panoply of full virtues. Without the right natural predisposition, they would sooner or later become distracted from the correct path by bodily desires which were not yet controlled by governing knowledge. Further, natural virtues remain part of the psychological make-up of those who ² Aristotle distinguishes the natural virtue or virtues which may be possessed as character traits from birth, from virtue ‘in the strict sense’ (Eth. Nic. 6.13, 1144b 3–4). While this basic contrast is the same, the explanation and use of the idea of natural virtues differs significantly between the Republic and the Nicomachean Ethics. Aristotle suggests that people are likely to have one natural virtue but not another (e.g., to be naturally courageous but not naturally self-disciplined), echoing an idea also found elsewhere in Plato’s thought, in the contrasting courageous and moderate natures of Plt. 306a–308b, and even in Republic VI itself, where it is stressed that passion and quickness on the one hand and stability and reliability on the other are only found together in rare individuals (503c1–d4). The burden of the Republic VI discussion in 485a4-87a8, which is recalled at 503b6–7 and which is the focus of the present chapter, is in contrast to show that the natural virtues of those few people who are natural philosophers will all be compatible and arise together. Plato’s emphasis there on the natural virtue not of the many, but of the few natural philosophers, caused by their natural drive to philosophize, contrasts tellingly with Aristotle’s association of natural virtue with ‘children and brutes [i.e. wild animals]’ (Eth. Nic. 6.13, 1144b 8). ³ ‘Hydraulic’ is used in the same context by Nails (2006: n.36) and Kahn (1987: 97–9). Kahn rejects the hydraulic reading of Resp. VI as incompatible with Platonic desire as rational. But his opposition is too stark: eros here is both part of a continuous flow of psychic and physiological energy, and also oriented to reason. Compare Lear (1990: 39.41) on how Freudian psychic energy is expressed ‘sometimes mentally, sometimes physically’ and ‘can be redirected’.

46

Melissa Lane

go on to develop full virtues. This plays a crucial role in the overall argument of the Republic, as the natural virtues of those with philosophical natures explain why the same people who are suited to do philosophy are also morally suited to rule. Nevertheless, as we shall see, the psychological structure of natural virtue differs from that of full virtue. The many virtues of the fully virtuous person pivot around her knowledge; those of the naturally virtuous person pivot around her temperance or self-discipline (s¯ophrosun¯e). These natural virtues explain how it is that Socrates appears virtuous; presented in Socrates’ own voice in the Republic, they can be seen as his riposte to Alcibiades’ flawed portrait of him in the Symposium. Crucial evidence for this is the fact that Socrates cites himself (496c), as well as Theages, as examples of those with philosophical natures. By implication, he attributes to himself the natural virtues listed at 485–7, and this is what gives us the solution to the Socratic problem that, according to Nehamas, Plato left unsolved. Socrates does not possess full knowledge, but due to his love of learning and the dispositional and then evaluative attitudes which it engenders, he has the natural virtues listed at 485–7. This enables us to explain why his contemporaries, like Alcibiades, thought he had true virtue, no less (remembering that Socrates was the person closest to full virtue whom they ever knew). To bear out these claims, the chapter is framed by the Symposium, with its central portion being an account of Republic VI. We begin by considering, and rejecting, Alcibiades’ proposed programmatic solution to the Socratic problem—that Socrates does have knowledge, and is concealing it. The alternative account which Socrates gives of himself in Republic VI, as one of the natural philosophers possessing the natural virtues, is then detailed, with an eye both to the light it sheds on Socrates and to its function in the context of the unfolding argument of the Republic. We then return to the Symposium to consider the detailed description which Alcibiades gives of Socrates’ behaviour, the actions that allegedly manifest true virtue: for although his solution to the Socratic problem is flawed, his speech contains the materials needed for the correct solution. We will find that, in fact, what he describes is remarkably similar to the natural virtue of Republic VI, 485–7, conforming better to this account than to his own proposed explanation. We conclude by reflecting on the role of natural virtue and the nature of the philosophers in Plato’s work. I I . A LC I B I A D E S ’ S O LU T I O N : D I AG N O S I N G S O C R AT I C CONCEALMENT Our understanding of Alcibiades’ interpretation of Socrates should be shaped ab initio by the fact that it is Alcibiades’ interpretation. This is the Alcibiades who freely admits that he follows Socrates only reluctantly and intermittently, even blocking his ears against him; he who will be discredited as a traitor to Athens, for

Virtue as the Love of Knowledge

47

pursuing personal ambition and thus showing that he has renounced the Socratic path of philosophy. Alcibiades’ mesmerizing effect on Athens, which continued well into the fourth century,⁴ makes it reasonable to assume that Plato portrayed him expecting his audience to bring their knowledge of his bad character to bear in evaluating his assertions. As Andrea Wilson Nightingale counsels, ‘Since Alcibiades’ character distorts his perceptions’—and, we may add, since Plato would have expected readers to expect this—‘we must be especially wary when he claims to reveal the ‘‘inner’’ Socrates.’⁵ Yet this does not mean that we must reject the content of the stories which he tells about Socrates. Plato’s text encourages his readers to adopt a hermeneutic of suspicion towards Alcibiades’ interpretation of Socrates, but not towards the veracity of the incidents that he recounts. For his intimacy with Socrates would lead the reader to expect him to know stories about Socrates which others do not, just as his flaws give reason to expect that he will misinterpret them.⁶ According to Alcibiades, Socrates presents himself as being erotically affected by beautiful youth, and as ignorant. By feigning these outward appearances, he conceals his inner ‘temperance’ (Symp. 216d7): i.e. his contempt for physical beauty, wealth, and all other usual sources of honour (a contempt associated with a belief that the young men present are worth nothing), and his ‘divine and golden and in every way beautiful and wonderful’ inner images which constitute knowledge (Symp. 216e7–217a1, my translation).⁷ Notice, however, that erotic entanglement and ignorance are not the sum total of the outward appearances with which Alcibiades credits Socrates. Instead he goes on to credit him with temperance, courage and other virtues as exhibited in a sequence of remarkable stories to be reviewed in section IV below. The seduction scene apart, all the incidents that will be recounted would have been widely observed by those who served with Socrates in the army, if not by the symposiasts whom Alcibiades ⁴ Duff (1999: 222) discusses fourth-century lawsuits involving Alcibiades’ son in which prosecutors and defendants argued over whether Alcibiades had been a would-be tyrant, including Lysias arguing in the affirmative and Isocrates in the negative. This is indicative of how potent the figure of Alcibiades remained. ⁵ Nightingale 1993: 126; see also her comments on Alcibiades’ bad character (p. 125). I am sympathetic to her argument that Socrates need not be either ignorant or wise, as Alcibiades’ inner/outer dichotomy posits; he was rather a tertium quid, a philosopher who loves but lacks knowledge (p. 127). But for the reasons given in the text, her translation of eir¯oneumenos (Symp. 216e) as ‘ironizing’, and the concomitant centrality she gives to the notion of Socratic irony, is crucially flawed, even though she glosses the verb as ‘wilful deception’ and says that it expresses a flawed Alcibiadean interpretation of Socrates which Plato ‘invites the reader’ to reject (p. 120 n. 28). ⁶ What Gifford (2001: 38) says about characters such as Euthyphro and the characters of Republic I applies to Alcibiades as well: ‘Plato has these characters, as well as Socrates himself, deliver lines which, while meant or taken in one sense by the self-ignorant interlocutor, serve the further function of exhibiting for intelligent readers of the dialogue the gulf that separates how things seem to that character and how they really are.’ ⁷ The ‘divinity’ of these images suffices to identify them as constituting knowledge. As Sheffield (2001: 197) observes, their ‘divinity’ and ‘goldenness’ also echoes Diotima’s description of the Form of Beauty.

48

Melissa Lane

is regaling. They are as much a part of Socrates’ public persona as are erotic entanglement and ignorance. According to Alcibiades, then, the outer shell of Socrates is actually complex. Part of it (erotic entanglement and ignorance) is feigned in order to conceal inner knowledge and virtue. But another, contradictory part of it manifests the putative inner virtues in actions where Socrates is perceived to act virtuously (hence, according to Alcibiades, to be acting on the basis of his hidden inner knowledge and virtues). Indeed, it is only this aspect of the outer shell that makes the diagnosis of hidden inner virtues and knowledge necessary at all. It is only if Alcibiades, his listeners, and Plato’s readers all perceive him as virtuous that they can be puzzled by the problem of how to explain this.⁸ Alcibiades’ explanation of the outer appearances of Socrates thus divides into two parts, corresponding to the two aspects of that outer shell. His outward manifestations of virtue are explained by his concealed inner virtues and knowledge; these are revealed in the stories which we will consider in section IV. Here, we are concerned with the other part of the shell: the profession of ignorance and the appearance of erotic interest in boys. How does Alcibiades explain this? He is most often supposed to be accounting for Socrates’ outer appearance and action in terms of irony, perhaps playful irony. On such views, Socrates is using irony surreptitiously, so that when he professes his ignorance and denies any erotic interest in boys, he should not be taken to mean what he literally says. The irony explanation is based on the claim (or in some cases assumption) that the meaning of eir¯oneuomenos, with which Alcibiades sums up the Silenus comparison, imports irony. In line with this explanation, for example, Nehamas and Woodruff (in Cooper 1997) translate 216e4–5 thus: ‘In public, I tell you, his whole life is one big game—a game of irony.’ However, as I have argued elsewhere,⁹ Alcibiades’ ascription of eir¯oneuomai to Socrates cannot be understood to mean that Socrates is an ironist who uses irony to announce his knowledge and temperance to the world.¹⁰ Rather, what Alcibiades would have ⁸ Rowe (1998: 206, n. to 215a5–222b7), suggests that scepticism about Alcibiades’ charge of Socratic h¯ubris should extend to Alcibiades’ claims about Socrates’ temperance, courage, and wisdom. I disagree, partly because the stories which will establish these are (except for the seduction scene) presented as public knowledge. ⁹ See Lane 2006, where this interpretation is derived from a general argument that everywhere else in Plato and Aristophanes eir¯oneuomai and its cognates mean ‘concealing by feigning’, together with a detailed account of the Symposium showing that there is no reason to depart from the presumption that it has the same meaning in this text. This argument contravenes the many translators who have rendered eir¯oneia as ‘irony’ in the Symposium, and those who have defended it, notably Vlastos (1991 [1987]) and Nehamas (1998: 46–69). See also Lane (forthcoming), for a more general discussion of ‘Socratic irony’. ¹⁰ The same is true of the other use of a word cognate with eir¯oneuomai, when Alcibiades in recounting his failed attempt at seduction describes Socrates’ habitual manner as eir¯onik¯os (Symp. 218d6–7). This scene reveals just how systematically Alcibiades misunderstands Socrates, giving us further grounds to suspect that his interpretation of Socrates may be more generally

Virtue as the Love of Knowledge

49

the symposiasts believe is that Socrates conceals his knowledge and virtue by feigning ignorance and intemperance. So I would translate an expanded version of the same passage thus: He [Socrates] believes all these possessions to be worth nothing and us to be nothing, I tell you; he spends his whole life in concealing [this] by feigning [in the ways described above, professing ignorance and denying erotic interest in boys], including toying with his fellow men. (216e2–5)¹¹

If eir¯oneuomai means here, as everywhere else in Plato, ‘concealing by feigning’, its ascription to Socrates imparts exactly the same structure of an external appearance concealing inner attributes as does the Silenus statue image. In using a version of this term twice, all that Alcibiades is saying is that Socrates is concealing his inner knowledge by feigning ignorance, and his inner temperance by feigning an erotic pursuit of boys. In other words, instead of eir¯oneuomai serving to explain the inner/outer contrast (as the ‘irony’ interpretation would have us believe), it merely reinforces it. Once Alcibiades’ solution is stripped of the unfounded imputation of irony to Socrates, we are forced to recognize that it offers no explanation of Socrates’ purported concealment of his knowledge and virtue. Socrates is alleged to conceal these for no apparent reason, a concealment that Alcibiades has been able to pierce only by happenstance (216e7). At least on the ‘ironic’ interpretation we can see why Socrates should feign ignorance: it is merely a playful way of revealing something important about himself. But if, as philological considerations strongly suggest, Alcibiades is attributing to Socrates something for which he has no obvious motive, and which is in tension with the disavowal of wisdom which Socrates makes in the Apology (20c3, 21b)—and we add to this Alcibiades’ signalled unreliability, especially in philosophical matters, with which this section began—we can take the further step of doubting whether Alcibiades is actually right. Maybe Socrates is not concealing his knowledge: rather, he should be taken seriously when he avers in the Apology that he has none to conceal. This thought would be strengthened if it could be supported by an alternative explanation of the Socratic problem summed up by Nehamas at the outset of this chapter. mistaken. I do not rehearse the details of this argument here, as they can be found in Lane 2006. Hunter (2004: 101) captures the same sort of point well: ‘The depth of Alcibiades’ misunderstanding is thus revealed when he explains that he assumed that, in return for sexual favors, he would ‘‘hear everything which Socrates knew’’ (217a4–5). This is not the model of Socratic inquiry with which we are familiar … Alcibiades, who claims to know what lies behind Socrates’ appearance, has once more been misled by appearances.’ But Hunter’s subsequent characterization of Socrates as someone who ‘never [sc. has] great enthusiasms and desires, [is] never out of control, lacking nothing’ (p. 101), is strikingly unfaithful to Alcibiades’ emphasis on Socrates’ passionate absorption in the pursuit of thinking. ¹¹ My translation as given in Lane 2006: 72–3, following in part (but with the large exception of the translation of eir¯oneia) that of Nehamas and Woodruff (in Cooper 1997).

50

Melissa Lane

Fortunately, there is such an alternative explanation of Socrates’ apparent virtue, and it is presented in the voice of Socrates himself, in Republic VI.

I I I . T H E N AT U R A L PH I LO S O PH E R I N REPUBLIC V I

The hydraulic model of disposition and its development Now, we surely know that, when someone’s desires (epithumiai) incline strongly for one thing, they are thereby weakened (asthenesterai) for others, just like a stream that has been partly diverted into another channel … then, when someone’s desires flow towards learning and everything of that sort, he’d be concerned, I suppose, with the pleasures of the soul itself by itself, and he’d abandon (ekleipoien) those pleasures that come through the body—that is, if he’s a true philosopher (philosophos). (Socrates to Glaucon, Resp. VI, 485d6–8, d10–e1)

In the passage from 485a4–487a8 on the nature of the philosophers (henceforth ‘NP’), and explained most succinctly in that part of it quoted above, Socrates explains the nature of philosophers by comparing their desire to a flowing stream of water: a comparison which leads me to call this a hydraulic model of desire (485d7–8). Philosophers are lovers (er¯osin) of learning (485b1), whose psychic energy flows so strongly into a loving pursuit of the truth (485d3) that it saps the flow into bodily desires. This does not originate in a reasoned judgement that bodily desires and experiences are unimportant, or in a deliberate policy of asceticism by which they are forcibly suppressed. It originates rather in a hydraulic redirection of psycho-physiological energy effected by the sheer power of the love of learning in the soul of the natural philosopher. That hydraulic effect, drastically canalizing desire, in turn generates a disposition to temperance. Socrates follows the introduction of the hydraulic effect by asserting this summarily: ‘surely such a person is temperate and (kai) not at all a money-lover’ (485e3), adding that it is appropriate for others to concern themselves with and ‘take seriously’ (spoudazein, 484e5; cf. e4) the things for which money is needed, but not for the natural philosopher to do so. The kai explains the attribution of temperance by giving an important example (which we will see recurring in the Symposium), but does not exhaust that attribution, which two sentences before was established as pertaining generally to the ‘pleasures that come through the body’ (485d12). The natural philosopher will divert as much energy as possible away from bodily desires; the inference is that he or she¹² will invest none at all in the greedy or lustful desires that would lead to what are generally judged to be the paradigmatically intemperate actions. ¹² The Greek uses the masculine throughout NP, but as the possibility of philosopher-queens has already been vindicated in book V, I will use gender-inclusive language for the philosophical natures discussed in book VI.

Virtue as the Love of Knowledge

51

Because the attribution of temperance immediately follows the introduction of the hydraulic effect, which itself is offered to explain the fact that lovers of the pleasures of the soul abandon the pleasures of the body, it is described initially in dispositional terms. Natural philosophers, pulled by the love of learning, simply do not desire the pleasures of the body (note that this does not deny that they will have to meet the body’s needs). But spoudazein is ambiguous between mere dispositional description and evaluative content. With their desires patterned as they naturally are, the natural philosophers become aware of the contrast between the pleasures of the soul that they pursue and those of the body; and they come to evaluate the former as matters that they care about in contrast with the latter. While Socrates presents the philosopher’s nature as originating in sheer hydraulics, one kind of attraction (to the pleasures of the soul) which deprives another kind (to the pleasures of the body) of psychic energy, he soon begins to speak in terms which suggest an interplay between this original disposition and an evaluative attitude which arises from and confirms it.¹³ The natural philosophers become aware of what it is they love, and reflect on why, and although they do not yet have knowledge, they endorse the value of the truth which is the object of their loving pursuit (485c4). It is this evaluative attitude to the value of the object, or content, of their love (i.e., truth and learning), which is primarily responsible for generating the next quality ascribed to the natural philosophers: that of courage. Socrates excludes slavishness and small-mindedness (aneleutherias … smikrologia, 486a4–6) as incompatible with the striving for grasping the whole, both divine and human, which characterizes the natural philosophers. He then asks rhetorically: ‘And will a thinker high-minded enough (megaloprepeia, contrasting with smikrologia above) to study all time and all being consider human life to be something important?’ (486a8–10). When Glaucon replies that this is impossible, Socrates infers that such a person will not consider death to be a terrible thing (486b1), and then concludes that a cowardly and slavish (aneleutheri¯o) nature will take no part in true philosophy (486b3–4). Here, the emphasis is placed not on the hydraulic canalizing of desire alone, but on the qualitative nature of what that desire seeks—although the natural philosopher does not yet have knowledge, she or he is attracted to a certain kind of knowledge: namely, knowledge of the whole, and with this as object, other objects pale by comparison. What pales in particular is the value of human life and the ¹³ Compare the identification in Woolf 2004: 98–110 and passim of two different ways of interpreting the Phaedo: an ‘ascetic’ interpretation of the philosopher’s practising of death (such as that given in Bostock (2000 [1986]) and an ‘evaluative’ interpretation. Woolf argues that while both are present in the Phaedo, the bulk of the text favours the latter. The distinction is helpful for our analysis of Republic VI (which Woolf does not discuss), which is certainly not to be read as giving an ascetic account of forcible suppression of the desires of the natural philosophers. But nor does it content itself with a solely evaluative account: as we have seen, Republic VI explains the evolution of such evaluative attitudes in terms of an original dispositional genesis in the hydraulic model, and traces the interaction between them.

52

Melissa Lane

concomitant fear of death—and so there is no investment of energy in avoiding death. The subsequent attribution of justice—strictly, the exclusion of injustice—is presented as a concomitant of the possessing of temperance and courage, and so as arising from a combination of dispositional and evaluative qualities. Socrates asks (again rhetorically) whether someone with the nature established so far—of not being money-loving, slavish, or cowardly; now adding to this list not being a boaster and being orderly (kosmios: a close ally of temperance)—could possibly become¹⁴ unreliable or unjust (adikos) (486b6–8). The natural philosophers will not be unjust, but their justice, such as it is, has no root or origin separate from their temperance and courage. Temperance, courage, justice: it can be no accident that these are the qualities focused on in Passage NP, though they are joined in 486b–e by a range of qualities such as orderliness, reliability, quickness in learning, good memory, measuredness, and grace. Yet temperance is the only one to be named by the unequivocal invocation of the name of a virtue (the philosopher is by nature s¯ophr¯on, 485e3), while Socrates subsequently seems to backpedal from an attribution of the full virtues simpliciter to the natural philosopher by remarking that he (or she) will be ‘a friend and relative of truth, justice, courage, and temperance’ (philos te kai suggen¯es al¯etheias, dikaiosun¯es, andreias, s¯ophrosun¯es, 487a4–5, my modification of the Grube–Reeve translation in Cooper (1997)). Do the philosophers by nature possess the (full) virtues (apart from wisdom), or are these some other sort of qualities which sometimes go under the name of full virtue, and sometimes don’t? And if the latter, what sort of qualities are they? So far in section III, I have taken care to use only terms used by Socrates in Passage NP. However, these terms are problematic, as Socrates slides between the names of the virtues and their intimation by exclusion of their opposites, and between the names of the virtues and the designation of these qualities as ‘friend and relative’ rather than identical to full virtues. While he does not introduce any consistent vocabulary to distinguish the qualities of the natural philosophers from the full virtues, I will now advert to the term introduced at the outset of this chapter—‘natural virtues’ and, considered collectively, ‘natural virtue’—in order to do so, using these terms for the qualities of temperance, notcowardliness, and not-injustice which are ascribed to the natural philosophers in Passage NP. In short, I will say that the philosophers by nature possess the natural virtues of temperance, courage, and justice. This vocabulary can intimate the same closeness, and yet absence of identity, between the virtues of the natural philosophers and full virtues as does Socrates’ chosen device of semantic slippage. Natural virtues very often appear to be full virtues, because both are commonly ¹⁴ Does ‘would [not] become’ (an … genoito, as a rhetorical question) signal that the virtue depends on further education? As argued below, our passage is expressly insulated from the effects of education. The phrase must therefore refer simply to how this quality will unfold in action.

Virtue as the Love of Knowledge

53

manifest in similar actions; because they share a dispositional and evaluative preference for activities and pleasures of the soul over those of the body; and because the former are also a precondition for development of the latter. (Thus the relationship is stronger than a family resemblance, because the one is necessary to allow development of the other, and because they share an explicable underlying structure.) Whereas Socrates speaks of the ‘natural philosopher’ but not of the ‘natural virtues’, it is helpful to introduce the distinction between natural virtues and full virtues as terms of art, as all the points which Socrates makes can be consistently interpreted along the lines of that distinction. Moreover, it aligns with a distinction which he does draw at the end of NP, in making clear that his account therein has been carefully restricted to a consideration of the nature of philosophers before they have ‘reached maturity [literally, ‘been perfected’] in age and education’ (487a7–8). This contrast between ‘nature’ and completion by ‘age and education’ evinces the ‘implicit distinction’ between two senses of ‘the nature of man (h¯e tou anthr¯opou phusis, 395b4)’ which Antony Hatzistavrou has identified in the Republic. The first sense, corresponding to Socrates’ use of the term ‘nature’ in NP, is that of ‘a cluster of a particular person’s natural capacities which are to be developed by a process of education’; the second, corresponding to Socrates’ contrasting reference to ‘completion’ by ‘age and education’, is that of ‘the developed personality of a particular person, i.e. basic character-traits this person has acquired after a process of education’.¹⁵ Socrates is thus in NP talking about ‘nature’ in the sense of natural qualities before they have been developed and perfected by education, without any corruption having set in.¹⁶ (The possibility ¹⁵ Hatzistavrou (2006), who himself uses the distinction to argue that the philosopher-kings can be shown not to sacrifice their happiness when assuming the task of ruling so long as that happiness is correctly understood in relation to their nature in the latter sense of ‘developed personality’. On ‘nature’ (phusis) in ancient Greek culture, see more generally Lloyd 1996, the remarks on its normative character (p. 6) and on its role in demarcating the sorts of explanations given by the phusikoi (p. 103). ¹⁶ That NP is explicitly and solely concerned with the first sense, that of the natural capacities which are to be developed by a process of education, is confirmed by a capacious back-reference in Book VII (535a–536b), when Socrates concludes his account of the education designed for the philosophers and returns to the question of to whom such education should be allotted. There he refers to the ‘sort of people we chose in our earlier selection of rulers’ (535a6–7), referring to those who are ‘the most stable, the most courageous, and as far as possible the most graceful’ (535a10–b1). In this list, courage and gracefulness come from NP (gracefulness as an additional aspect of the philosopher’s nature having been highlighted at 486d4–10), while the reference to maximum stability (bebaiotatous) picks up on a point which had been added at 503b–e, when Socrates made the point that it is rare to find people who combine the set of (in effect) NP qualities with the requisite ‘stability’ (versions of bebaios, 503c4, c5, c8). The book VII passage goes on to class having a good memory and being keen on study and hard work (535b5–c4), which are also natural virtues from Passage NP (486c3–d3), as qualities specifically conducive to education. These Passage NP natural qualities conduce to education; they cannot therefore be its fruits. The relationship between Passage NP and 535a is noted by Adam (1902: vol. ii) in his note to 485a–487a (NP), but he blurs the issue by also adumbrating 375a–c (which makes a much more general claim about the need to combine gentleness and spiritedness) and 377b–391e (which

54

Melissa Lane

of corruption arises, as we will see below, from a certain instability in the natural virtues, which is a further way in which they differ from full virtues.) When a natural philosopher is well educated, she will continue to display the same basic pattern of acts and omissions in her action, the explanation of which will become overdetermined. She would anyway refrain due to hydraulic causes (as it were, the genetic reason for her refraining) but she is also now able to give a reasoned basis for refraining (as it were, a motivating reason, which when based on knowledge will also qualify as normative reason). Because such omissions, in particular, are the most striking feature of virtuous characters, natural philosophers will appear virtuous to most observers. It may be that at the margins, the acts and omissions of the natural philosopher and those of the knowledgeable and fully virtuous person will diverge. The lover of a beautiful boy in the Phaedrus, for example, who is moved by beauty to ascend toward philosophy, and shares the indifference of the natural philosopher to money,¹⁷ ‘forgets mother and brother and friends entirely’ (Phdr. 252a2–3), while the fully virtuous person presumably would not. Yet they share significantly in disposition and evaluation despite the issue of knowledge which divides them. The natural philosopher who is not corrupted by education would never rob a bank, commit adultery, desecrate a temple, or commit any of the other paradigmatic wrongs of greed in which injustice and intemperance unite.¹⁸ Natural philosophers would have no motive to abuse power.

The hydraulic model in the context of the Republic’s argument It might seem a retrograde step in the Republic’s argumentative strategy to return to consideration of the philosophers’ nature before education, when that education has already been introduced (if not yet fully detailed). Why should Socrates not at this point be eager to expound further the philosopher with true knowledge introduced at the end of book V and the beginning (before NP) of book VI, rather than retreating temporarily to delineation of the natural philosopher who as yet lacks knowledge? In other words, why is the nature of the philosophers so carefully isolated here from their education? is irrelevant). The Grube–Reeve translation’s note to 535a in Cooper 1997 refers the reader to the largely irrelevant passage at 412b ff., while overlooking its specific back-reference (as qualified by 503c) to Passage NP. ¹⁷ According to Socrates, he ‘doesn’t care at all if [he] loses his wealth through neglect’ (Phdr. 252a3–4). In both texts, and in Alcibiades’ speech in the Symposium, money is viewed as the main instrument for procuring or assuaging bodily pleasures and pains. Hence a temperate attitude to money is essential to temperance in general, whether as a natural or a full virtue. ¹⁸ Compare Resp. IV 442e4–443a10, where injustice and intemperance are shown to be linked in the ordinary and most evident cases of unjust action—embezzlement, temple robberies, thefts, betrayals of friends, oath-breaking, adultery, disrespect for parents, neglect of the gods—although in this context, before the introduction of the philosopher and his psychic core of temperance, the explanatory emphasis is placed on justice.

Virtue as the Love of Knowledge

55

Textually, the appeal to the nature of the philosopher picks up the second half [b] of the two-part procedure outlined at 474b3–c3 to justify the claim that philosophers must rule: If we’re to escape from the people you mention [sc. the people whom Glaucon imagines rushing to attack them on hearing the claim that philosophers should rule], I think [a] we need to define for them who the philosophers are that we dare to say must rule. And once that’s clear, [b] we should be able to defend ourselves [in claiming that philosophers must rule] by showing that the people we mean are fitted by nature both to engage in philosophy and to rule in a city, while the rest are naturally fitted to leave philosophy alone and follow their leader.

[a] is established by introducing the philosophers as full knowers at end of book V and that part of book VI preceding NP. [a], in other words, focuses on the philosophers’ knowledge. But [b]—their fittedness by nature both to doing philosophy and to ruling—is explained in NP by an appeal to the moral character of those naturally attracted to philosophizing. The reasoning in NP runs thus: (i) Natural philosophers are those who love to do philosophy. (ii) Loving to do philosophy is what fits people for doing philosophy, because: (iii) loving to do philosophy subjects them to the hydraulic effect, from which arise the natural virtues of: (a) temperance, primarily as a disposition; (b) courage, primarily as an evaluative judgement following on the causes and establishment of the temperate disposition; (c) justice, as a result of (a) and (b); and (d) other good moral and temperamental qualities, also arising from the same sources. (iv) These natural virtues fit them for doing philosophy by ensuring that they will pursue it wholeheartedly and undistractedly. (v) These natural virtues also fit them morally for ruling others by ensuring that they will not have any psychological temptation to abuse their power. Thus (i)–(v) collectively establish [b]. And in an extended passage which we may call ‘NP-max’ (beginning with NP and extending through 497b), we find two conjoined implications of [b]: (vi) Natural virtue explains the fact that most of those who now study philosophy become corrupted, for it is due to the strength of their natural virtues that in a corrupt society they become corrupted (example, implicit: Alcibiades); (vii) It also explains why philosophy is now useless, because it is either (a) studied by those who are not naturally fitted to study it; or (b) studied by natural philosophers who, each for a peculiar cause, are unable to be useful to existing cities (examples named: Theages and Socrates, together with other unnamed hypothetical types)

56

Melissa Lane

Finally, outside the scope of NP-max, we make the final inference linking [a] and [b]: (viii) The other condition for rule—having knowledge [a]—will also be a result of their being fitted for philosophy, since it is only through pursuing philosophy that one can gain the requisite knowledge (the education of the philosophers in the remainder of book VI, after NP-max). So, in summary: (ix) the natural philosophers are naturally fitted for philosophy, and are therefore—through their hydraulically generated natural virtues—also morally fitted for ruling; they are also those who can become epistemically fitted for ruling, by gaining knowledge through philosophy. As we have already considered (i)–(v), and (viii) is outside the scope of this chapter, it remains to examine (vi) and (vii). At the end of NP, Adeimantus objects: someone might well say now that he’s unable to oppose you as you ask each of your questions [sc. the questions that have been asked in NP], yet he sees that of all those who take up philosophy … the greatest number become cranks, not to say completely vicious, while those who seem completely decent are rendered useless to the city because of the studies you recommend. (487c4–d5)

It is well known that to answer this objection, Socrates introduces the ship of state simile. Less noticed is that the moral he draws from the simile, including the attempt to explain why most natural philosophers become either vicious or useless, harks back to NP: ‘let’s begin our dialogue by reminding ourselves of the point at which we began to discuss the nature that someone must have if he is to become a fine and good person’ (489e4–490a1), a nature which is ‘completely contrary to the opinions currently held about him’ (490a5–6).¹⁹ Socrates will conclude that the ‘shame’ and ‘reproaches’ which are currently brought against philosophy (495c3–4) are the result of its being pursued by those who are naturally unfitted to do so, and who beget ‘sophistries’ rather than ‘true wisdom’ when they do pursue it (496a7–9). The focus on [b] now serves to explain the current corruption of philosophy by its unsuitable adherents, as well as to justify the claim that it should instead be restricted to the province of the naturally suited (because naturally philosophical) few. Socrates then illustrates each of these implications—the corruption which has made most natural philosophers vicious, and the causes which make the rest useless, in existing cities—by appealing to recognizable existing types of people ¹⁹ Admittedly, Socrates then briefly dilates on the goal which such lovers of learning seek (attainment of knowledge), but his invocation of their virtues clearly recalls the terms in which the nature of the philosopher was described in NP (‘[r]emember that courage, high-mindedness, ease in learning, and a good memory all belong to it’, 490c9–10).

Virtue as the Love of Knowledge

57

in each case. That is, he gives an explanation of the likes of Alcibiades (a paradigm of corruption), and then gives an explanation of the likes of Socrates himself (a paradigm of civic uselessness to Athens as it now is).²⁰ As for Alcibiades: the description of someone with a philosophical nature, born in a large city, who is ‘rich and well-born, in addition to being tall and good-looking’, who becomes filled with ‘impossible ambitions … empty pride and vain display’ (494c–d) has long been recognized as evoking if not naming Alcibiades. Socrates argues here that it is those with philosophical natures who are most prone to the fullest corruption. This is because the natural virtues themselves—although initially generated by the love of knowledge, which imparts a high-mindedness and disdain for petty material desires—may as character traits pull against the tendency to philosophize. Specifically, Socrates contends that the natural virtues such as ‘courage, temperance, and the other things we mentioned [sc. in Passage NP]’ (491b9–10) themselves ‘tend to corrupt the soul … and … drag it away from philosophy’ (Resp. 491b8–9). The high-mindedness, vigour, and even self-discipline which they impart to their possessor make him liable to attract other forms of gratification (e.g., political success) and in turn to become attracted by them. For this argument to work, the natural philosophers have to have possessed the natural virtues despite then undergoing a corruptive process which is the opposite of true education. And indeed, Alcibiades admits in the Symposium that when he hears Socrates speak, he is overwhelmed and transfixed, as if drunk, becoming more frenzied than the Corybants (Symp. 215d6–e4). Alcibiades is a natural philosopher in the sense that he is naturally and passionately moved by Socrates’ words. But having turned in the wrong direction, away from philosophy and toward the corrupting attractions of political leadership, he violently disrupts that natural affinity to philosophy: ‘I refuse to listen to him; I stop my ears and tear myself away from him’ (Symp. 216a6–7). The Symposium gives us Alcibiades’ developed self-portrait of the sketch drawn in Republic VI: the corruption of a natural philosopher by his very natural virtues. The hydraulic effect in natural philosophers will initiate the formation of the natural virtues, but these may not continue to harmonize with or reinforce the passion for philosophy which initially generates them. The natural virtues are unstable in relation to philosophy, liable to become diverted from the love of learning and harnessed instead to other sources of more immediate and sensual satisfaction. The very vigour, tenacity, and high-mindedness which they involve can be turned and corrupted to serve unvirtuous, unphilosophical ends. As for cases like those of Socrates, he includes himself and Theages among those ‘who spend their time on philosophy as of right’ (496b). For different reasons in each case, no one in the group of those born to date with philosophical ²⁰ Although in the Apology Socrates articulates a way in which he is useful to Athens (serving as its gadfly), the criterion of usefulness in the Republic is focused on usefulness as rulers.

58

Melissa Lane

natures has actually persisted in seeking political power in cities as they now are (a process that inevitably leads to corruption: witness the likes of Alcibiades), and so each has been able to preserve his passion for philosophy uncorrupted. The message about avoiding politics in existing cities has long been recognized in this section of the text, which is part of NP-max. What has not been recognized is that, in including himself among the small group of natural philosophers, Socrates is also applying to himself the natural virtues listed earlier in Passage NP. Having identified himself as a natural philosopher, he is thereby also endowing himself with the natural virtues of temperance, courage, and justice, as well as the host of other qualities he had identified as qualities of the natural philosopher in NP. Now, as we saw above, Nehamas holds that Plato nowhere gives a solution to the problem of how Socrates can be virtuous while still lacking knowledge. But we have seen that in fact Plato does give a solution, by explaining Socrates’ character in terms of natural virtue in Republic VI. And as we shall see in the next section, this explanation fits perfectly with the way in which Alcibiades describes Socrates’ virtues in the stories of the Symposium. We can go further. For the person with natural virtues is the only person able to set out on the road to full virtue. As Republic VI specifies, it is only someone with a philosophical nature—hence only someone who possesses the natural virtues—who is fitted to pursue a philosophical education. Thus one must have the natural virtues before being able to acquire the full, cognitively based virtues. This makes sense at a psychological level. By being undistracted by excessive physical desires, the natural philosopher clears crucial psychic space for the cultivation of reason. Without natural virtue, a person would lack the eros for knowledge that can guide him to knowledge of the Forms, and he would be assailed by strong bodily desires without yet having the knowledge and trained reasoning power to subdue them. Being a natural philosopher clears the soul of excessive desires and enables the soul to focus on the pursuit of knowledge before gaining the knowledge that could then tame the desires. It remains true that, as John Cooper states, the Republic centres on the view that ‘no one is just, strictly speaking, who does not have knowledge of what is best to do’.²¹ Yet no one can gain such knowledge who is not a philosopher by nature, while being a philosopher by nature leads one to act and refrain from acting in distinctive ways (if not necessarily all the ways) in which fully just and virtuous people will also distinctively act and refrain. As a result, it is reasonable for Socrates to talk of natural philosophers as having (natural) virtues (just as Alcibiades’ calling Socrates virtuous is accepted by the symposiasts without demur), even though he insists that they do not have the full virtue which would ²¹ Cooper 1999 [1977]: 140, emphasis original; he remarks likewise that ‘one sees that Plato consistently restricts justice, as a virtue of individuals, to those who possess within themselves knowledge of what it is best to do and be’ (p. 141, emphasis original). Cooper’s essay seeks to comprehend the views of both book IV and books V–VII on the subject, but does not mention Passage NP.

Virtue as the Love of Knowledge

59

come with education. Were it not so, no philosophical rule could be established. For it is only the natural philosopher who can escape the vicious circle of needing someone else (a philosopher-ruler) to tame one’s desires for one, and so who would be available to establish a virtuous circle of philosophical rule if the other conditions for doing so (including gaining full knowledge, and overcoming their natural reluctance to rule) were met. I V. A LC I B I A D E S ’ S TO R I E S O F S O C R AT I C V I RT U E S : A N A LT E R N AT I V E R E A D I N G We return now to the Symposium, to see whether the Republic’s account of natural virtues can illuminate the problem of Socrates portrayed by Alcibiades there, albeit being contrary to Alcibiades’ own flawed solution to that problem. Although Alcibiades initially calls Socrates’ inner temperance hidden, he subsequently recounts stories of his virtues which have been manifest and (except for the seduction scene) widely observed by others. In these stories, beginning with the seduction scene and continuing through the accounts of Socrates on military campaign, Alcibiades switches from insisting that Socrates’ virtues are hidden to insisting that they have been (in the military scenes) or can now be (in recounting the seduction scene) put on display for everyone to perceive. Alcibiades glosses the stories as examples of Socratic virtues. He programmatically introduces the virtues as he sums up the seduction scene by saying that in its aftermath, he recognized Socrates to be an exemplar of temperance and courage (s¯ophrosun¯e and andreia, Symp. 219d5). But the conventionality of this ascription of the virtues is undercut in two ways. First, because the ascription of temperance is mixed with the claim that it was justifiably to be resented: Alcibiades at once admires the temperance displayed by Socrates’ resistance to seduction and claims that it displayed contempt and hubris which deserve to be judged by the symposiasts as if they were court jurors (Symp. 219c2–6).²² Second, because instead of completing an ascription of the four cardinal virtues, Alcibiades says that he would not have thought it possible to ‘happen’ to meet a man who ‘happened’ to go further toward phron¯esis and karteria (endurance) (219d6–7).²³ This repetition of the verb ‘happen’ (entuchein) undercuts the usual association of phron¯esis with predictability, explicability, and the absence of (and contrast with) chance, while the strange pairing of phron¯esis and karteria—as if endurance were the fourth cardinal virtue instead of justice, which Alcibiades ²² The accusation that Socratic virtue is actually an expression of contempt is repeated in the tale of the soldiers who found his indifference to the cold a sign of contempt for them (h¯os kataphronounta, Symp. 220c1), though in that case Alcibiades seems to reject their accusation. ²³ He then remarks on Socrates’ indifference to money (Symp. 219e), parallel to the illustration of temperance in terms of money in Passage NP (Resp. 485e3; cf. 486b6) and the Phaedrus passage discussed above, n. 17.

60

Melissa Lane

never mentions²⁴—demonstrates that whatever Alcibiades may believe himself to be doing, he is not in fact describing Socrates in the orthodox terms of full virtue. Indeed, neither phron¯esis nor s¯ophrosun¯e nor andreia is mentioned by name in the rest of the stories. Where we would expect mention of temperance or s¯ophrosun¯e, Alcibiades uses the physiologically tinged karteria (for references, see below); where we would expect mention of andreia, Alcibiades uses the physiologically tinged adverb err¯omen¯os (Symp. 221b6). These naturalistic Socratic virtues do not, as named or portrayed in the stories, appear to be full knowledge-based virtues. Rather, they are depicted in terms which closely resemble the natural virtues delineated in Republic Passage NP. Karteria in verbal form plays a striking role in the first three post-seductionscene examples of Socratic temperance, featuring in Socrates’ unusual ability to withstand hunger (219e8–220a1), cold (220a6–c1), exhaustion and immobility (220c1–d4). Alcibiades emphasizes the exceptional ways in which Socratic karteria is displayed: he stood up to hunger better than anyone, yet could also enjoy a feast better than anyone; he could drink anyone under the table, but never become drunk; he walked barefoot on winter ice with only a light cloak. The karteria actions are never described as being based on the giving of reasons based on knowledge; indeed, it’s hard to see how reasoned decision could prevent one from becoming intoxicated upon drinking, or from feeling cold when walking barefoot on ice. Socratic temperance does not arise—as would a full virtue—from a cognitive judgement in which reason is deliberately governing desire. Rather, it arises—as would a natural virtue—from a hydraulic absorption of energy which is continuous between psychic and physiological. Socrates’ immunity to erection, despite physical stimulus, in the seduction scene, is continuous with his immunity to intoxication, despite the effects of drink, while on campaign (and again at the symposium of the eponymous dialogue itself ), and with his immunity to fatigue, as we shall shortly see. His failure to display these physiological responses demonstrates the extent to which his energy has been diverted away from investment in normal bodily desires and responses. We tend to consider moral virtue as distinct from any physiological responses. But for the Greeks, s¯ophrosun¯e was at its core about the control of physical pains and pleasures; and karteria functions here the same way, except that it stresses instinctive physiological hardihood in place of reasoned judgement. So it cannot be right to describe Alcibiades’ Socrates as ‘a picture of complete (and unnerving) rational control’:²⁵ the emphasis here is on physiological reactions, in which no work is done by the mention of phron¯esis and its associated inner/outer split in ²⁴ Except in saying that it is dikaion to credit Socrates with his behaviour in battle (Symp. 220d5–6). Rowe (1998: 206, n. to 215a5–222b7) proposes that it is having charged Socrates with h¯ubris which prevents Alcibiades from crediting him with the virtue of justice. ²⁵ As suggested by Rowe (1998: 211, n. to 220a5–6).

Virtue as the Love of Knowledge

61

the earlier part of Alcibiades’ speech. Not only does the hydraulic pull overwhelm the attraction of other desires, it even overwhelms the physiological effects of drink, overeating, and (we will now see) sleep deprivation. That physiological hardihood is not merely a quirk of physique. Instead, as the most important story of Socratic karteria shows, it is the result of a hydraulic attraction to philosophy.²⁶ In that story, Alcibiades recounts an episode of philosophical absorption which led Socrates to stand motionless and sleepless, from dawn to dawn. (It is implied that he then went off to spend an ordinary twelve-or-so-hour day before finally sleeping, just as will happen at the end of the Symposium itself, when after a sleepless night Socrates goes about his ordinary business.) This story is framed by reference to karterein by means of a quotation from the Odyssey (iv. 242). Whereas in the earlier stories we saw only indifference to pain and pleasure, here for the first time we hear of a positive desire acted upon by Socrates. It is because he was visibly absorbed in thinking (Alcibiades calls it sunno¯esas; the other soldiers recognized it as phrontiz¯on) that he was indifferent to the ordinary need for movement and for sleep, and to all the other cares and objects which he might otherwise have pursued that day. In short, although Alcibiades does not call it philosophy, it is Socrates’ philosophical nature—his overpowering desire to search for knowledge—which in this case explains his karteria (perceived also as his temperance) as a natural virtue. In this story, it is the commitment to doing philosophy which itself succeeds in enfeebling his ordinary physical needs and desires, in both dispositional and evaluative ways: they are ‘weakened’ (Resp. 485d7) in that they are hydraulically starved, and partly independently, partly as a result, in that they do not appear to him compelling or important. Dominic Scott puts it well in speaking of the Symposium: it is ‘his absorption in philosophy which underlies his [sc. Socrates’] indifference to physical pleasure or pain’.²⁷ When Alcibiades turns to what we normally regard as questions of Socrates’ courage—while studiously avoiding that word—we find that Socrates’ courage, as registered by Alcibiades, is a more conventional matter than his temperance. He exhibits it alongside and not incommensurably with the courage displayed by other brave men in battle. It is the extent and nature of his temperance that is truly remarkable about him (remember that it is temperance, not virtue ²⁶ In the course of discussing the Phaedo, Woolf (2004: 105) offers an insightful brief account of Alcibiades’ portrait of Socrates (p. 105) which is broadly compatible with mine: ‘What is striking about Socrates is not that he never indulged, or (contrarily) was never exposed to material hardship, but that it seemed to have so little effect on him … Rather than being immune or even especially abstemious, it seems best to read Socrates as fundamentally indifferent to his bodily pleasures and pains, in the sense of regarding neither as of great significance in his life. What matters is the quest for wisdom.’ However, Woolf emphasizes the evaluative over the dispositional in explaining this indifference, whereas the Symposium stories about Socrates stress his sheer psychological and physiological dispositions, an emphasis which accords with the origins of the hydraulic model in Republic VI. ²⁷ Scott 2000: 31, a point with which I agree, although my overall interpretation of the relationship between Alcibiades’ and Diotima’s speeches diverges from his.

62

Melissa Lane

in general, which Alcibiades originally posits Socrates to be concealing inside himself: endothen … s¯ophrosun¯es (Symp. 216d6–7)). Saving Alcibiades’ life at Potidaea is not an unheard-of act for a warrior, though it is undoubtedly brave;²⁸ the only characteristic Socratic touch here is his refusing an award for valour and urging that Alcibiades should have it, which is better described as temperance than courage. Similarly, what is striking in the retreat from Delium is not so much bravery as the fact that Socrates remains mindful, collected (emphr¯on, Symp. 221b1),²⁹ more so even than the famous general Laches.³⁰ On the natural virtue model (but not on the standard virtue as knowledge model), as we saw in the previous section, this predominance of temperance over courage makes perfect sense. For courage, as we saw in our Republic passage, is an effect of what one loves, while temperance is an effect of the fact that one loves—it manifests directly the hydraulic reshaping of desire. The hydraulic effect generates temperance as the psychic core, in the sense of origin, of courage (and, though Alcibiades does not recount it, justice). That Socrates is, of all his virtues, most characteristically temperate, underscores the unconsciously intimated effects of the hydraulic model in Alcibiades’ speech. The love of knowledge is a better explanation for the nature of the Socratic virtues Alcibiades recounts than is his own attempt to explain them by imputing concealed knowledge to Socrates. Alcibiades concludes these stories by asserting that Socrates is unique: ‘there is no man similar to him’ (221c4–5). How are we to understand this? For his part, Alcibiades characteristically muddies the waters in trying to explain his claim. First, he asserts that whereas Achilles was comparable to others such as Brasidas as a (great) warrior, and Pericles to others such as Nestor as a (great) orator, Socrates is not comparable to any other human being. Initially the point of this contrast would seem to be that Socrates has no identifiable or conventional role within the city. But Alcibiades then adds that the only viable comparison for Socrates is his own earlier comparison of him to the non-human Silenii and satyrs, a reference which implies that the point of the contrast is that Socrates has an inner/outer split, as do the Silenus figures, and a daimonic aspect, as do the satyrs. As in his initial outburst, Alcibiades here insists on seeing Socrates as unique in virtue of his inner/outer split, which explains his external virtues by reference to hidden internal virtue and knowledge. Socrates ‘is such an atypical ²⁸ It is an interesting question why Socrates bothers to save Alcibiades’ life on either a full knowledge or a hydraulic account, since on either line he should not think human life important. Scott (2000: 36 and passim) uses the parallels with Diotima’s speech to argue that Socrates saves Alcibiades’ life because he is his ex-er¯omenos, for whom Socrates can and does still feel a non-erotic, non-obsessive ‘powerful attachment’, even though he has advanced on the Diotimean ascent beyond the love of individuals. ²⁹ Rowe (1998: 212, on 221a8–b1) neatly translates emphr¯on as ‘with his wits about him’. ³⁰ Another sign in this material that Alcibiades does not fully understand the significance of what he is reporting, is his quotation—actually, slight misquotation—of the chorus’s description of Socrates from Aristophanes’ Clouds 362, as part of his attempt to describe this Socratic virtue (Symp. 221b3–4).

Virtue as the Love of Knowledge

63

man (atopian), both in himself and in his arguments (logoi), that the attempt to search for a comparison either among people today or in the past would not come close’ (my translation, to emphasize the atypicality of his strangeness). In contrast to Alcibiades’ emphasis on Socrates’ uniqueness and incomparability, Socrates in Republic VI (Passage NP) sees only the cause of his not having benefited the city as unique (that cause being his daimonion). On the more fundamental issue of his virtues, far from treating himself as incomparable or unique, we have seen that he includes himself in a small group of those with philosophical natures who manifest the natural virtues accordingly. Perhaps he is unusual in the strength and consistency of the hydraulic pull within him, and in the way that it permeates not only his psychological but also his physiological responses. But he is not unique in manifesting the hydraulic effect per se, as this is generated by his philosophical nature, a nature which he shares with those few other persons who are also natural philosophers. Neither is he unique in the natural virtues which originate in him as a hydraulic effect of that philosophical nature, though the forcefulness and consistency of his manifestation of those virtues may be unusual (and, since there are anyway not many natural philosophers around, may appear unique to many observers). It is true, and important, that Socrates is the only natural philosopher who made a powerful impression on the Athenians, at least as Plato reports matters. In this sense, absent a full philosopher whom Plato does not claim that we have ever seen, Socrates’ example as a natural philosopher is the best and only example of virtue (albeit natural) which he and his contemporaries knew (hence the closing remark of Phaedo 118a16–17 that Socrates appeared to his friends as ‘of all those we have known the best, and also the wisest and most just’). Yet the Republic insists that Socrates is in principle (and in reality) comparable to other natural philosophers, as indeed it must if it is to have a pool of natural philosophers potentially available to rule. Alcibiades’ absorption in Socrates blinds him to the fact that what he admires in Socrates is not fundamentally unique to Socrates at all. (One might say that Alcibiades has gotten nowhere on Diotima’s ascent.) The Republic is, inter alia, Socrates’ rebuke to Alcibiades, for insisting so much on his uniqueness as to fail to understand its generic cause in his philosophical nature. V. C O N C LU S I O N The Socrates described in the Symposium by Alcibiades is, when the latter’s biases and incomprehensions are discounted, compatible with the Socrates who describes himself as a natural philosopher in Republic VI. Moreover, we have seen that natural virtue engendered by the love of knowledge explains the virtues as possessed by the recognizably most virtuous (though not fully virtuous) person in Athens. That person, Plato’s Socrates, while displaying the natural virtues

64

Melissa Lane

to a degree which led his contemporaries to take him to be fully virtuous, nevertheless insists continuously from ‘early’ to ‘middle’ dialogues that only full virtue resting on knowledge could constitute either morality or happiness. For example, his argument in the Phaedo that only philosophers are courageous and temperate—which begins by contrasting lovers of wisdom with lovers of the body, wealth, and honours in terms seemingly congenial to the hydraulic model (Phd. 68c1–2)—ends by insisting that only wisdom makes the virtues true virtue (Phd. 69b1–3).³¹ But there is no inconsistency here, so long as we distinguish between natural virtues and true virtue, and recognize also that they fall on a continuum of which the latter is the teleological end. On that continuum, the Phaedo is most interested in full virtue, and so would deny the names of courage and temperance to any but the philosophers (Phd. 68b1–69a4), while oscillating between explaining this as due to the philosophers’ being ‘lovers’ of knowledge (as in the passage just quoted) and as due to their actual possession of wisdom (e.g. Phd. 69b1–3). This emphasis on the possessing knowledge end of the spectrum reflects the Phaedo’s concern with the death of the philosopher who has pursued education and reflection throughout his life. By contrast, Passage NP of Republic VI emphasizes the loving knowledge end of the spectrum, reflecting its concern with the birth and growth of the philosopher, who begins as a natural philosopher before going on to be either educated properly or corrupted. Without, then, being able to make a wholesale break between them, we may still ask: what would full knowledge-based virtue add to natural virtue? The answer is that it would add just what Plato’s Socrates (both ‘early’ and ‘middle’) insists it would: definition and understanding. For someone seeking a rational account of the world, natural virtue is unsatisfying: it rests on a psychophysiological genetic or causal pull rather than on an explanation citing normative reasons. Knowledge explains why the natural virtues are virtues, making them not simply the result of innate dispositions or of evaluative beliefs, but grounded in an account of their value, of what makes them good, and so virtues indeed.³² Natural virtue, moreover, is incomplete from the standpoint of the cardinal virtues, for it cannot by definition include phron¯esis (pace Alcibiades’ introduction of this term as part of the inner/outer split which he claims to see in Socrates). Nevertheless, accepting the transformative effects of definition and understanding, it does not follow that full virtue would result in wholly different actions, described extensionally, from natural virtue. While natural virtue may sometimes miss the mark, it will generate a cluster of temperance-driven actions which will be roughly coextensive with a cluster of the actions dictated by full virtue. And this is no accident. The person with full virtue must develop out of the ³¹ This is part of a notoriously difficult sentence, for a plausibly anti-ascetic reading of which see Weiss 1987. ³² I am grateful to Amber Carpenter for helping me to clarify this point.

Virtue as the Love of Knowledge

65

natural philosopher, because only the latter will have the psychic space and energy to pursue knowledge. The higher value on the pleasures and objects of the soul, over those of the body, to which the natural philosopher initially dispositionally and then evaluatively subscribes, will only be reinforced by the attainment of knowledge. And it is that preference for the soul, for truth and learning, over the bodily, which generates and eventually explains the philosopher’s consistent avoidance of the actions and attitudes most typical of intemperance, cowardice and injustice. Rooted in his or her temperance, the natural philosopher develops an evaluative attitude which supports the disposition to be courageous, and these qualities together mean that she or he will not act unjustly. This basic pattern of dispositions and evaluations will be reinforced by the attainment of knowledge, even if it is also corrected at the margins. But we must take note of one thing that the person with full virtue can do, which the natural philosopher cannot do. This is to rule over other people. The hydraulic model of dispositions and its associated evaluative attitudes works only within one’s own soul, independently of whether one knows the good or not. To rule others, one needs knowledge of the good in order to prune their desires and shape their actions. Pruning must come from without, and so involve reasons or reason-based interventions to substitute for the absent hydraulic impulse in the soul of the person needing to be ruled. The hydraulic model and its development of the natural virtues served Socrates well in his own life, albeit leaving him intellectually dissatisfied and still seeking knowledge to attain full virtue. But those who are to become philosopher-rulers need to supplement and complete it with education leading to knowledge of the good. Thus we see that in the Republic, Socrates is presented by Plato as the natural philosopher—he, or someone like him, would be the candidate for gaining full philosophical knowledge, but he does not yet possess it. And we see that the philosopher-rulers of the Republic will retain at their psychic core the hydraulic effects of the love of knowledge and its associated evaluative outlook, even when they have fully completed their education. Cicero remarks that Plato ‘says that they [sc. the philosophers] are just because they are busied with the pursuit of truth and because they despise and count as naught that which most men eagerly seek and for which they are prone to do battle against each other to the death’ (Off. I. 28). This diagnosis is strikingly akin to the account given in Republic VI of the natural philosophers. In offering this as an account of the Republic’s general argument, Cicero puts his finger on the indispensable role played by the nature of the philosopher in enabling him or her both to gain knowledge and to have the moral qualities necessary to rule.³³ Socrates was (naturally) virtuous ³³ Cicero’s further claim in this passage—that the philosophers would rule only by compulsion and so would fail to be just in so far as they did not voluntarily do what was right—is controversial, depending on the correct interpretation of why the philosophers rule, a question which has not been discussed here. But the point in the text holds independently.

66

Melissa Lane

even without a full philosophical education. Had he received one and become a philosopher-king, he would still have displayed the patterns of desire of a natural philosopher, even when possessing the full virtues and ability to rule of an educated one. Virtue as the love of knowledge explains both Plato’s Socrates and the path from him to the philosopher-kings and -queens envisioned in the Republic. REFERENCES Adam, J. (1902) (ed.), The Republic of Plato, 2 vols. Cambridge. Barnes, J. (1984) (ed.), The Complete Works of Aristotle, 2 vols. Princeton. Bostock, D. (2002 [1986]), Plato’s Phaedo. Oxford. Cicero, M. T. (2005 [1913]), De Officiis, trans. W. Miller, Loeb Classical Library edition of Cicero, ed. J. Henderson. Cambridge, Mass. (vol.21). Cooper, J. M. (1997) (ed.), with D. S. Hutchinson, associate ed., Plato: Complete Works. Indianapolis. With translations of the Phaedo by G. M. A. Grube; of the Symposium by A. Nehamas and P. Woodruff, and of the Republic by G. M. A. Grube, rev. C. D. C. Reeve. (1999 [1977]) (ed.), ‘The Psychology of Justice in Plato’, in his Reason and Emotion: Essays on Ancient Moral Psychology and Ethical Theory (Princeton), 138–49. First published in American Philosophical Quarterly, 14: 151–7. Duff, T. (1999), Plutarch’s Lives: Exploring Virtue and Vice. Oxford. Gifford, M. (2001), ‘Dramatic Dialectic in Republic Book I’, Oxford Studies in Ancient Philosophy, 20: 35–106. Hatzistavrou, A. (2006), ‘Happiness and the Nature of the Philosopher-Kings’, in F-G. Herrmann (ed.), New Essays on Plato, (Swansea), 95–124. Hunter, R. (2004), Plato’s Symposium: Oxford. Kahn, C. H. (1987), ‘Plato’s Theory of Desire’, Review of Metaphysics, 41: 77–103. Lane, M. (2001), Plato’s Progeny: How Socrates and Plato Still Captivate the Modern Mind. London. (2006), ‘The Evolution of eir¯oneia in Classical Greek Texts: Why Socratic eir¯oneia is not Socratic Irony’, Oxford Studies in Ancient Philosophy, 31: 49–83. (forthcoming), ‘Irony’, in D. Morrison (ed.), The Cambridge Companion to Socrates. Cambridge. Lear, J. (1990), Love and Its Place in Nature: A Philosophical Interpretation of Freudian Psychoanalysis. New York. Lloyd, G. E. R. (1996) Adversaries and Authorities: Investigations into Ancient Greek and Chinese Science. Cambridge. Nails, D. (2006) ‘Tragedy Off-Stage’, in J. H. Lesher, Debra Nails, and Frisbee C. C. Sheffield (eds) Plato’s Symposium: Issues in Interpretation and Reception, Centre for Hellenic Studies series (Cambridge, Massachusetts), 179–207. Nehamas, A. (1998), The Art of Living: Socratic Reflections from Plato to Foucault. Berkeley. Nightingale, A. W. (1993), ‘The Folly of Praise: Plato’s Critique of Encomiastic Discourse in the Lysis and Symposium’, Classical Quarterly, 43: 112–30. Rowe, C. J. (1998) (ed.), Plato: Symposium. Warminster. Scott, D. (2000), ‘Socrates and Alcibiades in the Symposium’, Hermathena, 168: 25–37.

Virtue as the Love of Knowledge

67

Sheffield, F. C. C. (2001), ‘Alcibiades’ Speech: A Satyric Drama’, Greece and Rome, 48: 193–209. Vlastos, G. (1991 [1987]), ‘Socratic Irony’, in his Socrates: Ironist and Moral Philosopher (Cambridge), 21–44. First published in Classical Quarterly, 37: 79–96. Weiss, R. (1987), ‘The Right Exchange: Phaedo 69a6–c3’, Ancient Philosophy, 7: 57–66. Woolf, R. (2004), ‘The Practice of a Philosopher’, Oxford Studies in Ancient Philosophy, 26: 97–129.

4 Equal Sticks and Stones David Sedley

It is a very special pleasure to be able to dedicate this chapter to Myles Burnyeat, from whom over the past three decades I have learnt more about ancient philosophy than from any other scholar, past or present. The chapter takes the form of a commentary on Plato, Phaedo 74a9–c6, a passage on which my views were partly formed in discussion with him, as a whole series of letters and postcards still in my possession testify. The passage in question is the centrepiece of Socrates’ main argument for Recollection, based on the example of seeing equal sticks or stones and being led by these to think of something distinct from them, the Equal itself. It has long been recognized as containing a pivotal argument for the separation of Forms from their sensible instances. But virtually every sentence of it has generated at least one interpretative crux. 74a9–b3 (74a9–12) ‘Then consider whether this is the case. We say, don’t we, that there is an Equal (φαμέν πού τι εἶναι ἴσον)—I don’t mean a stick to a stick, or a stone to a stone, or anything else of that sort, but some different thing beyond all those, the Equal itself (αὐτὸ τὸ ἴσον). Are we to say that there is one, or none?’ (74b1) ‘We most certainly are to say that there is one,’ said Simmias, ‘with a vengeance!’ (74b2) ‘And do we know what it itself is (ἐπιστάμεθα αὐτὸ ὃ ἔστιν)?’ (74b3) ‘Certainly.’¹

This chapter is a companion piece to Sedley (2006). A certain amount of overlap in content between the two pieces has proved unavoidable. Parts of the chapter were presented to audiences at the University of Edinburgh, UC Davis, Corpus Christi College Oxford, Emory University, and Reed College. I am grateful to participants in those discussions, and to Francesco Ademollo, Gail Fine, Inna Kupreeva, and Stefan Koller for written comments. ¹ Σκόπει δή, ἦ δ’ ὅς, εἰ ταῦτα οὕτως ἔχει. φαμέν πού τι εἶναι ἴσον, οὐ ξύλον λέγω ξύλῳ οὐδὲ λίθον λίθῳ οὐδ’ ἄλλο τῶν τοιούτων οὐδέν, ἀλλὰ παρὰ ταῦτα πάντα ἕτερόν τι, αὐτὸ τὸ ἴσον· φῶμέν τι εἶναι ἢ μηδέν; Φῶμεν μέντοι νὴ ∆ί’, ἔφη ὁ Σιμμίας, θαυμαστῶς γε. ᾿Η καὶ ἐπιστάμεθα αὐτὸ ὃ ἔστιν; Πάνυ γε, ἦ δ’ ὅς.

Equal Sticks and Stones

69

T H E E X A M P L E O F ‘ E QUA L’ Why choose the example of equality? The answer, I shall argue, lies in the fact that equality is a basic geometrical concept. Because it is primarily germane to geometry, its choice as a sample object of recollection supports the impression that recollection is once again, as in the earlier Meno (82a8–86c3), being presented as universally attainable. For in that earlier passage even an uneducated slave was shown to be capable of solving a geometrical problem—a demonstration that the relevant information is already present inside every human soul, awaiting retrieval through a process of expert interrogation. In the Phaedo, in order to contribute effectively to his final aim of showing that every human soul is immortal, Socrates needs once more to display recollection as universally available to human beings, and a simple geometrical piece of knowledge, available to everyone to uncover and display, is ideal for the purpose.² Back at Phaedo 65d4–e6, Socrates introduced the Forms into the discussion for the first time, although it turned out that the theory was already familiar to the present company. He there mentioned two groups of Forms: (1) Just, Beautiful, and Good (65d4–8); (2) Largeness, Health, and Strength (65d12–e1). The second group may occasion some surprise: isn’t health, after all, a rather more empirical concept than Plato would be expected to opt for when introducing those objects of pure thought, the Forms? But in the light of Plato’s previous work these two groups turn out to make excellent sense. The first group—Just, Beautiful, and Good—is the standard list of difficult value concepts, on which people can never agree, according to the Euthyphro (7c3–d8), because they have no established standard to judge them by, as they do by contrast for decisions as to what is, for example, larger and smaller. Just as clearly, the second group, Largeness, Health, and Strength, is designed to recall the early pages of the Meno (72d4–e9). There Meno will not readily accept that all virtues are virtues because of a single shared form—that all people, for example men and women alike, are good in the same way. In order to persuade him, Socrates raises the parallel cases of health, largeness, and strength, the very same list that recurs in the Phaedo: Meno will at least agree that men and women are healthy, large, or strong in the same way. This is a familiar type of Socratic move, using the analogy of an already successful discipline in order to found an inquiry in the still undeveloped discipline of ethics. Compare Laches 192a1–b4, where Socrates assumes his proffered definition of another simple mathematical ² I am persuaded by Scott (1995: ch. 2) that recollection is depicted as occurring during the intellectual process of learning, and not, as widely held, in the course of mere rational thought. But (cf. Sedley 2006: 314 and the section headed ‘Who Are ‘‘We’’?’ below) I believe that, as both the ‘equality’ example and the requirements of the broader argument confirm, the relevant learning is seen as starting with the most elementary mathematical studies, and not as restricted to the conceptual inquiries of Platonic philosophers.

70

David Sedley

concept, speed, to be uncontroversial, and therefore proposes it as the model for defining a difficult ethical concept, courage. Likewise in the Meno the merit of the analogy with health, largeness, and strength lies in the fact that these are the objects of already successful disciplines—medicine, measurement, and gymnastic training respectively—and that for this reason we already know that each is a unitary property. Evidently, then, at Phaedo 65d12–13 the occurrence of the group Largeness, Health, and Strength as a distinct set of Forms draws on the Meno, and thereby on the recognition that these concepts have already been mastered by existing disciplines, so that they can serve as models for investigating the difficult value Forms. The presence of largeness in the list is of especial importance, because largeness is interdefinable with equality, the specimen Form chosen for the Recollection Argument.³ For Plato, Equal is, more specifically, part of the triad Larger/Equal/Smaller (Phd. 75c9; Resp. 602e4–5), or Large/Equal/Small (Soph. 257b6–7): comparative and positive versions amount to the same thing, because ‘large’ means large in relation to something, i.e. larger than something. Being part of this regular triad, equality clearly serves Plato as a size relation, not (as it might have been) as a numerical relation intermediate between ‘few’ and ‘many’.⁴ Equal is therefore going to be interdefinable with large and small, in the sense that the definition of any one of the triad will contain, implicitly, the definitions of the other two. However ‘large’ is defined, ‘small’ will be defined as the converse relation, while ‘equal’ will be defined by the simultaneous absence of both largeness and smallness. As for a fourth item considered in the Recollection Argument, ‘unequal’, it will be equivalent to ‘large or small’⁵ (much as at Soph. 257b6–7 ‘not-large’ is equivalent to ‘small or equal’). This systematic interdefinability means that the epistemic standing of equality must be on a par with that of largeness. Anyone who knows the one will know the other. Largeness, as we shall shortly see confirmed, is treated by Plato as an easy Form, one that is in fact already widely understood. And it is because exactly the same goes for Equality that the recollection demonstration starts from it, so as to appeal to learning experiences which those present, and Plato’s readers, should already have had. It is because people really do already know what equality is, in the same sense of ‘know’ in which we also aspire to know the value Forms,⁶ that it, unlike the value Forms, can serve us as actual evidence about how knowledge is acquired. ³ For what follows regarding the definition of largeness etc., cf. also Sedley 1998: 127–8. ⁴ τὸ ἴσον serves Plato as a size relation at the vast majority of its occurrences. A rare exception, kindly brought to my attention by Stefan Koller, is Prm. 151b5–e2, where the plural form ἴσα is used for numerical equality (cf. also Tht. 155a2–5). ⁵ Or, if used in the plural, of a set of mutually unequal items, ‘unequal’ will mean ‘large or small in relation to each other’. ⁶ Cf. n. 18 below.

Equal Sticks and Stones

71

In that case, what do we know equality to be? Start again from the co-ordinate property largeness. Plato assumes it to be a familiar fact, not in need of argument, that a thing is large precisely in so far as it exceeds, so that we can say either that largeness is ‘exceeding’, smallness ‘being exceeded’, or that they are, respectively, ‘the power to exceed’ and ‘the power to be exceeded’. This assumed definition is apparently at work in the background at Phaedo 102b8–d2,⁷ part of Socrates’ Last Argument for the immortality of the soul, and is more or less formally stated at Parmenides 150c6–e1.⁸ Its status as a familiar and simple truth is well evidenced by a passage in the Hippias Major (294a8–b4): For what we were seeking is that thing because of which all beautiful things are beautiful—just as it is because of that which makes all large things large, namely exceeding (for it is because of this that they are all large), that even if they exceed without appearing to they must necessarily be large.⁹

Even someone as dim as Hippias is expected to understand and endorse this definition without argument, and, following the familiar Platonic pattern, it is then to serve as a model for the elusive and highly problematic value term ‘beautiful’. If largeness lies in exceeding, and smallness lies in being exceeded, it follows that equality lies in neither exceeding nor being exceeded. In our Phaedo passage, when it is agreed without discussion that ‘we’ know what equality is, this definition of equality—probably as ‘the power to neither exceed nor be exceeded’¹⁰—will no doubt be primary among the things that those present are assumed already to know about what equality is. It need not necessarily exhaust what we are assumed to know about it. Another thing we know about equality is presumably that it is a transitive property: if A is equal to B, and B is equal to C, A is equal to C. Arguably, even if you have never heard of the word ‘transitive’, and have never formulated this principle to yourself, you already know it to be true. Indeed, you could hardly count as a

⁷ At Phd. 102c1–2 it is because of his largeness that Simmias exceeds. It might be argued that if ‘exceeding’ were the definiens of ‘largeness’, we should have expected it to cause largeness, and not vice versa. However, the causal relation here is not one involving Forms at all, but one between immanent largeness and the fact that its possessor exceeds some other individual. Provided that I am right about Plato’s definition of largeness, to say that Simmias’ largeness causes him to exceed is equivalent to saying that his power to exceed causes him to exceed, and that sounds both plausible and Platonic. ⁸ At Plt. 283c11–d1, where he seems to class length and shortness as kinds of ‘exceeding and falling short’, these can be taken to be species of largeness and smallness respectively, viz. largeness and smallness in one dimension. ⁹ I have re-punctuated the text here, to read ἡμεῖς μὲν γάρ που ἐκεῖνο ἐζητοῦμενᾧ πάντα τὰ καλὰ πράγματα καλά ἐστιν —ὥσπερ ᾧ πάντα τὰ μεγάλα ἐστὶ μεγάλα, τῷ ὑπερέχοντι (τούτῳ γὰρ πάντα μεγάλα ἐστί), καὶ ἐὰν μὴ φαίνηται ὑπερέχῃ δέ, ἀνάγκη αὐτοῖς μεγάλοις εἶναι. Burnet’s punctuation in the OCT seems to me to make dubious sense by preventing the καί in b3 from meaning ‘even’, as it surely must. ¹⁰ My apologies for the split infinitive, without which the sense would risk being obscured.

72

David Sedley

rational being if you did not. But—and this is what makes the choice of equality such a felicitous one for Socrates’ current argument—you cannot possibly have got that knowledge about equality from sense perception. The senses provide many apparent counter-examples to the transitivity of equality. In a group of three bottles, bottle A looks equal to bottle B and bottle B looks equal to bottle C, but bottles A and C do not look equal. Yet no amount of such sensory counter-evidence could ever weaken your belief in the transitivity of equality, because you already know its truth independently of sense experience. Socrates remarks in the Hippias Major, as we have seen, that the facts about what makes things large remain necessarily the same even when things don’t actually appear that way; and much the same point carries over to the facts about the closely related notion of equality: what we know about it intellectually is such as to override the apparent counter-evidence of the senses. Plato’s choice of example is thus beautifully apt for conveying to us the Recollection theory’s main basis: there are some things which we seem to be born knowing, and which sensory experience can, at best, merely prompt us to recall.

F O R M T E R M I N O LO G Y The comments under this heading are linguistic and technical. Greekless readers, and those in a hurry, can safely skip them. I go along with the majority view that ‘the Equal itself ’ (αὐτὸ τὸ ἴσον) here is, from the start, the transcendent Form of Equal. But because it has been challenged from time to time,¹¹ it is worth setting out the reasons for endorsing it. ‘The F itself ’ (e.g. αὐτὸ τὸ ἴσον) is de facto Plato’s most favoured, even if not his most technical, locution for Forms. Its mere occurrence here does not yet guarantee that the reference is to a Form, but it does when one adds the following indicator. Socrates himself announces his preferred technical term for Forms soon after the present passage, at 75d2: they are all the things on which we impose the label or ‘set the seal’ (ἐπισφραγιζόμεθα)¹² ‘what it is’, ‘ὃ ἔστι’. This technical usage—for which the simple ‘what it is’, ὃ ἔστι(ν), and the enhanced ‘what it itself is’, αὐτὸ ὃ ἔστι(ν), are used interchangeably—is later (92d9–e1) reaffirmed by Simmias: ‘the being which is named after what it is’ (αὐτὴ … ἡ οὐσία ἔχουσα τὴν ἐπωνυμίαν τὴν τοῦ ὃ ἔστιν). The same terminology is referred to by Socrates as already familiar and current in the Republic (507b7, 597a2) as well. Its first introduction as a technical term appears in fact to have been at Cratylus 389b5–6, since there, uniquely, it is introduced to Socrates’ interlocutors as an ¹¹ Penner 1987: 57–62; Dimas 2003: 198–9. ¹² Cf. Rowe 1993: 174–5: ‘There could be no clearer way of indicating that ὃ ἔστι is a technical term.’

Equal Sticks and Stones

73

innovation.¹³ By the time of the Phaedo there can be no doubt that it is already meant to be familiar to Plato’s readers as established terminology. Note, then, that Socrates, in referring to the Equal itself, uses this very same expression ‘what it is’ throughout the Sticks and Stones argument: not just at 74d6 and 75b1–2, but as early as 74b2, in the present opening lemma. For there ‘We know what it itself is’ (ἐπιστάμεθα αὐτὸ ὃ ἔστιν) is almost certainly a deployment of the same locution. True, the relative pronoun translated ‘what’ here (ὅ) functions to introduce an indirect question, and the demonstrative pronoun I have translated ‘it itself’ (αὐτό) could in principle be a mere unemphatic ‘it’. Thus in another context the clause might have been dismissed as simply meaning, without technical overtones, ‘We know, of it, what it is’. But that seems most unlikely in the present context. Variants of the ‘what it (itself) is’ locution recur elsewhere in key contexts concerning Forms with the same relative pronoun likewise used to express an indirect question (Prm. 134b14; Phdr. 262b7–8, Resp. 533a10–b2), and with the same demonstrative pronoun sometimes used to express an emphatic ‘itself’ rather than an unemphatic ‘it’. We should not hesitate to recognize it in the present passage too. This reading of the sentence is in fact supported by the pointed echo of it at 75b5–6, ‘knowledge of the Equal itself, what it is’ (ἐπιστήμην αὐτοῦ τοῦ ἴσου ὅτι ἔστιν), where (a) the variant relative pronoun (ὅτι) this time unambiguously introduces an indirect question, and (b) the demonstrative pronoun (αὐτοῦ) equally unambiguously expresses an emphatic ‘itself’. Why then the indefinite pronoun in 74a9–10, φαμέν πού τι εἶναι ἴσον? It is best translated not, as it most frequently is,¹⁴ ‘We say that there is something equal’, but rather, with the force of an indefinite article, ‘We say … that there is an Equal’.¹⁵ This highly Platonic usage is in fact a standard way of making an existential statement about a Form, and is equivalent to the English ‘We say that there is a Form of such and such’. That the τι has this function as an indefinite article, although not the only possible grammatical construal, is confirmed by the fact that in Plato Form locutions with ‘itself’ (αὐτό) regularly use either the definite article or τι: Plato’s speakers are in most contexts either referring to ‘the’ Form of F or asserting existentially that there is ‘a’ Form of F. That is how τι is used by Aristotle throughout the fragments of his critique of the Platonic theory, the Peri ide¯on,¹⁶ and by Plato not only at Phaedo ¹³ In Sedley 2003: 6–14 I argued that the original version of the Cratylus was written around the time of the Phaedo, but overlooked this evidence that, in addition, it should in fact be dated slightly before the Phaedo. ¹⁴ e.g. Valgimigli 1924; Bluck 1955; Gallop 1975; Grube 1980; Bostock 1986: 66; Rowe 1993; Ebert 2004. Closer are Dixsaut (1991), with ‘Il existe de l’´egal’, and Jowett 1871, Fowler 1914, Hackforth 1955, and others, with ‘there is such a thing as equality’. ¹⁵ This correct translation is used by Dancy (2004: 266, 268), although he then disappointingly treats it as interchangeable with ‘there is something equal’. ¹⁶ Text and commentary in Fine 1993.

74

David Sedley

65d4–5, in the sentence which first introduces Forms to the dialogue, but also at 76d7–8 and 100b6, where virtually no one would dispute that the reference is to transcendent Forms. The same usage with the indefinite pronoun recurs in Cratylus (439c7–d1), Republic (476c7), Parmenides (130b7–8), and Timaeus (51b8). It is typically accompanied by ‘itself’ (αὐτό): e.g. καλόν τι αὐτό or αὐτό τι καλόν (‘a Beautiful itself’). The only slight abnormality about the present passage is that the αὐτό is supplied, not within the initial existential formula at 74a9–10, but at the end of the same sentence, in the phrase ‘the Equal itself ’ (αὐτὸ τὸ ἴσον) at 74a12. And even that omission of αὐτό finds a parallel in an unambiguous reference to the Forms at the end of the Recollection passage (76d8–9: ‘If the objects we’re always harping on exist, a beautiful, and a good and all such being’: εἰ μὲν ἔστιν ἃ θρυλοῦμεν ἀεί, καλόν τέ τι καὶ ἀγαθὸν καὶ πᾶσα ἡ τοιαύτη οὐσία). There should be no doubt that what we have in front of us is one version of Plato’s standard existential formula for introducing a Form.

WHO ARE ‘WE’? It is scarcely deniable that at the very beginning of the lemma (74a9–10), where ‘we say’ that there is a Form of Equal (φαμέν πού τι εἶναι ἴσον), the reference is to the Socratic circle, just as it was earlier, in the initial introduction of Forms back at 65d4–5, when a very similarly worded formulation was agreed between Socrates and Simmias: ‘Do we say that there is a Just itself, or none?’ (φαμέν τι εἶναι δίκαιον αὐτὸ ἢ οὐδέν;). But in the present passage Socrates quickly switches from the indicative to the subjunctive: ‘Are we to say’ that there is such a Form (74a12, φῶμέν τι εἶναι ἢ μηδέν;)? This subtle shift from the descriptive to the prescriptive allows the reference of ‘we’ to expand without restriction, potentially including anybody and everybody. Compare Cratylus 439c7–d1, where a closely matching locution for the existence of Forms is proposed to, and agreed by, Cratylus, who is certainly no voice of Platonism: ‘Are we to say that there is a Beautiful itself, and a Good, and likewise for each one of the things that are, or not?’ (πότερον φῶμέν τι εἶναι αὐτὸ καλὸν καὶ ἀγαθὸν καὶ ἓν ἕκαστῶν τῶν ὄντων οὕτω, ἢ μή;). This means that in the Phaedo passage the reference of ‘we’, now broadened, may in principle remain broad in the rest of the lemma, where Socrates proceeds to secure Simmias’ agreement that we know what the Equal is (74b2), and in the next lemma, where he will ask him how we obtained that knowledge. Moreover, if I was right earlier that in Equal an easily mastered geometrical Form has been chosen, it seems more than reasonable to take up this option, thus placing no very narrow restriction on the scope of the ‘we’ who know what

Equal Sticks and Stones

75

equality is, and, in particular, not limiting it to Platonic philosophers. The only reason for seeking to retain such a narrow restriction might be that Socrates is talking here, not just about knowing equality, but about knowing a separated Form, the Equal itself, and that Forms are often taken by Plato’s interpreters to be altogether inaccessible to non-philosophers. But Plato apparently has no qualms about attributing knowledge of Forms even to those who do not understand their transcendent status—such as the carpenter who looks to the Form of Table or Couch in Republic X. In the Phaedo the only criterion mentioned for knowing a Form is being able to define it (76b4–7), and there seems no reason to doubt that all or most people are able to define equality: whether or not they have all already formulated its definition, they can at least all easily do so under questioning. This consideration might suggest that ‘we’ in the present lemma means all rational beings. However, the pronoun’s range is likely to be somewhat narrower than that, if Socrates’ next question (74b4, see below), about how we acquired our knowledge of the Equal, is taken to imply that the ‘we’ in question are already actively conscious of it as a distinct object of thought. This narrows the field, perhaps, but still allows it potentially to include anybody who has studied geometry. It in fact comfortably accommodates the group later in the passage described as undergoing recollection: namely ‘those we speak of as ‘‘learning’’ ’ (76a6). It certainly does not exclude those who happen not to subscribe to the theory of Forms. Then what about the more elitist remarks made later in the Recollection argument, at 76b4–12, according to which few if any apart from Socrates himself know the Forms? Here Socrates, in order to show that we do not retain our prenatal knowledge of the Forms throughout our lives, gets Simmias to agree that having knowledge of something entails being able to give a definition of it, and that most people cannot formulate definitions of the Forms he has just now listed. Simmias in fact doubts whether, following Socrates’ execution, there will be anybody with that capacity. I follow Scott and Rowe¹⁷ in taking the reference here to be to the newly extended list of Forms, expanded at 75c7–d5 to include not only the triad of Large/Equal/Small but also value Forms such as Beautiful, Good, Just, and Pious. It is, as I have said, invariably these latter that Plato’s Socrates presents as elusive and difficult. In my view this reading remains by far the most satisfactory way to remove the apparent contradiction between the two passages.¹⁸ ¹⁷ Scott 1995: 67–8; Rowe 1993: 168. ¹⁸ Although I have resisted the alternative, to which scholars sometimes resort (e.g. Bostock 1986: 67–8), of distinguishing the later passage from the earlier as referring to philosophical as opposed to mundane knowledge, I accept that the simpler or more accessible a concept is, the simpler the knowledge of it will be, which is why at 73c5–d5 sensible particulars are the objects of—presumably very low-grade—knowledge. There is no equivocation on ‘knowledge’, but there are different grades of knowledge, depending on its object. I differ from the equivocation

76

David Sedley

74b4–6 ‘Where did we acquire the knowledge of it? Wasn’t it from the things we were just mentioning? On seeing sticks, stones or other things that are equal, wasn’t it from these that we came to think of that object, it being different (ἕτερον) from them?’¹⁹

E QUA L TO W H AT ? The subject of the example is first introduced as ‘the things we were just mentioning’, recalling 74a10 in the previous lemma, ‘a stick [equal] to a stick, or a stone to a stone’. The example is resumed under the description ‘sticks, stones or other things that are equal’. I find it extremely hard to doubt that ‘equal’ in the phrase ‘things that are equal’ indicates an internal relation obtaining within a set of things, and thus that Socrates means ‘things that are equal to each other’. That both captures the essence of the earlier example—stick equal to stick, and stone to stone—and respects the natural meaning of the Greek. If someone refers, whether in Greek or in English, to ‘equal things’, without specifying an external correlative, the expression will inevitably be taken to mean ‘equal to each other’. I therefore find it very hard to go along with the alternative interpretation, according to which Socrates’ expression refers to a plurality of items which are, each of them, equal to some further item, left unspecified. Such an interpretation is not only not required by the run of the passage, but assumes Plato to be departing from linguistic idiom in a way that might prove hard to parallel in his entire corpus,²⁰ let alone in the Phaedo itself. The far more natural interpretation, ‘equal to each other’, will find ample confirmation in the remainder of the argument. 74b6–c6 (74b6–9) ‘Or doesn’t it appear different (ἕτερον) to you? Look at it this way too: [1] don’t equal stones and sticks, the very same ones, sometimes appear equal to one [or at one time], but not to another [or at another] (τῷ μὲν ἴσα φαίνεται, τῷ δ’ οὔ / τότε μὲν ἴσα φαίνεται, τότε δ’ οὔ). (74b10) ‘Yes, certainly.’ (74c1–2) [2] ‘But did the Equals themselves (αὐτὰ τὰ ἴσα) ever appear to you unequal, or Equality Inequality?’ (74c3) ‘Never yet, at any rate, Socrates.’ interpretation by denying that in addition to the knowledge of Equal mentioned at 74b2–3 there is also ‘philosophical’ knowledge of it. ¹⁹ Πόθεν λαβόντες αὐτοῦ τὴν ἐπιστήμην; ἆῤ οὐκ ἐξ ὧν νυνδὴ ἐλέγομεν, ἢ ξύλα ἢ λίθους ἢ ἄλλα ἄττα ἰδόντες ἴσα, ἐκ τούτων ἐκεῖνο ἐνενοήσαμεν, ἕτερον ὂν τούτων; ²⁰ ‘The many doubles’ at Resp. V, 479b2 (kindly brought to my attention by Tim Clarke) is not comparably incomplete, since it picks out e.g. numbers with integer halves, i.e. the class of even numbers.

Equal Sticks and Stones

77

(74c4–5) [3] ‘Then those equals, and the Equal itself, are not the same.’ (74c6) ‘By no means, Socrates, in my view.’²¹

T H E E QUA L S S Y L LO G I S M Equal sticks and stones can appear in a certain way; the Form of Equal cannot appear in that way; therefore the Form of Equal is not the same thing as the equal sticks and stones. The need to prove this non-identity between Form and participant, rather than just continue to assume it, arises from the need to satisfy Socrates’ first criterion of recollection (73c7–d1), which among other things specified that the reminding object must be different (ἕτερον, 73c9) from the item recollected. The bare bones of Socrates’ semi-formal syllogism are as follows: [A1] F sensibles appear F and also un-F. [A2] The Form of F never appears un-F. [A3] Therefore F sensibles are not the same thing as the Form of F. The most notorious headache generated by this syllogism is (assuming for the present the textual reading most often printed) the meaning of ‘appear equal to one, but not to another’ in [1] (b8–9). By defending the reading of ‘equal’ (ἴσα) as ‘equal to each other’, I have already implicitly rejected the frequently favoured interpretation that the enigmatic expression means ‘equal to one thing, but unequal to another thing’. The equality at issue is from the start their equality to each other, not to any further thing or things.²² That throws us back on the alternative of taking the datives as masculine, construed with ‘appear’: to one person the sticks appear equal (to each other), to another they appear unequal (to each other). But here too there are grounds for hesitation. On this, which I shall call the ‘B reading’, the alleged difference between equal sensibles and the Form of Equal would lie in the fact that [B1] F sensibles appear F to one person, un-F to another person. [B2] the Form of F always appears F to Simmias (74c1, σοι). [B3] Therefore F particulars are not the same thing as the Form of F. ²¹ ἢ οὐχ ἕτερόν σοι φαίνεται; σκόπει δὲ καὶ τῇδε. ἆῤ’ οὐ λίθοι μὲν ἴσοκ καὶ ξύλα ἐνίοτε ταὐτὰ ὄντα τῷ μὲν ἴσα φαίνεται, τῷ δ’ οὔ; [or, on a variant reading, τότε μὲν ἴσα φαίνεται, τότε δ’ οὔ;] Πάνυ μὲν οῦν. Τί δέ; αὐτὰ τὰ ἴσα ἔστιν ὅτε ἄνισά σοι ἐφάνη, ἢ ἡ ἰσότης ἀνισότης; Οὐδεπώποτέ γε, ὦ Σώκρατες. Οὐ ταὐτὸν ἄρα ἐστίν, ἦ δ’ ὅς, ταῦτά τε τὰ ἴσα καὶ αὐτὸ τὸ ἴσον. Οὐδαμῶς μοι φαίνεται, ῶ Σώκρατες ²² Technically he could be saying, about a pair of sticks, that they sometimes, in addition to being equal to each other, also appear equal to a third thing but unequal to a fourth thing. But I cannot see how, if this is what Plato intended, he could have hoped to be understood.

78

David Sedley

F sensibles are liable to intersubjective disagreement, whereas the Form of F is not liable to intrasubjective disagreement. This does not, as stated, display a genuine incompatibility between the sensibles and the Form. It could in theory be the case, for example, both that the sensibles and the Form sometimes appear F to some people, not-F to others, and that to Simmias both the sensibles and the Form always appear F. In other words, nothing has been said to show that what is true of the sensible equals is false of the Form of Equal. Naturally, with enough interpretative licence we could eliminate this incoherence, for example by assuming, however implausibly,²³ that Simmias here represents the consensus of everybody. We would then take the two premisses as jointly saying that there is room for disagreement about the equality of the sticks and stones, but no room for disagreement about the equality of the Form of Equal. But the text can hardly be said to lend itself to such a reading. We might also worry that, thus construed, the argument would risk being fallacious. It would turn on the different phenomenologies of sensible and intelligible equality: the former can be a matter of disagreement, since it is sometimes mistaken for inequality, but the latter never could be; therefore they cannot be the same thing. This may import a fallacy comparable to that in the following syllogism: [C1] To some people it has appeared the highest mountain is Everest, but to others that it is not. [C2] To no one has it appeared that Everest is not Everest. [C3] Therefore the highest mountain is not Everest. Usually, if X is identical to Y, whatever is true of X is also true of Y. But in certain contexts—especially intentional contexts such as are introduced by the verb ‘appear’—this law fails. Both the Everest syllogism and, on the B version, the Equals syllogism seem to be examples of such a fallacy. Before settling for an interpretation which is unsatisfactory from both linguistic and philosophical points of view, we should consider the well-attested alternative manuscript reading in b8–9, τότε μὲν ἴσα φαίνεται, τότε δ’ οὔ; yielding the translation ‘ … appear equal at one time but not at another?’ This actually offers a significantly more coherent argument: [D1] F sensibles appear F at one time but un-F at another. [D2] the Form of F never appears un-F. [D3] Therefore F sensibles are not the same thing as the Form of F. On this, the D reading, the emphasis throughout is on the way that appearances do or do not change over time: the appearance of equal sticks and stones can ²³ Thus Mills (1957–8: 50) suggests that σοι here is to be unpacked as ‘to you as a representative of humanity in general’. An obvious objection is that in the context of [B1] it is hardly plausible that one person could simply be assumed to speak for everyone.

Equal Sticks and Stones

79

vary, the appearance of the Form of Equal never does. When Simmias is asked whether the Form of Equal has ever appeared unequal to him (c1), this will not be in order to distinguish him from any other witness, but simply to call on the best evidence available to him: namely, the memory of his own invariant thought processes about the Equal. The same premiss could have been obtained, but less authoritatively and more conjecturally, by asking him how things have appeared to other people. It is useful here to compare the very similar argument in the Theaetetus (190b2–d2) against the Other-judging model of false belief. On this model, false belief would amount to judging the F to be un-F, which in turn, Socrates argues, amounts to saying to yourself that the F is un-F. But, he asks, has Theaetetus ever expressed to himself the thought that, for example, the beautiful is ugly or the just unjust? Has he, even in a dream, entertained the thought that the odd is even? In further confirmation, the same style of questioning is then extended to other people too: has anybody else, sane or mad, entertained the thought that a horse must be a cow, or that two must be one? This argument resembles the present one in appealing to constraints upon what thoughts anyone does or does not ever entertain. The total non-occurrence of the thought that the beautiful is ugly, pointed out in the Theaetetus, is closely comparable to the total non-occurrence, according to the Phaedo, of its appearing to someone that the Equal itself is unequal, or that equality is inequality. In the Theaetetus Socrates uses the interlocutor himself as the prime witness, while going on to insist that the same answer as Theaetetus gave would have to be given by anybody. Likewise in the Phaedo passage, it is easy to assume that Simmias himself, as interlocutor, is invoked as his own most authoritative witness about what does and does not ever appear, but that his own evidence is assumed, as in the Theaetetus, to correspond to the experience of everyone else.²⁴ The alternative manuscript reading therefore has much to recommend it.²⁵ There is also a possible objection, however. In 74b8, can Plato really have written ‘sometimes … at one time … at another’ (ἐνίοτε … τότε … τότε)? The translation would be ‘don’t equal stones and sticks, the very same ones, sometimes appear equal at one time but not at another?’ It is, I admit, a little inelegant. But in mitigation, even the double temporal qualification makes perfect sense: there are times when—i.e. occasions on which—the very same sticks vary between appearing now F, now not-F. The ‘sometimes’ is needed, because it cannot be ²⁴ Since [D1] does not, as [B1] did, invoke differences between individuals’ viewpoints, the context this time presents no obstacle to taking Simmias, like Theaetetus, to be assumed to speak for everyone. Cf. preceding note. ²⁵ The reading is advocated by Verdenius (1958), and adopted by Vicaire (1983), Dixsaut (1991), Ebert (2004: 34 n. 2 and 210 n. 12), and Dancy (2004: 267), but not so far as I know by any English-language editor or translator of the Phaedo, or in the OCT either past (Burnet 1900–7) or present (C. Strachan in Duke et al. 1995), although the latter’s app. crit. now shows it to be the more strongly attested reading, in effect supported by two families of MSS, compared to one family offering the rival reading (I am very grateful to Bob Sharples for pointing this out to me).

80

David Sedley

assumed that this variation of appearance obtains in every case. It may depend, for example, on the fact that the judging subject changes position in relation to the sticks or stones during the period specified. Furthermore, the slight awkwardness about the double temporal qualification has at least one compensating advantage: it may be precisely what motivated the scribal correction (from τότε … τότε to τ ῷ … τ ῷ) found in the other part of the manuscript tradition. This temporal reading is therefore textually superior to the one previously considered. But does it fare any better from the philosophical point of view? Sensible F particulars, the argument now runs, sometimes appear F, sometimes un-F; the Form of F never appears un-F; therefore the Form of F is not the same thing as the sensible F particulars. Is this any less vulnerable to the Everest objection formulated above in the C version of the syllogism? The answer will depend on the kind of ‘appearing’ that is at issue? There seem to be three main possibilities to consider. On one construal, ‘appear’ is no more than ‘turn out to be’. In the muchdiscussed passage at the end of Republic V (476e4–479d9), when describing the compresence of opposites—the fact that the very same perceived subject is characterized by both of a given pair of opposite properties—Plato has the sensibles ‘appearing’ (φαίνεσθαι) both F and un-F, and this contradictory ‘appearing’ is treated as entailing that those same sensibles both are and are not: that is, both are and are not F. If we use this as evidence that the ‘appearing’ of particulars is simply their proving to be this or that, the Phaedo syllogism would be contrasting the objective invariability of Forms with the objective variability of sensibles. The context would not be an intentional one, and the Everest objection would lapse. But regardless of how we interpret the Republic’s assumptions, we cannot safely infer that ‘appear’ in the Phaedo means no more than ‘turn out to be’. For in our present passage, even after we have eliminated the personal datives at 74b8–9, as I have proposed we should, the same verb, ‘appear’ (φαίνεσθαι), is still construed with a personal dative, ‘to you’ (σοι), at 74c1. This shows that the appearances in question are indexed to judging subjects, and seems to make it unavoidable that we should regard them as having an intentional content. That realization may encourage us to consider a second option, according to which the appearances in question on the one hand are relative to one or more judging subjects, but on the other hand are of a veridical kind, such that a thing’s ‘appearing’ F would amount to its showing itself to be F (a well-attested use of the Greek verb). Sensibles really do have opposite properties, we might say on Plato’s behalf, but one—indeed the most typical—way in which sensibles possess their properties is by manifesting them to observers, a fact which makes those properties unavoidably context-dependent. Hence, looking ahead to the later stages of the Recollection Argument (74d4–75b9), the deficiency of equal sticks and stones with respect to the Form of Equal will there lie not just in

Equal Sticks and Stones

81

their appearing less stable in respect of their equality, but also in their being less stable in that regard. Forms, by contrast, not only do not change in their objective properties; they do not even manifest the contraries of their objective properties, because the kind of being that they have is altogether non-relative and context-independent. Thus interpreted, Socrates’ syllogism would rest on the variability of sensible properties over time, due to their necessary context dependence, by contrast with the context-free absoluteness and invariability of objects of pure thought. And, as such, it would capture one of Plato’s most fundamental grounds for separating the intelligible from the sensible realm. However, the signals from the text discourage this interpretation too. The sticks and stones are initially introduced, unequivocally, as being equal (74b5, 7–8). Their equality must therefore lie not in the way they manifest themselves, which for Plato is inherently indeterminate, but simply in their objective dimensions. That is in any case by far the most natural way to read the unqualified phrase ‘equal sticks and stones’. It follows that, for the purposes of the present argument, ex hypothesi the sticks’ appearance as equal is true, their appearance as unequal false. That is already enough to exclude the veridical reading of ‘appear’.²⁶ We thus arrive at the third and final available interpretation. The syllogism sets out to contrast two cases of equality, in both of which the equality is a matter of objective fact. In the first case that objective fact is not such as to prevent the false contrary appearance from occurring too; in the second case it is. Therefore they must be two different cases. Therefore the particular equality of empirical objects is not the same thing as Equality as such. It is no doubt a matter of historical regret that we are forced to settle for an interpretation which leaves Plato vulnerable to the Everest counter-example. But it would be anachronistic to think that he was fully alive to the nature of the particular fallacy involved, especially given its unsignalled occurrence in arguments that he puts to work elsewhere.²⁷ Besides, it is important to appreciate that the preceding interpretation in my list was just as vulnerable to the same objection. Opaque-context fallacies can arise with veridical ‘appearances’ too.²⁸ Consider, the following example: ‘It is a visible fact that the Thames is full of ²⁶ In this passage φαίνεσθαι is accompanied by a simple adjective, leaving it unclear whether the implicit full construction would be with the infinitive εἶναι or the participle ὄντα, of which the latter is normally veridical, the former not. Note, however, that in the resumption at 74d5–6 (ᾶρα φαίνεται ἡμῖν οὕτως ἴσα εἶναι ὥσπερ αὐτὸ τὸ ὃ ἔστιν ἴσον;) the non-veridical construction is used. If I am right, the sticks’ and stones’ failure to be fully like the Equal itself, which Socrates goes on to highlight, must be explained in terms of their failure to appear as consistently as it does, and not vice versa. Something which is such as to appear inconstantly is ipso facto inferior to something which is not. ²⁷ Cf. Grg. 467c5–468e5, and the Theaetetus argument summarized above, p. 79. ²⁸ The second interpretation sketched above has much in common with the helpful comments of Fine (1993: 331–2), who does, however, imply that all that is needed to save the syllogism from fallacy is that ‘appear’ be understood in its veridical sense, equivalent to ‘evidently be’. This, I am suggesting, would not in fact help.

82

David Sedley

water; it is not a visible fact that the Thames is full of H2 O; therefore water is not H2 O.’ The only fallacy-free version of the argument would have been the first, which simply appealed to objective instability versus objective stability, leaving appearances aside: sensible equality is inherently variable; the Form of equality is inherently invariable; therefore the Form of equality is not the same thing as sensible equality. Such an argument would indeed have been thoroughly Platonic. But it is not the argument used in the Phaedo.

‘ T H E E QUA L S T H E M S E LV E S ’ So much, then, for the structure of the argument. I turn now to the notorious plural at 74c1: ‘the Equals themselves’ (αὐτὰ τὰ ἴσα).²⁹ Certain scholars have suggested that this plural marks a reference to something other than the Form of Equal—perhaps to immanent equality, comparable to ‘the largeness in us’ which will be introduced later in the dialogue (102b8–103a3),³⁰ or to mathematical objects,³¹ or even to a broader class which includes not only separated Forms but also these further items.³² However, such a usage has no parallel in Plato, and the reader of the Phaedo has been given no help towards recognizing it here. Nor, finally, would it do anything but needlessly complicate the argument, by threatening to prove that, as well as separated Forms, we also recollect these other items, whatever they may be. That implication, once introduced, would then have needed either explaining, or blocking off by an additional argument. Considerations of economy recommend taking this plural as a variant designation for the Form of Equal, even though that Form is elsewhere in the passage (including the conclusion of the syllogism itself, 74c4–5) referred to by the singular, ‘the Equal itself’ (αὐτὸ τὸ ἴσον). Why then, at 74c1, does Socrates use the plural, ‘the Equals themselves’, in place of his usual singular designation? One suggestion³³ has been that the reasons for the plural are fundamentally grammatical: the word for ‘equal’, ἴσα, first appeared in the predicate position in step [1], where it was plural in form for purely local grammatical reasons.³⁴ The Form expression, ‘the F-itself ’, simply picks out such a predicate and recasts it as a subject. Hence, it is suggested by the grammatical answer, the plural is simply a hangover from the grammatical plural in the first premiss of the argument. An alternative answer takes it to be something about the concept of equality as such that leads to the plural. The Form of Equal is somehow construed as ²⁹ ³⁰ ³¹ ³² ³⁴

Dale 1987 includes a useful survey of past proposals regarding this crux. e.g. Bluck 1959; Wedin 1978; cf. reply by Smith (1980). Including Hackforth 1955: 69 n. 2; Bluck 1955: 67 n. 3. e.g. Dimas 2003: 197–203. ³³ Owen 1968: 114–15. For this plural, see n. 38 below.

Equal Sticks and Stones

83

an idealized relation between two or more equal things. There is strong support for this latter answer in the fact that such plural designations of Forms occur, outside this passage, only for the predicates Similar (αὐτὰ τὰ ὅμοια, ‘the Similars themselves’, Prm. 129b1) and Many (τὰ πολλά, Prm. 129d5, where Plato uses the simple expression ‘the Many’, but in a context where he must mean ‘the Many themselves’). Similarity, like Equality, is naturally viewed as an internal relation between two or more symmetrically related items. If x is equal to y, y is equal to x; and likewise, if x is similar to y, y is similar to x. As for the Many themselves, it would have been impossible to represent this in the singular, since it is precisely the property of being plural, and its grammatical equivalent in the singular, αὐτὸ τὸ πολύ, would have to be translated ‘the Much itself ’ and to mean something quite different, the property of being quantitatively much, not numerically many. If the grammatical explanation of the expression ‘the Equals themselves’ were correct, it would be an amazing coincidence that the plural locution occurs only for these very suitable Forms, which express what are in their very nature the shared properties of a set of two or more suitably related items, and never for Forms like Large or Beautiful.³⁵ Apart from ‘Equal’ and ‘Similar’, plus their opposites, the only other familiar Platonic Forms designating symmetrical relations are ‘Same’ and ‘Different’: if x is the same as y, y is the same as x; and likewise for ‘different from’. ‘Same’ does not occur anywhere in Plato in a plural form resembling ‘the Equals themselves’, but that is hardly a surprise, given that sameness is understood as reflexive—a symmetrical relation between a thing and itself, not between two or more discrete items. Hence the only potentially significant absentee is a pluralized form of ‘Different’, ‘the Differents themselves’ (αὐτὰ τὰ ἕτερα). But its non-occurrence is easily explicable: this Form is discussed only in the Sophist, where all the interest focuses on difference as an external relation (‘x is different from y’ = ‘x is not y’), not as an internal, mutual one. So the second explanation of the plural remains the most plausible. If one accepts it, there is no resultant need to envisage the Form of Equal crudely as a pair or set of ideally equal things (as proposed by Geach³⁶). No doubt Forms often function as paradigms or ideal standards which particulars imperfectly reflect, and indeed they certainly do just that in this passage. But they are not merely that: Forms are also universal essences, and, later in the Phaedo (100b1–101d3), the causes of the properties of particulars. In no way, then, can a Form be reduced to a mere perfect instance. Besides, according to the Affinity Argument a little later in the Phaedo, Forms are emphatically incomposite: it is because a Form is not analysable into parts that it cannot be dismantled and thus destroyed. So no Form can with much plausibility be identified with an ideal pair or set of items. ³⁵ e.g., there would have been ample scope for αὐτὰ τὰ καλά at Resp. 476b4–d3. ³⁶ Geach 1956. Vlastos 1956 followed Geach in this, but retracted his agreement in the 1965 reprint.

84

David Sedley

The plural is better explained as follows. The Form can be referred to as ‘the Equals themselves’ (αὐτὰ τὰ ἴσα), because the Form of Equal is what it is for things (plural) to be equal (to each other). Likewise in the Parmenides, the Form ‘the Similars themselves’ (αὐτὰ τὰ ὅμοια) is plural because it amounts to what it is for things (plural) to be similar to each other,³⁷ and the Form referred to in the Parmenides as ‘the Many’ is so called, not because it is many things, but because it is what it is for things (plural) to be many. This explanation thus coheres with, and lends support to, my earlier contention that right from the start, with the reference to equal sticks and stones, the plural ἴσα (‘equals’) has referred to a set of mutually equal things. Throughout the passage, including now the plural designation of the Form of Equal, there are clear advantages in taking equality to be treated thus, as a purely internal relation. Even granted the legitimacy of the plural designation for a Form, it remains to ask why Plato should choose to vary this particular Form’s designation from singular to plural and back again within the space of a single argument. The answer, I suggest, has to do with syllogistic form. The singular ‘the Equal itself ’ (αὐτὸ τὸ ἴσον) is so much the standard Form locution that it normally wins by default. But in the context of the present syllogism, Plato finds clear advantages in using grammatically matching plural predicates in both the first and the second premiss.³⁸ Equal sticks and stones can appear, but also fail to appear, as equals (ἴσα, plural); the Equals themselves never appear as unequals; therefore the Form is not the same thing as the equal sticks and stones. The grammatical isomorphism between major and minor premiss serves to confirm that it is the very same predicate, inequality, that on the one hand is impossible to associate with the Form but on the other tends regularly to appear in its sensible instances.

‘ . . . O R E QUA L I T Y I N E QUA L I T Y ? ’ What, finally, should we make of the end of step [2], ‘But did the Equals themselves (αὐτὰ τὰ ἴσα) ever appear to you unequal, or Equality Inequality (74c1–2, ἢ ἡ ἰσότης ἀνισότης)?’? This addition may appear misjudged. It is no doubt true that Equality never appears to be Inequality, but then equal sticks and stones never appear to be Inequality either; so no contrast emerges as it did in the first half of the premiss. To restore some measure of respectability, we could ³⁷ In favour of this, note how at Prm. 129a8–b1 the plural Form locution αὐτὰ τὰ ὅμοια has been immediately preceded by talk of mutually similar and dissimilar things: ὅμοιά τε καὶ ἀνόμοια αὐτὰ αὑτοῖς (where this last word probably means ‘to each other’, as often, rather than ‘to themselves’). ³⁸ The point can be made clear by a more literal translation than given above, preserving the plural adjectives, as follows: ‘[1] Don’t equals stones and sticks, the very same ones, sometimes appear equals to one, but not to another …? [2] But did the Equals themselves [αὐτὰ τὰ ἴσα] ever appear to you unequals, or Equality Inequality?’

Equal Sticks and Stones

85

take the point to be that, whereas ordinary Fs can appear as their own opposite, un-Fs, the Form F-ness can never appear as its own opposite, un-Fness. What, even so, does Plato hope to gain by adding this clause at all? Assuming, as I do, that both ‘the Equals themselves’ and ‘Equality’ refer to the Form of Equal, it is best to take the second clause as little more than an epexegetic gloss on the first. The variant locution ‘the Equals themselves’ was chosen to enhance syllogistic coherence, as argued above. The additional gloss ‘equality’ (ἡ ἰσότης) now serves to confirm that it is indeed still the Form, Equality itself, that Socrates is talking about: to think that the Equals themselves are unequal would be the same as thinking that Equality is Inequality. The grammatically singular gloss has the further advantage of easing a final transition which now occurs. In stating the syllogism’s conclusion, Socrates restores coherence with the broader context by switching back from the plural to the more familiar singular designation of the Form: ‘Then those equals, and the Equal itself, are not the same’ (74c4–5). The interposition of ‘Equality’ as another variant name for the Form, by both reintroducing the grammatical singular and reminding us that there are a variety of synonymous Form designations available, makes the transition a smoother one than it would otherwise have been. REFERENCES Bluck, R. S. (1955), Plato’s Phaedo. London. (1959), ‘Plato’s Form of Equal’, Phronesis 4: 4–11. Bostock, D. (1986), Plato’s Phaedo. Oxford. Burnet, J. (1900–07) (ed.), Platonis Opera, 5 vols. Oxford. Dale, A. T. (1987), ‘Αὐτὰ τὰ ἴσα, Phaedo 74c1’, American Journal of Philology, 108: 384–99. Dancy, R. (2004), Plato’s Introduction of Forms. Cambridge. Dimas, P. (2003), ‘Recollecting Forms in the Phaedo’, Phronesis, 48: 175–214. Dixsaut, M. (1991), Platon, Ph´edon. Paris. Duke, E. A., et al. (1995) (eds.), Platonis Opera, i. Oxford. Ebert, T. (2004), Platon, Phaidon. G¨ottingen. Fine, G. (1993), On Ideas. Oxford. Fowler, H. N. (1914), Plato, i: Euthyphro, Apology, Crito, Phaedo, Phaedrus. Cambridge, Mass. Gallop, D. (1975), Plato, Phaedo. Oxford. Geach, P. (1956), ‘The Third Man Again’, Philosophical Review, 65: 72–82; repr. in R. E. Allen (ed.), Studies in Plato’s Metaphysics (London), 265–77. Grube, G. M. A. (1980), Plato, Phaedo, 2nd edn. Indianapolis. Hackforth, R. (1955), Plato’s Phaedo. Cambridge. Jowett, B. (1971), The Dialogues of Plato, 5 vols. Oxford. Mills, K. W. (1957–8), ‘Plato, Phaedo 74b7–c6’, Phronesis, 1: 128–47; 2: 40–58. Owen, G. E. L. (1968), ‘Dialectic and Eristic in the Treatment of the Forms’, in Owen (ed.), Aristotle on Dialectic: The Topics (Oxford), 103–25; repr. in Owen, Logic, Science and Dialectic (London, 1986), 221–38.

86

David Sedley

Penner, T. (1987), The Ascent from Nominalism. Dordrecht. Rowe, C. J. (1993), Plato, Phaedo. Cambridge. Scott, D. J. (1995), Recollection and Experience. Cambridge. Sedley, D. (1998), ‘Platonic Causes’, Phronesis, 43: 114–32. (2006), ‘Form–Particular Resemblance in Plato’s Phaedo’, Proceedings of the Aristotelian Society, 107: 309–25. Smith, N. D. (1980), ‘The Various Equals at Plato’s Phaedo 74b–c’, Journal of the History of Philosophy, 18: 1–7. Valgimigli, M. (1924), Platone, Fedone. Bari. Verdenius, W. J. (1958), ‘Notes on Plato’s Phaedo’, Mnemosyne, 4/11: 193–243. Vicaire, P. (1983), Platon, Ph´edon. Paris. Vlastos, G. (1956), ‘Postscript to the Third Man: A reply to Mr Geach’, Philosophical Quarterly, 65: 83–94; repr. in R. E. Allen (ed.), Studies in Plato’s Metaphysics (London, 1965), 279–91, and in Vlastos, Studies in Greek Philosophy, ii (Princeton, 1995), 204–14. Wedin, M. V. (1978), ‘Αὐτὰ τὰ ἴσα and the Argument of Phaedo 74b7–c5’, Phronesis, 23: 191–205.

5 The Phaedo’s Final Argument Nicholas Denyer

What I here call the Phaedo’s final argument ends at 106e–107a with the conclusion ‘a soul is something immortal and indestructible, and our souls really will exist in Hades’. It began at 95e, with some prolonged criticism of various wrong theories about how to explain coming into and going out of existence. This paper takes up the final argument when it turns to expounding the theory that it asserts, or at least hypothesizes, to be right, and follows the argument through to its conclusion. This chapter will contend, in effect, that this theory is weak enough to be plausible, yet strong enough to come surprisingly close to yielding the conclusion that a soul is something immortal and indestructible.

1 . H Y P OT H E S I Z I N G F O R M S ‘Hypothesizing that there is something beautiful itself by itself, and good and big and all the rest of them’ (100b). When Socrates makes this hypothesis, he is, he claims, following the method of ‘hypothesizing on any particular occasion whatever theory I judge to be the most robust’ (100a). If the hypothesis is to be as robust as Socrates suggests, what can it amount to? Let us define the two terms ‘lilid’ and ‘soquid’ as follows. We say that a thing is lilid if and only if either it is in the Northern Hemisphere and liquid or it is in the Southern Hemisphere and solid; and we say that a thing is soquid if and only if either it is in the Northern Hemisphere and solid or it is in the Southern Hemisphere and liquid. Thus Australia is a lilid mass of land, surrounded by soquid seas; its inhabitants have lilid bones to stiffen their limbs, and soquid blood circulates through their veins. Britain is the other way round: soquid land, lilid seas, soquid bone, lilid blood. And a moment’s thought will enable us to describe, in terms of ‘lilid’ and ‘soquid’, what happens when a ship crosses the Equator. Acknowledgements are due to Dominic Scott and Laura Gleen for comments on this essay, and to Myles Burnyeat for general inspiration over the years.

88

Nicholas Denyer

The terms ‘lilid’ and ‘soquid’ are obviously somehow bogus. But what exactly is wrong with them? They are defined sharply enough; for the boundaries between the lilid and the soquid are no less sharp than the boundaries between the solid and the liquid and the Northern and the Southern Hemispheres. Nor is the problem with ‘lilid’ and ‘soquid’ any shortage of things to fall under them; for the North Atlantic brims with lilid, as the South Pacific does with soquid. Even so, we want to protest that there are no such things as lilidity and soquidity. And our protest will be a protest against lilidity and soquidity in particular. Our protest will not be a general protest against forms or universals or abstract objects. For, in the sense of ‘there are’ in which there are obviously no such things as lilidity and soquidity, there are such things as liquidity and solidity. Or, at any rate, the hypothesis that there are such things is pretty robust. The forms that we thus hypothesize are, in a way, separate from their particulars. At any rate, the form of so-and-so must be something in addition to the particular so-and-sos if the existence of particular so-and-sos does not guarantee the existence of the form of so-and-so itself. However, when we hypothesize forms that are in this way separate, we are not, or at least not yet, hypothesizing forms that are capable of existing without instances.¹ Perhaps this makes our hypothesis even more robust. Of course, ‘lilid’ and ‘soquid’ are not only much more bogus than ‘liquid’ and ‘solid’; they are also much less familiar. And this can hardly be a coincidence. But our reluctance to agree that there are such things as lilidity and soquidity is more a cause than an effect of the unfamiliarity of terms for those things. For consider the relative familiarity these days of two expressions coined in the nineteenth century: ‘od’ and ‘electromagnetism’. ‘Electromagnetism’ caught on; ‘od’ these days is hardly found outside the Oxford English Dictionary, which gives the following account: A hypothetical force proposed by Baron Karl von Reichenbach as pervading all nature and accounting for various physical and psychological phenomena. … The Od force was thought to manifest itself in certain persons of sensitive temperament by streaming from their fingertips, and to be exhibited especially by magnets, crystals, heat, light, and chemical action. It was held to account for the phenomena of mesmerism and animal magnetism, among other things.

If, of these two originally unfamiliar coinages, the term ‘electromagnetism’ has come to be more familiar than the term ‘od’, that is because we have discovered that there is such a thing as electromagnetism, and no such thing as od. Not all the forms that Socrates hypothesizes are in fact needed for the purposes of his argument that souls are imperishable. Even if there is no such thing as bigness, his argument might still work. Or, to be more precise, if his argument works at all, then it will still work even if there is no such thing as bigness. But, ¹ For this, and for various other kinds of separation, see Fine 2003: 255–7.

The Phaedo’s Final Argument

89

for his argument to work, there do have to be such things as life and animation. And the hypothesis that there are such things is more than simply the evident fact that some bodies are alive and animate. That is why it is a hypothesis, and why it might be rejected even by someone familiar with the evident fact, just as someone familiar with the existence of all the various things thought to manifest od—magnets, crystals, and so on—might nevertheless doubt the existence of od itself. Still, as hypotheses of this kind go, the hypothesis that there are such things as life and animation is hardly reckless. Let us therefore adopt it for the moment, to see where it leads.

2 . F O R M S A S C AU S E S ‘It is by the beautiful that all the beautiful things are beautiful’ (100d, 100e). When Socrates gives this explanation, he declares that it is ‘extremely safe’ (100d). What can such explanations be if they are to be as safe as Socrates suggests? What makes Mary a Newnhamite? There is such a college as Newnham College, and Mary is one of its members. Why are sugary things sugary? There is such a thing as sugar, and they contain it. How is it that solids are solid? Well, there is such a characteristic as solidity, and this is a characteristic that solids have. These explanations are no doubt ‘simple-minded’ (100d–e). To anyone who knows of the existence of Newnham College, sugar, and solidity, these explanations will sound fairly vacuous. For the existence of such things is more or less tantamount to their being available for use in explanations of this kind. This is why there is no such explanation of why lilids are lilid. For the nearest thing to such an explanation of why lilids are lilid would be that lilids in the Northern Hemisphere are lilid because they have the characteristic of liquidity, while lilids in the Southern Hemisphere are lilids because they have the characteristic of solidity instead. Still, the very fact that we have not automatically hypothesized a form of lilidity, even though we have concocted the term ‘lilid’, means that it is not utterly vacuous to explain why a thing is solid by saying that it has solidity.² For this fact means that there is some difference between a thing’s being solid and its having the characteristic of solidity. The difference may well be slight; indeed, the difference may well be no more than a difference between two formulations of the same truth, one compact (‘The thing is solid’), the other more articulated (‘The thing has the characteristic of solidity’). Nevertheless, even so slight a difference is enough to make it possible to explain the one by the other. Think, ² Thus I am not persuaded by Sedley (1998: 117): ‘It seems abundantly clear that Plato sees some causal relationships, of the ‘‘the F makes things F’’ type, as conceptually self-evident.’ What are conceptually self-evident are conditionals of the form ‘If there is such a thing as the F, then the F makes things F.’ The consequents of such conditionals are not even true in all cases, let alone self-evident.

90

Nicholas Denyer

for example, of ‘They are my in-laws because I am married to their offspring.’ This explanation may not go far or deep. For it does no more than spell out the way that their being my in-laws involves not just them and me, but also some third party, the offspring of theirs to whom I am married. But this explanation is, so far as it goes, correct. Likewise, if having solidity does not make a thing solid, what else would? Although explanation by forms sounds vacuous to begin with, it can soon become more substantial. There is a frame of mind in which one might ask ‘Why did the forest come to be ablaze?’, and feel that the answer ‘Because Anna dropped her cigarette end’ says both too much and too little: too much, because it hardly matters that it was Anna rather than Angela; too little, because without tinder and oxygen there would have been no blaze, no matter what Anna did with the end of her cigarette. When in this frame of mind, we want to find some apt necessary and sufficient condition for the forest’s coming to be ablaze. Of course, we are not always in this frame of mind. We are not always even clear that it is a sensible frame of mind to be in. (Hence the joke about the man who goes to the doctor complaining of a pain in his knee. The doctor takes a look and gives his diagnosis: ‘It’s simply old age, I’m afraid.’ The patient objects: ‘But the other knee is just as old, and it doesn’t hurt.’) Still, we are sometimes in this frame of mind, and we then find satisfaction only once we reach something like ‘The forest came to be ablaze because it contained a lot of readily oxydizable material, and a lot of oxygen, and there was a bit of heat, enough to set some of the oxydizable material oxydizing, releasing so much heat as to set more of the oxydizable material oxydizing, and therefore releasing heat.’ The explanation that we thus reach is just ‘The forest came to be ablaze because a fire started in it’—amplified by supplying the real definition of what it is for a fire to start. And the possibility of supplying a real definition stands or falls with the existence of the thing to be defined. The discovery of the real definitions of fire and liquidity and electromagnetism vindicates the hypothesis that there are such things, ready to be cited in such explanations. Similarly, the discovery that there is no real definition of od shows that there is no such form of energy, no one thing that all the various odic phenomena have in common. And settling whether there is or is not a real definition of yuppie flu would settle whether there is such an ailment to complain of, or only a bunch of people who complain. A difficulty often felt with explanation by forms is that, however vacuous it may sound at first, it soon becomes far too substantial. If we hypothesize a form of so-and-so to make the so-and-sos so-and-so, then, it might seem, we simply add to the so-and-sos that we are to explain; indeed, we seem liable to add, not just one more so-and-so, but infinitely many; for if our original so-and-sos required a form of so-and-so in addition to themselves to make them so-and-so, then how can the same not apply to any so-and-sos that we reach at any stage in this process? This difficulty received a classic formulation from Plato himself, in Parmenides 132a–133a.

The Phaedo’s Final Argument

91

This difficulty may overcome some theories about forms. But it leaves unharmed the theory hypothesized in the Phaedo. For that theory insists on a distinction between the way that a form of so-and-so is so-and-so, and the way that the particulars named after it are so-and-so. The form of big itself is just plain big, and is in no way also small. It is therefore distinct from Simmias, who can be called big (by comparison with Socrates), but who can also be called small (by comparison with Phaedo), and who therefore has in him both bigness and smallness (102b–c). Similarly, the form of equality is distinct from particular equal things in that its equality is free from the qualifications under which their equality labours; and this is one of those points that the Phaedo finds so obvious that their truth can shine through any of several diverse formulations.³ In general, the form of so-and-so cannot admit the opposite of so-and-so, while particular so-and-sos can only too readily admit their opposites. In short, the form is necessarily so-and-so, whereas the particulars are so-and-so only contingently. That is why we can say that the particulars require some special explanation, that the form helps give them such an explanation, and that the form requires no such explanation itself. To adapt an analogy from Republic 509d–510b, 514c–515a, we can point to a picture of Myles and say, with some truth, ‘That’s Myles’; but the picture is also not Myles; we can explain what makes the picture Myles by saying that, in addition to this and all the other pictures of him, there is such a man as Myles himself, that he simply and unqualifiedly is Myles, and that he is present in the picture; and we can stop right here, without positing a third Myles, who stands to Myles himself as Myles himself stands to the picture.

3. FORMS IN US We have hypothesized that there is such a thing as animation, and that if a body is animate, then that is because there is animation present in that body. Or, to put it another way, we have hypothesized that ensouled bodies are ensouled because there is such a thing as soul, and they each have a share of it. From this, we can easily argue that, even after Socrates has died, animation or soul will continue to exist. For if such a thing ever exists, then it will exist always: a form is not liable to perish; so when we discover that, for example, there is no such thing as od, we ³ Thus 74b–c argues that since sticks and stones have a property (ἆρ’ οὐ λίθοι μὲν ἴσοι καὶ ξύλα ἐνίοτε ταὐτὰ ὄντα τῷ μὲν ἴσα φαίνεται, τῷ δ’ οὔ; πάνυ μὲν οὖν) which the form equality lacks (αὐτὰ τὰ ἴσα ἔστιν ὅτε ἄνισά σοι ἐφάνη, ἢ ἡ ἰσότης ἀνισότης; οὐδεπώποτέ γε), it follows that they are not the same as it (οὐ ταὐτὸν ἄρα ἐστίν … ταῦτά τε τὰ ἴσα καὶ αὐτὸ τὸ ἴσον). The argument’s multiple expressions for the form, and multiple expressions for the crucial property, indicate that its point is intended to be independent of which of those expressions turn out to be the most apt. Cf. 100d εἴτε παρουσία εἴτε κοινωνία (both referring to whatever the relation is between the Form of Beauty and particular beautiful things), the passages cited in n. 4 below, and Tht.184c, where Socrates apologizes for a distinction between seeing ὀφθαλμοῖς and seeing δι’ ὀφθαλμῶν, in terms suggesting that such distinctions are not his usual style.

92

Nicholas Denyer

discover that there never was any such thing in the first place; and if we are right to hypothesize that there is such a thing as soul, then there always will be. But if we are worried that the soul of Socrates may soon perish, we will not be consoled by any argument that soul itself is imperishable. We will need an argument for the imperishability of Socrates’ soul. For if I fret that some expensive wine may soon be vinegar, since the alcohol in it may soon turn into acetic acid, it is no consolation to be reminded that alcohol itself, the chemical substance whose real definition is given by the formula C2 H5 OH, the thing whose presence makes alcoholic drinks alcoholic, is simply not the sort of thing that is liable to perish. According to the Phaedo, there are, in addition to the ‘forms themselves in nature’ (103b), and to the particulars that share in those forms, things of a third kind: forms ‘of ’ or ‘in’ particulars, forms that those particulars ‘have’.⁴ The idea is that, if, and only if, there is such a thing as the form so-and-so itself in nature, then there is such a thing as the so-and-so in this particular so-and-so for as long as, and only for as long as, this particular continues to be so-and-so. For example, there is such a thing as alcohol; and there will be such a thing as the alcohol in my wine only for as long as my wine remains alcoholic. Forms in particulars belong in some ways with particulars, and in other ways with forms themselves in nature. They belong with particulars, in that there are as many so-and-sos in particulars as there are particular so-and-sos: since my cocktail and your cocktail can each become more or less alcoholic independently of the other, the alcohol in mine is distinct from the alcohol in yours. They belong with forms themselves in nature, in that the so-and-sos in particulars are more or less as strictly and unshakably and necessarily so-and-so as the so-and-so itself: a tot of brandy is not perfectly alcoholic, and can become even less alcoholic as the alcohol in it is boiled or burnt away; but the alcohol in the tot is just plain alcohol, and can cease to be alcoholic only at the cost of altogether ceasing to be. The Phaedo’s argument for the imperishability of the soul will be an argument for the imperishability, not of soul itself in nature, but of our souls, or the souls in us. Socrates’ body is ensouled so long as, and only so long as, there is such a thing as the soul in Socrates’ body. The argument will be that this thing will continue to exist for ever, even after Socrates’ body has ceased to be ensouled, although this thing will of course then no longer be in Socrates’ body, but somewhere else instead. For the argument will be that, unlike the tot of brandy, which can cease to be alcoholic in either of two ways, either because the alcohol in it evaporates away unharmed or because the alcohol in it gets destroyed (102d–e), an ensouled body can cease to be ensouled in only one way, by the departure unharmed of the soul that was in it (106e). Our comparison between the soul of Socrates and the alcohol in a drink invites the objection that we are treating the soul as altogether too corporeal. ⁴ Sample turns of phrase from 102b–103b are τῷ μεγέθει ὃ τυγχάνει ἔχων, τὸ ἐκείνου μέγεθος, and τὸ ἐν ἡμῖν μέγεθος. See n. 3 above for the implications of this variety of expressions.

The Phaedo’s Final Argument

93

That would certainly not be Plato’s intention. For Socrates illustrates the points that he wants to make about the soul by making analogous points about a range of other things, from the most grossly corporeal (snow, at e.g. 103c) to the most etherially pure (threeness, at e.g. 104d), by way of various apparent intermediates (fire, at e.g. 103c; fever, at e.g. 105c). It is as if he wants to ensure that we can accept what he has to say, whatever degree of corporeality we take the soul to have.⁵ And there are indeed some plausible but incorporeal analogies for how the soul in a body must be conceived if the Phaedo’s final argument is to work. For example, mindful of the way in which Platonic souls are sources of motion (Phdr. 245c–e; Leg. 895a–896b), we might compare the soul of Socrates to the momentum of a body: when the body comes to a halt, it will have imparted its momentum to some other body, and the momentum can therefore outlast the body whose momentum it earlier was, even though momentum is not a stuff.

4. BRINGING ON FORMS There is more to be said about fire, snow, fever, and three than simply that they are fire, snow, fever, and three. For we can say also that fire is hot, snow cold, fever an illness, and three an odd number. These facts about those things cast their shadows lower down. For example, since fire itself is hot, anything else that is fiery—whether fiery particular or the fire in a fiery particular—will thereby also be hot. In short, fire brings on heat. And because of this, we can give to the question ‘What is such that any body in which it is present will be hot?’ a reply more fancy than ‘Heat’; for we can also reply ‘Fire’ (105b–c). Are these ‘fancier replies’ explanations? Do they give causes? Socrates does not actually call them by the term αἰτία and its cognates, terms that he had freely used earlier. Nor do his fancier replies use the term ποιεῖν and its cognates, unlike the simple-minded reply which held that nothing other than the presence or whatever of the form of beauty ποιεῖ a thing beautiful (100d). Nor, for that matter, do his fancier replies use the causal datives or the constructions with διά and the accusative that had earlier been so prominent. Perhaps the point is that even if the fancier replies go some way towards satisfying those left unsatisfied by simple-minded explanations, this does not mean that in any strict sense they are explanations themselves.⁶ In this case, we need not fret about how ⁵ Cf. 70e–71d, where some very diverse examples illustrate the principle that opposites come from opposites, to ensure that we will apply the principle to life and death, regardless of the sense in which those two are opposites. ⁶ If so, the point is missed by Gregory Vlastos, who wrote a section named ‘The ‘‘clever’’ aitia’ (Vlastos 1969: 317–25), without ever asking whether this might be a misnomer. The point is missed also in many of the renderings of 105b, ᾧ ἂν τί ἐν τῷ σώματι ἐγγένηται θερμὸν ἔσται, given by translators and commentators: ‘what must be inherent in a thing to make it hot’ (Archer-Hind

94

Nicholas Denyer

to reconcile the apparent qualms in 96e–97b over opposite explanations of a single phenomenon with the apparent commitment in 105c to accepting the two opposites ‘fever’ and ‘chill’ as fancier replies to the question ‘What is such that any body in which it is present will be sick?’ For if fancier replies do not even pretend to present causes, they can hardly be faulted for presenting things that are not causes.⁷ In any case, whether the fancier replies are meant as a kind of explanation or as a substitute for explanation, they should be congenial to those who view explanation as deductive-nomological. For the main way in which the Phaedo goes beyond standard deductive-nomological views would be by its answer to the question of which generalizations report laws, and are therefore, when used as premisses in deductions of truths to be explained, explanatory of those truths: the answer would be that it is a law that every so-and-so is a such-and-such when there are forms of so-and-so and of such-and-such, and the one form brings on the other.⁸ When one form brings on another, the connection between the two is stronger than simply that the one form never does occur without the other. The connection is, at very least, that it never can. For when the Phaedo says that nothing will admit the opposite of something that it brings on, that snow brings on cold, which has heat for its opposite, and that snow can therefore never admit heat (103c–d), it means at the very least that there cannot be something snowy that is not cold; or, to put it another way, that it is necessary that if a thing is snowy, then it is cold. And this will have the corollaries that if something snowy is only contingently cold (like, for example, a snowy and therefore cold drink), then it is only contingently snowy, and that if something snowy would have to perish were it to stop being snowy (like, for example, the snow in the drink), then it would have to perish were it to stop being cold.⁹ If we apply this to soul, on the understanding that soul brings on life, then we reach the result that Socrates’ soul cannot cease to be alive without perishing. 1883: 115); ‘what, when present in its body, will always make a thing hot’ (Bluck 1955: 124); ‘what must be present in anything, in its body (i.e. ‘in a thing’s body’), to make it warm’ (Burnet 1911: 122); ‘what is that which must be in the body to make it hot’ (Church 1951: 60); ‘what it is that must be present in the body to make the body warm’ (Cooper 1941: 177); ‘what causes anything in which it is to be hot’ (Fowler 1982: 363); ‘what it is, by whose presence in a body, that body will be hot’ (Gallop 1975: 59); ‘what, coming into a body, makes it hot’ (G. M. A. Grube, in Cooper and Hutchinson 1977: 90); ‘what must come to be present in a thing’s body to make it hot’ (Hackforth 1955: 158); ‘what that is, the inherence of which makes the body hot’ ( Jowett 1955: i. 464); ‘what it is that, when it comes to be in a body, makes that body hot’ (Rowe 1993:258); ‘what must be present in body to make it hot’ ( Tredennick 1978: 166). ⁷ I am therefore unpersuaded by Bostock (1986: 188, in a section entitled ‘ ‘‘More subtle’’ causes’) that the example of fever ‘seems ill chosen’ because ‘Fever, understood as the condition of having too much heat in the body, would seem to have an opposite, viz. having too much cold in the body’. ⁸ See Armstrong 1983: 85. ⁹ See Gow 1965: 109 and Arnott 1996: 426 for the use of snow to cool drinks. Aristotle would have thought the practice unhealthy: see Aulus Gellius 19. 5.

The Phaedo’s Final Argument

95

This result is no doubt true; but it has not got us much closer to the conclusion that Socrates’ soul cannot perish. There is, however, more to snow’s never admitting heat than just that the snow in a snowy thing cannot cease to be cold without perishing. For there is a difference between ceasing to be cold and coming to be hot, as should be obvious to everyone familiar with the distinctions drawn in Symposium 201e–202b and Sophist 257b–c between the negation of a thing and its opposite. And there is a difference between never coming to be hot, except at the price of perishing, and never coming to be hot at all, as should be obvious to everyone. If Socrates is conscious of these differences, then he claims not only that, for example, no snow can cease to be cold without perishing, but also that no snow will ever come to be hot, either before it perishes (for then it will be cold) or after (for then it will not be at all). If we apply this to soul, on the understanding that soul brings on life and that the opposite of life is death, then we reach the result that Socrates’ soul can never come to be dead, not even by perishing. This result does not quite take us all the way to the conclusion that Socrates’ soul cannot perish. But it certainly gets us closer than we were, when our only result was that Socrates’ soul cannot cease to be alive without perishing. 5 . F RO M I M M O RTA L I T Y TO I M PE R I S H A B I L I T Y So far, we have argued that no soul can ever come to be dead. This inability to come to be dead might be redescribed as immortality of a sort, and some such redescription is given at 105e. What, however, has this sort of immortality to do with imperishability? On the face of it, not very much. An ice cube can hardly come to be dead (not ever, not even if it melts), and so an ice cube has this sort of immortality; but an ice cube can perish nevertheless, for this sort of immortality will not stop it melting. Cebes does not appreciate this. For he says, ‘There could hardly be anything else that does not admit of perishing, if what is immortal, being eternal, is to admit of perishing’ (106d), as if the immortality which implies eternity were the same as the immortality which is implied by the inability to come to be dead. Can we do any better than Cebes? Let us invoke our earlier thought that, since soul brings on life, souls are alive. Let us add to this the new thought that if a living thing perishes, then it dies.¹⁰ This new thought looks incontestably true of living bodies; let us assume that it is true of living things generally, including souls. So if a soul perishes, it comes to be dead. But no soul, we have earlier argued, can ever come to be dead, not even at the price of perishing. So no soul ever perishes. ¹⁰ That the Phaedo’s argument can be clinched by this thought is pointed out by Dorothea Frede, who (1978: 32) formulates it as: ‘whatever is alive … can only pass out of existence by accepting death, by dying’.

96

Nicholas Denyer

REFERENCES Archer-Hind, R. D. (1883), The Phaedo of Plato. London. Armstrong, D. M. (1983), What is a Law of Nature? Cambridge. Arnott, W. G. (1996), Alexis: The Fragments. Cambridge. Bluck, R. S. (1955), Plato’s Phaedo. London. Bostock, David (1986), Plato’s Phaedo. Oxford. Burnet, J. (1911), Plato’s Phaedo. Oxford. Church, F. J. (1951), Plato: Phaedo. Indianapolis. Cooper, J. M., and Hutchinson, D. S. (1997), Plato: Complete Works. Indianapolis. Cooper, L. (1941), Plato on the Trial and Death of Socrates. Ithaca, NY. Fine, Gail (2003), Plato on Knowledge and Forms. Oxford. Fowler, H. N. (1982), The Euthyphro, Apology, Crito, Phaedo and Phaedrus of Plato. London. Frede, Dorothea (1978), ‘The Final Proof of the Immortality of the Soul in Plato’s Phaedo 102a–107a’, Phronesis, 23: 27–41. Gallop, D. (1975), Plato: Phaedo. Oxford. Gow, A. S. F. (1965), Machon: The Fragments. Cambridge. Hackforth, R. (1955), Plato’s Phaedo. Cambridge. Jowett, B. (1955), The Dialogues of Plato, 4th edn. Oxford. Rowe, C. J. (1993), Plato: Phaedo. Cambridge. Sedley, David (1998), ‘Platonic Causes’, Phronesis, 43: 114–32. Tredennick, H. (1978), The Last Days of Socrates. London. Vlastos, Gregory (1969), ‘Reasons and Causes in the Phaedo’, Philosophical Review, 78: 291–325.

6 Beauty of Body, Nobility of Soul: The Pursuit of Love in Plato’s Symposium Alexander Nehamas

I The last part of Socrates’s speech on er¯os in Plato’s Symposium opens with a strikingly enthusiastic declaration: ‘Only in the contemplation of Beauty is human life worth living’—a declaration made more striking by its similarities to the words that Socrates uses in his Apology to express what was perhaps his deepest conviction: ‘The unexamined life is not worth living for a human being’.¹ In fact, it should be even more striking because of its similarities to another famous declaration—this one by Aristotle in the Nicomachean Ethics: ‘What is done virtuously is noble and it is done for the sake of the noble.’ But no one without a knowledge of the language and culture of fifth- and fourthcentury Greece can know how close Aristotle’s sober and systematic treatment of virtue is to Plato’s inspired hymn to the power of love. For no one without such knowledge can know that both the nobility that makes action virtuous and the beauty that makes human life worth living—two ideas apparently A very early version of this text was delivered as the Gray Lectures at Cambridge University in April 2004. I am grateful to the Faculty of Classics at Cambridge for their generous invitation and their marvelous hospitality and to Richard Hunter, Malcolm Schofield, Dominic Scott, David Sedley, Frisbee Sheffield, and Robert Wardy for their personal and intellectual companionship during my stay. I have presented the core of the discussion of the ascent passage in the Symposium at several institutions, and I gained much from my discussions there; a version of it has appeared in The European Journal of Philosophy, 15 (2007). I am aware of the irony involved in submitting this essay, meant to honour Myles Burnyeat and thank him for all I have learnt from him since we first met in 1970, because I once again failed to exercise the patience with which he so admirably continues to review and revise his ideas before he is satisfied that they are ready to be seen by others. ¹ Symp. 211d1–3: ἐνταῦθα τοῦ βίου, ὦ φίλε Σώκρατες, ἔφη ἡ Μαντινικὴ ξένη, εἴπερ που ἄλλοθι, βιωτὸν ἀνθρώπῳ, θεωμένῳ αὐτὸ τὸ καλόν; cp. Ap. 38a5–6; ὁ δὲ ἀνεξέταστος βίος οὐ βιωτὸς ἀνθρώπῳ.

98

Alexander Nehamas

irrelevant to each other—are denoted by the very same Greek term: the word kalon.² Those who work regularly with classical texts may find it difficult to remember how surprising, jarring, and confusing it was to find out for the first time that a single word applies not only to the obvious attractions of Aphrodite³ and Alcibiades⁴—neither one a paragon of virtue—not just to the ingenious decorations of the helmet that Hephaestus made for Achilles,⁵ but also (though perhaps ironically) to a hill suitable for running,⁶ as well as Antigone’s death for rightfully burying her brother despite Creon’s prohibition.⁷ That kalon has such a broad and diverse range has gradually become a commonplace, and classical scholars are by now largely comfortable to use at least two different terms where the Greek has only one. Some prefer to make do with the single word ‘fine’, but that, I believe, has failed to make either the Greek understandable or the English idiomatic. For example, readers of the otherwise outstanding version of Aristotle’s Nicomachean Ethics by Sarah Broadie and Christopher Rowe (Broadie and Rowe 2002: 143) are bound to stop at a sentence like ‘actions in accordance with excellence are fine and for the sake of the fine’ (IV 1, 1120a 23–4). Similarly, it seemed perfectly natural to abandon ‘fine’, which Paul Woodruff had used throughout his own translation of the Hippias Major,⁸ when it came to rendering kalon in our collaborative translation of the Symposium: ‘Only in the contemplation of the fine is human life worth living’, to the extent that it makes any sense at all, would be a bizarre barbarism.⁹ Liddell and Scott’s Lexicon actually lists not two but three distinct senses for kalon: ‘beautiful, of outward form’, ‘with reference to use, good, of fine quality’, and ‘in a moral sense, beautiful, noble, honourable’.¹⁰ The first sense is ‘aesthetic’ in a strict traditional sense, and it is reasonable tow think that it is confined to how people or things look independently of their other qualities. So, although Nireus of Syme was second in beauty (kallistos) only to Achilles, nothing else about him was notable: ‘he was a lightweight, trailed by a tiny band’.¹¹ The ² Arist. Eth. Nic. IV 1, 1120a 23–4: αἱ δὲ κατ’ ἀρετὴν πράξεις καλαὶ καὶ τοῦ καλοῦ ἕνεκα. ³ The Homeric hymn to Aphrodite begins: Αἰδοίην χρυσοστέφανον καλὴν Αφροδίτην | αἴσομαι. Cf. Hes. Theog. 194–5. ⁴ Pl., Prt. 309a2–3. ⁵ Hom. Il. XVIII. 611–12: τεῦξε δέ οἱ κόρυθα βριαρὴν κροτάφοις ἀραρυῖαν | καλὴν δαιδαλέην. ⁶ Xen., An. 4. 8. 26. ⁷ Soph., Ant. 71–2: κεῖνον δ’ ἐγὼ | θάψω. καλόν μοι τοῦτο ποιούσῃ θανεῖν. ⁸ Woodruff 1982: p. xv: ‘kalos is ‘‘fine’’ throughout, though some contexts taken alone would call for ‘‘beautiful’’ or ‘‘good’’.’ ⁹ Nehamas and Woodruff 1989. In what follows, I cite that translation, with occasional alterations. ¹⁰ Liddell and Scott 1968: s.v. καλός. The Lexicon lists a fourth, ironic sense, ‘fine, specious’, which does not affect this discussion. ¹¹ Hom. Il. II. 671–5, trans. Fagles 1990. Note however, in connection with the rest of this essay, that Nereus flouts certain expectations regarding a man of his social position (and looks).

Beauty of Body, Nobility of Soul

99

second sense is functional, and refers primarily to how well something is suited for some purpose, although it has some aesthetic connotations as well. We find it in Xenophon’s Symposium when Socrates, perhaps in jest, praises his protruding eyes and pronounces them superior (kalliones) to Critoboulos’ because they allow him to see not only in front but sideways; also, in Plato’s Hippias Major when Socrates argues, quite seriously, that a thing is kalon only to the extent that it performs its function well.¹² The third is clearly illustrated by Theognis, when he lays out the three main objects of desire—the good, the useful, and the pleasant—that were to become, especially when kalon and agathon came to be more closely associated with morality,¹³ a staple of Greek philosophy: ‘The best (or noblest: kalliston) thing is justice; health, the most advantageous; | but the greatest delight is to win the object of one’s desire.’¹⁴ Sir Kenneth Dover, whose views on these issues are the starting point of so much of our thinking about them, simplifies only a little when he gives voice to the current consensus regarding the word’s two main senses (Dover 1980: 2): The word [kalon], when applied to a person, means ‘beautiful’, ‘pretty’, ‘handsome’, ‘attractive’, and its antonym is aischros, ‘ugly’. The words are also applied to objects, sights and sounds and whatever can be heard about and thought about, such as an institution, an achievement or failure, or a virtuous or vicious action; kalos expresses a favourable reaction (‘admirable’, ‘creditable’, ‘honourable’) and aischros an unfavourable reaction (‘disgraceful’, ‘repulsive’, ‘contemptible’).¹⁵

In another passage that conforms with Liddell and Scott’s first definition of kalon and to which we shall return, Dover draws a sharp distinction between the aesthetic and the moral senses of the term: ‘It must be emphasized that the Greeks did not call a person ‘‘beautiful’’ by virtue of that person’s morals, intelligence, ability or temperament, but solely by virtue of shape, colour, texture and movement’ (Dover 1974a: 16).¹⁶ Everything I have said so far, I am afraid, is awfully elementary if not downright boring. But, in a way, that is exactly the point: these ideas have become part of our common sense when it comes to understanding the Greeks—how else could we impose some order on such a heterogeneous field? I am certain that they ¹² Xen., Symp. V. 2–8; Pl. Hp. mai. 295c–e. ¹³ In Eth. Nic. II 3, e.g., Aristotle’s list of the objects of choice consists of the kalon, the expedient, and the pleasant (καλόν, συμφέρον, ἡδύ, 1104b 31); in VIII. 2, he says that people love the agathon, the pleasant, and the expedient, without any indication that the difference is significant (ἀγαθόν, ἡδύ, χρήσιμον, 1155b 18–19). ¹⁴ Thgn. El. I. 255–6: κάλλιστον τὸ δικαιότατον· λῷστον δ’ ὑγιαίνειν· | πρᾶγμα δε τερπνότατον, τοῦ τις ἐρᾷ, τὸ τυχεῖν. Translations from the Greek, unless credited otherwise, are mine. ¹⁵ Cf. Dover 1974b: 69–73. ¹⁶ Cf. Thgn. El. II. 1259: ὦ παῖ, τὴν μορφὴν μὲν ἔφυς καλός, ἀλλ’ ἐπίδειται | καρτερὸς ἀγνώμων σῇ κεφαλῇ στέφανος. That may seem to lend support to Dover’s view. But it is worth pointing out that by qualifying καλός with μορφή, the verse suggests that the contrast is not between beauty and wisdom, but between one kind of beauty—beauty of appearance—and another—beauty of mind, both being equally instances of κάλλος.

100

Alexander Nehamas

are largely correct. I also know that I am not competent to discuss Greek usage in general. I am specifically concerned with the role of kalon in the thought of Plato and Aristotle,¹⁷ and although even here I feel unable to make any confident generalizations, I must confess that the opening pages of Plato’s Charmides make me very uneasy.

II The subject of this short but very complex dialogue is the nature of temperance (s¯ophrosun¯e). Its location is a palaestra, where young men, naked, practice their wrestling. In keeping with both, the dialogue begins with some witty back-andforth regarding Socrates’ weakness for attractive boys: ‘I am a broken yardstick when it comes to beautiful boys (kalous),’ he admits; ‘just about everyone who is of the right age seems beautiful (kalos) to me’ (154b). Although he insists that he is in all cases more concerned with the soul than he is with the body, he has to confess that when Charmides took a seat beside him, his principles were challenged: ‘When I saw inside his tunic, I was on fire. I lost my head. I felt that in these matters nothing can match the wisdom of Cydias’ advice regarding beautiful (kalos) boys: ‘‘The fawn should take care, when he comes face-to-face with the lion, not to end up as his meal.’’ ’¹⁸ We know that Plato was deeply suspicious of sexual desire and pleasure. Yet, having just associated kalos so strongly with sexual excitement and the loss of self-control (cf. 155c), he goes on without a moment’s hesitation to use the same word to describe the virtue of self-control itself! Asked by Socrates what temperance is, Charmides first replies that it is a kind of calmness or quietness (h¯esuchiot¯es) and immediately agrees, as if it were the most natural thing in the world, that temperance is kalon, allowing Socrates to argue that we cannot identify it with quietness, which is not kalon but its opposite, aischron (159b–160d).¹⁹ That is not in itself surprising: Plato regularly describes the virtues as kala (e.g. La. 192c–d; Prt. 349e). But the contrast between the sexual magnetism of Charmides’ beauty and its effect on Socrates, which are crucial to the dialogue’s opening scene, and the nobility of temperance, on which the philosophical discussion depends, is extraordinarily stark.²⁰ Perhaps to emphasize ¹⁷ The present essay, however, will remain centered on Plato. I hope to address Aristotle’s treatment of kalon another time. ¹⁸ Chrm. 155d–e; cf. Dover 1974a: 155–6 with n. 6. ¹⁹ That is not quite accurate; see the next paragraph for a necessary qualification. ²⁰ The first use of καλόν independently of its sexual connotations occurs at 157a5, where Socrates applies it to the kind of argument or discussion (λόγος) that leads the soul toward temperance. That is only a few lines after Socrates has spoken yet again of Charmides’ effect on him (156d1–2). (I assume that the use of the adverb kal¯os at 156a9 is too conventional to carry any specific semantic weight.)

Beauty of Body, Nobility of Soul

101

the continuity between Socrates’ dialectic and the rest of his behavior, the text seems to envisage a seamless transition from excitement to repose, from passion to moderation. But the use of kalos to pay a compliment both to a boy who provokes an excess of sexual desire and to the virtue that produces the restraint of considered judgment undermines that transition: to contemporary ears, at least, it seems to depend on little more than a pun. The transition would be smoother if kalon applied to things in virtue of a more abstract feature that both beauty and nobility have in common. The Charmides contains a suggestion that it does, and a hint of what that feature might be. We find both in Socrates’ argument for the claim that every activity of the body and of the soul is kalli¯on (more noble) when it is fast and brisk than when it is calm and slow (159c3–160b5). Every step of his argument appeals to one or another form of kalon—except once: when, just before reaching his conclusion, he says that those who perform the activities of the soul the fastest are most worthy of praise.²¹ Charmides does not protest, and Socrates immediately returns to kalon and completes his argument. Kalon is often translated as ‘praiseworthy’ or ‘admirable’ when people’s character, rather than their appearance, is primarily involved.²² And despite Dover’s claim that when it is applied to a person its meaning is ‘beautiful’, it is hard to believe that when Pindar writes, For with your [the Graces’] help that all delightful and sweet things are accomplished for mortals, as long as there is a skillful, beautiful (?—kalos) and splendid man,²³

kalos refers simply to the way someone looks. Rather, it connects the poet’s skill (sophos) with his renown (aglaos) by introducing his admirable ability to produce, with the help of the Graces, works that, like Pindar’s own, bring sweet delight to the world.²⁴ The powers and accomplishments of a poet are clearly matters of praise and rightful pride. Is beauty, however, worthy of praise? Some might think so, especially if the praise is directed at the beauty itself without giving credit for having it to the person whose beauty it happens to be. Aristotle would not. He took praise to ²¹ ἐπαίνου ἄξιοι, 160a10. There are twelve explicit and three implicit occurrences of the term between 159c1 and 160b5. ²² See, however, Woods 1982: 47: ‘Common to its [kalon] various applications is that anything so characterised is represented as an appropriate object of admiration.’ ²³ Pind. Ol., 14. 5–7: σὺν γὰρ ὔμιν τὰ τερπνὰ καὶ | τὰ γλυκέ’ ἄνεται πάντα βροτοῖς, | εἰ σοφός, εἰ καλός εἴ τις ἀγλαὸς ἀνήρ. ²⁴ Patricia Easterling has suggested to me that perhaps the three adjectives refer to three distinct persons and not one only, as (on the strength of the single tis in l. 7) I assume here: sophos might then designate the poet, kalos the athlete, and aglaos the ruler, each one of whom is involved in the events that surround an epinician ode (although ‘sweet’ and ‘delightful’ would seem to apply primarily to poems). It is clear, however, that even if we accept this intriguing hypothesis, kalos still cannot imply merely appearance: it takes more than good looks to be a great athlete even if, as we shall soon see, athletes were often praised for their beauty.

102

Alexander Nehamas

be appropriate only to the results of effort and intentional action. He thought beauty was neither praiseworthy nor, conversely, subject to blame because it is something that comes by nature or luck, while ‘blame and praise are given not to things that occur of necessity or by luck or in the course of nature, but to all the things of which we are ourselves a cause … virtue, vice and the resulting deeds’.²⁵ Perhaps he was right. Perhaps we should agree with him that of the various goods (which include honour, wealth, good fortune, bodily excellences,²⁶ and various other capacities) only the ethical and intellectual virtues are kalai because only what is within our own power is truly praiseworthy (Eth. Eud. VIII 3, 1243b 18–30). But Aristotle’s philosophical theory was not a reflection but a revision of ordinary practice, and he admits that in ordinary practice wealth, bodily excellences, and the like are what people most want and consider the greatest goods, what they most often praise each other for having. Ever since the old men of Troy caught sight of Helen from their seats high on the city walls, beauty was for the Greeks an object of praise and a reason for pride.²⁷ That is a very deep-seated attitude. It represents, in fact, the feelings of most people today: not much has changed in that regard since the time of Homer. For the Greeks, though, the privilege accorded to beauty depends greatly on the complex social dimensions of kalon, which connect appearance, social status, and, as we shall see, a certain sort of behavior. These connections follow the word, with a slight democratic twist, even when it appears in graffiti or on pottery—in statements of the type Alkibiad¯es kalos or Melieus kalos (‘Alcibiades is beautiful’ or ‘The beautiful Melieus’). Franc¸ois Lissarague, who has studied such inscriptions, points out that acclamations applied to the figures portrayed on various items of pottery reveal that ‘as a general rule … the kaloi are prominent and fashionable young men who were known by everyone in Athens’ (Lissarague 1999: 365). By and large, beauty—good looks—and social class go hand in hand; Cinderella is really a fairy-tale.²⁸ Moreover, a group of graffito acclamations discovered on the island of Thasos locate kalos in a semantic space that includes not only terms referring to appearance (eupros¯opos, kallipros¯opos, eusch¯em¯on, euruthmos) but also words that introduce social qualities like sophistication or urbaneness (philok¯omos, asteos) and even a combination of both: when a boy is praised for his urbane face or appearance (asteopros¯opos), Lissarague writes, ‘it is not a matter of sexual desire pure and simple, but an appreciation of beauty which has a social and ethical side to it’ (p. 360). Nor is it such a matter when a character in Aristophanes tells the Athenians that the Thracian king, Sitalkes, ‘was so ²⁵ Arist. Eth. Eud. II. 6, 1223a 9–14, trans. Woods (1982), slightly modified. See also Eth. Nic. III 1, 1109b 31–2: ἐπὶ μὲν τοῖς ἑκουσίοις ἐπαίνων καὶ ψόγων γινομένων, ἐπὶ δὲ τοῖς ἀκουσίοις συγγνώμην, ἐνίοτε δὲ καὶ ἐλέου. ²⁶ The word Aristotle uses here for the bodily excellences, which include beauty, is aretai. ²⁷ Hom. Il. III. 188–90. ²⁸ Or else the story shows that when it doesn’t reflect social status, beauty is one of the best means of securing it.

Beauty of Body, Nobility of Soul

103

amazingly fond of Athens, so truly your lover that he would write ‘‘Ath¯enaioi kaloi’’ on the walls’.²⁹ At least in such a context, then, to praise someone for being kalos was to express not just a reaction to his appearance but also an appreciation of something more than the good looks on which the reaction was primarily based. Historically, that something more was a social attribute, which is why kalos was frequently joined with ‘high-born’ (gennaios) and ‘rich’ (plousios). Socrates himself makes that commonplace connection in this very part of the Charmides (157b8) and in a famous passage of the Meno (71b6–7). The social here shades into the ethical, Theognis’ rather desperate verses indicate: Wealth, most beautiful (?—kalliste) and most desirable among the gods, with your help a man becomes good even if he is bad.³⁰

It is exactly because of that interpenetration that Theognis’ conservative instincts lead him to his constant warnings against the good mixing with the bad, the rich with the poor, the noble with the commoner (e.g. I. 15, 71). For better or worse, not just in aristocratic but in democratic societies as well, the beautiful, the rich, and the high-born are expected to behave in specific ways; when they do not, and depending on their overall attitude toward them, the rest of the world is either disappointed or wickedly pleased. Plato, I believe, developed the view that ‘virtue does not come to be from possessions; virtue is what makes possessions and everything else in private or public life good for human beings’ (Ap. 30b) precisely in order to explain why such people so often contradict our expectations and why we should not, in any case, expect them to behave better than anyone else.³¹ He took such things to be perfectly neutral, neither good nor bad. Aristotle was somewhat more generous, but still gave them only his qualified approval; these are ‘external goods’ or ‘goods of fortune’, but they are not kala unless they are used correctly, as practical wisdom dictates (Eth. End. VIII 3, 1249a ). Unlike the virtues, these goods are not praiseworthy in themselves, but only on account of what they enable a virtuous agent to accomplish.

III If that is right, to describe people as kaloi is to go beyond the features of their appearance and indicate an assessment of their status and behavior as well. It is to make, to that extent, a forward-looking judgment. Although based on an ²⁹ Ar., Ach. 142–4: καὶ δῆτα φιλαθήναιος ἦν ὑπερφυῶς | ὑμῶν τ’ ἐραστὴς ὡς ἀληθῶς, ὥστε καὶ | ἐν τοῖς τοίχοις ἔγραφ’· Αθηναῖοι καλοί. Cited by Dover 1974a: 111 along with a broad overview of the available evidence (pp. 111–24). ³⁰ Thgn. El. I. 1117–18: Πλοῦτε, θεῶν κάλλιστε και ἰμεροέστατε πάντων, | σὺν σοι καὶ κακός ὢν γίνεται ἐσθλὸς ἀνήρ. ³¹ On the grammar of this phrase, see Burnyeat 2003.

104

Alexander Nehamas

appreciation of people’s appearance—on what we can tell about them on the basis of limited information—it is neither limited to nor exhausted by it. That, as I will try to explain in what follows, is as it should be. We do expect more of the people we find beautiful than we do of others.³² To find anything beautiful—a person, a work of art, even a landscape—is to have a sense that it holds more within it than we have been able to see so far, and that what it holds is worth coming to see. Its still unknown attractions draw us closer to it. In the presence of a beautiful person, I am not happy merely to look (although I will do so, gladly): I also want to approach that person, to get to know and spend time with them. There can be many reasons—psychological, social, professional, ethical—to resist that desire (and be happy merely to look) in particular cases, but the desire itself is real and springs from a vague sense that the person that attracts me would somehow make my life better in ways of which I am not yet aware if he were part of it. The qualifications are essential; they show that if I actually decide to approach a beautiful person (that is, a person I find beautiful), I am putting myself at risk, willingly giving him the power to make me think and desire things I have not thought or desired before without yet knowing what they are. I am willing to change as a result of our interaction, although I cannot possibly know what that will make of me. To find something beautiful is not simply to be attracted to its appearance. For I can make sense of ‘appearance’ in this context only by thinking of it as whatever we can come to think or feel about something on the basis of a relatively casual inspection. To be attracted simply to the appearance of something is to set in advance a limit to the features of something that are relevant to me. A person’s appearance is exhausted by what lies already before me, and therefore the desires it provokes are for ends I can specify in advance—sexual pleasure, social advantage, or another well-defined benefit (that is not to put such ends down—only to distinguish them from the ends of love, the reaction that beauty provokes). If what I want from you depends just on your good looks, I already know what it is, and I know (or think I know) exactly who I shall be as a result. But when I am attracted to your beauty, however certain I am that what you will give me will make my life better in some way, I am also aware that I do not know either what it is or what will become of me as a result. A recent trend in evolutionary psychology, which takes both beauty and our reactions to it to be biologically determined and reflections of the selection value ³² It is important to note that this applies mainly to the people we find beautiful: i.e. to those whose beauty actually strikes and affects us—not, generally, to those whose appearance satisfies some set of generally accepted criteria of beauty. It is quite possible not to find such people attractive, while, as we will see below, it is equally possible (and more likely) that the people we find beautiful and attractive most often fail to satisfy such criteria. The discussion that follows does not concern what beauty ‘objectively’ is, if such a thing exists, but only what it is to find someone or something beautiful. In order to avoid complicating my language unreasonably, however, I will often speak simply of beautiful things. A more extended argument about these issues is made in Nehamas (2007).

Beauty of Body, Nobility of Soul

105

of reproductive success, implies that my attitude is untenably naive. Roughly, the idea is that attractive women are those who look most fertile—capable of having the most offspring—and attractive men are those who look most dependable—capable of protecting both women and children. So, whether we are aware of it or not, we all know what we expect from beauty: a better chance of propagating our genetic material. The claim is, first, that by and large members of different cultures find the same human features attractive and that their judgments are therefore biologically based; second, that the features on which they agree are indicators of reproductive success. In various experiments designed to establish this claim, people are shown a series of photographs of human faces, some representing individuals and others composite faces, combinations of two or more digitized images. When they are asked to rank them according to their beauty, they generally judge composite images to be more beautiful than simple ones and more composite images to be more beautiful than less composite ones. Since larger combinations of images tend to get closer to the average of the features they incorporate, such studies conclude that beauty is averageness. This may seem very surprising: beautiful objects of every kind seem to stand out against their background, not to melt into it. But what is average here is not beauty itself, but the size, proportion, and distribution of facial features. That sort of averageness, the argument goes, reflects optimal design, outstanding health, and, since evolutionary pressures tend to affect those who are found at the extremes of biological populations, the best chances of survival and genetic propagation.³³ As far as I can tell, it is not clear whether people are responding to the faces the photographs are of or to the photographs themselves.³⁴ That may be a problem since, as Aristotle was already aware (Poet. 1448b9–12), a representation of a beautiful object is not necessarily a beautiful representation, and a beautiful representation of an object is not necessarily a representation of a beautiful object. But it may be a minor problem, which we can put aside, agreeing that judgments of the beauty of faces really converge among cultures. What follows from that? What follows, I think, is that whatever it is that these experiments measure is something other than beauty. For one thing, the evidence on which these judgments are made is severely limited: the faces, for example, are as expressionless as possible so that nothing will ‘interfere’ with their evaluation. But why suppose that the expression of a face interferes with the perception of its beauty instead of being one of its essential elements? Why, more generally, agree that the least possible evidence on which any judgment can be made is also the most basic ³³ The literature on this issue is extensive. See, among others, Cunningham et al. 1995; Jones and Hill 1993; Jones 1995; Langlois et al. 1987a, b, 1990; Langlois and Roggman 1990; Langlois et al. 1991, 1994, 1995; Perrett et al. 1994, 1998; Little and Perrett 2002. I discuss this research in detail in Nehamas 2007: ch. 2. ³⁴ e.g., Langlois and Roggman (1990: 116, 118, 119), speak of ‘digitizing faces’ rather than photographs and consistently refer to their photographs as ‘faces’.

106

Alexander Nehamas

evidence on which every judgment must be made? It is only natural that the most elementary and therefore most primitive judgments of attraction, based on the least possible evidence, must appeal to the most abstract and generic features of their objects and that those features are probably biologically determined: what else could they be? There is no reason to believe that such extraordinary judgments are also the most fundamental, that they are at the center and not at the edges of our experience, or that they reveal the essential nature of ordinary judgments. Moreover, the judgment of beauty is nothing like the detached evaluation that people are being asked to make here: it is one thing simply to describe one of these faces as ‘beautiful’ or ‘attractive’ (the terms are used interchangeably in these studies) and quite another actually to be attracted by it, one thing to believe that people are likely to be struck by a face and another to be in fact struck by it. The judgment of beauty has an inseparable affective component that is missing from the reactions that these experiments elicit and the relationship between them remains completely unclear. More important, the affective component of the judgment of beauty is forward-looking. It involves the desire to approach the beautiful person, a feeling that somehow we belong, and must continue to be, together—a bit like Aristophanes says that each half of our original divided self feels for the other (Symp. 189c–193d). That is the least beauty demands, and nothing like it occurs under these controlled conditions, which do not begin to touch its complexity. If these experiments do measure beauty, then, by their standards, most people in the world are not beautiful. Yet most people in the world have at one time or another both loved and been loved by someone—not always, unfortunately, the same person—and, as Diotima says, it is impossible to love what one does not find beautiful: er¯os is by nature a lover of beauty (Symp. 203c).³⁵ Shakespeare’s Dark Lady sonnets contain the only instance of a person who is in love with someone he actually finds ugly and, for just that reason, have caused no end of trouble to their readers.³⁶ To love someone—not as a Christian loves God’s children³⁷ but as an individual—is inseparable from finding them beautiful. Love has already died when one day I am no longer moved by my lover’s beauty, when I can look at her face dispassionately and measure, so to speak, the quality of her features. Love can survive the most bitter hatred—Catullus knew that³⁸—but cannot live for a moment with ugliness: hate is not its opposite; indifference is.³⁹ Beauty, ³⁵ In the Symposium, Socrates attributes his speech to the holy woman Diotima. In what follows, I will use their two names indiscriminately. ³⁶ More on this in Nehamas 2007: ch. 2. ³⁷ Not even as parents love their children, if the controversial understanding of love in Frankfurt 2004 is correct. ³⁸ Catull. Carmina, 85. ³⁹ One peculiar but suggestive asymmetry between love and hate, which suggests that they are not contraries, is that while it is impossible to fall in love with someone with whom you have

Beauty of Body, Nobility of Soul

107

moreover, is always manifested in a lover’s appearance, and we are only making things easy for ourselves when we say that some people love each other not for their looks but for their kindness, their sensitivity, or their intelligence instead. The psychological, mental, and moral qualities of the people we love are revealed in their face and bearing, in how they—literally—look to us. The ‘inner’ cannot be separated from the ‘outer’. When Emma Bovary, for example, finds herself attracted again to her mediocre husband because they both believe that he will perform an operation that will make him famous and give them the life Emma has always dreamed of, she notices, ‘with some surprise, that his teeth were not at all unsightly’. When, inevitably, Charles botches the operation, Emma finds that ‘everything about him exasperated her … his face, his clothes, what he did not say, his entire being, his very existence’ (Flaubert 2004: 157, 165).

IV Plato’s startling and original insight was to see what contemporary psychology missed: the essential connection of beauty and love, love’s forward impetus and the beckoning of beauty. The Symposium tells us that er¯os is always of the beautiful, never of the ugly (201a, 202d), that it impels the philosophic lover to reach ever further in his pursuit of beauty, which draws him ever closer toward it. It was Plato, through Socrates’ speech, who gave these features of love and beauty their first philosophical articulation. Socrates begins by defining er¯os as the desire to possess the beautiful (204d3). He then, controversially, replaces the beautiful with the good (agathon), redescribes the object of er¯os as the possession of the good, that is, as happiness (eudaimonia, 204e6–205a7), and adds that the desire to possess the good is actually the desire to possess it forever (206a5–12).⁴⁰ He continues, following in Diotima’s footsteps, to describe er¯os more specifically as the desire not simply to possess the beautiful but ‘to give birth’ in it (206b5–8, e5). Every human being, Diotima has told him, is pregnant, and the urge to give birth is part of our nature. But giving birth is possible only in the presence of beauty, because giving birth is a divine feature and an immortal aspect of mortal things, and only beauty never had any direct contact (which includes letters from or pictures of that person), simply on information supplied to you by a third party, it is quite possible to come to hate another through an account of their personality or actions. ⁴⁰ David Sedley has remarked to me that although Plato does believe that all good things are beautiful, there is no reason to attribute to him the stronger claim that all beautiful things are good: ‘the substitution [of ‘‘good’’ for ‘‘beautiful’’] assumes nothing about the identity of the two’, he wrote in a private communication. I don’t see, however, how Plato could have allowed that substitution if he did not also believe that all beautiful things are good. Apart from issues of logic, the substitution would then explain only what it is that we want in regard to the beautiful things that are also good (to possess them and so be happy). It would cast no light on our desire for those beautiful things that, on Sedley’s hypothesis, are beautiful but not good.

108

Alexander Nehamas

is appropriate to the divine. To the extent, then, that er¯os is the desire to possess beauty and goodness forever, it is also, he argues even more controversially, a desire for immortality (206e8–207a4). All mortal things—the beasts as well as every human being—desire immortality. But unlike immortal things, which remain forever unchangingly the same, mortal things can approach immortality only by leaving behind something that is like them and perpetuates them into the future (207a5–208b6).⁴¹ At this point, Socrates attributes to er¯os a complex hierarchical structure. Everyone, he has already said, is pregnant both in body and in soul (206c2–3). Those who are pregnant in body perpetuate themselves by the easiest and least admirable method—biologically, by leaving human children behind them (208e1–5). But some people are pregnant in soul even more than in body.⁴² Their desire for immortality manifests itself as a thirst for ‘immortal virtue and fame’ (208e5–209e4). These people turn to paederasty⁴³ and, in company with someone beautiful in body or in both body and soul, produce beautiful logoi —accounts, discussions, views—concerning virtue, especially the wisdom and temperance that make life in society possible. The best among them are the great poets and legislators, whose works—‘more beautiful and more divine’ than human children (209b8–e4)—have won them immortal fame and glory: their works are the offspring they are happiest to leave behind, an intellectual rather than a biological progeny. The highest kind of love, though, the love of Diotima’s ‘perfect mysteries’, is motivated not by the desire for fame but by the thirst for wisdom. Its progeny is also intellectual, but it is philosophy, not poetry or law, to which its initiates are devoted. And it contains a further hierarchy of its own, which begins with the physical beauty of a single boy and gradually rises through the beauty of the human body in general, the beauty of the soul, of laws and institutions and the sciences to the Form of beauty, the nature or essence of beauty itself, which turns out to have been the real object of er¯os all along.⁴⁴ But if every lover is ultimately drawn to the Form of beauty, glimpsed obscurely through everything else in the world that is to some degree beautiful, it might seem that nothing that lies below the Form in Plato’s ladder of love (not to mention ⁴¹ Some of the obscurities of this idea are well discussed in Burnyeat 1977. I will not be addressing the logical issues raised by this stretch of Socrates’ speech in this essay. ⁴² More on this below, p. 115. ⁴³ It is not absolutely clear that everyone at this level is a man, since Diotima uses the story of Alcestis, who offered to die in place of her husband, to introduce this category (208d2–6). Alcestis (like Achilles and Codrus) sacrifices herself for the sake of ‘her virtue remaining immortal and for glorious fame’ (ὑπὲρ ἀρετῆς ἀθανάτου και τοιαύτης δόξης εὐκλεοῦς, 207d7–8)—the motive of every other lover in these ‘lower mysteries’ (cf. ἀθάνατον κλέος καὶ μνήμην, 209d3). But everyone whom Plato refers to as worthy of fame on account of ethical or intellectual accomplishment in what follows is a man. Perhaps he is making an implicit distinction between courage and the other ethical virtues. ⁴⁴ That, at least, seems to be suggested by the idea that the lover’s efforts are all made for the sake of the Form of Beauty (τοῦτο … οὗ ἕνεκεν, 210e5–6; ἐκείνου ἕνεκα τοῦ καλοῦ, 211c2).

Beauty of Body, Nobility of Soul

109

the objects of the lower kinds of er¯os) is ever actually loved—at least not, as we often like to say today, ‘for itself’ and not for the hint of the Form, the trace of real beauty, within it. The possibility that whenever a lover realizes that one kind of beauty is greater than another (as the beauty of the human body in general, say, is greater than the beauty of a single one, 210b3–6) he chooses the higher over the lower kind without a second thought reinforces that suspicion. For it implies that, whatever the boy he leaves behind may feel, the lover perceives no change in that transition: he still loves the same thing—beauty, just more of it now than before. But that is a cold and cruel kind of love, a heartless passion. It was that thought which prompted Gregory Vlastos (1981) to criticize Plato for failing to see that love is first and foremost the love of individuals, and the questions he raised have ever since been central to the interpretation of the Symposium. Vlastos (1981: 31) attributes to Plato the view that ‘what we are to love in persons is the ‘‘image’’ of the Idea in them’. We love them, that is, only to the extent that they are good and beautiful, but since none of us is perfectly good or beautiful, love is not directed at us, blemishes and all: ‘The individual, in the uniqueness and integrity of his or her individuality, will never be the object of our love.’ Human imperfection, though, would imply that if I love you for your virtues, I cannot love you for yourself, only if Plato also believed that if I love you for your goodness, your beauty, or, for that matter, for your yellow hair, what I really love is not you but your goodness, your beauty, or your yellow hair instead. Although Plato may have thought so, we cannot just assume that he did: many people—and even some philosophers—believe that we love people for particular reasons without feeling that we do not therefore love them for themselves.⁴⁵ The issue is complex, and the question remains open: we may love the image of the Form in a person without, for that reason, loving the Form of which it is the image and not the person who bears it. Perhaps, though, the reason is inherent in Plato’s conception of the philosopher’s ascent, which Vlastos describes as follows: Persons evoke er¯os if they have beautiful bodies, minds, or dispositions. But so do quite impersonal objects—social or political programs, literary compositions, scientific theories, philosophical systems and, best of all, the Idea of Beauty itself. As objects of Platonic love all these are not only as good as persons, but distinctly better. Plato signifies their superiority by placing them in the higher reaches of that escalated figure that marks the lover’s progress, relegating love of persons to its lower levels. (Vlastos 1981: 26)

That is true: Plato considers the love of individuals inferior to the love of abstract programs or theories, and that, in turn, inferior to the love of beauty itself.⁴⁶ ⁴⁵ e.g. Brentlinger (1970) and Keller (2000) have denied that loving someone because of their features implies that the love is directed at their features and not at them. Kolodny (2003) claims that such a ‘quality’ theory of love is to be rejected, but does not address directly views like Brentlinger’s or Keller’s. See also Velleman 1999: 362–4. ⁴⁶ True, i.e., subject to an important qualification, as we shall see below.

110

Alexander Nehamas

But that is not to say that those who stop at the lower steps of the scala amoris do not love the person or the institution that inspires them. Even if the love of ‘impersonal objects’ cancels attachments to particular individuals, all that follows is that a life devoted to politics (nomoi kai epit¯edeumata) or learning (math¯emata) is better, more valuable, and, in the end, happier and more fulfilled than the private lives of most of the people in the world. That would not have been news to Plato’s Greek audience—although his reasons certainly were. Nor is Plato the only philosopher, in Greece or elsewhere, to think that purely private lives of no distinction are of little value. That is not to deny that individuals can be loved; it is only to claim, rightly or wrongly, that life is at its best when it is devoted to something else instead. Plato might have believed that the Form of beauty is the only object of love (and so have disqualified not only individuals but also ‘social or political programs, literary compositions, scientific theories, philosophical systems’) if he thought of desire in the Symposium as he had described it in the Gorgias. For he there has Socrates argue that if we engage in a course of action for the sake of something else, we desire only that for the sake of which (hou heneka) we engage in that action and not the action itself:⁴⁷ we undergo painful medical treatment, which we certainly do not want, for the sake of being healthy, which we do, and undertake dangerous voyages for the sake of becoming wealthy. ‘Wisdom, health, wealth and similar things’, he continues, are good and desired for themselves; their contraries are bad and never desired; but medical treatment, traveling, sitting, walking, running, and almost everything else besides are neither good nor bad: they lie between (metaxu) good and bad, and for that reason we never pursue them for themselves but only for the distinct good to which we believe they lead (467e–468c). We have already seen that the lover’s efforts are all made for the sake of the Form of beauty (210e5–6, 211c2),⁴⁸ and if means and ends here are related as they were related in the Gorgias, the Form of beauty would be the only thing with which every lover, knowingly or not, is in love. If that were so, though, it would have been natural for Plato to say at some point that nothing that is loved for the sake of beauty itself is either loved or beautiful in its own right—just as taking medicine or walking is neither good nor desired in its own right. Nothing other than the Form would then be beautiful.⁴⁹ But the Symposium gives no hint of ⁴⁷ Grg. 467d6–e1: ᾿Εάν τίς τι πράττῃ ἕνεκα του, οὐ τοῦτο βούλεται ὃ πράττει, ἀλλ᾿ ἐκεῖνο οὗ ἕνεκα πράττει. ⁴⁸ It may be worth asking whether ἐπανιέναι, also at 211c2, should be rendered as ‘going upwards’ (from ἐπανέρχομαι) or ‘letting go’, ‘dismissing’ (from ἐπανίημι). If the latter, the earlier stages of the ascent would appear to be valuable purely as a means to reaching the Form—just as medicines are valuable only as a means to health—and of no further use once they have served their purpose. ⁴⁹ It might be, of course, that the Form is not involved in the lower stages of er¯os but only in the ‘perfect’ mysteries that involve the philosopher (cf. 209e–210a). But Diotima’s statement that even

Beauty of Body, Nobility of Soul

111

such an attitude on his part, even when Diotima explains how superior the Form of beauty is to everything that has led up to it (210e2 ff.). This is particularly surprising, because the ways in which the Form is superior—it is beautiful in every respect, in relation to all things, at all times, and so on (211a1–5)—are the same features whose lack makes perceptible things neither beautiful nor ugly, neither large nor small, neither heavy nor light, and therefore in general between (metaxu) being and not being in the Republic (479d7).⁵⁰ Although the notion of ‘between’ plays a crucial role in the Symposium as well, the argument surrounding it is markedly different. Socrates reports that he used to believe that nothing separates beauty from ugliness or wisdom from ignorance until Diotima convinced him that er¯os is neither beautiful nor ugly, neither wise nor ignorant, neither mortal nor immortal, but—precisely—between such extremes. Her reason is that, since er¯os is the desire to possess the beautiful, and no one desires to possess what he or she already has, er¯os lacks beauty, and cannot therefore be beautiful.⁵¹ Its intermediate status, however, depends on its having a beautiful object, and so the argument presupposes that love is directed at things that are beautiful without the qualifications of the Gorgias or the Republic.⁵² Plato is not doing metaphysics here. Socrates, like Agathon (199c3–201c9), agrees that love is of beauty on the basis of everyday experience. And everyday experience, like the Symposium itself, tells us that beautiful things are everywhere around us.⁵³ Plato, of course, may be exploiting everyday experience in order to undermine it. Perhaps he is pulling us gently into agreeing with him on the basis of the sensible view that beautiful things inspire love and desire, so that he can gradually lead us to realize that the sensible view is wrong and that it is not the beauty of these things but the beauty of beauty itself that inspires love and desire. Plato could have adopted such a strategy in the Symposium. But he did not. He never denies that the objects of the two lower kinds of er¯os are beautiful, nor does he the ‘lower’ mysteries are themselves (to be pursued? to be understood?) for the sake of the perfect mysteries (209e5–210a2) suggests that the lower stages of er¯os, too, may bear such a relationship to the Form: see Dover 1980: ad loc. ⁵⁰ In fact, Plato’s position here is considerably more complex, but that doesn’t affect the point I am making. ⁵¹ Nor is it ugly, for then, in parallel with the argument to the effect that the ignorant never pursue wisdom, since they are not even aware that they lack it, er¯os wouldn’t desire, as it actually does, either beauty or wisdom (204a1–b5). ⁵² i.e. they do not belong to the third class of the Gorgias —things that are not good or bad, beautiful or ugly, but that occasionally lead to goodness and beauty—or to the polla of the Republic —things that are not either beautiful or ugly or both beautiful and ugly or neither beautiful nor ugly. The translation that best brings out the scope of Plato’s point in the Republic is that of Waterfield (1993: 201). ⁵³ Perhaps Agathon and Socrates agree that the objects of er¯os are beautiful because, in the end, the only object of er¯os is the Form of beauty itself, whose own beauty is never in question. But no one at this stage of the drama—neither Agathon nor Socrates nor the readers of the Symposium—has ever heard of beauty itself or the rest of the hoary metaphysics of the theory of Forms. Plato has given no reason to either the characters or the readers of the dialogue to agree that love is for beauty other than their everyday experience.

112

Alexander Nehamas

do so once he begins describing the philosophic lover’s ascent. The beauty of what lies on the way to the Form does not prove to be an illusion once the lover sees more beauty beyond it: only its brilliance diminishes, as the moon’s radiance wanes in the light of the sun. Yet, the lover is to ‘look down’ on his passion for a single body and think little of it when he discerns the beauty that is common to all bodies.⁵⁴ Doesn’t he then cease to find the boy that started him on his way beautiful? No—because it is, without a doubt, the intensity of his passion for the boy, and not the boy (and perhaps not even the passion itself), from which he must turn away.⁵⁵ That is, in fact, exactly what an important passage in the Republic (474c–475e) suggests. Socrates is trying to explain what a philosopher—a notion introduced here for the very first time—is, and he begins by describing Glaucon, with whom he is talking at the time, as ‘a lover of boys, an erotic man’ (philopais kai er¯otikos). Such men, he continues, always have a reason for finding every boy of the right age attractive: a snub nose is pert, a hooked nose is regal, one that falls in between is perfectly proportioned; dark boys are manly, pale ones are children of the gods, and the word ‘honey-yellow’ speaks for itself. Socrates may be speaking tongue in cheek here, but his point is serious: those who love the beauty of boys in general love the beauty of every individual beautiful boy; whether ‘true’ love is or is not exclusive in the manner that is now canonical is simply not an issue. His point is serious, because it allows him to introduce the idea that like the lover of boys, the lover of wine, of honor, of sights and sounds, and the lover of wisdom (especially the lover of wisdom)—the philoinos, the philotimos, the philotheam¯on, the phil¯ekoos, and finally the philosophos —is in love with everything that belongs, so to speak, to the ‘field’ to which his desire is directed. The philosopher loves all wisdom. The lover of bodily beauty, then, does not abandon the boy who first sparked his desire—he loves all boys, much as Don Giovanni, who also has a different reason in each case, loves every woman. Nothing could be more surprising than this extraordinary convergence between Plato and Mozart, unless it is the ⁵⁴ ‘When he grasps this, he must become a lover of all beautiful bodies, and he must think that this wild gaping after just one body is a small thing and despise it’ (ἑνός δὲ τό σφόδρα τοῦτο χαλάσαι καταφρονήσαντα καὶ σμικρὸν ἡγησάμενον, 210b5–6); cf. Price 1989: 39. ⁵⁵ This passage is discussed much less often than it deserves, especially by those who find in Socrates’ speech an impersonal, almost inhuman sort of love, despite the fact that most translations I have consulted render the passage correctly. One scholar, however, translates it ambiguously, allowing such an impersonal reading to insinuate itself in the reader’s mind as if it was part of Plato’s text and not the product of a wilful interpretation: the lover, Martha C. Nussbaum writes, ‘sees that he ‘‘must set himself up as the lover of all beautiful bodies, and relax his excessively intense passion for one body, looking down on that and thinking it of small importance’’ ’—leaving it unclear whether ‘that’ and ‘it’ refer to the lover’s passion for one body or that body itself. That makes it easier to charge the passage, on the assumption that ‘all beauty, qua beauty, is uniform, the same in kind’, with advising the lover to abandon the beauty of the boy’s body for the beauty of the body in general and, worse, with ‘making the related the same, the irreplaceable replaceable’ (Nussbaum 1986: 179).

Beauty of Body, Nobility of Soul

113

fact that Socrates’ introduction of philosophy in the Republic and perhaps into Western thought as well is the actual source of Leporello’s ‘Catalogue Aria’ in Don Giovanni!⁵⁶ The connection is established through Lucretius’ De Rerum Natura, where Socrates’ joke has already been given a heterosexual spin, through Molière, who translated all of Lucretius into French and used the passage in The Misanthrope, and Lorenzo da Ponte, who read widely in Molière and adapted his libretto for the opera from Molière’s Don Juan.⁵⁷ We should not allow this connection, however, to mislead us into thinking that Socrates is advocating a betrayal of the boy with whom the philosophic lover begins his ascent. Here I must disagree with A. W. Price (1989: 46–7), who thinks that at this stage the lover ‘is at least unfaithful to [the boy] and may desert him altogether,’ although he does not believe that the betrayal is sexual: ‘What is envisaged is not precisely sexual promiscuity: the lover was aim-inhibited (as Freud would say) from the beginning, for his attachment to one body only produced words (210a7–8). Hence the only Don Juanism in question is one of attraction, not of gratification.’ Price finds such ‘promiscuity’ in Socrates’ own susceptibility to beauty as Alcibiades describes it in his own speech in the Symposium: ‘He’s crazy about beautiful boys; he constantly follows them around in a perpetual daze’ (216d2–3). Yet even if we could describe Socrates as a lover of all bodies, his passion has nothing of infidelity: he never abandons one youth for another.⁵⁸ The fact is that the second stage of the ascent makes contemporary readers uneasy, for we assume that people who love more than one ‘body’ must do so for selfish and exploitative reasons: so strongly are our intuitions shaped through the values of monogamy. There is then a strong temptation to minimize the sexual contact involved or to convince ourselves that the lover is no longer interested in any particular body but only in body in the abstract—not a promising sexual object.⁵⁹ But instead of thinking of the ⁵⁶ Mozart, Don Giovanni, Act I, Scene 4: ‘Nella bionda, egli ha l’usanza | Di lodar la gentilezza— | Nella bruna, la constanza, | Nella bianca, la dolcezza! [etc.]’. ⁵⁷ Some will think, of course, that the Don is falsifying his experience simply in order to add more conquests to his ‘list’, but his attitude is much more complex. With a comic twist, he echoes Socrates’ description of Glaucon when he justifies his deceptions by attributing all of them to love: ‘Whoever is faithful to one betrays the others. I, whose emotions encompass all, love them all without exception’ (II. 1; cp. Molière, Don Juan, I. 217–25). Is he a hypocrite? The difficulty of answering this question is part of the reason why the opera continues to be fascinating. The intermediate links are Lucretius’ De Rerum Natura, IV. 1160–70 (‘nigra melichrus est, inmunda et fetida acosmos [etc.]’), which was Molière’s direct source, and The Misanthrope, II. 4 (‘La pˆale est aux jasmins en blancheur comparable; la noire à faire peur, une brune adorable [etc.]’). ⁵⁸ That is not Socrates’ place on Plato’s ladder of love. Alcibiades’ story about Socrates’ refusal to have sex with him (218d–219d) shows that Socrates, if we are to find a place on Plato’s ladder for him at all, has reached at least the stage where the beauty of laws and institutions has become apparent: only then do lovers realize ‘that the beauty of bodies is unimportant’ (σμικρόν τι, 210c3–6), and only then do they begin to give sex a secondary role in their relationships. ⁵⁹ Price (1989: 47) suggests both responses in his discussion. Regarding the first, I can’t agree with him that the lover’s ‘attachment to one body only produced words’—either in the case of the lovers of fame or in that of the beginning lovers of wisdom. He intimates the second when he writes,

114

Alexander Nehamas

lover seducing as many beautiful boys as possible, we would do better to imagine him ‘giving birth to beautiful logoi’ with as many beautiful boys as possible. Instead of suspecting that seeing the beauty of the body leads him to betray his rightful lover (and, of course, all the others as well), we would do better to insist that as long as he is wrapped up in one boy only, he is depriving others of his advice. Unlike its modern readers, the Symposium pays no serious attention to the question of whether he has sex with some, with most, or even with all of them. Nothing, in any case, prevents a lover from continuing to have a favorite while seeking the company of other young men; Socrates is constantly in pursuit of the young (though in his case, as we have seen, sex is not what he is after), but Alcibiades continues to have a special place in his life (Prt. 309a–b; Grg. 481d; Symp. 213b–d). Nothing Plato has said so far implies that the philosophic lover discards the objects he meets on his way as he continues his ascent. That is as it should be. Although I consider Dostoevsky a far greater writer than Ian Rankin, I do not for that reason dislike Rankin’s mysteries—nor did I, once I read Dostoevsky, stop reading mystery novels altogether. Although at moments Plato may have believed that it was wrong of me not to have done so, nowhere in the Symposium does he even suggest that it is wrong to love such lesser mysteries—only that those who do are not as happy as those who are devoted, say, to Crime and Punishment instead. As the Phaedo might have put it, the beauty the lover leaves behind neither withdraws nor is annihilated when a greater one emerges beyond it (102d–e). Plato never even suggests that the lover who realizes that the beauty of soul is ‘more valuable’ than the body’s also realizes that he was wrong to have valued bodily beauty in the first place. Vlastos (1981: 23) implies that he does: ‘At the next level, higher in value and still more energising, [Plato] puts the love of mind for mind, expecting it to prove so much more intense than skin-love that merely physical beauty will now strike the lover as a ‘‘small’’, contemptible, thing.’ But, first, ‘contemptible’ is too strong a term for smikron, which usually means ‘negligible’ or ‘unimportant’.⁶⁰ I know what it is to feel that to have loved some particular person was a mistake: that is not the feeling Plato attributes to the philosophic lover. Second, the passage to which Vlastos refers here, τὸ περὶ τὸ σῶμα κάλλος σμικρόν τι ἡγήσεται εἶναι (210c5–6), as I have said already (see n. 58), applies only to a higher stage of the ascent—to lovers who have discerned the beauty of laws and institutions (210c3–6). Only then does sex become at most of secondary importance to the philosophic lover (though it is not yet completely abandoned—that, as we shall see, occurs only at a higher in connection with the passage of the Republic we discussed above, that ‘the generosity of response [of the φιλόπαις] should inspire in a man pregnant in soul a non-particularized love-poetry inspired by, and intended for, ingenuous youth in general’. ⁶⁰ Liddell and Scott 1968: s.v. mikros, I. 3.

Beauty of Body, Nobility of Soul

115

level of the ascent). The ‘love of mind for mind’ is much more intimately tied to what Vlastos contemptuously dismisses as ‘skin-love’. The fact is, though, that nothing in Socrates’ speech can rightly be called ‘skin-love’. Not only the perfect, but also the lesser mysteries involve from the very first and throughout giving birth to kaloi logoi (209a ff., 210a7–8), without, of course, making er¯os purely ‘mental’ or ‘spiritual’. Plato writes very carefully: the men who turn to paederasty are ‘even more pregnant in soul than they are in body’ (ἐν ταῖς ψυχαῖς κυοῦσιν ἕτι μᾶλλον ἢ ἐν τοῖς σώμασιν 209a1–2), just as earlier he had Diotima say that ‘all human beings are pregnant both in body and soul’ (206c1–3). These men leave neither the body nor even women behind—how could they, if Athenian paederasty, unlike most modern (and some ancient) homosexuality, was perfectly compatible with marriage and children,⁶¹ while Socrates was himself married and the father of at least two sons? More important, sex is important to paederastic relationships even after the lover turns to the beauty of the soul;⁶² as we just saw, sex does not become unimportant until the next level of the ascent, when laws and institutions enter the picture. Why, in any case, should we assume that when the beauty of the body becomes ‘unimportant’, it ceases to matter altogether? A careful reading of the Symposium shows that the discussion of sex is itself less important to Plato and more important to his philosophic lover than the tradition surrounding him has been able to admit. The philosophic lover, the lover of fame, and the lover of women are united by their common desire to ‘give birth in beauty’, and are distinguished from one another by the motive that leads each to want to give birth. These three different motives correspond to the motives that Plato distinguishes in the Republic’s divided, tripartite soul, which makes in that way a clear if indirect appearance in the Symposium. Lovers of women, although Plato describes them as ‘pregnant in body’, are ruled by their soul’s appetitive part.⁶³ Appetitive desires are common ⁶¹ See, e.g., Cohen 1991: 195 and Dover 1980: 171. Dover (1980: 163) does not emphasize the inclusiveness of paederasty and writes, ambiguously, that ‘those men who are ‘‘fertile in body’’ fall in love with women and beget children … but those who are ‘‘fertile in soul’’ transcend that limitation’. ⁶² This is obscured if we assume that the lover who discerns the beauty of body in general turns to the beauty of the soul of one particular boy; so, e.g., Price (1989: 46) writes that at that stage, ‘I prize some boy’s soul more than his body. ( This could be the same boy as within [the first stage], but may well be a new one.)’ If the new relationship was sexual, it would be difficult to distinguish this higher level from the very first, which also involves sex with a single boy, unless the earlier relationship was purely sexual (true ‘skin-love’) and nothing else. And since it is clear that even at the first stage more than sex is involved, it becomes necessary to declare that the third level is independent of sex altogether in order to distinguish it from the first. If, on the other hand, we take it that the lover retains at this third stage all the youths with whom he surrounded himself at the second, the problem doesn’t arise—he now loves them for the beauty of their soul, and he can even love boys who are not as conventionally beautiful provided their soul is of the right kind (καὶ ἐὰν ἐπιεικὴς ὢν τὴν ψυχὴν τις κἂν σμικρὸν ἄνθος ἔχῃ, ἐξαρκεῖν αὐτῷ, 210b8–c1). None of these relationships need be asexual, although Plato has nothing to say about this one way or the other. ⁶³ Even in the Phaedo, where the distinction between body and soul is most pronounced and the body at first seems as if it were a completely self-motivating agent, all desires are ultimately due to

116

Alexander Nehamas

to every human being (which is another way of saying that Plato does not believe that philosophers leave them behind⁶⁴). Others, though, are not as widely shared—more accurately, they are not as strong in all as they are in some (all human beings are pregnant both in body and soul). There are two kinds of such desires. The first characterizes the lesser mysteries. Plato’s emphasis on the love of honor (philotimia, 208c3), the desire for glory (kleos, 208c5, 209d3), and the craving for fame (doxa, 208d8), and memory (208d5, 209d3) makes the connection between this sort of er¯os and the ‘spirited’ (thumoeides) part of the soul, which thirsts for tim¯e and doxa, unmistakable (cf. Resp. 580d–583a). The second kind of desire belongs to the lover of the perfect mysteries, who pursues beauty itself under the rule of the reasoning part of his soul, which is devoted to learning and philosophy (philomathes, philosophon, Resp. 581b9). Since each part of the soul, as we know from the Republic, has its own appropriate pleasure, so each part of the soul, the Symposium implies, has its own appropriate er¯os, directed at the objects whose beauty each different part of the soul appreciates. For that reason, just as the pleasures of the soul differ in immense degree but are still, for all that, pleasures, so every object of er¯os, however humble in comparison to the beauty of the Form, is still beautiful, and, however dimly, it reflects the Form’s light.⁶⁵ The parts of the soul may have different objects—in the Republic they are profit or gratification, honor or victory, and wisdom—but the pleasure each part finds in its object and the beauty of the object that inspires its love are the same for all. The Form of Beauty, then, may be more beautiful than everything else, and the intensity of the true philosopher’s love may dwarf our everyday feelings, but since er¯os is essentially the desire for beauty and is certainly felt by everyone, beauty is not the exclusive property of the Form. It is a feature of the world around us. Lovers of sex, lovers of fame, and all those lovers who start on the way to the Form but fail to reach it can be happy to some degree or another, because they all secure some measure of immortality for themselves.⁶⁶ Only the philosophic lovers, though, achieve a vision of beauty the soul. Some, however, occur only when the soul is embodied and can be described as desires of the body. See Beere 2003. ⁶⁴ In the Republic Plato makes explicit provisions for the sexual and reproductive life of the guardians (457b–462a, and note in particular 458c–d). Both guardians and philosophers are temperate, but by no means celibate. ⁶⁵ Plato may be joking in the Republic, when he calculates that the life of the virtuous king is 729 times more pleasant than the life of the tyrant (587e), but his desire to determine which of the three kinds of life (each corresponding to the dominance of a different part of the soul) is most pleasant (576b–588a) implies that the pleasures involved are the same in kind and differ only in degree. For a convincing derivation of the number 729, see Kang 2006. ⁶⁶ It might be tempting to say that the lowest lovers only believe (wrongly) that they secure happiness and immortality in view of 208e3–5: these people provide ‘themselves through childbirth with immortality and remembrance and happiness, as they think, for all time to come’. But the phrase ‘as they think’ (ὡς οἴονται) does not impugn their belief as a whole, but only their optimism as to the lasting nature of what they have secured.

Beauty of Body, Nobility of Soul

117

itself, which imparts genuine beauty and goodness to every one of their actions and, by permeating every single aspect of their life, imbues it completely with happiness.

V The connection between beauty, goodness, and happiness comes to the fore when Plato takes one of the most radical and difficult steps in his gradual but startling transformation of er¯os from an urge for reproduction to the practice of philosophy: when he characterizes er¯os as primarily a desire to give birth in the presence of beauty, shifting his emphasis from thinking of love as an urge to possess beautiful things to conceiving it as an impulse to create them instead (205b7 ff.). These two ideas together allow Plato to claim that the philosopher, who gives birth in beauty itself, ‘does not give birth to images of virtue, not being in touch with an image but, since he is in touch with the truth, to true virtue’ and becomes happy, dear to the gods, and as truly immortal as a human being can be (211e4–212a7). But the close connection between beauty and goodness that is so crucial to Plato is full of difficulties, and his transformation of er¯os from possession to production, from desiring something external to bringing forth something from within, remains deeply obscure. The very idea of desiring to possess the object of one’s love, with which Socrates begins his account of er¯os (200a ff.), is suspect. It calls to mind a wish to dominate, to exploit and manipulate, a lack of respect and regard that reinforces commonplaces about the ‘acquisitiveness’ and ‘egocentricity’ of Greek ethical thought. The desire to possess, one might say more generally, belongs to the consumer, not the lover; it reveals not love but its absence. How can we possibly want to own, and thus be free to use, what we value (as we say) for itself, as an end and not as a means? What would then distinguish us from the perverse character whose anatomy John Fowles presented in The Collector? Possession, though, is not identical with ownership—or, if it is, it is ownership of a different kind: I may possess something as a detachable piece of property, losing which will not affect who I am, or as a genuine part of myself, which I cannot lose without undergoing a serious change of my own.⁶⁷ To possess something as love requires—a person or a work of art I want to treat not merely as a means but also as an end in itself—I must want possession to be a mutual affair: I must want it to be mine as much as I want to become its own as well. To treat something as a means is to take my desires as given and expect it to satisfy them: I do not expect that what I want and value will change as a result of our interaction, except accidentally—certainly not as a result of any desires ⁶⁷ See Thomas Hobbes, Leviathan, book I, ch. 16.

118

Alexander Nehamas

or values it happens to manifest. But when I treat something as an end, I am willing to reconsider my desires and values as a result of taking its own desires and values into account. I treat it with respect. I also make myself vulnerable to it—vulnerable in that I am allowing it to steer me in new directions that I might not even have conceived without it. I engage in a prospective commitment—I make a guess about the future which expresses a desire not only to spend part of my life in the company of that object, but also a need to come to know it better and see how it is likely to affect me (and how, in turn, I can affect it best). For that reason, the pursuit of knowledge is always an element of love and an attendant of beauty. And in that way we have come back to the idea that to find something beautiful, far from reaching a stable conclusion about it, is to have a sense, as vague and inarticulate as it is sweeping and powerful, that there is more to that object (or person) than I can now envisage, and that my life would somehow be more worthwhile if that beautiful thing were to become part of it. But since it is impossible to know in advance what beauty promises to yield, when I act on what is no more than a promise of something valuable but unknown, I am taking a serious risk, for I do not now know how I will change as a result of my action, and whether that change will be for the better or for the worse. And so part of what I undertake when I try to make something mine is to come to know it as well as I can, to understand what it is, and to see how it will affect me and what it will be able to give me. To love something is always, in part, to try to understand what makes it beautiful, what drew me, and, as long as I still love it, continues to draw me toward it. Love cannot be separated from the desire to understand what exactly it is that I love and why I find it beautiful—questions that will not find an answer as long as love survives and that account for the sense that we love only what we do not yet fully possess. For ‘possessing’ something here includes understanding it well enough not to find it any longer surprising. Once that has happened, however, the promise—the beckoning—that is essential to beauty has disappeared, and so has beauty itself. For we then already know what to expect, and what was once beautiful, and an end in its own right, has now been transformed into a mere means—something that can only be owned. The urge to understand is obviously part of our feeling for the works of art that we love. To be overwhelmed by the beauty of In Search of Lost Time, as I am, is not simply to experience certain feelings while reading it. It is also to be willing to devote, literally, part of my life to it—not just to read it (although I will certainly be willing to do that, too), but also to come to know it better, to understand what Proust accomplishes in this work. This has required an effort to learn things I would not have been interested in otherwise—about Proust more generally, about the social, cultural, and political situation in Paris between the end of the nineteenth century and the first years of the twentieth, to improve my French, to understand more about the Dreyfus affair, anti-Semitism, and homosexuality, about the history of the novel in France and elsewhere, to look

Beauty of Body, Nobility of Soul

119

at Vermeer, to listen to Debussy, and much else besides. None of this has been simply a matter of sitting alone in my study. It has involved meeting people I would not have met before, traveling to new places, even changing my everyday habits (it is a very long book), leading my life in directions I could not have even imagined existed had Proust not led me to them. All this is part of my love for the novel, which is inseparable from my effort to understand it and, in fact, to see it as no one else has ever seen it before—for that is what ‘making it mine’ finally amounts to. My love for the novel is necessarily expressed in the urge to interpret it and to keep doing so as long as it attracts me, as long as I still feel that there is more to it than I have seen so far. And as long as I am still trying to interpret it, the more various the things to which I will relate it in order to understand it, to see perhaps how it accomplishes something that nothing else had accomplished before. There is, in other words, no difference between delving more deeply into the novel and wandering more widely in the rest of the world—the more I bring to my understanding of the novel, the more the things in the world to which I relate it, the better I can see how it is different from and how it resembles them and recognize its specific accomplishment, the features that distinguish it from everything else. The better, that is, I come to see how it is in itself, in its own right. To the extent that being involved with it has changed my life, that book has come to possess me; to the extent that I have found something new and unusual in it, I have made it mine; and to that extent I have become new and unusual myself. The same is true of love for persons. When I want to make someone I love mine, I also want her to want to make me her own as well. I am willing to allow her characteristics, many of which I do not yet know, influence who I will be, and I want her to let features of mine help shape her own future. More important, I am willing to let us both affect each other by means of features that do not yet exist but will come into being only as a result of our interaction. How can anyone know where such a process is likely to lead? The kinds of things we love—persons and objects both—and our reasons for loving them and finding them beautiful determine, and express, a large part of our character. To find something beautiful, I have been saying, involves the sense that life would be more worthwhile if that beautiful object were to become part of it. But I have said nothing about what makes life worthwhile, for, unlike Plato, who thinks that this is in all cases moral virtue, I do not think that there is anything both general and informative to say about it. I can only say, begging every question, that in the ideal case the various paths we follow through life on account of things we have loved, and what we have come to understand about them, will gradually transform us too into something no one has seen before and itself worthy of love, attention, and admiration in its own right—in a word, into something beautiful. The possibility that the pursuit of beauty may lead to creating more of it is the deep truth in Plato’s identification of er¯os with the desire to give birth in

120

Alexander Nehamas

beauty. Since he takes beauty to lead inevitably to goodness, Plato is not nearly as troubled by the risks inherent in that pursuit as I am. For myself, though, there is no guarantee that the things I find beautiful will lead me to a good or a successful life. Even if they do, it will always be possible to wonder whether, instead of immersing myself in Plato, Proust, or the people I have loved, I should have worked for Oxfam instead. Yet there, too, the problems would persist: how could we know what that would have led to in the long run—what, for example, if I had ended up embezzling its funds?

VI Whatever the answers to such questions and whatever their long-term effect on our lives, love and beauty are inseparable from the effort to understand what we love or—what is the same—to understand why we love it. As long as love persists, no answer will ever be complete; as long as something still strikes me as beautiful, the sense that there is something about it that is still worth coming to know and celebrate—that there is more to love—remains. That is why judgments of beauty are always at least partly prospective and why the most beautiful things always seem inexhaustible. This forward-looking element in the perception of beauty, the sense that beautiful things are constantly drawing us further, is the deep truth in the lover’s ascent in the Symposium. I have described that movement—the beckoning of beauty, the impetus of love—both as an absorbed immersion in each beautiful object itself and, simultaneously, as an expanding vision of the world to which it belongs. Plato describes it as an upward path. Does that ascent leave the object with which it begins behind? Does beauty, in drawing us further, also draw us away from what sparked its pursuit? The path to the Form, Socrates says, begins with the beauty of a particular boy in whose company a man gives birth to beautiful logoi (210a4–9). These logoi —to repeat: accounts, arguments, pieces of advice, and their results—are to Plato’s scheme what new understandings and interpretations as well as their consequences are to mine, except that Plato is convinced that as a result of their interaction both lovers change for the better. If the lover has a bent for philosophy, there comes a point when he realizes (katanoein, 209a9) ‘that the beauty of any one body is twin to the beauty of any other and that … the beauty of all bodies is one and the same’ (210a8–b3).⁶⁸ How he comes to that realization is not something Plato explains—we can only guess.⁶⁹ One guess, then, is the following. Pressed forward by er¯os and desiring what he does not already possess, the philosophic lover tries to make the boy’s beauty more completely his; that is, ⁶⁸ For the happy translation of ἀδελφόν at 210b1 as ‘twin’, I rely on Rowe 1998. ⁶⁹ See Price 1989: 42.

Beauty of Body, Nobility of Soul

121

he tries to understand why he loves the boy, what makes him beautiful. It is his interest in this question that makes this lover philosophical, for to ask what makes the boy beautiful is to ask what (bodily) beauty is—it is a form of the question that Socrates was constantly putting to his contemporaries.⁷⁰ The answer to that question, for Platonic reasons we may leave aside for now, is that what makes the boy’s body beautiful is what all beautiful bodies have in common.⁷¹ He then surrounds himself with beautiful boys, though he remains as far as I can tell in the company of the first: ‘loosening the bonds of wildly gaping after just one’ (210b5) may be prompted, as I have already tried to suggest, by his realizing that concern for the first boy does not exclude concern for others as well, rather than the view that concern for others excludes concern for the first. The image of an older man surrounded by and always in pursuit of beautiful boys—recall again what we know of Socrates, and his relation to Alcibiades—is much more satisfying than the image of one who abandons a boy so that he can prey freely on others, or abandons boys altogether so that he can contemplate the impersonal feature that their bodies have in common. The process continues. A lover whose passion for philosophy persists will now understand that the beauty of the soul is superior to the beauty of the body (210b–c) and, I believe, will now pay attention to a still larger group of boys—those with beautiful souls—and will even be satisfied to find among them someone whose soul is no more than decent, provided, Socrates says, that he is at least slightly ‘in bloom’.⁷² Plato has no sympathy for the commonplace of the beautiful soul trapped in a wizened body, and he may not even believe that it is ⁷⁰ I don’t think we can simply assume, as we generally do, that the lover described here is a philosophical type simply because Plato describes his ascent to the Form. The question is why some lovers undertake the ascent in the first place, and to answer that, unlike the lovers of the two lower levels, they are pursuing wisdom is simply to use an English instead of a Greek word to restate the original point. Φιλοσοφία, the love of wisdom, is manifested precisely in the fact that some people, unlike others, are intent on asking what is responsible for the things around them being as they are. ⁷¹ In a way, at this stage the lover is like Euthyphro, who claims that piety is avenging religious wrongs (Euthphr. 5d–e), or Laches, who believes that courage is the willingness to stand one’s ground in battle (La. 190e)—their accounts are no better at explaining piety or courage in general than bodily beauty is at explaining what beauty in general is. The difference is that the (successful) lover, as Plato describes him, does not stop with that explanation, but keeps asking similar questions until he reaches the only answer that can answer them all: the Form itself. ⁷² The interpretation of 210b8–c2 is no easy matter. It is often taken to indicate that a beautiful soul is enough to make a boy attractive, provided that he is not actually physically ugly. That is one of the interpretations offered by Rowe (1998: 195), although I think (and I believe Rowe agrees) that assuming that ‘decency of soul does constitute beauty’ is to read ἐπιεικής too generously. The word often denotes no more than a modicum of virtue, and Plato himself seems to think that ἐπιείκεια is not by itself sufficient to guarantee resistance to bad influences, which is why he needs to qualify it by πάνυ at Resp. 538c3 to ensure its effectiveness. I am inclined to think that when Plato writes that the lover is satisfied with a boy whose soul is ἐπιεικής but whose body is in ‘little bloom’ (σμικρὸν ἄνθος), he is not contrasting negligible physical beauty with considerable excellence in the soul. He is, rather, suggesting that the lover, who has realized the importance of the beauty of soul, will be satisfied with a limited quantity of it, likely as it is to be accompanied by a limited quantity of bodily beauty.

122

Alexander Nehamas

possible. Everything in the world, he writes in the Republic (401a1–8), including the human body, displays grace or inelegance, shapeliness or unshapeliness. The latter, he continues, are indications of bad character, the former signs of temperance and virtue. We must not let Alcibiades’ image of Socrates as Silenus—ugly on the outside, full of beauty within (215a5–b6)—mislead us. Like Socrates’ logoi, who seem ridiculous at first but reveal unexpected treasures when one pays enough attention to them (221d7–222a6), his own beauty, too, is evident in his features for those who take him seriously. Despite the ancient commonplace that only the boy, and not his older lover, need be beautiful, it is quite clear—from contemporary pottery, for example—that the men in question were nothing like the satyrs and the Sileni who are made to look ridiculously repulsive. Alcibiades did not and would not offer himself to such a man. Plato still does not say, though, what prompts the lover’s transition to a new focus on the beauty of the soul, and we must guess again. Here, then, is another guess, which I can offer only in the most tentative manner. I argued earlier that beauty of character is manifested in the body, affecting our perception and allowing us to find our lovers not only, say, wise, sensitive, or kind, but also good-looking: the soul really animates the body, the body truly incarnates the soul. Could Plato have held a view of that sort? He seems to accept it at Republic 401a, which we were just discussing, and to give it further expression when he says at 403d2–7: ‘I am not of the opinion that if the body is in a good condition, then this state of physical excellence makes the soul good too. I think it is the other way round: a good soul, by being in a state of excellence, allows a body to maximise its potential for physical goodness.’⁷³ This view makes an appearance in the Charmides, when Socrates says, perhaps in jest but perhaps not, that ‘the soul is the source of everything that is good and bad for the body and for the whole man; these things flow from the soul just as everything that affects the eyes flows from the [whole] head’ (156e6–157a1). I must admit that this is not overwhelming evidence.⁷⁴ But the hypothesis has an immense interpretative advantage: it allows exactly the same reasoning to apply to the transition to the soul that earlier applied to the transition to the body in general, the soul explaining the beauty of the body in general just as the latter had explained the beauty of the particular body that set the process in motion. It is also, if that matters, Plotinus’ view: ‘It is the soul’, he writes, ‘that makes every body that is called ‘‘beautiful’’ what it is.’⁷⁵ ⁷³ Translation by Waterfield (1993), slightly modified. Plato gives a realist version of the thesis I would endorse: he believes that the beauty of a soul will actually make a body in fact attractive and, in principle, an object of love, while I would say that the character traits of the people we love are manifest in their appearance—manifest to those who love them, not necessarily to others as well. ⁷⁴ It may conflict with 210b8–c1 (cf n. 62 above), and it seems to conflict with 209b4–7. ⁷⁵ Plotinus Enn. I. 6. 7.29–30. See the commentary in Kalligas 1994: 272. Plotinus, though, credits the soul with the beauty of ἐπιτηδεύματα as well, while for Plato laws and ἐπιτηδεύματα belong to the next higher step of the ladder of love. I believe that the conflict can be resolved if

Beauty of Body, Nobility of Soul

123

The questions continue and, armed with our hypothesis, we can now answer them readily. What makes beautiful souls beautiful? Being moulded by beautiful laws, customs, and institutions—a beautiful culture—that produce the good and decent behavior of the boys whom the lover is (still) pursuing. What is, in turn, responsible for the beauty of these laws, customs, and institutions? When do they lead unerringly to beautiful souls, to virtue and the good life? When they have been established not haphazardly and as tradition or the powerful prescribe, but as knowledge dictates—the knowledge that Socrates was looking for in Plato’s elenctic dialogues and the grand structure of which is articulated in the Republic. The beauty of knowledge leads to a great expanse, an ‘open sea’ of beauty, and the lover now ceases to care for the beauty that is present in one thing only (210c6–d3). The text may tempt us to think that this turn to the Forms, the objects of knowledge, leads away from all individual attachments, but it also gives us reason to resist the temptation. The lover, Plato writes, is no longer moved by ‘the beauty of a single thing, satisfied like a menial servant (oiket¯es) with the beauty of a boy, a man or a single occupation, contemptible like a slave (douleu¯on) and of no consequence’ (210d1–3). His language recalls precisely Pausanias’ earlier description of the self-abasement of lovers who, imploring a boy’s favors, are eager ‘to provide services (douleiai) even a slave (doulos) would refuse’ (183a), where the motive of the behavior is clearly sexual passion. What is in question, therefore, is the nature of the bond between lover and beloved, rather than the bond itself. The lover abandons sex with boys, not boys themselves. How else would the ‘many gloriously beautiful’ logoi to which the lover gives birth affect the young? He gives birth to them in the presence of the young, content neither to gaze around into the blue nor to isolate himself in his study—the first alternative is as unrealistic as the second is anachronistic. The lover—the philosopher—is actively engaged in a dialectical enterprise to which passionate interaction with others is absolutely crucial. But now he knows that the beauty of boys, their souls, the practices that make them beautiful, even the knowledge on which those practices are based, and their objects is one and the same: they have all become objects of his love; there is no reason to think that the philosopher pays for his expanding vision with a countervailing blindness.⁷⁶ At that point he also realizes that what makes them all beautiful is itself one and the same—the Form of beauty itself, ‘wonderfully beautiful in its nature’ (210e4–5), free of every blemish that has marred the beautiful things he has encountered so far, and, as we now know, on account of which the whole endeavour was undertaken in the first place. we understand Plotinus to take that term to refer to behavior or patterns of behavior, which he considers sensible objects, while in Plato it signifies customs or institutions. ⁷⁶ Contrast Price 1989: 45. Despite several disagreements, I have learned a lot from Price; I doubt, though, that he would consider me a worthy student.

124

Alexander Nehamas

In a very serious sense, as the philosopher gains a vision of the Form of Beauty, he falls in love with the world itself. It is the beauty of that Form, which is reflected in the beauty of everything else in the world, whose contemplation is the only thing that makes human life worth living (211d1–2). But contemplation is far from being detached, abstract, and theoretical. It is part and parcel of the philosopher’s active participation in his circle. It is manifested in (not merely responsible for) his giving birth to the most beautiful logoi possible, to virtue itself, and not merely images of it (211e–212a). His are the best children a human being can have, born in his dialectical interaction with those who spark his vision of the nature of beauty itself and through whom he sees the beauty that is reflected in every other beautiful thing in the world. How could the philosopher engage with them in that way if he did not love them?

VII Aristotle believed that fire is the hottest thing in the world, since it makes everything else in the world hot; similarly, the principles of being, which are the source (aition) of the truth of everything else, are the truest things in the world.⁷⁷ The principle that underlies that view states that the cause is greater than its effect, and it is absolutely central to the metaphysics on which much of Plato’s account of love and beauty depends. What explains the beauty of an object is more beautiful than the object whose beauty it explains, and what explains the beauty of everything is the most beautiful thing of all. Add to that the idea that beauty and goodness are essentially related if not identical to each other, and the success of erotic attraction is guaranteed.⁷⁸ In the presence of the greatest beauty, which is also the greatest goodness, the philosopher has the best and most beautiful life that a human being can have and confers the greatest benefits on his fellows. His ascent to the Form is a continuing effort to move forward, to understand further the beauty of the objects of er¯os, to determine what accounts for it, and it is at every stage inseparable from the generation of beautiful and good logoi, which are the other side of his vision and crucial to its ultimate perfection. Plato was the first to describe the forward movement of love, the urge to learn ever more about the things we love and to understand what makes them beautiful. I have said that it is provoked by the feeling that there is more to the beautiful things we love than we have seen so far, sparked by the sense that ⁷⁷ Arist. Metaph. a1, 993b 23–31. ⁷⁸ A notable exception is the ‘tyrannical er¯os’ that rules the life of the type of man who takes his name from it: the tyrant (Resp. 573d–575a). Plato describes it so negatively that he makes it impossible to use that term or the English ‘love’ to translate it; Waterfield 1993, e.g., translates it as ‘lust’; Reeve 2004 as ‘passion’.

Beauty of Body, Nobility of Soul

125

we can come to know them better and kept alive by the feeling that our future together is still open. For me, though, beauty bears no essential connection to virtue. Although I expect that beautiful things will make my life in some way better, I have no way of knowing what that way will be or whether I am right. That can be determined only in the course of time—if my interaction with something and our effect on each other prove to have been worthwhile. But what is and what is not worthwhile, what beneficial or harmful, can be known only in retrospect, which is often too late. It can also provoke intractable disagreement: you think that the woman with whom I have spent my life has sucked out all that was once good in me, that my best friend has debauched me, that watching television has corrupted my standard of taste; I don’t agree, and sometimes we may have to leave the matter at that. Has television, I have often asked myself, had a corrupt influence on me? At times I believe it has enabled me to produce some decent philosophy. But you find my argument that your contempt of the medium is a version of, neither better nor worse than, Plato’s rejection of poetry as repulsive as television itself, and my essay on the beauties of St. Elsewhere a disgrace. How are we to decide? How can I decide, for that matter? Before I was first attracted to television, I used to find it despicable; I looked down at those who enjoyed it with a mixture of pity and scorn. Now I have become able to appreciate its virtues—or have I? My feeling is that, other things being equal, I am better off now than I was then. But how can I tell for sure, since along with a taste for television I have also developed standards of judgment that, from the point of view of my own earlier self, are depraved and corrupt? By my earlier standards, I am now depraved, corrupt, and miserable, although I don’t know it. By those I currently accept, my earlier standards were silly, prejudiced, and deprived me of discerning a kind of beauty that has served me well in the meantime. Which standards are right? Plato, for whom no disagreement is ultimately intractable, answers: the standards of philosophy—the only standards that can establish whether a life has been worthwhile. And what those standards say is that life is worth living only in the contemplation of beauty, in giving birth to kaloi logoi, beautiful ‘accounts’ or actions, that promote virtue and goodness. Why actions? Because, as we have seen, the philosopher brings forth not ‘images of virtue’ but genuine virtue instead (212a2–5). But that is true of every stage of the ascent, and even of the lesser mysteries: virtue is exactly what legislators like Lycurgus and Solon—the greatest initiates of these mysteries—have given birth to (209e1–2). There is no real distinction here. The beauty of each logos depends essentially on its incorporation into the lives of those who engage with it: there is no other standard. The contemplation of beauty, which is at no stage a passive affair, is manifested in the creation of something beautiful—a logos and one or more persons who make that logos part of their life. This is the Platonic core of my idea that the pursuit of beauty leads us to see what we love in new ways that are distinctly our own. For if we succeed, we will reveal beauties that no one has seen

126

Alexander Nehamas

before within them and make them more beautiful than they were before; and as we make things beautiful, they give us beauty in return. In that way, we may perhaps one day become beautiful in our own right, objects that others may love and want to come to know in ways of their own. Like Plato, I am convinced that beauty is a spur to creation: it gives rise to new beauty. Unlike him, though, I do not believe that it has to direct us to virtue, and for that reason I think that it contains an ineliminable risk. It is not an infallible guide; it is ‘only a promise of happiness’.⁷⁹ It is also a promise that is not always fulfilled, and sometimes it may be worse if it is, because by then I may have become incapable of seeing the harm—aesthetic or ethical—it has done me. But sometimes it is, and a new beauty, a new spur to creation, enters the world. It is then that one can say, and say truly, that, quite apart from its moral worth, one’s life was a life worth living, that one is happy to have become who one is.

VIII Beauty is a matter of distinction, and beautiful things stand out among things of their kind. They push those things into the background and reveal new styles in which things and their constituents can be put together. But achieving distinction is never a matter of duty, and for that reason beauty—much more so than duty—is something for which one can be praised and, moreover, something of which one can be proud. In pride, I now want to suggest, we may finally have found the beginnings of a path to what we were looking for in the opening part of this essay: the element that is common to and connects the various senses of kalon with each other. Let us first think of kalon as applying to anything that one can be proud of either having or being. And let us consider that pride is directly correlated with admiration and praise (epainos)—what I am proud of is what you admire and praise me for—and inversely correlated with shame—the feeling I have when I have fallen short of what I had hoped for myself, what I had thought myself capable of.⁸⁰ The fact that aischron, which is the most common term for both ‘ugly’ and ‘shameful’, is also the most common contrary of kalon cements this last connection. It is, though, an obvious commonplace that not every occasion of pride will do. Since people can be proud or ashamed on the most inappropriate grounds, we need to determine when pride is rightful and shame justified. That, in turn, depends on the way in which the combination of features that gives rise to such ⁷⁹ ‘La beauté n’est jamais, ce me semble, qu’une promesse de bonheur’ (Stendhal 1919: 45–6); ‘La beauté n’est jamais que la promesse du bonheur’ (Stendhal 1926: 74 n. 1). ⁸⁰ The views of Williams (1993: ch. 4) are crucial to any examination of this topic.

Beauty of Body, Nobility of Soul

127

reactions is or is not itself appropriate, on whether things have been put together well. Whether it is the features of a living organism (including the elements of the human body), the parts of an artifact, or the actions of a person, it is only if these are fitting to one another, to the purpose they serve, and to the context in which they are manifested that pride in them is justified. Appropriateness—the second sense of kalon according to Liddell and Scott—provides the connection that ties its three senses together and explains its association with pride. Beautiful things accomplish their ends in the most fitting way: they are the best of their kind.⁸¹ Is it possible, though, to take pride in features that are, so to speak, god-given and require no effort at all, no intentional behavior of any sort on our part? That also seems inappropriate: the gifts of luck can produce all sorts of good feelings in us, but pride does not properly belong among them. We take pride in our accomplishments. And although every accomplishment may also depend on luck, opportunity, or natural talent, it must also be in serious measure the result of our own effort and activity. Pride in what is not one’s own accomplishment is hollow and vain.⁸² If so, however, how could one ever be proud of being handsome or attractive—the first and most common sense of kalon—which hardly qualifies as an accomplishment? And if one couldn’t, what is there to tie the various uses of kalon together? The word, once again, seems to be simply ambiguous. Yet it is a fact that the Greeks, like people everywhere and at all times, were willing to consider pride in beauty at least perfectly common, if not justified—Paris’ awarding the apple to Aphrodite rather than Hera or Athena would not have caused the Trojan War if it were otherwise. That might reestablish a sort of unity to kalon, but only at the cost of attributing to them a failure to understand one of the most common phenomena in their experience and their linguistic practice. Instead of thinking along these lines, however, we should recall that the evidence of the pottery and the graffiti discussed earlier suggests that beauty was commonly taken to be connected to one’s status and descent. Belonging to the right family and social stratum was crucially important and a sure mark of distinction to archaic Greek society, which, like all societies predicated on a clear hierarchical class structure, derived its standards of beauty from the features of its ruling class. So, for example, Achilles finds it quite natural to speak of his physical features and his noble lineage together when he rejects Lycaon’s plea to spare his life: ⁸¹ Admiration and pride, we should note, may diverge. Even those who are justifiably proud of their accomplishments may be praised for the wrong reasons and by the wrong people: to be admired by people one holds in contempt is not an honor. It is a major concern of Plato’s in the Republic that those who deserve admiration (the philosophers) are admired for the right reasons, by the right audience. See Nehamas 1999: 76–91. ⁸² Of course, that doesn’t prevent us from taking pride in the achievements of others—not only in the sense of being proud of them, but also if we bear certain relationships to them—they may be our students, our children, even our friends—on our own behalf.

128

Alexander Nehamas Come, friend, you too must die. Why moan about it so? Even Patroclus died, a far, far better man than you. And look, you see how handsome (kalos) and powerful (megas) I am? The son of a great (agathos) man, the mother who gave me life a deathless goddess.⁸³

At the other end of the same scale, the proverbial ugliness of Thersites—the only ‘commoner’ to speak in the Iliad —is certainly not unrelated to his social station (II. 212–19). Although even here noble lineage and handsome physique may sometimes seem no more than matters of luck or accident, those possessed of such gifts were supposed to behave in certain appropriate ways. Looks indicated status, and with status came expectations. Many of the connections involved are beautifully illustrated, once again, by Theognis: Although you are kalos, young man, your friends are bad (kakoi); on their account you find yourself surrounded by wretches (deiloi) and that has brought shameful (aischron) disgrace upon you.⁸⁴

The pride that goes with being kalos depends in great part on being prepared to act as a kalos should; failure to do so produces shame. Intention is part of this picture. Despite his deep disapproval of the old aristocratic system’s particular expectations, Plato, too, expects much of the kalos. While the old system derived its standards from social practice and assumed that different social classes will necessarily behave differently, Plato sees in beauty the capacity to behave according to rational standards of appropriateness which he believes apply in general, independently of a person’s social class and (in the Republic) even of gender.⁸⁵ He rejects the idea that the aim of education is to raise individuals to act as their social status dictates, and insists that they should be prepared to act as rationally as each is able to, and for that reason subordinates both to kalon and to agathon to reason and wisdom. Not for Plato Pericles’ contrast, however mild it may have ⁸³ Hom. Il. XXI. 106–9: ἀλλὰ φίλος θάνε καὶ σύ· τί ἦ ὀλοφύρεαι οὕτως; κάτθανε καὶ Πάτροκλος, ὅ περ σέο πολλὸν ἀμείνων. οὐχ ὁράᾳς οἷος καὶ ἐγὼ καλός τε μέγας τε; πατρὸς δ εἴμ’ ἀγαθοῖο, θεὰ δέ με γείνατο μήτηρ. ( Translation by Robert Fagles.) ⁸⁴ Thgn. El. 2. 1377–9: καλὸς ἐὼν κακότητι φίλων δειλοῖσιν ὁμιλεῖς ἀνδράσι, καὶ διὰ τοῦτο αἰσχρὸν ὄνειδος ἔχεις, ὦ παῖ. ⁸⁵ In the case of artifacts, the expectation is that a beautiful object will function well according to standards determined by those who know how to use it. That the quality of an artifact is determined by its users is stated at Resp. 601b–602a. That those who know how to use resources like wealth and other things ‘said to be’ good well are virtuous is implied at Resp. 495a4–8. That wisdom is required to make good use of resources is stated explicitly, among other places, in the Euthydemus (278e–281e). See Lear 2006, an important essay which arrived too late for me to take fully into account.

Beauty of Body, Nobility of Soul

129

been: ‘We love beauty (philokaloumen) without being extravagant and we love wisdom (philosophoumen) without becoming weakly.’⁸⁶ For Plato, beauty, virtue, and philosophy have become aspects of one and the same kind of character and mode of life: the person who is to become truly good (kalos te kagathos) must above all be ‘led by truth: if he does not pursue truth absolutely and wholeheartedly, he is bound to be a specious impostor, with nothing whatsoever to do with philosophy’.⁸⁷ Although at one point in the Republic Plato concedes that it is possible for the quality of body and soul to diverge, he does so in a strangely tentative and offhand manner (402d10–e4).⁸⁸ His considered position, for which he argues forcefully and explicitly, is that the most beautiful sight is of a body whose features are in harmony with the beautiful character of its soul (402d1–4). As a matter of fact, a frightening passage that follows shortly suggests that no one in the kallipolis (which is, incidentally, more than a merely pretty city) will possess either an unseemly body or a bad soul: law and medicine will make sure of that, caring for ‘the body and the soul of those of your citizens who are naturally well endowed in these respects—as for the rest, those with a poor physical constitution will be allowed to die on their own, and those with irredeemably rotten souls you will yourselves put to death’ (409e5–410a4).⁸⁹ Bodily beauty is valuable to Plato mainly to the extent that it is a sign of, and conducive to, a beautiful character that is defined without reference to social status:⁹⁰ we are witnessing here the transformation of the concept of social nobility into a concept denoting nobility of soul which Nietzsche describes as a crucial step in the emergence of the modern notion of moral virtue in the opening sections of On the Genealogy of Morality (Nietzsche 1968a). The idea of fit underlies the unity of kalon for both Plato and the aristocratic tradition he wants to replace: their difference is over the standards that determine ⁸⁶ Thuc., Hist., 2. 40. 1. ⁸⁷ Resp. 489e4–490a3, trans. Waterfield (1993), slightly modified. In the Phaedrus (249d3–5), Plato writes that the best human life is that of a philosopher, a lover of beauty (φιλόκαλος) or a man of culture and erotic desire (μουσικός καὶ ἐρωτικός). What exactly he means by this is a matter of considerable controversy. In Nehamas and Woodruff 1995: 35, we suggest that Plato’s distinction here is exclusive, specifying three different modes of life—a view of which my present interpretation of the Symposium is causing me to be less confident. ⁸⁸ When Glaucon, at d10–e1, broaches the possibility of an imperfect body with a good soul, Socrates grants him the point (συγχωρεῖν) because, he says, Glaucon ‘is or was’ in love with such a person, and immediately changes the subject. ⁸⁹ Translation adapted from Waterfield 1993. ⁹⁰ Plato, of course, must provide a revised conception of beauty along with his revised conception of virtue—otherwise, among other things, he would not be able to account for the nature of Socrates, who was, by traditional standards, proverbially ugly. He accomplishes that by thinking that discerned beauty of soul manifests itself in the body and constitutes a promise of more to come. Unfortunately, Alcibiades’ famous image of Socrates as Silenus, ugly outside but beautiful inside, in the Symposium (215a4–d1) has suggested just the opposite to most of Plato’s readers, but that is mostly due to forgetting that it is an image and taking it as a literal statement instead. In Nehamas 2007, I argue that Plato’s view is actually the reverse of Nietzsche’s charge that Socrates was ‘monstrous in appearance, monstrous in soul’ (1968b: 475): for Plato, he was, rather, wondrous in soul, wondrous in appearance.

130

Alexander Nehamas

correct fit and justify pride in possessing or being something kalon. For both, also, what seems beautiful can diverge from what is in fact beautiful. Right judgment in these matters requires some sort of knowledge, which Theognis thinks the right social background, upbringing, and long-term intercourse provide, while Plato aims to secure it through the right education of the soul and the body, the early stages of which the Republic claims to make available to everyone—or at least all those who may one day become the city’s guardians. The soul’s education, we know, begins (as it always has, Plato writes at Republic 376e2–4) with poetry which, suitably composed to represent and foster good character, is itself beautiful, and it is reinforced by a beautiful environment—pictures, artifacts, buildings, bodies, and everything else (401a1–4)—which has the same effect.⁹¹ The result of that education is, as I just said, a sort of knowledge, acquired before the onset of reason and its deeper understanding of beauty, and ready to welcome both out of what Plato, tongue in cheek, describes as the ‘philosophical’ love of learning that the young guardians share with dogs and causes them both to like what they are familiar with and despise what they do not already know (375e9–376a5). It is reasonably clear that Plato’s elementary education aims to acculturate the ‘spirited’ (thumoeides) part of the soul to taking pleasure in things whose beauty reason will eventually account for by means of establishing their goodness. Its primary goal, after all, is to make the young guardians themselves thumoeideis and prepare them to acknowledge (aspazesthai, 402a3) the superiority of reason, just as the second part of the soul is eventually said to be reason’s ally (summachos t¯oi log¯oi, 440b3).⁹² But what is very surprising is that Plato, who says that the thumoeides loves both victory and honour (philonikon, philotimon) and craves power, success, and fame (581a9–b3), never describes it as philokalon as well.⁹³ The reason, I believe, reveals another major difference between Plato and the tradition that precedes him. Victory, honour, power, success, and fame are all values firmly associated with aristocratic society and politics. Plato preserves them, but by attributing them to what we may well call an auxiliary part of the soul, makes them subordinate to the intellectual and moral virtues he takes to be paramount. The problem, though, is that while most people seem to be naturally motivated to pursue victory and honour, very few—if any—among them, as Plato had to acknowledge in the Gorgias, can see the attraction of wisdom or justice.⁹⁴ In beauty, he found a value whose desirability was not simply a logical fact about it but its most palpable feature: to find something ⁹¹ See Burnyeat 1999: 217–22. ⁹² Convincing evidence for this view is given by Lear (2006: 116–17). ⁹³ The term is in general rather rare. Aristotle uses it at Eth. Nic. I 8, 1099a 13; IV 4, 1125b 12; X 9, 1179b 9, and Hist. an. I 1, 488b 24. In Plato, it appears only twice, at Phdr. 248d3 (see above, n. 90) and Criti. 111e3. ⁹⁴ See Moss 2005.

Beauty of Body, Nobility of Soul

131

beautiful is inseparable from desiring it, while, as he admits in his discussion of akrasia in book IV of the Republic, we may fail to follow even what we know to be the wise or just course of action. And although the effort to show that virtue or goodness is also good for us, and we therefore have a motive for pursuing it, requires an argument as complex and controversial as the whole of the Republic, which, even if sound, may fail to instill such a motive within us, the desire for beauty enters life before reason and full understanding find their way, if they ever do, into the soul. But if wisdom and justice, virtue and goodness, were related to beauty in such a way that reaching for beauty was a way of reaching for these virtues as well, then all human beings, since all are lovers of beauty, would be in principle capable of feeling the attraction of virtue, whether or not they in fact do so. All lovers, as Socrates argues in the Symposium, begin at the same place, with the beauty of an individual; but where that leads them is in all cases a different and, in some, a much longer story. Part of that story, which is told in the Symposium, is that beautiful things awaken a desire for immortality, remembrance, and happiness (208e4; cf. 204d–205a). Beautiful things spark an urge to become like them and so, like them, to leave our own mark on the world. Some people, we have seen, try to do so through their children, others through poetry or law, still others through philosophy. The first group largely fails: their appetite leads them to ‘think’ (hegountai, 208e4) that children will secure what they desire, but no ‘temples’ (hiera, 209e4) have been erected for them. The second—Solon and Lycurgus, Homer and Hesiod: the legal and poetic authorities behind the aristocratic tradition and values that Plato assigns to the ‘honour-loving’ part of the soul—succeed; Greeks and barbarians both are still paying tribute to their accomplishments (209d6–e4). It is the third group, though, which are by far the most successful: the philosophers, unlike their older competitors, create true beauty and virtue and are loved by the everlasting gods themselves; they come as close to real immortality and happiness as human beings possibly can (211d1–3, 212a3–7). The reason is that er¯os is in their case a rational desire, which includes the urge to understand what makes beautiful things beautiful and guards against the attraction of what merely seems beautiful but in reality is not. Only the really beautiful, in Plato’s view, fulfills the promise that every beautiful thing, whether it really is so or not, makes to those whom it draws toward it. But since everyone is drawn toward some beauty or other, everyone already possesses a motive for the pursuit of virtue and happiness. The question is whether one’s soul can discern real beauty, and so genuine virtue and happiness, neither illusory nor subject to the inconstant and unreliable opinions of moral human beings, an accomplishment of which one can be perfectly proud, whether the world at large is willing to praise it or not. Beauty may in the first instance affect the thumoeides, as the Republic indicates, or the appetitive part of the soul, as we learn in the Symposium. But it is in the end and in its true nature a proper object of the desires of reason—and

132

Alexander Nehamas

that is why the Republic does not characterize the love of honor also as love of beauty.⁹⁵ Genuine beauty belongs to reason because reason, unlike the other parts of the soul, asks what makes things beautiful, and the answer that satisfies it distinguishes between things that may only seem beautiful and things that really are. When the philosophic lover, for example, establishes that what makes bodies beautiful is in all cases one and the same (Symp. 210a8–b2), he learns what to look for and will no longer find beautiful a body that lacks the feature in question or ugly a body that has it. When Laches, though, proposes standing one’s ground as one among several explanations of courage (La. 190d7–191c5), he can no longer tell between the genuine courage of the Spartans at Thermopylai and, had he but known of it, the foolishness that brought Pyrrhus both victory and destruction. What stands behind the contrast between the genuine and the apparent is always, for Plato, the good—in particular, the good of the human soul. The further the philosophic lover rises on the Symposium’s ladder of love, the more beautiful and beneficial to the soul the logoi he produces and the more beautiful and virtuous he himself becomes until, at the peak of his ascent, he reaches perfect beauty and so true virtue as well and not their ‘images’ (212a4), ‘the many things that each appears to be, despite its being one, because it turns up all over the place, combined with actions and bodies and all the others’ (Resp. 476a5–7). One may need nothing more than knowledge of the flute in order to determine whether a particular instrument is beautiful or artless (Resp. 601d4–602a10), but that is so only on the assumption that flute playing is itself a worthwhile practice. And that is for the wise and virtuous, for the philosopher, who knows which practices contribute to the human good, to determine according to the standards established, Plato thinks, not by society or tradition but by the nature of the human soul.⁹⁶ ⁹⁵ Here I disagree with Lear’s view that τὸ καλόν cannot be the proper object of reason. She points out (Lear 2006: 117) that Resp. 505d distinguishes τὸ καλόν from the good on the grounds that although no one desires to have what merely appears good, it is perfectly possible to be satisfied with what merely seems beautiful. That is true, but not, I think, telling. The same point holds for justice (505d6, 506a4), but that does not seem to me to show that the desire to be just is not a rational one (what else could it be?). She also argues that τὸ καλόν is the appearance of goodness, and that although ‘the beauty [of things] reveals their value … reason doesn’t pursue beautiful things because they manifest goodness, but because they are good. The love of beauty per se would, on the other hand, love beautiful things not because they are good, but primarily because they appear good.’ ‘To manifest,’ though, is ambiguous between ‘to seem but not to be’ and ‘to reveal’. And the love of beauty is a desire of reason precisely when it includes the effort to determine whether its object is genuinely beautiful and what, in that case, is the goodness that makes it so. ⁹⁶ The nature of the human soul, in turn, is a reflection of the nature of the world itself, whose goodness and beauty it also reflects to the extent that it is virtuous. Had this essay been more centrally concerned with the Republic, it would have connected the idea of fittingness explicitly with the idea of unity, and thus articulated Plato’s reasons for thinking that beauty and goodness can’t diverge—and would have done so to the extent that it reflected the virtues of the study of mathematics and the soul in Burnyeat 2000.

Beauty of Body, Nobility of Soul

133

Plato’s conception of kalon consists of many strands, all of them contestable and deeply controversial. One is an effort to transform the social fully into the psychological, and preeminence completely into virtue—is that really possible? is reason quite as autonomous as he imagined?⁹⁷ Another is the view that beauty is a real feature of things to which, to the extent that it is genuine, love is the proper response—but beauty seems to me no less love’s creature than its cause. There is his confidence that beauty is genuine only when it bears an extraordinarily close connection to the good, when its attraction is in effect the attraction of virtue itself—and yet it is the promise and not the delivery of perfection and happiness, which are in any case vastly more variable and multiform, that is the mark of beauty. Even those features of his vision which strike me as the most profound—the promise of perfection and happiness issued by every beautiful thing, drawing its lovers ever forward; the desire to create beauty and become oneself beautiful that is the other side of the urge to possess it—are difficult to reconcile with the sense, most forcefully expressed by Schopenhauer, that beauty brings a sense of satisfaction and completeness and leads not to continued effort but to a restful contemplation instead. In beauty, Plato saw not just a promise of happiness but a pledge of virtue as well. The pursuit of beauty brings both the philosopher and his companions closer to human perfection, a perfection determined purely in terms of the soul’s inherent nature and not the accretions it has acquired as a result of its embodied state and its historical place. The ascent thus accomplishes something inherently good—an achievement and a distinction of which one can be rightly proud, whether or not anyone else knows of it or not. The philosopher’s pride has nothing to do with reputation or fame. That stunning vision, though, has reached us because, ever since it was composed, the Symposium, along with Plato’s other works, has been drawing swarms of interpreters, eager to put their own stamp upon it. Was that among Plato’s goals? Who knows? The fact is that Socrates, whose logoi it describes, and Plato, whose logos it is, are no less famous than Homer or Hesiod, Lycurgus or Solon (209d–e), in large part because the Symposium has proved inexhaustible. The life that Plato honors may not be the only good human life, and pride may not be totally independent of praise, which is why Homer, for one, still stands by his side. But if to be beautiful is to provoke the creation of beauty, the proliferation of beautiful logoi, little can be compared to the Symposium. Like a philosopher’s work, it is a ‘gloriously beautiful’ logos (kalos kai megaloprep¯es) and, like a poet’s, ‘a child which, being immortal itself, has provided its father with immortal remembrance and glory’ (Symp. 209d3–4). Whether or not it has ever led its readers to virtue, the Symposium has proved to be an offspring of which Plato—in every sense of these words—can rightly be proud. ⁹⁷ For a sketch of a negative answer, see Williams 1993: 99–101.

134

Alexander Nehamas

REFERENCES Beere, J. (2003), ‘Psychology, Virtue, and Immortality in Plato’s Phaedo’. Princeton. Brentlinger, J. T. (1970), ‘The Nature of Love’, in ‘The Symposium’ of Plato (Amherst, Mass.), 113–29. Broadie, S., and Rowe, C. (2002), Aristotle: Nicomachean Ethics, Translation, Introduction and Commentary. Oxford. Burnyeat, M. F. (1977), ‘Socratic Midwifery, Platonic Inspiration’, Bulletin of the Institute of Classical Studies, 24: 7–16. (1999), ‘Culture and Society in Plato’s Republic’, The Tanner Lectures on Human Values 20: 215–324. (2000), ‘Plato on Why Mathematics is Good for the Soul’, in T. Smiley (ed.), Mathematics and Necessity: Essays in the History of Philosophy, Proceedings of the British Academy, (Oxford), 103: 1–81. (2003), ‘Apology 30b 2–4: Socrates, Money, and the Grammar of γίγνεσθαι’, Journal of Hellenic Studies, 123: 1–25. Cohen, D. (1991), Law, Sexuality, and Society: The Enforcement of Morals in Classical Athens. Cambridge. Cunningham, M. R., Roberts, A. R., et al. (1995), ‘Their Ideas of Beauty Are, on the Whole, the Same as Ours: Consistency and Variability in the Cross-Cultural Perception of Female Physical Attractiveness’, Journal of Personality and Social Psychology, 68/2: 261–79. Dover, K. (1974a), Greek Homosexuality. New York. (1974b), Greek Popular Morality in the Time of Plato and Aristotle. Berkeley. (1980), Plato: Symposium. Cambridge. Fagles, R. (1990), Homer: The Iliad. New York. Flaubert, G. (2004), Madame Bovary. Oxford. Frankfurt, H. G. (2004), The Reasons of Love. Princeton. Hobbes, Thomas (1962), Leviathan, or the Matter, Form and Power of a Commonwealth Ecclesiastical and Civil, ed. Michael Oakeshott (New York: Collier Books, 1962). Jones, D. (1995), ‘Sexual Selection, Physical Attractiveness and Facial Neoteny: Crosscultural Evidence and Implications’, Current Anthropology, 36/5: 723–48. and Hill, K. (1993), ‘Criteria of Facial Attractiveness in Five Populations’, Human Nature, 4/3: 271–96. Kalligas, P. (1994), Plotinou Enneas Prote. Athens. Kang, S.-H. (2006), ‘Continuity and Discontinuity between ‘Protagoras’ and ‘Republic’: A Gradual Development’. ( Thesis, Princeton). Keller, S. (2000), ‘How Do I Love Thee? Let Me Count the Properties’, American Philosophical Quarterly, 37: 163–72. Kolodny, N. (2003), ‘Love as Valuing a Relationship’, Philosophical Review, 112: 135–89. Langlois, J. H., Ritter, J. M., et al. (1991), ‘Facial Diversity and Infant Preferences for Attractive Faces’, Developmental Psychology, 27/1: 79–84. (1994), ‘What Is Average and What Is Not Average About Attractive Faces?’, Psychological Science, 5/4: 214–20. Langlois, J. H., and Roggman, L. A. (1990), ‘Attractive Faces Are Only Average’, Psychological Science, 1/2: 115–21.

Beauty of Body, Nobility of Soul

135

et al. (1987a), ‘Infant Preferences for Attractive Faces: Rudiments of a Stereotype?’, Developmental Psychology, 23/3: 363–9. (1987b), ‘Infants’ Differential Social Responses to Attractive and Unattractive Faces’, Developmental Psychology, 26/1: 153–9. (1995), ‘Infant Attractiveness Predicts Maternal Behaviors and Attitudes’, Developmental Psychology, 31/3: 464–72. Lear, G. R. (2006), ‘Plato on Learning to Love Beauty’, in G. Santas (ed.), The Blackwell Companion to Plato’s ‘Republic’ (Malden, Mass.), 104–24. Liddell, H. G., and Scott, R. (1968), A Greek–English Lexicon, rev. edn. Oxford. Lissarague, F. (1999), ‘Publicity and Performance: Kalos Inscriptions in Attic VasePainting’, in S. Goldhill and R. Osborne (eds.), Performance Culture and Athenian Democracy (Cambridge), 359–73. Little, A., and Perrett, D. (2002), ‘Putting Beauty Back in the Eye of the Beholder’, The Psychologist, 15/1: 28–32. Moss, J. (2005), ‘Shame, Pleasure, and the Divided Soul’, Oxford Studies in Ancient Philosophy, 29: 137–70. Nehamas, A. (1999), The Art of Living: Socratic Reflections from Plato to Foucault. Berkeley. (2007), Only a Promise of Happiness: The Place of Beauty in a World of Art. Princeton. and Woodruff, P. (1989), Plato: Symposium. Indianapolis. (1995), Plato: ‘Phaedrus’. Indianapolis. Nietzsche, F. (1968a), On the Genealogy of Morals. New York. (1968b), Twilight of the Idols, in W. Kaufmann (ed.), The Portable Nietzsche. New York. Nusssbaum, M. C. (1986), The Fragility of Goodness: Luck and Ethics in Greek Tragedy and Philosophy. Cambridge. Perrett, D. I., Lee, K. J., et al. (1998), ‘Effects of Sexual Bimorphism on Sexual Attractiveness’, Nature, 394: 884–7. May, K. A., et al. (1994), ‘Facial Shape and Judgements of Female Attractiveness’, Nature, 368: 239–42. Price, A. W. (1989), Love and Friendship in Plato and Aristotle. Oxford. Reeve, C. D. C. (2004), Plato: ‘Republic’. Indianapolis. Rowe, C. J. (1998), Plato: ‘Symposium’. Warminster. Stendhal, H. M. B. (1919), Rome, Naples et Florence. Paris. (1926), De l’amour. Paris. Velleman, D. J. (1999), ‘Love as a Moral Emotion’, Ethics, 109: 328–74. Vlastos, G. (1981), ‘The Individual as an Object of Love in Plato’, in Platonic Studies (Princeton), 3–42. Waterfield, R. (1993), Plato: ‘Republic’. Oxford. Williams, B. (1993), Shame and Necessity. Berkeley. Woodruff, P. (1982), Plato: Hippias Major. Indianapolis. Woods, M. (1982), Aristotle: Eudemian Ethics, Books I, II and VIII. Oxford.

7 Er¯os, Philosophy, and Tyranny Dominic Scott

I N T RO D U C T I O N Standing at opposite ends of the moral spectrum, the philosopher and the tyrant of the Republic seem worlds apart from one another—the one just, ordered, and harmonious, the other lawless, bestial, and wild. And yet they have one thing in common: both are gripped by an obsessive er¯os. At various points in book VI, philosophers are described as lovers, whether of learning, truth, or philosophy itself (485b1, 490b1–7, 499c1–2, and 501d2). According to the channel analogy of 485a–487a, their philosophical er¯os is like a flow of water directed into a single stream, drying up their other desires for more worldly goods. But the channel argument also applies very well to the tyrant, described not just as lawless, but also as having an er¯os that he pursues with complete single-mindedness, an er¯os that informs all aspects of his character, dominating his beliefs, desires, and actions. Many readers may question whether there really is an interesting similarity here. As if to resist the comparison, some translators use different expressions for the er¯os of the tyrant and that of the philosopher. Though happy to talk of the ‘love’ that the philosopher has, they prefer such expressions as ‘master passion’ or ‘lust’ when they come to translate the tyrant’s er¯os at 572e5, even though Plato himself uses one and the same word.¹ Another way of challenging the significance of the I am grateful to audiences at Brown, Cambridge, Charlottesville, Edinburgh and Toronto for helpful and incisive comments on earlier drafts of this chapter, particularly Gabriel Richardson (my commentator at Toronto), Charles Brittain, Eric Brown, Thomas Johansen, Geoffrey Lloyd, Christopher Rowe, Theodore Scaltsas, Malcolm Schofield, Frisbee Sheffield, Richard Stalley, Gisela Striker, and Terry Penner. I owe particular thanks to Myles Burnyeat for his inspirational approach to the Republic over the years. ¹ For ‘master passion’, see Jowett 1953; Cornford 1941; Lee 1974; ‘ruling passion’ and ‘spirit of desire’, Shorey 1930; ‘overpowering passion’, Richards 1966; ‘lust or passion’, Ferrari and Griffith 2000; ‘lust’, Waterfield 1993 and Grube 1971, but emended by Reeve to ‘erotic love’ in Cooper 1997. Among the commentators, Nettleship (1901: 314) translates the tyrant’s er¯os as ‘bestial

Er¯os, Philosophy, and Tyranny

137

comparison is to claim that all the principal characters of the Republic are driven by er¯os for a specific object: the timocrat for honour, the oligarch for money, and the democrat for freedom. If so, there would be no special link to be made between philosopher and tyrant. But the work does not in fact analyse the timocrat, oligarch, or democrat in terms of er¯os, and their desires are more dilute than those of either the philosopher or the tyrant. The democrat flits between all sorts of activities in a somewhat nonchalant way (561c6–d7); indeed, his character is based on a compromise between frugality and lawlessness (572d1–3).² Something analogous applies to the timocrat, who settles on a life of honour as a compromise between the lives of reason and appetite (550a4–b7). Nowhere is he actually said to be driven by er¯os for honour; he does not pursue it obsessively or even very single-mindedly—a point underscored by the fact that his appetitive desires grow in strength as he grows older (549a9–b4). Nor is the oligarch actually said to have er¯os for money. He is in fact a deeply conflicted character, who experiences the pang of unnecessary appetites but forcibly restrains them (554b7–e1). The philosopher and the tyrant stand out for their complete single-mindedness. But how deep does the comparison go? To answer this question, I shall start with the description of the tyrant in book IX, focusing upon the exact reason why Plato characterizes him in terms of er¯os. I then turn to the philosopher and, after a brief excursus into the Phaedrus, examine passages from the central books of the Republic that help to reveal the strand linking philosophical and tyrannical er¯os.

I. THE TYRANT The passage on tyranny follows the pattern underlying the earlier descriptions of timocracy, oligarchy, and democracy in book VIII, 545c–562a. In each case Socrates begins with the state before moving to the individual; he also presents his classification as a narrative of degeneration: timocracy evolves from the just state, oligarchy from timocracy, and democracy from oligarchy. In the case of the individual, we are to imagine a philosopher’s son succumbing to bad influences and becoming a timocrat. In turn, his son turns into an oligarch, whose son goes on to become a democrat. This means that throughout book VIII both political and psychological accounts are divided into two parts, one describing the evolution of the particular form, the other describing its nature once developed. passion’, Adam (1963: ii. 322) as ‘master passion’ and ‘lust’, and Annas (1981: 302–4) as ‘lust’. Bloom (1968: 424) translates er¯os at 572a5 as ‘love’ and takes the comparison between the tyrant and the philosopher seriously, although his remarks on the subject are brief and somewhat incautious. See n. 25 below. ² It is true that the oligarchic and democratic states are said to have an unquenchable desire (apl¯estia) for wealth and freedom respectively (562b6–c5), and at 590b6–9 the appetitive individual has an apl¯estia for money. But having a desire that never reaches satisfaction is not the same as having an erotic obsession.

138

Dominic Scott

The same applies to the analysis of tyranny. Socrates describes the emergence of the tyrannized state in VIII, 562a–566d, before going on to describe what it is like at 566d–569c. Then in book IX, he turns to the emergence of the tyrannical individual (571a–573c) and describes the quality of his life at 573c–576b. The account of the tyrannized soul develops the theory of the divided soul originally set out in book IV and used throughout book VIII. While the philosopher (i.e. the just person) is ruled by the rational part, the timocrat is ruled by spirit. The oligarch and democrat both appear to be appetitive, but Socrates introduces a subdivision of this part to account for the difference between them (554a and 558d–559d). Some appetites are necessary, others unnecessary. Necessary appetites lead to good effects (e.g. health) and cannot be removed by training; unnecessary desires can be eliminated and have harmful consequences for both body and soul. The oligarch experiences unnecessary appetites, but restrains them in the interests of preserving his wealth. In contrast, the democrat gives free rein to appetites of both kinds. When it comes to the tyrant, Socrates introduces yet another division among the appetites: of those that are unnecessary, some are lawless, others are not (571a5–572b8). The democrat may experience lawless desires (more than the timocrat or oligarch, perhaps) but restrains them, despite his overt policy of total freedom and equality of desires (572b10–d4). The tyrant differs from him by pursuing such desires without any inhibition or shame. So far, we have heard nothing about er¯os in the account of the tyrant. This emerges when Socrates comes to focus on the way in which his character emerges from his democratic upbringing (572e4–573c10). We may be expecting Socrates to apply exactly the same pattern of analysis as before and to say that, just as the democrat emerged by rebelling against the restraint of his father, the tyrant does so by rejecting his own father’s restraint of lawless desire (572b–573c). But the process by which this happens turns out to be slightly complicated. When Socrates describes the son being tempted by the purveyors of lawless desires, what follows is not that he simply gives in. Quite the reverse: like his father before him, he proves resistant. So the ‘tyrant makers’ devise some stratagem (572e5–6) to complete the change they helped to start. The means they use is er¯os, which hereafter becomes the key player in the description of psychological tyranny: er¯os is to the soul of the tyrant as a tyrant is to his city.

Lawless desires Although my main interest in this section is er¯os and the role it plays in the tyrant’s development, we should pause to take note of a problem about the nature of lawless desires. Socrates thinks that they are present in almost everyone (to a greater or lesser extent), but are kept under restraint. Asked by Adeimantus to explain what sort of desires he has in mind, he replies: Those that are awakened in sleep, when the rest of the soul—the rational, gentle, and ruling part—slumbers. Then the beastly and savage part, full of food and drink, casts off

Er¯os, Philosophy, and Tyranny

139

sleep and seeks to find a way to gratify itself. You know that there is nothing it won’t dare to do at such a time, free of all control by shame or reason. It won’t shrink from trying to have sex with a mother, as it supposes, or with anyone else at all, whether man, god, or beast. It will commit any foul murder, and there is no food it refuses to eat. In a word it omits no act of folly or shamelessness.³

Lawless desires clearly constitute some sort of ‘no-go’ area for the vast majority of people. But it is difficult to spell out precisely what a lawless desire is meant to be, and how it differs from a lawful one. The passage quoted might suggest a characterization in terms of objects: lawless desires are for things that might shame most people even to mention. Incest is explicitly cited; cannibalism might be another example.⁴ But this merely invites the question of what distinguishes lawless from lawful objects of desire, and so the original problem remains. The fact is that Socrates relies on his interlocutors’ intuitions when introducing the category of lawless desires. Contrast the way he distinguishes necessary and unnecessary appetites at 559b8–11. Here he gives us two criteria: unnecessary desires can be eliminated by the right kind of training; they also harm one’s rationality and self-control. The existence of these (empirical) criteria means that we do not have to rely on mere intuition to distinguish the necessary from the unnecessary. By contrast, Socrates gives us no equivalent means for distinguishing lawless from non-lawless desires in book IX. For the purposes of this chapter, we do not actually need a solution to this problem, although I would like to suggest a direction in which to pursue one. Socrates’ point may well be that lawless desires are inherently destructive of society and of the relations that should exist between human beings. Of course, ‘lawful’ appetites and spirited desires can damage the social fabric by generating competition for money and power. But at least there can be something we would recognize as a human society amid such competition, however unstable. Lawless desires involve the transgression of boundaries that would tend to destroy the very possibility of human relationships, family, and society, not just destabilize them (cf. 580a3–4). And it is not just human relationships that are at issue. The point is larger in scope: at 572d2 Socrates talks of the lawless element desiring sex even with a god, and this suggests that we should not limit the discussion to relations within the human realm. Lawless desires are blind to the sacred as well as to the social, and the tyrant has cut himself off equally from both.⁵ ³ 571c3–d4. Translations of the Republic are by G. M. A. Grube, rev. C. D. C. Reeve, in Cooper 1997 with slight modifications. ⁴ By way of further illustration, Nettleship (1901: 313) refers to Aristotle’s brief remarks on bestiality in Eth. Nic. VII 5, 1148b 19–25: ‘by bestial states I mean e.g. the female human being who is said to tear pregnant women apart and devour the children; or the pleasures of some of the savage people around the Black Sea who are reputed to enjoy raw meat and human flesh, while some trade their children to each other to feast on … ’ (trans. Irwin 1999). ⁵ See Grg. 507e3–508a4.

140

Dominic Scott

The tyrant’s er¯os Whether this is the best way of understanding the nature of lawless desires is a question best left for another occasion. I now wish to return to the main concern of this section, the role that er¯os plays in the tyrant’s development. At 572d5, Socrates imagines the democrat’s son experiencing the same temptations as his father and initially resisting the efforts of his corrupters. The text continues: .: Then, when those clever enchanters (magoi) and tyrant-makers have no hope of keeping hold of the young man in any other way, they contrive to plant in him a powerful er¯os, like a great winged drone, to be the leader of those idle desires that spend whatever is at hand. Or do you think that er¯os is anything other than an enormous drone in such people? .: I don’t think it could be anything else. .: And when the other desires—filled with incense, myrrh, wreaths, wine and the other pleasures found in their company—buzz around the drone, nurturing it and making it grow as large as possible, they plant the sting of longing in it. Then this leader of the soul adopts madness as its bodyguard and becomes frenzied. If it finds any beliefs or desires that are thought to be good or that still have some shame, it destroys them and throws them out, until it’s purged him of moderation and filled him with imported madness. (572e4–573b4)

On their own, lawless desires are unable to push the youth over the edge. This is one thing that makes this stage in Plato’s account of degeneration different from its predecessors. In all the other cases, the son’s character evolved more smoothly. True, there were usually forces helping from outside, but they did not have to push too hard given the psychological background already in place.⁶ The tyrant-makers, on the other hand, have to find some further means to bring about the transformation. What must qualify er¯os for this role is that, as a species of desire, it has a peculiar intensity that transforms its subject’s attitude to other concerns, in this case blinding him to the importance of social norms and boundaries. The democrat’s son has already weakened such respect by allowing the growth of lawless desires, though he has not yet gone so far as to destroy it. But in developing a powerful er¯os in him, the corrupters find a mechanism that will do so, once and for all. Socrates clearly thinks that such a conception of er¯os will strike a chord with intuition. Er¯os, he reminds Adeimantus, is often thought to be a tyrant (573b6–7): the point is that its intensity may create an attitude of indifference towards others’ concerns. In the same spirit he makes a connection between er¯os and madness (573c3). ⁶ See 549c8–550b7 and 559d7–561a4. Nettleship (1901: 314) and Annas (1981: 303) seem to ignore this point by suggesting that the slide from democracy to tyranny in the soul is inevitable.

Er¯os, Philosophy, and Tyranny

141

Socrates and Glaucon have already associated sexual er¯os with madness in a passage about sexual er¯os in book III: .: Is excessive pleasure compatible with moderation? .: How can it be, since it drives one mad just as much as pain does? .: What about the rest of virtue? .: No. .: Well then is it compatible with insolence (hubris) and licentiousness? .: Very much so. .: Can you think of a greater or keener pleasure than sexual pleasure? .: I can’t—or a madder one either. (402e3–403a7)

In general, pleasure, like pain, has the power to destroy one’s beliefs about social norms (nomim¯on: cf. 413b9–c3 with 430a1–b5). But er¯os is in a league of its own, and it is its connection with hubris that we should take seriously. Hubris is something that concerns one’s relations with others, whether human or divine. It is a matter of overstepping social or religious boundaries. This is exactly the point implicit in book IX: what the tyrant-makers need to do is to destroy the inhibitions that still prevent the democrat’s son from sliding into the lawless life. He has lawless desires, but they are not yet strong enough for him to abandon the tenets of his upbringing. Hence the tyrant-makers engender a specific kind of desire that has a unique capacity to transform the soul. ¯ I I . T Y R A N N I C A L A N D PH I LO S O PH I C A L EROS C O M PA R E D Anyone wishing to pursue the comparison between the tyrant and the philosopher might be tempted to focus on the connection drawn in the Republic IX between er¯os and madness (573c3–5; cf. III, 403a7). This resonates with the Phaedrus, which first talks of a harmful form of er¯os, linked to hubris, and described as a form of madness (cf. 241a4 and 265a9–10), but then goes on to discuss other forms of madness—many of them beneficial—and other forms of er¯os, including philosophical er¯os, a form of beneficial madness. If the Republic also conceives of the philosopher’s er¯os —not just the tyrant’s—as a form of madness, the comparison between the two starts to go deeper. The conception of philosophical er¯os in the Phaedrus will indeed turn out to be crucial for our purposes, but we need to sound a note of caution. In the passage quoted above from Republic III, Socrates links sexual er¯os to madness but immediately goes on to distinguish a better form of er¯os: .: But the right kind of er¯os is by nature that of order and beauty, and is moderated by education in music and poetry? .: That’s right.

142

Dominic Scott

.: Therefore, the right kind of er¯os has nothing mad or licentious about it? .: No, it hasn’t. .: Then sexual pleasure mustn’t come into it, and the lover and the boy he loves must have no share in it, if they are to love and be loved in the right kind of way? (403a8–b3)

Although this passage does not mention philosophical er¯os, what it says must apply to any form of er¯os of which Socrates approves, including the philosophical, which only appears on the scene at the end of book V. This seems to suggest that the Republic does not treat all forms of er¯os as madness, and that the idea might be peculiar to the Phaedrus. If so, forging a link between the er¯os of the philosopher and that of the tyrant within the Republic is not as straightforward as we may have thought. Nevertheless, before we give up the comparison altogether, we should turn to examine the account of philosophical madness in the Phaedrus.

The Phaedrus The speech characterizing er¯os as a kind of divine madness (244a–257b) comes as a riposte to an earlier one that described it as essentially harmful (237b–241d). The account of er¯os sketched there took it to be a desire of some kind, having as its object something beautiful or, more specifically, pleasure taken in the beautiful. What marks out er¯os, though, is its relation to hubris. Our innate desire for pleasure can always come into conflict with reason. Wherever it does so, the result is hubris, which can take different forms. Where the pleasure is for food, we call it gluttony. Where it is for physical beauty, we call it er¯os. The danger inherent in er¯os, therefore, is that it is an irrational desire overpowering reason—hence a case of hubris (238a1–c4). The point of the speech as a whole is to show that one should never gratify anyone affected by er¯os. The riposte to this comes as soon as the next speech begins: There is no truth in the speech that claims that when a lover is present one should rather gratify the person who does not love, on the grounds that the former is mad and the other is in control of himself. If madness were bad simpliciter, the speech would have been fine. But in fact the greatest of goods come to us through madness, as long as it is a divine gift. (244a3–8)

Socrates then proceeds to show that not all forms of madness are bad, and some—those that come from the gods—are highly beneficial. Examples of benign madness include prophecy, possession, and poetry. But the greatest kind is divinely inspired er¯os. The connection between such er¯os and madness is then explained through the myth of the winged charioteer, which begins at 246a6. In its discarnate state, the soul, pictured as a charioteer leading two horses, caught a glimpse of the forms by following in a procession with the gods. Those souls that did not catch an

Er¯os, Philosophy, and Tyranny

143

adequate view lose their wings and fall down to earth. But as long as they had enough of a glimpse, they are able to recollect the vision of beauty in a beautiful body. The sight of physical beauty causes the wings to start re-growing, a process that involves the most intense mixture of pleasure and pain (251a1–252b1). This awakening of the incarnate soul brings about a marked change of behaviour: what previously seemed important—family, friends, money, conformity to social norms—ceases to be so: [The soul of the lover] will not willingly leave [the beloved] and doesn’t value anyone more highly than the beautiful boy, but it forgets completely about mother, brothers and companions and isn’t concerned in the slightest if it loses its property through neglect. As for social norms (nomim¯on) and seemly behaviour (eusch¯emon¯on), in which it used to take pride, now it despises them all (kataphron¯esasa). (252a1–6)

The lover leaves all this behind to spend as much time as possible with the beloved. Such behaviour is what gains the lover a reputation for madness. The point is recalled at 265a6–11: We said that er¯os is a sort of madness, didn’t we? … And that there are two forms of madness, one caused by human illness, the other by a divine release from socially accepted norms (nomim¯on)? (265a6–11)

In sum, the argument of the second speech is that madness comes in different types, which may or may not be harmful. Socrates agrees with the implicit claim of the previous speech that er¯os is a form of madness, but disagrees with its assumption that madness is always harmful. So far, we can see that the Phaedrus has something to tell us about er¯os quite generally, regardless of whether it is of the harmful or beneficial variety: on first impact, at least, er¯os affects its subjects with a sort of tunnel vision that may divert them from other concerns. We can see such concerns dividing into various types. One might loosely be termed prudential: lovers cease to care about their personal possessions (252a3–4); there is also, perhaps, an indifference towards reputation: they no longer care if they appear unconventional.⁷ But there is something else from which er¯os diverts its subject: social and familial relations. People with whom one has special ties of kinship or friendship become pushed to the margins. To turn now specifically to divinely inspired er¯os, this is not something that all humans experience, only those who had a sufficient view of the forms beforehand. But all inspired lovers experience the powerful mixture of pleasure and pain described at 251a1–252b1. However, within this group there is a distinction to be made. After they initially recognize the form in the beautiful particular, some souls follow the process of recollection through to the end. These are the souls of the philosophers, the ones who saw most in their previous ⁷ The word eusch¯emon¯on (252a5) suggests an interest in appearances.

144

Dominic Scott

life (248d2–3). As a result of continued recollection in this life and in two subsequent human incarnations, they re-grow their wings completely, and return to the divine procession. There are also others who experience the initial impact of recognition, but do not complete the process.⁸ We shall return to these inspired, but nonphilosophical souls in a moment, but first I wish to focus on the philosopher and his attitude to the social. The crucial passage comes at 249c4–e1: Therefore it is right that only the philosopher’s mind should grow its wings: through memory, it always remains as close as possible to those things by proximity to which the gods are divine. A man who makes correct use of these reminders is always at the most perfect level of initiation, and he alone becomes truly perfect. Standing outside human concerns and coming close to the divine, he is considered by the many to be out of his mind, although what they don’t realise is that he is divinely possessed. This is the point to which the whole discussion of the fourth kind of madness has been leading. Whenever someone sees the beauty that exists down here and is reminded of real beauty, he grows his wings and lifts them in his desire to fly up, but is unable to do so. Like a bird, he looks upwards and neglects what is down below. This is why he is accused of being mad.

This passage is central to our interests. It shows that the madness of the philosophical lover, i.e. of someone who has followed the course of recollection to the end, consists in leaving the realm of the human behind. We already know from 252a1–6 that the philosopher neglects ties of family and friendship. I take the present passage to include this point, and to show that the philosopher is similarly indifferent to the affairs of the city. Qua lover, the philosopher becomes less and less of what Aristotle was to call a ‘social animal’,⁹ and I shall use the word ‘asocial’ to capture the way he is disinclined to engage in political activity, or to see social life as something worth pursuing for its own sake. (He is not a complete loner, of course, because he seeks out the company of his beloved. Yet this does not show that he is drawn to human interaction as such, but that he aspires to the divine.) Despite the evidence of the text just quoted, some might challenge the conclusion that the philosopher in the Phaedrus is asocial by pointing to a later passage, 252c3–253c6. Here, Socrates distinguishes different types of people among the divinely inspired lovers. In the myth, the twelve gods lead their own processions to the forms. Once incarnated, those humans who start to see something of the form in the particular differ from one another because they emulate the character of the particular god they originally followed. They also seek out a beloved who potentially has a similar character, which they then try to nurture. Anyone with a Zeus-like nature is philosophical. But in the same sentence (252e3), we learn that he is also ‘commanding’ (h¯egemonikos), a word ⁸ See Ferrari 1992: 267: ‘the inspiration of love opens up a potential path but does not ensure that it will be followed’. ⁹ Plt. I 2, 1253a 3; Eth. Nic. I 7, 1097b 11.

Er¯os, Philosophy, and Tyranny

145

that Hackforth (1952: 99) translates as ‘being disposed … to the leading of men’. This seems to imply, contrary to what I have been arguing, that such a lover will be as naturally inclined to political leadership as to philosophy, and will seek to mould his beloved in this image. But what does ‘commanding’ really mean? Shortly after this point, the Heralike nature is said to be ‘kingly’ (basilikon, 253b1–2). If we had already taken ‘commanding’ in the sense Hackforth proposes, what would be the difference between the followers of Zeus and Hera? I suggest that, in the context of the lover and the beloved, h¯egemonikos alludes to the leadership of the charioteer over his horses. Of all the different characters, it is the Zeus-like one who is most able to steer his horses in the right direction. The same word appears in the Protagoras when Socrates asks whether knowledge is powerful enough to rule (h¯egemonikon) in the soul (352b4). Similarly, the knowledge of the Zeus-like character controls the passions and appetites of the soul, ‘showing them the way’.¹⁰ Further confirmation of this approach comes from 248d2–e3, where Socrates lists the different incarnations people will enter depending on the extent of their view of the forms in their previous existence. Whoever saw the most will enter the bodies of those who become philosophers, lovers of beauty and devotees of the Muses.¹¹ Next are those who become lawful kings or military commanders; the third group includes statesmen—and so on until one reaches the last group, the tyrants.¹² Quite how this list maps on to the distinction drawn later in terms of who followed which god is not fully clear.¹³ But for our purposes, it is enough that the first group is distinguished from the political souls in the second and third; nothing is said about any inclination to rule on the part of the philosophers. This fits well with the exclusively contemplative and other-worldly account of the philosophical lover that follows almost immediately (249c4–e1, quoted above). The overall point is that some divinely inspired lovers are drawn towards politics, while others retreat in the opposite direction towards the divine. It is worth dwelling in more detail on the distinction between the political and the philosophical lovers, all of whom are divinely inspired. As I have argued, all inspired lovers glimpse something of the form when they first see the beloved. Some see less than others, so their recollections will differ, but they must all experience some degree of recognition. But after this initial frenzy, they do not ¹⁰ My approach is broadly in agreement with Thompson (1868: 69–70), who takes h¯egemonikos in what he calls an ‘esoteric’ sense, construing the difference between Zeus and Hera, between the terms h¯egemonikos and basilikos, as between the speculative and practical intellect. ¹¹ As Hackforth (1952: 83) notes, these are different aspects of the same character. ¹² Hackforth (1952: 83) takes the reference to a ‘lawful’ (ennomou) king in 248d4 to suggest the contrast of the Statesman between the ideal ruler who does not need law and second-best rulers who do (cf. 297d–e and 302e). See also Rowe 1986: 181. In my view, however, we do not have to import a theory from the Statesman to make sense of the word ennomou. If we do take it to imply a contrast with another group on the list, the contrast may lie in the opposite direction, i.e. with the tyrant (248e2), who is of course lawless. ¹³ Perhaps the second group includes the followers of both Hera and Ares. See Ferrari 1987: 267.

146

Dominic Scott

all behave in the same way. Recall again the passage quoted above, 249c4–e1. This is concerned only with souls that complete the initiation to the end—the philosophers, the followers of Zeus, and the only ones to regain their wings. They achieve a complete distancing from the world of human affairs, and do so because they are so keenly aware of being in the presence of the divine. But the others, e.g. the followers of Hera, take a different route. They do not complete the initiation and become immersed in human affairs. Perhaps they will always stand apart—or out—from others, but not to the same extent as the philosophers. Their madness is most marked at the initial stage—not, as with the philosopher—all the way through to the end.¹⁴ In all this there is a general point about er¯os: it involves an obsessive desire that has the power to block out other concerns. But it is in the case of the philosophical lover that this phenomenon reaches its culmination. It is not only at the initial impact that he becomes asocial, but he continues in this way and develops the attitude to an extreme the more he recollects. I stress again that, although all inspired lovers are described as neglecting family and friends, only the philosopher is said quite generally to neglect the human in favour of the divine (249c9–d8). His attitude to the human, the social, and the political is continuous with—indeed, the culmination of—the indifference to convention that all inspired lovers initially show.

Back to the Republic Does the Republic also see the philosophical lover as gripped by a form of madness? Perhaps not, on a straightforward reading of Republic III, 403a8–b3: after all, this does say that good forms of er¯os have nothing to do with madness. On the other hand, it is possible that the word ‘mad’ (manikon, 403a10) is employed here only in a narrowly pejorative sense, and that the passage should not actually be used to rule out the existence of another form of madness—a benign one—that is associated with er¯os. Even so, the fact remains that Socrates never talks explicitly of philosophical er¯os as being a form of madness in the rest of the Republic.¹⁵ But my argument does not depend on claiming that Socrates actually calls the philosophers in the Republic mad. What matters most in comparing the tyrant and the philosopher is whether the features in virtue of which the philosopher of the Phaedrus is called mad also apply to the philosophers of the Republic;¹⁶ in ¹⁴ Here I agree with Ferrari (1992: 267). ¹⁵ One exception is 539c6, where Socrates talks pejoratively of young men with a ‘mania’ for refutation. But he is not, of course, talking about the activities of true philosophers in this passage. (Nor, incidentally, does this mania have anything like the psychological power of erotic love; the young men are compared to puppies tearing away anyone who approaches them.) ¹⁶ See Irwin 1995: 305. Nussbaum (1986: 203–23) drives a wedge between the two dialogues, contrasting the purity of reason in the Republic with the impassioned philosophic er¯os of the

Er¯os, Philosophy, and Tyranny

147

particular, whether the philosophers of the Republic are asocial in the same way as philosophers are in the Phaedrus. There are various places one might look to show this. In the channel analogy of 485a–487a, Socrates talks of the intensity of the desires associated with er¯os, especially philosophical er¯os. At first sight, this sits well with the way in which er¯os in the Phaedrus diverts one from various concerns. Both texts explicitly mention the concern for money, which gets pushed to one side (485e3–5 and 253a3–4). However, I do not think this passage on its own can be used as evidence that the Republic shares the same asocial conception of philosophical er¯os as the Phaedrus. Two questions would have to be settled for the link to be established. First, the Phaedrus talks of indifference (amelia), the Republic merely of weakness of desire: should the two be assimilated? Second, does the Republic include social concerns among those that er¯os has a tendency to weaken, as it does with money and physical pleasure? Another passage that might support the similarity between the two dialogues is Republic 500b8–c1, where the philosopher is said not to be interested in human affairs (anthr¯op¯on pragmateias), but instead to look to the divine order of the forms, on which he wants to model his soul. At Phaedrus 249c8–d3, Socrates talks of the philosopher standing outside human concerns (anthr¯opin¯on spoudasmat¯on) as he draws near to the divine. One might object that, whereas this sentence does refer to a quite general disregard for human affairs, Republic 500b8–c1 could be talking about something more specific: the philosopher’s disdain for the particularly competitive aspects of human associations and political life. But if it is difficult to use 500b8–c1 on its own as firm evidence for the more generalized social indifference that we find in the Phaedrus, the same cannot be said of another text, this time a remark made about death at Republic 604b12–c1, where the philosopher thinks that human affairs (anthr¯opin¯on) are not worth much concern (spoud¯es).¹⁷ This certainly does seem to echo the philosopher’s indifference to the social in the Phaedrus. But I think the best place to turn if we are to find the view of the Phaedrus alive in the Republic is the notorious issue of the philosophers’ return to the cave. When Socrates announces that the philosophers of the ideal state will not be permitted to contemplate full time, but will be compelled to rule, Glaucon objects that this constitutes an injustice towards them (519d8–9). In what follows, Socrates persuades Glaucon that it is just to require them to rule, though Phaedrus. But this ignores the fact that the Republic treats reason as a motivating force, no less than spirit or appetite; cf. 580d8, where all three have their own desires and pleasures. Also, it is difficult to see what sense she can make of the channel analogy (485a–487a), whose whole point is that the motivating force of reason, its er¯os, is so powerful that it leads to the weakening or even abandoning of other desires (cf. also 581b6–7). ¹⁷ Cf. also 486a8–b2: ‘will a thinker high-minded enough to study all time and all being consider human life to be something important?’

148

Dominic Scott

he acknowledges—and emphasizes—that the philosophers will rule reluctantly and under compulsion. This passage has exercised a number of scholars. The usual point of contention is whether the requirement to return creates a contradiction in the Republic as a whole. To return is the just thing to do, but some would argue that it is against the philosopher’s interests, in which case we seem to have a counter-example to the thesis that Socrates purports to defend from book II onwards: that justice is always in the agent’s interest. I shall not tackle this question head-on here, but focus on the philosophers’ motivation as they return to administer the affairs of the city. What I wish to show is that the way in which Socrates describes their motivation supports the claim that there is something essentially asocial about philosophical er¯os. For the sake of argument, let us assume that there is no contradiction with book II, and that the return is in the philosophers’ interest. Nevertheless, why are they reluctant to return, and in what sense are they ‘compelled’ to do so? Socrates mentions compulsion several times in this context (500d4, 519e4, 520a8, 520e2, 521b7, 539e3, and 540b5), and he goes out of his way to stress their reluctance to rule: the state will be better governed if there is something its rulers would rather be doing than ruling (520e4–521a8). Although it may seem as if the philosophers return simply because they think it just to reciprocate for the education that they have received, the argument is more complicated.¹⁸ The founders (Socrates and his interlocutors) will impose a law on the philosophers, requiring them to rule. They do not actually impose this law out of a demand for reciprocity, but because of utility: the state will be best run by reluctant rulers, educated in the way described in the work so far. Glaucon’s objection is that this law is unjust because it will make the philosophers lead a less happy life than they would otherwise do. It is to counter this that Socrates introduces the point about reciprocity: because the philosophers have received an education from the state, it is just to impose the law. As Eric Brown has argued (2004: 282), he does not say that the founders are required to impose the law because of reciprocity, but that they are allowed to do so. (Nor does Socrates argue that the rulers are required to rule directly by considerations of reciprocity; rather, they are required to obey a justly imposed law.) The founders’ reasons for imposing the law follow directly from utility, rather than reciprocity, which means that, if there were no law (e.g. because the founders had found a better way of setting up the state), the philosophers would not return to the cave. The compulsion at issue is one of legal necessity. The crucial point to draw from all this is that, in returning, the philosophers are diverted from what they would otherwise most naturally do—were it not for the existence of the law. This helps to support the idea, suggested by the other ¹⁸ See Brown 2000 and 2004: 277–83, esp. 282. Contrast Kraut 1999: 247–9, who places the emphasis more directly on reciprocity.

Er¯os, Philosophy, and Tyranny

149

passages mentioned above (485a–487a, 500b–c, 604b–c), that philosophical er¯os is by its nature asocial.¹⁹ To clarify the point, contrast an interpretation that makes ruling fall naturally out of philosophical er¯os. This interpretation, proposed in slightly different forms by Peter Vernezze and Terence Irwin, seeks to explain the philosophers’ preparedness to rule in the Republic by importing elements of Diotima’s speech in the Symposium.²⁰ There she tells us that all lovers strive after immortality and seek to extend their existence by ‘giving birth in the beautiful’ (206e5). Some people do this at the bodily level, by childbirth; others are ‘more pregnant in soul’, and give birth to offspring of a very different kind (208e2–209a4). According Vernezze and Irwin, what this means is that they ‘propagate’ their virtue in others. A pedagogical lover, for instance, aims to educate his beloved, and legislators like Lycurgus and Solon instituted laws that created virtue in successive generations (209d4–7). As for the philosophical lover, Diotima tells us that he ascends the ladder of love, rung by rung, and at the top gives birth to ‘true virtue’ (212a3–6), which these interpreters again take to mean the generation of virtue in someone else: having attained philosophical wisdom (‘true virtue’), i.e. knowledge of the form of beauty, they then seek to help others to do the same. Applied to the Republic, the point is supposed to be that the philosophers will want to propagate their virtue both by educating others to be philosophers (cf. 540b5–6) and by creating demotic virtue in the citizens at large (cf. 500d4–8). Only by doing this—by engaging in politics—will they secure their continued existence beyond the grave. In stark contrast to the view I have espoused, political activity has now become a direct and natural outcome of philosophical er¯os so that it would be quite wrong to see any discontinuity between philosophy and politics. There are, however, some serious objections to this interpretation that, taken together, render it very implausible. First of all, these interpreters owe us a good explanation as to why so much has to be imported into the Republic. Once Glaucon has voiced his concern about the happiness of the philosophers, Socrates makes a long and detailed reply (519d8–521b11). Yet, according to Vernezze and Irwin, he manages to miss out the most crucial element of that reply: there is not even the faintest whisper about er¯os, let alone about the propagation of virtue in this passage. Why? Secondly, the reading of the Symposium adopted by Vernezze and Irwin is itself highly contestable, particularly the claim that the philosophical lover seeks to propagate virtue in another person. Before we go any further, we should underline the distinction that Diotima makes between the lovers of the lower and higher ¹⁹ As in the Phaedrus, the philosophical lover of the Republic is not a complete recluse, but welcomes the company of fellow philosophers (cf. 520d8). Nevertheless, philosophers do not, qua lovers, seek company for company’s sake, but to further their pursuit of contemplation through dialectic. ²⁰ Vernezze 1992: 340–8 and Irwin 1995: 302–16.

150

Dominic Scott

mysteries (209e5–210a2). Both groups are ‘more pregnant in soul’ than those who seek immortality through biological offspring (cf. 209a1–2). The lovers of the lower mysteries, who include the poets and lawmakers, are described in 208e5–209e4, the passage immediately preceding the reference to the two kinds of mysteries. The lovers of the higher mysteries include those who ascend the ladder of love methodically and in the right order—the philosophers. Now it is true that the first group do seek to propagate virtue in certain others (cf. 209c1–2 and e2–3), though we have to be careful not to exaggerate the importance of this point. As 209d1–e4 makes clear, the real vehicle for immortality is honour and memory, a point already established at 208c2–d2: all lovers seek to lay up immortal fame (kleos) for themselves. Thus the poets are remembered because of their poems, and the legislators because of their laws. They might also be remembered because they propagated virtue in others, but that is not the primary focus of this passage. Socrates is not claiming that one achieves immortality directly by replicating one’s own psychological state in others—the psychological equivalent of spreading one’s genes. If the propagation of virtue does play a role in this passage, it is as one means among others (e.g. the composition of poetry, the making of laws) of ensuring that one is remembered after death. So when we turn to the lovers of the higher mysteries, the philosophers, we cannot simply assume that they too will be intent upon spreading their virtue (i.e. their wisdom) in others. We have to look for explicit signals in the text for such a view—yet there is no mention of their producing virtue in anyone else in the whole of this passage. True, at the culmination of the ascent Diotima says that the philosophical lover produces genuine virtue (212a4–5), but it would be begging the question to assume that this refers to virtue created in another, as opposed to the virtue (i.e. wisdom) they themselves acquire in apprehending the form. Of course, the question arises as to how they achieve immortality, but it is not as if the Vernezze–Irwin answer is the only solution possible. Diotima may simply mean that, like the lawgivers and the poets, they are remembered long after their death for their wisdom (and not just by men, but by gods as well: 212a6). So in my view, Vernezze and Irwin have to read the ‘propagation’ model not just into the Republic, but into the Symposium as well,²¹ which makes it a very fragile basis for any solution to the problem of the philosophers’ return. Particularly troubling for this interpretation are Socrates’ references to ‘necessity’ and ‘compulsion’ in discussing the philosophers’ requirement to rule. If propagating virtue in others is the natural expression and outcome of philosophical er¯os, these references become very puzzling indeed. Although various attempts have been made to explain them away,²² there is one fact that poses an insuperable difficulty for the propagation view. Socrates does not just say that the return is necessary, but that they, the founders, will compel or ‘necessitate’ the ²¹ For a more detailed critique of the interpretation, see Sheffield 2006: ch. 4. ²² For a discussion of these, see Brown 2000: 6–9.

Er¯os, Philosophy, and Tyranny

151

philosophers to return. If the philosophers wanted to engage in ruling in order to secure their continued existence, why would they need legal compulsion from the outside? All the necessary motivation ought to come from within.²³ In the Symposium, by contrast, lovers will do absolutely anything to ‘lay up immortal glory forever’ (208c5–6); they hardly need any external compulsion. The fact that the philosophers have to be ‘compelled from the outside’ to rule makes a great deal of sense if we take philosophical er¯os to be inherently asocial, as I have argued. Left to its own devices, this er¯os would lead one away from the human and political towards the divine, exactly as we find in the Phaedrus. It is only by the creation of an artificial law—a contingent device—that the stream of er¯os can be diverted to pass over political terrain. As soon as it has completed this diversion, it reverts to its original path, leading away from the human and the social, back to pure philosophical contemplation. One point of clarification: I do not wish to say that the philosophers’ er¯os blinds them to social concerns. This would be to suggest that philosophical understanding could be the source of a mistake—an underestimation of something’s true worth. The philosopher sees things in their true perspective; and from that perspective, cosmic in its scope, political activity is not worth very much, though it is worth something. As Iris Murdoch (1977: 13) put it: ‘human affairs are not serious, though they have to be taken seriously.’

C O N C LU S I O N I started this chapter by claiming that both the tyrant and the philosopher are obsessive in a way that the other characters in the Republic are not: both are driven by er¯os. But as we went on, we realized that the comparison goes much further. Their er¯os makes them, in different ways, asocial. Surprising as it may sound, there is a link between the tyrant’s disregard of social boundaries and the philosopher’s natural reluctance to return to the cave. In all this Plato is exploiting an entirely familiar point: in the moment of erotic obsession, there is a neglect of convention and society. Both the tyrant and the philosopher take this feature to an extreme, though in entirely different ways. In the tyrant’s case, such neglect—temporary in the case of most lovers—becomes transformed into a permanent blindness to the existence of norms and boundaries, whether social or religious; in the case of the philosopher, a permanent devaluing of the human, the social, and political per se, in favour of the divine.²⁴ What the ²³ This is the point of the word prosanangkazontes (520a8). For a nice parallel (in a different context), see Symp. 181e5, which uses the same word for the external imposition of a law to modify the direction that one’s internal motivations would otherwise take. ²⁴ A further point of comparison is that the philosopher is godlike, and hence above mere mortal nature (500c9–d1 and 518e2), and the tyrant sees himself as such (360c and 568b). On this aspect of philosophy, see Vlastos 1994: 63.

152

Dominic Scott

two figures have in common is that they represent one of the most basic features of er¯os, but pushed to an extreme. Finally, we should mark out some differences between the two characters, even at the risk of stating the obvious. Clearly, we need a way of showing that the philosopher is in no way lawless,²⁵ which is not too hard: qua lover, the philosopher is asocial, but not anti-social. When describing the tyrant’s nature, Socrates focuses on his anti-social tendencies: because the tyrant is firmly within the appetitive realm, where the objects of desire involve physical gratification, he needs to plunder other people’s physical resources. This is a theme familiar from the Republic, which sees pleonexia as something often caused by appetitive desire. Physical resources are limited, so a competition between parties becomes established. In this competition there are winners and losers, and the tyrant takes all: his victory is everyone else’s loss. By contrast, the philosopher disdains the physical objects of appetite, as the channel analogy makes clear (485d12). Precisely because of his er¯os, he has no interest in engaging in the competition for material goods or indeed for honour. Other passages in the Republic illustrate this point well: as well as the channel argument, 500b–c paints a distinctly other-worldly portrait of the philosopher, something repeated in the description of the timocrat’s father with his disregard for family fortunes or honour (459c2–550b2). Again, towards the end of book IX, Socrates repeats how the love of learning makes one indifferent to money and honour (581b5–8). So there is no danger that the philosopher’s er¯os will make him anti-social in the manner of the tyrant. Furthermore, it will increase his sense of reverence for the divine. Above, I pointed out that lawlessness not only ignores the norms of human society, but also disregards our ties and duties towards the gods. Exactly how Plato’s metaphysical approach to the divine translates into a more conventional religious attitude is too large a subject to take on here, but he would surely insist that contemplation of the forms increases one’s commitment to religion and piety.

REFERENCES Adam, J. (1963), The Republic of Plato, 2 vols., 2nd edn. Cambridge. Annas, J. (1981), An Introduction to Plato’s Republic. Oxford. Bloom, A. (1968), The Republic of Plato. New York. Brown, E. (2000), ‘Justice and Compulsion for Plato’s Philosopher-Rulers’, Ancient Philosophy, 20: 1–18. ²⁵ Bloom (1968: 424) writes: ‘both the tyrant and the philosopher depreciate law … ’. Although he is right in general to stress the similarity between the two characters, the claim that the philosopher ‘depreciates’ law requires some qualification, especially in the light of the philosophers’ willingness to obey the law that sends them back down to the cave.

Er¯os, Philosophy, and Tyranny

153

(2004), ‘Minding the Gap in Plato’s Republic’, Philosophical Studies, 117: 275–302. Cooper, J. M. (1997) (ed.), Plato, Complete Works. Indianapolis. Cornford, F. M. (1941), The Republic of Plato. Oxford. Ferrari, G. (1987), Listening to the Cicadas. Cambridge. (1992), ‘Platonic Love’, in Kraut (1992), 248–76. Ferrari, G. (ed.), and Griffith, T. (trans.) (2000), Plato: The Republic. Cambridge. Fine, G. (ed.) (1999), Plato II: Ethics, Politics, Religion and the Soul. Oxford. Grube, G. M. A. (1971), Plato: The Republic. London. Rev. edn. by C. D. C. Reeve in Cooper (1997). Hackforth, R. (1952) (ed.), Plato’s Phaedrus. Cambridge. Irwin, T. (1995), Plato’s Ethics. Oxford. (1999) (trans.), Aristotle: Nicomachean Ethics. Indianapolis. Jowett, B. (1953) (trans.), The Dialogues of Plato, 4 vols., 4th edn. Oxford. Kraut, R. (1992) (ed.), The Cambridge Companion to Plato. Cambridge. (1999), ‘Return to the Cave: Republic 519–521’, in Fine (1999), 235–54. Lee, D. (1974), Plato: The Republic, rev. edn. Harmondsworth. Murdoch, I. (1977), The Fire and the Sun: Why Plato Banished the Artists. Oxford. Nettleship, R. L. (1901), Lectures on the Republic of Plato. London. Nussbaum, M. C. (1986), The Fragility of Goodness. Cambridge. Richards, I. A. (1966), (ed.), Plato’s Republic. Cambridge. Rowe, C. J. (1986), Plato: Phaedrus. Warminster. Sheffield, F. C. C. (2006), Plato’s Symposium: The Ethics of Desire. Oxford. Shorey, P. (1930), Plato’s Republic, Loeb edn., 2 vols. Cambridge, Mass. Thompson, W. H. (1868), Plato: Phaedrus. London. Vernezze, P. (1992), ‘The Philosophers’ Interest’, Ancient Philosophy, 12: 331–49. Vlastos, G. (1994), Socratic Studies. Cambridge. Waterfield, R. (1993), Plato: Republic. Oxford.

8 Virgil’s Sacred Duo: Phaedrus’ Symposium Speech and Aeneid IX Robert Wardy

In our times, no scholar in the field of ancient philosophy has contributed more than Myles Burnyeat: spanning the complete temporal range of the subject (with forays beyond), across every department of philosophy, his studies have both helped to revivify the most traditional and worked-over authors and topics, and also stimulated sustained interest in the unjustly neglected. But Burnyeat’s magisterial impact has not been limited to philosophy ‘proper’. What is philosophy ‘proper’? When in self-reflective mode, we read Burnyeat with immense profit and pleasure, especially his interpretations of Plato and Platonism. But however one deals with the perennial issues of philosophical self-definition, we also recognize that philosophical doctrines and attitudes make themselves felt in literary writings not themselves strictly philosophical, on any definition of ‘philosophy’. How best to approach such movements across genres? Such questions interest Myles Burnyeat,¹ and my foray into philosophical and literary interaction is a shot at an answer I hope he might enjoy. In the following I attempt two things. First, I try to show that Virgil’s Aeneid is influenced by Plato’s Symposium.² In the nature of the case, such arguments can never be probative. Critics who claim to perceive striking continuities Earlier versions of this paper were delivered at Princeton University, The Central European University in Budapest, and the Center for Hellenic Studies in Washington DC. I wish to record my gratitude for my audiences’ helpful questions and suggestions, especially the input of Mark Buchan, Denis Feeney, Andrew Feldherr, Andrew Ford, Richard Hunter, Robert Kaster, Emese Mogyorodi, Gregory Nagy, Carole Newlands, Christopher Rowe, Nick Rynearson, and Victoria Wohl. Michael Reeve’s and David Sedley’s charitable criticism is much appreciated. My greatest debt is to Philip Hardie and John Henderson, whose generous, imaginative, and learned reactions have improved this work immensely. ¹ One notes Kenney’s (1990) acknowledgement of his advice on antiquity’s most celebrated literary Platonist. ² For quite some time literary theorists in general, and Latinists in particular, have discussed whether and to what end one might discriminate between e.g. ‘allusion’ (often, but not invariably, taken to import a subjective relation, wherein the later author consciously alludes to a predecessor, to some intended effect) and e.g. ‘intertext’, ‘intertextual’, etc. (often, but not invariably, taken to prescind from subjectivity, either in circumspection or in avowed hostility to appeals to authorial

Virgil’s Sacred Duo

155

between texts placed within the same tradition or featuring in different ones amass cumulative evidence for real connection. With rare exceptions, they rely, negatively, on the weight of their case rendering dismissive appeal to brute coincidence or mere generic similarities (‘but this is simply a topos in the public domain’) unattractive; positively, on how fruitful we find the experience of rereading these texts in the light of the purported intertextual allusion. This will never amount to proof positive; and my claim about Plato and Virgil will remain vulnerable to sceptical pressures. Nevertheless, this is the relatively easy bit of the operation. Second, I shall conjecture about what Virgil does to incorporate his Platonic material. Virgil, most learned and allusive of authors, never superficially tosses in a citation or generates an echo for its own sake: for him, writing is rewriting, as he harmoniously or polemically engages with the multiple traditions which he so spectacularly enriches. His encounters with Homer, Hesiod, Callimachus, Ennius, Lucretius, Catullus, and others are intricate and subtly ambitious affairs; but his tacit dialogue with Plato is another sort of conversation altogether. Or is it? Philosophers keen on staking out their special intellectual territory might be exercised by Virgil’s tapping into Plato, on the assumption that this borrowing is different in kind from, say, epic and elegiac cross-fertilization. When philosophy goes afield, does it go astray? I am convinced that the Nisus and Euryalus episode in Aeneid IX, the most captivating and problematic section of Virgil’s poem, draws curious inspiration from Phaedrus’ speech in Plato’s Symposium. The minimum of which I am confident is that the writing of Aeneid IX is informed by a reading of Phaedrus’ speech: this bald causal hypothesis is readily defensible. But how might one helpfully specify the character of that causal relationship? So far I have done no better than resort to woolly alternatives, that the Symposium ‘inspires’ Virgil, or that Virgilian writing is ‘informed by’ Platonic reading. Meaning? My core conviction is that I have read a section of Virgil which is somehow rooted in some Plato; that a part of the Symposium is a partial blueprint, model, or template for a section of the Aeneid (please note: ‘partial’). How to make even modest progress in articulating that relation? This is, of course, not just a task in the explication of these two masterworks, but rather one instance of the general challenge we face in mapping the classical tradition. However learned we Classicists might become, it is as well to remind ourselves that the tradition has always already been constituted by the very learned authors who make it up—and I do mean ‘make it up’ in both senses of the phrase. Finally, what about movement in the other direction? No, I am not suggesting that Plato pulled a copy of the Aeneid off the shelf of some quasi-Dantesque and Borgesian library, wherefrom to draw intention). So far as I can see, the fate of my readings is largely independent of such considerations, and I employ more or less indifferently the whole gamut of terms designating intertextual relations. ‘Largely’, not ‘entirely’: at least some of my Virgilian allusions to Plato cannot be comfortably rephrased in an impersonal idiom. So much the worse for readings which pretend to eliminate the author without remainder.

156

Robert Wardy

tips for Phaedrus’ speech. Instead, I am wondering whether exposure to Virgil’s reaction to Phaedrus could or should feed into our reaction to the Symposium, as we reread that primary text in retrospect—so long as we are on our guard against flagrant, question-begging circularity.³ To begin with Phaedrus. Since the objective is to analyse the interplay between Plato and Virgil, I shall not attempt even a partial interpretation of his speech for its own sake; the following fragmentary comments are strictly limited by relevance to the project in hand. Some methodological remarks. Neither the general run of readers nor the experts seem able to summon up much enthusiasm for his performance.⁴ Is this critical attitude warranted? Not at all obviously. First, on the very pro tem. concession that the speech is but paltry fare, we should nevertheless feel obliged to wonder: if the speeches are of variable quality not only for the trivial and inevitable reason that no such sequence could possibly be homogeneous, but rather also because Plato designedly composed them that way, what are the implications? One should fumble after something more illuminating than the good old chestnut that the speakers represent an erotic hierarchy, etc. But second, and absent the concession that Phaedrus is feeble, his disquisition on positive erotic and virtuous emulation and negative erotic shame before vice (178d1–179b3)⁵ is analysis of common mores of the highest order—that is, ‘analysis’ in the sense of lucid explication of some widespread Greek ideology, bearing comparison as such with much of the Nicomachean Ethics. Third, one must register what follows from the fact that what Phaedrus delivers is not a neutral explication, but rather an exhortation to cultivate erotic virtue as a phase in his encomium of ἔρως. We should resist the temptation to take at immediate face value Socrates’ subsequent bloody-minded dismissal of his predecessors’ statements as no more than a pack of meretricious lies (198d ff.): we are not entitled to the assumption that the character Socrates speaks for the author Plato. Rather, what we must bear in mind is the parodic and often self-parodic complexion of much symposiastic performance:⁶ which is to say, we must not neglect the possibility that Phaedrus’ speech is to be read as humorous quasi-exhortation, a clarion call to the acme of erotic valour on the part of a not conspicuously heroic individual. To take this possibility seriously is not to conclude that (much of) the Symposium is not to be taken seriously: that many of its participants’ assertions are anything but straightforward is consistent ³ On such uncustomary reconfigurations of ‘tradition’, see Hinds 1998: esp. ch. 4, ‘Repetition and Change’. ⁴ e.g. ‘Phaedrus confines himself to certain ingredients of standard encomia’ (Dover 1980: 90); ‘the main characteristic of Phaedrus’ contribution is perhaps the way it tries to be clever, without quite pulling it off ’ (Rowe 1998: 137). ⁵ I have the impression that αἰσχύνη, not φιλοτιμία, dominates the little essay—inhibiting embarrassment, not liberating rivalry, has pride of place: why should that be so? ⁶ This approach to the dialogue is most tellingly developed in Hunter 2004.

Virgil’s Sacred Duo

157

with their possessing serious as well as comic aspects. This strain of seriocomic (self-)parody has everything to do with why this dialogue has remained enduringly enigmatic—we shall see that the notoriously enigmatic tone of the Nisus and Euryalus episode is not unrelated to the systematic obscurity of the Symposium. Next some remarks concerning points of detail which will prove relevant to the Aeneid episode. First, Phaedrus blandly collocates state and individual as alike spurred to glorious accomplishment by ἔρως (178d2–4). In the light of the superabundance of lessons in Greek culture warning of how all too easily erotic passion can erode ἀρετή, and specifically the sociopolitical virtues, this optimistic thesis cannot be intended to pass on the nod, whether one thinks of the character Phaedrus’ intentions vis-`a-vis his co-symposiasts, or Plato’s intentions vis-`a-vis us. We can imagine the mock-solemn nods of acquiescence at Agathon’s party, or entertained smiles at suppression of destructive ἔρως (to return when Pausanias competes); and we can glimpse a Platonic subtext meant to unsettle our unthinking presumptions about what types of behaviour (true) ἔρως might, or might not, stimulate or inhibit. Second, Phaedrus is concerned to magnify the putative wholesome effects of ἔρως beyond anything φιλία, non-passionate, primarily familial attachments, might achieve (178d6–7, 179b8–c2): perhaps a further invitation to wonder whether the centrifugal, isolationist tendencies of the erotic couple really do harmonize so nicely with civic participation. Third, Phaedran ἔρως operates in the strenuous public arena of military spectacle: that is, the ethos is archaic, ἀρετή is seemingly confined to ἀνδρεία, courage, and what one sees and how one is seen are decisive for ethics (what was once dubbed ‘shame culture’). Therefore, fourth and inevitably, Homer has a prominent place in Phaedrus’ conception of ἔρως and its military accomplishments (179a7–b3). Are we breezily to accept that erotic inspiration really does suffice to make Homeric heroes of all us lovers (οὐδεὶς οὕτω κακός, 179a7)? No matter what mediocre stuff one is made of, erotic transformation issues in something like the best φύσις.⁷ According to the archaic, exhibitionistic standard, seeming is (as good as) being: hence the supposed adequacy and primacy of φιλοτιμία and αἰσχύνη in shaping behaviour. On this view, that ἔρως is an external, possessing, and inspiring force is a congenial idea. But reference to φύσις, real intrinsic nature, invites the philosophically inclined to wonder whether imposition and constraint could actually constitute virtue and vice, above and beyond profoundly affecting states of character. None of this passes muster unproblematically; and we need not suppose that any of it is meant to appear unproblematic, in the fiction, as Phaedrus performs for the other characters, or as we read Plato. ⁷ Similarity is not identity. Might this be a very compressed reference to the debate over the wellsprings of virtue? Might we read Phaedrus as tacitly conceding that nature takes pride of place, but slyly suggesting that ἔρως supplants knowledge and practice as a producer of ἀρετή?

158

Robert Wardy

Fifth, what are we to make of the coy reference to the Theban Sacred Band—if that is what it is? ‘So were there some contrivance to bring about a city or army of lovers and their beloveds, there is nothing that would enable them to run their own city better than by abstention from all shameful things and mutual competition for honour; and were such men really to fight beside one another, although few in number, they would conquer virtually all of humanity’ (178e3–179a2). That Phaedrus’ seeming reference is cast as a future less vivid conditional⁸ has impressed scholars who believe in the possibility and value of dating the composition of Platonic dialogues; the routine inference is that the Symposium must have been written prior to the formation of the Sacred Band.⁹ Nothing whatsoever follows. Sometimes it suits Plato to avoid anachronism; and sometimes he indulges in it flamboyantly. On what basis can we be confident that in this instance employment of the future less vivid conditional does not result from avoidance of anachronism? Even if we accept that 378 is a firm historical reference point, the dating exercise seems to be at an impasse; but it may well be that the reference point is illusory, so that the exercise is anyway moot. David Leitao has pressed the case that ‘… the historicity of an erotic Sacred Band rests on the most precarious of foundations’.¹⁰ The marshalled grounds for extreme scepticism are very impressive,¹¹ and would make Plato’s grammar unremarkable (how else to refer to a fiction?). True, accepting Leitao’s conclusions would not mean that geneticists need throw in the towel at once: they retain the terminus post quem of 384 furnished by Aristophanes’ historical allusion to the Spartan division of Mantinea in 385 (193a2 ff.), which might somehow be coupled to an estimate of the time when the legend of the Sacred Band gained currency (although promulgation of a legend would not be a punctual event, to put it mildly). So let us suppose that Leitao is right, that the Sacred Band was always an ideological construct rather than a historical fact: what might that suggest about Phaedrus’ invocation of this ideal? The answer may be contained in his supposition that μηχανή (178e3), some ‘contrivance’ or ‘device’, would be required for the formation of this erotic troop: why should that be necessary? Discovery or use of a μηχανή is called for only when there is an obstacle to ⁸ εἰ οὖν μηχανή τις γένοιτο … οὐκ ἔστιν ὅπως ἂν ἄμεινον οἰκήσειαν … καὶ μαχόμενοί γ¯’ ἄν … νικῷεν ἄν … ⁹ ‘Phaedrus speaks of an army composed of ἐρασταί and their παιδικά in extravagant and entirely hypothetical terms; but there are reasons for thinking that the ‘‘sacred band’’ of Thebes, composed in just such a way, was formed in or very soon after 378’ (Dover 1980: 10). Dover infers that the terminus ante quem for composition of the dialogue is 379. ¹⁰ Leitao 2002: 143. ¹¹ But James Davidson, for one, is not impressed: ‘predictably, the existence of the Sacred Band has been questioned on a priori grounds by David Leitao … . The evidence he must discount, however, much of it from contemporaries, is overwhelming’ (Davidson 2005: 17). Davidson does not explain the surprising characterization ‘a priori’; and since his piece is not written for the specialist, the controversy cannot be pursued profitably. Doubtless the (a)historicity of the Sacred Band will remain contentious.

Virgil’s Sacred Duo

159

overcome or circumvent, especially if the problem is severe;¹² why should the recruitment of the couples be especially problematic? True, Phaedrus does think that some lovers are better than others; however, since he does not lay special emphasis on discriminating between valid or superior and counterfeit or inferior ἔρως (contrast e.g. Pausanias), it is prima facie surprising that any ‘contrivance’ is called for—why not enlist recognizably pugnacious pairs? But not if we take proper account of the disjunction ‘city or army’ (178e4), for then the proposal is indeed radical. ‘City-or-army’ is a virtual hendiadys, given the complete identification of adult male citizen with citizen-soldier in the classical πόλις. There are, accordingly, two possibilities for explaining why a μηχανή is required. Phaedrus does not merely envisage an elite unit within an ordinary army: the lovers would not only constitute the entire army, but also conduct the affairs of the city (178e5). So perhaps the idea is that citizenship itself would be conditional on erotic status, that non-lovers would be debarred from civic participation, or even denied citizenship outright—it is not hard to see that such an ideal is unrealized, and likely unrealizable. An alternative extension of this possibility would not merely exclude women from participation in public, political life, as in Athens, but also deny them, as well as male non-lovers, citizenship. Phaedrus evinces none of the misogyny expressed by other speakers—for him, Alcestis is a heroine. Nevertheless, his focus on a paradigm of military virtue inevitably elevates the masculine over the feminine;¹³ the heroism of Alcestis is noteworthy precisely because it demonstrates that the transformative potential of ἔρως suffices to make this woman more courageous than the unmanly Orpheus (ἀλλὰ καὶ αἱ γυναῖκες, ‘and even women’, 179b5). Since, apart from the oddity of Alcestis and Admetus, Phaedrus throughout his speech is focused on paederastic couples, one might well conclude that they alone would make up his citizen-soldiers. On this reading, a ‘contrivance’ is again certainly required, for this is the ideal of a City of Men: women must discharge their biological function, but citizenship will be (impossibly?) severed from the circumstances of birth; it will no longer be the (defeasible) inheritance of the offspring of two citizen-parents.¹⁴ ¹² e.g., it is the term employed to describe Zeus’ shifting of the genitals to initiate sexual reproduction in Aristophanes’ myth (191b6). ¹³ Not inevitable, of course, for another speaker in another dialogue, who also espouses a revolutionary ideal of military politics, but includes women by revamping their education; and, unlike Phaedrus in the Symposium, Socrates in the Republic is nothing if not explicit about why a stupendous μηχανή is required! ¹⁴ Rowe is judiciously circumspect about evidence for dating composition, and also pays attention to the disjunction: ‘ ‘‘an army of lovers’’ is usually taken as a covert, anachronistic reference to the Theban ‘‘sacred band’’, formed on just such a basis in the early 380s; but the combination ‘‘city or army’’ might more readily have recalled the Spartan example—itself partly reflected in the army regulations of the best city at Republic 468b–c’ (Rowe 1998: 138). I resist the Spartan connection, since it neither accounts for the μηχανή, nor allows for the possibility that Phaedrus might be hinting at a state in which women are invisible, as it were (Spartan women were notoriously visible). Leitao too speculates about ‘whether an idealised image of Sparta does not also lurk behind the proposal made by Plato’s Phaedrus’, but infers that ‘the paederastic bond, then, may have been

160

Robert Wardy

Sixth and last, Phaedrus’ catalogue of self-sacrificial erotic paragons and cowardly failures. On Alcestis, I cite Richard Hunter’s interpretation, since it contrasts very cleanly with mine: ‘that Alcestis’ willingness to die in her husband’s stead is a paradigm of the power of ἔρως to induce self-sacrifice is at least not the most obvious interpretation, given the usual presentation of that emotion … ; that she surpassed her parents-in-law ‘‘in φιλία because of her ἔρως’’ (179c1–2) merely calls attention to the manipulation. φιλία and ἔρως may, of course, coexist, even in a marriage, but Phaedrus’ lack of explanation or theorising … induces scepticism.’¹⁵ Does it? Phaedrus simply isn’t in the business of theoretical explanation; with a vengeance, that task falls to later members of the speech sequence. If we suppose the function of his speech within the economy of the dialogue as a whole to be deliberate problematization (to which other speaker does Diotima refer so frequently in her tacit critique?), then what should be induced is aporia, not impatient scepticism, and there is no ‘manipulation’. Despite Hunter’s concession, Phaedrus is deliberately courting paradox: if not incompatible, ἔρως and φιλία remain in potential conflict, given that their objects will at least normally be distinct, and so capable of making claims at best difficult to reconcile (one obvious sense in which erotic passion is destructive);¹⁶ and one imagines that the reunion of Alcestis and Admetus with his parents may not have been an occasion abounding with feelings of happy, comfortable φιλία. The complaint ‘why an action which would have led to Alcestis’ permanent separation from her husband, Admetus (contrast Achilles’ deliberate death), should be an act of ἔρως is left quite unclear’¹⁷ is much to the point. I presume that the remark about Achilles explains why Hunter is provoked, since it might indicate that the train of thought is something like this: ‘the ἐραστής, the lover, necessarily seeks possession of, or some manner of union with, the ἐρώμενος, the beloved; thus whatever prompted Alcestis to volunteer to lose Admetus, it couldn’t have been ἔρως’. Phaedrus’ self-imposed duty is certainly not to praise ἔρως solely within the vague limits of ordinary assumptions: more or less implicit serio-comic paradox is to be anticipated, to take a leaf from Hunter’s book. Later speakers at the symposium will have a great

chosen as a metaphor for civic cohesion not because of the hierarchical nature of the bond, but in spite of it’, simply because of Spartan military prowess (Leitao 2002: 161). This cannot be right. Phaedrus does nothing to make lover and beloved symmetrical; indeed, their marked asymmetry remains an unresolved tension in his account, and is an important contribution to the dialogue’s further agenda. Were we to accept Leitao’s thesis, we should have to convict Phaedrus of gross incompetence, since asymmetrical paederastic relations would figure inappropriately and faute de mieux in his sociopolitical vision. ¹⁵ Hunter 2004: 40. This highly uncharacteristic lack of generosity is somewhat mitigated on the following page: ‘later examples of self-parody from the other guests may, however, incline us to give Phaedrus the benefit of the doubt’. ¹⁶ This is woefully inadequate, but tracing out the various popular and Platonic conceptions of how ἔρως and φιλία might be related is far too large and complex a task to pursue here. ¹⁷ Hunter 2004: 40; cf. Rowe 1998: 139.

Virgil’s Sacred Duo

161

deal to say about what ἔρως might rightly or wrongly, optimistically or forlornly, desire; and about what successful lovers might be obliged to renounce. On Orpheus, the effeminate artiste, and Achilles, the embodiment of erotic perfection. The former’s culpable softness is made manifest because he contrives to enter Hades alive—this is presumably grounds for condemnation rather than praise for ingenuity because he was afraid to die (179d5–6: and otherwise would have abandoned his wife?). The latter’s unadulterated heroism emerges from the purity of his self-sacrificial intentions; since Patroclus is already lost to him, Achilles’ death can only be added to that of his lover in eventual joint extinction, a protracted Liebestod.¹⁸ But, of course, all readers of Homer know that what Achilles gave up was a lengthy, inglorious life in exchange for immortal fame. Taken as a group, the exempla yield a strange, unsettling pattern. Divine intervention may have returned Alcestis to the living and spared Achilles Hades; but embracing death willingly remains the criterion of erotic authenticity. The exempla associate ἔρως with violent conflict likely to culminate in a glorious death, paradoxically using voluntary separation and selfdestruction to illustrate commitment to erotic union. Within Phaedrus’ speech, his suggestively problematic donation to the programme of the Symposium, this all remains opaque to a degree; and ‘problematic’ will be our keyword in forging links with the Aeneid, as we assess Virgil’s thematic inheritance from Plato. The terms in which Virgil couches his introductory description of the erotic relationship between Nisus and Euryalus unmistakably label them as a classic ἐραστής / ἐρώμενος couple: the former is acerrimus armis (176:¹⁹ ‘this formidable warrior’, in West’s translation), that is, a military adult, and a successful one to boot; the latter, his comrade, is iuxta, ‘together’ with him (179; cf. 223). Euryalus, second to none in looks, is ‘a boy showing the first signs of youth’s down on his unshaven cheeks’ (181, Hardie’s translation)²⁰—that is, the ἐρώμενος non pareil. Cf. gratior et pulchro veniens in corpore virtus, V 344 (‘the manly spirit growing in that lovely body’, West’s interesting version): the modest Euryalus is that seductive Symposium package, physical beauty incorporating psychic beauty. Or is this mere wishful thinking on the part of a Symposium fanatic? They are a pair; they are more than good friends; one is senior, the other junior. These features suffice to establish that Virgil is using the model of Greek paederasty; that he ¹⁸ ‘Homer does not portray the mutual affection of Achilles and Patroclus as a homosexual relationship, but it was so interpreted in classical times’ (Dover 1980: 94; he further conjectures that Aeschylus in the Myrmidons may have been responsible for the innovation). The scholarly communis opinio is indeed that an erotic reading of their relationship in Homer is anachronistic; but we should nevertheless not be over-hasty to categorize such passionately intense male bonds, especially in a military context, as either sexual or not (cf.: who knows (all of ) what went on between Butch Cassidy and the Sundance Kid?). True, the Iliad is not explicit; but some things go without saying. We should keep an open mind on the antiquity of an ‘erotic’ Homer. ¹⁹ Unless otherwise indicated, all Virgilian references are to Aeneid IX. ²⁰ Hardie 1994: 107.

162

Robert Wardy

does so to characterize a military couple narrows down the field of possible sources to Phaedrus’ speech and a few others. In Theocritus XIII, that Heracles and Hylas are ἐραστής and ἐρώμενος could not be more explicit (l. 6; Heracles is assiduous in educating the youth, ll. 8–9), and they are usually inseparable (l. 10—until disaster strikes). Furthermore, the loss of Hylas prompts Heracles to abandon the Argo, much to the derision of his more orderly shipmates (ll. 70–3); thus here too Virgil might have discovered the topic of paederasty working with or against an epic endeavour. Similarly in Apollonius, in whom Virgil is so deeply versed and to whom he recurs so often, Heracles’ and his beloved Hylas’ introduction, setting off together on their adventure, seems reminiscent of the Virgilian text (cf. Argonautica I, 131–2 with Aeneid 179–80).²¹ But I feel no inclination to argue for the Symposium as Virgil’s unique inspiration; it is perfectly typical of his learned re-creations that, in an unparalleled manner, they allude to, combine, or play off different sources. My argument will seek to assemble a gradually strengthening cumulative case for the prominence of Plato. From the outset, complications set in. Nisus is guarding the gate—that is in fact the very first thing we learn about him in this book. During the siege, there is no more responsible position to occupy (nec non trepidi formidine portas | explorant, 169–70: ‘anxiously checking the gates’, West). But, in the course of that same sentence, Nisus the mature and disciplined guardian morphs back into a ranging hunter.²² Wild hunter or trained soldier?²³ So, fittingly enough, we meet this liminal character in a gateway; and in this book, gates are not good for Trojans (think ahead to Pandarus and Bitias). As for Euryalus, I am much taken with Hardie’s collocation of Iliad XXIV 348, ‘[a young man] with his first beard, when youth is at its most attractive’, the locus classicus of the sexy down topos, with the warning that ‘the first down may also be regarded as marking the end of boyish beauty’:²⁴ whether this can or should be squared with Euryalus’ bloom of youth back in book V²⁵ is another matter. That is, perhaps one leaves the question of whether Euryalus is blooming or about to fade unresolved, so that his status, no less than that of Nisus, remains shifting and unstable. Or might one hazard the guess that a delectable, au point παιδικά is curiously like a pear (with a nod in Eddy Izzard’s direction):²⁶ there in the bowl for what seems forever and ²¹ Richard Hunter kindly suggested that I inspect the Hellenistic sources. ²² Nisus erat portae custos, acerrimus armis, | Hyrtacides, comitem Aeneae quem miserat Ida | venatrix iaculo celerem levibusque sagittis (176–8). Both Fitzgerald and West spoil the effect by splitting up the sentence. ²³ On this and much else one should consult Philip Hardie’s commentary. ²⁴ Hardie 1994: 107–8. ²⁵ Euryalus forma insignis viridique iuventa (V, 295: ‘Euryalus standing out for the bloom of his youthful beauty’, West). ²⁶ And in Seneca’s: gratissima sunt poma, cum fugiunt; pueritiae maximus in exitu decor est (Ep. XII 4). Alcibiades is, perhaps, the exception which proves the rule: περὶ μὲν οὖν τοῦ κάλλους ’Αλκιβιάδου οὐδὲν ἴσως δεῖ λέγειν, πλὴν ὅτι καὶ παῖδα καὶ μειράκιον καὶ ἄνδρα πάσῃ συνανθῆσαν τῇ ἡλικίᾳ καὶ ὥρᾳ τοῦ σώματος ἐράσμιον καὶ ἡδὺν παρέσχεν (Plut. Vit. Alc. I 3).

Virgil’s Sacred Duo

163

a day, hard as oak or rock; but if one glances away distractedly for what seems no more than an instant, the fruit immediately turns to mush? Which is to say that maybe a boy’s perfection is so fleeting that there is all too little to separate under- from over-ripe, so that downy cheeks will convey conflicting signals.²⁷ Line 182 should be dwelt on carefully: his amor unus erat pariterque in bella ruebant (‘they were one in love, and side by side they used to charge into battle’, West). This union of love with heroic valour is, I contend, Phaedrus’ evocation of the Sacred Band ideal in miniature, synthesized in the high Virgilian manner with Homeric and other materials; the two halves of the symmetrical, evenly balanced verse give us ἔρως matched by, or expressing itself in, superb military impulse. Notice that this line has them together in love and war—indeed, together in responsibility, standing watch at the gate (183)—for now. ‘Probably ‘‘common love [for each other]’’, rather than ‘‘shared passion [for war]’’, as Servius Dan. and Henry take it’, says Hardie ad loc.²⁸ No. Not only is there no need to make a stark either/or choice; the inextricable involvement of love and war for Phaedrus should encourage us to concede Virgil his careful ambiguity.²⁹ A sceptic about Platonic influence³⁰ might insist that the famous opening gambit Nisus ait: dine hunc ardorem mentibus addunt, | Euryale, an sua cuique ²⁷ A further speculation about Euryalus’ transient bloom: is the striking double flower simile (purpureus veluti cum flos succisus aratro | languescit moriens, lassoque papavera collo | demisere caput pluvia cum forte gravantur, 435–7) saying that he loses his ὥρα, his flos, only at the moment of death? Or might it be that a disconcerting aspect of the Liebestod is that he is most attractive in death? ‘Bion’s Adonis and the Virgilian figures that reflect him present less a contrast than a morbid superimposition of love and death, with the elements of the horrible and macabre enveloped in the sensual’ (Reed 2004: 32). ²⁸ Hardie 1994: 108. ²⁹ As Hardie then acknowledges: ‘but the reference of the word amor is often problematic in Virgil; in the present episode erotic and martial passion are difficult to disentangle’ (1994: 108). For myself, I should prefer to say ‘deliberately impossible to disentangle’. ³⁰ Makowski is no sceptic: ‘it will be seen that the two episodes have so many affinities with the Platonic dialogue, particularly the speech of Phaedrus, that this work must be taken into account as influencing Virgil’s conceptualisation of Nisus and Euryalus at every level’ (Makowski 1989: 4). So far, so good. But Makowski ignores the dialectical relations obtaining between the speeches so as to extract a sort of ‘identikit’ portrait of paederasty (p. 4–5), and is blind to the possibility that Virgil might have been a sufficiently acute reader of Plato to appreciate that it is hardly the case that everything (anything?) Phaedrus claims is endorsed Platonic doctrine; and his defensive ardour for Nisus and Euryalus is set at so high a pitch that he rejects outright any suspicion that some of their behaviour is, at best, questionable. His tactics are, first, to insist against the prudes, with obvious success, that Nisus and Euryalus are not just friends; and then to whitewash the pair vigorously. But, as with the recommended diffidence towards the Homeric Achilles and Patroclus (see n. 18), the assurance that Nisus and Euryalus are erotically attached cannot definitively settle on a fixed character for the amor between the two, as if clear blue water separates camaraderie from homosexuality. On the other hand, Makowski’s gloss on Nisus’ question is valuable and thought-provoking: ‘we, of course, know in retrospect that the question, with its vocabulary of passion (ardor) and desire (cupido), refers to the longing of Nisus to engage in some action, but as they [the lines] stand (that is, before the hearer or reader gets to line 187) they are ambiguous, especially after the introduction of the two as lovers. I would argue that Virgil has consciously made the words ambiguous, or to put it another way, that he has made the words function on two levels, those of ἔρως and military ἀρετή, and that they apply equally to love relationship and to the

164

Robert Wardy

deus fit dira cupido? (184–5: ‘this urge to action, do the gods instil it in our minds, or is each man’s fierce desire a god to him, Euryalus?’, after Fitzgerald)³¹ taps exclusively into the hoary Epic tradition of puzzling over human and/or divine motivation.³² But it would be typical Virgil to be doing that and simultaneously aligning ardor and cupido with the inspirational ἔρως avid for witnessed glory which Phaedrus adduces for his encomium. In this instance, one can have one’s generic cake and eat it: for remember Phaedrus’ assurance that ‘just as Homer says, the god ἔρως inspires some heroes with force’³³—that is to say, Plato has already obligingly made the Homeric connection for us.³⁴ Plato’s engagement with Homer is anything but superficial; and it is blindingly obvious that this is also true of Virgil’s. Virgil’s recognition and appreciation of the philosopher’s manifold Homeric interests would provide one very sound reason for his thoughtful reworking of Platonic texts. And there are proper grounds for discerning a tighter affiliation with the Symposium. Commentators on the dialogue routinely puzzle over what the ἔρως praised in the speeches (really, ultimately) is: god, δαίμων, personification of psychological force or condition? Nisus speaks initially of a psychological impulse, albeit one instilled by autonomous gods (or is that merely a fac¸on de parler?); but in the formulation desire for an ἀριστεῖα …. Furthermore, as in the Symposium, where there is a sustained ambiguity between ἔρως the divinity and ἔρως the love between two men which, Phaedrus tells us, inspires them to ἀρετή, so here in this passage Virgil employs the word cupido with the same ambivalence’ (ibid. 8–9). ³¹ The semantics of dira cupido are difficult, since the connotations of dira are often, or even usually, negative. True, the dira cupido for rebirth of souls languishing in the underworld (VI 721) might seem an extreme, but blame-free longing—although even here that the discarnate are miseris injects a note of doubt: are they neutrally ‘pitiable’, or culpably ‘wretched’ for entertaining so perverse a desire? In any case, when the Sibyl castigates the dira cupido of Palinurus (VI 373), the expression must be pejorative; as also in Georgics I. 37, nec tibi regnandi veniat tam dira cupido, denying that the to-be-deified Augustus would ever plump for Hell as his special domain (‘dira has an element of ‘‘shocking’’ as well as ‘‘intense’’ ’ (Mynors 1990: 9) ). Perhaps the best way to respect the semantic ambivalence is to read Nisus’ question ironically: as far as he and Euryalus are concerned, they are invaded by a desire whose fierce intensity is only natural, since it is for aliquid magnum; but the diction sets off alarm bells for the reader, who wonders whether the great thing will turn out to be an enormity. ³² e.g. οὐκ οἶδ’ εἴ τίς μιν θεὸς ὤρορεν ἠὲ καὶ αὐτοῦ θυμὸς ἐφορμήθη ἴμεν ἐς Πύλον, (Hom. Od. IV 712). ³³ ἀτεχνῶς, ὃ ἔφη ῝Ομηρος μένος ἐμπνεῦσαι ἐνίοις τῶν ἡρώων τὸν θεόν (179b1–2). ³⁴ Not that, for either Plato, as we have seen, or Virgil, the Homeric connection is simple. Hardie’s r´esum´e of Virgilian deviations from the Homeric exemplar is very useful for our purposes: ‘Nisus’ proposal combines elements from both Nestor’s speech at Iliad X 204–17, asking for volunteers for the night expedition into Trojan lines, and Diomedes’ reply at 220–6. But there the old counsellor asks for others to volunteer and promises both renown and material reward; Diomedes is keen to go, but asks for a companion. Nisus conceives his great adventure himself, and thinks of reward only for his constant companion who is, exceptionally, not to accompany him on this mission’ (Hardie 1994: 108). Throughout book IX, as elsewhere in the Aeneid, whenever Virgil’s Trojans mirror Homer’s Greeks, the reflections are always meaningfully distorted. In this instance the rewriting sets up questions pertinent to the themes inherited from Plato: in what circumstances can or should Nisus and Euryalus remain together? Who hopes to get what from their nocturnal foray?

Virgil’s Sacred Duo

165

of the second disjunct, the desire itself is divinized—as it were. ‘Dira and Cupido when capitalised do of course both become divine beings.’³⁵ Of course. ‘When and when not to capitalize ἔρως?’, agonizes every editor, if not quite every reader, of the Symposium. Both Plato and his student Virgil prefer to leave us guessing. A narrative of pairing (iuxta, 179), twinning (pariter, 182), singular versus plural (unus, 182, and the di —deus slide, 184–5) quickly establishes itself. The binary story-line (each of the names Nisus and Euryale occupies the beginning of 184 and 185, and the arming scene is doubled (303–7) ) will eventually strain against fissiparous movement: when Ascanius is falling over himself to shower our daring duo with prodigal gifts, that bina, geminos, duo, and bis (263–72) are all for Nisus, Euryalus asking for unum (284), can be read as a quiet anticipation of their final, fatal splitting.³⁶ (Blame such complications on the grammar: if Latin had a dual, Virgil could have written another Iliad; I mean, other than the one he did.) At this juncture, I claim no more than that these structures are evocative for the reader who has meditated on erotic (dis)union in Phaedrus’ exempla. The reaction of the beloved: obstipuit magno laudum percussus amore | Euryalus (197–8: ‘Euryalus was overcome, pierced to the heart with a great love of glory’, West). But whose laudes? For himself? For Nisus? For them both, despite Nisus’ jarring segregation of prizes for Euryalus from his own fama (194–5)?³⁷ Hardie quips: ‘when it comes to heroics Euryalus is a copy-cat’.³⁸ As he should be, since the ἐραστής inculcates virtue by setting before his παιδικά a proper model for emulation. Euryalus’ reaction to the threatened desertion bears this out: comrades are to share in major deeds, dangers are not to be confronted alone; and with his reference to the doughty Opheltes’ acknowledging him as his son and subsequent military experience, Euryalus depicts himself as the good soldier in the making, ready for greater things, if Nisus will lead him and show the way (199–204). Should one detect any disquieting note? est animus lucis contemptor (205: ‘a spirit that disdains mere daylight’, Fitzgerald). The sortie will require the cover of darkness (sed celerare fugam in silvas et fidere nocti, 378), and light (lux inimica, 355) will prove fatal to Euryalus; for what Nisus proposes is not a regular daytime military operation, wherein he might be evident to his beloved. Phaedrus’ condition for exemplary behaviour from military lovers, that they be constantly under each other’s eyes, cannot be fulfilled. ³⁵ Hardie 1994: 109. ³⁶ To these organizing structures we should add the polarity group versus individual, so convincingly and elegantly exposed in Saylor 1990. ³⁷ ‘Either ‘‘[Euryalus’] love for [Nisus’] heroism’’ … , or ‘‘love, desire for praise [for Euryalus himself ’’ … But 182 amor unus should remind us not to discriminate too finely between aspects of the couple’s passions’ (Hardie 1994: 111). I am not sure that this is quite right. Given that Nisus is making as if to separate himself from Euryalus, and thereby to monopolize any fama which might accrue from the exploit, perhaps the nuance should be that just for now Nisus’ threatened withdrawal forces a choice over which we can only hesitate. With his unresisting capitulation to Euryalus’ protest, the problem disappears—temporarily. ³⁸ Hardie 1994: 111.

166

Robert Wardy

Nisus’ fantasy of a return in triumph (ovantem, 208) reminds us of his state just before he tumbles in the foot race (hic iuvenis iam victor ovans, V 331); and now we should briefly draw in some materials from the earlier book. The problem was that he was not yet a victor —he just thought he was. In book V, a physical slip results in nothing worse than getting messed up, but nevertheless receiving a consolation prize, as Aeneas soothes ruffled feelings; in IX, a strategic slip results in separation and horrific death, in the absence of Aeneas—or rather, as the direct consequence of failure to seek him. And it is not as if they can really be permanently, stably together, any which way, even in the burlesque competition of book V. Had Nisus not tripped, he would have won, leaving Euryalus defeated in the rear; as it is, by tripping up Salius, he puts Euryalus in the winning lead, and stays behind. But there is no scenario in which Nisus and Euryalus can share the victory (I am excluding a dead heat, of course). As so often in the Aeneid, lovers find it enormously difficult to occupy the same space at the same time. Nisus’ cheeky loser’s complaint that he has been left empty-handed claims that fortuna inimica has unfairly robbed him of his just deserts (V 355–6). In the idyllic context of that book, this impudence raises a smile from Aeneas (V 358); but in the dark context of book IX, such fatal ignorance of the hazards of athletic and military competition, of fortune and of fate, has inevitable consequences. It is common ground, of course, that the mission is a failure; controversy smoulders over the nature of that failure, pathetic or noble, shameful or poignant and glorious. But exactly what was the mission’s intended goal in the first place? Nisus’ words to the Trojan leaders are difficult to construe and interpret: si fortuna permittitis uti | quaesitum Aenean et moenia Pallantea, | mox hic cum spoliis ingenti caede peracta | adfore cernetis (240–3). The majority construe quaesitum as a supine³⁹ and ‘understand’ nos as the subject of adfore:⁴⁰ but who are ‘we’? Lennox has plausibly argued that reference to Nisus and Euryalus alone and the havoc which they in fact go on to wreak produces glaringly incompetent rhetoric;⁴¹ he recommends either ‘understanding’ Aenean as the unexpressed subject, or—his preference—supplying nos, but giving it ‘wider reference, to ³⁹ Courtney advocates the transposition 240–242–243–241–244, on the grounds that ‘where the manuscripts offer 241, it presents a use of the supine that is wholly unexampled in Latin literature’ (Courtney 1981: 17–18). We should not be intimidated—this is Virgil, after all; in any case, the transposition would not scotch our problem. ⁴⁰ ‘If you let us take advantage of this to find our way to Aeneas and Pallanteum, you’ll see us back with plunder before long, and slaughter done’ (Fitzgerald 1984); ‘if you allow us to take this opportunity to go and look for Aeneas and the city of Pallanteum, you will soon see us coming back laden with booty and much slaughter done’ (West 2003). ⁴¹ ‘And consider the audience themselves; what do they want? Of course it is that a message should reach Aeneas and that he should come and relieve them. They are not interested in the safe return of Nisus and Euryalus, except in so far as it would mean that Aeneas had been apprised of the situation. Still less are they interested in any booty which the two could win … . Could we really expect the leaders not to notice, or fail to admonish, such a glaringly irrelevant and reckless statement of intent?’ (Lennox 1977: 338).

Virgil’s Sacred Duo

167

include Nisus, Euryalus, Aeneas and the allies which he has gone to seek’.⁴² For Lennox this establishes that Nisus has no antecedent or subsequent interest in irresponsible pillage: Euryalus alone bears responsibility for the fiasco.⁴³ This will not wash, although I agree with his assessment of the rhetorical impropriety of nos with narrow reference. I have already commented on Nisus’ surrendering all material rewards to Euryalus in his original proposal. In his exaggerated and over-enthusiastic reaction to the plan, Ascanius pledges to reward Nisus with all manner of fabulous prizes (263–74); Nisus is silent, but we presume he acquiesces. In marked contrast, Ascanius offers Euryalus no prizes, but rather the prince’s future companionship (275–80); to which Euryalus replies with a single petition, that his mother’s security be guaranteed. Thus the prizes, initially associated with Euryalus alone, seem to become attached to Nisus during the audience; and finally, during the sortie, Euryalus will despoil a Rutulian corpse, leading to catastrophic discovery.⁴⁴ I conclude that we cannot exactly identify the mission’s goal. Doubtless Ascanius et al. ‘understand’ a wide reference for nos; but if sufficiently impressed by the way in which the prizes float about, we readers are at liberty to leave the scope of nos indeterminate, whatever beliefs or intentions we ascribe to Virgil’s characters.⁴⁵ However that may be, it is more important to recognize that the interpretation of 240–3 has attracted so much attention because many have deprecated pillage as such, undisciplined barbarism unworthy of proper soldiers; the slaughter of the enemy has not attracted this opprobrium. But if the expedition takes us into epic territory, then slaughter-and-spoliation is a single Homeric activity. Other defenders of Nisus and Euryalus, alive to this very consideration, have cited it in exculpation of what, to their detractors, has seemed inexcusable behaviour. This defensive manœuvre meets with at best limited success. Throughout, Virgil deliberately, anachronistically ‘Romanizes’ many of his descriptions of the ⁴² Lennox 1977: 338. ⁴³ ‘He [Nisus] would have succeeded in his mission were it not for the excess of his companion, his bosomlove, whose desire for trinkets brings them both to disaster’ (Lennox 1977: 341). This phrasing reduces the epically plundering, if blundering, Euryalus to something like a silly girlfriend, childishly besotted with ‘trinkets’, who ruins the show by entangling the nobly rational Nisus in a fatal conflict of duties. ⁴⁴ Just before Euryalus grabs the belt and helm which will give him away, glimmering in the moon’s lux inimica, Nisus calls him off, sensit enim nimia caede atque cupidine ferri (354): what does this mean? ‘For he noticed that Euryalus was being carried away by bloodlust and greed’ (West 2003); or ‘for he saw his friend carried away by slaughter and lust for blood’ (Fitzgerald 1984)? ‘caede atque cupidine is best taken as an example of hendiadys (rather than understanding cupidine as ‘‘lust for booty’’ … ), but the separation of the two elements is more than a merely stylistic device: the experience of killing creates the desire for more. The cupido of 185 for heroism and fame has degenerated into a crazed lust for blood’ (Hardie 1994: 137). Once more I would prefer to hesitate: we know that Euryalus is indulging himself excessively, but in what? Virgil prefers that we not be sure what to think. ⁴⁵ Reading ad rather than et in 241 and re-punctuating, the ‘understood’ subject of adfore becomes quaesitum Aenean ad moenia Pallantea. This too would absolve Nisus of rhetorical ineptitude, but it remains preferable to retain a nos of uncertain scope as the subject.

168

Robert Wardy

Trojans;⁴⁶ the presence of embryonic institutions, importing as they do a host of ‘civilized’ conventions, attitudes, and inhibitions, effectively problematizes the epic material, even before Plato is added to the composition. Not that this can be a decisive consideration: until the Latins have been bested, only to absorb their Trojan conquerors, civilization, especially of an Asiatic origin, is bound to appear at a serious disadvantage in military conflict. Thus the excuse that Homeric mores dictate that going after gold is a perfectly acceptable side activity for heroes on the rampage does get somewhere—although those who believe that the participants in the Doloneia, the Iliadic ancestor of our adventure, behave atypically and disturbingly by Homeric lights,⁴⁷ will not concede that it is a precedent of unproblematic nocturnal heroism. When one takes proper account of Virgil’s generic blending and occasional confrontation between the ἔθη which the genera variously represent, that excuse cannot get very far: as is universally recognized, successful Virgilian fighters are in transition, moving beyond the (or a) Homeric world-view. When all is said and done, any time spent dispatching Rutulians is time wasted, if the paramount aim is to reach Aeneas. The d´enouement is bristling with curiosities. Imprudens (386): Nisus’ (even temporarily) forgetting Euryalus,⁴⁸ leaving him in the lurch, is just what a Phaedran ἐραστής would never do to his ἐρώμενος (and it might be appropriate to remember that Socrates on the battlefield does not desert Laches and Alcibiades—although he manifests no compunction at abandoning Aristodemus to mere social embarrassment). Now of course such a judgement might elicit the brisk retort that this simply goes to show that if we are indeed in epic territory, no or at any rate very little Platonism is discernible—certainly not anything so specific as Phaedrus. I reply that it goes without saying that his Sacred Band ideal has unmodified application only in a phalanx under the sun, not in a nocturnal guerrilla situation. Nevertheless, Nisus and Euryalus are playing at being Odysseus and Diomedes, to the utter neglect of their essential mission. Ascanius’ passionate reaction to Nisus’ proposal is studded throughout with asseverations that his father’s return is everything not only to him, but also to the Trojans, set to become Romans (see, especially, 257, 260–2);⁴⁹ the couple’s dereliction is a betrayal of the state and civilization to be. They should have been displaying conspicuous but, crucially, intelligent bravery to one another, for the sake of the commonweal; that would have been Phaedran. They are paederastic fighters; but ⁴⁶ e.g. ‘Virgil continues to suggest a more formal world than the Homeric: Nisus and Euryalus request an ‘‘audience’’, as if in the consilium of the Roman general’ (Hardie 1994: 116). ⁴⁷ Some ancient critics were moved to expunge the Doloneia from the authentic Homeric canon on the grounds of ἀπρέπεια, ‘unseemliness’. ⁴⁸ Contrast the wonderful line non tamen Euryali, non ille oblitus amorum (V, 334). ⁴⁹ Virgil firmly associates this filial concern with the boy’s maturation (esp. 311): a typical paradox, in that maturity-to-be reveals itself in still immature dependence on, and longing for, the father.

Virgil’s Sacred Duo

169

Nisus and Euryalus venture out into a dark environment quite unsuited for the performance of exploits to be apprehended and imitated. So they are Symposium lovers, but badly defective ones. Or are they? In the Aeneid, the Sacred Band shrinks into a sacrificial, Sacred Duo. Nisus not once, but repeatedly indulges his erotic devotion to Euryalus, at the expense of his military and civic duty, leaving the Trojans in fearful jeopardy (the Rutulians have not caught sight of him; even at the end, the way to Pallanteum remains open).⁵⁰ Phaedrus, so far from perceiving any conflict between ἔρως and πόλις (or anyway pretending to perceive none), assures us that coupled lovers under each other’s eyes and the gaze of other erotically motivated citizen-soldiers will further communal interests as they vie to cut a magnificent figure. This coincidence of private passion and public virtue is, of course, just that: it is only because the lovers are contained within a battle formation that what they (do not) wish for themselves motivates action likely enough to help their army. One might, therefore, consider excusing Nisus and Euryalus, those Phaedran lovers avant la lettre Romaine, as victims of (imagined) history not yet evolved to a stage wherein they might flourish: before the conquest of Latium, the conditions necessary for diverting ἔρως into useful courage do not obtain (in conclusion, we shall think about why there cannot be a Roman version of the Sacred Band). Or one might launch a much more aggressive defence. It is not that such love is out of step with the times. Rather, Virgil can see much more sharply than Phaedrus—or, outside the symposium, need not persevere in the pretence that amor, however organized or constrained, ever abets military and citizen virtue. On this interpretation, the latent target of Virgil’s criticism might be not Nisus and Euryalus as Phaedran lovers, but rather the Phaedran ideal itself, inasmuch as it remains doubtful whether the couple’s erotic fervour could be directed outward for the larger good in any historical circumstances. If anything like this might be on the right track, here we have found an opportunity to read Virgil reading Plato, reforming or deforming Phaedrus’ paederastic ideal. It is a critical commonplace that Nisus’ turning back in a doomed attempt to rescue Euryalus is an analogue of Orpheus’ descent to Hades in search of Eurydice. In which case this might be an elaboration of the implied negative judgement of the couple arising from the Symposium linkage (on the assumption that the viability of Phaedrus’ ideal is not currently in question), not only, or even primarily, a nod in the direction of the Georgics. At first blush, this might appear to be a most unlikely reminiscence: for if Phaedrus roughly upbraids Orpheus as μαλακός, an effeminate softie, incapable, unwilling, or both to sacrifice himself for Eurydice (or is this more symposiastic play?), Nisus, who is anything but soft, might seem to be a highly inappropriate stand-in for the cowardly artist. Yet, if his hard ardour results in abandoning Euryalus no less surely than the singer’s softness loses Eurydice, Virgil has his opening for further ⁵⁰ Cf. Bleisch 2001: 186–7.

170

Robert Wardy

ironic reflection⁵¹ on the mishaps of a Phaedran couple. According to Phaedrus, ἐραστής and ἐρώμενος are acting on their passion when they die rather than break free from the phalanx. In Virgil, this purported erotic unity of private and public splits into two, incompatible moments: fleeing and forgetting Euryalus, in pursuit of military duty; remembering Euryalus and returning—which is to turn one’s back on Aeneas. Erotic self-immolation never serves the public weal.⁵² When considering Euryalus’ reaction to Nisus’ plan, we asked ‘whose laudes?’; then it seemed best to hang fire. In the moments before the Rutulians put the boy to death, Nisus himself recommends an unexpected answer. He invokes their pity: nihil iste nec ausus nec potuit (428–9: ‘he neither ventured nor could have done anything’). Never mind that this pathetic invocation is transparently implausible: the glaring fact remains that with these final, impotent words, Nisus is hardly contributing to Euryalus’ post mortem laudes, as one would expect a Phaedran military lover to do (that might—or might not—be a job which Virgil usurps in the σφραγίς, as we shall see). What is he up to? Were he credible, the name ‘Euryalus’ would become at best an extinguished cipher, at worst an ignominious byword: if there is any good fama going, it will not be shared. But again there is a twist. To the notoriety of Orpheus Phaedrus opposed the celebrity of Achilles; and, in what we now recognize as a typical Virgilian oscillation, the final reunion of their Liebestod (444–5)⁵³ is a reminiscence not of shameful ⁵¹ Or, if I was correct to suggest that the handling of erotic togetherness and separation in Phaedrus’ exempla is already, and designedly, freighted with paradox, Virgil’s hommage is more of an expansion than a simple critique. ⁵² To recur to the Orpheus of the Georgics: perhaps one might read the Aeneid recasting as, in some respects, an artful inversion of its predecessor. In the earlier text, Orpheus restitit, Eurydicenque suam iam luce sub ipsa | immemor heu! victusque animi respexit (G. IV, 490–1: ‘Orpheus stopped; forgetful—alas!—and overcome he looked round at his Eurydice, almost in the light’; cf. Nisus: ut stetit et frustra absentem respexit amicum, Aen. IX 389). Nisus, forgetful of Euryalus’ situation, presses ahead and pays him no heed; Orpheus, forgetful of Proserpina’s injunction, stops stock still and unhappily attends to Eurydice, when obedient inattention would have returned her to the light, in the Georgics the realm of life, in the Aeneid the betrayer and destroyer of the lives of Nisus and Euryalus. And, of course, that the name Euryalus echoes the name Eurydice is itself an intertextual modification of the intratextual Eurydicen-Eurydicen-Eurydicen echo of G. IV 525–7. Also, since Orpheus in the Georgics is a (Virgilian?) artist, his sympathetic treatment there might be something of a riposte to the Orpheus of the Symposium, soft and cowardly ‘because he was a musician’ (179d5). Finally, let us also take note of the intriguing report that in earlier traditions, both Greek and Latin, Aeneas’ wife was named not ‘Creusa’, but ‘Eurydice’ (Paus. X 26 1; Enn. Ann. 37). Why the alteration? Virgil ‘had already written about another Eurydice in his fourth Georgic, and for him (and doubtless for his readers) the name could not be dissociated from the Orpheus-legend’ (Austin 1964: 289). I am not convinced; since Aeneas’ separation from Creusa is indubitably intended to read for us as a version of the loss of Eurydice, Virgil can hardly intend to ‘dissociate’ the narrative of Aeneid II from its prototype (Michael Reeve suggests to me that giving them the same name might have struck him for that very reason as too crude). Virgil, the allusive poet par excellence, puts together a set of variations on a theme; members of the set will exhibit different similarities to and divergences from the model. ‘Nisus is, in many ways, a type and anti-type of Aeneas; his return into danger to seek Euryalus parallels and alludes to Aeneas’ journey back through burning Troy to save Creusa, and to Orpheus’ journey through the Underworld to rescue Eurydice’ (Bleisch 2001: 187). ⁵³ Often in the Aeneid death seems the only way to avoid separation: Juturna bemoans her immortality, since she would rather accompany her brother Turnus in death to the underworld

Virgil’s Sacred Duo

171

Orpheus, but rather of splendid Achilles.⁵⁴ But the allusion is problematic. Phaedrus takes Aeschylus to task for representing Achilles as the ἐραστής, while, insists Phaedrus—in another symposiastic joke?—Achilles’ looks show that he must instead be the ἐρώμενος. But no matter how enigmatic or even impenetrable Virgil makes his episode, one thing, at least, is unambiguous: Nisus is the senior partner. Therefore he cannot be altogether like Phaedrus’ Achilles—which is the type of provocative inconcinnity which should by now be familiar.⁵⁵ At the beginning I stated that arguments for allusion rely on the accumulation of intertextual similarities productive of fruitful (re)interpretation: paederasty in warfare, (in)visibility, the (dis)harmony of private passion and public duty, Orpheus, Achilles … What is the tone of the σφραγίς or poetic ‘seal’ of the Nisus and Euryalus episode (446–9), a direct authorial address to the dead pair set at the highest emotional pitch?⁵⁶ Fortunati ambo! –why the felicitation?⁵⁷ In a unique reference (XII 880–1); Anna complains that Dido did not share death with her (IV 677–9); not to mention Aeneas’ desperate recklessness with his own life as he loses much of what he loves in book II. ⁵⁴ Well spotted by Makowski 1989: 13. ⁵⁵ To complicate an already baroque interpretation, could it be that Virgil is ‘correcting’, or at any rate qualifying, Phaedrus’ correction of Aeschylus on whether Achilles was ἐραστής or παιδικά? i.e., immediately before the finality of the Liebestod, its fusion and nullification, Nisus at least momentarily becomes the ‘junior’ partner—rôle reversal is, after all, not uncommon in erotic relationships. ⁵⁶ The σφραγίς has some peculiar, patterned echoes in the next book. First, there is the impassioned second-person address to Pallas, slain by Turnus (X 507–9), and matched by the apostrophe to Lausus (X 791–3), himself already a mirror image of Pallas (X 433–6). Second, interestingly, both these youths are likewise intimately paired up, although their couplings differ from the relationship of Nisus and Euryalus. Pallas is (nearly, would-be) attached to Aeneas; and this attachment is of a paederastic stamp, as amply attested by both Pallas’ reprisal of the Dido rôle (X 160–2), and Aeneas’ berserker furor in reaction to Pallas’ death, a` la Achilles wreaking vengeance for Patroclus (X 510 ff.). Lausus, of course, is attached to his father, Mezentius, not an ἐραστής. However, in the light of the nurturing, ‘paternalistic’ concern of ἐραστής for ἐρώμενος, this need not be an absolute, categorial difference; so we descry a Virgilian penchant for marked, heightened authorial—authoritative?—memorialization of fallen lovers, of destroyed, or relinquished, intimacy. ⁵⁷ ‘It is vital to understand this particular part of the intervention correctly. It has been misunderstood, in more than one way. It is not (for example) uttered with reference to Nisus’ and Euryalus’ prospective felicity in being hymned by Virgil … . it is a reference to Nisus’ and Euryalus’ felicity in love, in particular to their elegiac union as lovers in death … we must detect a further voice: it is invention and arrangement of a most provoking kind to offer this elegiac and romantic opinion at this point in an epic story’ (Lyne 1987: 235–6). If I am correct, Nisus and Euryalus are not lovers tout court, or lovers only by elegiac convention. They are also Phaedran lovers; and, if much less than perfect, according to Phaedran standards, then their erotic felicity is likewise defective, the felicitation accordingly ironic, and most likely based on their prospective immortalization, pace Lyne. Not that Plato’s presence in the Aeneid ousts Homer or ejects elegy from the text. But ‘one of the reasons for the difficulty in deciding on the meaning of the episode is precisely the conflation of a number of different genres, each of which presupposes different expectations on the part of the reader’ (Hardie 1994: 29). Since the various generic implications are disparate and not infrequently in tension, should we conceive of Virgil as content to leave his ‘voices’ in dissonance? If so, it would not follow that the Aeneid is an easy read for complacent readers: on the contrary. For, as Plato need not select a clear dialectical winner for us, yet nevertheless it remains incumbent on the engaged

172

Robert Wardy

to what his poetry can bring about, Virgil pledges that their fame will prove undying—or rather, that is the implication of its being coeval with Rome itself, if the City will prove eternal. On my and related readings, the ‘seal’ is dripping with strange and inscrutable ironies. Bleisch’s ingenious notion that in the σφραγίς Virgil is likening himself to the conciliatory Aeneas of book V is a lovely idea.⁵⁸ And perhaps our author is being even more outrageously authoritative. That Nisus and Euryalus were already there in the tradition, albeit as non-entities, must remain a possibility one cannot exclude, but it is far more congenial to suppose that they are pure invention⁵⁹—in which case Virgil ventures to immortalize what he has made up, investing his erotic fiction with Roman credentials (or vice versa?).⁶⁰ And another dollop of irony. Let us imagine a world in which our boys acquitted themselves impeccably, resisting the temptations of slaughter and booty, giving the Rutulians the slip, and making their way safely to Pallanteum. Would their mission have been a success? In the epic manner, the chronology is not really perspicuous (X 147–50), but we lack grounds for eliminating the depressing scenario according to which Aeneas has already left Evander for his rendezvous with Tarchon. If so, whom would Nisus and Euryalus have found in Pallanteum? The super-annuated Evander, under- and over-age males, and the womenfolk; thus a wretched failure in this world too, if no fault of their own.⁶¹ Naturally, Virgil could dispose of his creations as he pleased, and was at liberty to award the couple a distinguished fate in some other future; but he didn’t. Therefore one might float the paradox that it is only by virtue of their failure boldly exalted in the σφραγίς that ‘Nisus’ and ‘Euryalus’ survive as such evocative names. Phaedrus wrapped up his speech with a distribution of post-mortem glory (a distribution not, perhaps, unleavened by paradox and irony): Virgil’s reprise with variation dares to glorify an inglorious failure, endowing the ἔρως with a κλέος it so singularly does not deserve, from the perspective of either Homer or Phaedrus. and competent philosophical reader to attack (some of ) the outstanding problems, so might Virgil rely on his astute reader to resolve (some of ) the Aeneid ’s multiple ambiguities of message and ambivalence of tone. ⁵⁸ Bleisch 2001: 189. ⁵⁹ ‘The fiction of the Aeneid must be asserted with so much power that it will itself become a tradition. The poem faces head-on the fact that it is a fiction, yet one that has its own achieved power, effect and truth: it can make the invented characters of Nisus and Euryalus live’ (Feeney 1991: 186). ⁶⁰ To pursue the ‘making up’ of our strand of the tradition further into its Statius chapter would take us too far afield; but let us at least recognize that despite the careful modesty of his own σφραγίς (vos quoque sacrati, quamvis mea carmina surgant | inferiore lyra, memores superabitis annos. | forsitan et comites non aspernabitur umbras | Euryalus Phrygiique admittet gloria Nisi ( Theb. X 445–8) ), Statius is more than clever enough to produce his own variations on the Platonic original, to set beside Virgil’s. For when Dymas commits suicide and throws himself atop another corpse, declaring hoc tamen interea certe potiare sepulcro (Theb. X 441), we should be put in mind of not only Aen. IX, but also Achilles and Patroclus in Phaedrus’ speech. ⁶¹ Thanks to Robert Kaster for encouraging me to think this through (if I’m thinking about the number of Lady Macbeth’s children, he is not to blame).

Virgil’s Sacred Duo

173

But is it all irony? Virgil’s celebration of these lovely losers is not so unlike Tennyson’s Light Brigade, a cock-up which nevertheless shows you what the British Army is made of: hence Roman discipline, given the chance, could work with and on the raw material of Nisus and Euryalus, for all their ludicrously comic butchery and botchery. In this sense, to this imprecise extent, perhaps erotic self-immolation does somehow, sometimes, serve the public weal.⁶² Then again, one might take a different tack. That Nisus and Euryalus fail to meet Phaedrus’ standards has everything to do with their liminality, black hunters never to mature into hoplites:⁶³ at best a larval form of a possible template for a Roman version of the Sacred Band, had legendary history been different.⁶⁴ Rome was in fact—that is, in fantasy—founded not by male lovers united in death, but on the separation of fratricidal murder. That is, by invoking the parallel between the concluding, conclusive death of Turnus and the death of Volcens⁶⁵—followed immediately by the iuncta mors of our couple—one might interpret the death of Nisus and Euryalus as an abortive foundation story for an impossible city whose heroes are impotent to lay real foundations—which is where Virgil so helpfully steps in. Hence the σφραγίς: if ἔρως and πόλις are reconcilable, it is only within the confines of what Virgil himself has made up, and by subverting Phaedrus’ assurance that love protects the city; in these verses, Rome itself is harnessed, even subordinated, to the celebration of a private passion.⁶⁶ Nisus and Euryalus inhabit a masculine world. Only after their death does a woman burst on to the scene: the lamenting mother of Euryalus. In apostrophe to his severed and displayed head,⁶⁷ she laments that her son failed to bid her farewell. In her unrestrained grief she has let fall her quintessentially woman’s ⁶² With many thanks to John Henderson. ⁶³ Convention has it that an ἐραστής, while not very much older than his adolescent παιδικά, would nevertheless have attained full adulthood in some respects (not all: paederastic lovers are (normally) unmarried). How comfortably does this convention sit with Nisus’ immaturity? There is not really a difficulty. As I have repeatedly argued, he is a liminal character: thus in some respects, at some times, he clearly plays the part of the supervising, senior lover; elsewhere he does not. ⁶⁴ This is not the only cue for counterfactual erotic speculation. Let us reconsider Ascanius’ invitation to Euryalus to become his privileged, inseparable comes (275–80): what sort of friendship is in prospect? Since Nisus and Euryalus are inseparable, are we to envision a bizarre threesome? Surely Ascanius cannot imagine himself supplanting Nisus in Euryalus’ affections. Ascanius is only slightly younger than Euryalus (275–6). At least with reference to the paederastic paradigm, the mathematics are awkward: Ascanius cannot be added to the existing couple, Nisus cannot be subtracted from it, the difference between Euryalus and Ascanius is too small. So is it that Ascanius naively overlooks how perplexing intimacy with Nisus’ beloved would prove? Might this aspiration, at best inchoate, at worst incoherent, be a symptom of the lingering remnants of the dangerous immaturity from which he begins to emerge in the final sections of the book? ⁶⁵ Note condidit (443, of Nisus ‘burying’ his sword in the mouth of Volcens) and condit (XII 950, of Aeneas ‘plunging’ his sword into the breast of his last enemy). ⁶⁶ I owe this train of thought to Philip Hardie, who pithily concludes that all that is left is laus amoris, the only surviving product of amor laudum. ⁶⁷ In the grim light of the next day the remains of Nisus and Euryalus are all too conspicuous to the besieged Trojans: visu miserabile (465).

174

Robert Wardy

work, weaving (476), and flown to the walls to expose herself to the Rutulians, where no woman should be seen: this is a place for the fighting men alone, albeit for soldiers, not the perished hunting and hunted pair. Right order—that is, some approximation to the order of Phaedrus’ City of Men—is restored when she is escorted within. This grotesque finale to our episode supplies two more links from the Aeneid back to the Symposium. First, we might ponder antiquity’s most famous severed head, Orpheus’. That head talks or sings. This one remains dumb, of course. If, before their death, Nisus temporarily took on the guise of Orpheus, post mortem it is Euryalus’ turn—or rather, in his silence, he is an anti-Orpheus.⁶⁸ Second, my delineation of the allusive structure joining Plato and Virgil has repeatedly appealed to one or another city which might (not) have been: what about the pre-eminent other city of the Aeneid, Carthage, and the (non-) pair Aeneas–Dido? What of heterosexual love? Is there room for an ideal City of Lovers of Women? The short answer: no, not in this text; Aeneas himself is (temporarily) feminized in Carthage, and in book IX Numanus Remulus taunts the Trojans with the charge of effeminacy; he is dispatched by Ascanius, from the ramparts recently vacated by Euryalus’ mother (590–663). A nod in the direction of a much longer answer: is any City of Lovers a tenable vision? The Guards of the Republic enjoy sexual rewards for military prowess, but this is satisfaction of an appetite, not licence for the formation of erotic bonds; and Socrates’ communism puts familial loyalty distributed across a whole social class in the place of Phaedrus’ paederastic devotion limited to a single couple. The Platonic Socrates expresses his longing for a dynamic description of the ideal city in action (Ti. 19b–c). In the Aeneid Virgil demonstrates how imagined or imaginable, idealized or ideal cities equipped with competing erotic and military foundation myths proliferate in his text, or beneath its surface. This is no philosophical demonstration; but, for once, that may not have been what Socrates was looking for. REFERENCES Austin, R. G. (1964), P. Vergili Maronis Aeneidos Liber Secundus. Oxford. Bleisch, P. R. (2001), ‘Nisus’ Choice: Bovillae at Aeneid 9.387–8’, Classical Quarterly, 51: 183–9. ⁶⁸ The mother of Euryalus alone of the elderly and frail women had dared to remain with the Trojan army for the duration (217–18, 285–6); provision for her welfare was her son’s sole concern, eliciting from Ascanius the promise that he would treat her as his very own mother, except for the name ‘Creusa’ (297–8: is it coincidental that he swears by his own head (300)?). Despite her important rôle, she is the only anonymous parent of Nisus and Euryalus (on the likely assumption that Ida (177) is not the mountain, but the eponymous nymph who mothered Nisus). I find it difficult not to venture on to the bridge too far, and propose that an appropriate, suppressed name for Euryalus’ mother would be that other name for the mother of Ascanius: Eurydice (a ‘suppression’ occurring only in the reader’s imagination if, as seems very likely, none of these characters was to be found in any pre-Virgilian tradition). This identification would have the most pleasing consequence that in this version a live Eurydice mourns a vanished Orpheus.

Virgil’s Sacred Duo

175

Courtney, E. (1981), ‘The Formation of the Text of Vergil’, Bulletin of the Institute of Classical Studies, 28: 13–29. Davidson, J. (2005), ‘Mr and Mr and Mrs and Mrs’, London Review of Books, (2 June), 13–18. Dover, K. (1980), Plato: Symposium. Cambridge. Feeney, D. C. (1991), The Gods in Epic: Poets and Critics of the Classical Tradition. Oxford. Fitzgerald, R. (1984), The Aeneid: Virgil. London. Hardie, P. (1994), Virgil Aeneid Book IX. Cambridge. Hinds, S. (1998), Allusion and Intertext: Dynamics of Appropriation in Roman Poetry. Cambridge. Hunter, R. (2004), Plato’s Symposium. Oxford. Kenney, E. J. (1990), Apuleius: Cupid and Psyche. Cambridge. Leitao, D. (2002), ‘The Legend of the Sacred Band’, in Nussbaum and Sihvola (2002), 143–69. Lennox, P. G. (1977), ‘Virgil’s Night-Episode Re-examined (Aeneid IX, 176–449)’, Hermes, 105: 331–42. Lyne, R. O. A. M. (1987), Further Voices in Vergil’s Aeneid. Oxford. Makowski, J. F. (1989), ‘Nisus and Euryalus: A Platonic Relationship’, Classical Journal, 85: 1–15. Mynors, R. A. B. (1990), Virgil Georgics. Oxford. Nussbaum, M. C., and Sihvola, J. (2002) (eds.), The Sleep of Reason: Erotic Experience and Sexual Ethics in Ancient Greece and Rome. Chicago. Reed, J. D. (2004), ‘A Hellenistic Influence in Aeneid IX’, Faventia, 26: 27–42. Rowe, C. J. (1998), Plato: Symposium. Warminster. Saylor, C. (1990), ‘Group vs. Individual in Vergil Aeneid IX’, Latomus, 49: 88–94. West, D. (2003), Virgil: The Aeneid. London.

9 Plato on War Angela Hobbs

This chapter addresses two interconnected issues. The first is whether Plato regards war as an inevitable feature of human coexistence. Could there be a human society that avoided war altogether? If so, what would it be like? The second is whether Plato thinks the removal or avoidance of war in toto (as opposed to the avoidance of a particular threat) would in any case be desirable: even if it were possible, might the price be too high? Specifically, I shall be asking whether the development of civilized society in the Republic can only take place concurrently with the development of war. Does Plato think that humans can enjoy a life enriched by the arts and philosophy and certain material luxuries only if they are prepared to accept war? If one does have to choose between a peaceful life without philosophy and a war-prone life with it, which life should win? In trying to decide whether Plato thinks that such a choice really is unavoidable, we will have to examine his view of the roots of aggressiveness and warmongering in the human psyche,¹ and ask whether he believes aggressiveness itself to be innate, no matter what political and social conditions prevail, or what education and training are offered. For reasons of space, I shall focus on the Republic, especially the account of the origin of war in Republic II, though I shall also discuss passages from the Politicus, Laws, and elsewhere which illuminate that account. It is not my aim to give a comprehensive survey of every view expressed on war in Plato’s dialogues. When I do discuss dialogues other than the Republic, I shall not assume that a unitarian I am grateful to audiences in Cambridge, Sheffield, Warwick, and the University of North Carolina at Chapel Hill, where I gave earlier versions of this paper, and for comments and help received from Paul Cartledge, Stephen Houlgate, Stephen Makin, M. M. McCabe, Penelope Murray, Fabienne Peter, David Reeve, Malcolm Schofield, David Sedley, and Robert Stern; particular gratitude must go to Dominic Scott for his wise and astute editorial advice. Above all, I should like to express my profound thanks to Myles Burnyeat, whose lectures and doctoral supervision did so much to nurture my love of ancient philosophy, and to broaden and illuminate my philosophical horizons. ¹ Clearly, the majority of aggressive acts do not come under the heading of ‘war’, and wars may be undertaken for defensive as well as offensive purposes. For a definition of the cultural phemomenon termed ‘war’, see Dawson 1996: 13–14.

Plato on War

177

reading—or for that matter a developmentalist one—will apply to questions regarding war throughout the corpus. I certainly believe that it is worthwhile to reflect on whether the different passages tell a reasonably consistent story, but I do not assume in advance either that they do or that they do not. By ‘Plato’, I mean the author of the Platonic corpus, as it has come down to us. In this chapter, I do not suppose that Plato necessarily endorses the views of any of his characters, whether Socrates or the Eleatic or Athenian Strangers or anyone else. Plato’s views need to be reconstructed carefully, with due attention to what these characters say, but also with regard to the contribution of other interlocutors and the surrounding context.

P L ATO ’ S V I EW O F WA R Before the main issues can be explored, however, we first need to consider the critical question of Plato’s attitude to war itself—which is, we should note, a different question from whether he thinks that a community that lives without war is always preferable without qualification to a community that lives with it. In terms of the overt comments, the answer is clear: any careful scrutiny of the dialogues shows that most of his main characters’ explicit appraisals of war are negative.² Its origin in the Republic (which we shall be discussing in detail below) is said by Socrates at 373e to be the same as that of most evils (kaka), namely acquisitiveness, a point also made bluntly at Phaedo 66c: ‘all wars are made to get money’. And at Republic 378b–c all stories of wars and battles amongst the gods are to be censored, because ‘quarrelsomeness is one of the worst of evils’.³ In Republic 8–9 the timocratic state is criticized for preferring spirited, ‘thumoeidic’ men who opt for war over peace (547e–548a), and the tyrant is condemned for stirring up one war after another, both to make his people feel the need for a strong ruler and to reduce them to poverty and hence powerlessness (566e–567a). In the Politicus, too, the undiluted brave and manly types (the andreioi) are chastised for being excessively warlike and leading their state to destruction or slavery,⁴ while in the Laws, the Athenian Stranger criticizes states which focus exclusively on promoting courage and manliness (andreia), since they subordinate peace to war rather than war to peace (628c–e): it is peace, not war, that is unambiguously declared to be the highest good. The same point is made at 803d, where we are told: ‘in war there is neither play nor education ² The most obvious possible exception that I can find is Ti. 23c–d, where Athens was originally established by a ‘war-loving’ (philopolemos) as well as ‘wisdom-loving’ (philosophos) Athena. See n. 32 below. For Plato’s use of polemos (‘war’) and cognates throughout the dialogues, see Brandwood 1976. ³ All translations from the Republic are those of Lee 1974, unless otherwise stated. ⁴ Plt. 308a; though one should note that the opposite, decorous (s¯ophrones) types are also said to put their state at risk by not being warlike enough and being unable to offer robust defence.

178

Angela Hobbs

worthy of the name, nor will there ever be; it is the life of peace that everyone should pursue, as much and as well as he can’.⁵ Such explicit appraisals, of course, are not the whole story. As we shall see, no main character advocates pacifism in the face of aggression,⁶ and Socrates consequently devotes painstaking attention in Republic II and III to the education of an Auxiliary class of Guardians whose chief function is military defence (this emphasis on defence rather than attack will be discussed below). The central characters also acknowledge that war can be an excellent training-ground, display case, and test for certain virtues, most notably andreia. Witness, for example, Republic 429e–430c, where civic (politiké) andreia is defined by Socrates as ‘the ability to retain safely in all circumstances a judgement about what is to be feared, which is correct and in accord with law’—an ability, in other words, which can clearly be well examined and displayed on a battlefield. And at 466e–467e provisions are made for allowing the children of Guardians to witness and assist in wars as part of their training; this training, furthermore, is not just in the skills of fighting but also in the development both of a sense of their future civic responsibilities and of the temperament to fulfil them (467a). Clearest of all, perhaps, is Laws 707c, where the Athenian Stranger unequivocally says that the land battles of Marathon and Plataea have ‘made the Greeks better (beltious)’.⁷ It is also worth noting in this context how both Socrates and the young mathematician Theaetetus are portrayed in military settings to advantageous effect in, respectively, Symposium 219a–221c, Laches 181a–b, and Theaetetus 142b. It is worth remembering, too, that in the Republic properly developed and controlled military courage (i.e. controlled by the reason of the PhilosopherRulers) is to receive state-sanctioned social rewards: at 460b, for example, the state’s best young male defenders are to be honoured in various ways, including being offered more, and more enticing, sexual opportunities at the official mating festivals. This honour is of course also intended as a means of breeding a new generation of excellent warriors for the state (468c); but it is clear that it is in addition part of a general policy of encouraging the right kind of soldier and military andreia. In similar vein, 468b highlights how those who excel on the battlefield will receive a potent cocktail of honours, embraces, and food.⁸ It ⁵ Trans. Bury 1926. It is worth emphasizing these passages, as a number of classical scholars have portrayed Plato in a considerably more militaristic light. A case in point is Craig (1994), who, though commendably challenging, is decidedly selective in his reading. In discussing Rsp. 373e, e.g., Craig simply reports (p. 6) that Socrates says that they are not currently concerned with whether the effects of war are good or bad; he does not mention Socrates’ telling claim in the same passage that the origin of war—acquisitiveness—is the same as that of ‘most evils’. ⁶ See also n. 4 above. ⁷ In contrast to sea battles such as Salamis which, despite its successful outcome, in the Stranger’s view made the Greeks worse: sailors, he claims, do not stick boldly at their posts, but escape without shame when the going gets tough (706b–d). He may also be hostile to Salamis because of its acknowledged role in strengthening Athenian democracy: see Arist. Pol. 1304a 18–29. ⁸ It would be intriguing to know whether Socrates envisages successful female warriors—and there are women in the state’s army (457a)—being rewarded in the same way.

Plato on War

179

further appears that human warriors of the appropriate kind⁹ are to be fitting subjects for the heavily censored art-forms of books II and III: in 389e Socrates speaks approvingly of Iliad III. 8 and IV. 431 where ‘The Achaeans moved forward, breathing valour, in silent obedience to their officers’, and in 399a he asks for a musical mode that will represent a brave man on military service who faces injury or death with steadfast fortitude. Select warriors may even continue to be eulogized in book X: though by now almost all art is actually banished from the state, the fortunate citizens are still permitted ‘hymns to the gods and paeans to good men’ (607a), and the latter group can presumably include the state’s defenders. Nevertheless, we need to be careful here. Such passages celebrate the courage, resolution, and obedience of a certain type of soldier, and acknowledge the role that war can play in forming and testing the corresponding type of courage; they do not overtly celebrate or foster war in itself (not even Leg. 707c). Overtly, Plato’s attitude to war is still that it is an evil, albeit an evil which can allow for the display and nurture of certain goods. In this, if nothing else, his position on war bears some comparison with that of Hegel in Philosophy of Right, who maintains in 330–40 that though war is not an absolute evil, it is an evil nevertheless. One cannot simply go around starting wars in order to purify the nation’s moral health. Apart from the initial land appropriation of 373d which introduces war into the polis (to be discussed in detail below), we do not hear of Plato’s Guardians making an unprovoked assault on another state;¹⁰ on the contrary, all the emphasis is on their ability to defend their country if attacked. To take just four examples: at 414d–e Socrates proposes that all the citizens be told that they are, in fact, descended from Mother Athens and must protect her, while at 415d–e the Guardians are enjoined to pick a site for the polis which will be easy to defend; 422–3 outlines ways of defending the republic against wealthier states, and in 421–3 we are told that the state must not grow beyond a certain size or exceed a certain measure of wealth, both clearly policies which will prevent aggressive expansionism.¹¹ In addition, as far as the Rulers are concerned, wars arguably divert time and energy away from working towards and contemplating the Forms of Justice, Beauty, and the Good. One could argue, in a slight concession to Craig,¹² that by devoting considerable attention to war and warriors Plato is implicitly —perhaps inadvertently—celebrating and even nurturing war. There may be something in this: consider, for example, Republic 521d, where it is said that all the higher studies undertaken by trainee Philosopher-Rulers must be relevant in war and ⁹ And perhaps divine ones too: the ban on portraying battling gods at 378b–c applies explicitly only to gods fighting amongst themselves and against ‘heroes and their friends and relations’. ¹⁰ It is not even clear that we are supposed to think of this first assault as being undertaken by Guardians—certainly not trained ones; the text is open to interpretation on this point. ¹¹ 460a also states that the polis must not be allowed to grow too large. ¹² See n. 5 above.

180

Angela Hobbs

appropriate to soldiers. It may be thought that, instead of praising military courage, we should be working towards a world in which the virtue of military courage is no longer necessary. Yet there is, I believe, far more to be said for the undeniable fact that Plato makes a sustained effort throughout all his works to extend andreia into non-martial spheres, such as philosophy. In the Republic alone, we are told that the philosopher has to endure hostile attacks, and that philosophy requires boldness and courage; doing philosophy is also compared to waging battle for truth in the face of hostility, ridicule, and contempt.¹³ Admittedly one could argue that such metaphors can work two ways: instead of being applauded for taking traditional agonistic activities and reworking them in a less aggressive fashion, Plato could be charged with promoting an agonistic culture by portraying philosophy as a contest in the first place, albeit one aimed at vanquishing falsehood rather than specific opponents (even if some of the opponents themselves continue to be motivated by personal animosity). As I have written on these matters extensively elsewhere, however,¹⁴ I shall not dwell on them here; the key point for my present thesis is that Plato believes that it is perfectly possible for all the virtues, including andreia, to flourish in peacetime. This is the vital move, irrespective of how one reads his depiction of the manliness of philosophy. The courage of the soldier is not the only kind of courage. War can certainly provide a backdrop for the exhibition of valour, but it is by no means the only field, or even necessarily the best field, for its display.

T H E O R I G I N O F WA R We are now in a position to turn to the question of whether Plato regards war as an inevitable feature of human existence. The first issue we need to explore is his analysis in Republic II of the origin of war. In the attempt to define justice, and consider whether it is helpful or harmful to the individual agent, Socrates proposes that they first consider justice on the larger canvas of the state, and that this can most clearly be achieved by constructing a community from scratch, to see precisely when, why, and how political justice emerges. So in 369b–372d he constructs a simple, classless community based on economic need and exchange.¹⁵ He and Adeimantus agree that rather than each person trying to supply all their own needs through their own labour, it is easier for each individual to specialize and devote their time and energies to one particular job, such as cobbling, and exchange the products of their craft with those of, for ¹³ e.g. 450b, 472a, 473e (enduring hostile attacks), 486b, 535a, 357a (need for boldness and courage), 534c (practice of philosophy compared to making an assault). See Hobbs 2000: 232–49, esp. 243–4 with n. 69–74. ¹⁴ Hobbs 2000. ¹⁵ This simple community is helpfully discussed in Barney 2001; Schofield 1993; and Cross and Woozley 1964: 75–93. The notes of Adam (1963) on this section remain illuminating.

Plato on War

181

example, the weaver or the farmer. Such a system also allows for the fact that in all crafts situations arise when certain tasks need to be carried out immediately: one must seize the kairon, the appropriate moment. Somewhat conveniently, Socrates appears to envisage (a) that each individual is naturally suited to one job and one alone and (b) that the distribution of natural skills amongst the population matches that population’s requirements. Such happy coincidences may well strike one as in need of justification, but fortunately they are not our present concern.¹⁶ Let us grant that in this classless society everyone is both a producer and a consumer of basic goods, living a simple bucolic existence of shared meals of bread and wine and uplifting community hymn-singing. Crucially, it is a community at peace, both with itself and its neighbours: there is no poverty and no war. Nevertheless, it is important to notice that the inhabitants are still aware of the possibility of both these things and are keen to avoid them (372b): hence their endeavours to keep their community small and to live within their means. Glaucon, however, complains at this juncture that ‘that’s just the fodder you would provide if you were founding a community of pigs’ (372d)—doubtless a reference to the inhabitants’ permitted dessert of roasted acorns. In response, Socrates, despite emphasizing that this is the ‘true’ and healthy polis, allows the introduction of more sophisticated but explicitly ‘unnecessary’ (373a–b) goods and activities into what is now termed a ‘luxurious’ city suffering from feverish inflammation. The inhabitants are to be granted furniture, expensive foodstuffs, courtesans, cosmetics, and the fine arts, including artefacts made from gold and ivory.¹⁷ Unnecessary desires, however, are also crucially unlimited, and so too are the goods they seek (373d10); their indulgence will thus lead to a great expansion of the polis. Artists, beauticians, hunters, cooks and doctors, for instance, will all now be required, and their introduction in turn requires an expansion of the land needed to house and feed them. Before long, the community will start to appropriate some of its neighbours’ territory, the neighbouring states will grow jealous of the community’s increasing wealth and try to appropriate it in turn, and war will result. According to the one person, one job principle already established, this will entail the creation of a separate warrior Guardian class. And this is also the point where we are told, as we saw earlier, that the cause of war is ‘the same as that of most evils’ (namely, acquisitiveness).¹⁸ ¹⁶ Annas (1981: 73–6) discusses the principle of specialization; see also White 1979: 85–7. Schofield (2000: 209) suggests that the idea is that the relevant specialists are to be collected to form the first society, thus avoiding the problem of unlikely coincidences. ¹⁷ Burnyeat (1999: 231–6) emphasizes the importance of couches and tables in particular in the Greek ideal of public dining as an essential component of civilized society. ¹⁸ It is interesting to compare Socrates’ account in the Republic with modern archaeological theories which locate the origins of organized warfare during the first significant land settlements of the neolithic period, on the grounds that land settlements both offer possibilities for land disputes and provide a little more leisure for activities other than hunting and gathering. It is not coincidental that it is in the neolithic period that we start to see increasing craft specialization,

182

Angela Hobbs

Do we now have at least possible answers to our opening questions? What can we learn from Republic 369–75 about whether Plato thinks humans can live together in any form of community without war, and, further, whether he thinks they can live without war in political structures embellished by the arts? At first glance it does seem that some form of very simple human association might be possible on peaceful terms; that is suggested by the first pre-political community. But, in answer to the second question, the price of peace will be high, and too high for some: war thus appears to be a feature (or at least a necessary possibility) of any reasonably developed constitution; war and civilization are part of the same process. This is what is suggested by the luxurious state that Socrates introduces at Glaucon’s request. It may help to illuminate this initial interpretation of Republic II if we contrast it with the position of Hobbes in Leviathan, Elements of Law, and De cive.¹⁹ For Hobbes, the natural state of man (that is, the state of man without political organization) will quickly degenerate into a state of war. Very briefly, the chief reason for this lies in his belief that ethical values are both relative and subjective. It is not simply that humans do differ widely, as a matter of fact, in their views about what constitutes good and bad, right and wrong; it is also the case that such differences cannot be resolved by simple appeal to an independent and natural criterion. In nature there is no such criterion. So discord and wars will naturally tend to arise, particularly over disagreements as to when it is appropriate to assert one’s right to self-protection. The only way out of this impasse, Hobbes thinks, is for man to supply what nature has not: namely, objective standards of right and wrong in the shape of civil laws made by the sovereign ruler, or rulers. So, in Hobbes’s eyes, we have a clear choice between war and law, which are presented as stark alternatives. Whereas in Republic II war appears at first sight to be depicted as an inevitable part of civilization in the sense of a politicized life informed by the arts, for Hobbes civil society or the commonwealth is the only conceivable solution to war. I S WA R AVO I D A B L E ? T H E EV I D E N C E O F T H E BU C O L I C C O M M U N I T Y To what extent, then, is Republic 369–75 to be taken at this initial face value? I shall take each component of this interpretation in turn, starting with the claim including specialist weapons production. See O’Connell 1989: 30–44; Dawson 1996: 30; Ferrill 1985: 12, 18–31 (Ferrill also gives evidence for possible warlike activity in late palaeolithic times, but argues persuasively that most features of warfare (‘true warfare’ as he puts it, p. 19) arrived with the technological innovations of the neolithic period). For a judicious evaluation of the relation between war and the economy, and particularly on the war potential inherent in both staple and luxury goods, see Schofield 1993: 188–9. ¹⁹ See, e.g., Elements of Law 11.10.8 and De cive 1.13.

Plato on War

183

that this first community shows how humans can live together without war (whether or not that is desirable). The crucial question to consider here is clearly whether Plato ever really does envisage the war-free pre-political community as a genuine possibility, and this, of course, depends on the vexed and intriguing issue of precisely how to interpret both the tone and the content of Socrates’ description of it. Let us begin with the question of tone. Is Socrates’ account supposed to be a serious analysis, either of the actual historical development of the polis (or a possible version of it), or at any rate of its economic and psychological origins? Or is it supposed to be simply some sort of joke, perhaps a satire on currently fashionable stories of a distant golden age or Age of Cronos?²⁰ Such legends of former easeful idylls had been popular in Greek literature at least from the time of Hesiod,²¹ and depending on both writer and immediate context, their tone can range from wistful nostalgia through critique of contemporary mores to absurdist parody.²² Or are both seriousness and satire in play? The two readings need not be mutually exclusive: Socrates could be genuinely attracted to this first, austere community while simultaneously being amused by certain literary treatments of similar pastoral scenes.²³ It has sometimes been assumed that Glaucon’s derisive snort—that Socrates appears to be founding a community of pigs—is conclusive proof that this first community is intended solely as a parody, but this does not necessarily follow: if Socrates is no straightforward mouthpiece for Plato, then Glaucon most certainly is not. Whatever Glaucon’s underlying stance, his explicit comments often challenge Socrates’ proposals, and he is always portrayed as a forceful advocate of a sophisticated and urbane way of life.²⁴ ²⁰ In Hes. Op. 108–110, the men who lived in the blessed and leisured Age of Cronos are specifically called ‘golden’. For a general discussion of the tradition of a golden age in Greek literature, see Baldry 1952: 83–92 and Guthrie 1957: 69–79. ²¹ Hes. Op. 105–202. Empedocles may also have described a golden age in his Purifications: see Barnes 1987: 198–201. ²² For nostalgic uses, see Arist. Ath. pol. 16. 7; Ath. Deipnosophists VI, 267e–270a (although some of the texts that Athenaeus mentions are satirical). See also the vegetarian Peripatetic Dicaearchus appealing to the myth in support of his views in Guthrie 1957: 74 and Vidal-Naquet 1986: 285–7. Aristophanes mocks characters seeking to restore the Age of Cronos at Clouds 398 and 1070 and Plutus 581. Vidal-Naquet 1986: 285–301 contains a general discussion of ambivalent responses to the golden age; see too Brisson 1970: 402–38 and Dillon 1992: 21–36. ²³ Burnyeat (1999: 229–30), e.g., focuses on the ‘teasing’ tone of Socrates in this passage and his liberal use of irony, while Schofield (2000: 211) calls Socrates’ account ‘tongue-in-cheek’ and ‘a comic explosion’. Barney (2001: 216–17) also emphasizes the picture’s ‘parodic’ nature, and Dillon (1992: 26) writes that ‘Plato appears to go somewhat out of his way to make fun of the simple level of society he has postulated’. Yet all these writers hold that the passage has serious import as well, or at least fulfils a serious role within the structure of the dialogue as a whole (see also Schofield 1993: 183–96). In contrast, Crombie (1962–3: 89–90) straightforwardly takes the first city to be the true ideal, with no irony in play. ²⁴ For overt criticism see, e.g., 473e–474a, where Glaucon exclaims that if Socrates is going to advise that philosophers should become kings, or all kings become philosophers, then he must not be surprised if a crowd takes up arms and pursues him. For Glaucon’s love of the arts, see 399e, where he is called mousikos, ‘cultured’.

184

Angela Hobbs

The strong probability that Socrates is gently mocking a tradition of golden age literature does not in itself, therefore, preclude the possibility that he is also offering us some sort of serious analysis. But if some kind of serious analysis is in operation here, the question now becomes: analysis of what? Is Socrates providing us with an account of what he considers to be a real possibility (whether set in the past or the future) of how empirical humans might live together in peace, or is he concocting a fantasy designed to explore the economic and psychological origins of society? It seems to me that the true challenge to the former, empirical option comes not from the teasing tone of the passage but from the question of what sort of psychological make-up is possessed by the inhabitants of this first community. Are they envisaged as having tripartite psyches comprising reason, spiritedness (thumos), and the appetites, anticipating the model of Republic IV? Or are their psyches supposed to consist only of appetites, and (initially at least) the necessary and limited appetites at that?²⁵ If the latter reading is preferred, then it does look as if 369–73 is not intended to be an account of recognizable, empirical humans, but is being put forward as a deliberately unrealizable fantasy, one of the functions of which is to clarify the complexity of actual humans and locate the origins of aggression and war more precisely. If, on the other hand, the book II producers and consumers are conceived by Plato as possessing tripartite psyches—are real, empirical humans, in other words—then this would strengthen the view that Socrates is offering the genuine possibility of a simple human community living without war (although always, we should remember, with an awareness of war and a desire to avoid it) (372b–c). This interpretation, however, raises problems of its own. If this is supposed to be an empirical possibility, then (a) why is there no explicit reference to tripartition at this stage of the dialogue, and (b) why do the (implicit) rational, spirited, and appetitive parts of the psyche operate so differently in this first community from how they are said to operate after 373? Their rational abilities, for instance, appear to be confined to means–end reasoning: there is no suggestion that they ever practise philosophy or speculate on any subject in a general way. Their appetites—their epithumiai —are likewise confined to the necessary and limited desires, those desires aimed at securing physical survival and (very) modest comfort. And their spirited elements, their thumoi, are not explicitly mentioned at all. If the thumos is nevertheless still supposed to be present in them, why is it so far removed from the raw self-assertive drive that constitutes the untutored thumos immediately after this passage at 375 (the first mention of thumos as a psychic function in the text)? Their appetitive and spirited elements would appear to be ideally restrained and docile, even though they do not have the ²⁵ It is by no means clear whether the economic producers of 369–73 are supposed to be an exact match for the Producers in the ideally just tripartite state. In any case, even from book IV onwards, the structure of the Producers’ psyches is not always easy to determine.

Plato on War

185

benefit of Philosopher-Rulers to guide them and the educational institutions of the republic to form them. Taken together, these problems suggest strongly to me, far more than Glaucon’s comment, that Socrates is not putting this community forward as a serious depiction of how fully realized, tripartite humans might once have lived at peace, or could live at peace in the future, but is rather offering it as a fantastical means both of clarifying the nature of actual human societies and the origin of war, and of commenting on the current state of affairs, in Athens and elsewhere.²⁶ It cannot therefore be taken as support for the view that empirical, psychologically tripartite humans can live without war in certain circumstances.

T H E P R I C E O F PE AC E So far we have mostly been addressing the first of our main issues—namely, whether war is an inevitable feature of human coexistence—and we shall shortly be considering the evidence of the second, luxurious, war-prone polis in this regard. First, however, we need to reflect more carefully on how the bucolic community can assist with our second main question, that of whether the removal or avoidance of war in toto would in fact be desirable. Certainly the character of Socrates is not likely to care much about forgoing cosmetics, prostitutes, and fancy foodstuffs; and given the amount of artistic censorship he is soon to propose for the ideally just state, he may even be happy enough about forgoing all or most of the fine arts. Yet Plato the consummate artistphilosopher may well have more doubts. Does he view the bucolic community as any sort of ideal, however fantastical? To help answer this question, we shall make a brief excursus into other examples of associations which exist without war in the dialogues, before turning back to the bucolic community of Republic II.

Excursus: other war-free communities in Plato There are three passages we need to consider: one from the Politicus, the other two from the Laws. At Politicus 271–2 the Eleatic Stranger recounts a myth concerning a previous Age of Cronos—as opposed to the current, more troubled Age of Zeus—when fruits grew plentifully without any need for agriculture and the climate was so mild that houses and most clothing were ²⁶ At the very most, the problems suggest that Socrates is implying that such a halcyon idyll could never have lasted very long, as its survival depends on certain key aspects of the human psyche remaining in abeyance. However, this interpretation seems to me far less plausible than supposing that Socrates does not intend to describe empirical humans in this passage. See also Barney 2001: 218–20.

186

Angela Hobbs

unnecessary.²⁷ Animals were tame and could converse with humans, who were, partly in consequence, vegetarians. Above all, there were no political constitutions (politeiai), and no wars or strife of any kind. The Eleatic Stranger now asks his interlocutor, Young Socrates, whether life in this peaceful, leisured community would be happier (eudaimonesteros) than in contemporary war-prone human states. Perhaps surprisingly, Young Socrates is not sure, and the Stranger says it all depends on whether the beings²⁸ of the Age of Cronos used their leisure to engage in philosophical discussion, both amongst themselves and also to gain the viewpoints of the talking animals. The very strong implication is that if they did not, then even a society without war is not worth the price of forgoing philosophy.²⁹ The two other main passages in which Plato envisages a war-free society, Laws 677–80 and 713a–714b, are rather more difficult to interpret in respect of our central questions.³⁰ At 677a, the Athenian Stranger asks Megillus and Clinias whether they believe there to be any truth in the ‘ancient stories’ (palaioi logoi) which tell of the frequent devastation of human societies by floods, plagues, and other catastrophes. In particular, he asks them to consider the effects of a mighty flood which destroyed cities and most of their inhabitants, and which only a few herdsmen living in the hills were able to survive.³¹ The Stranger proceeds to describe the simple and mostly artless life of these scattered survivors—artless because the flood has swept away not only tools but also knowledge of skills and inventions, as well as statecraft and legislation. However, despite their fear and desolation, they are also free from the greed, rivalry, and corruption of cities, ²⁷ For thoughtful discussions of the Politicus myth see Brisson 1995; Lane 1998: 99–136; McCabe 2000: 141–9; Rowe 2000: 239–44. ²⁸ ‘Beings’, as they are not depicted by the Stranger as fully human: they are, e.g., earth-born (271a). ²⁹ Although the Stranger emphasizes that the myth is ‘child’s play’ (paidia, 268d–e) and that it is ‘over-long’ (277b), this description does not prevent it from raising profound questions about what is of real value in contemporary human life (as well as in life for the beings of the Age of Cronos). The seriousness of paidia is often emphasized by Plato (see e.g. Resp. 424e–425a and Leg. I, 643b–d and VII, 793e–794b, 797a–798d). ³⁰ Cambiano 2002 contains a fruitful discussion of Leg. 677–80. See also Barker 1918: 356–7. ³¹ Ti. 22a–23a and 25c–d also refer to periodic cataclysms, the worst caused by fire and flood, which destroy large numbers of mankind and all the arts, sciences, and records of the civilizations affected; there is mention of one great cataclysm in particular, the flood survived by Deucalion and Pyrrha, at 22a. The Critias, too, makes repeated references (109d–e, 111a–b, 112a, 112c–d) to the periodic destructions of Ti. 22–3, and Criti. 112a again specifically mentions the great flood endured by Deucalion. These passages, however, do not then go on to discuss how war in general arises in the consequent rebuilding of civilization, concentrating instead on the supposed war between an earlier, more illustrious version of Athens and the island of Atlantis. ( The Timaeus and Critias passages do, however, assume that any well-ordered city will be thoroughly prepared for war and expert in its arts; more disturbingly, as we saw in n. 2 above, the Timaeus emphasizes (23c–d) that Athens was originally established by a ‘war-loving’ (philopolemos) as well as ‘wisdom-loving’ (philosophos) Athena: it is arguably the most obvious exception to Plato’s generally negative stance towards war). Finally, the Politicus myth discusses periodic cosmic reversals (270b–d) which bring great destruction to humankind, but does not mention a flood or floods.

Plato on War

187

and, critically, civil strife and war are unknown to them (678e). This is for a number of reasons: on the one hand, the catastrophic situation that they face means that they need to stick together to survive; on the other, they possess sufficient resources in terms of food, clothing, shelter, and clay to prevent squabbling. It is also highly significant that all the mines were flooded: there is no gold or silver to fight over—no one is rich, and there is thus no cause for jealousy—and in any case there is no iron or (the copper and tin required for) bronze with which to make weapons.³² The interpretative challenges of this story fall into two main groups. First, although the Stranger and his interlocutors agree that the tale is ‘very credible’ (panu … pithanon, 677a), it is still hard to know the extent to which Plato intends the account to be read as historical. Is he really offering it as a plausible description of an empirical society which is genuinely—if temporarily—free from war? Or is he simply trying to illuminate the conditions which give rise to war and peace? The diction of the passage could support either reading: the Stranger uses ambivalent phrases such as ‘let us conjecture’ (no¯es¯omen, 677a) and ‘let us assume’ (th¯omen, 677c) and his interlocutors reply that his conjectures and assumptions are ‘certainly probable’ (eikos goun, 677b)³³ and ‘most likely’ (kai mala prepei touth’ houtos, 678c). Such phrases could simply indicate that they are discussing a period of which there are no written records, or they could be taken to suggest that Plato wants us to see the account as a fiction.³⁴ Secondly, it is again hard to determine the precise tone of the passage and how we are supposed to view this society of survivors. It is true that all three speakers agree that such a society, neither rich nor poor, is the kind in which the noblest characters (gennaiotata ¯eth¯e) are formed, and they speak approvingly of the simplicity and moral innocence of their imagined (if not imaginary) community.³⁵ Unlike the Republic or Politicus, none of them expresses any particular concern that the absence of the art of war is concomitant with the absence of all the arts and sciences, including the art of writing. But this does not mean that Plato, a writer adept in a number of arts and sciences, is so unqualified in his approval, or intends us to be. Admittedly there is no such clear flagging of doubts as there is in both the Republic and Politicus, but there is surely a warning sign in the comparison (680b) made between the patriarchal customs of the clan-based hill-dwelling communities of Laws 677–80 and the patriarchy ³² Although Legs. 677–80 does not purport to depict a golden age, it reminds us of the irony that gold itself does not, and cannot, feature in golden age literature, except metaphorically to describe the inhabitants of more plentiful and leisured times. All metals and metallurgy were seen as a source of strife. ³³ See also 681a. ³⁴ Guthrie (1957: 67–8) tends to the latter reading, though he adds that the text precludes certainty. I see no compelling evidence to sway one in either direction. ³⁵ ‘Imagined’ is not intended to rule out the possibility that such a community once existed, or could exist in the future; it simply denotes that the community so described does not currently exist, and has not recently existed.

188

Angela Hobbs

practised by the unlovely Cyclopes in the Odyssey.³⁶ Yet even here the message appears to be mixed: while Megillus talks of the Cyclopes’ ‘savagery’ (agriot¯es), the Stranger describes such patriarchy as ‘of all kingships the most just’ (680e). Nevertheless, it seems safe to suppose that Plato would hardly regard the brutal, man-eating Cyclopes, who live without philosophy, state laws, or reverence for the gods, as an ideal to which humans should aspire.³⁷ At the very least, Plato seems to be inviting us in this passage to consider whether a community without war is necessarily always preferable to a war-prone one. The final, apparently war-free community, described in Laws 713a–714b, is hard to interpret chiefly because it is so brief, and because its main function is not to examine the development of the polis and the origin of war, but rather to emphasize how quickly human societies disintegrate when left in merely human control. As in the Politicus, we are returned to the blessed Age of Cronos, when human affairs were guided by wise and benevolent daimons, and humans lived a life of plenty and peace. Although this peaceful state is explicitly said to be a happy one, no mention is made of whether these peaceful, happy humans studied philosophy or not, or indeed practised any of the arts or sciences. However, we are told that both their homes and their states were directed by immortal reason, enshrined as law, and this suggests that the practice of philosophy would at least not have been out of place in such a society. Perhaps, then, these mythological beings did not have to choose between civilization and peace, although this does not necessarily help those of us living in the more challenging, human-misdirected, Age of Zeus.

Is the avoidance of war desirable? The evidence of the luxurious city Beyond the Republic, then, there is some evidence that Plato would not see a war-free community which lacked philosophy (and the arts and sciences) as desirable. So if he were to see the bucolic community in Republic II as some sort of (fantastical) ideal, we should expect it to include philosophy, at the very least. But, as we have seen, this appears very implausible. Although we are not explicitly told that the acorn-munching hymn-singers do not engage in philosophy, there is absolutely nothing to suggest that they do or would, or perhaps even could: there is no evidence that they possess anything more than means–end reasoning powers.³⁸ It certainly appears that the practice of philosophy in the Republic does not come ³⁶ Hom. Od. IX. 105–566; their practice of patriarchy is cited as one of their distinguishing characteristics at 112–15. ³⁷ This is the only passage in the dialogues in which Plato mentions the Cyclopes, but there can be no doubt about his lack of sympathy with their way of life. For further discussion of their role in this passage, see Dillon 1992: 30–1. ³⁸ Barney (2001: 213) also holds that there is no philosophy or serious intellectual inquiry in the First City.

Plato on War

189

about until after the development of the fevered society and the emergence of war. The term philosophos first appears at 375e10, in a discussion of how to select the right children for training as future warrior Guardians: they must possess both spirited (thumoeidic) and wisdom-loving (philosophic) elements in their psyches, and the education of these young warriors is directed chiefly at achieving the correct balance between these thumoeidic and philosophic aspects through a judicious mixture of literature and music (mousik¯e) and physical training. Let us now turn to the ‘luxurious’ city. As we have just noted, it is only at this point that philosophy and the other sciences are introduced. Does this mean that to pluck the fruits of philosophy one has to accept war as inevitable? This is the crux of the matter, and I believe that the answer is no, though assessing the evidence is far from straightforward. The main point to emphasize is that the introduction of war into the community at 373d–e is not necessarily final: there is, after all, most of the dialogue still to come, and there is plenty of time for the relationship between the polis and war to change. So let us consider what happens next, particularly to the newly demarcated Guardian class. After their early studies in mousik¯e and gymnastic, those amongst the tyro warriors who are found to be most suited to the study of philosophy will be siphoned off when they are about 21 and given an arduous higher education in mathematics, astronomy, harmonics, and dialectic as the core of their training to become Philosopher-Rulers, a process which culminates when they are 50 in a revelation of the Form of the Good.³⁹ Until they are 50, they must intersperse these studies with periods of military training and duties; after 50, it appears that they are now fully-fledged Philosopher-Rulers and absolved of military obligations, though clearly not political ones (540b). So the critical question now is: can a state actually ruled by philosophers ever be free from the greedy and feverish unnecessary desires which are said in 373d–e to lead to expansion, appropriation, and war? And can it also be free from the kind of misdirected and unchecked thumoeidic desires for status, victory, and power which at 547e–548a are criticized for encouraging a preference for war over peace? In short, can citizens ruled by the beneficent reason of philosophers be free from the desires which are the root cause of aggressive warmongering? A key quote at 399e suggests that perhaps they can: the fevered, landappropriating, and somewhat bellicose state depicted in 372–5 is not, we learn, to be its final development.⁴⁰ In the intervening section Socrates has been outlining the rules by which literature and music are to be composed and performed—rules which we later learn are to be established and maintained by Philosopher-Rulers.⁴¹ ³⁹ Early studies in mousik¯e and gymnastic: 376c–412a; higher education for selected Guardians: 521c–541b (selection at 537b and d); revelation of the Form of the Good: 532b and 540a. ⁴⁰ A point highlighted by Guthrie 1975: 448 and Burnyeat 1999: 231. ⁴¹ Guardians are not divided into Auxiliaries and Rulers until 412b; the concept of the Philosopher-Ruler first appears at 473c–e (though Guardians were said to require ‘philosophic’ qualities at 375e). That the Philosopher-Rulers are supposed to supervise the training of both mind

190

Angela Hobbs

At 399e Socrates then claims that the proper implementation of these rules will have the effect of ‘thoroughly purging’ (diakathair¯o) the old swollen state. Perhaps one of the effects of such purging will be to remove those feverish acquisitive impulses and unchecked and perverted thumoeidic desires that lead to war. Such speculation is strengthened by the fact that at 431a–b bad desires are said to be the result of ‘bad upbringing or bad company’. If so, then we can at least say, in answer to our second main question, that offensive war is not necessary to civilization. This seems to be confirmed by a fact we have already noted: namely, that after the belligerent land appropriation of the fevered state at 373d, all the emphasis is on the Guardians’ ability to defend, not attack: witness 414d–e, 415d–e, 421–3.⁴² These passages, of course, underline an uncomfortable truth: the fact that even a purified, peaceloving state will always remain at risk from attack by neighbouring non-ideal states, and as long as these neighbouring states remain non-ideal there will thus always remain the need for a highly trained army, whose soldiers (as we have also seen) are to be kept motivated by a number of state-bestowed honours⁴³ and honed by the correct balance between gymnastic and mousik¯e. At 543a we are still told that the ‘kings’ should be those who are best at philosophy and war. So it remains the case that however purely civilized a state may be, it will always be faced with the possibility of having to engage in defensive war. Yet, even if this reading is broadly correct, 399e still poses a fundamental problem of interpretation, depending on how one understands Socrates’ claim that they have, without noticing it, been thoroughly purging their bloated and luxurious (truph¯o) state. Does the ‘purification’ refer simply to the cleansing modifications they have made to the original picture of the feverish polis (arguably the literal reading of the text), or does Socrates mean that their regulation of the arts (to be put into practice by the Guardians) will have the effect of cleansing the psyches of the state’s inhabitants, and purging them of their feverish desires? Or are both meanings in play? The issue matters, because if the second option is taken and Socrates is thought to be referring to the purificatory work of the Guardians, then this implies that the feverish, aggressive desires which lead to war are innate. On the first option, in which Socrates is simply discussing the purification of a picture, the psychological model that he is working with is left open, and we cannot necessarily infer that aggression is hard-wired into us: it may arise only from the wrong kind of culture and upbringing. On this first interpretation, unhealthy desires would not be innate in the citizens of the philosophic state and would have no chance to develop. and body is made clear at 546d (though they are here called ‘Guardians’, 546b makes it plain that it is only the Rulers that are meant), and there is no reason to suppose that this training is intended to be very different from that outlined in 379–399. ⁴² See p. 179 above for a discussion of these references. ⁴³ See p. 178 above.

Plato on War

191

The decision between these two psychological options is fundamental to our whole discussion. For if aggressive desires are innate, then it is open to an opponent of Socrates to say that Socrates is wrong even to aim for the purification of feverish desires: on the contrary, perhaps the ideal state needs to engage in war, including offensive war, as some sort of outlet for the innately aggressive desires of its citizens. A weaker version of this criticism might hold that if aggression is innate, then we should just accept that war will always be a necessary possibility in even the ideal state. One of the difficulties in deciding which psychological model Socrates has in mind is that in the event of there being only one ideally just state, war will still continue, as we have seen, and it is hard to determine the precise degree of freedom from aggressive tendencies in people who are actually fighting. The issue is considerably illuminated, however, if we consider what would happen if, hypothetically, all states on earth were to be ruled by philosophers: this is, after all, Socrates’ aspiration in his visionary speech on how to cure the ills of mankind at 473c–e. If one hypothetically removes the threat of war, and hence the need for a defence force, does Plato still think that even ideally educated appetites and thumoi will, or could, retain the aggressive impulses which can lead to war? This would be tantamount to saying that such aggressive impulses are both innate and ineducable. The desire actually to wage war (as opposed to a desire for unspecified aggression) cannot itself be innate, as war is a cultural phenomenon, and the desire for it has to be learnt;⁴⁴ but we might be innately predisposed to form desires for unspecified aggression, which could lead to a desire for war. So the ultimate questions of our whole discussion are: first, does Plato think that the belligerent (in the literal sense of ‘war-waging’) aspects of human nature stem from innately aggressive impulses or not? And, secondly, if he does, does he think that such innately aggressive impulses are educable? I submit that on most of the evidence of the Republic, particularly books II–IV, VIII and IX, Plato thinks it possible that such aggressive, and potentially bellicose, tendencies are not necessarily innate, and hence would not come to exist at all in an ideal world in which all states are ruled by philosophers, and all citizens are raised and educated according to the intellectual, emotional, and physical training that these Philosopher-Rulers devise. The two most obvious possible exceptions are 571a–2b, where Socrates says that some of the unnecessary desires are lawless and that we may perhaps (kinduneousi, 571b) all be born with them (eggignesthai), and 588b–9d, where the thumos is described as a lion and the appetites as a many-headed monster. Yet in the latter passage it is not entirely clear whether the image is supposed to represent the psych¯e of a new-born babe, or that of an adult (or at least a child) already infected by a corrupt society; while in the former the ‘kinduneousi’ shows that Socrates is not prepared to commit himself entirely to the notion that savage desires are innate in all of us, or indeed ⁴⁴ See Ferrill 1985: 11 and n. 1 above.

192

Angela Hobbs

in any of us.⁴⁵ Nor is it at all certain that ‘eggignesthai’ has to mean ‘innate’ in any case; it might just mean ‘occur naturally’ when certain conditions arise; and in a philosophically ruled world, those conditions may never arise.⁴⁶ In addition, we are not told that these lawless desires, even the ones which lead to dreams of committing murder, are necessarily the kind of desires with the potential to impel us to go to war (though of course some murders might provoke others to go to war against us). In short, there is no incontrovertible evidence in the Republic that aggression is innate, and there is the evidence of 431a–b that bad desires are the consequence of bad upbringing or bad company.⁴⁷ It is important to be precise here. My thesis that Plato does not think that aggression is necessarily innate (at least not in everyone) does not commit him to believing that one can ever eradicate, even in an ideal world, the basic selfassertive (thumoeidic) and acquisitive and pleasure-seeking (appetitive) impulses from the perversion and thwarting of which aggression and thence belligerence ultimately arise, when they do arise: as Republic 580d–581b makes very clear, the epithumiai (appetites) will always pursue physical satisfaction and material gain, and the thumos will always seek honour and (worldly) success, just as assuredly as reason will always desire truth and knowledge. And 611b–d makes it plain that so long as the psych¯e is embodied, the acquisitive appetites and status-regarding thumoeidic impulses will remain. Acquisitive, self-assertive, and hence potentially aggressive tendencies (‘aggressive’, but still not necessarily ‘belligerent’ in the strict sense) will always exist in embodied human nature, and will always be in need of careful and knowledgeable channelling and rechannelling, and, in the last resort, occasional suppression. But there is nothing, I maintain, in Plato’s theory of human nature as outlined in the Republic, especially books IV, VIII, and IX, which essentially and automatically precludes the possibility of such channelling, in the right global environment and with the right goals, role models, and rewards on offer: indeed, this is true even if one takes 571–2 as evidence for the existence of innate aggression, as Socrates says clearly in that passage that if reason rules us while we are awake, then such lawless desires will be quietened, even in our sleep. On my reading, according to which Plato believes that we are born not with innate aggression but with an innate potential for aggression, the consequence of his psychology is that this potentiality does not necessarily have to be realized, and ⁴⁵ A point highlighted by Deslauriers 2001: 234. In Deslauriers’s response to Barney’s (2001) paper, the question of whether Plato regards aggression as innate is one of the two main issues she raises. As will be clear from my discussion here, I am largely in agreement with Deslauriers on this matter. ⁴⁶ See Liddell, Scott, and Jones 1953: s.v. eggignomai (2): ‘of things, qualities, etc., spring up, appear in or among e.g. Plato Rep. 351d … Gorgias 526a’. ⁴⁷ Perhaps the strongest evidence in the Platonic corpus for innately fierce desires is Ti. 69c–d, where the ‘mortal soul’ is said to contain ‘terrible and necessary passions’ (deina kai anagkaia … path¯emata). Yet even here it is not obvious that these desires will inevitably include aggression, or even inevitably result in it, let alone result in war: Timaeus singles out pleasure, pain, rashness, fear, anger, hope, and lust.

Plato on War

193

in a cosmos run by philosophers, will not be realized. In our current imperfect world, of course, it almost certainly will be. The initial apparent contrast between Plato and Hobbes on the relation between civilization and war has thus turned out to be deceptive: Plato’s message is that civilization and war develop together only if civilization is already in the wrong hands; under the guidance of wise and beneficent philosophers with the power to educate and temper the psychological wellsprings of aggression, it is at least conceivable that there could be a civilization without (at least) offensive war, and perhaps even a cosmopolis without war at all. Furthermore, it is worth attempting such channelling and rechannelling of the roots of aggression and war even if one is sceptical about the possibility of total success. For even if the education of appetitive and spirited desires only diminishes, rather than entirely removes, the impulses which can eventually lead to war, then it would still be possible for humans, even in our compromised and radically uncertain world, at any rate to improve our chances of living at peace in a reasonably sophisticated political constitution, a constitution which contains both the (admittedly censored) arts and philosophy itself. And this possibility should, I submit, be cause for at least a little hope.⁴⁸ REFERENCES Adam, J. (1963) (ed.), The Republic of Plato, 2nd edn.. Cambridge. Annas, J. (1981), An Introduction to Plato’s Republic. Oxford. Baldry, H. C. (1952), ‘Who Invented the Golden Age?’, Classical Quarterly, n.s. 2: 83–92. Barker, E. (1918), Greek Political Theory. London. Barnes, J. (1987), Early Greek Philosophy. Harmondsworth. Barney, R. (2001), ‘Platonism, Moral Nostalgia and the City of Pigs’, Proceedings of the Boston Area Colloquium in Ancient Philosophy, 17: 207–27. Brandwood, L. (1976), A Word Index to Plato. Leeds. Brisson, L. (1970), ‘De la philosophie politique à l’épopée, le Critias de Platon’, Revue de métaphysique et de morale, 4: 402–38. (1995), ‘Interprétation du mythe du Politique’, in Rowe (1995), 349–63. Burnyeat, M. F. (1999), ‘Culture and Society in Plato’s Republic’, The Tanner Lectures on Human Values, 20: 215–324. Bury, R. G. (1926), Plato: Laws, trans. with introduction. Cambridge, Mass. Cambiano, G. (2002), ‘Catastrofi naturali e storia umana in Platone et Aristotele’, Rivista Storica Italiana, 114/3: 694–714. Craig, L. H. (1994), The War Lover: A Study of Plato’s Republic. Toronto. Crombie, I. M. (1962–3), An Examination of Plato’s Doctrines, 2 vols. London. Cross, R. C., and Woozley, A. D. (1964), Plato’s Republic: A Philosophical Commentary. London. ⁴⁸ It is interesting that Jonathan Shay 2000: 31–56, writing about the necessity or otherwise of aggression and war from a psychological perspective indebted to evolutionary biology, also reaches a conclusion of cautious optimism.

194

Angela Hobbs

Dawson, D. (1996), The Origins of Warfare in the Western World. Boulder, Colo. Deslauriers, M. (2001), ‘Commentary on Barney’, Proceedings of the Boston Colloquium in Ancient Philosophy, 17: 228–36. Dillon, J. M. (1992), ‘Plato and the Golden Age’, Hermathena, 153: 21–36. Ferrill, A. (1985), The Origins of War: From the Stone Age to Alexander the Great. New York. Guthrie, W. K. C. (1957), In the Beginning: Some Greek Views on the Origin of Life and the Early State of Man. London. (1975), A History of Greek Philosophy vol. 4. Cambridge. Hansen, M. H. (1993) (ed.), The Ancient Greek City-State, Acts of the Copenhagen Polis Centre, 1. Copenhagen. Hegel, G. W. F. (1967), Philosophy of Right, trans. T. M. Knox. Oxford. Hobbes, T. (1991), Leviathan, ed. R. Tuck. Cambridge. (1969), The Elements of Law, Natural and Politic, ed. F. Tönnies, 2nd edn. by M. M. Goldsmith. London. (1983), De cive: The Latin Version, ed. H. Warrender. Oxford. Hobbs, A. (2000), Plato and the Hero: Courage, Manliness and the Impersonal Good. Cambridge. Lane, M. S. (1998), Method and Politics in Plato’s Statesman. Cambridge. Lee, D. (1974), Plato: Republic, trans. with introduction. Harmondsworth. Liddell, H. G., Scott, R., and Jones, H. S. (1953), A Greek–English Lexicon, 9th edn. Oxford. Lloyd, A. B. (ed.), 1996, Battle in Antiquity. London. McCabe, M. M. (2000), Plato and his Predecessors. Cambridge. O’Connell, R. L. (1989), Of Arms and Men: A History of War, Weapons and Aggression. Oxford. Rowe, C. (1995) (ed.), Reading the Statesman: Proceedings of the Third Symposium Platonicum. Sankt Augustin. (2000), ‘The Politicus and Other Dialogues’, in Rowe and Schofield (2000), 233–57. Rowe, C., and Schofield, M. (2000), The Cambridge History of Greek and Roman Political Thought. Cambridge. Schofield, M. (1993), ‘Plato on the Economy’, in Hansen (1993), 183–96. (2000), ‘Approaching the Republic’, in Rowe and Schofield (2000), 190–232. Shay, J. (2000), ‘Killing Rage: physis or nomos —or Both?’, in van Wees (2000), 31–56. Vidal-Naquet, P. (1986), The Black Hunter. Baltimore. Wees, H. van (2000) (ed.), War and Violence in Ancient Greece. London. White, N. (1979), A Companion to Plato’s Republic. Oxford.

10 Language in the Cave Verity Harte

The analogy of the Cave is by far the most elaborate of Socrates’ three analogies, offered in lieu of an account of the form of the Good. Even so, his exposition of the analogy takes up scarcely more than three Stephanos pages (514a1–517a7), of which the first is truncated. This first page (514a1–515a3) is devoted to the mise en scène: indicating the general features of the cave and the situation of its prisoners. There follows a brief exploration of the depicted prisoners’ state of mind (515a5–c3), offered in explanation of Socrates’ claim that, even if the prisoners are strange, as Glaucon would have it (515a4), they are nevertheless ‘like us’ (515a5). This exploration takes the form of four questions and answers and a resumptive assertion at 515c1–3. The first two questions invite Glaucon to consider what he thinks the prisoners have seen (ἑωρακέναι, 515a6), first, of themselves and each other (515a5–b1) and, second, of the objects being carried along behind them (515b2–3). The remaining two questions invite Glaucon to draw consequences¹ from the answers to the first two questions, subject to reflection on two additional conditions: first, that the prisoners be envisaged as having the ability to talk to one another (515b4–6);² and second, that the cave should have an echo that the prisoners can hear (515b7–10). This invitation to It is a particular pleasure to include this paper in a collection in honour of Myles Burnyeat, trespassing as it does in areas on which he has written so magnificently: most obviously, the Republic and the cave, in particular; less obviously, but no less important in the background to this paper (see n. 39 below), the Theaetetus. The puzzles that motivate this paper reach back to some of my earliest conversations with Myles, as his M.Phil. student in Cambridge; this present essay upon them benefits, most recently, from typically astute and fruitful questions of his in response to an ancestor of a related paper of mine (Harte 2006). An earlier version of the present paper was given at the Leventis Foundation Conference on ‘The Good and the Form of the Good in Plato’s Republic’, at Edinburgh University, 2–5 March 2005. I am grateful to the audience on this occasion for helpful discussion. For written comments and/or further discussion, my particular thanks also to Peter Adamson, Ursula Coope, Gail Fine, M. M. McCabe, Stephen Menn, Terry Penner, Frisbee Sheffield, and Mark Textor. ¹ Note the inferential particle οὖν, 515b4. ² Bestor (1996) has argued forcefully that it is implausible to suppose that the prisoners, given their situation, could have language, and thus that Plato must have an alien approach to the foundations of language. It is unclear to me whether the details of what is, after all, an analogy are meant to stand up to the sort of scrutiny Bestor gives them as to the plausibility of the supposition

196

Verity Harte

draw consequences continues their exploration of the depicted prisoners’ state of mind by considering the nature of their use and understanding of language. This is the focus of the present paper: the prisoners’ use and understanding of language; what this reveals about their state of mind; and what, in turn, this contributes to an explanation of the way in which these prisoners are ‘like us’. In particular, the paper is focused on the three lines in which Socrates and Glaucon consider what would be true of the prisoners if they could talk (515b4–6). I . 5 1 5 B 4 – 6 : T H E P R I S O N E R S ’ S I T UAT I O N , I F T H EY C O U L D TA L K At 515b4–5, Socrates invites Glaucon to agree to a specific account of what would be true of the prisoners’ use of language, given the situation they are in, if they were able to talk to each other. The account to which he invites him to agree, or so I shall argue, is this. The prisoners are equipped with a vocabulary, whose terms—at least in the mouth of a knowledgeable speaker—refer to what I shall call simply ‘real things’. (This claim about their vocabulary is not yet controversial, I take it, with the proviso.) In the terms of the analogy, they refer to the objects, manufactured copies of which cast shadows on the wall of the cave that the prisoners face. They refer, that is, to the objects the analogy depicts as being outside the cave, which are analogues, I take it, of Forms.³ If you asked a prisoner to point to the object to which a term he uses refers, he would point to one of the shadows on the wall of the cave. Nevertheless, I shall argue, the prisoners’ terms, in the prisoners’ usage, correctly refer—unbeknownst to the prisoners themselves—to the objects and not to the shadows. (This is where the controversy begins.) This is the first feature of the prisoners’ condition as regards the use of language to which I wish to draw attention. The prisoners, in talking and using certain names or terms, refer to real things (the objects models of which cast the shadows). They do so despite the fact that they do not take themselves to do so; they themselves would point to the shadows (that is, to what in fact are shadows)⁴ as that to which they refer. And they do so despite the fact that there that the prisoners use language. In any case, his arguments are, by and large, orthogonal to the matters on which I will be focusing. ³ In general, my discussion will be framed within the terms of the analogy and will not get embroiled in the controversial topic of what the various features of the analogy correspond to. This much, however, and that the focus is (at least, mainly) on values (cf. Resp. 517d4–e1), I take to be uncontroversial. ⁴ This parenthetical clarification should be borne in mind throughout. Given their situation as I interpret it, the prisoners do not have (de dicto) thoughts about the shadows as shadows. Is there then a question, regarding their relation to the shadows, as to how the prisoners accomplish such demonstrative acts as, e.g., pointing at them? Such acts are possible for them, I take it, because the shadows—what are in fact shadows—are given to them directly in perception.

Language in the Cave

197

is no evidence of (nothing in the situation depicted in the analogy to represent) their having any sort of experience of the real things (experience that would seem that way to them).⁵ Their physical situation as represented in the analogy precludes their having been, or seen, anything outside the cave. Indeed, their experience, as so far described—what they have seen, as regards themselves, each other and anything else—consists simply of the shadows (515a6–7; implicitly again at 515b2).⁶ This first feature of the prisoners’ condition requires some defence as an interpretation.⁷

Defending the interpretation: the text and context of Resp. 515b4–5 A first line of defence concerns the text of 515b4–5, whose reading is problematic. Here, for comparison, are the textual variants that appear in the editions of Adam, Burnet, and Slings: (Adam 1902:, v. 2 ad loc.): Εἰ οὖν διαλέγεσθαι οἷοί τ’ εἶεν πρὸς ἀλλήλους, οὐ ταῦτα ἡγεῖ ἂν τὰ παριόντα αὐτοὺς νομίζειν ὀνομάζειν ἅπερ ὁρῷεν; (Burnet 1902:, v. 4 ad loc.): Εἰ οὖν διαλέγεσθαι οἷοί τ’ εἶεν πρὸς ἀλλήλους, οὐ ταῦτα ἡγῇ ἂν τὰ ὄντα αὐτοὺς νομίζειν ἅπερ ὁρῷεν; (Slings 2003:, ad loc.): Εἰ οὖν διαλέγεσθαι οἷοί τ’ εἶεν πρὸς ἀλλήλους, οὐ ταῦτα ἡγῇ ἂν τὰ ὄντα αὐτοὺς ὀνομάζειν ἅπερ ὁρῳεν;

Slings’s apparatus reads as follows: b5 ὄντα Iambl. Procl.: παρόντα ADF: παρίοντα Laur. 80. 7 pc ὀνομάζειν Iambl.: νομίζειν ὀνομάζειν AD: νομίζειν F Procl. Adam’s text is the best suited to the interpretation proposed. It is the text explicitly adopted by the translation of Grube–Reeve, who translate as follows: ‘And if they could talk to one another, don’t you think they’d suppose that the names they used applied to the things they see passing before them?’⁸ This text, translated thus, gives the interpretation proposed, on the plausible assumption that, in context, we are also invited to understand that the prisoners’ supposition is wrong: the names they use in fact apply to things other than the things they ⁵ The prisoners have experience (of the shadows) for which the real things—assuming these are what the models are models of—are in some way ultimately responsible. But it is a moot question whether the experience they do have (of the shadows) itself counts as experience of (what are, as a matter of fact) the real things. It is, I take it, uncontroversial that it would not seem that way to the prisoners themselves. ⁶ I take no stand here on how one should understand the claim about their experience of themselves; for discussion, contrast Burnyeat 1999: 238–43 with Brunschwig 2004. ⁷ The defence involves detailed discussion of the text in Greek. Those without Greek may wish to forgo the defence (and hence, perforce, to take the interpretation as given, at least in reading the remainder of the paper) and skip ahead to the next subsection entitled ‘Back to the prisoners’ situation in general’. ⁸ Reeve 1992, ad loc., with footnote.

198

Verity Harte

see passing before them, the real things of which these are shadows of models. The same text appears to underlie the translation of Griffith (in Ferrari 2000, ad loc.). The text is explicitly endorsed and given an interpretation I take to be like mine in a typically terse, but loaded footnote by Owen (1973: 144–5 n. 20). Burnet’s text is found in MS F and supported by an inference from the reading of Proclus. It is, from the point of view of my interpretation, the worst, since it drops all direct reference to language and naming. It would translate, I take it, as follows: ‘And if they could talk to one another, don’t you think they would take as beings these very things they see?’⁹ This is a perfectly plausible view of a claim to which Glaucon might here be invited to agree. However, precisely because it drops all reference to language and naming, this text seems to me worst from the point of view of likely sense, on grounds independent of my preferred reading. Without reference to language in the consequent of the condition, it is hard to see the point of its antecedent ‘if they could talk’. They could presumably all take things to be thus and so, in their own minds, whether or not they could talk to each other. Slings’s text has no direct MS authority, but is an inference from the reading of Iamblichus. I have not been able to discover his reason for its adoption.¹⁰ The interpretation I propose could be obtained from this text, with a bit of work: take τὰ ὄντα as a shorthand way to refer to the (real) bearers of the names that the prisoners use, standing as proxy for the names, and translate: ‘And if they could talk to one another, don’t you think they would name these very things they see the things there are [sc. with the names that are in fact names for the things there are]?’ The mistake this involves as to their actual reference would be supplied by the context. The naming in view (on this or any other reading) would be, not acts of christening, but general use. Adam’s νομίζειν ὀνομάζειν has the benefit of building the implication of error directly into the sentence; and has the authority for its occurrence in two of the three most important manuscripts, including the one to which Slings (and others) attribute greatest authority (A).¹¹ Further, the occurrence of two infinitives ending in –ζειν makes it easy to see how one or other might be omitted in the process of copying.¹² The weak point in Adam’s text is παριόντα. Practically all the manuscripts have παρόντα. For my purposes, there is less at stake here: the central interpretative point goes through whether the prisoners erroneously think that they name the passing things they see (reading παριόντα) or the present things they see (reading παρόντα). The reading παρόντα, however, ⁹ LSJ, s.v. νομίζω II. 1 and cf. 515c2 for comparable use. ¹⁰ It is not among the passages commented on by Slings 2005. See, however, Boter 1989: 76, 237, and n. 12 below. ¹¹ See preface to Slings 2003: esp. p. ix. ¹² By the same token, as Stephen Menn suggests to me (pers. comm.), one could explain the duplication as the incorporation of a variant reading. Boter (1989: 76) speculates along these lines, but without considering the philosophical merits of the apparent duplication.

Language in the Cave

199

is one that none of these editors favour. Adam’s variant—from which παρόντα is an easy corruption—is found, but only in a late Florentine manuscript. It also has parallels in the passage elsewhere. Related participles turn up in the immediately succeeding context, at 515b8, 515a9, and 515d5. Each has a different application: to the people carrying things along, to the shadow passing along, and to the manufactured objects being carried along. This variety of application in the succeeding occurrences means that any difference between their reference and that of the putative occurrence at 515b5 need not trouble us. As regards the reading of the text, then, the interpretation of the prisoners’ situation put forward, while it is not forced, is given considerable support. A second line of defence, both for the reading and for the interpretation, comes from consideration of the parallel to be established between the question and answer at 515b4–6 and its companion at 515b7–10. These two questions and answers form a natural pair. Of the four questions and answers by means of which Socrates explores the prisoners’ state of mind in 515a5–b10, and whose conclusion he resumes, at 515c1–2, this latter pair together explore the consequences of the answers to the first questions as to what the prisoners have seen.¹³ They do so, in each case, under an additional assumption, assumptions that complement each other. In the first question and answer, the additional assumption is that the prisoners can talk; they can make use of language. In the second, it is that the cave has an echo; they can hear language in use. Here is Slings’s text and a translation of the second question and answer, designed to beg as few questions as possible: Τί δ´εἰ καὶ ἠχὼ τὸ δεσμωτήριον ἐκ τοῦ καταντικρὺ ἔχοι; ὁπότε τις τῶν παριόντων φθέγξαιτο, οἴει ἂν ἄλλο τι αὐτοὺς ἡγεῖσθαι τὸ φθεγγόμενον ἢ τὴν παριοῦσαν σκιάν; ‘And too, what if the prison had an echo from the facing wall? Whenever one of those passing along uttered a sound, do you think that they [the prisoners] would think what sounded were anything other than the passing shadow?’ (515b7–10)

Those ‘passing along’ I take to be the people carrying the objects whose shadows are cast. We have already been told (515a2–3) that some of these are uttering sounds, while some are silent. It is their sound that echoes off the facing wall and, Glaucon is invited to agree, that the prisoners will erroneously take to originate from the shadows. The view that this sound is (or includes) linguistic sound is not forced by the verb used both at 515a2 and here at 515b9 (φθέγγομαι). But this verb is certainly used for linguistic sounds, when applied to the sounds that are made by people.¹⁴ The possibility of its including sounds other than linguistic sounds may be designed to accommodate what might otherwise seem like a weakness in the analogy. If the only sounds the prisoners were to hear—the sounds made by themselves and by these carriers—were human linguistic sounds, we would be forced to conclude that they would attribute either no sound or ¹³ Again, note οὖν, at 515b4.

¹⁴ LSJ, s.v. φθέγγομαι.

200

Verity Harte

human linguistic sound to shadows that include shadows of models of animals and of inanimate objects. Use of this verb allows that the sounds whose echo they hear include sounds appropriate to both animals and inanimate objects.¹⁵ Of course, no analogy is perfect: Socrates could have envisaged the models themselves as making noises;¹⁶ and if the noises the carriers make do have this range, then we must envisage that they, on occasion, imitate animals. These wrinkles aside, the main point is clear: following the question about the prisoners’ own use of language, this second question includes a corresponding question about their reception of language. With this in mind, consider how close is the parallel between the questions, if we take Adam’s reading of the first question and the interpretation of it that I have proposed. In each question, Socrates asks Glaucon what he thinks the prisoners think, so that there are two verbs of thinking in all (οἴει … αὐτοὺς ἡγεῖσθαι, at 515b8–9; ἡγῇ … αὐτοὺς νομίζειν, on the Adam reading of 515b5). This repeating pattern itself offers some additional support for Adam’s reading of the double infinitive in b5. On the interpretation proposed, the first question invites Glaucon to agree that the prisoners (erroneously) think that they name the passing shadows they see (τὰ παριόντα … ἅπερ ὁρῷεν). Given this interpretation, the second question makes the corresponding point about sound. Whereas the first question is asked in such a way as to invite the answer ‘yes’, the second invites the answer ‘no’. This variation aside, the parallel is complete: Glaucon is invited to say whether he thinks that the prisoners would think that the source of the sound that they hear is anything but the passing shadow (τὴν παριοῦσαν σκιάν). In each question, Glaucon is invited to agree that the prisoners mistakenly identify the shadows as what they name/hear, whereas what they in fact name/hear is the actual bearer of the name/maker of the sound.

Back to the prisoners’ situation in general In defence of my interpretation of the prisoners’ situation so far, I have argued that both the text of 515b4–5 (especially on what seems the most favourable reading of it, which I will be referring to as ‘Adam’s reading’) and its immediate context may be taken to support the first feature of the prisoners’ condition as regards the use of language to which I have drawn attention. This was that the prisoners’ terms, in the prisoners’ usage, refer, unbeknownst to them, to their proper referents: the objects, not the shadows. There is also a second feature of the prisoners’ condition to which I wish to draw attention. Whereas the first concerned the de facto reference of the prisoners’ terms, whatever their own ¹⁵ Again LSJ, s.v. φθέγγομαι. LSJ’s survey of uses and itemization of senses suggest that the term’s range is precisely well suited to capture the characteristic sounds associated with people (where this is vocal, linguistic sound), animals (animal calls), and inanimate objects. ¹⁶ But perhaps the idea of artefacts that make a sound (not instruments, but, e.g., models of animals that make the noise of the corresponding animal) is a relatively modern one.

Language in the Cave

201

confusion regarding this matter, the second concerns the relation between the de facto referents of their terms and the items that the prisoners would be mistakenly inclined to point to as the referents of their terms: the shadows. Although the prisoners have mistaken views about the reference of the terms they use, and hence make any number of mistakes about the nature of the objects to which they in fact refer,¹⁷ it would appear that, by and large, when they use a term—the term ‘ox’, for example—under the mistaken impression that the thing they are talking about is one of the shadows on the wall, then the shadow in question is as a matter of fact a shadow of a model of an ox, as opposed to a shadow of a model of some other animal. The second feature of the prisoners’ situation to which I wish to call attention, then, is the systematic correspondence between the items the prisoners would be mistakenly inclined to point to as the referents of the terms that they use and the actual referents of the terms that they use. By and large, the former turn out to be shadows of models of the latter. That this is true, by and large, seems to me to follow from the way that the cave analogy is set up. That it is only true by and large —that, on some occasion, some individual prisoner might use the term ‘ox’ with the mistaken impression that it refers to the shadow of a model of a horse—seems to me the implication of the sort of competitions that the prisoners are said to have amongst themselves. At 516c8–d2, Socrates imagines that the cave-dwellers might give out honours and prizes to the person who is quickest to observe the passing shadows and on the basis of a memory of their customary sequence is best able to predict what will come next. The quickness of observation at issue, I take it, is a quickness to identify, not merely to see.¹⁸ And there is point to such competitions only if people sometimes get things wrong. Three passages here benefit from being taken closely together and in a way that may offer yet further support for Adam’s reading of 515b4–5. To begin with 516c9: prizes are awarded—amongst other things—for being the quickest, I say, to identify ‘the things passing’, that is, the shadows. The expression ‘the things passing’, τὰ παριόντα, at c9, is an exact parallel for Adam’s reading of 515b5, and gives a particular slant to the sort of conversational naming that the prisoners’ envisaged activity may be thought to involve. If the prisoners’ envisaged talking is focused on identifying the things that pass before them—giving their names, saying what they are—then we can pick up on the resonance, in the Republic especially, of Socrates’ use of the verb διαλέγεσθαι, at 515b4, to characterize their ¹⁷ Two different sorts of mistake are at issue here: (i) if you ask a prisoner to point to the referent of ‘ox’, they will point to the wrong thing; (ii) the prisoners have a mistaken conception of the underlying nature of ox. These two types of mistake stand in no necessary relation. For the prisoners, however, the reason why they have a false belief about the nature of ox is their mistaken belief that the item they refer to as ‘ox’ is something that is in fact merely a shadow of a model of the actual referent. ¹⁸ καθορᾶν (516c9) at least indicates some kind of intensive study beyond the simple act of seeing.

202

Verity Harte

putative talking: this verb is used for conversation, but also for the specifically philosophical conversation that is dialectic. This, then, is the sort of conversation one might expect to have with Socrates, the sort of conversation directly alluded to when, at 515d5–7, we again envisage a prisoner being asked to identify ‘the things passing’ (τὰ παριόντα), which are now, not the shadows, but the models that cast them. This prisoner, forcibly removed from his seat, is compelled, on being questioned, to answer what each of them is and finds himself in the familiar aporia of those who engage in such conversation with Socrates (515d6). Taken together, then, we have the following picture of the prisoners’ situation, as regards language use. (1) The terms the prisoners use—and in their usage—correctly refer to the objects models of which cast the shadows. But the prisoners have a mistaken conception of the referents of the terms, supposing them to refer to what are in fact shadows. (2) There is, nevertheless, a systematic correspondence between the items they mistakenly suppose to be the referents of their terms and their actual referents. By and large, the item they erroneously think they refer to is a shadow of a model of their actual referent. I I . T H E P U Z Z L I N G C H A R AC T E R O F T H E P R I S O N E R S ’ S I T UAT I O N The prisoners’ situation as regards their use of language is puzzling. We should, I think, be surprised by both features of the prisoners’ use of language to which I have drawn attention. From a perspective external to that of the prisoners themselves, their language use involves a measure of ‘success’, albeit success amidst a mass of confusion, and it is this measure of success that should strike us as puzzling. Consider, first, the striking fact that the prisoners, in talking, successfully refer to what—from this external perspective—are the ‘proper’ referents of the terms that they use, but do so unbeknownst to themselves. There is no obvious story to explain this apparent success of their reference. Ex hypothesi, the prisoners have had no direct contact with the real things to which they refer and seem entirely unaware of their existence. Nor is there any provision within the analogy for them to have been introduced to the terms they use by an authoritative or knowledgeable user, such that we might use this introduction as the basis for the success of their reference.¹⁹ Further, the prisoners are portrayed—however implausibly²⁰—as though, all by themselves, they constitute an active linguistic community, such that it is a surprise that they are not themselves taken to be straightforwardly authoritative about the reference of the terms they use. ¹⁹ In sect. III, I will consider various possibilities for explaining the success of their reference. ²⁰ Cf. n. 2 above.

Language in the Cave

203

The puzzling character of this feature of the prisoners’ condition may be brought out by contrasting the prisoners with the denizens of two rather more recent thought experiments with implications for reference, each put forward by Hilary Putnam. Begin with his brain-in-a-vat hypothesis.²¹ Putnam invites us to a consider a world in which every occupant is a brain in a vat: every one consists of a brain, like the brain of an ordinary person, kept in a vat of nutrients and fed stimuli through its nerve endings from a computer such that each brain has ‘experiences’ which might be exactly phenomenologically like the experiences we are having at this very moment. And these brains in a vat might have a vocabulary just like ours, which they deploy in a way that corresponds, mutatis mutandis, to the way in which we deploy ours. Prima facie, Plato’s prisoners are in a condition very like that of Putnam’s brains. But this prima facie likeness is an illusion, and one that helps to bring out the surprising character of the prisoners’ small measure of success. In the development of the brain-in-a-vat hypothesis, Putnam makes a very different decision about reference. Indeed, this is his central point. He asks whether, on the hypothesis of a world made up of brains in a vat, these brains in a vat could say or think that they are brains in a vat; and he argues that they could not, not even when formulating a ‘thought’ of the form we are brains in a vat. For they cannot, he argues, refer to what we can refer to.²² Put the point another way: if we asked Socrates’ question at 515b4–5 (on Adam’s reading) of the brains in a vat—if they could somehow point to what they take their terms to refer to—the brains, unlike the prisoners, would be right. The contrast between the prisoners and the brains in a vat both underlines the mystery of the prisoners’ contrasting success and explains its importance for them: the prisoners in the cave—unlike the brains in a vat—can recognize, and hence can seek to change, their condition. The prisoners are also unlike the residents of another hypothetical world put forward by Putnam: Twin Earth.²³ Twin Earthlings use the term ‘water’ for a substance whose chemical constitution differs radically from that of water on Earth, but which is otherwise indistinguishable from water on Earth. Where the prisoners, we might imagine, have a conception of the referent of their terms which is in many ways very unlike the conception of a knowledgeable user, Twin Earthlings—at least the non-scientists among them—have a conception of the referent of their term ‘water’ pretty much exactly like that of a non-scientific Earthling user of their own term ‘water’. Nevertheless, Twin Earthlings differ, as having a different reference for the parallel term. The prisoners are most like the occupants of Earth (not Twin Earth) c.1750 and prior to the discovery of the chemical composition of water. Putnam argues ²¹ Putnam 1981: ch. 1. ²² Putnam 1981: 8. ²³ In, e.g., ‘The Meaning of ‘‘Meaning’’ ’, in Putnam 1975: ch. 12.

204

Verity Harte

that these eighteenth-century Earthlings use the term ‘water’ successfully to refer to water on Earth. They do so despite their apparent similarity to Twin Earthlings of a comparable era (prior to the discovery of the chemical composition of Twin Earth water). And they would continue to do so (at least in the short term) if, unbeknownst to them, they were transported to Twin Earth and, when asked to point to the water they mean, would point to the Twin Earth water around them. Like the prisoners, the eighteenth century Earthlings combine successful reference with lack of knowledge of the nature of the substance to which they refer (and, in the switching case, with a local confusion as to which is the object to which they refer). But their relation to the de facto referent of their term is considerably more direct than that of the prisoners, for whom the relation is mediated by shadows and models (a fact I shall return to later). Putnam’s cases bring out the surprise in the fact that, despite being unaware of their doing so, the prisoners—unlike the brains or the Twin Earthlings, but like the eighteenth-century Earthlings in the switching cases—successfully refer to the actual referents of the terms they use, and not to their shadowy counterparts. No less surprising is the apparently reliable correspondence between these actual referents and the shadows to which the prisoners would be mistakenly inclined to point as their referents. While we can easily explain why the prisoners would think that the referents of their terms were the shadows, we cannot so easily explain their lighting on the very shadow that is a shadow of a model of the actual referent, and no other. In particular, we cannot explain the prisoners’ mistaken inclination to think that the referent of a term they use is a shadow of a model of its actual referent by appealing to the relation between referent and shadow. We cannot do so, because this gets things the wrong way round. Ex hypothesi, the prisoners in the cave have no experience of the actual referents, the originals to the shadows. Hence there is no obvious reason why, in using the terms they do, and in having a mistaken impression that their referents are the passing shadows, it should happen that they pick on the shadow that is as a matter of fact a shadow of a model of the actual referent. (There is more to be said here, and I shall return to it later.) The prisoners’ situation is puzzling, but it is not absurd. (Outright absurdity would constitute a powerful objection to my interpretation of their situation, despite the defence of sect. I.) As we have seen, their situation is comparable to that of eighteenth-century Earthlings in Putnam’s switching cases. These Earthlings use the term ‘water’ to refer to Earth water, despite their ignorance as to its chemical nature, and, Putnam argues, they would continue to do so even if, unbeknownst to themselves, they were to be transferred to Twin Earth, so that the substance to which they might mistakenly point as their referent would be the Twin Earth counterpart of water, and not their actual referent, water on Earth. Putnam’s views about the nature of reference in the envisaged scenario

Language in the Cave

205

are not uncontroversial, but the situation he envisages has not been rejected as outright absurd.²⁴ Putnam’s switching cases provide the nearest comparison known to me for the prisoners’ situation, as I have described it. There is, however, one important element of their situation that supports the drawing of further comparisons and provides a context in which to consider what resources we may draw on to resolve the puzzling character of their situation. Like Putnam’s eighteenth-century Earthlings, the prisoners combine reference with ignorance of the (essential, identifying) nature of that to which they refer. The view that reference is possible, in combination with ignorance of this sort, is maintained by, amongst others, Putnam and Tyler Burge. Burge (1991) also attributes this view to Frege. According to Burge, Frege had the view that, even the most competent users of, for example, a mathematical expression—mathematicians of earlier centuries—might have an unclear or incomplete grasp of the expression’s sense. This goes along with a model of mathematical or scientific progress according to which ‘[p]rogress is a matter of obtaining a better, clearer grasp of thoughts that one is already dimly thinking and unperspicuously expressing. Better theory results in deeper understanding and clearer explication of some of one’s own thoughts and senses.’²⁵ This idea—that, especially in the case of technical or scientific expressions, it is possible to refer to something, to think and talk about it, whilst suffering from various conceptual confusions about that about which one thinks and talks—underlies some central features of Burge’s own views about mental content, which depend on the claim that our beliefs, as he puts it, may, in a very wide number of cases, be ‘infected by incomplete understanding’.²⁶ His example is a person who thinks (correctly) that he has arthritis in his fingers, wrists, and ankles, and who, in addition, thinks (evidently, incorrectly) that he has developed arthritis in his thigh. At least in certain circumstances,²⁷ Burge argues, we may attribute to this person thoughts about arthritis —despite their misunderstanding of it as a disease that can sometimes occur outside of joints—and not thoughts about ‘tharthritis’, a term Burge suggests we might coin to cover both arthritis and whatever condition affects this person’s thigh.²⁸ Like the Burge and Putnam cases, Plato’s prisoners talk about things of which they are (in their case, apparently quite radically)²⁹ ignorant. But there are (at least) two important differences between the prisoners’ situation and, ²⁴ For criticism of Putnam, see Segal 2000. ²⁵ Burge 1991: 42. According to Burge, this is not Frege’s only model of progress; sometimes old concepts are replaced, together with their determining senses; but the two models are, Burge argues, ‘complementary’ (pp. 42–3). ²⁶ Burge 1991: 79. ²⁷ This is an important qualification, and one to which I shall return below. ²⁸ Burge 1991: passim; for ‘tharthritis’, see pp. 94–5. ²⁹ Below, I address the question of what, if any, true thoughts the prisoners might be thought to have about their actual referents (and how they might have come by them). It is a moot question how much ignorance reference can survive. Nevertheless, since the scope of the prisoners’ ignorance

206

Verity Harte

for example, that of the person in Burge’s arthritis example. These differences complicate, but do not, I think, undermine the broader comparison. The first difference is that Plato’s prisoners are not only ignorant about the nature of that to which they refer, they are also subject to confusion as to which item in their environment is the item to which they refer. This feature of the prisoners’ situation finds comparison only in Putnam’s Twin Earth switching examples, which is the reason why these examples provide the closest comparison for the prisoners’ situation. This first difference makes it more puzzling that reference to the object about which they are ignorant should be held to obtain; puzzling, but, as we have seen, not absurd. The second difference builds on the first, and concerns the two apparently competing referents in Plato’s scenario: the real things to which the prisoners in fact refer and the shadows of their models to which the prisoners mistakenly think they refer. Thus far, I have presented the situation as involving a stark choice: either the terms refer (and, from an external perspective, successfully) to the real things or (as the prisoners themselves think) they refer to the shadows of models. The Putnam and Burge cases are set up starkly in this way: eighteenth-century Earthlings refer either to Earth water or to Twin Earth water; one or the other, not both. It is possible, however, that the situation that Plato envisages is not so stark. We might think of the prisoners’ sentences as having the form ‘… is F’.³⁰ Invited to point to an occupant of the subject position in this sentence, a prisoner will point to a shadow. This shadow is not (strictly and properly) F. But, since it is a shadow of a model of what is strictly and properly F, one can say of it derivatively that it ‘is F’. Now, when the prisoners point to a shadow as ‘being F’, they presumably think that this shadow is what, strictly and properly speaking, ‘is F’ (whether or not they make such thoughts explicit). In this way, one might recast my account of their situation as follows. The terms that the prisoners use, in their usage, strictly and properly refer to the real things and only derivatively to the shadows. The prisoners mistakenly think that the terms that they use refer strictly and properly (indeed exclusively) to what are in fact shadows of models of the strict and proper referents. Support for this view of the prisoners’ situation might be drawn from what Socrates says about a prisoner forced to rise up from his seat, turn and try, while still dazzled by the light of the fire, to make out (καθορᾶν) the models that cast the shadows (515c4–d2). Socrates invites Glaucon to consider what such a prisoner would say ‘if someone were to say to him that, whereas then he saw nonsense (φλυαρίας), now, having been turned closer to what is and towards remains unclear, but the fact of their reference, on the interpretation defended, is clear, their ignorance should not be taken to call the reference into question. ³⁰ Is this a statement of identity or a predicative statement? I’m not sure whether this distinction should be taken to apply here, but it seems to me that the driving focus is on the identification of the nature of things.

Language in the Cave

207

what more are, he should see more correctly (ὀρθότερον βλέποι)’ (515d2–4). He imagines such a prisoner having each of the passing models pointed out to him and being asked what they are (515d4–6). ‘Don’t you think,’ he says, ‘that he would be at a loss (ἀπορεῖν) and would think that the things then seen were more true/real (ἀληθέστερα) than the things now shown?’ (515d6–7). Introduced to the models of which he then saw shadows, our forcibly turned prisoner is expected to be able to have a go at saying what they are. He took the shadows to be, respectively, let us say, F, G, H. He is invited to see that the models (as well or instead) are, respectively, F, G, H. The point to be noted about this passage is its use of comparatives. The prisoner, in being invited to identify the models as, respectively, F, G, H, is invited to think of himself as seeing more correctly (515d3) than before. He himself will think of the things seen before as being more true or more real than what he is now shown (515d7). Neither side presents the situation as one in which one set of judgements or observations as to what is F, G, or H eradicates, rather than simply downgrades, the other. This opens up a possibility in line with the revised view of the prisoners’ situation suggested above: that identification as F, G, or H, has some truth or correctness at every level, but that, contrary to the forcibly turned prisoner’s initial mistaken judgement, correctness improves as one moves away from the original imprisoned condition.³¹ On the revised view of the prisoners’ use of terms, it does not involve the stark choice implied in the way I had previously presented their situation. But this should not be thought to diminish the puzzling character of their situation. The prisoners are still referring, strictly and properly speaking, to real Fs, while being ignorant of them. Such improvement in their situation as there may be concerns their relation to the shadows. In talking of Fs, and referring, strictly and properly speaking, to the real Fs of which they do not have knowledge, the prisoners mistakenly think of themselves as referring to the corresponding shadow. Since, on this view, the shadow can, albeit in a derivative manner, be referred to as F, their mistake about their referents seems partly relieved. But this, I think, is not something in which our prisoner should take much comfort. Since the success of a reference to a shadow of a model of F derives from its strict and proper reference to real Fs, one can only knowledgeably refer to a shadow as F if one does so with knowledge of its being a shadow of a model of what is strictly and properly the referent of F, and this is knowledge that our prisoner is clearly lacking. Thus, since the revision merely complicates, without relieving, the puzzling character of the prisoners’ situation, in talking of their situation, I shall continue to set such complications aside. ³¹ There is an alternative possible explanation of these comparatives. Since the models, no less than the shadows, are not the real things themselves, then a prisoner’s observation and identification of the models can be considered an improvement in their condition, but they are still variously mistaken. On this view, the comparatives indicate a condition closer to, but not yet arrived at, correct vision and truth.

208

Verity Harte

I I I . E X P L A I N I N G T H E P R I S O N E R S ’ S I T UAT I O N Since the situation of the prisoners, on the interpretation proposed, though not absurd, is certainly puzzling, it invites explanation. (The more or less comparable cases of Putnam and Burge are offered as inviting explanation; indeed, they are offered in the course of proposing and defending particular sorts of explanation.) What is the explanation of the prisoners’ reference to the (strict and proper) referents of the terms that they use—the real things—despite their ignorance of the fact that they do so? And what is the explanation of the fact that, in mistakenly thinking that they refer to one of the shadows, it happens that the shadow in question is a shadow of a model of their actual referent? One might think that we could unite these two puzzling aspects of the prisoners’ situation in such a way as to build up some kind of story as to what may explain them. Begin with a prisoner’s use of ‘ox’, which in fact refers to whatever is really ox, with the mistaken impression that it refers to a shadow of a model of an ox. Now, the shadow of a model of an ox stands in a distinctive relation to a real ox. The shadow of a model of an ox has some features that are explained by it being an ox that it’s a shadow of a model of, and not a hippopotamus, for example: one such feature might be its observable shape. In this respect, one might say that the shadow of a model of an ox has some aspect of the character of real ox, although it is, of course, not a real ox. Not being a real ox, it also has some features that are not explained by its being the shadow of a model of an ox. It has, for example, whatever features are generic to shadows as such. And it has some features that are somewhat distorted versions of features found in real ox. An example might be its visible size—the size of a shadow is only partly a function of the size of the object whose shadow it is, being a function also of the relation between shadow-casting object and light source.³² One might seek to ground the prisoners’ reference to real ox through the occurrence of these ox features within the shadow that they are mistakenly inclined to take to be the referent of their term ‘ox’. At the same time, the partial character of the occurrence of features in the shadow and the inevitable distortion would explain the extreme inadequacy of their conception of ox. Nevertheless, the prisoners refer to ox via the ox features of what they mistakenly take to be the referent of the term ‘ox’. However, while this may be the right way to indicate the ‘tracks’, as it were, on which their reference runs, we should not be deceived into thinking that this has given us an explanation of how the reference comes about. That we have no such explanation here is what the second puzzling feature of the prisoners’ ³² This is complicated further by the fact that the shadows in question are shadows of models of the real things, which models could all be roughly the same size. Let us imagine that the models reflect differences in size amongst the real things.

Language in the Cave

209

situation brings out—the systematic correspondence between the actual referents of their terms and what they mistakenly think to be their referents. We have no explanation of why it should be that it’s the shadow with ox features that our prisoners should mistakenly take to be the referent of their term ‘ox’. Having no explanation of this, we have no explanation of the fact that this shadow is merely a shadow of a model of their actual referent, and not their actual referent itself. My own proposal is that, in order to explain both of the puzzling features of the prisoners’ situation, we must assume that the prisoners start off with some cognitive grip on the actual referents of their terms, however they may be thought to have obtained it. This cognitive grip both secures the reference to the real things of which they appear to be utterly ignorant and explains the fact that, by and large, in the distorting context of their environment, they mistakenly take to be the referent of their term ‘ox’ what happens to be the only available item having (some) relevant features, albeit they are unaware of its reason for doing so. It is important to see which way round this goes on my story. Suppose one thought that the prisoners could be shown to have some cognitive grip on some of the features that, in fact, characterize ox. There would be no need to tell the sort of story I am telling just to explain their possession of such a grip. Their possession of an awareness of some such features could be explained by their contact with shadows of models of ox and the possession of these features by the shadows in question. What this contact cannot in and of itself explain is the prisoners themselves thinking of such features as being features of ox. Once again, it’s the second puzzling aspect of their situation that makes this clear. And this is why my story goes the other way around. It’s the occurrence within the prisoners of some cognitive grip on the actual referent of their term ‘ox’ that explains both why their term does in fact refer to ox (and not, for example, to ox-shadow) and why it should turn out to be a shadow of a model of an ox that they mistakenly think it refers to. I offer two sorts of argument in favour of my proposal: an argument through elimination of obvious alternatives, and an argument as to this being a plausible Platonic strategy for resolving the problem. First, consider what alternative strategies might be available. One way of securing reference is via some sort of causal theory. Here is Evans’s account of the view that he attributes to Kripke and dubs the ‘causal theory’ (1973: 298): ‘A speaker, using a name ‘‘NN’’ on a particular occasion will denote some item x if there is a causal chain of reference-preserving links leading back from his use on that occasion ultimately to the item x itself being involved in a name-acquiring transaction such as an explicit dubbing or the more gradual process whereby nick names stick.’ Evans emphasizes the need for a ‘reference-preserving’ link. He does so, he says, to highlight a condition he finds in Kripke: ‘a speaker S’s transmission of a name ‘‘NN’’ to a speaker S’ constitutes a reference-preserving link only if S intends to be using the name with the same denotation as he from whom he in his turn learned the name’ (Evans 1973: 298).

210

Verity Harte

Stated like this, it seems obvious that the causal theory cannot provide the explanation of how the prisoners refer to the real things, rather than the shadows. In so far as we have any evidence about dubbings or christenings, they must be performed by the prisoners. In so far as the prisoners have the intention to use the names they use with the same denotation as one from whom they have learnt it, this is an intention to use the name with the same denotation as other prisoners like themselves. The prisoners are an active linguistic community, all of whom (mistakenly) take their terms to refer to the shadows. If antecedent dubbings and reference-preserving intentions must count as decisive, then the prisoners should be taken to refer to the shadows and not the real things. Perhaps this is too swift. What if we were to say that the prisoners use names with the intention to refer simply to ‘whatever it is that is responsible for this experience I’m having’. When they do this in respect of the shadow of a model of an ox, then, unbeknownst to themselves, they thereby refer to the real ox, which stands as the causal origin of their experience.³³ Although this answer holds out the promise of explaining both of the puzzling aspects of the prisoners’ situation, I think it vulnerable to two different sorts of objections. First, why should the real thing take priority as the causal foundation of the prisoners’ reference? It is, of course, true that the real things stand at the causal origin of the prisoners’ acts of naming—they are the ultimate causes of the things to which the prisoners point when they name, as that of whose models the shadows are shadows. However, even setting aside the sorts of problems I have considered as regards causal chains of namers, any causal chain at issue must be reference-preserving. So, whether or not the real things are the referents of the prisoners’ names depends upon whether or not this rather convoluted causal chain (via shadows and models) is itself reference-preserving. But that point is moot. And there are at least two competitors: the shadows and models. Second, while this deflationary view would make it possible to explain how it comes to be a shadow of a model of their actual referent that the prisoners mistakenly think they refer to, it must suppose it a brute fact about the prisoners that they should choose, for the purpose of naming, to organize their experience in such a way as to carve the world in ways that correspond, per chance, to the ways in which it carves for real. I think it more likely that Plato takes this sort of fact to be itself in need of explanation. Like the causal theorist, Putnam and Burge, each in a different way supposes that responsibility for securing the reference need not (in their view, cannot) rely solely on what is in an individual speaker’s head: the world plays a role. On the view of Putnam and Burge, that world is partly social.³⁴ Consider once more ³³ If I understand him correctly, this is the view of Terry Penner, and the point where his and my views diverge, despite a number of common interests. I am grateful to him for discussing his (largely unpublished) view with me; for some indications of it, see now Penner 2005. ³⁴ See e.g. Putnam 1973 and Burge 1979.

Language in the Cave

211

Burge’s example of a man who thinks correctly that he has arthritis in his fingers, wrists, and ankles and who thinks incorrectly that he has developed arthritis in his thigh. According to Burge, what makes it the case that this man does indeed have an incorrect thought about arthritis and not a correct thought about ‘tharthritis’, Burge’s coinage for something that includes both arthritis and whatever is wrong with his thigh, is the linguistic community of which he is part—in particular, the existence in that linguistic community of experts who know very well that arthritis is a disease that does not occur in thighs, and the willingness of the man with the misunderstanding to defer to such experts when corrected.³⁵ However, we cannot avail ourselves of this sort of strategy here. Within the prisoners’ community, there are no experts to whom a prisoner might defer as regards the misunderstanding in question, since the misunderstanding is shared throughout the community.³⁶ If the reference-securing factor is not to be found in the physical environs of the prisoners, or in their social, linguistic community, what is left, but for the explanation to turn on what’s in their heads? By this, I do not mean that the prisoners’ reference is secured by the content of just any description they would give of their referent or by the aggregate of all the descriptions they might give; given their mistaken impression about their position, this would secure reference to the shadows alone. Rather, what I have in mind, if I understand him correctly, is something similar to what Christopher Peacocke (2003) calls an ‘implicit conception’. Peacocke proposes implicit conceptions as being in the tradition of the idea attributed to Frege by Burge, discussed above, that one might have an inadequate cognitive grasp of a nevertheless fully determinate sense.³⁷ Unlike Burge’s own development of this idea—through the division of linguistic labour across the linguistic community—the possession of an implicit conception is a feature of individuals, considered in themselves, and does not require the existence of deference to other members of their linguistic community presently able to articulate the content of the conception (Peacocke 2003: 122). According to Peacocke, a thinker may have ‘an implicit conception, but [be] unable to make its content explicit’ (p. 121). Indeed, like Plato’s prisoners, ‘[t]he thinker may even be unable to formulate principles distinctive of the concept his possession of which consists in his having that implicit conception’ (p. 121). Thus, ‘[t]o make an implicit conception explicit can … on occasion be a major intellectual achievement’ (p. 122), an achievement not only for the individual, but also for the community of which they are part. ³⁵ Burge 1991. ³⁶ What of the mysterious individual who turns the prisoner round and asks him questions about the models? The prisoner, in some sense, defers to this person eventually (since he continues the ascent). But this, I think, is not enough to do the work that Burge requires. His idea is of an active linguistic community in which some people actually, presently, are experts, and that is not what is envisaged here. Further, if the prisoner does eventually defer, it comes after some considerable resistance, so deference to already acknowledged authority does not seem to be what does the work here. ³⁷ Peacocke 2003: 117, 135, with reference to Burge 1991, discussed in sect. II.

212

Verity Harte

Possession of an implicit conception, according to Peacocke, involves the thinker in a psychological relation to a content specifying the nature of the conception (p. 124). Peacocke appears to envisage the content of the conception as a definition; and this might be an appropriate model for the Platonic story, in which we are interested in the identification of real natures, in saying what things are, where this means answering the sort of ‘what is it?’ questions that Socrates was wont to ask. For Peacocke, a thinker’s possession of an implicit conception explains, for example, their ability to make and recognize valid inferences involving the concept of which they have an implicit conception and thereby come to accept certain truths about it. In the case of the prisoners, this ties in with the role that an appeal to implicit conceptions can play in explaining the cave’s picture of cognitive progress (on which more below). Notice that, importantly, from the point of view of following Peacocke’s sort of line in explaining the puzzle about Plato’s prisoners’ linguistic situation, Peacocke claims that ‘[a] thinker’s explicit endorsement of an incorrect definition does not mean that he does not have an implicit conception whose content is the correct definition’ (p. 123). While it is undoubtedly the case that thinkers in the deeply mistaken position of Plato’s prisoners are by no means typical of the sort of thinker that Peacocke himself has in mind, his framework is nevertheless useful in explaining the sort of thing that one might say in order to relieve the puzzle about their situation. Of course, to say that the prisoners have something like an implicit conception is not yet to say very much about what things are like, in their heads. And it is hard to know how much flesh one can put on this. Possession of an implicit conception, the content of which is the (real) definition of the terms they use, secures their reference to real things and underwrites their ability to sort the world in ways that correspond to the actual divisions between things with the result that, by and large, they use the term ‘ox’ to refer to what is as a matter of fact a shadow of a model of real ox. In addition, it seems reasonable to suppose, the prisoners have some true, explicit thoughts about ox, as well as a large measure of false ones. That these are true thoughts about ox is explained by their having the implicit conception. But the question of which such true thoughts come with the implicit conception and which derive from their situation and from the ox features of the shadow to which the implicit conception directs them must, I think, remain an open question. The argument for my proposal that I have offered so far takes the form of an inference to the best explanation. If the prisoners’ situation is the one I have defended and is puzzling for the reasons I have given, then an explanation in terms of the prisoners’ possession of something like an implicit conception is the best explanation available. Of course, this strategy of argument is not decisive. I have argued against some other possible explanations. But these arguments may not persuade, or there may be other, better alternatives not here considered. There is, however, reason to think that the explanation offered is the sort of

Language in the Cave

213

explanatory strategy that Plato might favour. And this is the second line of argument I offer in support of my proposed explanation. There is evidence that the explanation offered here is the sort of explanatory strategy that Plato would favour in the fact that, elsewhere in his work, he puts forward the thesis that learning is recollection. At its most general, the thesis that learning is recollection of knowledge previously known and then forgotten is a thesis in which the starting condition of an (embodied) soul is such that it is intellectually conditioned in such a way as to make knowledge possible, whatever one might wish to say about what such intellectual conditioning involves.³⁸ Note that I am not suggesting that, behind the depiction of the situation of the prisoners as regards the use of language, is an implicit appeal to the thesis that learning is recollection. Rather, my claim is that the explanatory strategy at work in the proposal that learning is recollection is of broadly the same sort as that which I’m proposing as the explanation of the prisoners’ situation, and that this fact offers additional support for my proposal.

I V. P RO G R E S S A N D I N QU I RY: P R I S O N E R S ‘ L I K E U S ’ The prisoners’ linguistic condition involves a peculiar combination of success and failure. They succeed in referring to real things, despite their apparent utter ignorance about them. They fail in misidentifying the shadows as the real things to which they refer. Their success is integral to their failure. If asked what an ox is, the prisoners will point to what is as a matter of fact a shadow of a model of an ox. In doing so, they make a mistake. They make a mistake in doing so, because when they talk of ox, they mean ox, whatever they may point to. Succeeding and failing as they do, they fail to have (proper, identificatory) knowledge of ox or, for that matter, of shadows of models of them. It is the prisoners’ peculiar combination of success and failure that is crucial to the pay-off of their position—the way in which it feeds into the educational moral of the cave and its implications for the Republic’s overall interest in philosophical progress. In this way, my account of their condition both supports and is supported by the overall context of the cave analogy. The cave is introduced as an image in which we find some semblance of our nature ‘as regards education and its lack’ (514a2). And it yields a moral about education: that education is not a matter of putting knowledge into a soul in which it is presently lacking; the power to learn and the instrument by which we learn are present in the soul of each of us; education is the art, not of implanting this power to learn, but of turning around the soul so that its existing ³⁸ For my money, this is another case in which the specifics of the intellectual condition given from the beginning are an open question. In the case of recollection, they are also a matter for disputes that I do not need to get involved in here.

214

Verity Harte

power is oriented correctly (518b7–d7). Obviously enough, the Cave analogy is concerned to emphasize the enormity of this educational task; what the prisoners’ minimal success ensures is that the intellectual journey—if made—involves no radical discontinuity or change of subject. Just this, I think, is (at least part of) the sense in which the prisoners are ‘like us’, as Socrates says at 515a5. Consider our position, reading the Republic and, like Socrates and his companions, investigating the nature of justice. At the beginning of such an inquiry, we already identify certain things as ‘just’, but—like Socrates and his companions—we don’t know what justice is. We don’t know what justice is, Socrates says, even when it has been ‘rolling around at our feet’ all along, unrecognized, in the guise of the ‘each to his own’ principle (432d8–433a4). Here, we may note in passing, is at least a nod in the direction of a recognizably modern problem in the background of our discussion: namely, that of how an identification could ever be informative, famously addressed by Frege (1952) in distinguishing sense from reference.³⁹ For present purposes, however, what is important is that avoidance of radical discontinuity in our, nonetheless ignorant, talk about justice allows that, while we could expect, upon achieving knowledge of justice, to revise our conception of it very considerably, we will still be focused on the same object as at the start of our inquiry. Finally, I think that the picture proposed can make sense of what would happen if a returning prisoner were to take up his seat in the cave and re-enter their competitions. The habitual prisoners’ successful reference ensures that the escapee and they can indeed have some kind of conversation together. But the difference in their situation can be captured even so. The habitual prisoners are like people looking for something in a distorting mirror who have the mistaken impression that the object in question is in the mirror in the sense of its being its location. The returning prisoner is like someone looking for this same object in the mirror in the sense that the mirror is the medium for its identification. The latter must correct for distortion; the former need not. Hence the initial contest might be simplest for those who have spent their lives in front of the mirror and have never been apprised of their mistake. REFERENCES Adam, J. (1902), The Republic of Plato, 2 vols. Cambridge. Bestor, T. W. (1996), ‘Plato’s Semantics and Plato’s Cave’, Oxford Studies in Ancient Philosophy, 14: 33–82. Boter, G. (1989), The Textual Tradition of Plato’s Republic. Leiden. ³⁹ The relation between Frege’s problem and Plato, especially in the related context of the Theaetetus’ puzzles about misidentification, is explored by Rudebusch (1985), Burnyeat (1990), and Woolf (2004). In this connection, in the Republic context, compare also Socrates’ complaints against Thrasymachus’ strictures on answers to the question about justice in the form of an identity statement (Resp. 337a8–b4).

Language in the Cave

215

Brunschwig, J. (2004), ‘Revisiting Plato’s Cave’, in Proceedings J. J. Cleary and G. M. Gurtler, SJ (eds.), of the Boston Area Colloquium in Ancient Philosophy, 19 (2003), 145–77. Burge, T. (1979), ‘Individualism and the Mental’, in P. A. French, T. E. Uehling jun., and H. K. Wettstein (eds.), Midwest Studies in Philosophy, IV: Studies in Metaphysics (Minneapolis), 73–121. (1991), ‘Frege on Sense and Linguistic Meaning’, in D. Bell and N. Cooper (eds.), The Analytic Tradition: Meaning, Thought and Knowledge (Oxford), 30–60. Burnet, J. (1902), Platonis Opera, iv. Oxford. Burnyeat, M. F. (1990), The Theaetetus of Plato. Indianapolis. (1999), ‘Culture and Society in Plato’s Republic, The Tanner Lectures on Human Values, 20: 215–324; available at: http://www.tannerlectures.utah.edu Evans, G. (1973), ‘The Causal Theory of Names’, Proceedings of the Aristotelian Society, suppl. vol. 47: 187–208; repr. in A. P. Martinich (ed.), The Philosophy of Language, 4th edn. (Oxford, 2001), 296–307. References are to the latter. Ferrari, G. R. F. (2000) (ed.), Plato, The Republic, trans. T. Griffith, Cambridge Texts in the History of Political Thought. Cambridge. Frege, G. (1952), ‘On Sense and Reference’, trans. Black, in P. Geach and M. Black, (eds.), Translations of the Philosophical Writings of Gottlob Frege (Oxford). Originally published in German in 1892. Harte, V. (2006), ‘Beware of Imitations: Image Recognition in Plato’, in F-G. Hermann, (ed.), New Essays on Plato (Swansea), 21–42. Owen, G. E. L. (1973), ‘Plato on the Undepictable’, in E. N. Lee, A. P. D. Mourelatos, and R. M. Rorty, (eds.), Exegesis and Argument, Phronesis suppl. vol. 1 (Assen) 349–61; repr. in Owen, Logic, Science and Dialectic: Collected Papers in Greek Philosophy, ed. M. C. Nussbaum (London, 1986), 138–47. References are to the latter. Peacocke, C. (2003), ‘Implicit Conceptions, Understanding and Rationality’, in M. Hahn and B. Ramberg, (eds.), Reflections and Replies: Essays on the Philosophy of Tyler Burge (Cambridge, Mass.), ch. 7. Penner, T. (2005), ‘Plato’s Theory of Forms in the Republic’, in G. Santas (ed.), The Blackwell Guide to Plato’s Republic (Oxford), 234–62. Putnam, H. (1973), ‘Meaning and Reference’, Journal of Philosophy, 70: 699–711; repr. in A. W. Moore (ed.), Meaning and Reference (Oxford, 1993), 150–61. References are to the latter. (1975), Philosophical Papers, ii: Mind, Language and Reality. Cambridge. (1981), Reason, Truth and History. Cambridge. Reeve. C. D. C. (1992), Plato, Republic, trans. G. M. A. Grube, rev. C. D. C. Reeve. Indianapolis. Rudebusch, G. (1985), ‘Plato on Sense and Reference’, Mind , n.s. 94: 526–37. Segal, G. M. A. (2000), A Slim Book about Narrow Content. Cambridge, Mass. Slings, S. R. (2003), Platonis Rempublicam. Oxford. (2005), Critical Notes on Plato’s Politeia, ed. G. Boter and J. Van Uphuijsen. Leiden. Woolf, R. (2004), ‘A Shaggy Soul Story: How not to read the Wax Tablet Model in Plato’s Theaetetus’, Philosophy and Phenomenological Research, 69/3: 573–604.

11 Metaspeleology Malcolm Schofield

1 . A D I AG N O S I S A N D A R E M E DY Of all Plato’s memorable images, the Cave is the most compelling, and perhaps indeed ‘the most famous metaphor in the history of philosophy’.¹ Yet it is a headache for philosophical scholarship. A determinate interpretation has eluded commentators—not for want of their trying. The Cave was notoriously the one topic on which R. C. Cross and A. D. Woozley, in their textbook guide to the Republic of 1964, found themselves in disagreement.² In An Introduction to Plato’s Republic (1981), Julia Annas concluded flatly that ‘the imagery, memorable though it is, has no consistent overall interpretation’.³ In the subsequent quartercentury studies of the Cave from one perspective or another have proliferated. But I have no sense of incipient or imminent consensus. In fact I agree with Julia Annas: it is impossible to give one single consistent account of everything Socrates says in developing the Cave analogy. So what I propose in this paper is not another treatment of the Cave that will attempt to square the circle. Instead I offer an exercise in what might be called metaspeleology: a diagnosis of the crux of the difficulty that generates interpretative impasse, and an associated remedy for interpretative Angst. The diagnosis is a simple one. The Cave communicates not one philosophical I hope Myles Burnyeat will enjoy this return to the Cave. Its remote origins lie in his lecture course at Cambridge on the central books of the Republic —which for me were a decisive moment in trying to understand Platonic epistemology and metaphysics. The more immediate origin is the opportunity I have recently enjoyed of giving the same lecture course myself. An earlier version of the argument of the paper was presented to the University of Edinburgh’s Classics Seminar in January 2006. I am grateful to Ulrike Roth for the invitation, and to the members of my audience for comments. Christopher Rowe and David Sedley kindly made some observations on the penultimate draft—which forced on me a considerable amount of rethinking and rewriting, particularly in sect. 3. Dominic Scott has furnished excellent editorial suggestions: my thanks to him too. ¹ So Blackburn 2006: 101. ² See Cross and Woozley 1964: 227–8. ³ See Annas 1981: 256.

Metaspeleology

217

vision but two—both of great interest, but different. One is developed in the Cave as initially narrated. The story tells its own compelling story, dropping plenty of hints—varying in directness or mysteriousness—on how it is to be read. The other vision is articulated mostly in the philosophical commentary on the Cave which Plato’s Socrates supplies when he tells Glaucon how to decode it.⁴ The remedy? We get that from seeing that what the commentary proposes is a rereading of the narrative offering no more than an alternative construction of its meaning, not the single compulsory version. Consistency of interpretation as between original narrative and subsequent commentary is not mandatory. A little more precisely, we should take the commentary as enunciating inter alia a set of instructions not on what the narrative means as originally articulated, but on how it is to be reread as an allegory of the trainee philosopher’s education. Those instructions for rereading are what link the Cave with the Sun and Line analogies. Moreover, the parallel signalled between Cave and Line may be seen as limited to one rather general point that need not require interpretative contortions on the part of the reader.⁵ The Cave as narrated begins as a moral and political allegory of the condition of ordinary people in the city—in the first instance, the democratic city—and of their need for redemption from it.⁶ The Cave as reinterpreted in philosophical commentary is an image of the reorientation of the soul which can be achieved ⁴ For the crucial observation, see Burnyeat 1999: 243: ‘It is only in retrospect that we learn that the Cave has to do with mathematics as well as cultural values (532bc)’. I am not persuaded, however, by his suggestions for harmonizing the two projects, and in particular for seeing progress in mathematics as the way in which the prisoners in the Cave achieve ‘a better understanding of justice than they had before’ (Burnyeat 2000: 45). The proposal doesn’t seem to me to take the full measure of the incommensurability between Plato’s two projects in the Cave. ⁵ The contortions are usually occasioned in large part by the need interpreters have felt to find a way of explaining the original intellectual condition of the prisoners as a form of eikasia (imagination, conjecture), with the state of mind they achieve upon release from their chains accordingly equated with pistis (conviction; in principle not so tricky to achieve, despite Socrates’ emphasis on puzzlement rather than conviction: VII, 515d). Segments L1 and L2 in the bottom half of the line are thus correlated with two stages in a similar—but highly contentious—division of the prisoners’ progress into four: viz. with those represented as occurring inside the cave itself (conventionally labelled C1 and C2). On my proposals, any such move would be a mistake. (I am inclined in any case to suspect the passages which schematize a quartet of cognitive states of different relative clarity (VI, 511d–e, VII, 533e–534b) of being late additions to the text, whether made by Plato or by someone else intervening at an early point in its transmission, but in any case not altogether well fitted to their immediate contexts. I entertain a parallel suspicion with regard to the closely related clause at Ti. 29c3.) When Socrates bids us connect the Cave with the Line (VII, 517a–b), all that he explicitly enjoins is that we marry the upward journey and the viewing of what is above (as in the Cave narrative) with the ascent of the soul to the realm of the intelligible (in the Line). In other words, he says only: connect ascent with ascent, and contemplation of what is higher with contemplation of what is higher. ⁶ It ends with a celebrated account of the philosophers’ return to exercise rule in the cave, now understood as the ideal city in which they have been nurtured and educated (VII, 519d–521b). This might be considered as introducing yet a third function to which Plato puts the Cave allegory, and one which presupposes incorporation of features drawn both from the narrative as first developed

218

Malcolm Schofield

by the practice of mathematics, as prescribed in the educational curriculum of the ideal city devised for the ruling intellectual elite. Morals and politics versus mathematics; ordinary people versus the ruling elite; the democratic versus the ideal city: these three key polarities strike and puzzle all readers of the Cave. My argument is that they become altogether less puzzling once we notice that one set of elements functions primarily within the Cave as first narrated, the other in the Cave mostly as presented in Socrates’ philosophical commentary, which dwells on elements in it that receive little subsequent attention in the original telling.⁷ The body of the paper has a tripartite structure. First comes a relatively brief section on the Cave’s place in the developing argument of the Republic, and on the interpretation of the image Socrates supplies in order to explain the way it contributes to that argument. Second is a rather longer section on the Cave as actually narrated. Finally, I offer some brief concluding reflections: on why Plato may have decided to combine two such very different projects, and on how his instructions for rereading nonetheless leave them different. My aim is to liberate readers of Plato from the tyranny of thinking they have to find significance simultaneously ethical and mathematical in every detail of the Cave narrative, harnessed to just one overarching interpretation.

2 . T H E C AV E I N T H E A RG U M E N T O F T H E R E P U B L I C From (roughly) the middle of book VI (502c) to the end of book VII the Republic discusses the studies which the philosophers who are to rule the ideal city must pursue if they are to be equipped for the task. The image of the Sun (VI, 507a–509c) introduces the ultimate and most important subject of study, the Form of the Good, in its relation to knowledge and reality quite generally; the Line (VI, 509d–511e) then develops that topic further. The Cave (VII, 514a–521b) launches a related but different project: explaining the education that will be needed if a person is to be in a position to gain an understanding of true reality and above all the Good. Education (paideia) is announced as its topic at the outset (VII, 514a); and when Socrates attempts an interim summary of its moral at 518b–d, he talks again of education—as rightly conceived, not as the ability to put sight into blind eyes that some claim to possess. Once the Cave has illuminated the general nature of that education, the next item on the agenda will be to identify the studies it will need to consist in if it is to achieve what it promises (VII, 521c–d; cf. 518d). and from the subsequent philosophical reinterpretation. I shall, however, have little to say about this section of the passage in the present paper. ⁷ This is particularly true of the way Socrates works out the relationship with the Sun analogy (VII, 517b–c; cf. 516b–c).

Metaspeleology

219

About education the Cave has a simple message.⁸ Education is conversion: a conversion of the whole soul from concern with what is in process of becoming—the sensible world—to a focus on intelligible reality, and ultimately the Good (VII, 518c–d). This is what Socrates means to tell us when he imagines prisoners turning around from darkness (in the Cave below) to illumination (in the sunlit world above). The ascent to reality that is constituted by conversion is what he goes on to call ‘true philosophy’ (VII, 521c). In the immediate sequel to the Cave he goes on to explain something that the narrative itself has done nothing whatever even to intimate. He identifies the intellectual practice that will enable someone with a natural gift for philosophy to undergo this conversion. It is to be mathematics—arithmetic and plane geometry pretty much as currently practised, the new science of solid geometry, and a reformed astronomy and musical theory which will treat the study of the heavens and harmonic ratios just as geometers treat the triangles they draw on their whiteboards—in other words, as approximations to purely intelligible motions and ratios (VII, 522b–531d). The telling of the story of the Cave at the point in the Republic at which it is told is only relevant because it dramatizes the power of mathematics to effect the reorientation of our minds which is to be discursively explicated for each of the mathematical sciences in turn in the pages that follow. In a notorious anacolouthic sentence near the end of the whole discussion in book VII, Socrates says (532b–c): The release from the chains, the turning round away from the shadows to the images [i.e. the images which cast many of the shadows the prisoners take to constitute reality] and the firelight, the upward journey from the underground cave into the sun, and in the world above the inability yet to look in the direction of the animals and plants and the light of the sun, but instead at their divine reflections in water⁹ and shadows of real things (rather than shadows of images cast by a light which is itself a shadow in comparison with the sun)—all the practice of the sciences we have just described [i.e. the mathematical sciences] has this power of leading the best element in the soul up to the contemplation of the best among the things.¹⁰

Or to adapt a more careful formulation adumbrated by Socrates himself a little earlier (VII, 518c–d), mathematics is the technique needed to orientate the soul’s own intrinsic capacity for truth in the right direction. To be sure, the ⁸ Or more exactly, the message as it is represented in Socrates’ commentary on the Cave is a simple one. As we shall see, the commentary omits much that the actual narrative suggests, in this as in other respects. ⁹ ‘Divine’ is predicated of mere appearances, doubtless to emphasize that everything in the upper world is changeless intelligible reality. ¹⁰ Myles Burnyeat has rightly insisted (Burnyeat 1987: 227 n. 37) that mathematics is here made responsible for achieving everything in the Cave narration that Socrates lists—including the initial release from chains. Plato presumably wants primarily to convey the commensurately general message that mathematics is what enables us to look up to intelligible reality (albeit through images or appearances of one sort or another), instead of down to the realm of the sensible (cf. VII, 529a–c).

220

Malcolm Schofield

anacolouthic sentence suggests a rather more complex point than that. It implies that the power of mathematics is despite that limited: it is associated with an inability to look at reality other than through images (cf. VII, 533b–c).¹¹ This nuance will prepare us for the reintroduction of dialectic as the capacity we need to take us all the way to understanding first principles and grasping the account of what each thing is, above all the Good itself (VII, 532d–e, 534b–d; cf. VI, 511b–c). But initial orientation towards what is real and intelligible is represented as a job for mathematics, not dialectic. The decision to stake so much on mathematics might seem to be a mistake—on Plato’s own premisses. Isn’t Socrates failing to attend to something he has himself stressed at some length in book VI? Isn’t he forgetting that what is likely to divert a mind naturally equipped for philosophy from pursuit of the truth is moral corruption—the pleasures of the flesh and the varieties of ambition, in short the entire false value system fostered by democracy and its creature sophistry? People who have fallen victim to all that are in need of urgent moral reclamation before there is any chance of interesting them in the intellectual problems that fascinate mathematicians. Mathematics may be necessary but can hardly be sufficient for conversion. In fact, the point is not lost on Socrates. The passage on the need for a technique to engineer conversion is followed at once by a discussion of the problem of the person whose intellectual faculties are keen enough, but who is ‘compelled to pay service to vice’ (VII, 519a). The solution is expressed in counterfactual terms. If the soul of someone like that had been hammered into shape from childhood, it would have had struck from it the leaden weights of becoming which otherwise grow into it through gluttony (presumably Socrates wants to exploit the thought that such people get overweight), and turns the soul’s vision downwards. Those Socrates envisages as suitable potential rulers of the ideal city are persons who have been hammered into shape before they begin an education in mathematics. He makes it clear—in passages just before the Sun, Line, and Cave analogies are developed and just after—that they will have been tried and tested by methods described in book III to ensure that they have acquired the unshakeable moral character that education in music and gymnastics was designed to produce (VI, 503a–e, VII, 521d–522b; cf. III, 412d–414a). In the second of the passages, in particular, he is also insistent that music and gymnastics cannot themselves constitute the art or science that will ‘act as a magnet to the soul, drawing it away from the world of becoming towards the world of what is’ (VII, 521d): physical education because in its concern for the body it presides ¹¹ It may also hint at mathematical significance in the difference marked in the narrative between (a) looking at images in the cave and (b) looking at reflections in the world outside. Perhaps (a) relates to the way mathematicians use visible diagrams (cf. VI, 510d–511a), (b) to the way their apprehension of intelligible reality has a dreamlike obliqueness about it (VII, 533b–c): see Burnyeat 1987: 227–9.

Metaspeleology

221

over what comes to be and perishes, music because it trains by good habits, not on a basis of knowledge. They may not drag the soul down with leaden weights, but they do not have the power to turn it upwards. So, while the Cave itself does not address the question of the moral preparedness of the elite who are to be given the education that will prompt their intellectual conversion, the wider argument to which it contributes certainly does. To be suitable material for intellectual conversion from concern with the sensible world to a new focus on intelligible reality, a person needs not just the native equipment that Socrates has detailed—a bent for learning, a good memory, a courageous disposition, largeness of spirit (VI, 490c, 494b)—but a properly and securely developed moral character, ‘inculcated by custom and practice’, as he puts it in his commentary on the Cave (VII, 518e). That development is the subject of the discussion of music and physical training that occupies much of books II and III. And it is emphatically training and development, not moral reclamation or moral conversion. There is a great deal in Plato’s narrative of the Cave that this brief sketch of its contribution to the argument of books VI and VII has omitted. That is partly because brevity is what I have been aiming at. But it is mostly because much of the narrative is evidently preoccupied with different concerns that have no obvious or direct connection with mathematics and its role in the education of the rulers of the ideal city. To these concerns we must now turn.

3 . T H E C AV E A S N A R R AT I V E The story the Cave actually tells or intimates is in many ways ill fitted to function as the allegory of philosophical education that its role in the developing argument of the Republic requires it to be. More precisely, what the Cave tells us about the world outside the cave, and about what dawning realization of its nature and structure is like, fits the requirements of the allegory well enough (complementing what we have already learned from the Sun and the Line).¹² But with the world within the cave it is another matter. In the first place, what the cave represents is at the outset not the sensible world in general (as the philosophical commentary will urge us to assume: VII, 517b), but the city: the city as it is, above all the democratic city, not the ideal city—although it will apparently transmute into that when the issue of the philosophers’ obligation to descend again to it is broached.¹³ Conversion is certainly the crux of the narrative, ¹² Here as elsewhere I write ‘Cave’ (‘Sun’, ‘Line’) when referring to the entire analogy or image, ‘cave’ (‘sun’, ‘line’) when referring to the cave as it figures as an element in the narrative at various points. ¹³ The narrative seems to want to have it both ways at this point. For example, Socrates denies philosophers the option of refusing to ‘come back down again to the prisoners we were talking

222

Malcolm Schofield

and that certainly means intellectual conversion. But it involves primarily the shock of disillusionment about the moral values current in the world of the city as it is, and pre-eminently of justice, not a better grasp of the realities studied by mathematics. In the first instance, in fact, it is a conversion from a state of almost total illusion and delusion about the world of the democratic city to a faltering grip on the ‘realities’ of that world—not enlightenment about true intelligible reality, although of course that is the ultimate destination of the intellectual journey that the prisoners are to undertake. Conversion is not here a metaphor for the shift in intellectual focus from the sensible to the intelligible which Socrates’ commentary on the Cave represents it as constituting. That shift is marked in the narrative by emergence from the cave into the sunlit world outside. Conversion is something that happens within the cave, when someone who could to begin with see only shadows is compelled to turn around, and to look with great difficulty and uncertain success at the objects—more real than the shadows, but still only themselves images—that cast them. It will help to see the force of these points if we look at the narrative in more detail. When we do so, we find a lot of mystery, no doubt much of it deliberate on Plato’s part.¹⁴ What exactly is it that the shadows on the wall of the cave and the statuary throwing the images represent? What precisely is symbolized by the prisoners’ inability to see anything except the shadows, and by their belief that reality is constituted by the shadows? Why are they represented as imprisoned in the first place? A lot is destined to remain fairly obscure. None of the questions I have just posed is answered by Socrates’ philosophical commentary on the Cave—again, no doubt deliberately. Interpreters scrabble over the clues Plato seems to be letting slip at various points. But consensus notoriously remains elusive. The passage that gives most away is perhaps the following one (VII, 517d–e):¹⁵ And here’s another question. Do you think it’s at all surprising if a person who turns to things human from contemplation of the divine cuts a sorry figure, and is made to look a complete fool—if before he can see properly or can get acclimatized to the darkness around him, he is compelled to compete, in the law courts or anywhere else, over the shadows of justice or the images which cast those shadows, or to get into argumentative conflict about whatever assumptions regarding these things are made by those who have never seen justice itself?

The scenario Socrates portrays is of course dense with allusion to his own plight before the court at his trial for impiety and moral corruption. That theme is about, or share in their hardships and rewards—whether of the more trivial or the weightier kind’ (VII, 519d). This presupposes that they started life in the world of the city as it is described in the opening stretches of the Cave (514a–516e; the echo of ‘rewards’ (timai), 516c, is especially notable). But a few lines later on he is explicit that the philosophers have been brought up and educated in a very different environment: the ideal city (520a–c). ¹⁴ Contrast the painstaking and indeed rather schoolish explanations of the different elements of the Sun and Line analogies that Socrates supplies as he develops them. ¹⁵ Translations are mostly based on Tom Griffith’s version in Ferrari and Griffith 2000.

Metaspeleology

223

one to which Plato often returns. And the mention here of ‘the law courts or anywhere else’ is echoed almost exactly in a much expanded version of this analysis of the philosopher’s plight in the ‘digression’ of the Theaetetus (174c), where it picks up a reference to ‘the law court or the council chamber or any other public gathering of the city as a body’ (173d). The evocation of forensic and deliberative rhetoric conveyed by ‘compete’ and ‘conflict’ over shadows and images of justice inescapably brings democratic Athens to mind—not only the trial of Socrates, but such episodes as the debate over policy towards Mytilene between Cleon and Diodotus (as represented in book 3 of Thucydides’ History (3. 37–48)). In this Republic passage, as in the Theaetetus digression, Plato is working with a simple polarity between things human and things divine as focus of cognitive interest, with justice itself—the Form of Justice—the prime example of an object of contemplation, and images and shadows of justice treated (apparently indifferently) as what the unenlightened argue about and jostle over. These are presumably human conceptions of what is just or unjust uninformed (as Socrates says) by understanding of what justice really is, as his reference to people’s ‘assumptions’ makes fairly plain. Democratic politics is also suggested by an earlier passage in which Socrates describes the kind of intellectual activity characteristic of the prisoners in the cave (VII, 516c–d): Back in the cave they [i.e. those who have escaped from it] might have had rewards and praise and prizes for the person who was quickest at identifying the passing shapes, who had the best memory for the ones which usually came earlier or later or simultaneously, and who as a result was most capable at predicting what was going to come next. Do you think someone who had escaped from the cave would feel any desire for these prizes? Would he envy those who won esteem and exercised power among the prisoners? Or would he feel as Achilles does in Homer? Would he much prefer ‘to labour as a common serf, serving another man with nothing to his name’, putting up with anything to avoid holding those opinions and living that life?

Some readers think that Plato must here be writing about the populace as they watch the passing show of politics (like a theatre audience, as Cleon complains in the Mytilenean debate: History 3. 38). But prizes go to the politicians, not ordinary members of the d¯emos. Perhaps everybody in the assembly does engage in a sort of guessing, spotting, and remembering game, but politicians are distinguished by being the ones best at it, by that being recognized to be their strength, and by their exercising power in consequence.¹⁶ I am put in mind of a different passage of Thucydides: the eulogy of Themistocles for his supreme ability to form judgements instantly with minimal deliberation, his no less impressive success at guessing for the most part the future turn of events, and his capacity to interpret what is on the city’s hands at the moment (Hist. 1. 138). ¹⁶ This point is well made by James Wilderbing (2004: 121–8), but to my mind his further suggestion that the puppeteers are ‘the multitude’ is misconceived.

224

Malcolm Schofield

I take it that this is the sort of intellectual performance that Socrates has in view in the extract I have just quoted. Themistocles is the kind of person who has ‘what passes for wisdom there [i.e. in the cave]’ (516c), and on that account wins esteem and comes to wield significant power in the city. One further passage confirms the view of the Cave that has been emerging so far from our examination of the narrative. When Socrates imagines himself persuading those who have become true philosophers that they must come back down among the prisoners, this is how he concludes his address to them (VII, 520c–d): So you must go down, each of you in turn, to join the others in their dwelling-place. You must get used to looking at things covered in darkness. When you do get used to it, you will see a thousand times better than the people there do. You will be able to identify each of the images there, and recognize what it is an image of, because you have seen the truth about what is beautiful and just and good. In this way the city—ours and yours—will enjoy government that is really awake, rather than the kind of dream in which most cities live nowadays, governed by people fighting one another over shadows and engaging in factional conflict over ruling, as if it were some great good.

Once again the focus is on values—the good, the beautiful, the just—and on the difference between the truth about them (seen outside the cave) and images (what philosophers will be able to recognize inside it once their eyes have got used to the darkness). Once again the suggestion is that in the ordinary pattern of things the city which the cave represents will be the city as it is, riven by faction (that would fit oligarchies as much as democracies) and ‘fighting over shadows’, unless it is reformed by philosophy. What sort of conversion is possible within such a world? That question takes us back to the intensely mysterious opening passage of book VII. It and its details have been the subject of minute scrutiny and speculation in buckets. I am not going to try to recapitulate all of that. I shall focus on just one cluster of elements in the narrative. What are the shadows that the prisoners are looking at?¹⁷ What are the images that cast them? What is symbolized by their turning from the one to the other? Why does the reorientation leave them so confused? What did their imprisonment consist in? If we can get no reasonably clear idea what answers to give to these questions, we are left simply baffled by the Cave’s narrative. And given that the clues to Plato’s meaning we have been identifying and following so far have yielded a reasonably coherent and intelligible interpretation, our best hope of understanding what the opening passage means by conversion is likely to rest in the assumption that it, too, is telling us something about the values of the city as it is. ¹⁷ I shall offer no discussion of the remarkable information that the prisoners have never seen anything of themselves or each other except the shadows cast by the fire on the wall of the cave in front of them (VII, 515a). For a fascinating exploration of what might be being suggested, see Brunschwig 2003.

Metaspeleology

225

Readers who share that general assumption may well have their own favourite conjectures as to how my list of questions is to be answered. In what follows I offer a hypothesis of my own which does not exactly duplicate any other I have come across. But I should stress that its status within my overall argument is merely exemplary: to indicate the sort of story that needs to be told to give substance to details that the narrative clearly means to invest with symbolic significance. And it is a non-exclusive hypothesis: I don’t want to deny that there may also be validity in other answers to the questions. Indeed, I take it that Plato’s writing is of set purpose multiply suggestive. Here is the hypothesis. Most people in the city as it is hold values—above all about justice—that are mere reflections of something else. What is that something else? Later in the text Socrates will refer to them compendiously as images (eid¯ola, 520c; agalmata, 517d). But when he first introduces them, he refers to ‘all sorts of vessels and implements … and statues of men and other animals, artefacts made of stone and wood and all kinds of materials’ (514c–515a). The intent must surely be to emphasize the immense variety and comprehensive range of the craftsmanship and artifice that is represented in these images. If we transpose this idea to the realm of human values, we get the thought that the values most people hold are reflections of the all-embracing culture they inhabit—which is itself a hugely complex human artefact. When Socrates goes on to explain that the prisoners think or name what they see ‘reality’ (515b), he is in effect also making the point that—confined as they are to looking at shadows—they are completely unaware that what they treat as reality is nothing other than a reflection of human cultural artifice. Let me offer an illustration of the point.¹⁸ Suppose some of the prisoners hold that justice is telling the truth and repaying your debts. Socrates will be saying: That opinion is something they take to be the truth about justice—what justice really is. But it is in fact nothing more than the shadow in their minds left by the culture in which they have been raised. That culture exercises universal control over their minds. And because they have no cognitive resources that might enable them to extricate themselves from its control, they are actually its prisoners. Conversion will require a complete reversal of outlook, sparked by the acutely painful and reluctant realization that what they had been taking to be reality is not reality. Suppose that this is achieved (how I shall discuss shortly). What the prisoners then need to come to appreciate is that there is something which has more reality than what they had taken reality to be. Socrates represents the process as extremely difficult and confusing (515d): When someone is released and compelled suddenly to stand up and turn his neck round and walk and look towards the light, and finds all these things painful to do, and because of the flickerings of the fire can’t see those things distinctly whose shadows he was seeing ¹⁸ In using this example (and in subsequent reflections on it) I am indebted to the stimulus of Smith 1997.

226

Malcolm Schofield

before, what do you suppose he’d say if he was told that what he used to see before was just nonsense, whereas now he is seeing better, since he is closer to what is, and turned towards things that have more reality? Suppose further that each of the passing objects was pointed out to him, and he was asked what it is, and compelled to answer. Don’t you think he’d be puzzled? Wouldn’t he think the things he saw before to be more true than what was being pointed out to him now?

Within the terms of the Cave’s own narrative, Socrates’ final suggestions here are perhaps not hard to account for. In their original condition it was well-nigh inevitable that the prisoners would take the shadows to be reality. In fact they had every reason to take them to be living creatures. All of the shadowy figures moved, and some of the figures (humans and animals, presumably) uttered sounds, as it must have appeared to the prisoners. There was nothing in their cognitive lives to make them doubt that these appearances constituted reality. In confronting the objects they are now being shown, they experience understandable difficulties. The main problem is fully described here by Socrates. The released prisoners have to look in the direction of the fire, and that makes it hard for them to see the objects in front of it clearly at all: it dazzles and it flickers. They find it hard to accept that what is now being presented to them has more truth or reality than the shadows they used to observe. There is something distinctly familiar about this new condition of theirs. The puzzlement (aporein) which Socrates attributes to them has obvious affinities with the sense of helplessness induced in interlocutors in early Platonic dialogues by his own characteristic form of questioning. Indeed, if we ask what technique could bring someone to turn around, as the prisoners are made to do, from illusions about reality to a degree of realization that they are illusions (coupled with a consequential disabling numbness of mind), then the questioning practised in Socratic elenchus is surely the obvious candidate. The light it casts initially brings confusion and dismay, and an inability to see the way forward. Socrates’ interlocutors typically start with the confident or complacent assumption that X is F (courage is standing in line and not running away, s¯ophrosun¯e is doing things with due order, justice is telling the truth and paying your debts), and are brought to see that that’s not the reality—because sometimes X is not-F instead. They don’t thereby get much understanding of what the status of their initial convictions is, or of how to make headway beyond them. Characteristically they make fresh attempts to articulate their ideas, but the eventual effect of the illumination they have gained is more like the paralysis that car headlights induce in rabbits and stoats that get caught in their glare or, as Meno complained, like the numbness the sting of the sting-ray fish produces (Men: 80a–b). The ex-prisoners’ problems continue even when particular objects are pointed out to them. Comparison with the passage already quoted on the return of the philosophers to the cave suggests why. Once the philosophers’ eyes adjust to the darkness, they will see ‘a thousand times better than the people there

Metaspeleology

227

do’: philosophers would be able to ‘identify each of the images there, and recognize what it is an image of’, because they have ‘seen the truth about what is beautiful and just and good’ (VII, 520c). We infer that the released prisoners who haven’t yet emerged from the cave would be just puzzled by these images. Perhaps they couldn’t even become aware that lifeless images are what these things are—given that they have never encountered any of the originals. Certainly any identifications they might make would have to be based on experience in unpropitious cognitive conditions—and experience that they don’t yet have much of. No wonder if the shadows they once looked at still—to begin with—seem to them truer realities. But we —the dialogue’s readers—know that the objects they now see are artefacts manipulated by handlers. On the interpretation I am proposing, Plato thereby intimates to us something the ex-prisoners themselves do not grasp: that the values that were reflected in the judgements they formed in their original state are no more than nomoi, human cultural artefacts. Let me recur once more to the example of justice conceived as truth telling and debt repayment. Because this conception is widely believed, but because telling the truth or repaying your debts turns out—on the examination Socrates gives that proposal in book I of the Republic —to be just sometimes, but sometimes unjust (I, 331c–332a), the principle qualifies as one of those many cultural norms (nomima) about what is beautiful and the rest which people in general accept, but which in book V’s formulation ‘roll about between not being and pure being’ (V, 479d). Thus we find the materials for the idea that a failed candidate for the truth about justice may actually be a culturally acceptable norm available—and so available for allegorization in the Cave—elsewhere in the dialogue.¹⁹ Conversion to a perspective from which illusions about, for example, justice can be seen as such does not mean, however, that the converts themselves will perceive them as the shadows of cultural norms that they are. (It is one thing to grasp that X is not-F as well as F , and that being F is therefore not the reality of what X is, another to appreciate that the idea that X is F is merely a nomos.) Still less does conversion presuppose or constitute a grasp of the intelligible reality of justice itself (of what X really is). To revert to the Cave narrative, the prisoners are released from their chains and forced to turn around before they make their journey out of the cave up into the world outside. Even if they are at any rate now pointed in the right direction, the journey they still have to make is portrayed as a pretty long haul, and it is particularly emphasized that they will have to be dragged forcibly and painfully over a steep and rough path to get outside. In ¹⁹ Socrates does not think such nomima have no validity at all. In book IV he allows that the person whose soul is just will go in for what he nonetheless describes as ‘the usual rubbish’ (ta phortika, 442e–443a). He won’t embezzle gold or silver deposited with him for safekeeping. He won’t have anything to do with temple robbery, theft, or betraying people. He will stick by his oaths and other agreements. The artefacts that constitute the furniture of the cave are fairly solid stuff—within that limited framework (it is from higher perspectives that they are ‘rubbish’).

228

Malcolm Schofield

other words, conversion was one thing, ascent is quite another. Nor does the narrative indicate that there is much relationship between the two. In particular, there is no suggestion that what will help to drag the released prisoners out of the cave into the sunlight bears any relationship to further reflection on the nature of the objects of whose existence they became aware when first forced to turn their necks around.²⁰ We are now in a position to return to the simple, basic question: what was it that shackled the prisoners in the first place—that made them prisoners? The answer I have already suggested is: culture itself. Culture—the culture of the city as it is, and especially the democratic city—is what imprisons them.²¹ It imprisons them because they are quite unaware that their beliefs about justice and other values are not straightforwardly true, but the reflexes of something else; and because, left to their own devices, they are quite unable to achieve awareness of that. On my reading, the prisoners’ shackles are inseparable from the manipulation of images and other artefacts which brings about the shadowplay on the wall of the cave, and so effects the only cognitive input they ever receive. Only something like Socratic questioning has the power to unsettle their beliefs in such a way as to release them from their intellectual shackles, albeit at the cost of leaving them thoroughly puzzled about what is real and what isn’t.²²

4 . P L ATO ’ S A LT E R N AT I V E C AV E S The Cave, therefore, is a picture which incorporates alternative instructions—one set explicit, the other largely implicit—on how it is to be read. Unsurprisingly they generate rather different philosophical models of intellectual imprisonment and liberation. The official instructions advise us to imagine a Cave whose contribution to the argument of the central books of the Republic is clear and determinate. From the end of book V to the end of book VI Plato has been making us think about the philosopher—latterly about the intelligible reality for which philosophy has a consuming passion, and about its dependence on the Form of the Good. Glaucon had coped well enough with the image of the Sun; he had wanted to know more, had listened to Socrates developing the image of the Line by way of response, and despite something of a struggle ²⁰ There is nothing here resembling the use of visible forms to think about the intelligible which is ascribed to geometers in the Line (VI, 510d–e). ²¹ I need hardly add that my argument here—as in this section of the paper generally—owes much to Myles Burnyeat’s illuminating observations on the Cave (Burnyeat 1999: 238–43). ²² It is worth noting that Plato gives us no reason to think that he conceives of the ruling elite of the ideal city as having been imprisoned by the culture they absorbed through poetry, music, and gymnastics—which includes myths, notably the myth of the Noble Lie—controlled and controlling though it is. The higher education they are to get through the practice of mathematics and dialectic seems to leave that culture intact, although they presumably acquire a new understanding of its proper basis. Cf. Wilson 1976: 126–7.

Metaspeleology

229

had finally got an adequate grip—so Socrates assures him—on what it was saying. By the end of book VII he will be in no doubt about what the Cave contributes to this ongoing discussion of philosophy and the philosopher. It is supposed (see Section 2) to symbolize the conversion from preoccupation with the sensible world to understanding of intelligible reality that mathematics will enable someone to make—if they have the right natural aptitude for philosophy, and if as a potential ruler they have thoroughly assimilated the training in moral virtue which will be provided in the ideal city. But the Cave is a piece of writing that ambushes the reader. What it actually portrays in its narrative at the beginning of book VII (see Section 3) is something quite different: another sort of conversion altogether, from being wholly captive to reflexes of cultural norms of justice and the good construed as the real truth (as people typically are captive in the city as it is), to a numbed and faltering appreciation that the truth lies elsewhere. This form of conversion is something which anybody might experience if subjected to Socratic questioning—although there would then have to be a long and difficult further process of intellectual struggle if such a person were to achieve an adequate grasp of the real nature of justice and other values. In short, Plato brings the argument back to us —the ordinary person in the world as it is—and to something like the everyday preoccupations with justice and injustice which launched the dialogue back in book I. It announces an image which will illuminate our nature when it comes to education and being uneducated (VII, 514a). And though we might have been expecting more on philosophers and their education, for a while at least Socrates is going to talk about us. His prisoners are strange prisoners, says Glaucon. ‘Like us’, says Socrates (515a).²³ This is a brilliant stroke on Plato’s part. The Cave narrative gives us something less intellectually strenuous and more immediately accessible (despite its manifold mysteriousness) than the Sun, and above all the Line: what is more, something no reader can ever forget. And in turning attention in the first instance to us, it manages to suggest that reflection on the requirements of philosophical education needs to be set in the broader context of self-examination about the human condition, and about its imprisonment by the cultural limitations imposed by human society as it actually is. The price Plato pays is a degree of confusion. The conversion he specifically has in mind for philosophers (e.g. VII, 521c, 524e–525a) is something they would need to undergo even if they had escaped or avoided that sort of imprisonment and enjoyed the benefit of moral acclimatization and training in an ideal city.²⁴ In the terms first elaborated ²³ Does Socrates mean to suggest that even he was once the prisoner of the culture in which he was raised? Certainly his picture of the difficulties a philosopher will experience on descending once more to the cave from which he has escaped has—as we have observed—a strong autobiographical dimension. ²⁴ I say ‘even if they had escaped or avoided’ the kind of imprisonment imposed by living in the social and cultural environment of the city as it is. But of course book VI had argued that those

230

Malcolm Schofield

in the narrative, it is not really conversion but ascent; and it is liberation—if liberation it is—not from cultural assumptions, but from reliance on the senses, and their focus on the realm of becoming, instead of the intellect, with its access to being.²⁵ Narrative and philosophical commentary upon it work out ideas that are not merely different, but in terminology as well as in substance at odds with each other. Nonetheless, Plato has Socrates give some directions which ought to have done much to minimize the confusion, had they been taken as they should be: as instructions on how to reread the Cave as an allegory of the philosopher’s education. These directions are spelled out in the famous passage where we are told to fit the image of the Cave as a whole to what has been said before—that is, to the Sun and the Line (VII, 517a–c). That needn’t mean in every detail.²⁶ The assumption that we are being told to bludgeon everything in the Cave to fit whatever parallels could be identified in Sun and Line has caused much of the interpretative damage. In fact, Socrates is extremely selective. The material in the narrative that we explored in Section 3 above is completely ignored. Socrates now urges us just to make some very general connections with elements in the two previous images. He picks out two. First, we are to compare or assimilate the cave dwelling to the visible region illuminated by the power of the sun (from the Sun analogy). Socrates doesn’t say it is the realm of the visible—with good reason, since as we saw in Section 3 on the Cave narrative, the cave in fact there represented the city (the city as it is). Second, his hope is that the upward journey and the view of things in the world above can represent the soul’s journey up into the realm of the intelligible (as in the Line), though he expresses qualified confidence about claiming that this is what coming to understand the intelligible would be like. The story he has told, however, about gazing on the sun itself and figuring out that it is what sustains seasons and years and everything in the realm of the visible—and indeed is in a way the cause of all of them—captures the way things appear to him. The Form of the Good is the last thing to be understood in the realm of the knowable, and we need to reason out the way that it is cause with the natural potential for philosophy are—in all but exceptional cases—so corrupted by that sort of environment that they can never fulfil their potential (VI, 490e–497a). From its point of view, escape from politics in the cave of the democratic city to philosophical contemplation in the realm of the intelligible beyond it comes close to counterfactual fantasy. ²⁵ Yet it is important that the two conversions on my interpretation share a key fundamental feature. What liberates the prisoner of the culture of the democratic city is realization that X is not-F as well as F . Similarly, it is famously appreciation that e.g. the middle finger is both large and small that illustrates the way in which engagement with the predicates typically studied in mathematics turns the soul away from the sensible towards the intelligible (VII, 523b–525b). ²⁶ So Burnyeat 1987: 228 n. 38: ‘The solution [to the intractable problem of trying to establish a one–one correspondence between the four stages of the Cave and the four sections of the Line] is to let Plato’s stage-directions tell us what kind of relationship he means us to establish between his images.’

Metaspeleology

231

of all that is right and beautiful whether in the visible or the intelligible realm, if we are to act wisely in private or public business. Reread the Cave like that, and you won’t make it consistent, but you won’t be left terminally disconcerted or deciding that it’s all baffling conundrum. In fact, you’ll be poised to extract what you are to be told you need for the main business in hand in this part of the Republic: the educational method by which someone can be put on the upward path from the realm of the sensible to the realm of the intelligible (VII, 521c–d; cf. 518c–d)—which proves to be mathematics. REFERENCES Annas, J. (1981), An Introduction to Plato’s Republic. Oxford. Blackburn, S. (2006), Plato’s Republic: A Biography. London. Brunschwig, J. (2003), ‘Revisiting Plato’s Cave’, Proceedings of the Boston Area Colloquium in Ancient Philosophy, 19: 145–77. Burnyeat, M. F. (1987), ‘Platonism and Mathematics: A Prelude to Discussion’, in A. Graeser (ed.), Mathematics and Metaphysics in Aristotle (Bern), 213–40. (1999) ‘Culture and Society in Plato’s Republic’, The Tanner Lectures on Human Values, 20: 215–324. (2000), ‘Plato on Why Mathematics is Good for the Soul’, in T. J. Smiley (ed.), Mathematics and Necessity: Essays in the History of Philosophy (Oxford), 1–81. Cross, R. C., and Woozley, A. D. (1964), Plato’s Republic: A Philosophical Commentary. London. Ferrari, G. R. F. (ed.) and Griffith, T. (trans.) (2000), Plato: The Republic. Cambridge. Smith, N. (1997), ‘How the Prisoners in Plato’s Cave are ‘‘like us’’ ’, Proceedings of the Boston Area Colloquium in Ancient Philosophy, 13: 187–204. Wilderbing, J. (2004), ‘Prisoners and Puppeteers in the Cave’, Oxford Studies in Ancient Philosophy, 27: 117–39. Wilson, J. R. S. (1976), ‘The Contents of the Cave’, in R. A. Shiner and J. King-Farlow (eds.), New Essays on Plato and the Pre-Socratics (Guelph), 117–27.

12 Why no Platonistic Ideas of Artefacts? Sarah Broadie

I . A N O N - B I O G R A PH I C A L QU E S T I O N Plato’s followers from early on disowned Ideas for human artefacts. In the Phaedo the case is stated in this way—that the forms (eid¯e) are causes both of being and of becoming; yet when [i.e. in the cases in which] the forms exist (kaitoi t¯on eid¯on ont¯on),¹ still the things that share in them do not come into being, unless there is some efficient cause (to kin¯eson); and many other things come into being (e.g. a house or a ring), of which we say there are no forms. Clearly, therefore, even the other things can both be and come into being, owing to such causes as produce the things just mentioned. (Arist., Metaph. A, 991b 3–9; cf. B, 999b 17–20; A, 990b 8–11)²

The ‘forms’ which ‘we’ are here said to deny for house and ring are the metaphysically separate entities against which Aristotle argues.³ Here for convenience I use ‘Ideas’ to refer to these, and ‘forms’ to cover both them and Aristotelian forms indifferently. Other passages from the Metaphysics confirm that it is because house and ring are artefacts that Ideas for them were ruled out.⁴ The point is clear in the Peri This paper is a token of my appreciation of Myles Burnyeat’s example as pathfinder and communicator for our discipline: also of gratitude for the Cambridge May Week seminar in Ancient Philosophy, an event which I particularly associate with Myles’s standards and from which I have benefited enormously over the years. ¹ Irwin and Fine (1995) have ‘even if the forms exist’. ² Translations from the Metaphysics are by W. D. Ross in Barnes 1984, except that I have not followed his use of ‘Forms’ (capital ‘F’) where the reference is to the Platonistic Ideas. ³ ‘Metaphysically separate’ (sc. from worldly particulars) is notoriously hard to pin down. Defining separateness as ontological priority (cf. Fine 1984) fails for Platonisms that teach the necessary eternity of the physical world, if ‘ontological priority’ entails that the prior can exist without the posterior. Perhaps the separateness of Ideas which Aristotle targets involves what below I refer to as ‘a se causativity’. ⁴ H 3, 1043b 18–23, where the point is tied to the non-substantiality of artefacts; Λ 3, 1070a 13–18, a difficult passage.

Why no Platonistic Ideas of Artefacts?

233

Ide¯on. Here is a Peri Ide¯on objection to one of the Arguments (for Ideas) from the Sciences (epist¯emai): And this last , in addition to the fact that, like the other arguments, it does not prove that there are ideas, will seem to establish that there are also ideas of things for which they do not want ideas. For if, because medicine is the science not of this health but of health without qualification (hapl¯os), there is some health itself (esti auto tis hugieia), then this will also apply to each of the crafts (technai). For none of them is of the particular or the this either, but each is of the without qualification that it is about. For example, carpentry is of bench without qualification, not of this bench, and of bed [or: couch⁵] without qualification, not of this bed. And sculpture, painting, house-building, and each of the other crafts is related in a similar way to the things that fall under it. Therefore there will be an idea of each of the things that fall under the crafts as well—which they do not want. (Peri Ide¯on, recensio vulgata 79. 22–80. 7)⁶

This passage assumes an uncharacteristic (for Aristotle) contrast between epist¯emai and technai. Normally Aristotle either uses ‘epist¯em¯e’ broadly to cover all expertises, including the teachable crafts (e.g. at Metaph. A, 990b 13), or he uses it in the technical sense of the Posterior Analytics to mean a purely theoretical hexis apodeiktik¯e. The latter definition figures in Nicomachean Ethics VI as template for (inter alia) the similarly scholastic definition of techn¯e as hexis meta logou al¯ethous po¯etik¯e (Eth. Nic. VI 3–4). Usually in Aristotle medicine figures as a paradigm techn¯e: always counting as one by the scholastic definition of techn¯e, sometimes included under the broad sense of epist¯em¯e; and sometimes, as a non-theoretical discipline, excluded by the narrow sense of the latter.⁷ But here in the Peri Ide¯on, techn¯e and epist¯em¯e are mutually exclusive on some different principle. For medicine (along with geometry, 79. 13) is given as a prime example of epist¯em¯e, while the examples of techn¯e are carpentry, sculpture, painting, and house building. (The recensio altera adds weaving.) This division, let us suppose, is based on the thought that medicine aims to restore the body’s natural state—which for the most part exists by nature or by cosmic nous —whereas benches, houses, and sculptures are essentially manufactured.⁸ In the early Academy, the embargo on Ideas of artefacts is associated with Xenocrates, who defined Idea ‘as the paradigmatic cause of the things that are ⁵ ‘Couch’ is correct (standing not for what is slept in at night but for what is reclined upon at an ancient Greek symposium), as we philosophers have been led to see by Myles Burnyeat’s brilliant disquisition on ‘Greek couches’ in his Tanner Lectures (1999). But this does not affect the argument of the Peri Ide¯on. ⁶ Apart from a small change of punctuation, the translation is by G. Fine (1993: 13–14), based on the critical edition of D. Harlfinger in Leszl 1975. ⁷ See also An. post. I 19, 100a 9. ⁸ In the recensio altera, t¯on techn¯et¯on occurs at 79. 19, before mention of techn¯e. The adjective means ‘manufactured’, thereby restricting the scope of techn¯e to suit the argument. It seems that techn¯etos does not occur in Aristotle. There may have been different grounds for distinguishing epist¯em¯e and techn¯e in ways that fit the Peri Ide¯on examples. The argument of this paper does not depend on how that goes.

234

Sarah Broadie

always [or: in any given case] constituted according to nature’ (t¯on kata phusin aei sunest¯ot¯on).⁹ It is natural, although not mandatory, to take the last phrase as referring to natural phenomena.¹⁰ But what about the author of the Phaedo himself? Here is Aristotle again: Now in some cases the ‘this’ does not exist apart from the composite substance, e.g. the form of the of house does not so exist, unless the art of building exists apart … but if it does [i.e. if the ‘this’ does exist apart] it is only in the case of natural objects. And so Plato was not far wrong (ou kak¯os eph¯e) when he said that there are as many forms [sc. Ideas] as there are kinds of natural things (hoposa phusei) (if there are forms at all) … (Metaph. Λ, 1070a 13–19)

It is natural to understand this too as evidence for a restriction of Ideas to non-artefact kinds, naming Plato¹¹ as agreeing with the restriction. However, the Platonic dictum which Aristotle reports here is logically compatible with admitting artefact-Ideas. ‘As many A’s as B’s’ does not entail, where C is some sort of contrary to B (as ‘artefact’ is to ‘natural thing’ in one obvious sense of the latter), that there are no A’s corresponding to any C’s. Moreover, the fact that a contrast between natural objects and artefacts is on Aristotle’s mind here is no guarantee that Plato, in the reported remark, meant ‘natural things’ in the same sense. Nor is it even a guarantee that Aristotle believed Plato to have meant it thus. Aristotle is perfectly capable of appropriating someone’s pronouncement to use it, knowingly, with his own quite different interpretation. Suspicion about the natural interpretation of this last passage is of course reasonable, since one would never guess from Plato’s dialogues that artefact Ideas are excluded; some of the most emphasized examples of forms are the Couch and Table of Republic X, the Shuttle and Auger of the Cratylus.¹² (Instances from Plato’s texts aside, the ordinary experience from which, apart from doing mathematics, Platonistic intuitions draw most strength, is that of planning and producing according to plan. Something ideal and beyond anything we can see or touch seems to be guiding our thoughts and actions. Not everything produced through planning or intelligent aiming is rightly called an artefact: that I owe my health to medical science does not make it my physician’s artefact. Still, no one can doubt that artefacts are paradigms of planned outcomes. Hence it may well seem that forms of artefacts ought to count as Ideas par excellence —‘if there are Ideas at all’, as Aristotle would say.) ⁹ Procl. In Plat. Parm. 691, Stallbaum. For doubts about the text (most significantly, about aei), see Cherniss 1962: 257. ¹⁰ Aristotle standardly uses the same or a cognate formula to refer to natural as opposed to artificial things; see Bonitz, Index s.v. sunistanai, 731a 18–26. ¹¹ Although Alexander’s text may have attributed it to hoi ta eid¯e tithemenoi. ¹² Resp. 596b ff.; Cra. 389a ff. Burnyeat 1999: 245–9, is illuminating on artefact-forms in the Republic, although he is not concerned (any more than Plato himself in the Republic) with purely ontological issues.

Why no Platonistic Ideas of Artefacts?

235

Efforts abound to relieve the discrepancy between evidence in the dialogues that Plato endorsed artefactual Ideas and evidence in Aristotle’s Metaphysics that he excluded them and was therefore vulnerable to the objection levelled in our first two passages above.¹³ Most of the discussion centres on Plato: what did he hold when, concerning artefact-Ideas? But nearby looms a non-biographical question: why were any early Platonists against artefact-Ideas? That is the first question of this paper. It arises not simply because of examples in Plato’s dialogues, but because, as we see from the first two passages, the Platonists who rejected artefact-Ideas did so in the teeth of the fact that some of their own arguments, if sound, opened the door to such entities. For they must have realized that their arguments for Ideas from the Sciences can be paralleled by exactly similar arguments from the Crafts. Plainly, then, their rejection of artefact-Ideas was neither casual nor easily renounced. Speculation about the reasons behind it might help us piece together what was going on in the early Academy.

I I . I N T E R P R E T I N G T H E TIMAEUS It would be foolish to assume that there was a unique reason why early Platonists rejected Ideas for artefacts. Just as there were many reasons for postulating Ideas at all—considerations which do not all sit well together or support a single coherent theory—so it may have been with the tenet against artefact-Ideas. Here I suggest one story. It has its roots in the Timaeus, but gets off the ground with Xenocrates: Xenocrates who, we already have some reason to believe, defined Ideas so as to rule out any for artefacts.¹⁴ Aristotle mentions contemporaries who regard as not literal the Timaean account of the cosmos starting to exist from pre-existing materials. According to Aristotle, they speak of themselves as having used the notion of a cosmic beginning purely as an expository device.¹⁵ This shows how close was their self-identification with the author of the Timaeus. Simplicius states that the reference of Aristotle’s remark seems to be Xenocrates and the Platonists.¹⁶ In Xenocrates’ vision¹⁷ of the physical domain as timelessly flowing from the Ideal and the mathematical, and all flowing ultimately from the One and the Indefinite Dyad, each level surely ‘gives rise’ to the next below by a sort of mathematical necessity: thus the orderly physical world could not have started to exist any more than its eternal ¹³ For coverage and discussion see e.g. Cherniss 1962: 240–60; Leszl 1975: ch. 5; Fine 1993: ch. 6. ¹⁴ There may be another story, perhaps not entirely independent of the one told here, whose key fact is that it was common ground in the Academy that artefacts are not substances (Arist. Metaph. Z 17, 1041b 28–30; H 3, 1043b 18–23; cf. Z 2, 1028b 8–13; H 1, 1042a 6–11). ¹⁵ Arist. Cael. I, 279b 32–280a 2. ¹⁶ Simp. in Cael. 303. 33–4. ¹⁷ See Theophr. Metaphy. III, 6a 23– b 9; cf. Arist. Metaph. Z 2, 1028b 24–7. See Ross 1958: i. pp. lxxiv ff.

236

Sarah Broadie

incorporeal principles. Or, while holding to the Timaeus doctrine that the cosmos will last for ever, Xenocrates may have been convinced by the general arguments which convinced Aristotle that what is destined never to perish can never have been begun.¹⁸ There were also the more specific arguments which convinced Aristotle that orderly physical motion always has gone on and always will.¹⁹ There were plenty of reasons why these thinkers—whether accompanied or led by Plato himself, but at least not held back by him—should decide for an eternal cosmos. We are about to see that, given this decision, their self-identification with the author of the Timaeus required them to distance themselves from a literal understanding of the text.²⁰ Consider the difference for Platonic metaphysics between a cosmos that came into being from pre-existing materials and an ungenerated cosmos composed of materials but never preceded by them. The ‘ungenerated’ account has no place for a Demiurge understood literally as an efficient causal principle distinct in nature from the intelligible paradigm to which he ‘looks’.²¹ In the Timaeus, literally understood, the unformed matter and the cosmic form are two principles which cannot unite to produce a cosmos unless brought together by a third one, the Maker. But if the matter has necessarily always been imbued with the form, then either no explanation is needed for why it is thus formed, or the mediating link would have to integrate them rather in the logical way a syllogistic middle term unites the extremes, or an intermediate conclusion in a demonstration is a conduit to the final one. The same with the question of what unites the incorporeal elements of the World-Soul into the Soul itself. Even if we are not trying philosophically to understand the causation of the cosmos from simpler ingredients as a timeless relation, the sheer absence of a cosmic beginning beggars our resources for imagining a distinct Demiurge. When we seem to grasp the thought of an artificer distinct from the ideal model and from the cosmos itself, don’t we imagine him starting to attend to the model, or gradually arriving at his full vision of it,²² or turning from merely contemplating that form as a thing of beauty to treating it as a model for something else? Perhaps, to bulk out an ontological distinction between Demiurge and model, we picture him withdrawing when his share of world making is done.²³ But how could he let go if his completed work was not also in some sense begun? If he had always been at it without beginning, it could not consist of a completable set of stages. But then there is no culmination upon which he can step back. ¹⁸ Arist. Cael. I 10–12. ¹⁹ Arist. Ph. VIII 1–2. ²⁰ For the history of contributions to this literalist versus non-literalist debate, see Cherniss 1962: 421–31, and Sorabji 1983: 268–76. ²¹ Pl. Ti. 28a, 29a. See Menn 1995: 1–5 and 9 for detailed discussion of the necessity for the distinct divine efficient cause. However, Menn seems to see no difficulty in retaining this notion in an eternalist framework (p. 61), by contrast with e.g. Dillon 2003: 81. ²² Ti. 30b. ²³ Ti. 42e.

Why no Platonistic Ideas of Artefacts?

237

If we are forbidden to imagine along any of those lines, the distinct Demiurge vanishes.²⁴ Either he merges, Stoically, with matter to become an aspect of the cosmos, or, if Platonism is retained, he merges with the intelligible cosmic form.²⁵ This form, then, has to be a changeless no¯eton accessible to some extent by minds like ours, since this is how forms were conceived of from the very first; but it must also now be understood as self-sufficiently²⁶ causative, whether as a paradigm inspiring something primally indeterminate to imitation or as a template generative in some other way. Now what about forms of human artefacts? It is vital for any sort of Platonist that these still count as intelligibles and remain changeless in every way necessary for intelligibles as such. For nothing has altered the fundamental fact behind Plato’s original argument for there being an intelligible structure of the cosmos at all. The argument was that since the cosmos is as good as such a thing can be, its causation must have been such as to ensure this excellence of result. Its causation, then, is to be understood on the model of the best human production. But, as a matter of familiar fact, best human production works, where possible, from a conception of the nature of the object desired which someone has thought out from relevant first principles, rather than by simply copying, without independent analysis of the nature of the goal, an already existing and for all one knows damaged example of the kind. Where the object to be made is a physical one, an already existing example would of course be a sensible thing, and the Platonist immediately brings this into comparison and contrast with the intelligible object that guides intelligent, good, human production.²⁷ Thus it would be absurd to question whether human artefact-forms are intelligibles, no¯eta. But a human craftsman’s no¯eton is of itself inert. It may be productive in the sense that the thinking which brought it to light was done with a view to production, and the thought content will reflect this (the geometer thinks of a sphere, the billiard-ball maker of a synthetic ivory sphere); but even granted suitable matter, this intelligible can do nothing on its own. My suggestion at this point is, first, that in the Xenocratean programme of deriving the eternal physical universe from incorporeals, the focus on Platonistic Ideas was necessarily a focus on intelligibles that are a se causative of the corresponding physical objects; and secondly, that from this perspective the intelligibles of human productive crafts stopped counting as real Ideas at all. The early debates over Ideas were about the relationship of Ideas to the objects of mathematics, about the relationship of Ideas and mathematicals to ‘the heaven ²⁴ Here I welcome confirmation from David Sedley’s forthcoming book, which brings out the intimate connection between cosmic beginning and divine craftsman in the Timaeus. ²⁵ Thus the Timaeus itself at 52–3, where in the formation of the elementary particles god and the forms are not distinct principles. ²⁶ i.e. along with matter if this is assumed as an independent principle. ²⁷ Ti. 28a–29a. On not taking a broken object as model, cf. Cra. 389b.

238

Sarah Broadie

and the world of nature’,²⁸ and about the relationship of all the above to the One and the Indefinite Dyad. In these contexts, if the term ‘Idea’ needed illustration, one would hardly reach first for Couch or Shuttle as examples. As the role of Ideas as a se cosmogonic causes became more and more an entrenched tenet, worries that Plato’s own Couch and Shuttle cannot possibly be a se cosmogonic causes would have seemed increasingly old-fashioned and irrelevant. To those at the cutting edge of the new debate, the Couch and its congeners, whatever they are, would be outside the field of discourse on such questions as whether Ideas are numbers. (A corollary: to admit Ideas in that sense for Shuttle and Couch—a sense shaped by the increasingly cosmological role of Ideas—would be to fall into a ‘pit of nonsense’.²⁹ It would be to postulate transcendent causes regularly generating shuttles and couches in their times and seasons world without end, rendering human craftsmen idle from all eternity except perhaps for those, such as they are, who only make superficial semblances of existing particulars. It is reasonable to proportion effect to cause, hence to reserve eternal principles, such as the Ideas are supposed to be, for causing what endures for ever in the natural world: immortal individuals such as the heavenly bodies, and fixed kinds of mortals perpetually instantiated.³⁰ But artefact-types come and go patchily in history, even if lost skills and sciences are always eventually rediscovered, as Aristotle held.) We have been looking at the consequences of removing the literal Demiurge, distinct in being and nature from the Idea of the cosmos, and replacing him with an Idea fitted to generate the cosmos on its own. Along the way we have moved from talking about the Idea of the cosmos in the singular to talking about Ideas in the plural, which is how we meet them in Aristotle’s Metaphysics and Peri Ide¯on. This slide needs to be justified. The justification, I suggest, is that removal of literal Demiurgy at the top carries with it removal of the divine Demiurgic ancillaries which are an essential part of the Timaean story. These gods were created to craft the mortal parts of the perishable creatures within the cosmos, because without them it would be incomplete, and the first Demiurge in his goodness could not destroy (or, presumably, build perishability into) anything ²⁸ Arist. Metaph. Λ 7, 1072b 14. ²⁹ Pl. Prm. 130d. ³⁰ Cf. the problematic of Arist. Metaph. B 4, 1000a 5–1001a 2. My emphasis on perpetual instantiation may seem at odds with Aristotle’s objection at Gen. corr. II 49, 335b 18–20: ‘If Ideas are efficient causes, why don’t they always generate their objects continuously (dia ti ouk aei gennai sunech¯os), rather than sometimes and sometimes not (pote men, pote d’ou), given that there always exist both Ideas and participants?’ However, the point there is surely that factors which are always available cannot alone explain the seasonal intermittency of generation. aei at 18 may mean ‘in every case’; the point then is: certain physical substances, i.e. the heavenly bodies, are being continuously generated by Ideas plus whatever participates; so how in other cases can exactly the same sorts of principle generate in a way that allows their effects to be merely seasonal? (It is not clear whether pote men, pote d’ou is temporal (cf. 335b 3), contrasting with sunech¯os, or non-temporal and contrasting with aei in its non-temporal sense.)

Why no Platonistic Ideas of Artefacts?

239

he made himself.³¹ In the completely depersonalized perspective of Xenocratean metaphysics, these Demiurgic deputies become identified each with the Idea of a natural kind.

I I I . A N I M P L I C AT I O N That these Ideas are causes without benefit of distinct craftsmen, whereas no Ideas are postulated for human artefacts, surely tells us that the notion ‘human craftsman working from a plan reached through thinking’ and the notion ‘a se causative Idea’ perform the same explanatory function in their respective areas. If, therefore, the Platonists excluded Ideas from the areas of human artifice, this arguably is because the contribution here of distinct crafting agents renders Ideas absurd because gratuitous.³² Ideas were a theoretical necessity for Platonism because, if you mentally subtract the distinct semi-personal Demiurges, the natural world and its organic contents still remain as data to be explained—and it is still true that matter by itself ³³ cannot account for them. As for pots, statues, shoes, ships, houses, couches, and shuttles: if you think away the human demiurges of these, you are left with nothing to explain, since on that hypothesis there would never have been such objects in the first place. Even so, pots, statues, etc. are kinds whose being and coming to be would definitely need some Idea-like explanation if they occurred in the absence of human craftsmen. They are not objects of which we, or at any rate the Platonists, could imagine a rational person saying, in that hypothetical situation, either ‘There is no explanation’ or ‘Matter by itself is explanation enough’. Thus (a) one does see those objects as needing explanation, and (b) when, after mentally subtracting the human artificers, one adds the latter back in, this is enough: one does not bring in Ideas, for one sees no work for Ideas to do. In short: one is treating human artificers and what they do as a genuine and satisfactory explanation of the being and coming to be of pots, couches, statues, and so on. Reader, if like me you feel that I am belabouring this point, you see how obvious it is. I wish to suggest—what is perhaps slightly less obvious—that the same point would have been equally obvious to everyone in the ancient debates about Ideas. That is, everyone including the Platonists was happy to accept the skill and activity of human artifice as a genuine and satisfactory explanation (aitia, aition) of the coming to be and being of pots, statues, and the rest. Someone may say: sure, outside their seminar, Platonists must have accepted the truth of ‘There are pots because potters make them’—they too ordered their ³¹ Pl. Ti. 41a–c; 42d–e. ³² The content of this reasoning is captured by Fine (1993: 290 n. 12), and well stated by Judson (2000: 133, except that line 9 has ‘with matter’ instead of ‘without matter’, a particularly unfortunate lapse at the centre of a dense exposition). Both scholars ascribe the reasoning only to Aristotle, not also (the point for which I am about to argue) to the Platonists. ³³ Whether empirical elements or the Receptacle.

240

Sarah Broadie

pots from the potter—but why should we suppose that as philosophers they took such a truth to constitute a satisfactory explanation? Well, one reason is that this hypothesis makes excellent sense of their very philosophical embargo on artefact-Ideas. The embargo is neatly explained on the assumption that from the Platonists’ own perspective, when it came to accounting for artefacts, explanation in terms of ordinary this-worldly crafting agents had never been dislodged. Another piece of evidence: In some instances we see that the cause is other than the form (the¯oroumen allo to aition on). For it is the doctor that implants health and the man of science who implants science, although health-itself and science-itself are as well as the participants (methektika, i.e. the patients being healed or taught), and the same principle applies to everything else that is produced according to a capacity (dunamis). (Arist. Gen. corr. II 9, 335b 20–4).³⁴

The context is an argument against the Platonists (335b 9 ff.). I take Aristotle to be speaking as someone close to them, and on their behalf as well as his own, when he says that in some instances ‘we see’ that the cause is other than the form (i.e. Idea). If so, this was a point which the Platonists accepted in the context of philosophical debate. Of course, Aristotle’s meaning may only be that common sense sees that the doctor and the teacher are effective. At whom, then, is he brandishing this piece of common sense? At the Platonists, on the assumption that philosophically they deny it? A reminder that the denial would baffle common sense is unlikely to change their minds. Or is he addressing persons similar to most of his appreciators today, who set great philosophical store by common sense—is he underlining a piece of common sense to them? If so, then again he is not trying to change minds. But if, as one is surely entitled to assume, Aristotle is here arguing to bring round some highly non-commonsensical people (who were his friends³⁵), then we should read him in this passage as appealing to a premiss which he and they both find philosophically plausible.³⁶ The examples at Gen. corr. 335b 20–4 are interesting for our purposes because they fall under the epist¯emai as contrasted with technai according to the Peri Ide¯on division (sect. I above). Since the Platonists postulate an Idea for every epist¯em¯e, and medicine is an epist¯em¯e, they postulate Ideas where everyone in the debate sees that in some instances the Idea does not do the causal work (or not all by itself ). Even if it does all the work when health or recovery owes nothing to human medicine, we also see instances where a human practitioner makes the difference. But if it is obvious on all sides of the debate that when a practitioner heals, ‘the cause is other than the form’, the same is even more obvious to all when what we have is a house being built or a shuttle shaped, since with such objects human causation is always necessary. In this way, then, Gen. corr. 335b 20–4 is ³⁴ Trans. H. H. Joachim in Barnes 1984, except that I have not followed Joachim’s capitalization. ³⁵ Arist. Eth. Nic. I 6, 1096a 11–17. ³⁶ See also Arist. Metaph. H. 3, 1043b 18–21, where he says that it is plain (d¯elon) that there cannot be separate substances for such things as house and utensil.

Why no Platonistic Ideas of Artefacts?

241

further evidence that Platonists too saw the being and coming to be of pots, houses, etc. as entirely adequately explained by the various artificers. Finally, the Timaeus helps prove the point. Plato chose his concept of the divine Demiurge partly because it gives a highly satisfactory explanation of a certain explanandum literally understood: namely, the genesis of the cosmos. The eternalizers’ misgivings were directed at this literal understanding. No sign of anyone complaining that, even granted a literal start to the cosmos, Plato’s divine craftsman would fail as a proper explanation of it. Given the topic, we can be sure that any fault found on this score would have been recorded as an important element in the case for the eternity of the world. If, then, on the to them false assumption that the world began, the Platonists saw nothing wrong with the divine Demiurge as an explanation, they must have seen nothing wrong with human demiurges as explanations of the coming to be of their products. If they thought that Plato reached his cosmic story by analogy from a prior judgement that human demiurges are adequate explanations, then in accepting the cosmic story as explanatorily adequate on its own misguided terms, the Platonists were accepting Plato’s prior judgement about the human case. If, rather, they thought that the explanatory adequacy of the cosmic case was self-evident, then working by analogy in the opposite direction, they would take it that the human demiurge is explanatorily adequate too.³⁷

I V. A W E A P O N AG A I N S T T H E P L ATO N I S TS ? If this account is correct so far, we see what a powerful argument Aristotle has against the Idea-theorists³⁸ if only he can get them to accept that the genesis of organic natural substances is on a par with that of artefacts. That would be one premiss of the argument. The other premiss the Idea-theorists already accept: that there are no Ideas, because no explanatory need to postulate them, for artefacts. They would then be forced to conclude that the same holds for organic natural substances.³⁹ ³⁷ There are passages such as Metaph. A 9, 992a 24–6, where Aristotle accuses the Platonists (including himself here) of failing to say anything about h¯e aitia hothen h¯e arch¯e t¯es metabol¯es: i.e. about what we know as Aristotelian efficient causality, which is exercised by individuals in this world. But if my argument in sect. III is right, then Aristotle certainly does not think that contemporary Platonists completely fail to recognize ordinary efficient causality. Their problem is that they recognize it only for human artificers, or perhaps human thoughtful agents in general. The context of 992a 24–6 shows that Aristotle is complaining about the inadequacy of the Platonistic principles to explain the natural phenomena. Similarly, when he describes first philosophy as ‘lisping about everything’ (A 10, 993a 15–16); he means: it is inarticulate about the principles in terms of which we should investigate the objects of theoretical inquiry, i.e. the heavens, the whole of nature, and any incorporeal substances. This is compatible with having a reasonable understanding of the house-builder as efficient cause of the house. ³⁸ Powerful but not knock-down, since the Ideas-theory was overdetermined. ³⁹ That this argument was at hand should shed light on Metaph. Z 7–9, 1032a 12–1034b 19, and on the question of these chapters’ contribution to Z. They are about natural, artificial, and

242

Sarah Broadie

If we read Aristotle as appealing where possible to positions common to him and the Idea-theorists, then our first passage above, from Metaphysics A, suggests that those men did not disagree with Aristotle that in natural genesis there has to be a down-to-earth activator of change, just as there is for artefacts. I quote this time a slightly modified version of the Ross translation: In the Phaedo the case is stated in this way—that the forms are causes (aitia, neuter) both of being and of becoming; yet [point A] when [i.e. in the cases where] the forms exist, still the things that share in them do not come into being, unless there is that which will get the change to occur (to kin¯eson [Ross translates ‘unless there is some efficient cause’]); and [point B] many other things come into being (e.g. a house or a ring), of which we say there are no forms. Clearly, therefore even the other things [i.e. the ones for which there are supposed to be Ideas] can both be and come into being owing to such causes (aitias) as produce the things just mentioned. (991b3–9)

The cases where ‘the forms exist’ are natural objects. Point A, I take it, as well as point B, is something the Idea-theorists accept. The reason for supposing this is as before. Aristotle will then be making an argument that might actually get somewhere with his opponents. But it is important that he and his translator not represent the latter as giving to ‘that which will get the change to occur’ the honourable status of explanatory cause (aitia, aition). This would go a good way towards contradicting their defining doctrine, and they could complain of ignoratio elenchi. On the other hand, they do not think—how could they?—that spontaneous comings to be. They are generally held to have been written independently of the rest of Z and inserted later by Aristotle. Burnyeat, in his learned and exciting 2001, sees evidence for this in the fact that Z 7–9 focus on artefacts: which to Burnyeat is out of kilter with the context. Burnyeat holds, and I do not disagree, that Z ‘has a single positive purpose, to which all its dialectical ingenuity is directed: to show that substantial being is form’ (Burnyeat 2001: 3, emphasis added). (‘Form’ here is, of course, Aristotelian form.) Hence a ‘careful reader’, says Burnyeat, ‘should be left gasping with surprise when they find that Z’s first explicit enunciation of the thesis that primary substantial being is form occurs in a parenthesis [1032b 1–2] about artefacts’ (p. 30, emphasis added). Moreover, since Z 4–6 ‘have laboured to tie essence, in its primary and strictest sense, to [substance]’, it is ‘scarcely consequential for Aristotle to proceed immediately in Z 7 … to equate essence with forms such as the form of a house (1032b 12–14)’ (p. 30). This is because in the central books of the Metaphysics Aristotle is clear that artefacts are not substances (for references, see n. 14). I agree that Z 7–9 are an insertion, but I question whether the emphasis there on artefacts makes the insert, as Burnyeat supposes, unsuited to its present context. Prominent throughout Z is a running battle with the proponents of separate Ideas (for this in 7–9, see 8, 1033b 19–1034a 5). On this front the artefact theme would directly succour Aristotle, given the Platonists’ acceptance that artefacts come to be without benefit of corresponding Ideas—if only they could also be persuaded to accept that the genesis of organic natural substances is relevantly similar. But the latter is one of the main things that Aristotle is trying to show in Z 7–9. He makes it the ground of his anti-Platonist conclusion at 8, 1033b 26–9. This is reached via an obscure argument at 19–24 securing the answer ‘No’ to ‘Does crafting a (bronze) sphere or a house entail that there is an Idea of Sphere or House?’. Aristotle then infers on the basis of nature–art similarity (h¯osper, 24) that there is no Idea for Man, etc.; cf. 1034a 2–5 and Λ 3, 1070a 13–30. ( The similarity he invokes to swing the argument is that in both cases producer and product are the same in form—the ‘synonymy principle’, as Burnyeat calls it. Burnyeat sees this as the ‘thread that runs all the way through’ Z 7–9 (Burnyeat 2001: 35). Likewise Judson 2000.) Arguably, then, the prominence of artefacts in these chapters was a large part of the reason for inserting them.

Why no Platonistic Ideas of Artefacts?

243

each instance of natural genesis just starts by itself. They think, like everyone else, that for the most part natural comings to be require down-to-earth agents to get them going. They do not expect saplings where no seed has dropped, or children without coition. How is this supposed to be combined with the metaphysical causality of the Idea? Well, presumably the Idea is in some ways like a modern law of nature according to a realist view of laws. Or it is a sort of commandment inscribed in the unshaken heart of incorporeal Truth, and physical things of the relevant kind ‘obey’ it. Such metaphysical entities would never be postulated at all if doing so clearly left no role for ordinary objects and ordinary events in implementing natural processes. One could even, reaching for a different model, think of the Idea as using the seed or the parent animal as a means (one sort of sunaition or sine qua non) whereby a new creature of the kind comes to be.⁴⁰ The upshot at this point is that in effect the Idea-theorists, like everyone else, accept such truths as ‘Man generates man’. Presumably, they know that they themselves had parents. But they do not accept that the existence and activity of parents satisfactorily explains, is aitia or aition of, the genesis of progeny.⁴¹ On the other hand, they do accept ‘Housebuilders make houses’, not only as true but as perfectly explanatory of the genesis of houses. Aristotle’s task is to get them to see that the cases are on a par; and he sets about it by famously stressing how, in both, the form of the result is already present in the agent.⁴² (A general moral: when in the Metaphysics Aristotle invokes artefacts to help explain his theory of natural substances, this is not always only because they provide beautifully clear illustrations of the four Aristotelian causes, particularly the form/matter distinction, and of the way in which actuality and potentiality are intertwined in perishable things. These are concerns of the physical works too; but in the Metaphysics the parallel with artefacts sometimes sends a message peculiar to first philosophy: namely, that artefacts do not require Ideas, and natural substances are relevantly like artefacts.) The interpretation which I have been suggesting raises the question: what hindered the Idea-theorists from seeing for themselves the force of the considerations which make so clear to Aristotle the needlessness of Ideas for natural substances? First, they, like everyone, accept the down-to-earth artificer and his activity as a perfectly satisfactory explanation of the genesis of an artefact. Secondly, they, like everyone, are aware of the regular fact that man generates ⁴⁰ Cf. Pl. Soph. 265c, where ‘everything that grows from seeds [this includes semen] and roots’ is said to come to be through the action of a god d¯emiourgountos. See also Pl. Phdr. 272b2–5 for an implicit ‘separation’ of the aitia (fem. noun) from the obvious agent of change: the change is that of being persuaded; the aitia is, or is in, the condition of recipient soul plus the nature of the speech; the obvious agent is the orator. ⁴¹ On things true but not explanatory cf. Arist. Eth. Nic. VI 1, 1138b 25–6, with 32–4; Eth. Eud. I 6, 1216b 32–9. ⁴² Arist. Metaph. Z 7–9, Λ 3–4. For recent detailed studies of the Lambda chapters see Judson 2000 and Crubellier 2000.

244

Sarah Broadie

man, lion lion, and so on. Thirdly, it cannot have escaped them, since it is simply dressed-up common sense, that the form of the biological offspring is determined by that of its parent just as the form of a brick house is determined by the form in the soul of the builder (for the builder is not a house). So, since this is all so obvious, why are they not already Aristotelians? What do artificial comings to be have and natural ones lack, or vice versa, that makes it all the same plausible or attractive to these philosophers to hold that without Ideas there is no proper explanation of natural genesis? For clues to an answer to this question we shall turn again to the Timaeus. Only an answer, given the likelihood of overdetermination. Nor can we take it for granted that the Idea-theorists themselves would have had any definite response to our question. They may have been so deeply committed to the Ideas of Man, etc. that even if they could give no clear independent reason for taking contrary attitudes about the explanatory value of ‘Man generates man’ and ‘The builder builds a house’, they would have continued to reject the first as inadequate, since otherwise they risked forfeiting the Idea. A great deal was at stake. Arguably, Xenocrates’ entire programme of deriving the world of nature from incorporeal substances depended on postulating causative Ideas for the natural kinds.⁴³

V. T H E TIMAEUS AG A I N , A N D T H E T H R E P T I C S O U L The fact that man comes from man, lion from lion, is not just a matter of like from like, or we should have house from house and no need for human house-builders. What artefacts lack and living mortals possess that explains the difference is to threptikon, as Aristotle calls it: that mode of soul whose functions are nutrition, growth, and reproduction.⁴⁴ At the philosophical juncture we are studying, an educated sense that ‘man [referring to particular individuals] generates man’ is genuinely explanatory stands and falls, I think, with confident command of the concept of to threptikon. The Platonists lacked this. What the Timaeus offers on this score is rudimentary and unpromising, as we are about to see. Here are the familiar essentials of Aristotle’s developed concept: (a) each individual organism (plants included) has and is animated by its own nutritive soul; (b) this psychic mode is not mental, let alone intellectual; (c) it and its activities are not properties, powers, or effects of basic matter as such: a true explanation of the fundamental characteristics of the elements of the ⁴³ Supporting this claim must be postponed. If it is correct, one can admire the argument of Metaph. Λ for the way it ensures nature’s dependence on the incorporeal without compromising the explanatory adequacy of ‘man generates man’. ⁴⁴ Arist. De an. II 4.

Why no Platonistic Ideas of Artefacts?

245

universe would not include mention of any sort of soul; (d) the nutritive soul is formative: it controls the shaping, differentiation, and arrangement of the parts of the organism throughout life, and it propagates this physical form through reproduction. Prima facie such a principle is simply absent from the Timaeus’ account of the mortal living creature. In the first place, the creature’s soul has just the parts familiar from the Republic: immortal intellect, spirit, and appetite. The last two are presented in terms of emotions and desires. Appetite of course relates to nutrition and reproduction: it desires feeding and sex, and thereby contributes to the survival of organism and kind.⁴⁵ However, the design problems posed by appetite, along with spirit, are mainly about ethical as distinct from biological well-functioning. (How ensure for intellect minimal disturbance from the unruly other two? Locate it in the head and them at one and two physical removes away. How bring it about that they can still be affected by reason’s requirements? Create organs—specifically, heart, veins, and liver—to mediate the communication.⁴⁶) Secondly, the body’s frame, organs, and tissues are all crafted by the divine ancillary Demiurges. Thirdly, the basic physiological processes of nutrition, respiration, and ‘irrigation’ are explained as resulting from the natural behaviour of elemental matter under confines of relevant parts of the anatomy.⁴⁷ For example, nutrition is due first to the action of fire, which by its natural sharpness chops food into blood and by an oscillation takes this from demiurgically constructed belly into demiurgically constructed veins, and then to the tendency of the elements in the blood (following the law that prevails in the universe at large) to cluster, kind to kind.⁴⁸ Fire and the other elements are of course inanimate and unintelligent. The nutritive oscillation of fire is due to the in-and-out of air in respiration, which in turn is due to the impossibility of a vacuum together with the natural tendencies and forces of hot and cold air in a confined space.⁴⁹ If this were all, one could conclude that the Timaean mortal organism has nothing like the Aristotelian threptikon. All the biological formation mentioned so far is done by divine intellect, not by anything belonging immediately and exclusively to the mortal individual itself. In addition, the latter’s three-part soul as so far presented has no power to effect biological formation. Our intellect resembles the divine enough for us to invent and make artefacts, improve on nature (as in medicine and agriculture), and study the sciences: but ‘without our aid’ were our ⁴⁵ Plants too have this lowest psychic level, which even in them is characterized by reference to pleasant and painful sensations, and appetites (epithumiai). Seeds and roots are mentioned, but (pace Taylor 1928: 542) no connection is drawn between the soul of a plant and its growth, feeding, and reproduction (Ti. 77a–c). ⁴⁶ Pl. Ti. 69c–72a. See Johansen 2004: ch. 7 for an illuminating discussion of this and related themes. ⁴⁷ Since the anatomy has been designed with a view to these effects and the benefits they bring, the explanation is teleological as well as mechanical. ⁴⁸ Pl. Ti. 70e, 80d–81b. ⁴⁹ Ti. 79a–e.

246

Sarah Broadie

bodies actually constructed.⁵⁰ Appetite and spirit set going, and in some cases have as objectives, various bodily processes, but these, from what we have seen, are mechanical effects of the properties of pre-organic elements placed under divinely crafted anatomical conditions. The explanations of these structures and processes have unfolded strictly within the boundaries set by the two great principles of Timaean cosmology, Intellect and Necessity.⁵¹ This Intellect is the maker of the world,⁵² and this Necessity is the matter, first and foremost, of the world.⁵³ There is nothing belonging to the individual organism itself as such that could be the source of its, or any, biological formation. On this basis one could not endorse ‘man generates man’ as a real explanation of the genesis of human beings—unless one understands the subject term not as referring indefinitely to particular members of the human race but as naming an Idea or Demiurgic paradigm, with the verb correspondingly referring not to a physical process that begins and ends in sensible particulars, but to a relation between different metaphysical levels. But this is not all. Timaeus’ brief remarks on reproduction hint at a different direction. The fullest passage occurs only at the end of the cosmology,⁵⁴ but not as an afterthought. We know this because the triple motif of birth, nutrition, and growth, with the associated theme of death, has already been introduced in words of deep solemnity at one of the great turning points of the Timaeus: the moment where the principal Demiurge hands over to his ancillaries the task of completing the cosmos by creating mortals.⁵⁵ Let us look now at the passage at the end. It shows the divinely crafted male and female organs as quasi-autonomous living beings⁵⁶ and agents of coition. The male system incorporates channels for the semen, which (on the theory adopted by Plato) is marrow from the brain that passes through the spine. The marrow, like the sex organs, is said to be ensouled (empsuchon).⁵⁷ Frenzied eros and appetite bring male and female organs together: ⁵⁰ However, in the Phaedo it is suggested, although not by Socrates, and as part of an extended simile, that the immortal soul, which is intellectual, weaves its body for the next incarnation (Phd. 87b–e). By contrast, at Ti. 41d it is the task of the ancillary Demiurges to weave the mortal body on to the immortal soul. ⁵¹ Ti. 47e–48a. ⁵² The master Demiurge gives his ancillaries a sketch of the whole of which their work will be part (Ti. 41a–d); thus they are not blinkered instruments but offshoots of world-making intellect as such, and their operations are not uncoordinated (cf. Menn 1995: 3–4). ⁵³ Cf. Ti. 42e–43a, where material for the bodies of mortals is said to be only ‘borrowed’ from the cosmos. ⁵⁴ On this placing see n. 59. ⁵⁵ Ti. 41d1–3. I understand gennate (‘bring them to birth’) not as a variant on apergazesthe (‘make them’), but as meaning ‘give them the wherewithal of reproduction’, just as ‘feed them’ means ‘give them the wherewithal of nutrition’ (including something to eat; cf. 77a–c). The same triple motif occurs at Resp. VI, 509b3–4, in connection with the Sun. ⁵⁶ They are called z¯oa three times: Ti. 91a2–3, b6, c2. Likewise the appetite for nutrition was earlier shown as an untamed animal (thremma agrion) in its own right, tethered to its manger in the belly (70d–e); but this seems more deliberately metaphorical. ⁵⁷ Ti. 91b2; cf. a3.

Why no Platonistic Ideas of Artefacts?

247

And they sow the ploughland of the womb with living creatures still unformed (adiaplasta) and too small to be seen, and again differentiating (diakrinantes) their parts nourish them till they grow large within, and thereafter by bringing them to the light of day accomplish the birth of living creatures. (Ti. 91c–d)⁵⁸

Here, then, the agents of biological formation are the most unintellectual, mortal, and corporeal objects in the world: they are the organs of reproduction.⁵⁹ Since lust for coition serves the same end as the embryological shaping to which it leads, lust and the physical activity of the organs appear really as stages of a single process.⁶⁰ Consequently, since the lusts on either side are very clearly properties of the creatures themselves or their organs, the embryological shaping is likewise theirs, i.e. done by them, not by the gods, however much it implements the divine goal of a cosmos unceasingly complete with mortal animals. It would seem that divine intellect crafts prototype creatures, whereas the intricate business of propagation is assigned to witless mortal subsystems. This general distinction is confirmed by our next passage. Here Plato introduces marrow as the primary seat of soul in mortal body: With bone, flesh, and all substances of that sort the case stands thus. The starting point for all these was the formation of the marrow, for the bonds of life, so long as the soul is bound up with the body, were made fast in it as the roots of the mortal creature; while the marrow itself is formed of other things. The god set apart from their several kinds those triangles which, being unwarped and smooth, were originally able to produce fire, water, air, and earth of the most exact form. Mixing these in due proportion to one another, he made out of them the marrow, contriving thus a universal seed for every mortal kind (panspermian panti thn¯et¯oi genei). Next he planted and made fast therein the several kinds of souls (ta t¯on psuch¯on gen¯e); and from the first, in his original distribution (euthus en t¯ei dianom¯ei t¯ei kat’ archas), he divided the marrow into shapes … And he moulded into spherical shape the ploughland, as it were, that was to contain the divine seed; and this part of the marrow he named ‘brain’ (enkephalon, lit. ‘in-the-head’), signifying that, when each living creature was completed, the vessel containing this should be the head. That part, on the other hand, that was to retain the remaining, mortal, kind of soul he divided into shapes (sch¯emata) at once round-and-elongated, naming them all ‘marrow’. From these, as if from anchors, he put forth bonds to fasten all the soul … (Ti. 73b–d; trans. based on Cornford) ⁵⁸ Trans. Cornford (1935: 357). (diakrinein was to be used in the same connection by Diocles of Carystos; cf. Cornford 1935, ad loc.) ⁵⁹ Yet the connection with intellect and the divine is kept, in that brain marrow is the seat of the immortal, intellectual, part of the soul. Cornford (1935: 291–3), comments that this is why Timaeus does not deal with reproduction earlier, along with nutrition, as a function of to epithumetikon. On biological reproduction and immortality, see Symp. 206e–208b; Leg. IV, 721b–c. ⁶⁰ In fact, Plato virtually identifies the male and female organs with eros and appetite respectively (see Cornford 1935: 357 ad 91b, n. 1). Grammatically, ‘they’ in ‘they sow … nourish … accomplish’ at 91d refers to eros and appetite. By a similar functional blurring of physical distinctions, the male organ (in the form of semen) is at work inside the womb post coitum.

248

Sarah Broadie

The marrow functions as interface between the soul and the rest of the body. It is because the soul is ‘planted’ and ‘anchored fast’ in the marrow (or, perhaps, to the rest of the body by the marrow) that the body is a living animal. This must be squared with the earlier account locating the parts of the soul. So we are now told that one portion of marrow is shaped into a globular mass which is the source of semen: this, the site of intellect, will be encased by the cranium; and another portion, in which will be sited the non-rational sides of the soul (these for the moment being treated as one), is shaped into slim cylindrical masses destined for the other bones. This much is clear, and may exhaust what is said here about shapes of marrow. However, some take lines c1–6 to state a pre-formationist doctrine: the marrow seed contains the soul types and body types of all future mortal animal species.⁶¹ There are reasons for doubting this interpretation; but present space forbids that discussion, and the point hardly affects the main argument. More important for now is the clear message of the above passage that what the divine crafting principle made was a prototype of marrow and its distribution. We are told that the arrangement, whatever exactly it included, was made ‘in the beginning’.⁶² We are told, by implication and via a single illustration, that an instruction was given to be followed when ‘each living creature was completed’ in future. For the god’s ‘naming’ the original brain matter as he did is surely tantamount to his leaving an instruction that this particularly precious kind of marrow is to be situated in the head, the location of immortal intellect. The formation in this respect of future animals, human or subhuman, will turn out right if the god-given name is lived up to, and encephalic marrow not mis-assigned to a pelvic or thoracic bone.⁶³ The instruction expressed by the name takes over now from the divine inventiveness that created the prototype.⁶⁴ In general: the Timaeus shows us divine wisdom creating prototypes of all the mortal kinds and their parts, but not itself doing the humdrum work of biological reproduction.⁶⁵ Even though this work crucially involves formation, the agents are the animals ⁶¹ Cornford 1935: 294–5, encouraged by Rivaud 1925. The evidence on which they lean includes 73c4–5, which I have not fully reproduced in the quotation above as being too ambiguous to present without detailed commentary. ⁶² Ti. 73c6; cf. 44a8–b1. ⁶³ Cf. Pl. Cra. 391d: ‘Socrates: Surely, the gods call things by their naturally correct names?’ ⁶⁴ We are not told that the name instruction was inscribed as a shape in the marrow, or that it comes with instructions about future animal species; so this passage does not support Cornford’s pre-formationist interpretation. ⁶⁵ Cf. Burnyeat 1999: 248 n. 66. For the creation of the prototype anatomy, physiology, and psychology of human reproduction, see Pl. Ti. 91a–d. For the first bodies of birds, non-upright land animals, and fishes and molluscs, see 91d–92c. Some of the language is consistent with the theory that these new forms evolved in an automatic (although divinely sanctioned; cf. 41e–42e) way from imperfect humans (see also 76d–e), but divine construction is spoken of at 91a, and ‘refashioning’ at 92b3; cf. a3, b5. Note that what I am calling the physical prototypes—the immediate work of the gods—are presented by Plato as pluralities: from the first, the Demiurge envisages quite a multitude of human beings (42a ff.).

Why no Platonistic Ideas of Artefacts?

249

themselves, even as they themselves, not divine wisdom, engage in their digestion and respiration.⁶⁶ Between them, the two passages just quoted suggest that the author of the Timaeus intended to welcome ‘man (e.g. Coriscus or Callias) generates man’ as a genuine explanation of the genesis of humans in the already instituted cosmos. It kicks in only in relation to descendants, hence does not compete with the explanation that refers to divine Demiurgy and guidance by a purely intelligible paradigm. For the latter is an explanation of the genesis of the prototypes. Discussion about whether Plato really meant to show the cosmos as beginning has focused on whether this conception is coherent, and modern discussion of its coherence has focused on whether sense can be made (consistently with the text) of one or both of the following notions: (N1 ) that the materials of the cosmos pre-existed their formation into the whole, and (N2 ) that formation occurred in successive stages.⁶⁷ However, the idea of a real beginning would also be supported by the notion (N3 ) (if it is viable) that mortals today are descended from physical prototypes. N3 would give us a set of creatures (some for each species) that are (i) non-arbitrarily first, and (ii) in each case temporally succeeded by their descendants. (This only applies once prototype reproductive systems have appeared.) To a scientific theist like Plato, N3 would surely be attractive prima facie. For it permits naturalistic explanation of comings to be within the already instituted cosmos while insisting on full and immediate divine responsibility for the original kinds, mortal as well as immortal. But can we in the Timaean context be satisfied with N3 ? I have compared reproduction with digestion and respiration: these processes were divinely designed, the wherewithal for them divinely built, but the animal itself then conducts them. It has to do its own living, and these are aspects of that. Plato’s detailed accounts of these processes make it intelligible how the agency for them is the animal itself (or an organic subsystem) operating as a machine driven by materials from the environment. But what parallel story can be told about reproductive formation? The mechanical accounts of ‘irrigation’, respiration, and nutrition refer to already formed cavities and channels in which fire, air, etc. behave as you would expect. How could such an account be given of embryonic formation? If there are genetic ‘instructions’, how do they work? Are they minute shapes inscribed in the marrow? But how would a set of shapes in marrow cause and govern formation of an entire organism? Do the adult reproductive organs and secretions which, according to Plato, differentiate the parts of the unformed foetus ‘heed’ the instructions?⁶⁸ If the instructions are ⁶⁶ This in some ways parallels the mythic situation of the Politicus, where the whole cosmos is first tended by the gods, then left to fend for itself. The contrasting idea of continuous (conveyed by the present participle) divine production of intra-cosmic objects is flashed at Soph. 265c. ⁶⁷ See Vlastos 1939 and 1964. ⁶⁸ Even though the male member is described as ‘disobedient and autonomous, like an animal incapable of listening to reason’ (91b6).

250

Sarah Broadie

incorporeal meanings immanent in physical shapes, what is it that understands them? If they are just physical shapes that work mechanically, how is it that they not merely inflict some shaping on something else but do so in accordance with the god-made prototypes? If these prototypes function as such by being examples, again what is it that ‘follows’ them? Moreover, these are physical prototypes: not merely of mortal creatures but perishable themselves. How can they or details within them continue to guide when they no longer exist? Or is it that biological form is transmitted from individual to individual rather as echoes, concentric ripples, and light rays are propagated through space or a physical medium? This would imply that biological form may get fainter and fainter as it is handed down. Is that consistent with an immortal cosmos guaranteed to be as zoologically complete as possible? VI. FINAL REMARKS Considering these difficulties about the Timaean physical prototypes as prototypes, one may wonder whether they contributed to the early Academy’s motivation for postulating a world sans physical prototypes: with species eternal in both temporal directions. At any rate, that postulate for whatever reasons came to be the conviction of many. I speculated on the metaphysical effect, regarding the Demiurge and his paradigm, of switching the Timaeus cosmogony into eternalist mode. We may ask a parallel question about the genesis of mortals. We saw that Timaeus’ N3 story mixes full divine responsibility (for the prototypes) with would-be naturalistic explanation (of their descendants). What happens to these features when the N3 story is exchanged for one that recognizes no absolute physical firsts? No mixing is possible. One or other feature takes over completely. The monodic alternatives are, respectively, Aristotelian and Platonic (Xenocratean). (1) Individuals in every generation are the effects, genuinely and intelligibly, of their natural parents, just as they are the genuine causes of their own offspring. Divine Demiurgy, or per se causative Ideas, contribute nothing, since there are no physical firsts for them to be immediately responsible for. Alternatively (2), there is a unique set of prototypes—they are none other than the incorporeal Idea-causes—which are equally immediately responsible for every physical individual of their respective kinds. Each physical individual stands to its Idea as ectype, and notwithstanding empirical relations of genealogy, none is the true formative origin of any other. Both alternatives entail the eternity of the cosmic animal itself, since it must always have been there to house the mortal generations. (One may wonder too where Plato stood in all this, both when he composed the Timaeus and later. Xenocrates, presumably meaning well, suggested that Plato all along was presenting an eternalist position in non-eternalist guise. That possibility cannot be ruled out. But now we see that it is false that Timaeus’ non-eternalist

Why no Platonistic Ideas of Artefacts?

251

account purveys, only more lucidly, the very same theories as an eternalizing Platonic one (alternative 2 above) would have done. On the contrary, it carries quite different metaphysical consequences concerning causation of mortal genesis in the world today. Thus Plato would have been mistaken if he had composed the Timaeus in the spirit claimed for him by Xenocrates. So on this point the actually more charitable interpreter was Aristotle, who insisted on taking Timaean non-eternalism at its word.⁶⁹ But later, Plato may have quietly decided not to stand by that aspect of the Timaeus.) How, to return to alternative 2, do Ideas cause physical particulars? Well, metaphysical details may be vague, but, broadly speaking, Ideas, first and foremost objects of intellect, can cause only through being somehow intelligized. In light of the Timaeus Xenocrates may have thought that a primal material principle ‘receives’ and so in some sense intelligizes or is made intelligent by them, and thereby comes to be formed into physical creatures.⁷⁰ Alternatively, even such an early Platonist as he might have tried to make sense of his commitments by postulating Ideas as self -intelligizing:⁷¹ an Idea is both an essence and an intellectual fiat that self-projects that essence into matter. Let us return to the question asked near the end of section IV. Given that the Platonists themselves (as I argued) accepted the house-builder and his art as a perfectly good explanation of the existence and coming to be of a house; and given that they, like everyone else, could see in a common-sense way that both the construction of human artefacts and the generation of biological kinds proceed by a sort of replication of form, why did they not just accept, with Aristotle, as a conclusion from those premisses, that ‘man generates man’ is not only true but a perfectly good explanation of a human’s genesis and existence, so that there is no need to invoke an Idea? Aristotle would then have been pushing at a door readier to open, and the Platonists would have moved closer to ridding themselves of the burden (as surely any metaphysician would regard it) of having to ‘separate’ each particular human substance (the Platonists did allow Callias and Coriscus this status) from the Substance of those substances. Their obduracy despite costs drew strength, I suggest, from the assumption that an adequate explanation must refer to something intellectual. This is what is missing from ‘One physical human or lion or coconut palm generates another’. It fails to cut deep. It purports to explain one sensible substance by another on metaphysically the same level; but how can that not lead to a pointless regress rather than an explanation? This is not so with ‘The builder builds a house’, precisely because, here, intelligized form or essence is translated into an external shape realized in matter. Although only humanly ⁶⁹ Arist. Cael. I 10, 280a 28–33. Sedley (forthcoming) makes a strong case for holding that the Timaeus-Critias is literal about a cosmic beginning. On charity in this connection, cf. Menn 1995: 61–2. ⁷⁰ Cf. Pl. Ti. 50a–53b on the Receptacle; Arist., Ph. I 9, 191b 35–192a 25; Metaph. A 6, 988a 7–14. ⁷¹ Such a view might seem to be suggested by Pl. Prm. 132b3 ff.

252

Sarah Broadie

intelligized and non-transcendent, the form ‘without matter’⁷² in the builder’s intellectual soul is metaphysically different enough from the explanandum to be a terminus of explanation, whereas biological parent and offspring seem equally non-intellectual effects of something, so that neither is source of form to the other. Metaphysically, it is as if the parent is sign rather than cause of what the offspring will be, or an indication to us of which Idea to refer to as the cause.⁷³ Aristotle’s metaphysics of organic generation (and mutatis mutandis of the organism’s self-maintenance) sweeps away the Idea by re-conceiving the natural generator. The particular generator lion which we see is the expression of a non-spatial interiority of that organism that specifies prescriptively the essence of lion, not by thinking but by a non-mental sort of intentionality. The generator lion, a particular, generates not qua external shape or shape system actualized in sensible matter. As such, this lion is an explanandum like its offspring (it is natura naturata), and an explanation which invoked just that would indeed be inadequate. No, the creature generates as its own and its offspring’s natura naturans, from the essentially non-sensible but non-transcendent prescriptive form that defines its nutritive soul. This is a sui generis kind of cleverness, a source of effects not due to matter nor yet to any sort of intellection.⁷⁴ REFERENCES Barnes, J. (1984) (ed.), The Complete Works of Aristotle, The Revised Oxford Translation. Princeton. Bonitz, H. (1870), Index Aristotelicus. Berlin. Burnyeat, M. (1999), ‘Culture and Society in Plato’s Republic’, the Tanner Lectures on Human Values. 20. (2001), A Map of Metaphysics Zeta. Pittsburgh. Cherniss, H. (1962), Aristotle’s Criticism of Plato and the Academy. New York. Cornford, F. (1935), Plato’s Cosmology. London. Crubellier, M. (2000), ‘Metaphysics Λ 4’, in M. Frede and D. Charles (eds.), Aristotle’s Metaphysics Lambda (Oxford), 137–60. Dillon, J. (2003), in G. Reydams-Schils (ed.), Plato’s Timaeus as Cultural Icon (Notre Dame), 80–94. Fine, G. (1984), ‘Separation’, Oxford Studies in Ancient Philosophy, 2: 31–87; repr. in Fine (2003), 252–300. (1993), On Ideas. Oxford. (2003), Plato on Knowledge and Forms. Oxford. ⁷² Arist. Metaph. Z 7, 1032b 11–14; Λ 3, 1070a16–17; cf. 29–30. ⁷³ There were probably circular currents: while technological examples fed the intuition that formation is due to intellect, thereby endorsing the role of Ideas in biological formation, general commitment to Idea-theory may have muffled the force of obvious non-intellectual examples: ‘No, the spider’s action doesn’t really explain the web, because spiders don’t act from thought’; cf. Arist. Ph. II 8, 199a 20–30. ⁷⁴ In writing this I have had the benefit of acute comments from colleagues in Edinburgh, London, and Cambridge.

Why no Platonistic Ideas of Artefacts?

253

Irwin, T., and Fine, G. (1995), Aristotle Selections. Indianapolis. Johansen, T. K. (2004), Plato’s Natural Philosophy: A Study of the Timaeus–Critias. Cambridge. Judson, L. (2000), ‘Metaphysics Λ 3’, in M. Frede and D. Charles (eds.), Aristotle’s Metaphysics Lambda (Oxford), 111–35. Leszl, W. (1975), Il ‘De Ideis’ di Aristotele. Florence. Menn, S. (1995), Plato on God as Nous. Carbondale, Ill. Rivaud, A. (1925), Platon, x: Timée, Critias. Paris. Ross, W. D. (1958), Aristotle’s Metaphysics, 2 vols. Oxford. Sedley, D. (forthcoming), Creationism and its Critics in Antiquity. Berkeley. Sorabji, R. (1983), Time, Creation and the Continuum. Ithaca, NY. Taylor, A. E. (1928), A Commentary on Plato’s Timaeus. Oxford. Vlastos, G. (1939), ‘The Disorderly Motion in the Timaeus’, Classical Quarterly, 33: 71–83; repr. in R. E. Allen (ed.), Studies in Plato’s Metaphysics (London, 1965), 379–99. (1964), ‘Creation in the Timaeus: Is it a Fiction?’, in R. E. Allen (ed.), Studies in Plato’s Metaphysics (London), 401–19.

13 Plato on What is Not Noburu Notomi

1 . WHAT IS NOT I N A N C I E N T PH I LO S O PH Y What is not (τὸ μὴ ὄν) was scarcely discussed in ancient philosophy before Plato. Although this phrase, or concept, made occasional appearances in philosophical arguments, it did not figure as their primary subject. Parmenides and his followers allowed no space for what is not, while the atomists, in contradiction to the Eleatics, equated it with the void. But in neither case is the concept subjected to any sustained discussion. The only exception in early Greek philosophy seems to be Gorgias’ treatise entitled On What Is Not (Περὶ τοῦ μὴ ὄντος), which is a parody of Eleatic arguments. Yet what Gorgias actually contends in the treatise is that there is nothing (οὐδέν), instead of what is not. This is clearly the upshot of the first of the three proofs, which runs as follows: if there is something, either what is or what is not (or both) is; however, neither of them is; therefore, nothing is. Admittedly, Gorgias deploys several dubious arguments concerning what is not in the course of trying to prove that nothing is. Nevertheless, the fact remains that he never investigates what is not in its own right. By contrast, its counterpart, what is (τὸ ὄν), occupied the spotlight ever since Parmenides identified it with the Truth. Melissus is reported to have written a This paper is a development of my previous study, Notomi 1999. The monograph was a revised version of my Cambridge Ph.D. thesis, submitted in 1995, under the supervision of Myles Burnyeat. In 2002 I also published a Japanese version of my English book, with two new appendices: the final appendix deals with the ‘Parity Assumption concerning what is and what is not’, which explores the line of the argument presented more fully here. I thank Christopher Gill heartily for his help in revising my English. I also thank Dominic Scott for the improvement of the style, and Jan van Ophuijsen for his valuable comments. In writing this article, I had an excellent opportunity to exchange my views on the Sophist with Michael Frede, Lesley Brown, Verity Harte, L´aszl´o Bene, Katerina Ierodiakonou, Paul Kalligas, Eyj´olfur Emilsson, Myles Burnyeat, and several Latvian scholars in Kolympari, Crete, 18–26 Sept. 2005. I am grateful to the organizer of the seminar, Arnis Ritups-Redovics (AD Fontes, Riga, Latvia), and the advisor, Myles Burnyeat.

Plato on What is Not

255

treatise called On Nature or On What Is,¹ and Protagoras may also have published a treatise entitled On What Is, about whose content little is known.² These works are probably what Gorgias had in mind when he wrote the treatise with the competing title.³ In the history of ancient philosophy, these two concepts have opposite fortunes, one bright and the other dark. While one forms part of the mainstream of philosophy (giving us the term ‘ontology’), the other is nearly submerged. When the same phrase, On What Is (Περὶ τοῦ ὄντος), reappears as the traditional subtitle for one of Plato’s later dialogues, the Sophist,⁴ the author is clearly aware of both aspects. Plato associates what is and what is not with light and darkness. When Socrates presents the theory of forms in Republic V, he describes opinion as darker (σκοτωδέστερον) than knowledge, which is correlated to what is, but brighter (φανότερον) than ignorance, which corresponds to what is not (478c). The metaphor is appropriate in this particular context, because the Good, the form that makes all beings what they are, will shortly be compared to the sun, and the rest of the intelligible realm to the world illuminated by the sun. The light metaphor appears again in the Sophist. The very brightness (τὸ λαμπρόν) of what is, at which the philosopher aims, makes it hard for us to apprehend it, while the darkness (ἡ σκοτεινότης) of what is not prevents us from discerning the sophist, who takes refuge there (253e–254b). Although the same metaphor appears in both dialogues, there is a crucial difference in the way it is used (as discussed below). Modern philosophers often assume that Plato treats what is not merely as the privation of being and that he dismisses the idea of absolute nothingness from the inquiry altogether, although the latter always remains a real philosophical problem. Pointing to the way in which Plato in the Sophist describes what is not as ‘different from what is’, these philosophers fault him for reducing the problem of absolute nothingness to that of something lacking particular properties. Against this interpretation, which at first sight seems to give an adequate account of the argument of the dialogue, I suggest that Plato tackles a more profound problem. What is not is no more trivial or easy to deal with than its counterpart, what is. It is perhaps a more perplexing concept, since it seems to prevent any discussion (λόγος). This feature takes us to the heart of the problem that Plato faces in the Sophist. There he works out a new strategy to overcome the difficulty: what is not ¹ Cf. DK 30A4, cf. A2; the title may be an invention of later editors, but for this, see Kirk, Raven, and Schofield 1983: 391–2, and Wardy 1996: 15, 153 n. 16. ² Cf. DK 80B2: Eusebius reports that Porphyry happened to read Protagoras’ On What Is, and noticed that the sophist plagiarized monism. Yet this might be the second title of some other work, e.g., the Truth. ³ For this reason I believe that the title was given by the author, Gorgias. ⁴ Cf. Diog. Laert. Vitae Philos. III. 58. The subtitles to the Platonic dialogues, which are usually associated with Thrasyllus, may have originated in the Old Academy. ‘On what is’ is regarded as the skopos of the dialogue by some Platonists, including Olympiodorus (cf. Notomi 1999: 1, 15–16).

256

Noburu Notomi

can only be clarified together with what is. The purpose of my paper is to clarify the implication of this strategy.

2 . WHAT IS NOT B E F O R E T H E SOPHIST Before examining the Sophist, let us briefly survey the occurrences of what is not in Plato’s earlier dialogues. Although he uses the expression what is not in certain arguments, he never analyses it before the Sophist. First, what is not appears in the ontological and epistemological argument of Republic V, as we have just seen. Knowledge deals with what is, and ignorance with what is not; therefore, what both is and is not must be the object of the intermediate faculty: namely, opinion. However, this argument, which clearly echoes Parmenides’ conceptual framework, is not developed any further in the passage. Second, what is not is exploited in some sophistic arguments.⁵ The sophistic brothers, Euthydemus and Dionysodorus, resort to this problematic concept to generate a number of sophisms: e.g. that it is impossible to state a falsehood, since there is no speaking or thinking of what is not (Euthd. 283e–284c). Protagoras puts forward a similar argument at Theaetetus 167a–b, which is restated later in the dialogue (188c–189b; cf. Cra. 429d). Yet despite the fact that what is not leads to serious philosophical difficulties in these dialogues, it is never discussed in its own right. However, this by no means implies that Plato was unaware of the solution to those difficulties before writing the Sophist. On the contrary, as Myles Burnyeat has shown, he intimates the basis of these fallacious arguments and even suggests a right direction for their solution in those passages;⁶ nevertheless, he postpones a full consideration until the Sophist. Likewise, Socrates postpones scrutiny of the ideas of Parmenides, which Theaetetus requests of him in passing, until the Sophist (cf. Tht. 183c–184b). Here, the principal speaker, the Eleatic visitor, takes up the challenge and embarks on the formidable project of examining what is, which involves refuting Parmenides and so might count as a form of ‘parricide’ (241d–242a). The question that the dialogue confronts is whether the sophist can be defined and distinguished from the philosopher in terms of what is not. Unless what is not is properly examined in relation to what is, the sophist will escape the inquiry. For the inquirers allege that he is a producer of images or falsehood; yet, he argues, these cannot exist unless what is not is. Thus, the problem of what is not lies right at the heart of the dialogue. ⁵ For the arguments concerning the impossibility of falsehood, see Notomi 1999: 179–83. ⁶ For this, see Burnyeat 2002.

Plato on What is Not

257

3 . J O I N T I L LU M I N AT I O N The apparent independence of the two subjects, what is not and what is, in the Sophist often misleads interpreters into treating them separately. They normally assume that there is a clear break in the argument after the series of difficulties concerning what is not, images, and falsehood (237b–241b), and tend to see a fresh start in the discussion of what is (242bff.). Yet such an interpretation misses the point of Plato’s overall project in the middle part of the dialogue (236d–264b). By looking at the transition from the difficulties concerning what is not, we can find a closer connection with the subsequent discussion of what is. First, the main task for the inquirers is said to be to prove, contra Parmenides, that what is not somehow is, and that what is in a way is not (241d–242a). Here the focus is on the possibility of combining what is and what is not; the two concepts should not be treated in stark isolation from each other. Second, at 243b–c, a new stage is introduced with a surprising comment: the inquirers should become aware of their own cognitive state concerning several basic terms whose meaning previously seemed obvious. This is the point where the Eleatic visitor introduces what is alongside what is not: Text 1: 243b–c Eleatic Visitor: By the gods, Theaetetus, do you understand anything of what they say each time, when one of them utters and says that many or one or two things are or have become or are becoming, or when another speaks of hot mixed with cold, and posits separations and combinations? When I was younger, I used to think I understood exactly what someone said when he said just what we are in the difficulty about now, namely what is not. Now you see where we are in the difficulty about it? Theaetetus: Yes, I do. EV: Perhaps we are in the same state of the soul equally (ἴσως)⁷ concerning what is. We insist that we understand without difficulty whenever someone utters this word, although we do not understand the other. Perhaps we are in a similar state about both. Tht: Perhaps.

In this passage the Eleatic visitor mentions several instances of expressions that he thought he understood, but actually did not. These include numbers of things—‘one’, ‘two’, and ‘many’—ways of being—‘is’ (ἔστιν), ‘has become’ (γέγονεν), and ‘is becoming’ (γίγνεται)—and their relations—‘separations’ (διακρίσεις) and ‘combinations’ (συγκρίσεις). Since these expressions and their related words appear in the subsequent inquiry into what is, the difficulties surrounding them must have a bearing upon, or even coincide with, the main ⁷ The word ἴσως means ‘equally’ and ‘perhaps’; in this passage, the first may fit the later pairing (in Text 2). Hence, I change slightly the translation given in Notomi 1999: 212. Whereas most scholars translate ἴσως together with τάχα just as ‘perhaps’, White (1993: 33) gives the translation ‘equally’.

258

Noburu Notomi

questions raised in that inquiry: whether there is only one being, or two, or many; how they combine and separate from each other; whether things that become or have become are included among the things that are. We should note that the initial sign of the confusion about these concepts and lack of awareness of our own ignorance already couples in a way what is not and what is. Why on earth does a full investigation into what is become necessary, after presenting the difficulties concerning what is not? The preliminary remarks in Text 1 anticipate the difficulty that the inquirers have to face concerning what is. It is already indicated that what is and what is not are ‘equally’ (ἴσως) difficult. The same point—that equal difficulties are involved in both cases—is stated again, in the middle of the examination of what is (245e–246a). Later the Eleatic visitor concludes his statement of the difficulty concerning what is, in the way hinted at in Text 1: Text 2: 250e–251a EV: Here, let us put this matter down as a complete difficulty. Since what is and what is not have an equal (ἐξ ἴσου) share of difficulty, now our hope is that, if one of these will turn up, whether faintly or clearly, the other will also turn up in this way. And if we cannot see either of them, we shall push the argument through between both of them at once (ἅμα) in the way we can proceed most plausibly. Tht: Fine.

The inquirers have examined a variety of people who have at any time discussed what is: the monists insist that only one is, and the pluralists that two or many are; also, the materialists believe that sensible becomings are, and the friends of the Forms that intelligible beings are. The examination has revealed that none of them understands what is at all, and eventually the inquirers reach their own conclusion, that what is is both motion and rest. Yet this is also rejected on the same basis as the argument against the dualists. This leaves them completely at a loss where to direct their thought concerning what is (249c–250d). This is the moment when the Eleatic visitor recalls the previous difficulty concerning what is not and couples it with the present difficulty concerning what is (250d–e). Dubbed the ‘Joint Illumination’ or ‘Parity Assumption’ by G. E. L. Owen, this strategy is well known to scholars.⁸ But its significance has not yet been fully appreciated. Some may regard the structural parity of the arguments about what is not and what is merely as a literary device. In my view, however, it is the key to understanding what Plato is doing in the Sophist. Owen originally drew attention to the ‘Joint Illumination’ in order to reject the traditional reading of the Sophist offered by, for instance, Cornford and Moravcsik. Their reading makes two assumptions. First, the difficulty concerning what is not arises, in so far as ‘is not’ means ‘does not exist’. Second, the dialogue’s ⁸ Owen 1971: 229–31 (= Owen 1986: 110–11) explains the assumption as ‘that the one cannot be illuminated without the other’.

Plato on What is Not

259

main aim is to distinguish between the complete use (existence) and the incomplete use (copula) of the verb ‘to be’. However, as Owen rightly points out, this reading fails to meet the requirement of parity between the two concepts. For it banishes ‘is not’ in the sense of negation of existence, and retains it in the sense of negation of copula, while still allowing both existential and copulative senses of ‘is’. In respect of the two uses of ‘to be’, what is and what is not do not form a pair. The traditional reading, which ascribes the complete use only to what is, should be rejected in the light of the ‘Joint Illumination’. Lesley Brown develops Owen’s point and rightly concludes that Plato’s occasional use of the complete use of ‘to be’ is not a distinct use, but constitutes an elliptical form of the incomplete use.⁹ The difficulty concerning what is not should not be read as a matter of existence (the complete use), as distinct from the copula (the incomplete use). On the other hand, it may be too hasty to conclude from this syntactical consideration that the concept of absolute nothingness has no place in Plato’s argument. As Charles Kahn has brought out, although ancient Greek had no equivalent for the word ‘to exist’, some uses of the verb ‘to be’ imply what we now call ‘existence’.¹⁰ The whole issue needs to be handled with great caution, since the verb ‘to exist’ as one of complete use appeared only later in the history of Western philosophy. The issue here is not as simple as people often assume: namely, that Plato and the other Greek philosophers ignored what is not in the sense of absolute nothingness, and that it was the Judaeo-Christian tradition, in its discussions of God’s creatio ex nihilo, that addressed the issue for the first time. I shall challenge this assumption, arguing instead that Plato also struggled with this profound problem. If we take the ‘Joint Illumination’ more seriously, we find that the difficulty of what is not does not arise from the nature of negation either, although modern logicians might expect to locate the difficulty there.¹¹ For whereas the concept of what is not may lead to some confusion about negation, the same can hardly be said of the other concept, what is. How to understand negation is, of course, one important issue to be clarified in the Sophist, but it is not the issue here. Rather, the ‘Joint Illumination’ suggests that the crucial issue is the possibility of combination and separation of kinds, which explains negation by means of one kind of relation: namely, ‘difference’. Thus, this line of interpretation is confirmed when we see how the negation in what is not is explained as ‘different’, rather than ‘contrary’, at 257b. For ⁹ Cf. Brown 1986. We should also note that Owen understands Plato’s analysis as ‘the direct parent of Aristotle’s’, when he fully examines Aristotelian ontology (Owen 1965: 71 (= Owen 1986: 260)). ¹⁰ For this, see his classic monograph, Kahn 1973/2003. ¹¹ Pace Owen (1971: 231 (= Owen 1986: 110)), who says that the problems ‘are to be resolved … by giving up a confusion about negation’.

260

Noburu Notomi

this is not a new analysis, but an implication of the whole argument about the combination and separation of kinds. The argument takes the form of modus tollens: Motion and rest are contrary, and they never combine with each other (250a, 255a–b) [this implies that, if two things are contrary, they cannot combine]. But what is and what is not can combine (256d–257a). Therefore, they are not contrary, but only different (257b). Let us now take a closer look at the Eleatic visitor’s comments in Texts 1 and 2, where the ‘Joint Illumination’ is proposed. The following three points are made clear: 1. What is not and what is partake equally (ἴσως, ἐξ ἴσου) in difficulty. 2. The investigation into these should treat them as a pair: if one of the two is made clear, the other becomes clear as well. 3. If we cannot see either, we should advance the argument (λόγος) between the two. The latter two points are not independent; while the third proposes a strategy for the present inquiry, the second makes a suggestion about its outcome, were it to be successful. As for the second point, the expression ‘whether faintly or clearly’ (εἴτε ἀμυδρότερον εἴτε σαφέστερον), together with the main verb ‘turn up’ (ἀναφαίνεσθαι), strongly indicates the image of light (Text 2). Accordingly, the metaphor of ‘Joint Illumination’ suggests that what is not, associated with complete darkness, should be brought to light, together with what is.¹² The last sentence in Text 2 (251a1–3) is notoriously difficult to interpret, but I follow Campbell’s comment and read it as presented above.¹³ Campbell defends the traditional reading διωσόμεθα, attested by all the manuscripts βTW, against other modern editorial suggestions.¹⁴ It seems to me a mistake to reject this subtle reading and to make another conjecture, as David Robinson does in printing διακριβωσόμεθα (‘we shall make our account accurate’) in the new Oxford Classical Text.¹⁵ ¹² Owen (1971: 229–30 (= Owen 1986: 108–9)) properly emphasizes the light metaphor here. This is why I prefer the expression ‘Joint Illumination’ to the perhaps more famous ‘Parity Assumption’. ¹³ Campbell 1867, Soph. 135–6; Owen (1971: 230 (= Owen 1986: 109)) comments that ‘Since Campbell there has been general agreement on the sense of the words’. ¹⁴ Despite the manuscript reading διωσόμεθα, several editorial emendations are proposed: (i) διοισόμεθα, Wagner, Apelt (from διαφέρω: to bear to the end, go through with); (ii) διωξόμεθα, Heindorf (from διώκω: to pursue an object, seek after); (iii) διασωσόμεθα, Stallbaum (from διασῴζω: to preserve through a danger); (iv) διαθησόμεθα, Hermann (from διατίθεμαι: to arrange or settle mutually, set forth). However, all these nineteenth-century emendations lack textual support. For the newest conjecture, see the next note. ¹⁵ Duke et al. 1995: 441. Robinson puts in the apparatus criticus ‘διακριβωσόμεθα vel διαβεβαιωσόμεθα’, probably stimulated by M¨uller’s suggestion διορθωσόμεθα. In any case, I strongly suggest that the supplied letters be indicated with angle brackets, as ‘διωσόμεθα’. However, the

Plato on What is Not

261

The middle voice of διωθέω means ‘force one’s way through’ or ‘break through’; Campbell translates it as ‘we shall fend off our argument from both’, or ‘steer clear of them’. Elsewhere this verb is used with the accusative; for example, Democritus says that ‘one will avert not a few evils in one’s life’.¹⁶ If we keep this verb, the genitive of the neutral dual ἀμφοῖν should be taken with its prefix δια-. The object ‘both’ which we need to avert must be the two difficulties concerning what is not and what is.¹⁷ Campbell also suggests that this sentence signifies the famous image of the dangerous voyage between the two monsters Scylla and Charybdis, or between the two floating rocks called Symplegades.¹⁸ Such images were familiar in antiquity. See, for example, the way in which Euripides describes Medea’s voyage: ‘you broke through the twin rocks of Pontos’.¹⁹ Before the ship moves, a pigeon is released and flies in between the twin rocks, which immediately crash toward each other.²⁰ Notice also that the author of the Seventh Letter (whether Plato or not) uses the image of Scylla and Charybdis to illustrate the dangerous voyage to Sicily.²¹ In these examples the image of ‘twin’ is as perspicuous as in the Sophist.²² The Sophist does not invoke these mythical images directly. But if we take into account other famous passages from the Phaedo, which compare advancing an argument (λόγος) at a critical moment of inquiry to a voyage, our sentence can be interpreted in a similar way.²³ The image of logos as a voyage between two floating rocks or monsters gives us an important clue as to how to advance the inquiry further. If the inquirers, at a loss before the twin difficulties concerning what is not and what is, want to avoid those obstacles and proceed safely, they must move forward between them; otherwise, the ship will crash against the rocks, and the voyage will fail. The way transmitted reading is grammatically defensible, as Robinson (1999: 154) admits. Of the two proposals by Robinson, διακριβωσόμεθα printed in OCT seems worse, when we consider the similar word διακριβολογουμένους at 245e (pace Robinson 1999: 155). ¹⁶ DK 68 B191: οὐκ ὀλίγας κῆρας ἐν τῶι βίωι διώσεαι. ¹⁷ Since the word ‘difficulties’ (ἀπορίαι) is feminine, the neutral ‘both’ may well refer explicitly to what is not and what is, and obliquely to their difficulties. On the other hand, Cornford (1935: 251), reads the simile differently as ‘with both elbows at once’. The image of ‘hand(s)’ certainly appears in 226a and 231c, but is rejected here by many scholars (e.g. Owen 1971: 230 n. 16 (= Owen 1986: 109 n. 15)). ¹⁸ For Scylla and Charybdis, see Hom. Od. XII. 73–126, 222–62. For the crashing rocks, see Hom., Od. XII. 59–72; Eur. Med. 1–2, 1263–4; Apollonius of Rhodes, Argonautica II. 317–40, 549–97; Apollod. Bibl. I. 9. 22: ‘ἦν δὲ ἀδύνατον … δι᾿ αὐτῶν διελθεῖν’. cf. Hdt., IV. 85. ¹⁹ Eur. Med. 432–3. ²⁰ Cf. Apollonius of Rhodes, Argonautica II. 564–5: ‘ταὶ δ’ ἄμυδις πάλιν ἀντίαι ἀλλήλῃσιν / ἄμφω ὁμοῦ ξυνιοῦσαι ἐπέκτυπον’. ²¹ Ep. VII 345d–e, with a citation from Hom. Od. XII. 428. For Scylla the monster, see also Resp. IX. 588c and [Pl.], Ax. 369c. On the other hand, there is no reference in the Platonic corpus to Charybdis except in the Seventh Letter. ²² The words ‘both’ (ἄμφω, ἀμφότερα) appear thirty-two times in the dialogue, as emphasized by Klein (1977: 60–4). ²³ Cf. Pl. Ph. 85c–d (logos as drifting boat) and 99c–d (inquiry in logos as the second voyage).

262

Noburu Notomi

to avoid the present difficulties is to take the two of them as a pair and find a middle way between them. This strategy is repeated in the course of the subsequent argument: Text 3: 254c EV: We should choose some greatest things, first to see what each is and then how they have power to combine with each other, so that if we cannot grasp both what is and what is not with full clarity (σαφηνείᾳ), we should not lack any argument (λόγου) about them, in so far as the present way of inquiry allows.

The ‘Joint Illumination’ suggests a way for the inquirers to escape the situation in which they are now trapped. Since the two difficulties are coupled, they must share a common root, and hence a common solution. Besides, since the difficulties are compared to monsters, the conclusions that have led to them must be wrong in themselves: those conclusions are that what is not is in itself unthinkable and unspeakable (238c–239a), and that what is is both motion and rest (249c–d).

4 . T H E D I F F I C U LT Y C O N C E R N I N G WHAT IS NOT The ‘Joint Illumination’ suggests that the two arguments concerning what is not and what is should not be read in isolation from each other. However, this principle is often ignored in the traditional interpretation. F. M. Cornford and, following him, Denis O’Brien read the argument concerning what is not in the following way:²⁴ 1. The difficulty concerning what is not (237b–239c) presents Plato’s positive view: what is not in a specific sense, namely ‘what in no way is’ (taken as ‘the totally non-existent’) should be banished from all inquiry. 2. Plato here follows and reinforces Parmenides’ position on what is not. 3. This position is the same as that presented in Republic V. If this reading is correct, the ‘Joint Illumination’ will collapse, since we cannot find the same line of argument in the discussion of what is. We must take Plato’s overall strategy into account. Let us examine these three points. (1) The traditional reading distinguishes two stages in discussing what is not. At the first stage, Plato presents the difficulties concerning what is not in the sense of ‘the totally non-existent’, and he approves and reinforces Parmenides’ ²⁴ Cornford 1935: 203–9 and O’Brien 1993, 1995, 2000. O’Brien (2000: 63–7, 70–2) criticized my interpretation of the argument concerning what is not in Notomi 1999: 173–9; I give a full argument against his view in Notomi (forthcoming).

Plato on What is Not

263

prohibition against what is not.²⁵ It is only at the second stage that Plato sets out on the criticism of Parmenides; he finally proves that what is not in another sense is, and rejects Parmenides’ position in part. These two stages are, on the traditional reading, clearly contrasted as Plato’s positive and negative attitudes toward Parmenides. This reading takes the initial difficulty concerning what is not, not as something to overcome, but as the truth, to be confirmed and put aside. However, this reading contradicts our view of the strategy of the ‘Joint Illumination’ (Text 2). The inquirers are in serious difficulties concerning both what is not and what is. The difficulty first raised concerning what is not should be solved and not left untouched. (2) It should be remembered that Cornford is one of the main targets of Owen, when he appeals to the ‘Joint Illumination’.²⁶ Cornford tries to distinguish between the two senses of the verb ‘to be’: what is not is first treated in the sense of ‘the non-existent’, but is later examined in the sense of ‘different from what is’ in the criticism of Parmenides. Thus, the two-stage view is closely tied to the traditional understanding of the two senses of ‘to be’. But taking in the sense of ‘non-existence’ is no longer plausible in the light of Owen’s article. (3) Finally, Cornford strongly associates the passage on what is not with the Republic.²⁷ On his reading, Plato keeps in the Sophist the same ontological and epistemological theory as in the middle dialogues. However, we shall see a definite difference in the treatment of what is not and what is in the two dialogues. Plato does not condone or support Parmenides’ doctrine of total non-existence. The Eleatic visitor demonstrates, through the self-contradiction that is generated, that it is too naive to speak of the unspeakability of what is not, as Parmenides does in his poem. For the very activity of speaking about what is not shows the absurdity of this attempt. This implies that the way in which Plato himself treated what is not, as corresponding to ignorance in Republic V, was also too naive. The Eleatic visitor’s presentation of the difficulties concerning what is not identifies and subverts the very prohibition of Parmenides. Let us observe the way in which the Eleatic visitor starts the argument that leads to the difficulty concerning what is not: Text 4: 237b EV: This is testified by him, and above all logos itself (ὁ λόγος αὐτός) will show it, if properly examined. ²⁵ Cf. Cornford 1935: 203: ‘Plato is not criticising, but confirming this (sc. Parmenides’) doctrine’; O’Brien 2000: 56–9, 68–75, 79 n. 224, 94–7. ²⁶ Cf. Owen 1971: 224 n. 3, 227–8 (= Owen 1986: 105 n. 2, 107–8); as Owen rightly points out, Cornford’s interpretation of what is not contains inconsistency. ²⁷ Cf. Cornford 1935: 202–3, 208.

264

Noburu Notomi

Here it is important to determine what ‘logos itself’ means, which the Eleatic visitor goes on to examine.²⁸ Some interpreters take it to mean the phrase ‘what is not to be’ (τὸ μὴ ὂν εἶναι) in 237a3–4.²⁹ However, the phrase is not called logos in this context. Instead, it will soon turn out that what is examined is not the problematic phrase ‘what is not to be’; rather, it is assumed twice that what is not cannot be combined with what is (237c7–9, 238a7–9; cf. 238d9–e2), in the way Parmenides maintains (DK 28 B7. 1–2, quoted in 237a8–9).³⁰ Considering these points, I propose, with reference to the recurrent use of ‘logos’ and ‘speak’ (λέγειν) in the subsequent argument,³¹ that the Eleatic visitor attempts to examine ‘logos in general’ instead of any particular statement, since this becomes the main issue in the difficulty concerning what is not.³² Plato uses the emphatic word ‘itself’ (αὐτό) to highlight his focus here.

5. ISSUES CONCERNING NAMING After the difficulties concerning what is not and what is have been set out, the subsequent inquiry seeks to resolve both difficulties together. The Eleatic visitor begins this new inquiry by asking ‘in what way the same thing is called by many names’ (251a). A new focus, ‘name’ (ὄνομα),³³ is thus introduced to illuminate the root of the twin difficulties. However, some considerations concerning naming have already been offered in the arguments concerning what is. For example, against the monists who insist that One is, the Eleatic visitor first points out that they are calling the same thing by two names, that is, ‘one’ and ‘is’; besides, he shows that they contradict their own monism by positing two things, that is, the name and its object (244b–d). ²⁸ I examine this crucial passage fully from the philological and contextual points of view, in reply to Denis O’Brien’s criticism, in Notomi (forthcoming). The Greek word λόγος is hard to translate in English, since there is no single word that covers the full range of its connotations in the same way as ‘Sprache’ in German or ‘kotoba’ in Japanese. Therefore, I transliterate this key word hereafter. Its basic meaning is whatever one states (λέγειν): sentence, statement, speech, argument, discussion, discourse, definition, etc. ²⁹ E.g., Campbell 1867, Soph. 82–3, as well as Cornford 1935: 220 n. 3 and O’Brien 2000: 55–75. The phrase is referred to by ‘this’ (τοῦτο) in 237a6; it also corresponds to ‘this’ (τοῦτο) in 237a8 (= Parmenides, DK 28 B7. 1). On the other hand, the similar expression, ‘this logos’ (ὁ λόγος οὗτος) at 237a3, which refers to the statement ‘that falsehood is’ (236e4), cannot be identical with ‘logos itself’ (pace Cornford and O’Brien). ³⁰ In my view, Plato argues against Parmenides, even before the ‘parricide’ passage where the Eleatic visitor declares that he dares to criticize Parmenides’ logos (241d–242a). ³¹ For λόγος, see 237b4, e7, 238b5, 239a6; the verb λέγειν appears twenty-three times, together with many uses of related words, from 237b to 239c. ³² White’s translation, ‘our own way of speaking itself’, seems closest to my reading (White 1993: 25). ³³ The word ὄνομα usually means ‘name’ or ‘word’, in contrast to ‘statement’ (λόγος); the corresponding verbs, ὀνομάζειν and λέγειν, are contrasted in the same way. But ὄνομα in the sense of ‘noun’ is coupled with ῥῆμα in the sense of ‘verb’ for the first time in Soph. 261e–262d.

Plato on What is Not

265

This may seem a minor issue in the whole inquiry, but we should remember that the original problem arises from such basic words as ‘one’ and ‘is’ (cf. Text 1: 243b). The basic question turns out to be what we understand by uttering (φθέγγεσθαι) the words translated ‘what is’ and ‘what is not’. We remember that the Eleatic visitor opened the argument about what is not by imagining someone being asked—not for eristic purposes or for fun, but in all seriousness—what the name (τοὔνομα) ‘what is not’ should be applied to (237b–c). Later, as soon as the difficulty concerning what is is concluded, he reminds his interlocutor (and readers) of this previous difficulty in connection with the present one: Text 5: 250d–e EV: When we were asked to what we should apply the name ‘what is not’, we were in complete difficulty. Do you remember? Tht: Of course. EV: And now aren’t we in a smaller difficulty about what is? Tht: To me, Visitor, we seem to be in a greater one, if it is allowed to say so.

At first glance, this seems to refer only to the first stage of the difficulty concerning what is not (237b–e), but the context clearly indicates that it covers the second (and third) difficulties as well (238a–239b), which result in the unthinkability and unutterability of what is not. Thus, the whole difficulty concerning what is not is now paired with the other difficulty concerning what is, as regards naming. Let us examine the argument about what is not in more detail. Both the initial question and the concluding remark in the first stage focus on the act of uttering (φθέγγεσθαι, 237b, e). Utterance is concerned with naming, in contrast with speaking a logos. Since by uttering some words we speak (λέγειν), the difficulty in uttering ‘what is not’ forces us to the conclusion that we do not even speak (237d–e). The original question concerning naming leads to the denial of any logos. In the second stage, the Eleatic visitor argues that, if no number, whether singular or plural, can be attached to what is not, the impossibility of utterance follows: Text 6: 238b EV: Then, how can anyone utter ‘things which are not’ (τὰ μὴ ὄντα) or ‘what is not’ through his mouth, or grasp them in thought, apart from number? Text 7: 238c EV: Do you understand, then, that it is impossible to utter, speak, or think ‘what is not’ itself by itself (τὸ μὴ ὂν αὐτὸ καθ᾿ αὑτό) correctly? But it is unthinkable (ἀδιανόητον), ineffable (ἄρρητον), unutterable (ἄφθεγκτον), and unspeakable (ἄλογον).

A pragmatic self-contradiction is immediately directed against uttering and speaking of the very statement of Text 7. The impossibility of uttering the name ‘what is not’ now reaches the ultimate difficulty: how can we utter and speak of this very difficulty?

266

Noburu Notomi

The phrase ‘itself by itself’ (αὐτὸ καθ᾿ αὑτό, 238c) indicates the complete separation of the object from everything else.³⁴ If what is not is separated from any other kind and is isolated, it becomes ‘what in no way is’ (τὸ μηδαμῶς ὄν, 237b). However, it is misleading to take this particular phrase to indicate another kind of what is not: namely, ‘absolute non-being’ (i.e. non-existence), as distinct from ‘relative non-being’ (i.e. different from something). Cornford and O’Brien assume that Plato here rejects the former kind, while accepting the latter. But if the Eleatic visitor discusses two different kinds of non-being, it follows that the difficulties raised concerning the one kind (i.e. ‘absolute non-being’) have nothing to do with the subsequent argument concerning the other (i.e. ‘relative non-being’). Since this distinction does not correspond to the argument concerning what is, their reading again violates the parity of what is not and what is. Instead, we should understand the adverb ‘in no way’ (μηδαμὣς) used here as emphasizing the absoluteness of separation from everything else. Returning to the examination of what is, we find it significant that when the Eleatic visitor targets Parmenidean monism, he first criticizes it in terms of naming. Parmenides in his poem admits only ‘Is’, but presents several ‘signs’ (σήματα) to illustrate it: ungenerated, unperished, unchanging, and a single whole like a sphere.³⁵ In what way can he state these, if using names contradicts his monist position? Also, the goddess tells (φράζω) the truth to Parmenides, that ‘you cannot tell (φράσαις) of what is not’,³⁶ so that Parmenides can judge by logos.³⁷ But how on earth is logos possible, which declares the truth of ‘Is’? The apparent separation of the two subjects, namely what is not and what is, in the first half of the middle part (236d–251a), indicates the root of the twin difficulties. If each is treated separately, neither name performs its proper function, since both what is and what is not in isolation prevent the possibility of being uttered. This in turn depends on the possibility of speaking.

6 . I S S U E S C O N C E R N I N G LOGOS The Sophist contains many uses of the words λέγειν (‘speak’) and logos. While they are not always focused on before the final analysis of statement (λόγος) into noun (ὄνομα) and verb (ῥῆμα), this issue remains a main theme throughout the ³⁴ The phrase ‘καθ’ αὑτό, appears later in the separationist’s claim along with ‘apart’ (χωρίς), as we shall see below (252c). Also, a much-discussed passage in 255c–d contrasts ‘themselves by themselves’ (αὐτὰ καθ αὑτά) with ‘in relation to others’ (πρὸς ἄλλα). On the other hand, this phrase is famous as an indication of the theory of forms in the middle dialogues. But we should notice its original meaning: e.g. when death is defined as the separation of the soul from the body, the body comes to be separated itself by itself (αὐτὸ καθ’ αὑτό) apart from the soul, and the soul is separated itself by itself (αὐτὴν καθ αὑτήν) apart from the body (Phd. 64c). ³⁵ DK 28B8. 2–4; note the negative character of the signs. ³⁶ DK 28B2. 6–8. ³⁷ DK 28B7. 5–6.

Plato on What is Not

267

dialogue. The question of how to speak of what is and what is not amounts to the question of how speaking is possible. We find that the concern with logos prevails in the whole inquiry of the dialogue, especially in its middle part. In facing the sophist’s counter-attack, the inquirers have to be content if they can extricate themselves in any degree from the sophist’s formidable logos (241c). To this end, they need to examine Parmenides’ logos, in relation to false logos, and to fight it out in logos (241d–242a); then they have to set out on an adventurous logos (242b). The inquiry into what is starts again by examining how their predecessors speak (λέγειν) of this concept. The ways they speak about what is are like telling a story (μῦθος). The Eleatic visitor casts doubt upon his own way of speaking of it, as well as theirs. He used to believe that he understood well what others meant whenever they uttered ‘what is’, just like ‘what is not’ (Text 1: 243b–c). But the way of speaking of the matter turns out to involve total confusion. The inquirers wish to learn what their predecessors meant when they spoke of what is (244a). In the examination of different positions, the Eleatic visitor constantly asks how they should speak. For example, he asks the tamed materialists whether they say that there are mortal animals, souls, and justice (246e–247c). The issue is how people speak concerning what is. The Eleatic visitor tries to elicit the essence of what is from their logoi. Also, in the imaginary conversation with the friends of Forms, he says, ‘when being is known by knowledge, according to this logos, then in so far as it is known it is changed by having something done to it, which we say would not happen to something that is at rest’ (248e). What is at issue is not a simple fact, but the way to speak about what is. The Eleatic visitor suggests that a philosopher should fight, by using every logos, against anyone who does away with knowledge, understanding, and intelligence; the inquirers eventually state (λέγειν), like a child making a wish, that what is consists of both things unmoving and moving (249c–d). They initially believe that with this logos they grasp what is; but it soon turns out that this logos fails to match what they expected (249d–250d). At this crucial moment, the Eleatic visitor proposes the ‘Joint Illumination’: ‘if we cannot see either of them, we shall push the logos through between both of them at once’ (Text 2: 251a). Now we clearly understand why the inquiry must advance the logos in the middle of the two extremes. For both difficulties result from the common failure in logos. The real issue concerning what is and what is not turns out to be whether we can properly speak by using these names. The difficulty as to where the inquirers should direct their thought concerning what is (250c) is to be answered only when they properly secure its logos. It becomes necessary to reconsider what logos is, in relation to name (ὄνομα). All Plato’s predecessors and contemporaries, including Parmenides, fail to speak properly of what is and what is not. Their failure in logos endangers philosophy itself, and the sophist’s counter-attack exploits this dangerous situation.

268

Noburu Notomi

7 . S AV I N G T H E LOGOS The basis of any logos is to call a single object by many names. However, the ‘late-learners’ deny this possibility by appealing to the one–many problem; they do not admit any logos, such as that man is good, but call man man, and good good. They fail to understand the essence of logos, and only enjoy denying any possibility to logos (251a–c). The late-learners’ manifestly naive argument is introduced so as to illustrate what results from this failure. They treat logos improperly as a sort of naming, which consists in the one-to-one relationship between a word and its referent. They admit no possibility of relating plural words to a single object so as to constitute a logos. The inquirers pursue their own logos by classifying all kinds of discussion of what is into three positions in respect of combination and separation: A Nothing is in any way combined with anything else whatsoever. However, this first choice leads its proponents to self-contradiction: they cannot even state their own thesis. In particular, the late-learners themselves turn out to present their thesis in the most ridiculous way, if they cannot combine the words, ‘apart’, ‘from’, ‘anything else’, ‘in itself’ (καθ᾿ αὑτό), and ‘is’, into a single statement (252b–d). We can observe that this result corresponds to the previous difficulty concerning what is not. This is the failure of those who can be called ‘separationists’. B Next, those who try to combine everything with everything (let us call them ‘combinationists’) will produce such absurd or impossible statements as ‘motion is at rest’ or ‘rest moves’, by combining even opposites (252d). This position muddles everything up, and again makes logos impossible. For it reduces all things to one. C The only possibility left is to admit both combination and separation in a proper way: some kinds combine with some, while others do not combine (252d–253c). This is the position of ‘proper-combinationists’. By repelling the first two, as if they were twin rocks or monsters, the inquiry seeks a middle way between them to save the possibility of logos. This matches the image of ‘Joint Illumination’ (Text 2). How then are these two extremes related to the twin difficulties concerning what is not and what is? The difficulty concerning what is not comes from the strong assumption made by Parmenides that what is not can by no means be combined with what is. This thesis forbids any number to be ascribed to what is not, and concludes that what is not is ‘unthinkable, ineffable, unutterable, and unspeakable’ (Text 7: 238c). However, this conclusion itself commits both theoretical and pragmatic self-contradictions. We must remember that the

Plato on What is Not

269

Parmenidean position causes a similar difficulty in grasping what is: monism cannot state its own thesis, since it makes logos impossible. This consequence is shared by the late-learners. Since Parmenides keeps ‘Is’ away from any other things, especially what is not, he ends with the same result as the separationists. On the other hand, the difficulty concerning what is arises from the inquirers’ wish to admit that what is constitutes both motion and rest (249c–d). This easy combination leads them to the failure in grasping what is. If what is constitutes both motion and rest, such absurd statements may arise as ‘motion is at rest’ or ‘rest moves’ (cf. 250a–b). It turns out that both separationist and combinationist positions end in total disaster. The search for the proper combination of what is and what is not eventually secures the possibility of logos in general. This paves the way for the solution to the twin difficulties. By repudiating both separationist and combinationist positions, the inquirers find the right way in the middle—that of the propercombinationist. This inquiry requires the combination of the greatest kinds, a task that is ascribed to the art of dialectic. In this argument, one of the pair, namely what is, turns out to be the principle of combination, and the other, namely what is not, turns out to be the principle of separation (253b–e).³⁸ These two together save logos. The Eleatic visitor later appreciates that the inquiry which admits the proper combinations of kinds saves the possibility of logos and of the greatest thing, namely philosophy itself (260a). Text 8: 259d–e EV: Indeed, my good friend, it is improper to try to separate everything from everything else. And that is a mark of someone completely uncultured and unphilosophical. Tht: Why? EV: To dissociate each thing from everything else is the total destruction of all logoi. For logos becomes possible for us in the interweaving of the kinds.

The two problematic concepts, what is not and what is, now turn out, taken together, to be the basis of the possibility of logos and of philosophical inquiry in general. What is and what is not are grasped not separately by themselves, as Parmenides and others attempt, but jointly in the field of logos. Logos comprises both together; or rather, what is and what is not, as the principles of combination and separation, interweave the kinds into logos. Here we should not take the inquiry into what is not and what is as being pursued on a firmly established basis, but regard the philosophical inquiry through logos itself as the main target to be secured. This is the essential message of the twin difficulties of what is not and what is. ³⁸ For this interpretation, see Notomi 1999: 234–7. What is not is later to be analysed into ‘different’ and ‘what is’ in 257c–258b.

270

Noburu Notomi

8 . T H E S O PH I S T A N D WHAT IS NOT It is in pursuit of the sophist that the inquirers confront the difficulties concerning what is not. For when they attempt to define him as a maker of false images or a speaker of false logos, the sophist counter-attacks and tries to destroy the basis of their attempt. He resorts to the Parmenidean thesis that what is not cannot be combined with what is, and thereby denies the possibility of falsehood and images altogether. What position, then, does the sophist take in respect of the combination and separation of kinds? Based on the total separation of what is not from what is, the sophist, on Parmenides’ authority, denies the existence of false logos, images, and sophistry. On the other hand, the late-learners separate each word from everything else by reference to the one–many problem. Therefore, it is natural to suppose that the sophist is, in one respect, associated with the separationists, including the late-learners. The Eleatic visitor concludes that the late-learners and the sophist converge as those who manipulate argument; they are compared to a ‘new-born child’ (νεογενής, 259b–d).³⁹ The other characteristic of the sophist is to mix everything up. By denying the possibility of falsehood and of contradiction, he tries to abolish the distinction between the sophist (i.e. himself) and the philosopher (i.e. inquirers). This position of confusing everything amounts to that of the combinationist. The sophist thus makes full use of both extreme positions, the separationist and the combinationist, in each argument, and tries to entrap his opponents in difficulties. He nullifies all possibilities of logos and philosophy. The inquirers must discern a proper, intermediate way, and advance the logos between these extremes. Their primary task is to save the possibility of logos. In the short digression in the middle of the dialogue, the knowledge which properly discerns combination and separation is ascribed to a free person, namely the philosopher, and that knowledge is called dialectic (253b–e). The inquirers achieve this kind of knowledge in the subsequent discussion of the combination of the greatest kinds. In this way, the proper-combinationist position is correlated with the philosopher, and realized through the course of the actual inquiry. The main task of the dialogue—namely, how to distinguish the sophist from the philosopher—turns out to depend on this possibility of dialectic. The difficulty in defining the sophist is coupled with the difficulty in grasping the philosopher. On the other hand, to secure the proper combination and separation, by discerning the proper roles of what is and what is not, enables the inquirers to overcome the twin difficulties. ³⁹ Homer describes Scylla as having a voice ‘of a new-born (νεογιλῆς) whelp’ (Od. XII. 86–7); if Plato has Scylla and Charybdis in mind as the twin monsters, this might be a hint that the separationist corresponds to Scylla.

Plato on What is Not

271

The possibility of logos jointly illuminates the sophist and the philosopher.⁴⁰ Text 9: 253e–254b EV: But you won’t assign dialectic, I suppose, to anybody other than the one who purely and justly does philosophy. Tht: How can one assign it to anyone else? EV: We will find out the philosopher now or later in a place like this if we search for him, although it is difficult to see him clearly; but the difficulty in the case of the sophist is different from the difficulty in this case. Tht: How? EV: The sophist escapes into the darkness of what is not, feeling his way in it by knack, and it is because of the darkness of the place that he is hard to see. Tht: It seems so. EV: The philosopher, on the other hand, always clings, through reasoning, to the kind of what is, and this time again he is not easy to see because of the brightness of the space. For the eyes of many people’s souls are unable to endure to look at what is divine. Tht: That is as reasonable as the last one.

The sophist takes refuge in the darkness of what is not, but is to be dragged out into the sunlight, when he is properly defined. Later the Eleatic visitor remembers that they have suffered dizziness (σκοτοδινία) in search of the sophist (264c). The image of darkness is associated with the sophist and his hiding place: namely, what is not. However, the treatment of what is not and what is in this digression (Text 9) illustrates a somewhat different aspect of the pair; for here these are treated not as the two kinds to be combined, as we saw in the main inquiry, but as things which occupy opposite places, bright and dark, without any communication (see fig. 13.1).

What Is

(brightness) --- philosopher

Joint Illumination what is   what is not

What Is Not

(darkness) --- sophist

Figure 13.1. Relation between what is and what is not. ⁴⁰ My main contention in Notomi 1999 is that no independent dialogue, the Philosopher (?), is intended in defining the other member of the pair (cf. 23–5, 238–40, 296–301).

272

Noburu Notomi

The proper combination of what is and what is not (the horizontal line in the middle) is the main issue pursued in the middle part of the dialogue. On the other hand, What Is and What Is Not (tentatively with initial capitals, above and below the line) indicate two opposite directions, towards which the philosopher and the sophist aim respectively, although each in total isolation is beyond logos: What Is Not is unspeakable, so is What Is (though in truth we cannot state these). However, What Is jointly illuminates what is and what is not (the principles of combination and separation), and through their proper combination, it makes logos possible, whereas What Is Not makes it impossible. This is how the two difficult concepts, what is not and what is, are related, and how, on this basis, the sophist and the philosopher are distinguished.⁴¹

9 . P L ATO A N D D A M A S C I U S O N WHAT IS NOT Finally, let us see how Plato concludes the difficulty concerning what is not. Just after the Eleatic visitor settles the nature of what is not (258d–e), he announces: Text 10: 258e–259a EV: Then, do not let someone say that we dare to say what is not is, by declaring that what is not is something contrary to what is. For we have already said good-bye to something contrary to what is, whether it is or not, and whether it has logos or is completely unspeakable (ἄλογον).

Cornford regards this ‘having already said good-bye’ as a reference back to the argument on what in no way is (τὸ μηδαμῶς ὄν) in 238b ff.⁴² Although that early argument is certainly included in this reference, it is not until the combination of what is not with what is is demonstrated that what is not as something contrary to what is is clearly rejected (cf. 257b). Thus, admitting the proper combination and separation between kinds, especially between the pair of what is and what is not, generates the final clear announcement of Text 10. ⁴¹ This scheme of what is and what is not in the Sophist can be contrasted with the ontologicalepistemological scheme of Republic V: What really is: Knowledge What is and is not: Opinion What in no way is: Ignorance At the bottom, ‘what in no way is’ as the object of the cognition, and ‘ignorance’, as its cognitive state, are virtually empty, since there is nothing corresponding to these, while at the top, ‘what really is’ signifies the forms, which are the object of knowledge and dialectical logos. This scheme may be closer to Parmenides’ framework, which the Sophist criticizes. ⁴² Cornford 1935: 295 n. 1: he understands it as ‘the simply non-existent’.

Plato on What is Not

273

However, we should notice that the argument (logos) concerning what is not may not be conclusive, but may leave something open. The last concessive clause of Text 10, ‘whether it is or not, and whether it has logos or is completely unspeakable’, may be an indication of this inconclusiveness of the issue. Following Text 10, the Eleatic visitor suggests that an opponent should refute their argument (λόγος), or otherwise accept it (259a). The inquirers have argued for the possibility of logos itself; therefore, unless anyone fights against this logos without using any logos, he must accept the inquirers’ logos. But this still leaves a possibility that someone can stick to the complete impossibility of saying what is not.⁴³ In this sense, the issue is not conclusive, whereas the Eleatic visitor has given a conclusive logos. What is not, taken as what in no way is, cannot be spoken or thought, but unspeakability already escapes this very statement. What is not is always beyond our speech and thought. This is the issue which the sophist exploits, but only with a self-contradiction. The passage on the unspeakability of what is not is mentioned a few times by the later Platonist Damascius in the beginning of the treatise called Problems and Solutions on the First Principles.⁴⁴ What Damascius takes from Plato’s argument is the idea that what is not always remains beyond our grasp through logos. For when we state that it is beyond our grasp, it already escapes us. He calls this self-contradictory feature ‘reversal’ (περιτροπή), apparently with Plato’s description of the sophistic method in mind.⁴⁵ Here Damascius is aware that the ‘unspeakable’ appears in two opposite directions: the one is above all the universe, and the other below it. He knowingly uses Plato’s argument concerning the latter (which he also calls ‘matter’) in order to illuminate the former, which is the primary object of his inquiry. For both have the same intractable nature: namely, unspeakability beyond our grasp. While Damascius directs his argument about the unspeakable at the first principle beyond all things, his reading of the Sophist may elucidate the ‘darkness’ which Plato suggests in connection with the slippery sophist. In this interpretation, Plato deliberately leaves open, as Text 10 may indicate, the question about the being of what in no way is, or speaking of its unspeakability. For he is aware that he cannot speak that. ⁴³ An ultimate sophist? How can we notice him? But this person can no longer be a sophist, since the sophist is above all a manipulator of logoi. ⁴⁴ Cf. I. 8. 3–4, 9. 20–2, 15. 18–21, 16. 1–6, 23. 1–4, 23. 22–24.2 (ed. WesterinkComb`es). I have learned a great deal about this interesting work from my colleague Satoshi Hori´e. ⁴⁵ Cf. e.g. I. 9. 3, 9. 21, 16. 5, 18. 9, 22. 19, 23. 3, 23. 25, 26. 3; although this word does not appear in the Sophist, it is proposed that Damascius implies the passage around 239d; cf. Dillon 1996: 129 n. 22.

274

Noburu Notomi

REFERENCES Brown, L. (1986), ’Being in the Sophist: A Syntactical Enquiry’, Oxford Studies in Ancient Philosophy, 4: 49–70. Burnet, J. (1905), Platonis Opera, recognovit brevique adnotatione critica instruxit, i, Oxford Classical Texts, 2nd edn. Oxford. Burnyeat, M. F. (2002), ’Plato on How Not to Speak of What is Not: Euthydemus 283a–288a’, in M. Canto-Sperber and P. Pellegrin (eds.), Le Style de la pens´ee: Recueil de textes en hommage a` Jacques Brunschwig (Paris), 40–66. Campbell, L. (1867), The Sophistes and Politicus of Plato, with a revised text and English notes. Oxford. Cornford, F. M. (1935), Plato’s Theory of Knowledge: The Theaetetus and the Sophist of Plato translated with a running commentary. London. Dillon, J. (1996), ’Damascius on the Ineffable’, Archiv f¨ur Geschichte der Philosophie, 78: 120–9. Duke, E. A., Hicken, W. F., Nicoll, W. S. M., Robinson, D. B., and Strachan, J. C. G. (1995), Platonis Opera, Tomus I, Tetralogias I–II continens, Oxford Classical Texts. Oxford. Kahn, C. H. (1973/2003), The Verb ’Be’ and its Synonyms, Foundations of Language Supplementary Series, 16. Dordrecht, 1973; The Verb ’Be’ in Ancient Greek, rev. edn. Indianapolis, 2003. Kirk, G. S., Raven, J. E., and Schofield, M. (1983), The Presocratic Philosophers, 2nd edn. Cambridge. Klein, J. (1977), Plato’s Trilogy, Theaetetus, The Sophist, and The Statesman. Chicago. Moravcsik, J. M. E. (1962), ’Being and Meaning in the Sophist’, Acta Philosophica Fennica, 14: 23–78. Notomi, N. (1999), The Unity of Plato’s Sophist: Between the Sophist and the Philosopher. Cambridge. Japanese version, revised with two new appendices, Nagoya, 2002. (forthcoming), ’Plato against Parmenides: Sophist 236D–242B’, in S. Stern-Gillet and K. Corrigan (eds.), Reading Ancient Texts: Essays in Honour of Denis O’Brien (Leiden). O’Brien, D. (1993), ’Non-being in Parmenides, Plato and Plotinus: A Prospectus for the Study of Ancient Greek Philosophy’, in R. W. Sharples (ed.), Modern Thinkers & Ancient Thinkers (London), 1–26. ˆ Deuz Etudes ´ (1995), Le Non-Etre, sur le Sophiste de Platon. Sankt Augustin. (2000), ’Parmenides and Plato on What is Not’, in M. Kardaun and J. Spruyt (eds.), The Winged Chariot, Collected Essays on Plato and Platonism in Honour of L. M. de Rijk (Leiden), 19–104. Owen, G. E. L. (1965), ’Aristotle on the Snares of Ontology’, in R. Bambrough (ed.), New Essays on Plato and Aristotle (London), 69–95; repr. in Owen (1986), 259–78. (1971), ’Plato on Not-being’, in G. Vlastos (ed.), Plato: A Collection of Critical Essays, i: Metaphysics and Epistemology (Notre Dame, ind.), 223–67; repr. in Owen (1986), 104–37.

Plato on What is Not

275

(1986), Logic, Science, and Dialectic: Collected Papers in Greek Philosophy, ed. M. Nussbaum. Ithaca, NY. Robinson, D. B. (1999), ’Textual Notes on Plato’s Sophist’, Classical Quarterly, 49: 139–60. Wardy, R. (1996), The Birth of Rhetoric, Gorgias, Plato and their Successors. London. Westerink, L. G., Comb`es, J. (1986), Damascius, Trait´e des Premiers Principes, 1, Paris. White, N. P. (1993), Plato, Sophist, translated, with introduction and notes. Indianapolis.

14 The Soul as an Inner Principle of Change: The Basis of Aristotle’s Psychological Naturalism Thomas Johansen

One of the central lessons of Myles Burnyeat’s ‘De Anima II 5’ is that we need to read Aristotle’s account of sense perception within the framework of his physics.¹ We can only understand what perception is for Aristotle by first understanding what alteration is in general—which the physical works teach us—and then refining that notion of alteration in view of the peculiar features of perception.² If by psychological ‘naturalism’ we understand the view that the soul is to be explained within the framework of one’s natural philosophy, whatever this may be, then Aristotle’s account of perception in De anima II. 5 is clearly a naturalistic one, on Burnyeat’s interpretation. Much in the same spirit, I want in this paper to consider Aristotle’s claim that the soul is an inner principle of change against the background of his account in the Physics of nature as an inner principle of change. That for Aristotle the study of the soul is part of the study of nature is clear already from the opening lines of the De anima, where he announces that ‘knowledge of the soul seems to contribute greatly to all truth but most of all the truth in relation to nature: for the soul is such a thing as a principle of animals’ (402a 4–7).³ The study of the soul is valued, then, for its contribution to our knowledge of nature. The A relative of this paper was presented to the conference on Teleology, Ancient and Modern at Toronto in 2005. I am most grateful to the members of the audience there for their constructive comments. ¹ Burnyeat 2002. ² See, e.g., Burnyeat’s explanation of Aristotle’s insistence on taking perception as an alteration, despite its peculiar features. ‘We can now see what Aristotle would lose by giving up the language of alteration. He would cut the links with the dialectic of De Generatione et Corruptione I 7 and the categorical analysis of change in Physics III 1–3. He would be set adrift, not merely from the reputable opinions he began with, but from the entire project of comprehending perception within the framework of the physics he develops in the De Generatione et Corruptione and Physics by analysis, systematisation and refinement of reputable opinions from the earlier tradition. He would have to tear up the De Anima and start again’ (Burnyeat 2002: 58). ³ δοκεῖ δὲ καὶ πρὸς ἀλήθειαν ἅπασαν ἡ γνῶσις αὐτῆς [sc. τῆς ψυχῆς] μεγάλα συμβάλλεσθαι, μάλιστα δὲ πρὸς τὴν φύσιν· ἔστι γὰρ οἳον ἀρχὴ τῶν ζῴων.

The Soul as an Inner Principle of Change

277

reason for that claim is in turn that the soul is such a thing as ‘a principle of animals’. We may think that ‘such as (hoion)’ serves to warn us that the claim is not literally true, that the soul is only like a principle (arkh¯e) of animals, but not really so. But, as we shall see, Aristotle thinks that the soul really is a principle of animals. The ‘such as’ here serves rather to put epistemic distance to the claim: we do not yet know what it means for the soul to be a principle, so the vagueness is appropriate. This paper seeks to overcome some of that vagueness. What is it for the soul to be a principle of living beings? In the process of answering this question we shall learn more not only about the soul but also, as Aristotle promised, about nature.

L I V I N G B E I N G S A S S E L F - M OV E R S : T H E PH Y S I C A L WO R K S For Aristotle living beings are natural beings and they are to be explained as such. So in De anima I 1 he recommends that we approach the soul physik¯os, in the manner of the natural philosopher. Since natural beings are composites of form and matter, studying living beings as natural beings means studying them as composites of form and matter. Aristotle’s ‘most common account’ of the soul is accordingly that of the natural philosopher: the soul is ‘substance in the sense of the form of a natural body potentially having life’ (De an. II 1 412a 19–21). Soul, then, stands to body as form to matter. Because living beings are composites of form and matter, they are subject to change. However, as we learn in Physics II, natural substances are not simply subject to change. They also determine some of their own changes, for they have within themselves principles (arkhai) of change and rest. It is this feature that sets them apart from other composite substances such as artefacts. Aristotle reminds us of the point in De an. II 1. He has himself used the analogy of an axe to illuminate the relationship between form and matter in living beings. But he then adds that the soul is the formula (logos) not of such a body, i.e. the body of an artefact, but rather of a natural body, which, he says, is a body ‘having a principle (arkh¯e) of change and rest in itself’ (412b 15–17). Soon afterwards, it emerges that he understands such principles as capacities (dunameis) for change. So he says of plants that they all seem to live, ‘for they clearly have within themselves (en hautois) a capacity and principle (dunamin kai arkh¯en) of such a kind, through which they take growth and diminution in opposite places’ (413a 25–7).⁴ The inner principle by which plants change themselves is thus a nutritive capacity, whereas for other living beings the ⁴ φαίνεται γὰρ ἐν αὑτοῖς ἔχοντα δύναμιν καὶ ἀρχὴν τοιαύτην, δι᾿ ἧς αὔξησίν τε καὶ φθίσιν λαμβάνουσι κατὰ τοὺς ἐναντίους τόπους.

278

Thomas Johansen

inner principles include also other capacities for perception, locomotion, or thinking.⁵ We have here an application of the Physics’ notion of nature as an inner principle of change: for living beings the soul is an inner principle of change. What makes living beings natural is that they have such an inner principle of change. What makes them living is that for them their inner principle is a soul or part of a soul: to be a living being is to have an inner principle of change and rest, where that inner principle does at least one of the things that only souls do. It is worth stressing that where Aristotle’s naturalism about the soul properly begins is not with the claim that living beings are composites of form and matter, or with the claim that living beings are subject to change, but with the claim that they have within themselves an inner principle of change and rest, and that the soul serves as such a principle.⁶ Natural beings have an inner principle of change; artefacts do not. This difference can be explained further in terms of the ‘four causes’. In Physics II 3 Aristotle says that there are four different kinds of cause: form, matter, ‘that towards which’ (final cause), and ‘that from which’ (efficient cause). To be a natural being is to have a principle of change within oneself not only as the material cause is a principle of change but also as the formal, final, and efficient causes are principles of change. All material beings have qua material the potential for change in Aristotle’s universe. So a slab of marble has the potential for change, the potential to be turned into, say, a statue or a bench. For this potential to be realized, however, the marble requires an efficient cause—an artist, say, who brings to bear the form of statue on the materials. For artefacts the efficient cause is external. That is why no slab of marble left to its own devices will ever, except by accident, become a statue. Natural beings are different: they are able to initiate changes to themselves because, it seems, they have the efficient cause within themselves. For it is the efficient cause that sets things moving. It is responsible for actualizing the potential of any given matter to have such and such a form. Aristotle thus refers to the efficient cause not just as ‘that from which the change originates’ but also as ‘that which moves’ and ‘that which produces the actuality or activity (energeia)’.⁷ When we are asked why this slab of marble was turned into a statue, whereas that slab wasn’t, given that both had the potential, we appropriately point to the efficient cause, the sculptor who initiated a change in ⁵ οὐδεμία γὰρ αὐτοῖς δύναμις ἄλλη ψυχῆς. τὸ μὲν οὖν ζῆν διὰ τὴν ἀρχὴν ταύτην ὑπάρχει τοῖς ζῶσι, τὸ δὲ ζῷον διὰ τὴν αἴσθησιν πρώτως (413a 33–b2). ⁶ This point is brought out clearly in the following passage from the Part. an. I. 1: ‘However, it is not the case that all soul is an origin [principle, arkh¯e] of change, nor all its parts; rather, of growth the origin is the part which is present even in plants, of alteration the perceptive part, and of locomotion some other part, and not the rational; for locomotion is present in other animals too, but thought in none. So it is clear that one should not speak of all soul; for not all of the soul is a nature, but some part of it, one part or even more’ (641b 4–10, trans. Lennox 2001). ⁷ Cf. e.g. Ph. 194b 29–32; An. post. 94b 7–8; Gen. an. 778b 1; De an. 417b 20.

The Soul as an Inner Principle of Change

279

the one slab but not in the other. The difference when it comes to explaining how natural beings are capable of initiating their own changes but artefacts aren’t seems then in the first place to be the fact that natural beings have the efficient cause within themselves. Note, however, that the efficient cause is capable of actualizing the potentiality of the matter to have a certain form because it already itself has that form.⁸ In this way, having the efficient cause within oneself also points to having the formal cause within oneself. However, we also know from Physics VIII 4–6 that the claim that natural beings initiate their own changes has to be qualified in view of their dependence on the environment. In these chapters, Aristotle argues that everything that is changed is changed by something. But if natural beings are able to initiate their own changes, do they change without being changed by something? Aristotle, in response, clarifies how natural beings are changed. He distinguishes between two kinds of natural being: living beings such as animals and other natural beings such as fire and earth. The motions of both kinds of natural being are, indeed, caused by something, but in the case of living beings the moving cause is internal, whereas in the simple bodies it is external. In animals ‘that which causes motion is separate from that which suffers motion, and in this way the animal as a whole causes its own motion’ (254b 31–3). The simplicity of the simple bodies, in contrast, does not allow for a distinction within them between a moving and a moved part, and so they cannot initiate their own motions. But, we are still free to think of the motions of the simple bodies by an external cause as natural to them, if it brings to ‘actuality the proper activities that the bodies naturally possess’ (255a 28–30). For instance, it may require a push for a stone to move down, but the motion still counts as natural because stones as heavy naturally tend to move downwards. At the end of Physics VIII 4 we may conclude that even if the simple bodies are not really self-movers, at least living beings are. However, in Physics VIII 6 Aristotle seems to go back also on this. For here he maintains not just that everything that moves is moved by something, but also that it is moved by something other than it. Or more precisely, he claims that the motion of all perishables is caused by something else. His motive for making this claim is that he wants to show that the continuous and eternal motion amongst the perishables requires the existence of something imperishable which always and continuously causes motion without itself being moved.⁹ But Aristotle now (259b 3–6) faces the objection that, by his own account, living beings are capable of moving themselves, without, it seems, being moved by others. So perhaps ⁸ Cf. Ph. VIII 5 257b 6–10: ἔτι διώρισται ὅτι κινεῖται τὸ κινητόν· τοῦτο δ᾿ ἐστίν δυνάμει κινούμενον, οὐκ ἐντελεχείᾳ, τὸ δὲ δυνάμει εἰς ἐντελέχειαν βαδίζει, ἔστιν δ᾿ ἡ κίνησις ἐντελέχεια κινητοῦ ἀτελής. τὸ δὲ κινοῦν ἤδη ἐνεργείᾳ ἔστιν οἷον θερμαίνει τὸ θερμὸν καὶ ὅλως γεννᾷ τὸ ἔχον τὸ εἶδος. ⁹ Gen. corr. I 3 318a 2–8 makes it explicit that Aristotle takes this argument specifically to concern the efficient causation of motion.

280

Thomas Johansen

we do not, after all, need a first unmoved mover to guarantee the continued motion of perishables. In reply, Aristotle denies that animals strictly speaking move themselves:¹⁰ We must understand that animals change themselves with respect to only one change [sc. locomotion], and even this one not strictly. For the cause is not from (ex) the animal itself. Rather, there are other changes natural to animals according to which they are not changed through themselves (dia haut¯on), such as growth, diminution, and respiration, according to which each of the animals is changed when it is at rest and not being changed according to a change by itself. The cause of this is the environment and many of the things entering from without. Thus in some cases the cause is nourishment—when it is being digested animals sleep, and when it is being distributed they awake and move themselves, the first principle being external. Therefore animals are not always being changed continuously by themselves. (Ph. VIII 6 259b 6–15)¹¹

Aristotle appears to be saying here that locomotion is not strictly speaking a case of self-motion since it is triggered by other changes in which the animal is passive, i.e. in which the animal does not initiate the change. So when food enters from without, it sets off digestion, which puts the animal to sleep, a state that continues until the food has been digested. The efficient cause, ‘the first principle of motion’, of the nutritive changes is thus external, and since the nutritive changes make us stop and start moving, the efficient cause of locomotion in this case also counts as external. Read in this way, however, this passage clashes with another in On Generation and Corruption I 5 where Aristotle is explicit that the efficient cause of nutrition (including growth) is not the food but the nourished living being: But what does the altering and the principle of the motion is in what grows and what is being altered, for what does the changing is in these things … but this [sc. the food] having undergone the change has passed away and what does the changing is not in it. (321b 6–10)¹² ¹⁰ I fail to see why Graham (1999: 112–15) thinks the argument could be clarified through the distinction between voluntary and involuntary action. The contrast that is relevant for the overall argument is clearly that between changes initiated by oneself and those initiated by external factors, or, put differently, between those changes where the actualizing or efficient cause lies within oneself and those where it lies outside oneself. Voluntary actions would form only a small subset of those changes that one initiates oneself, and would properly apply only to humans, whereas Aristotle begins the passage quoted by referring to the motions of animals. In fairness to Graham, it should be said that he offers this reading as only one option. ¹¹ τοῦτο δὴ δεῖ λαβεῖν, ὅτι μίαν κίνησιν αὑτὰ κινεῖ, καὶ ὅτι ταύτην οὐ κυρίως· οὐ γὰρ ἐξ αὐτοῦ τὸ αἴτιον, ἀλλ᾿ ἔνεισιν ἄλλαι κινήσεις φυσικαὶ τοῖς ζῷοις, ἃς οὐ κινοῦνται δι᾿ αὑτῶν, οἷον αὔξησις φθίσις ἀναπνοή, ἃς κινεῖται τῶν ζῴων ἕκαστον ἠρεμοῦν καὶ οὐ κινούμενον τὴν ὑφ᾿ αὑτοῦ κίνησιν. τούτου δ᾿ αἴτιον τὸ περιέχον καὶ πολλὰ τῶν εἰσιόντων, οἷον ἐνίων ἡ τροφή· πεττομένης μὲν γὰρ καθεύδουσιν, διακρινομένης δ᾿ ἐγείρονται καὶ κινοῦσιν ἑαυτούς, τῆς πρώτης ἀρχῆς ἔξωθεν οὔσης, διὸ οὐκ ἀεὶ κινοῦνται συνεχῶς ὑφ᾿ αὑτῶν. The passage as a whole responds to Ph. VIII 2 253a 8–21. On the relationship between the two passages, cf. Furley 1994: 6–7 and Morison 2004. ¹² ἀλλὰ τὸ ἀλλοιοῦν καὶ ἡ ἀρχὴ τῆς κινήσεως ἐν τῷ αὐξανομένῳ καὶ τῷ ἀλλοιουμένῳ· ἐν τούτοις γὰρ τὸ κινοῦν, … ἀλλ᾿ ἔφθαρταί γε τοῦτο παθόν, καὶ τὸ κινοῦν οὐκ ἐν τούτῳ.

The Soul as an Inner Principle of Change

281

Aristotle could hardly be clearer that he takes the efficient cause in nutrition to be found in the living being and not in the food. It seems, then, that if Aristotle in Physics VIII 6 means to use nutrition as an example of changes of which animals are not the efficient cause, then he has, by his own lights, chosen a singularly bad example. I suggest therefore that we return to the Physics passage to see if another reading is possible. Let us begin by pointing to some terminological distinctions that Aristotle himself carefully set up in Physics VIII 6. Given that all things that move are moved by something, he distinguished between (a) what is not moved because of itself (di’ hauto) and (b) what is moved because of itself. Group (b), in turn, allows for a distinction between (i) what is moved because of itself immediately and (ii) through several things (dia pleion¯on), like the stone which is moved by (kineitai hypo) the stick which is moved by the man.¹³ Aristotle also expresses the same notion of instrumentality by means of the dative: the man moves the stone with the stick (t¯ei bakt¯eriai).¹⁴ With these distinctions in hand let us return to our passage in chapter 6. It now appears significant that nutrition is here used to illustrate changes that are ‘not through the animals themselves’ (dia with the genitive). For, as we now know, this does not mean that nutrition does not happen ‘because of the animals themselves’ (dia with the accusative).¹⁵ It means that nutrition happens by means of an instrument. So Aristotle is not denying that nutrition is a case of self-motion, only that it is a case of self-motion by means of the animal itself. For nutrition happens by means of something external, food. We can compare the food with the stick in the example: the nutrition happens through or with the food, much as hitting the stone happens through or with the stick. However, the difference is that the instrument of nutrition acts on the animal itself rather than on some external object, like the stone. To understand better how the animal moves itself by means of the food, we need to bring in Aristotle’s distinction between the soul and the body, as does Aristotle himself in the following lines: For the mover is distinct (allo), and it is itself moved and causes change by interacting with each self-mover. In all these cases the first mover and cause of the thing moving itself is moved by itself, but incidentally. For the body changes place, so that what is in the body also changes place, moving itself through leverage. (Ph. 259b 15–20, trans. D. Graham)¹⁶ ¹³ Ph. 256a 4–10: τοῦτο [i.e. that all things that move are moved by something] δὲ διχῶς· ἢ γὰρ οὐ δι᾿ αὐτὸ τὸ κινοῦν, ἀλλὰ δι᾿ ἕτερον ὃ κινεῖ τὸ κινοῦν, ἢ δι᾿ αὐτό, καὶ τοῦτο ἢ πρῶτον μετὰ τὸ ἔσχατον ἢ διὰ πλειόνων, οἷον ἡ βακτηρία κινεῖ τὸν λίθον καὶ κινεῖται ὑπὸ τῆς χειρὸς κινουμένης ὑπὸ τοῦ ἀνθρώπου, οὗτος δ᾿ οὐκέτι τῷ ὑπ’ ἄλλου κινεῖσθαι. ἄμφω δὴ κινεῖν φαμέν, καὶ τὸ τελευταῖον καὶ τὸ πρῶτον τῶν κινούντων, ἀλλὰ μᾶλλον τὸ πρῶτον· ¹⁴ Ph. 256a 22–5: πᾶν γὰρ τὸ κινοῦν τί τε κινεῖ καὶ τινί. ἢ γὰρ αὑτῷ κινεῖ τὸ κινοῦν ἢ ἄλλῳ, οἷον ἄνθρωπος ἢ αὐτὸς ἢ τῇ βακτηρίᾳ, καὶ ὁ ἄνεμος κατέβαλεν ἢ αὐτὸς ἢ ὁ λίθος ὃν ἔωσεν. ¹⁵ Contrast Furley’s translation: ‘but there are other natural motions in animals, which they do not have because of themselves’ (Furley 1994: 6). ¹⁶ ἄλλο γὰρ τὸ κινοῦν, αὐτὸ κινούμενον καὶ μεταβάλλον πρὸς ἕκαστον τῶν κινούντων ἑαυτά. ἐν πᾶσι δὲ τούτοις κινεῖται τὸ κινοῦν πρῶτον καὶ τὸ αἴτιον τοῦ αὐτὸ ἑαυτὸ κινεῖν ὑφ᾿

282

Thomas Johansen

Aristotle here continues to call the animal a ‘self-mover’ and the soul the ‘first mover’, notwithstanding the reference in the previous sentence to the food as ‘the first principle’. ‘The mover’ referred to here as ‘distinct’, i.e. distinct from the self-mover, I take to be the food, which not only changes but is also changed in nutrition. The soul, meanwhile, is not moved, except accidentally, in so far as it is in the body, which is moved.¹⁷ Animal motion counts as self-motion even when the animal is affected by external factors such as food, first, because the living being also in these motions moves the food which moves the body,¹⁸ and, secondly, because it is only the body that is properly speaking moved by the food, while the soul is only moved accidentally in so far as it is in the body, which is moved. As an illustrative comparison, one might use the case of my hitting a tennis ball moving towards me. The ball moves towards me, I swing my racket, make contact with the ball, and send it towards my target. The motion of the ball, coming towards me is here first in the sequence of events in the sense that it precedes the motion of my racket: I respond to it. However, the motion of my racket is not caused by it, but by me, and the outcome, my sending the ball towards its target, is the result of the impact of the motions of my racket on the motions of the ball. My hitting the target thus presupposes the prior motion of the ball, but it is the motion of my racket which makes it hit the target. Similarly, we might say that the motion of the food, as it enters the body (cf. polla t¯on eisiont¯on, Ph. 259b 11–12), precedes my digesting it, but it is the action of my soul on the food that makes it nourish the body. If this is right, we may read ‘first’ in a temporal way when Aristotle says that the first principle of the change leading to the cessation and resumption of locomotion is external. The entrance of the food (like the oncoming tennis ball) clearly precedes my digestion of it, and any subsequent changes, but it is what the soul does to the food that makes αὑτοῦ, κατὰ συμβεβηκὸς μέντοι· μεταβάλλει γὰρ τὸν τόπον τὸ σῶμα, ὥστε καὶ τὸ ἐν τῷ σώματι ὂν καὶ τῇ μοχλείᾳ κινοῦν ἑαυτό. ‘Interacting’ may be a slight over-translation on Graham’s part, but it brings out clearly the message of the passage that the food in affecting the animal is also changed by the animal. ¹⁷ The focus on locomotion here might suggest an alternative reading whereby it is only with respect to locomotion that the animal remains a self-mover. The nutritive changes would then be entirely passive, with the food as their first mover, and the soul as the first mover at the point where those nutritive changes ceased to impede locomotion. However, this makes for a strained reading of the quoted lines. Taking ‘each self-mover’ and again ‘the thing moving itself ’ to refer to the animal, the mover that is ‘distinct’ (allo) can only be something external to the animal, which in the context is most naturally read as the first external principle mentioned in the previous sentence, i.e. the food. But then when we are told that this distinct mover is itself changed and is changed by each self-mover, it is most natural to take the movement referred to as that of nutrition, in which the food whilst acting on the self-mover is also, as we have just been told, digested and distributed around the body. That the body as a consequence of nutrition also changes place is of course the point that Aristotle seeks to establish, so for him to go on to say ‘for the body changes place, etc.’ is to be expected. ¹⁸ Cf. De an. I. 3 405b 31–406a 12.

The Soul as an Inner Principle of Change

283

the change to it in digestion, and consequently the change in the body.¹⁹ The motion of the food may be temporally first, but the motion of the soul is first in terms of efficient causation, imposing its form on the food in digestion so that it may nourish the body. This interpretation of the passage as a whole differs significantly from our first reading: it presents an interactive picture of the changes in nutrition as being caused both by external factors and by the soul. However, the primary agency clearly remains with the soul rather than the food: for whilst the food both changes and is changed, the soul remains the ‘first mover’ and is essentially unchanged.²⁰ This conclusion is the one that Aristotle needs for his overall argument in Physics VIII. On the one hand, it shows that animals are not absolutely (kuri¯os) self-movers, but require for their continued motion the temporarily prior agency of external movers, and ultimately the first unmoved mover. On the other hand, the conclusion maintains a clear sense in which animals are self-movers, and a sense in which the simple bodies are not, precisely because animals in these motions display the distinction between a moving part, the soul, and a moved part, the body, and because the soul remains unmoved, except accidentally, in the motions whereby the animal is nourished. In On Generation and Corruption, as we saw, Aristotle says that the moving cause of nutrition lies in the living being. Whilst he does not call the food ‘a cause’ of nutrition, he does use the instrumental dative for the role of the food, the same construction that we saw him use in the Physics VIII 5 for the stick by which the man hit the stone. If we take it that the food both acts on the body and is acted on by the soul in nutrition, it seems possible to make the two works consistent. However, there remains a tension between the two texts: in Physics VIII 6 Aristotle puts the emphasis on the agency of food in nutrition, because his overall argument requires him to show the way in which locomotion depends on changes initiated by the environment. The interactive reading suggested that this point was compatible with ascribing to the soul the role of first mover in nutrition and subsequent spatial motions. But even so, Physics VIII 6 may still seem to grant too much agency to the food from the point of view of On Generation and ¹⁹ Sleep and awakening happen, after all, as a result of what the organism does with the food, digesting and distributing it in the body (cf. the passives πεττομένης and διακρινομένης, 259b 13–14). Graham (1999: 114) aptly notes: ‘The fact that some factor from outside the animal plays a role in the animal’s behaviour does not by itself prove that the related behaviour is not self-caused. A robin eats a worm, becomes drowsy, and sleeps. It is true (assuming Aristotle’s account of sleep) that without the worm the robin would not have slept. But we must also consider the contribution that the robin’s digestive system makes; we could put the same worm in the stomach of a herbivorous animal, and that animal would not go to sleep.’ ²⁰ Notice also that this ‘interactive interpretation’ fits much better than the first reading with Aristotle’s other example at Ph. VIII 6 259b 9, respiration. On Youth and Old Age, etc. 27 (21) 480a 17–b12 makes it clear that the inhalation of air is caused by the expansion of the animal’s lung, which in turn allows the cold air to cool down the animal’s internal heat.

284

Thomas Johansen

Corruption with its simple identification of the soul with the efficient cause of nutrition. I hope to have shown so far that the issue of the efficient causation of nutrition is one of central importance to Aristotle’s physics, in particular to the right understanding of the claim that there are natural beings—in particular, living beings—which contain within themselves the principle of their own motions. The physical works, however, left us with a sense of uncertainty as to both the way in which and the extent to which this claim involved ascribing the efficient cause of their motions to the living beings themselves. I want now to seek further clarification from the work which, as we have seen, specializes in accounting for the principles and capacities of living beings as such, the De anima. I shall argue that Aristotle’s account of nutrition in the DA gives us a more nuanced version of the interactive picture I have suggested for Physics VIII 6.²¹ However, in addition to nutrition, I shall also have something to say about Aristotle’s account of perception, from the point of view of understanding animals as self-movers. Perception merits attention in this context because Aristotle says in De anima I 3 that ‘if the soul is moved, the most plausible view is that it is moved by the sensible objects’ (406b 10–11). What makes this view ‘the most plausible’ is no doubt Aristotle’s own belief that the animal is somehow affected in perception by the sense objects. However, I shall argue that even the case of perception illustrates the causal efficacy of the animal soul.

THE SOUL AS AN INNER PRINCIPLE OF CHANGE I N T H E DE ANIMA We saw Aristotle rehearse the claim in De anima II 1 that the living bodies are natural by having an inner principle of change and rest. Moreover, we saw him identify the soul as an inner principle of change in living beings. Given that Aristotle articulated nature in terms of the four causes in the Physics, we would expect him also in the De anima to tell us that the soul as an inner principle of change is a cause in one or more of those senses. Just this happens in De anima II 4, where Aristotle says that the soul is the formal, final, and efficient cause of the living body: ²¹ Anybody writing on this topic now has the benefit of Gill and Lennox 1994. Many of the articles in this illuminating volume, beginning with D. Furley’s classic ‘Self-Movers’, take their starting point in the problems posed by Ph. VIII for the claim that living beings are self-movers. Perception and intellectual thinking receive extensive treatment from the contributors. However, it is noticeable that none of the papers discusses Aristotle’s account of nutrition in the De an. in any detail, despite the fact that Aristotle’s prime example in Ph. VIII 6 of an apparently passive change is nutrition. Gill 1994: 21–2, has a paragraph on the efficient causation of nutrition in De an. II 4, which, however, discusses only the relationship between the soul and the inner heat, and not that between the soul and the nutriment.

The Soul as an Inner Principle of Change

285

The soul is the cause (aitia) and principle (arkh¯e) of the living body. But these are said in many ways, and similarly the soul is cause in the three of the defined ways: for the soul is the cause both as that from which the change is and as that for the sake of which and as the substance of ensouled bodies. That it is the cause as substance is clear: for the substance is the cause of being for all things, and living is being for living beings, and the soul is the cause and principle of this. Moreover, the actuality is the formula (logos) of what is potentially. (De an. II 4 415b 8–15)

Note that when Aristotle says here that the soul is the cause as ‘that from which the change is’, i.e. the efficient cause, he has in mind not just change in the sense of locomotion, but more generally change in the categories of quality and quantity.²² This becomes clear later in the text when Aristotle, clarifying the three ways in which the soul is cause, says with relation to the efficient cause: And the soul is the first thing from which the change according to place is. But this capacity does not belong to all living beings. But alteration and growth are also due to the soul. For perception seems to be a kind of alteration, and nothing perceives which doesn’t participate in soul, and something similar holds for growth and diminution, since nothing diminishes or grows naturally unless it is being nourished, and nothing is nourished which does not share in life. (De an. II 4 415b 21–27)

Aristotle here suggests the reason why he does not just want to say that the soul is the efficient cause of locomotion: It is only certain animals which have locomotion, so if the soul were the efficient cause of locomotion only, then the claim that the soul is the efficient cause of the living body would be true only for some of the fancier kinds of living being. But Aristotle wants to make a general claim that applies to all living beings. The best way of establishing as general the claim that the soul is a cause in all three senses, formal, final, and efficient, is to show that it holds true of the kind of soul that all living beings have—that is to say, the nutritive soul. That is why, I would suggest, Aristotle chooses to make the claim that the soul is a final, formal, and efficient cause in the chapter where he analyses the nutritive soul. He shows the soul to work as a formal cause in an engagement with Empedocles. Against Empedocles, Aristotle insists that the material elements of the plant are incapable themselves of imposing a limit and a logos, i.e. a determinate form, on the growth of the plant. This is, rather, the job of the soul. He shows the soul to be the final cause of nutrition in so far as nutrition aims to keep the living being alive, whilst generation (also a function of the nutritive soul) aims to make another living thing such as itself, thereby participating in the divine as far as possible for a mortal being. ²² Again the De anima’s account of the soul is tracking the Physics’ account of nature, cf. Ph. II 1 192b 13–15, which allows for nature to be an inner principle of change, not just of locomotion but also of growth and diminution or of alteration: ‘each [of the natural beings] has within itself a principle of change and rest, in respect of place, or of growth and decrease, or by way of alteration.’

286

Thomas Johansen

T H E S O U L A N D T H E E F F I C I E N T C AU S E OF NUTRITION The efficient causation of nutrition is a more complicated issue. Nutrition, like change in general, is a case of one thing being acted on by another such as to become like it, having previously been unlike it. In nutrition ‘what is nourished’ (to trephomenon) first acts on the food so as to make it like itself.²³ The change is all in the food. What is nourished changes no more than does the builder when he builds; there is merely a switch from inactivity to activity in what is nourished. The efficient cause of nutrition is thus ‘what is nourished’ and not the food. The body is nourished qua ensouled. It is qua ensouled, then, that the living body acts as an efficient cause on the nutriment. This formulation allows for a more nuanced picture of the efficient causation of nutrition. For, on the one hand, it is body qua ensouled which—by means of our natural heat—digests the food. This preserves the claim made earlier in the chapter that the soul is the efficient cause of nutrition. On the other hand, the ensouled body may qua body also be said to be acted on in nutrition: it is nourished by the soul. The ensouled body, then, acts qua ensouled on itself qua body so as to nourish itself. More precisely (416b 20–3),²⁴ the soul is what nourishes (to trephon) and the body what is nourished (to trephomenon). Where does this leave the nutriment (h¯e troph¯e)? The opening of chapter II 4 led us to expect that the nutriment would somehow be involved in the account of the actuality of nutrition—indeed, that an account of the actuality of the nutriment would somehow be prior to an account of the activity of nutrition.²⁵ At 416b 17–19 Aristotle seems to address this expectation: ‘the upshot is that this kind of principle of the soul is a capacity (dunamis) to preserve what has it qua being such, whilst the nutriment furnishes the activity’.²⁶ Here, finally, is an indication that the nutriment (h¯e troph¯e) plays a role in the activity of nutrition, such that the capacity of nutrition will remain unactualized without it. What that role is, however, is still not clear. Simplicius takes the nutriment to be the

²³ ἔτι πάσχει τι ἡ τροφὴ ὑπὸ τοῦ τρεφομένου, ἀλλ’ οὐ τοῦτο ὑπὸ τῆς τροφῆς, ὥσπερ οὐδ’ ὁ τέκτων ὑπὸ τῆς ὕλης, ἀλλ’ ὑπ’ ἐκείνου αὕτη. ὁ δὲ τέκτων μεταβάλλει μόνον εἰς ἐνέργειαν ἐξ ἀργίας (416a 34–b3). ²⁴ Albeit not entirely precise, cf. 416b 30. Burnyeat’s warning (2002: 51–3) about the interim status of the discussion of perception in De an. II 5 applies also to the discussion of nutrition in II 4. ²⁵ εἰ δὲ χρὴ λέγειν τί ἕκαστον αὐτῶν, οἷον τί τὸ νοητικὸν ἢ τὸ αἰσθητικὸν ἢ τὸ θρεπτικόν, πρότερον ἔτι λεκτέον τί τὸ νοεῖν καὶ τί τὸ αἰσθάνεσθαι· πρότεραι γάρ εἰσι τῶν δυνάμεων αἱ ἐνέργειαι καὶ αἱ πράξεις κατὰ τὸν λόγον. εἰ δ᾿ οὕτως, τούτων δ᾿ ἔτι πρότερα τὰ ἀντικείμενα δεῖ τεθεωρηκέναι, περὶ ἐκείνων πρῶτον ἂν δέοι διορίσαι διὰ τὴν αὐτὴν αἰτίαν, οἷον περὶ τροφῆς καὶ αἰσθητοῦ καὶ νοητοῦ (De an. II 4 415a 16–22). ²⁶ ὥσθ᾿ ἡ μὲν τοιαύτη τῆς ψυχῆς ἀρχὴ δύναμίς ἐστιν οἵα σώζειν τὸ ἔχον αὐτὴν ᾗ τοιοῦτον, ἡ δὲ τροφὴ παρασκευάζει ἐνεργεῖν.

The Soul as an Inner Principle of Change

287

material cause of nutrition, comparable to the builder’s timber.²⁷ This allows us to see why nutrition does not happen without food. However, it is harder to make sense of the way in which the nutriment on this reading ‘furnishes the actuality’ for the nutritive faculty’s potentiality. For normally we associate the material cause with the potentiality to undergo a change, and not with what actualizes that potentiality.²⁸ Aristotle goes on in the next line to use the language of instrumentality to describe the nutriment. It is that by which the body is nourished (h¯oi trephetai) and later (416b 25) it is that by which the soul nourishes the body (h¯oi trephei). The latter expression Aristotle says is used in two ways, comparable to the hand and the rudder on a ship, the hand moving whilst being moved, the rudder simply being moved. He suggests that the hand is like the natural heat, changed by the soul and in turn changing the food, whereas the food is like the rudder being changed, whilst not itself changing another thing. This again makes the nutriment look entirely passive. However, what Aristotle describes here is specifically the process of digestion (pepsis), the process that leads up to the body’s assimilation of the final nutriment. He is, in other words, not concerned with the entire nutritive process.²⁹ For at De an. II 4 416b 3–7 Aristotle distinguished between the undigested first nutriment and the digested nutriment which is the final thing that is added to the body (to teleutaion prosginomenon, 416b 3).³⁰ We may take it that the first nutriment is passive in the process of digestion, being moved by the natural heat without moving anything else in turn. But we may also see the last nutriment as acting on the body in so far as it is the last instrument used by the soul in maintaining the body. Simplicius may still be right that the way to understand the first undigested nutriment is as matter for nutrition.³¹ But Aristotle’s imputation of agency to the nutriment at 416b 11–17 fits better as referring to the final nutriment, which serves not just as matter but also as an instrument of nutrition. The sense of agency appears when the food is referred to as ‘capable of producing growth’ (aux¯etikon, 416b 13), in so far as the ensouled body is a certain quantity (poson ti), while in so far as the ensouled body is ‘a certain kind of thing or substance’ (tode ti kai ousia, 416b 13) the food qualifies as nutriment (troph¯e). In the ²⁷ ἡ τροφὴ παρασκευάζει ‘‘ὡς ὔλη ἐπιτηδεία πρὸς τὴν ἐνέργειαν παρακειμένη καθάπερ τὰ ξύλα τῷ τέκτονι· οὐ γὰρ ἄνευ μὲν τροφῆς, ἔνδοθεν δὲ ἐφ᾿ ἑαυτῆς ἐνεργεῖ ψυχή. δι᾿ ὅ τὸ τρέφον ἐστὶν ἡ πρώτη ψυχή’’ (in De an. 115, 18). ²⁸ Cf. Gen. corr. II 9 335b 29–31 τῆς μὲς γὰρ ὕλης τὸ πὰσχειν ἐστὶ καὶ τὸ κινεῖσθαι, τὸ δὲ κινεῖν καὶ ποιεῖν ἑτέρας δυνάμεως. It may of course still be true that the materials ‘furnish the actuality’ in the sense that they provide the builder with the opportunity to exercise his art. ²⁹ Similarly, the moving of the rudder is hardly the end of the causal process, for the rudder moves the ship, which is surely the proper end of the entire process! ³⁰ πότερον δ᾿ ἐστὶν ἡ τροφὴ τὸ τελευταῖον προσγινόμενον ἢ τὸ πρῶτον, ἐχει διαφοράν. εἰ δ᾿ ἄμφω, ἀλλ᾿ ἡ μὲν ἄπεπτος ἡ δὲ πεπεμμένη, ἀμφοτέρως ἂν ἐνδέχοιτο τὴν τροφὴν λέγειν· ᾗ μὲν γὰρ ἄπεπτος, τὸ ἐναντίον τῷ ἐναντίῳ τρέφεται, ᾗ δὲ πεπεμμένη, τὸ ὅμοιον τῷ ὁμοίῳ. ³¹ A view supported, and perhaps suggested to Simplicius, by the craftsman analogy at 416a 34–b3, in particular, ὥσπερ οὐδ᾿ ὁ τέκτων ὑπὸ τῆς ὕλης …

288

Thomas Johansen

capacity of nutriment, the food is referred to as ‘saving’ the substance of what is nourished, whilst in generation Aristotle says that it is ‘capable of producing generation’ (genese¯os poi¯etikon, 416b 15)—that is, capable of producing another substance like the one nourished.³² The view of the food as in some sense active in growth and nutrition fits better if we think of the final nutriment. For the final nutriment may be thought of as acting on the body as the final step in the efficient causation of nutrition, as the last instrument that the soul uses in nourishing the body.³³ Finally, as we saw, Aristotle says at 416b 17–19 that the nutriment furnishes (paraskeuazei) the nutritive capacity with the activity of saving the living body. Here again a reference to the final nutriment seems more appropriate, since it is only in so far as the food has been digested and is added to our bodies that it contributes to our survival.³⁴ On this reading we can combine two features of Aristotle’s account that otherwise seem to be at odds: on the one hand, his implication at the beginning of De anima II 4 that the nutriment has priority in actuality over the nutritive soul, and, on the other, his insistence that the nutritive soul is the agent and efficient cause in relation to the nutriment. My suggestion is that the final nutriment is prior in actuality to the nourished body qua nourished, whilst the soul is the efficient cause in relation to the first nutriment, acting on it so to make it like the nourished body. Where does this account leave us with respect to the argument of Physics VIII 6? The account clearly confirms the causal efficacy of the soul in nutrition. But does it do so to such an extent as to undermine Aristotle’s suggestion in Physics VIII 6 that the nutriment moves the living being? Not on the interactive reading I suggested for Physics VIII 6. On this reading, the soul acted on the body in nutrition by the agency of the food, which was thereby both moving and moved. The food was thus understood as the instrument by which or through which the soul nourishes the body, along the lines of the example of the man moving ³² The suggestion that the food somehow produces change or rest in the nourished body is echoed by the De sensu, which identifies specific causal properties in the food (heat and sweetness) as responsible for growth and nutrition: ‘Now, among the perceptible elements of the food which animals assimilate (προσφερομένης), the tangible produce (ποιοῦντα) growth and decay; it is qua hot or cold that the food assimilated (προσφερόμενον) produces (ποιεῖ) these; for heat and cold produce growth and decay. It is qua tastable, however, that the assimilated (προσφερόμενον) food nourishes. For all organisms are nourished by the sweet, either in simple state or in a mixture’ (Sens. 4, 441b 27–424a 2, Oxford revised trans.). ³³ In his commentary on 416b 25–9 Ross seems right to say that the food as ‘that by which the body is nourished’ has to be the same as one of the things ‘by which the soul nourishes’. It is less clear, however, given that Aristotle has said now that the food is only moved, how he can also mean to say that the nutritive soul, as Ross puts it, ‘acts on the food … and in turn the food acts on τὸ τρεφόμενον, i.e. on the body’ (l. 22) (my italics). My suggestion is that the first nutriment is passive in the digestive part of the nutritive process, yet the final nutriment can also be seen as acting instrumentally in maintaining the body. On this reading we can preserve the impression given by ᾧ τρέφεται that the food in some sense affects the body. ³⁴ The verb παρασκευάζω is naturally used of the role of an instrument, cf. De motu an. 702a 17: τὰ μὲν γὰρ ὀργανικὰ μέρη παρασκευάζει ἐπιτηδείως τὰ πάθη …

The Soul as an Inner Principle of Change

289

the stone by means of the stick in Physics VIII 5, or the example of my hitting the target with the moving tennis ball. De anima II 4 similarly presented the nutriment as ‘that by which’ the soul nourishes the body. It seems appropriate, indeed, to understand Aristotle’s threefold distinctions in De anima II 4 between the soul as ‘what nourishes’, the body as ‘what is nourished’, and the food as ‘that by which the body is nourished’, as a direct application of Aristotle’s three-fold scheme in Physics VIII 5 of the mover, the moved, and the means by which the mover moves. The account of nutrition in De anima II 4 thus left room for ascribing a kind of causal efficacy to the nutriment as an instrument used by the soul to act on the body. I argued, furthermore, that Aristotle, when ascribing instrumental agency to the nutriment, particularly had in mind the final nutriment. To be sure, it was hardly nutriment in the sense of final nutriment that Aristotle had in mind when talking about the first external mover in Physics VIII 6. But this difference is at least partly explicable by the fact that Aristotle in Physics VIII 6 was interested not so much in nutrition as such but in a by-product of nutrition, the cessation and resumption of locomotion, which is not clearly related to the formal changes that the food undergoes in digestion from first to final nutriment.³⁵ In so far as we worry about whether the account of nutrition in De anima II 4 presents the living being as too much of a self-mover for the liking of Aristotle in Physics VIII 6, the point seems worth making that the nutriment has some instrumental agency also in De anima II 4, even if it is only that of the final nutriment. On the other hand, in so far as we worry that the ascription of any efficacy to the nutriment is too much from the point of view of the GC, we may point out that the agency is only as an instrument and therefore not strictly as a first moving cause, a role reserved for the soul. Moreover, the instrumental agency was only that of the final nutriment, which has already been affected by the soul in the manner suggested by On Generation and Corruption. In De anima II 4, then, we seem to have a version of the interactive picture of the efficient causation of nutrition which goes a considerable way towards relaxing the tension between the Physics and On Generation and Corruption. Much of my argument concerning nutrition has hinged on the premisses that the soul is an unmoved mover in nutrition, and that whatever causal efficacy is ascribed to the food in nutrition has to be subordinate to that of the soul as the first mover. We saw that Aristotle in De anima II 4 claimed that the soul was the moving cause of both nutrition and perception. Yet I also noted Aristotle’s suggestion in De anima I 3 that the soul might in fact be moved in perception ³⁵ At Ph. 259b 12–13 we are told that locomotion ceases when the food is digested, and restarts when the food is being distributed. This may be an echo of the distinction between the first and final nutriments, but we are given no indication what it is about the different stages of food that would cause sleep and awakening. For an account of how food relates to the processes that bring about sleep and awakening we have to wait until Div. somn. 3.

290

Thomas Johansen

by the sense objects. Does perception, then, turn out to be an exception to the claim that the soul is an efficient cause?

T H E S O U L A N D T H E E F F I C I E N T C AU S E O F PE RC E P T I O N Two differences from nutrition stand out immediately when we turn to perception in De anima II 5. One is that perception is a kind of alteration—that is, a change in quality—whereas growth and nutrition were changes in the categories of quantity and substance respectively. Another is that whereas the nutritive soul was described as active in relation to its proper object, making the nutriment like the nourished body, the perceptual part is said to be passive in relation to its proper object, the sense objects, suffering (paskhein) in perception so as to become like the sense object.³⁶ Later in the chapter (417b 19–21) Aristotle seems explicitly to identify the sense objects as the efficient causes of the activity of perception by calling them ‘the producers of the activity’ (ta poi¯etika t¯es energeias). Perception, then, is a passive change in the perceiver caused by an external efficient cause, the sense object. This claim seems to rule out that the soul is also the efficient cause of perception. But if so, perception also fails to exemplify Aristotle’s claim in De anima II 4 that the soul is the cause of the body as the final, formal, and efficient cause. I want to resist this conclusion, however, by arguing two points: first, that the sense object is the efficient cause of actual perception only in a qualified sense, because actual perception is an alteration only in an attenuated sense; and secondly, that the soul is the efficient cause of the ordinary alterations that are involved in perceiving. On the first point I shall be brief and refer the reader to Myles Burnyeat’s comprehensive treatment in his ‘De Anima II 5’ (2002). In ordinary alteration, as Aristotle reminds us, the efficient cause makes the patient have in actuality a quality that it had only in potentiality prior to the change. However, though Aristotle thinks that perception is in some sense an alteration, he does not believe that perception is an ordinary alteration. The clarification begins at 417a 21. As we know from Physics III 1, change for Aristotle is in general the actuality of a potentiality qua potential. But, we are now told, there are two ways in which something can have a potentiality and, correspondingly, two ways in which a potentiality can be actualized in change. Aristotle thus distinguishes the two sorts of being changed or suffering (paskhein): one is the transition from first to second potentiality, the other from second potentiality to an actuality. Knowledge serves as the illustration. It is one thing to undergo the sort of change that a child ³⁶ De an. II 5 416b 33–5: ἡ δ᾿ αἴσθησις ἐν τῷ κινεῖσθαί τε καὶ πάσχειν συμβαίνει, καθάπερ εἴρηται· δοκεῖ γὰρ ἀλλοίωσίς τις εἶναι.

The Soul as an Inner Principle of Change

291

experiences when he learns arithmetic. It is another thing to change into being active with respect to knowledge that you already have, in the manner of the working mathematician. From these examples, Aristotle generalizes: Being affected is not a simple thing, but in one case it is a kind of destruction by the opposite, and in the other it is rather a preservation by what is in actuality of what is potentially and is potentially thus as potentiality is to actuality. For what has the knowledge comes to be contemplating (the¯oroun), something which is either not being altered (for the transition is into itself and into completion³⁷) or another kind of alteration. (417b 2–7)³⁸

Aristotle’s objection no doubt turns at least in part on the thought that altering (alloiousthai) ought to involve becoming ‘other’ (allo).³⁹ In the case of thinking about something you already know, you are not becoming different: you are showing yourself for what you already are (cf. epidosis eis hauto): namely, knowledgeable. Aristotle gives us the choice between describing the shift from second potentiality to actuality either as not an alteration or as another kind of alteration. The implication of the choice seems to be: if you are going to call the shift an alteration at all, you had better make sure you specify that it is a different kind of alteration from the ordinary one! Having distinguished between the alteration from first to second potentiality and that from second potentiality to actuality, Aristotle goes on to make further distinctions. Not only is actual thinking not an ordinary alteration; neither is learning. Learning too contrasts with the sort of alteration whereby one quality is destroyed by another, say hot in favour of cold. We might of course say that when you learn your ignorance is being destroyed, in that sense the alteration would count as a change to a state of privation, just as the change from hot to cold is a change to a state of privation of heat. But two things should discourage us from assimilating learning to such alterations. First, by learning, you develop your own nature, your rationality as a human being.⁴⁰ Secondly, by learning, you ³⁷ I take it as no accident that Aristotle uses the word entelekheia here: contemplation is the state in which knowledge has reached its end. ³⁸ οὐκ ἔστι δ᾿ ἁπλοῦν οὐδὲ τὸ πάσχειν, ἀλλὰ τὸ μὲν φθορά τις ὑπὸ τοῦ ἐναντίου, τὸ δὲ σωτηρία μᾶλλον ὑπὸ τοῦ ἐντελεχείᾳ ὄντος τοῦ δυνάμει ὄντος καὶ ὁμοίου οὕτως ὡς δύναμις ἔχει πρὸς ἐντελέχειαν· θεωροῦν γὰρ γίνεται τὸ ἔχον τὴν ἐπιστήμην, ὅπερ ἠ οὐκ ἔστιν ἀλλοιοῦσθαι (εἰς αὑτὸ γὰρ ἡ ἐπίδοσις καὶ εἰς ἐντελέχειαν) ἠ ἕτερον γένος ἀλλοιώσεως. ³⁹ As Sorabji (1992: 221) notes. ⁴⁰ One may perceive a tension between this denial that learning is a change to privative temporary conditions and the earlier claim that the learner alters many times and often changes from an opposite state, hexis (417a 31–3). However, two points should reduce the tension. First, Aristotle at 417a 31–3 does not refer to the repeated changes as ‘changes to privative states’, as at 417b 15. Taking a leaf from Burnyeat (2002: 53), we may understand the process as a repeated series of changes from ignorance in respect of particular items of knowledge within a body of knowledge as a whole: e.g. you learn arithmetic by changing from a state of ignorance first with respect to addition, then with respect to subtraction, and multiplication, and so on. Secondly, as Burnyeat shows, Aristotle is making a contrast between diathesis and hexis at 417b 15–16, which there specifies the sense of hexis as ‘a fixed dispositional state’, whereas at 417a 33 hexis is left vague.

292

Thomas Johansen

develop a hexis, a ‘fixed dispositional state’ in Burnyeat’s phrase,⁴¹ rather than a temporary condition (diathesis) of the sort ordinarily associated with privations.⁴² How do these distinctions apply to perception? Aristotle tells us: The first change of the sense faculty comes about by the parent, but whenever one is born, one already possesses (ekhei), similar to the state of knowledge, perceiving (to aisthanesthai). Perceiving in its activity is spoken of similarly to thinking, but is different in that in its case what produces the activity is external, the visible and the audible, and likewise also the other sensibles. (De an. II 5 417b 16–21)⁴³

Aristotle refers to the state one is in at birth with regard to perception as already possessing perceiving (ekhei ¯ed¯e … to aisthanesthai. Here he uses a grammatical form, the articular infinitive, to refer to ‘perceiving’), a form that he elsewhere reserves for the activity of perceiving (De an. 415a 18; cf. 402b 13). However, earlier in chapter II 5 (417a 9–12) he had warned us that aisthanesthai is said in two ways, of the activity of seeing and hearing and of the potentiality of seeing and hearing, of the sort we are characterized by when we are asleep. So he told us that also what is asleep is said to be hearing and seeing. We now know that the potentiality of the sleeper should be understood as a second potentiality, and we can therefore appreciate that when we say that he who sleeps is perceiving, there is good reason for our saying so, given how close second-potentiality perceiving is to actual perceiving. Somebody who is in a state of second-potentiality perceiving is in a sense already perceiving. For as a perceiver he needs to develop no further in order actually to perceive: he already (¯ed¯e) has perceiving, as Aristotle puts it at 417b 18. Aristotle’s view that the agency of the sense object actualizes a second potentiality throws up two significant parallels with nutrition. Aristotle told us in De anima II 4 that ‘the food is affected somehow by what is nourished, but

⁴¹ Cf. Burnyeat 2002: 62–3, 77. ⁴² The overall result of the distinctions is Burnyeat’s (2002: 66) triple scheme of potentialities and actualities: ‘(Pot1) the ordinary potentiality of a hot thing to be cold or of a cold thing to be hot; (Pot2) the first potentiality, grounded in a thing’s nature, to be a fully developed thing of its kind, capable of exercising the dispositions which perfect its nature; (Pot3) the second potentiality of a developed thing to remain a fully developed thing of its kind by exercising, and thereby preserving, the dispositions which perfect its nature … (Act1) is the actuality of (Pot1), and proceeds towards the replacement of (Pot1) by a contrary potentiality of the same kind; (Act2) is the actuality of (Pot2), and develops the dispositions which perfect the subject as a thing of its kind; (Act3) is the actuality of (Pot3), and contributes to the continued preservation of the dispositions which perfect the subject as a thing of its kind.’ ⁴³ τοῦ δ᾿ αἰσθητικοῦ ἡ μὲν πρώτη μεταβολὴ γίνεται ὑπὸ τοῦ γεννῶντος, ὅταν δέ γεννηθῇ, ἔχει ἤδη, ὥσπερ ἐπιστήμην, καὶ τὸ αἰσθάνεσθαι. τὸ κατ᾿ ἐνέργειαν δὲ ὁμοίως λέγεται τῷ θεωρεῖν· διαφέρει δέ, ὅτι τοῦ μὲν τὰ ποιητικὰ τῆς ἐνεργείας ἔξωθεν, τὸ ὁρατὸν καὶ τὸ ἀκουστόν, ὁμοίως δὲ καὶ τὰ λοιπὰ τῶν αἰσθητῶν.

The Soul as an Inner Principle of Change

293

this is not affected by the food, just as the builder (ho tekt¯on)⁴⁴ does not suffer by the matter, but it by him. The builder changes only from inactivity into activity’ (416a 34– b 3). This is the very same analogy that Aristotle uses in De anima II 5 (417b 8–9) to characterize the activity of knowing, and, I take it, the activity of perceiving.⁴⁵ Given the use of the builder analogy in De anima II 4, what is striking about its deployment in II 5 is that it now applies to what we expect to be the patient in change, the perceptual part. The implication is that in actual perception the patient changes no more than does the efficient cause in the case of standard changes, like building. It seems, therefore, that the change in the sensory part, which is passive in relation to the sense object, is no greater than the change in the nourished body in the case of nutrition, where the nourished body qua ensouled is active in relation to the nutriment. For in both cases the ‘change’, so-called, is one of bringing into activity an attribute by which one is already qualified, just as a builder is already a builder before he starts building, and is merely made to manifest his skill in the act of building. Secondly, there is a parallel between nutrition and perception that may help explain Aristotle’s choice of the obscure language of ‘preservation’ and ‘transition into the thing itself ’ in De anima II 5.⁴⁶ Aristotle in De anima II 4 contrasted the role of food in maintaining the living being and in generating new offspring, for nutrition is a matter of preserving (s¯ozei) the living being, whilst generation is the creation of another thing like the living being. Aristotle then summed up the role ⁴⁴ τέκτων can mean specifically ‘carpenter’, rather than generally ‘builder’ or ‘artisan’, but there is nothing in Aristotle’s argument in De an. II 4 to show a specific interest in carpentry. ⁴⁵ It may be objected that the builder analogy applies only to actual knowing, and not to actual perceiving, since the analogy serves to explicate the claim that it is not right to say that knowing is being altered, whereas, as Burnyeat argues (2002: 59–61), it is right in a sense to say that perceiving is being altered. I am less convinced than Burnyeat, however, that Aristotle means to make a difference between knowing and perceiving at this point. It seems to me for two reasons that Aristotle keeps in play the claim that both actual knowing and actual perceiving can be treated either not as an alteration or as a different kind of alteration from the ordinary ones. The first reason is that διὸ οὐ καλῶς ἔχει λέγειν τὸ φρονεῖν, ὅταν φρονῇ ἀλλοιοῦσθαι at 417b 8 seems to draw a consequence from the choice given (ἢ οὐκ ἔστιν ἀλλοιοῦσθαι … ἢ ἕτερον γένος ἀλλοιώσεως) rather than from just one of the alternatives. The implication is: it is not right simply to talk about actual knowing as being altered, given that actual knowing is not in any standard sense an alteration. Similarly, Aristotle goes on at 417b 13–15 to present us with the very same choice for the development from first to second potentiality knowing: either it should not be said to be a case of being affected (paskhein) at all, or it should be said to be a different kind of alteration from the ordinary sort. So as in the case of actual thinking, either don’t talk about alteration or make the required qualifications, which leaves both, in different ways, a kind of alteration. Secondly, I would emphasize that De an. II 5 is about perception, not knowledge. We should expect, therefore, that the job of the analysis of knowing in this chapter is to present distinctions within actuality and potentiality that apply to perception. It seems to me that for Aristotle to describe the activity of thinking in a way that will then after all not apply to the activity of perceiving because actual thinking is no alteration at all, whereas actual perceiving is an alteration in a way, runs counter to the explanatory purpose of the chapter. The explications of the similarities and differences between thinking and perceiving I take to be a job that Aristotle rather leaves for De an. III 4–5. For these reasons I am also inclined to take the artisan analogy to apply to both thinking and perceiving. ⁴⁶ The second of the two is noted by Burnyeat (2002: 56) as ‘a surprisingly lyrical phrase’.

294

Thomas Johansen

of food in nutrition as follows: ‘so the principle of the soul is such a capacity as to save (s¯ozein) what has it qua such, whilst the food furnishes the activity: that is why deprived of food it is not possible to be’ (416b 17–20).⁴⁷ Now compare the description of actual perception. Aristotle holds that actual perception, like actual knowing, is ‘a preservation (s¯ot¯eria) by what is in activity of what is potentially and is potentially thus as potentiality is to activity’ (417b 3–5). The role of the sense object is to produce this activity in the sense faculty, but it is not, as we saw, to alter the sense faculty in the ordinary sense of ‘alter’. When the sense objects are referred to as what produce the activity (ta poi¯etika t¯es energeias), we therefore need to understand it as a case of the sense objects helping preserve what the sense faculties already are at the level of second potentiality. But, so described, the role of the sense object seems analogous to that of the final nutriment in nutrition: food too furnishes the activity (paraskeuazei energein), as Aristotle put it (416b 19), and it does so not by altering the living being but by preserving what the living being already is. Notice here also the parallel between actual thinking, and by implication actual perception, as a transition into the thing itself (eis hauto) and nutrition as a preservation of the thing itself (auto heauto, 416b 17). Now there is clearly a difference between nutrition and perception, in that the actuality of perception does not keep us alive, the way nutrition does. Ceasing to nourish ourselves kills us, ceasing to perceive does not. But this difference can be understood in terms of the categorical difference between perception and nutrition: perception is a change in the category of quality; nutrition is a change or rest in the category of substance. With this difference in mind, we may understand the sense object as preserving or sustaining the living being qua perceiver—that is, as capable of undergoing a certain sort of qualitative change—in a parallel way to that in which the nutriment preserves or sustains the living being qua substance. One way of taking this point is to think that our powers of perception are lessened if we do not regularly exercise them, just as we lose our ability to use our knowledge if we do not regularly exercise it.⁴⁸ However, Aristotle’s point is not so much that the second potentiality of knowledge and perception are preserved by the activities but that they are preserved in the activities themselves. The perceptual capacity is thus preserved rather than altered in the activity of perceiving, because, I take it, actual perceiving is perceiving in the fullest sense, just as actual knowing is knowing in the fullest sense. It is part of my suggestion here that Aristotle’s peculiar choice of the terms ‘destruction’ and ‘preservation’ at 417b 3, terms which sensu strictu apply only to change and rest in the category of substance, is meant to invite us to draw this parallel with nutrition, which really is, as we know from De anima II 4, a change in the category of substance. The pay-off of drawing this parallel is to see the action of ⁴⁷ ὥσθ᾿ ἡ μὲν τοιαύτη τῆς ψυχῆς ἀρχὴ δύναμίς ἐστιν οἵα σώζειν τὸ ἔχον αὐτὴν ᾗ τοιοῦτον, ἡ δὲ τροφὴ παρασκευάζει ἐνεργεῖν. ⁴⁸ Cf. Burnyeat 2002: 55.

The Soul as an Inner Principle of Change

295

the sense object on the perceiver more clearly as a special case of an action that does not alter but preserves. For that is what the final nutriment does in relation to the nourished body: it preserves in activity what the living body already is. The analogies are of course only partial. The concluding lines of De anima II 5 (418a 3–6) remind us that actual perceiving is a case of becoming assimilated to the sense object. But whatever story we tell about the nature of this assimilation, it has to be one that is consistent with Aristotle’s message that the action of the sense object does not make the sense faculty other than what it was, but preserves it. For our purposes the important point is that given Aristotle’s view of the activity of perception, the sense object cannot therefore be understood as an efficient cause in the normal sense of the term. Where does this observation leave us with respect to the challenge posed by Physics VIII 6 to our view of animals as self-movers? David Furley observed in ‘Self-Movers’ that in perception ‘the soul is not moved by the objects in the external world in any of the senses enumerated in [De an.] I 3 (φορά, ἀλλοίωσις, φθίσις, αὔξησις), except that it experiences this highly specialized form of ἀλλοίωσις’. He then asked, Is this qualification sufficient to allow Aristotle to maintain his distinction between the movements of animals and the natural motions of inanimate bodies? It is certainly not sufficient in itself, because he uses the same pattern in his explanation of natural motion (Phys. VIII. 4, 255a 30– b 13). In this case too we can distinguish two stages: the change from (say) water, which is potentially air, into air, through an external agency; and then the full actuality of the element in attaining its natural place. Here too Aristotle uses the simile of the man first learning, and then exercising, his skill. So if animals are self-movers but inanimate natural bodies are not, the difference in the explanation of their motions is not to be found in this point.⁴⁹

We may support Furley’s claim by adding a further point. In Physics VIII 4 Aristotle argues that the external cause which allows the air, say, to actualize its second potentiality to rise up, counts as accidental, because it simply removes an obstacle to the air’s own natural motion.⁵⁰ Similarly, Aristotle says of knowledge that ‘its exercise follows at once upon the possession of it unless something prevents it’ (255b 22–3). However, the sense object is not an accidental but an ⁴⁹ Furley 1994: 10. ⁵⁰ ‘Thus not only when a thing is water is it in a sense potentially light, but when it has become air it may be still potentially light; for it may be that through some hindrance it does not occupy an upper position, whereas, if what hinders it is removed, it realizes its activity and continues to rise higher. The process whereby what is of a certain quality changes to a condition of actuality is similar: thus the exercise of knowledge follows at once upon the possession of it unless something prevents it. So too what is of a certain quantity extends itself over a certain space unless something prevents it. The thing in a sense is and in a sense is not moved by one who moves what is obstructing and prevent its motion, e.g. one who pulls away a pillar or one who removes the stone from a wineskin in the water is the accidental cause of motion … . So it is clear that in all these cases the thing does not move itself, but it contains within itself the source of change—not of changing something or of causing change, but of suffering it’ (Ph. VIII 4 255b 17–31, revised Oxford trans.).

296

Thomas Johansen

essential cause of actual perception. The presence of a sense object is always and necessarily required for actual perception, whereas an external cause is required in order to allow the air to move up only in those cases where it happens to be trapped. Hence the reason for seeing the activity of perception as dependent on the action of external objects is even better than the reason for seeing the activity of thinking or of air moving upwards as depending on external agents. Furley is correct, then, that the point that actual perception is the actuality of a second potentiality is insufficient in itself ‘to allow Aristotle to maintain his distinction between the movements of animals and the natural motions of inanimate bodies’. However, from this it does not follow that the point is irrelevant to the distinction.⁵¹ For the point, as we saw, comes with Aristotle’s claim that the ordinary changes in perceiving are those leading to the second potentiality. As we shall see, it is the animal’s own soul that is responsible for these changes. It is the point that the sense object does not play the role of an ordinary efficient cause in actual perception combined with the role that the soul plays in the ordinary changes leading to the formation of the second potentiality that allows Aristotle to maintain his distinction in the case of perception. If all the ordinary changes involved in perceiving have already taken place at the time of birth, then we want to know what the efficient cause of those changes was. Aristotle’s language at 417b 16–18 suggests that he sees two changes as required before an animal has the second potentiality of perceiving.⁵² The first change is that initiated by the male parent. The second is the development that the embryo undergoes from conception until birth. For more information about the stages of foetal development, we naturally turn to the Generation of Animals. Here it emerges that the first change referred to in De anima II 5 is the change that happens at the point of conception when the male semen first acts on the female menstrual fluid to form the conception. At this stage the father acts as the efficient cause, transmitting through the motions of the semen the animal form to the female’s matter. Aristotle explains in Generation of Animals II 1 (735a 12–25) that the movements of the semen are those of the nutritive soul.⁵³ The effect of the action of these movements on the ⁵¹ Nor does Furley say that it follows. However, he does abandon the question of the efficient causation of perception at this point, choosing instead to address the problem in terms of the intentionality of phantasia and thought: even where our actions follow as a result of stimulation by an external object, it is properly speaking the object as an intentional object (rather than as external), and therefore as formed by our psychological powers, which brings about the desire and the action. But this answer, if correct, applies only to those cases where the animal motion is caused by desire (orexis) and therefore addresses neither Aristotle’s specific example of nutrition in Ph. VIII 6 nor his specific worry in De an. I 3 that it is perception that most of all makes the animal soul appear passive. Nor, therefore, does Furley’s account qualify as a general answer to how living beings as self-movers are distinguished from other natural beings such as fire and air. ⁵² Note the μέν … δέ contrast in ἡ μὲν πρώτη μεταβολή … ὅταν δὲ γεννηθῇ. ⁵³ Cf. Gen. an. 737a 18–20. Thus the embryo grows to have the perceptual parts by the agency of its nutritive soul which, because it is the nutritive soul of an animal, is such as to produce or preserve a being with a perceptual soul. This explains why the perceptual part does not have to

The Soul as an Inner Principle of Change

297

menstrual fluid is the formation of the nutritive part, for blooded animals situated in the heart. Once the nutritive part has been formed, this part continues the process of growth that leads to the formation of the complete embryo, including the sense-organs. That the sense-organs (in particular the eyes)⁵⁴ should develop later in the formation of the animal is appropriate, since perception is the end for which animals live qua animals.⁵⁵ On this story, the efficient cause of the perceptual faculty appears in the first instance to be the father, because he sets up the first change. Recall here Aristotle’s description of the efficient cause in the canonical statement of the four causes in Physics II 3: the efficient cause is ‘the first principle of change or rest, e.g. the man who deliberated is a cause, the father is cause of the child, and generally what produces of what is produced and what changes of what is changed’.⁵⁶ However, it is also clear that Aristotle assigns an important role to the offspring’s own nature in the formation of the senses. It is the foetus’s own nutritive soul that propels the development of the sense-organs. It is important to observe the role that the offspring’s own soul plays in its development, because it is this role that genuinely offsets the efficient causation by which living beings are produced from that of the production of artefacts. Both in artistic and in natural production the movements of the efficient cause contain the form, but in art, Aristotle says in Generation of Animals II 1 (735a 2–4), ‘the movement exists in something else, whereas the movement of nature exists in the product itself ’. The movements of the offspring may derive from those of the semen, but once the embryo has been formed, the movements are internal to the offspring in its development. The point is clarified in Metaphysics IX 7,⁵⁷ where Aristotle discusses the question of when and when not to say that something is potentially such be formed at the same time in the embryo’s development as the nutritive part. Cf. also Gen. an. 740b 29–34: ‘As later on in the case of mature animals and plants this soul causes growth from the nutriment … so also from the beginning does it form the product of nature’ (revised Oxford trans.). ⁵⁴ Cf. Gen. an. 743b 33–744b 11. ⁵⁵ Cf. Gen. an. 742a 19–742b 17 with Part. an. I 1. ⁵⁶ ἔτι ὅθεν ἡ ἀρχὴ τῆς μεταβολῆς ἡ πρώτη ἢ τῆς ἠρεμήσεως, οἷον ὁ βουλεύσας αἴτιος, καὶ ὁ πατὴρ τοῦ τέκνου, καὶ ὅλως τὸ ποιοῦν τοῦ ποιουμένου καὶ τὸ μεταβὰλλον τοῦ μεταβαλλομένου (194b 29–32). ⁵⁷ ‘It is necessary to be clear about when each thing is potentially, and when not, for it is not at any time. For example, is earth potentially a man? Or not, but rather when it has already become a seed, and perhaps not even then? So just as not everything would be healed by the art or medicine or even by luck but there is something which is capable, and this is what is healthy potentially. And the definition of what comes by thought to be actually from being potentially, is that it comes to be when it is wanted and nothing external hinders it. And there in the person being healed, it is when none of the internal conditions prevent it. Similarly, there is potentially a house, if nothing in the thing acted on—i.e. in the matter—prevents it from becoming a house, and if there is nothing which must be added or taken away or changed; this is potentially a house, and the same is true of all other things for which the source of their becoming is external. And in the cases in which the source of the becoming is in the very thing which suffers change, all those things are said to be potentially something else, which will be it of themselves if nothing external hinders them. E.g. the seed is not yet potentially a man; for it must further undergo a change in a foreign medium. But when through its own motive principle it has already got such and such attributes, in this state it is already potentially a man; while in the former state it needs another principle, just as

298

Thomas Johansen

and such. Aristotle is arguing here that we talk differently about a seed being potentially a man from the way in which we talk about the building materials being potentially a house. In the case of the seed, we say that it is potentially a man when nothing external hinders it from growing into a man, whereas in the case of the building materials we say that they are potentially a house when nothing internal prevents them. And the reason for the difference is that we think that the seed has a principle of coming into being in itself. Or rather, the seed has this principle not at first on its own, but when it has been placed in another thing—presumably the female menstrual fluid—and there undergone a further change. Once the seed has changed in the female matter, it has within itself the motive principle, i.e. the efficient cause. From then on, its development into a mature human being is a matter of nothing external interfering with it. In contrast, the building materials have no such inner moving principle: they will require the continued assistance of the builder. Hence their potential is of a different kind from the seed’s: the building materials’ potential to be a house depends on their continuing to present no obstacles to what the builder does to them, whereas the seed’s potential to be a man depends on its surroundings continuing to present no obstacle to its natural motions.⁵⁸ The crucial point for my purposes is that the animal’s own nature, particularly its nutritive soul, is the efficient cause of the development of the second potentiality of perceiving. The father instils the movements of the nutritive soul in the female matter, which brings about the formation of the embryo, but the motions of the embryo’s own nutritive part then bring about the sense-organs and with them the second potentiality of perceiving. When we read therefore in De anima II 5 417b 17 that the father is the efficient cause of the first change (pr¯ote metabol¯e) that leads to the actual of perception, the qualification ‘first’ is significant. The movements of the father’s semen lead to the formation of the nutritive part in the embryo. But the embryonic animal is itself the efficient cause of the subsequent changes that specifically lead to the formation of the sense-organs. It is those changes that lead to the formation of the animal’s second potentiality of perceiving, its ‘firmly fixed dispositional state’ of perceiving. The contribution that the offspring’s nature makes to its own formation sets off the efficient causation of perceiving from the motions of artefacts which, as we saw, required continued external impetus. But it also distinguishes the efficient causation of perceiving from that of the motions of non-living natural beings such as air. In Aristotle’s own example, when the water acquires the second potentiality of moving upwards by turning into air, it does so because something else turns the water into air—say, something earth is not yet potentially a statue, for it must change in order to become bronze’ (Metaph. IX 7 1048b 37–1049a 18). ⁵⁸ Cf. Gen. an. II 1 734b 36–735a 4.

The Soul as an Inner Principle of Change

299

hot makes the water evaporate. The simple bodies do not acquire second potentialities as a result of internally caused changes. Indeed, the lack of an inner differentiation between a mover and a moved within the simple bodies seems firmly to exclude the possibility that the simple bodies could orchestrate their own development. If we consider the second potentiality and actuality of perception as stages of perceiving (both referred to as aisthanesthai), the first change is that brought about by the parent, whilst those leading to the formation of the senseorgans and thus the second potentiality of perceiving are specifically due to the animal’s own nature. The action of the sense object actualizes or completes the process initiated by the father and continued by the animal’s own nature. In this way, it is wrong to call the sense object the efficient cause of perceiving, full stop. For the moving cause of the changes leading to the second potentiality of perceiving, which are the only ordinary changes in perceiving (aisthanesthai), is the animal’s own soul. It may be no accident that Aristotle refers to the sense object at De an. 417b 20 as ‘what produces the activity’ rather than as the ‘moving cause’. For the sense object only brings to the final stage of activity a series of changes in the animal of which the animal has itself, from conception on, been the efficient cause. To conclude, the accounts of nutrition and perception in De anima show in different ways that the activities of living beings depend on environmental factors. Living beings, as Physics VIII 6 argued, are not absolute self-movers. However, the accounts also show how the animal’s own soul brings about the activities of nutrition and perception: in nutrition by assimilating the food to the nourished body before it is incorporated, and in perception by producing before birth all the ordinary changes required for actual perception. The efficient causation of both nutrition and perception is best captured by the interactive picture according to which the external factors act on the animal, but the soul remains the first mover. Aristotle’s emphasis on the dependence of animals on the environment in Physics VIII 6 is understandable given his argument for an absolute unmoved mover. The De anima, however, restores a robust sense of the power that the souls of living beings invest in them to initiate their own changes. REFERENCES Burnyeat, M. F. (2002), ‘De Anima II 5’, Phronesis, 47: 1–64. Furley, D. (1994), ‘Self-Movers’, in Gill and Lennox (1994), 3–14. Gill, M. L. (1994), ‘Aristotle on Self-Motion’, in Gill and Lennox (1994), 15–34. and Lennox, J. G. (1994) (eds.), Self-Motion: From Aristotle to Newton, Princeton. Graham, D. (1999), Aristotle, Physics Book VIII. Oxford. Lennox, J. G. (2001), Aristotle, On the Parts of Animals. Oxford.

300

Thomas Johansen

Morison, B. (2004), ‘Self-motion in Physics VIII’, in A. Laks and M. Rashed (eds.), Aristote et le mouvement des animaux (Lille), 67–79. Sorabji, R. (1992), ‘Intentionality and Physiological Processes: Aristotle’s Theory of Sense-Perception’, in M. C. Nussbaum and A. O. Rorty, (eds.), Essays on Aristotle’s De Anima (Oxford), 195–226.

15 Aristotle’s De Interpretatione 8 is about Ambiguity Susanne Bobzien

My goal in this paper is to show that, contrary to the prevalent view, in his De Interpretatione 8, Aristotle is concerned with homonymy; more precisely, with homonymy of linguistic expressions as it may occur in dialectical argument. The paper has two parts. In the first part, I argue that in Soph. el. 175b 39–176a 5 Aristotle indubitably deals with homonymy in dialectical argument; that De Interpretatione 8 is a parallel to Soph. el. 175b 39–176a 5; that De Interpretatione 8 is concerned with dialectical argument; that, hence, De Interpretatione 8, too, deals with homonymy in dialectical argument. In the second part I discuss objections that have been put forward against the view that De Interpretatione 8 is about homonymy and demonstrate that they do not succeed. 1 . H O M O N Y M Y I N SOPHISTICAL REFUTATIONS 1 7 A N D DE INTERPRETATIONE 8 In chapter 17 of his Sophistici Elenchi (or Sophistical Refutations), Aristotle makes the following remarks about fallacies of homonymy: If nobody ever made two questions into one question, the fallacy based on homonymy and amphiboly would not have come about, but either a refutation or no refutation. For how does asking whether Callias and Themistocles are musical differ from if both, though being different people, shared a single name? For if the name indicated ¹ more than one thing, asked more than one question. Now, if it is not right to ask to be given without qualification one answer to two questions, it is clear that it is not proper to answer without qualification any homonymous . (Arist. Soph. el. 175b 39–176a 5, my italics) Myles Burnyeat has been a source of inspiration for me for many years, both through his boundless expertise and enthusiasm in ancient philosophy and as a friend, and I am pleased to dedicate this paper to him. ¹ I translate δηλόω by ‘indicate’ and σημαίνω by ‘signify’, but I believe that Aristotle uses them interchangeably in the passages I discuss; see below n. 11.

302

Susanne Bobzien

Aristotle here gives some explanation about the ‘fallacy based on homonymy and amphiboly’. In modern terms, these are, roughly, the fallacies that come about as a result of either lexical ambiguity or ambiguity of phrases.² The example Aristotle uses is one of lexical ambiguity only, but what he says may be applied, mutatis mutandis, to structural ambiguities, too. There can be absolutely no doubt that in the passage quoted Aristotle discusses homonymy: he mentions homonymy three times.³ In the passage, Aristotle’s explanation of the fallacy of homonymy works by analogy. He starts with the assumption that someone who asks ‘Are Callias and Themistocles musical?’ puts forward two questions. These are, we can confidently assume, the questions ‘Is Callias musical?’ and ‘Is Themistocles musical?’. So here we have what Aristotle seems to regard as an obvious case in which two questions are asked with one sentence.⁴ And he states that a questioner in a dialectical game who asks ‘to be given without qualification one answer to two questions’, i.e. who demands the answerer to answer without qualification either ‘yes’ or ‘no’ to ‘Are Callias and Themistocles musical?’ does something that is not right. All this Aristotle appears to take as given in the passage quoted. Aristotle then draws the analogy by moving from sentences with two subject expressions to sentences with one, ambiguous, subject expression of the kind that occurs in fallacies of homonymy. Imagine that the questioner asked, ‘Is Callias musical?’, but there are actually two Calliases present, one musical and one not. Call the musical one Calliasm , the unmusical one Calliasn . Then we have a ‘homonymous question’, potentially leading to a fallacy of homonymy.⁵ Aristotle argues that here, too, we have two questions asked with one sentence, since the name (‘Callias’) signifies more than one thing (Soph. el. 176a2). These two questions are ‘Is Callias musical?’ and ‘Is Callias musical?’, where in one question ‘Callias’ signifies Calliasm , and in the other question ‘Callias’ signifies Calliasn . Aristotle draws the conclusion that, hence, in this case, too, the questioner does something wrong, if he demands the answerer to answer without qualification either ‘yes’ or ‘no’ to ‘Is Callias musical?’. Thus, according to Aristotle, if, as in the case of fallacies of homonymy, we have as a premiss or as a conclusion a question sentence that contains an ambiguous term, the questioner has asked more than one question: two questions if the term has two significations, three questions if the term has three significations, etc. ² Sometimes Aristotle seems to use ‘amphiboly’ (ἀμφιβολία) in a wider sense, where it also covers cases in which we have an (obscured) lack of specificity; cf. e.g. Rh. 1407a 33–9, where the Delphic Oracle is unspecific about whose realm Croesus is going to destroy if he crosses the Hades. ³ The third time is at the end of the section, at Soph. el. 176a 15. ⁴ Aristotle has a test for whether what is asked with one question sentence is two questions. If we could give different answers to whether Callias is musical and whether Themistocles is musical, there must be two questions in the one sentence (e.g. Soph. el. 177a 9–15). ⁵ Aristotelian homonymy covers both cases in which one proper name has more than one referent and cases in which one common noun has more than one meaning.

Aristotle’s De Interpretatione

303

Aristotle’s De Interpretatione, chapter 8, is a close parallel to the Sophistici elenchi passage, with the non-essential differences (i) that instead of proper names we seem to have common nouns, and (ii) that the focus is on statements rather than on questions: But if one name is assigned to two things which do not make up one thing, there is not a single affirmation, nor is there a single negation.⁶ Suppose, for example, that someone assigned the name cloak to horse and to human being; then a cloak is white would not be a single affirmation. For to say this is no different from saying a horse and a human being is white, and this is no different from saying a horse is white and a human being is white.⁷ So if these last signify more than one thing and are more than one , clearly the first also signifies either more than one thing or else nothing—for it is not the case that some human being is a horse. (Arist. Int. 8 18a 18–26, my italics)

Again, Aristotle argues by analogy from a sentence with two subject expressions to a sentence with one ambiguous subject expression. Someone states the sentence (S1) ‘A horse and a human being are white’. By stating (S1), that person makes two affirmations: ‘A horse is white’ and ‘A human being is white’. This is so, because (S1) signifies more than one thing: it signifies both that a horse is white and that a human being is white. Hence (S1) is more than one affirmation: it is two affirmations. The case to be explained by analogy is this: It is assumed that someone has determined that the word ‘cloak’ is to mean both horse and human being. Now suppose someone states the sentence (S2) ‘A cloak is white’. Then there are two alternatives. The first is that (S2) signifies more than one thing: it signifies both a horse and a human being. And by stating (S2), the person makes two affirmations: (A1) ‘A cloak is white’ and (A2) ‘A cloak is white’, where in (A1) ‘cloak’ signifies a horse that is white and in (A2) ‘cloak’ signifies a human being that is white. Alternatively, (S2) signifies nothing. For the only other possibility would be that (S2) signified some kind of horse–human compound that is white (a case of ‘two things that make up one thing’, Int. 8 18a 18). But such compounds don’t exist, and the second alternative can hence be dismissed. Thus, if read as a parallel to Soph. el. 175b 39–176a 5, De Interpretatione 8 is about homonymy of expressions, and in the clause ‘one name … given to two things which do not make up one thing’ (Int. 8 18a18) Aristotle talks about lexical ambiguity. Moreover, this interpretation is internally consistent and makes perfect sense of the entire passage. ⁶ I agree with Weidemann (1994: 222), that ‘nor is there a single negation’ (οὐδὲ ἀπόφασις μία) from 18a 21, which is rightly excised by Minio–Paluello, was most probably originally placed after ‘affirmation’ (κατάφασις) at 18a 19. ⁷ The text is ambiguous between (a) ‘ ‘‘A horse is white and a human being is white’’ ’ and (b) ‘ ‘‘A horse is white’’ and ‘‘A human being is white’’ ’.

304

Susanne Bobzien

The interpretation fits the immediate context of the passage from De Interpretatione 8⁸ It is sometimes assumed that De Interpretatione 8 cannot be about homonymy, because in chapters 7, 8, and 9 Aristotle introduces three exceptions to the Semantic Principle of Excluded Middle (SPEM), that of contradictory pairs of statements exactly one is true and one is false, and the first and the third exceptions introduce types of statements that can be differentiated at least in part syntactically (indefinite statements in De Interpretatione 7, and future contingents in De Interpretatione 9), whereas homonymy is a pragmatic feature. It can occur in any type of statement and you can’t find out from looking at the structure of a sentence whether it contains a lexical ambiguity. But Aristotle is fully aware that what he discusses in De Interpretatione 8 is of a different kind than what he discusses in Chapters 7 and 9. The result in Chapter 7 is (i) that for particular, universal, and singular statements, necessarily one of a contradictory pair is true and the other false; but (ii) that this does not hold for contradictory pairs of indefinite statements like ‘(a) human being is white’, ‘(a) human being is not white’; they can both be true (Int. 18a 8–12) at the same time. This result, i.e. (i) and (ii), is based on the assumption that in the pairs of contradictories, exactly one affirmation is contradictorily opposed to exactly one negation (Int. 17b37–40, 18a 8–9: cf. Int. 18a 13–14, 18a 18–19). Call this the singularity requirement for SPEM. At the beginning of De Interpretatione 8 Aristotle explicates the singularity requirement,⁹ and then discusses cases in which this singularity requirement is tacitly violated: that is, cases in which, syntactically, it looks as if we have exactly one (μία) affirmation opposed to exactly one (μία) negation, but nonetheless this is not so. These are (or at least include) cases of homonymy. Thus we can understand the clause that concludes De Interpretatione 8, which follows immediately after the passage quoted above: Hence in these ,¹⁰ too, it is not necessary that the one of a contradictory pair is true and the other false. (Int. 8 18a 26–7) ⁸ This section is boring and may be skipped without much loss. ⁹ The first sentence of Int. 8 does not introduce a new topic, as the medieval chapter division and Ackrill’s translation may make one think. (Ackrill (1963) suppresses the connective particle ‘δὲ’ in 18a 12.) Rather, with ‘but single are those affirmations and negations which … ’ (μία δέ ἐστι κατάφασις καὶ ἀπόφασις ἡ … ) Aristotle picks up ‘a single’ (μία) from the beginning of the previous sentence: ‘a single (μία) affirmation is contradictorily opposed to a single (μία) negation’. Thus all of Int. 8 is, in a way, a gloss on Int. 18a 8–9. (Perhaps the δέ in 18a 12 picks up the μὲν from 18a 8, and we don’t have a μὲν οὖν? I owe this suggestion to Jonathan Barnes.) ¹⁰ What is the antecedent of ‘these’ (which is in the feminine, ταύταις (Int. 8 18a 26) )? The context requires it to pick up ‘affirmations and negations’ from 18a 19. It is sometimes assumed that the immediate antecedent of ταύταις is ‘these’ (αὗται) in Int. 18a 24, since it is the closest feminine ˜ is the antecedent of ταύταις. This plural. But this can’t be. Let us assume by hypothesis that αὑται ˜ raises the question: what is the antecedent of αὑται? The answer depends on how one takes (i) ἔστιν ἵππος λευκὸς καὶ ἔστιν ἄνθρωπος λευκός (18a 23) from the preceding sentence (18a 21–3):

Aristotle’s De Interpretatione

305

Any statement that contains a homonymous subject expression (or predicate expression, for that matter), is a possible exception to SPEM. Take the syntactically contradictory pair (A3) ‘Every cloak is rational’ and (N1) ‘Some cloak is not rational’.¹¹ According to De Interpretatione 8, each is two statements; (A3) is two affirmations, and (N1) is two negations. One of the affirmations is true, one false, and one of the negations is true, one false. It is quite unclear what semantic value that gives to (A3) and (N1), if any, in Aristotle’s eyes. But we can rule out that Aristotle regarded either of them as true. Hence, SPEM doesn’t hold for the syntactically contradictory pair (A3), (N1). However, as statements like (A3) and (N1) violate the singularity restriction, they had effectively already been ruled out as candidates for SPEM in De Interpretatione 7. All Aristotle does in chapter 8 is to make such cases explicit. De Interpretatione 9 begins thus: With regard to what is and what has been it is necessary for the affirmation or the negation to be true or false. And with universals taken universally it is always necessary for one to be true and the other false, and with particulars too, as we have said. (Int. 8 18a 28–31, trans. Ackrill)

If De Interpretatione 8 were on a par with chapters 7 and 9, and had introduced simply another case of exceptions to SPEM on a par with those in chapters 7 ˜ must refer to this utterance together with (ii) ἔστιν (a) If one takes it as one utterance, αὑται ἵππος καὶ ἄνθρωπος λευκός from 18a 22. (b) If one takes it as two utterances, separated by καὶ, αὗται either refers solely to these two utterances (distributively), or to them together with (ii). ˜ refers to the analogon of the analogy Aristotle uses Regardless of whether (a) or (b) is correct, αὑται in Int. 8. This fact can be used to show the absurdity of the hypothesis that the immediate antecedent of ταύταις is αὗται. Aristotle introduced (i) and (ii) to draw an analogy; by contrast, (iii) is the example for the kinds of statement that are the topic of Int. 8, i.e. the statements in which δυεῖν ἓν ὄνομα κεῖται, ἐξ ὧν μή ἐστιν ἕν. And it would be absurd if, instead of drawing a conclusion about what the topic of the passage is, Aristotle drew a conclusion about the analogon he introduced in an example in order to elucidate the topic of the passage. ˜ is hence not the immediate antecedent of ταύταις. What is it, then? Aristotle’s conclusion αὑται in the ὥστε clause should be about the topic of Int. 8, i.e. the statements in which δυεῖν ἓν ὄνομα κεῖται, ἐξ ὧν μή ἐστιν ἕν. Therefore I suggest that Aristotle speaks loosely, and that the immediate antecedent is ἡ πρώτη (18a 25), whose immediate antecedent in turn is the illustrative statement (iii). Aristotle speaks loosely in so far as he uses ‘Hence, in the case of these, too, it is not the case … ’ (ὥστε οὐδ’ ἐν ταύταις) as short for saying ‘Hence, in the case of statements like this one, too, it is not the case … ’. Read in this way, the ταύταις of Int. 8 18b 6 fits perfectly with the rest of the passage Int. 8 18a 18–27. ¹¹ Aristotle’s own choice of ‘A cloak is white’ (if that was his choice) is unfortunate, in so far as it is an indefinite statement, and SPEM does not hold for it anyhow. If we read, instead, ‘Cloak is white’ and ‘Cloak is not white’ (with cloak as a generic), we don’t fare much better. First, Aristotle would have changed how he understands indefinite statements quite suddenly and without warning. Second, both his examples, ‘Human being is white’ and ‘Horse is white’ (Int. 8 18a 23) would be blatantly false, which goes against the almost universal practice in ancient logic of using illustrative statements that are true. The quirky interpretation that Aristotle uses ἱμάτιον as a name (‘Cloaky’) fares best here.

306

Susanne Bobzien

and 9, this sentence would be evidently false. For what Aristotle has shown in chapter 8 implies that even for universals taken universally there are exceptions to SPEM—take the example of (A3) and (N1). However, if, as I suggest, chapter 8 is an extended gloss on ‘μία’ from chapter 7, there is nothing odd with the first sentence of chapter 9. Aristotle simply picks up his thought from before his little detour. Statements with homonymous expressions have been excluded because they violate one of the requirements for SPEM, not because they are exceptions to SPEM. The immediate context of the controversial passage from De Interpretatione 8 hence fits nicely with the assumption that Aristotle discusses linguistic homonymy in that passage.

The differences between Soph. el. 175b 39–176a 5 and Int. 8 18a 18–26 We still need to explain the four differences between the parallel passages from Sophistical Refutations and De Interpretatione, and why they are immaterial for the question as to whether Aristotle deals with linguistic homonymy in De Interpretatione 8.¹²

(i) Proper names versus common nouns In the Sophistici Elenchi passage Aristotle uses proper names as examples for homonymous expressions. In De Interpretatione 8, arguably, Aristotle uses a common noun in his example. Does this give us reason to doubt that De Interpretatione 8 is a parallel to Soph. el. 175b 39–176a 5? It does not. First, although at some point Aristotle becomes aware of the difference, he often lumps proper names and common nouns together in one category: names, ὀνόματα (e.g. Int. 1 and 2). Second, Aristotle’s notion of homonymy equally covers common nouns and proper names. Third, we have examples with common nouns of fallacies of homonymy in the Sophistical Refutations, and nothing in 175b 39–176a 5 suggests that what Aristotle argues there is restricted to those fallacies of homonymy that contain a homonymous proper name.

(ii) ‘are’ (εἰσιν) versus ‘is’ (ἔστιν) In Soph. el., Aristotle uses ‘are’ in the analogon, whereas in De Interpretatione he uses ‘is’:¹³ ¹² I discuss three of these differences in this section. A fourth difference is that in Int. 8 Aristotle restricts consideration to ‘things which do not make up one thing’, but in Soph. el. 17 there is no such clause. I discuss this restriction below together with the prevalent view. A fifth difference is that in the Soph. el. passage Aristotle uses ‘to indicate’ (δηλόω), whereas in the De Interpretatione passage he uses ‘to signify’ (σημαίνω). This difference is not significant, however, as is shown by the fact that in several parallel passages in Soph. el. Aristotle uses ‘to signify’, and in Int. 17a 16 he uses ‘to indicate’. (These passages are quoted below.) ¹³ Here, I disregard distinctions (i) and (iii), as they seem irrelevant to (ii).

Aristotle’s De Interpretatione

307

For how does asking whether Callias and Themistocles are (εἰσιν) musical differ from … (Soph. el. 175b 41–176a 1) For to say this is no different from saying a horse and a human being is (ἔστιν) white, and this is no different from saying a horse is white and a human being is white … (Int. 18a 22–3)

Does this matter? First, there is a natural explanation for this difference from the direction in which Aristotle presents the argument. In the Soph. el. sentence, Aristotle works from ‘a and b are F’ to ‘c is F’, where ‘c’ has two significations. In the De Interpretatione sentence, Aristotle works from ‘c is F’ to ‘a and b is F’. Thus he may have simply opted each time for the minimal change from one sentence to the next, without intending anything different. Second, we have four different syntactic structures, (a) to (d): (a) (b) (c) (d)

εἰσιν x καὶ y F ἔστιν x καὶ y F ἔστιν x F καὶ ἔστιν y F ἔστι z F

x and y are F x and y is F x is F and y is F z is F

(explanandum) (explanandum) (explanandum) (explanans)

In De Interpretatione Aristotle works from (d) via (b) to (c). In Soph. el. he works from (a) directly to (d), although (d) is not explicitly given. Rightly or wrongly, Aristotle assumes that both (a) and (c) are evidently saying two things, and that (b) and (c) are evidently equivalent, and argues from there that the explanans, too, is saying two things, if not evidently so. So although (a), (b), and (c) may all differ in some important respect, this is not germane to the point that Aristotle is making.¹⁴

(iii) Affirmations versus questions In Soph. el. 175b 39–176a 5 Aristotle considers homonymy in questions. This is so, because in the Sophistici Elenchi he discusses fallacies directly as they occur within the dialectical game, where every premiss and conclusion is asked by the questioner, and the answerer is expected to reply.¹⁵ By contrast, in De Interpretatione Aristotle generally considers not questions but declarative statements (λόγοι ἀποφαντικοί, ἀπόφανσεις), and their two basic types, affirmation and negation. Does this mean that in De Interpretatione Aristotle is not discussing dialectic? Certainly not. As C. W. A. Whitaker has shown in great detail, in De Interpretatione Aristotle discusses affirmations and negations with dialectic at least in the back of his mind.¹⁶ Dialectic is not all questions. Rather, dialectic is crucially concerned with truth-evaluable items. For Aristotle, truth-bearers, at least ¹⁴ The context in Soph. el. strongly suggests that Aristotle parses (a) ‘x and y are F’ as ‘x is F and y is F’, or ‘ ‘‘x is F’’ and ‘‘y is F’’ ’, and thus that he takes it to be equivalent to (c). ¹⁵ For dialectic at Aristotle’s time see e.g. Smith 1997, introduction. ¹⁶ Whitaker 1996: passim. However, Whitaker, following Ackrill, believes that the passage quoted does not cover homonyms. I consider his arguments below.

308

Susanne Bobzien

those relevant to dialectic discourse, are linguistic items (Arist. Int. 16a 9–11, 16b 33–17a 3). Thus, statements (ἀπόφανσεις), in particular affirmations and negations, are the entities that are either true or false. In the context of dialectic, affirmations and negations qua truth-bearers enter in at least four ways: First, they are the answers to dialectical questions. For example, if the question is ‘Is animal the genus of human?’, the possible answers are the affirmative statement ‘Animal is the genus of human’ and the negative statement ‘Animal is not the genus of human’. ‘Yes’ and ‘no’ are abbreviations for the affirmative and negative answer respectively.¹⁷ Second, Aristotle also sometimes says that what is said in a dialectical question is ‘true’ or ‘false’.¹⁸ We can assume that what is, in this sense, said in a question could be expressed by the corresponding affirmative statement if the question is in positive form and by the corresponding negative statement if the question is in negative form. Third, we can imagine an entire fallacy explicitly put in declarative sentences rather than questions—for instance, when someone tries to solve it by himself, at his leisure, without being subjected to questions. Aristotle mentions this possibility for example at Soph. el. 177a 6–8. Fourth, the questioner may at any point give a summary of the premisses admitted up to then, and this would be done in the form of affirmative statements rather than questions.¹⁹ Even if Whitaker’s conclusion that De Interpretatione is all about dialectic may be a little too narrow, there can be no doubt that elements from dialectic are sprinkled throughout the work, and that De Interpretatione would have been useful for participants in dialectical games. Thus, we can conclude, in De Interpretatione 8, Aristotle may well discuss cases of homonymy as they would occur in dialectic; and that is first and foremost, in fallacies of homonymy. He holds that in dialectic, (i) someone using a homonymous expression when putting forward a question sentence may ask two questions,²⁰ and (ii) someone using a homonymous expression when stating a declarative sentence may make two statements—either two affirmations or two negations.²¹ He discusses (i) in Sophistici Elenchi 17 and (ii) in De Interpretatione 8. ¹⁷ Cf. e.g. Whitaker 1996: 101. ¹⁸ e.g. Top. VIII 7 160a 25. ¹⁹ We find this frequently in Plato’s dialogues. ²⁰ Cf. also Top. 160a 23–9, quoted below. ²¹ Thus, for Aristotle, one statement is more than one statement (more than one affirmation or more than one negation), and one question is more than one question. How can this be? Statements as well as affirmations and questions are linguistic items; but evidently the one statement that is two statements (two affirmations or two negations) must differ from those two, since it is subject to different rules of individuation from the latter. For the same reason, the one question that is two questions must differ from those two. We could say that by putting forward one question sentence the speaker asks two questions; and by putting forward one declarative sentence the speaker makes two statements (affirmations, negations) or states (affirms, negates) two things. We could think of the sentences as grammatical items, the statements (affirmations and negations) and questions as semantic items (that have a certain grammatical form). This fits the fact that Aristotle mostly talks as if it is the latter who are (or correspond to) truth-bearers. But Aristotle is not consistent on this point, and we are left with a muddle.

Aristotle’s De Interpretatione

309

Homonymy in ordinary discourse versus homonymy in dialectic This result suggests that we cannot infer that what Aristotle says about homonymy of expressions in De Interpretatione 8 is what he thinks about homonymy of expressions generally. (Nor can we rule it out.) In everyday life, people normally use ambiguous expressions in such a way that either (i) the non-linguistic or (ii) the linguistic context disambiguates the expressions. Moreover (iii), metaphors, poetry, and deliberate deception aside, speakers will intend exactly one signification of a homonymous expression. (i) When I state (S3) ‘I’m going to the bank’ and wield a cheque that needs to be deposited, this will make it clear that ‘bank’ refers to a monetary institution. (ii) When I state (S4) ‘The local bank doesn’t sell Euros’, the linguistic context makes it clear that ‘bank’ refers to a monetary institution. (iii) Each time I intend the signification ‘monetary institution’, and presumably the notion of the verge of a river doesn’t even cross my mind. It seems at least possible that Aristotle thought along these lines and would have taken (S3) and (S4) each to be just one affirmation, especially since Aristotle’s logic mostly deals with oral statements as primary truth-bearers.²² However, I am not here concerned with Aristotle’s view on homonymy in ordinary discourse. Regardless of whether he thought that in ordinary discourse context disambiguates, in dialectic the situation is different. Dialectic is a game which the questioner has won when he has led the answerer to contradict themself. And although sophistry is not allowed officially, if the questioner can smuggle it in unnoticed to produce a contradiction, he will, to all intents and purposes, have won the game.²³ Fallacies of homonymy and ambiguity are particularly useful for this purpose. And here we have the—interesting—case that the questioner may put forward a statement and leave it deliberately unclear which, if any, is the intended meaning. Most probably, the questioner has no intention one way or the other, since the response of the answerer may determine which way the argumentation takes. In addition, dialectical discourse differs from ordinary discourse in that both linguistic and non-linguistic context tend to be, by comparison, impoverished. Aristotle himself never comes quite clear about what he takes to be the criterion for when a speaker who uses a homonymous expression in a sentence asks more than one question or makes more than one statement. Mostly speaker intention appears to be irrelevant.²⁴ But there is evidence in Aristotle’s logical writings that ²² Cf. e.g. Int. 1. If, in Int. 8, Aristotle doesn’t discuss homonymy in general, this fact may explain also why in Int. 8 Aristotle does not state that he discusses homonymy. He is not interested in the general phenomenon of homonymy, but in a specific phenomenon that we can observe in certain cases in which homonymous expressions are used, as we encounter it—among other things—in dialectic. ²³ Cf. Arist. Soph. el. 17 175a39–41, where Aristotle notes that we have to beware of seeming to be refuted. ²⁴ There is one passage, Soph. el. 178a 27–8, where ‘listener decoding’ seems to matter to whether or not an answerer has been refuted.

310

Susanne Bobzien

it is a sufficient condition for a case of ambiguity in which someone makes two statements simultaneously by stating one declarative sentence²⁵ when (i) we have an ambiguous expression in the statement and (ii) it is unclear to the listeners which of the two possible significations of the ambiguous expression, if any, is intended by the speaker. Here are some passages from Sophistici Elenchi chapters 17 and 19 and from Topics VIII 7 which support the point that Aristotle believes that in some fallacies of homonymy and amphiboly the questioner simultaneously asks two questions, and that in those cases two things are said simultaneously with the question sentence. In Sophistici Elenchi 19 Aristotle begins with some general remarks about such fallacies: (1) Now, of the refutations that depend upon homonymy and ambiguity some have one of the premise-questions with more than one signified thing … e.g. … in the that the one who knows does not understand one of the premisequestions is ambiguous. (2) And that which is said in two ways is in the one case and in the other it isn’t; and that which is said in two ways signifies something that is and something that is not. … (Soph. el. 177a 9–15)

In (2), Aristotle picks up on what he said at Soph. el. 175b 39–176a 18 (quoted above)—that is, that the questioner asks two questions in one, and that the ambiguous expression has two ‘signified things’ at the same time. The first clause of (2) makes sense only if we assume that with the phrase ‘is in one case and in the other it isn’t’ (ὁτὲ μὲν … ὁτὲ δ’ …). Aristotle intends the two ways in which the thing is said; or more precisely, these two ways as they are simultaneously signified when the premiss question is uttered. For only then is it reasonable to say that one is (true), the other isn’t.²⁶ This is confirmed by the way the passage continues: (3) Whenever ²⁷ lies in the premise-questions, it is not necessary to begin by denying that which is said in two ways; for argument is not for the sake of this, but through this. (4) At the beginning one should reply concerning that which is said in two ways, whether it is a word or a phrase, in this way, that in one sense

²⁵ Or in which someone asks two questions simultaneously by putting forward one question sentence. ²⁶ If Aristotle had meant to use ‘ὁτὲ μὲν … ὁτὲ δ’ … ’ temporally, to be translated as ‘at one time … at another time …’ or similarly, then we would have expected him to say that both are : i.e. at one time, in one context, one of the two things said in the question sentence is (true), at another time, in another context, the other is (true); that is, that Aristotle would have alerted us to the fact that ambiguous expressions have different meanings in different contexts. But Aristotle does not say that. Thus ‘ὁτὲ μὲν … ὁτὲ δ’ …’ must here be used non-temporally to pick out the two things said by the two questions asked with one question sentence; only then does what Aristotle says make sense, fit the context, and comes out true. (Aristotle makes a similar point at Top. 160a 26, using ἐπὶ τὶ μὲν … ἐπὶ τὶ δ’ …; cf. also Arist. Soph. el. 177a 21–2; both passages are quoted below.) ²⁷ λέγεται πολλαχῶς.

Aristotle’s De Interpretatione

311

it is so, and in another not so, (5) for example that speaking of the silent is possible in one sense but not in another. (Soph. el. 177a 18–23)

The premiss here is ‘Speaking of the silent is possible’ (cf. Soph. el. 4 166a 12–14). This is a case where it is unclear to the listener which signification, if any, the questioner is intending, and the context does not disambiguate.²⁸ Hence it falls to the answerer to disambiguate: that is, to say that the premiss has one signification that makes it true and another that makes it false. Thus evidently Aristotle takes the premiss to have two significations at the same time.²⁹ Two short passages from Sophistici Elenchi 17 confirm that Aristotle discusses cases of homonymy and ambiguity in which one premiss has two significations at the same time: Now, if it is not right to ask to be given without qualification one answer to two questions, it is clear that it is not proper to answer without qualification any homonymous . (Soph. el. 176a 3–5)

This passage implies that homonymous questions may have two significations simultaneously. For otherwise one could give one answer without qualification. Now, if one should not give a single answer to two questions, it is evident that in the case of homonyms one should not say ‘yes’ or ‘no’ either; for the one who says has not given an answer, he has merely spoken. (Soph. el. 176a 14–16)

This passage, too, implies that homonymous questions may have two significations simultaneously. For otherwise, by saying ‘yes’ or ‘no’ the answerer would have given an answer. Finally, a passage from the Topics corroborates the same point: If understands the question, but it is said in several ways, then … if what is said is, in one way, false and, in the other, true, he should indicate that it is said in several ways, and that in one it is false, in the other true. For if he makes the distinction only later, it is unclear whether he saw the ambiguity at the beginning. (Top. 160a 23–9)

Here Aristotle discusses a case of fallacy of ambiguity (τὸ ἀμφίβολον) in which only one question is asked, and it is said in several ways (i.e. it is ambiguous), and this one question simultaneously has two significations—or is two questions, as ²⁸ In fact, we may have some sort of ‘second-order’ intention: the questioner may intend that their intention is unclear to the answerer. ²⁹ Cf. Arist. Soph. el. 4 166a 12–14: ‘Moreover, ‘‘speaking of the silent is possible’’; for ‘‘speaking of the silent’’ has a double meaning. It may mean that the speaker is silent and that the things of which he speaks are so.’ The passage in Soph. el. 19 continues: ‘and that in one sense one should do what must be done, but not in another; for what must be done is said in several ways’ (Soph. el. 177a 23–4). The corresponding fallacy is ‘Things that must be are good. Evils must be. Hence evils are good.’ In this case, the immediate linguistic context is likely to make the innocent answerer take the meaning to differ from premiss to premiss. Still, there is no way of guessing the meaning(s) intended by the questioner. The fact that Aristotle wants the answerer to determine explicitly in which way the ambiguous expression is to be understood in each premiss implies that he thought that otherwise each premiss might have two meanings at the same time.

312

Susanne Bobzien

Aristotle would put it. For if it had only one signification when asked, there would be no reason for the answerer to disambiguate the premiss before answering it. These quotes taken together support the following interpretation: Aristotle’s criterion for whether we have two questions or statements in one seems to be independent of speaker intention. This is further confirmed by the fact that Aristotle considers the case that we try to solve a fallacy at our leisure without anybody actually asking the questions (Soph. el. 177a 6–8), and thus without any questioner having any intentions. Rather, at least in the context of dialectic, a sufficient condition for having two questions or statements in one, with two simultaneous significations, seems to be that neither the linguistic nor the nonlinguistic context³⁰ disambiguates.³¹ In actual question-and-answer situations, a consequence of this is that the speaker’s intention is unclear to the listeners.

2 . T H E A RG U M E N TS AG A I N S T T H E V I EW T H AT DE INTERPRETATIONE 8 I S A B O U T H O M O N Y M Y

The prevalent view and the challenge it poses to my interpretation Before I discuss the arguments against the view that De Interpretatione 8 is about homonymy, I should briefly state what the prevailing view is. The most famous proponent of this view is John Ackrill. The most recent defendant is C. W. A. Whitaker. This view assumes that Aristotle wants us to give ‘(a) cloak’ the conjunctive sense ‘(a) horse and (a) human being’, which, however, is not to be understood as equivalent to ‘(a) horse-and-human-being’;³² the latter we might characterize as giving ‘(a) cloak’ a unifying as opposed to a conjunctive sense. ³⁰ Interestingly and oddly, Soph. el. 175b 15–24 suggests that for Aristotle the non-linguistic context of our pointing at an object (a Coriscus) while using a demonstrative phrase (‘this Coriscus’) does not serve to disambiguate the statement ‘This Coriscus is musical’ made in the presence of two Coriscuses. ³¹ Can the facts (i) that at Soph. el. 4 166a 4–5 Aristotle uses ὁτὲ μὲν … ὁτὲ δ’ … to explain double signification, and (ii) that at Soph. el. 4 166a 20–1 he uses ἢ … ἢ … when saying that an expression signifies two things, be used to rebut my claim that for Aristotle, in dialectical contexts, question sentences and declarative sentences containing ambiguous expressions have two significations at the same time? I believe not. Above (in n. 26) I have shown that Aristotle uses ὁτὲ μὲν … ὁτὲ δ’ … non-temporally for double signification. At 166a 4–5 he may do just the same. And as in English the two sentences ‘ ‘‘bank’’ means both ‘‘verge of river’’ and ‘‘financial institution’’ ’ and ‘ ‘‘bank’’ means either ‘‘verge of river’’ or ‘‘financial institution’’ ’ do not usually allow any inference as to whether the speaker assumes that the word has both meanings at the same time, so for Greek sentences with ἢ … ἢ … . (Cf. also in the same passage on ambiguity the use of καὶ … καὶ … at 166a 8 and of καὶ at 166a 14 in sentences that contain double signification.) Alternatively, one has to assume that what Aristotle says about homonymy in Soph. el. 4 at 166a 4–5 and 166a 20–1 is about cases like the fallacy with ‘τὰ δέοντα’, in which each premiss could be interpreted as providing sufficient linguistic context for disambiguation, and the fallacious element comes in only when the questioner pretends to draw the conclusion. But it is quite unclear whether Aristotle took this line. ³² Ackrill 1963: 130–1.

Aristotle’s De Interpretatione

313

Ackrill thinks that on this interpretation, Aristotle’s two ways of understanding ‘A cloak is white’ are ‘reasonable’;³³ but he doesn’t expand on what the conjunctive sense of ‘(a) horse and (a) human being’ is. Here Whitaker chimes in and suggests that ‘a statement about cloak involves making a statement both about man and about horse’, and that this implies ‘that ‘‘cloak’’ is meant as a word that is adopted to stand for two separate things misleadingly taken together as if they were a single unit, such as a horse and a rider’.³⁴ This still leaves me wondering what it is to adopt a word ‘to stand for two separate things misleadingly taken together as if they were a single unit’. Taking a horse and a rider as ‘a single unit’ we may state ‘a horse and rider is a pleasant sight’. But this is not equivalent to ‘a horse is a pleasant sight and a rider is a pleasant sight’. So that is not the sort of unit that Aristotle can have in mind, as in this case the speaker doesn’t make a mistake. Let’s try again: taking a horse and a rider as ‘a single unit’ we may state ‘a horse and rider is jumping over the fence’. But this is not equivalent to ‘a horse is jumping over the fence and a rider is jumping over the fence’. So that is also not the sort of unit Aristotle can have in mind, as in this case, too, the speaker doesn’t make a mistake. Let’s try again: ‘a horse and rider is white’. This could be equivalent to ‘a horse is white’ and ‘a rider is white’. What, then, would be the mistake that Whitaker assumes Aristotle to take the speaker to make? Perhaps that the speaker assumes that there is one specific whiteness in which the horse and rider share (as it would be if someone painted a horse and rider white with body-paint), whereas each of them has their own specific whiteness? I have to admit that both Ackrill’s and Whitaker’s versions of the prevailing view leave me puzzled. No doubt, the prevalent view owes its existence in part to Aristotle’s restrictive clause at the beginning of our passage from De Interpretatione 8: But if one name is assigned to two things which do not make up one thing, there is not a single affirmation, nor is there a single negation. (Int. 18a 18–19)

This clause suggests that it is somehow germane to the cloak example that (a) horse and (a) human being do not make up one thing. Thus I owe the reader an explanation of how this phrase fits in with my own interpretation, and I may as well make good on this point here. My explanation goes as follows. Underlying De Interpretatione 8 is Aristotle’s assumption that there are two ways in which two things can be called by one name.³⁵ Take the sentence ‘A and B are called C’. This can in principle be taken in the following two ways: (i) Here’s one word, ‘C’; it’s a name for what is A, and it’s a name for what is B. (ii) Here’s one word, ‘C’; it’s a name for what is A and B. ³³ Ackrill 1963: 131. ³⁴ Whitaker 1996: 97–8. ³⁵ This is Ackrill’s assumption, too; but I differ in what I think these two ways are. So does Weidemann (1994: 220).

314

Susanne Bobzien

Both (i) and (ii) are covered by a clause like: ‘ ‘‘C’’ is assigned to A and B’. In other words, that clause is ambiguous between (i) and (ii). The same holds for its Greek counterpart. Now, according to Aristotle, (ii) is what is typically used in definitions. For example, ‘human being’ is a name for what is rational and mortal and living; (i) is what we typically find in all other cases.³⁶ Aristotle’s course of argumentation in our passage from De Interpretatione 8 is then as follows: he asks the reader to take a case in which both A and B are called C (εἰ δὲ δυεῖν ἓν ὄνομα κεῖται, Int. 8 18a 18), but to set aside those cases in which A belongs essentially to all Bs or B belongs essentially to all As (ἐξ ὧν μή ἐστιν ἕν, ibid.). For the remaining cases it holds that if we have a statement ‘C is D’, we do not have just one affirmation (οὐ μία κατάφασιν , Int. 8 18a 18–19). Aristotle then provides an example to illustrate those remaining cases (οἷον … δῆλον ὅτι καἰ ἡ πρώτη ἤτοι πολλὰ, Int. 8 18a 19–25, to be interpreted as suggested above, with reading (i) ). At Int. 8 18a 25, Aristotle reverts to reading (ii) and pre-empts a possible objection, which is: ‘but what if you take ‘‘a cloak is white’’ in way (ii), wouldn’t we then have just one affirmation?’ Aristotle’s retort is that if we take his example in way (ii), then we have no affirmation, since the subject term would not refer to anything (ἢ οὒδεν σημαίνει, οὐ γὰρ ἐστιν τὶς ἄνθρωπος ἵππος’—, Int. 8 18a 25–6).³⁷ This is an argument ad hominem; that is, it considers an objection tailored to Aristotle’s example. And as such, it succeeds. Had Aristotle’s example been ‘assign the name ‘‘cloak’’ to human being and walking’, and were we to construct an exact parallel, at Int. 8 18a 25 he would have had to say: in this case, even with reading (ii) we have more than one affirmation, since on reading (ii) (S5) ‘a human and walking thing is white’ is equivalent to (S6) ‘a human being is white and a walking thing is white’. And—if this was Aristotle’s view³⁸—someone saying this would make (at least) two affirmations, with the difference to reading (i), that in this case the two affirmations may apply to the same individual. We know this from De Interpretatione 11, in which sentences like (S6) are discussed. Thus, with Aristotle’s restricting clause ‘which do not make up one thing’ in place, it holds for all remaining cases of one name assigned to two things that, whether you choose reading (i), as Aristotle discusses in Int. 8 18a 19–25, or reading (ii), as Aristotle discusses for ³⁶ One might suggest that no one would ever assign the same name to two different (kinds of) things. Yet, the multiple use of one and the same proper name proves this wrong; and in the case of common nouns, the Stoics, e.g., are a wonderful counter-example, as their philosophy is full of terms which they deliberately use in more than one way. In any event, it is irrelevant to Aristotle’s point whether the one name is assigned to two things or happens to designate two things or to hold of two kinds of thing. ³⁷ My interpretation requires the clause ‘οὐ γὰρ ἐστιν τὶς ἄνθρωπος ἵππος’ to explicate ‘ἢ οὒδεν σημαίνει’ only. ³⁸ Int. 11 20b 18–22 rather suggests that Aristotle would consider (S5) equivalent to (S7) ‘a human being is white and a walking thing is white and the human being and the walking thing are the same thing’ or something along these lines.

Aristotle’s De Interpretatione

315

his particular example in Int. 8 18a 25–6, you never make exactly one statement, and never signify exactly one thing. You either overshoot or undershoot. This is how my interpretation integrates the clause ‘ἐξ ὧν μή ἐστιν ἕν’. I now turn to the arguments in favour of the view that in De Interpretatione 8 Aristotle does not discuss homonymy. I will consider Ackrill’s and Whitaker’s arguments in turn.

Ackrill’s arguments Ackrill nowhere directly argues for his view. He bases it on the assumption that in De Interpretatione 8 Aristotle discusses the same issue that he discusses in chapter 5 and at the beginning of chapter 11.³⁹ However, there is no evidence for this assumption. In chapter 5 Aristotle does not in any way refer to what he says in chapter 8; and neither does he give any indication in chapter 11 that he has discussed before what he discusses there, nor does he give any indication in chapter 8 that he has discussed what he discusses there before in De Interpretatione or that he will discuss it again later in De Interpretatione. Let us briefly see what it is that Aristotle discusses in the presumed parallels in chapters 5 and 11. The relevant sentence in De Interpretatione 5 is this: A piece of statement-making discourse is one, if either it indicates one or it is one by a connective; a piece of statement-making discourse is many, if either it many and not one or it is connective-less. (Int. 5 17a 15–17)⁴⁰

Aristotle distinguishes two criteria for whether a piece of statement-making speech is one or many: First, are one or more things indicated? Second, are the relevant parts of the statement-making discourse connected by a connective or not? The second criterion is fairly straightforward and not relevant here. Cf. (S8) ‘Callias is white, Socrates is not white’ with (S9) ‘Either Callias is white or Socrates is not white’. (Recall that for Aristotle sentences or statements (ἀποφάνσεις) are primarily utterances.⁴¹) (S8) is a piece of statement-making discourse that makes two simple statements. (S9) is a piece of statement-making discourse that makes one compound statement.⁴² The first criterion is the relevant one: Are one or more things indicated? In simple statements, we have one piece of statement-making discourse if only one thing is indicated. We have more than one piece of statement-making discourse if more than one thing is indicated. In De Interpretatione 5, Aristotle says nothing ³⁹ Cf. Ackrill 1963: 126–7, 130–2, 145–6. ⁴⁰ ἐστι δὲ εἷς λόγος ἀποφαντικὸς ἢ ὁ ἓν δηλῶν ἢ ὁ συνδέσμῳ εἷς, πολλοί δὲ οἱ πολλὰ καὶ μὴ ἓν ἢ οἱ ἀσύνδετοι. ⁴¹ Moreover, when it comes to written language, there was no punctuation at Aristotle’s time. ⁴² Cf. Int. 17a 20–3: ‘is compounded of simple statements and is a kind of composite sentence’ (trans. Ackrill).

316

Susanne Bobzien

further about these cases in which more than one thing is indicated; nothing at all. This is not surprising: these cases are not the topic of this chapter. They are only mentioned because they are contrasted with what is its topic: simple pieces of statement-making discourse that are one because they indicate one . Theoretically, there are four possibilities regarding what cases Aristotle has in mind for those cases where more than one thing is indicated, if he has anything specific in mind, that is: 1. He has in mind only what he discusses in chapter 11. 2. He has in mind only what he discusses in chapter 8. 3. He has in mind both what he discusses in chapter 8 and what he discusses in chapter 11, but chapter 8 is not about homonymy. 4. He has in mind both what he discusses in chapter 8 and what he discusses in chapter 11, and these are two different kinds of thing.⁴³ The text of chapter 5 neither favours nor precludes any of these. I believe that Aristotle did not have (3) in mind. The relevant part of chapter 11 is: To affirm or deny one thing of many, or many of one, is not one affirmation or negation unless the many things together make up some one thing. (I do not call them one if there exists one name but there is not some one thing they make up.) For example, man is perhaps an animal and two-footed and tame, yet these do make up some one thing; whereas white and man and walking do not make up one thing. So if (a) someone affirms some one thing of these it is not one affirmation; it is one spoken sound, but more than one affirmation. Similarly, if (b) these are affirmed of one thing, that is more than one affirmation. (Int. 20b 12–22, trans. Ackrill)

The aim of chapters 11 and 12 is to determine when ‘To affirm or deny one thing of many, or many of one, is one affirmation or negation’ and when not. This requires Aristotle to figure out when many things make up one thing and when they don’t. It is the ‘when they don’t’ things that are pertinent to us. Aristotle distinguishes two cases, (a) and (b): An example of (a) would be ‘(a) white walking human is musical’. An example of (b) would be ‘Callias is a white walking human’. The reason we have ‘more than one thing’ turns out to be that ‘white’ and ‘walking’ are accidental to ‘human being’ (cf. Int. 12). The—otherwise important—details are immaterial to my purpose. What matters is this: according to Aristotle, both in (a) and in (b) we have grammatically one sentence but semantically three affirmations. It is reasonable to assume that both (a) and (b) give us ‘a piece of statementmaking discourse’ that is ‘many’ because it ‘ many ’ of the kind mentioned in chapter 5. But there is nothing in the passage in chapter 11 ⁴³ This doesn’t preclude that each is a special kind of what Aristotle would call ‘double question’ when it is put in the form of a ‘yes’–‘no’ question. Cf. Soph. el. 17, where Aristotle classifies the fallacy of homonymy as a kind of the fallacy of the double question.

Aristotle’s De Interpretatione

317

that suggests that Aristotle here discusses something he has already discussed before in chapter 8, or that implies that he thinks that he has already discussed it anywhere in De Interpretatione.⁴⁴ There is only one sentence in chapter 11 that may harp back to chapter 8: ‘I do not call one if there exists one name but there is not some one thing they make up.’ In chapter 8 Aristotle excluded from his discussion the cases in which two or more things that share one name make up one thing. Now, in chapter 11, what was excluded from discussion—namely, the case of many things that make up one thing—is the topic. And in the case where essential properties are predicated, like ‘Human beings are two-footed, tame, living beings’, the shared name (‘human being’) features as subject expression, and the things that make up one thing (two-footed, tame, living being) are predicated. The function of the ‘I do not call …’ sentence is to prevent the reader from wrongly assuming that those things which share one name but which do not make up one thing (be that in reading (i) or (ii) above) are under discussion here, too.⁴⁵ These were exactly the things Aristotle discussed in chapter 8. Thus, far from discussing the same topic in chapters 8 and 11, careful reading shows that Aristotle discusses different topics, and takes care each time to alert the reader to what he is not discussing in the respective chapter. In short, the results of scrutinizing the relevant passages in chapter 5 and 11 are: •

•

•

The passage in chapter 5 leaves it entirely open what the cases are in which even a simple piece of statement-making discourse indicates more than one thing. The passage in chapter 11 discusses cases in which one sentence has multiple subject or predicate expressions, and determines when these make up more than one statement and why. These cases are clearly different from those discussed in chapter 8, whichever interpretation of chapter 8 one prefers. chapter 5 in no way refers to chapter 8. The passage in chapter 11 mentions cases as discussed in chapter 8 only in order to preclude them from consideration in chapter 11.

⁴⁴ Aristotle does say that he has discussed the issue in the Topics, though: the passage quoted above continues: ‘So if a dialectical question demands as answer either the statement proposed or one side of a contradiction (the statement in fact being a side of one contradiction), there could not be one answer in these cases. For the question itself would not be one question, even if true. These matters have been discussed in the Topics’ (Int. 20b 22–26, trans. Ackrill). The passage Aristotle refers to is most probably Soph. el. 175b 41–176a 1 and 176a 6–14; the Sophistici Elenchi were originally part of the Topics. ⁴⁵ I put the ‘I do not call …’ sentence in brackets to indicate that it is not part of Aristotle’s main thought, but simply serves to exclude certain cases from discussion. That this is indeed the case is easily seen if one asks oneself what the example introduced by ‘for example’ (οἷον) is meant to illustrate. It can only illustrate the case Aristotle puts in the ‘unless’ (ἐὰν μὴ) clause (20b 13–14), that many things together make up some one thing. The ‘I do not call them one …’ sentence (20b 15–16) is a side remark on the ‘unless’ clause, to prevent the reader from wrongly taking certain cases as belonging to the class Aristotle introduces in the ‘unless’ clause.

318

Susanne Bobzien

Thus Ackrill does not succeed in backing up his claim that in chapter 8 ‘Aristotle is not discussing ambiguity of names’ (1963: 131).⁴⁶ The merits of his interpretation of chapter 8 thus have to be judged entirely by how well it does justice to the text of chapter 8 itself. Ackrill himself argues that ‘Aristotle fails to show’ what he intends to show in chapter 8. That is, if interpreted in the way Ackrill does, Aristotle fails to show what, on that interpretation, he is taken to intend to show. This should make Ackrill’s interpretation less plausible than interpretations like the one suggested above. Moreover, the one suggested above has the advantage of providing a clear parallel to the Topics, whereas there seems to be no parallel anywhere in Aristotle’s writings for Ackrill’s first reading of ‘(a) cloak is white’.

Whitaker’s argument from the meaning of ‘ambiguity’ Unlike Ackrill, Whitaker argues explicitly against the possibility that De Interpretatione 8 is about homonymy. He outlines three interpretations for Chapter 8: Is ‘cloak’ supposed to be ambiguous between the two senses, so that in certain contexts it would mean ‘man’ and in others ‘horse’? Or is ‘man and horse’ supposed to denote a compound entity, like a horse and rider, which might be talked about as a unit, without counting as one in the proper sense? Or, again, might ‘cloak’ be intended as the genus of man and horse? (Whitaker 1996: 96, my italics)

He argues against the first alternative thus: Aristotle cannot mean cloak to be an ambiguous word, which might signify either man or horse. Had he meant this, 1 [‘cloak is pale’] would not be equivalent to 3 [‘Horse is pale and man is pale’, as Aristotle argues]; instead, ‘cloak is pale’ would either mean ‘horse is pale’ or ‘man is pale’, depending on which sense of ‘cloak’ was intended, just as ‘bank’ does not mean ‘a financial institution and a hill’, but one or the other depending on context. We see instead that a statement about cloak involves making a statement both about man and about horse, not an ambiguous claim about one or the other. (Whitaker 1996: 97, my italics)

Whitaker’s argument against Chapter 8 being about ambiguity fails, because his notion of ambiguity is too restricted, both from a contemporary and from Aristotle’s point of view. Whitaker assumes that if an expression used in a statement is ambiguous (has more than one meaning), it means exactly one thing in that statement. As to which meaning it has, Whitaker first says that this is determined by the context, then that it is determined by the intention of the speaker, and then again that it is determined by the context. Perhaps ⁴⁶ There is also a whiff of circularity in the way Ackrill proceeds. In his commentary on chapter 8 he refers the reader to his commentary on chapter 5 (and chapter 11), but in his commentary on chapter 5 he simply says it’s clear from chapter 8 (and chapter 11) that Aristotle is not speaking of ambiguity. In his commentary on chapter 11 all he says about chapter 5 and chapter 8 is that ‘the question what constitutes a single affirmation or negation has already been discussed’ there.

Aristotle’s De Interpretatione

319

he means that the determining element of the context is the intention of the speaker. Now, it is true that in most cases in which an ambiguous expression is used in ordinary discourse, both the speaker intends exactly one meaning and the context disambiguates. But the fact that this is true does not make speaker intention and contextual disambiguation the same thing. Nor does it make either speaker intention or contextual disambiguation a defining element of a statement containing an ambiguous expression. There are certain situations in which a speaker may utter a sentence containing an ambiguous expression, in which (i) the speaker does not intend either meaning and/or (ii) the context does not disambiguate. These include situations of dialectical discourse.⁴⁷ As we have seen above, it is Aristotle’s view that in situations of dialectical discourse, statements containing an ambiguous expression may preserve both meanings of the expression at the same time, and that this is standardly so in fallacies of homonymy. Now, the De Interpretatione is concerned with dialectical discourse.⁴⁸ This is the main thesis of Whitaker, which he argues in his book passim. That is, De Interpretatione is precisely about situations in which, when ambiguous expressions are used, the speaker may not intend exactly one signification, and in which there may be no disambiguating context. Thus Whitaker is not successful in his argument against the view that De Interpretatione 8 is about ambiguity. Rather, he himself provides us with the perfect reason why Aristotle is in fact discussing ambiguity here: the De Interpretatione is—among other things—concerned with dialectical discourse, and it is in dialectical discourse that we are in constant danger that a sentence containing an ambiguous expression may be stated without the context disambiguating it, and hence where the stated sentence involves two statements: in Aristotle’s example, one about a horse and one about a human being. ⁴⁷ Whitaker himself admits as much where he discusses homonymy in Soph. el. and describes Aristotle’s Coriscus example thus: ‘ ‘‘Coriscus’’, which stands for two men of the same name (175b 15ff.) … ‘‘Coriscus’’ is … a name which might apply to either one of the two namesakes, although it is left unclear which’ (1996: 103, my italics). If it is left unclear which of the two Coriscuses the name applies to, then either speaker intention does not disambiguate or context does not disambiguate, or both. ⁴⁸ Whitaker claims that De Interpretatione was ‘meant to provide theoretical underpinning for dialectic, and so should be read closely with the Topics and Sophistici Elenchi, rather than with the Categories and Prior Analytics’ (1996: 2) and that it ‘does not take as its subject propositions, seen as the components of the syllogism, but rather contradictory pairs, which are central to the workings of dialectic’ (ibid.). This seems to me to be a false contrast entirely. In Aristotle’s syllogistic, it is important in many ways to know what the contradictory of a proposition is: examples are Aristotle’s ‘rejection proofs’ (I take the term from Smith (1989: p. xxii) and deductions through impossibility, both central to Aristotle’s syllogistic; moreover, contradictories feature in syllogisms from a hypothesis generally, and the reader’s knowledge of what the contradictory, or what the contrary, of a proposition is presupposed in the Analytica priora repeatedly. I also believe that Whitaker’s claim that the topic of De Interpretatione is contradictory pairs is too narrow.

320

Susanne Bobzien

Whitaker’s argument from the wider context in De Interpretatione Whitaker attempts a second argument in favour of his view that chapter 8 is not about ambiguity due to the wider context. Whitaker argues (1996: 104–5) that in chapter 5 Aristotle restricts the scope of the discussion to simple assertions, ruling out compound assertions; and that in chapter 6 he restricts the scope further by ruling out statements which assert and deny homonymously. ‘So both obviously compound assertions and those in which homonymous terms appear have been excluded from the discussion in chapters 5 and 6’ (1996: 105). From this Whitaker infers that since ambiguity has already been excluded before chapter 8, in this chapter Aristotle does not discuss ambiguity (cf. 1996: 105 and again 107). However, Whitaker’s claims are not borne out by the text. The topic of chapter 5 is ‘simple single-statement-making sentences’ (using Ackrill’s terminology). Accordingly, in chapter 5 Aristotle works toward an account of simple single-statement-making sentences. To that end, he contrasts them both with compound single-statement-making sentences and with simple non-single-statement-making sentences. He does not discuss compound singlestatement-making sentences as such anywhere in the De Interpretatione. But he does discuss two types of simple non-single-statement-making sentences, one in chapter 8 and one in chapter 11. Thus the fact that in chapter 5 Aristotle contrasts ‘simple single-statement-making sentences’ with two other types of sentences does not mean that he restricts the discussion to the former in the sense that he will not later discuss the latter. The topic of chapter 6 is contradictory pairs (ἀντιφάσεις). We find the passage in which homonymy is mentioned at the end of the chapter: Let us call an affirmation and a negation which are opposite a contradiction. I speak of statements as opposite when they affirm and deny the same thing of the same thing—not homonymously, nor whatever other such things that we add to counter the troublesome objections of the sophists. (Int. 17a 33–7, trans. Ackrill)

This passage needs careful reading. In the sentence in which the word ‘homonymously’ occurs, Aristotle explains what he means by ‘opposite’ (ἀντικείμενον). He explains what he means by ‘opposite’, because he uses the expression in his definition of contradictory pair. Thus, by mentioning homonymy, he is not restricting the topic of the discussion of the De Interpretatione at all. Rather, he is explaining how we are to understand a term he uses in the definition of contradictory pair. He glosses his phrase ‘affirm and deny the same thing of the same thing’ from 17a 35: for there to be true opposites, it is not enough that we have two sentences with the same subject and predicate expression, like ‘The bank is over there’ and ‘The bank is not over there’, but the expression (here ‘bank’) must be used non-homonymously. Thus we have no reason to assume that by the end of chapter 6 ‘homonymy, ambiguity, and other sophistic tricks [have been] excluded [from the discussion]

Aristotle’s De Interpretatione

321

as impermissible’ (1996: 107) and hence cannot be discussed in Chapter 8. Thus Whitaker’s second argument against the view that in Chapter 8 Aristotle discusses ambiguity (homonymy) fails, too.

C O N C LU S I O N I conclude that it is time that we emancipate ourselves from the bonds of Ackrill’s authoritative interpretation of De Interpretatione 8 and see it afresh in the light of the result of Whitaker’s contribution to our understanding of Aristotle’s work. De Interpretatione is—among other things—written with a view to dialectical argument; and De Interpretatione 8 is about homonymy of linguistic expressions as it may occur in dialectical argument.⁴⁹ REFERENCES Ackrill, John (1963) (trans. and comm.), Aristotle’s Categories and De Interpretatione. Oxford. Smith, Robin (1989) (trans. and comm.), Aristotle’s Prior Analytics. Indianapolis. (1997) (trans.), Aristotle Topics Books I, VIII, and Selections. Oxford. Weidemann, Hermann (1994) (trans. and comm.), Aristoteles, Peri Hermeneias. Berlin. Whitaker, C. W. A. (1996), Aristotle’s De Interpretatione: Contradiction and Dialectic. Oxford. ⁴⁹ I am most grateful to Jonathan Barnes for his incisive comments on a draft version of this paper.

16 Sextan Scepticism Jonathan Barnes

When you are investigating any matter it probably comes about either that you make a discovery, or that you deny discovery and agree that the matter is inapprehensible, or that you persist in the investigation. No doubt that is why in the case of philosophical investigations some have said that they have discovered the truth, some have asserted that it is not possible for the truth to be apprehended, and some are still investigating. Those who are called dogmatic, in the narrow sense of the word, think they have discovered the truth—Aristotle, for example, and Epicurus, and the Stoics, and some others. Clitomachus and Carneades and other Academics have asserted that things are inapprehensible. The inquirers are investigating. So it is reasonably thought that the highest kinds of philosophy are three—the dogmatic, the Academic, and the inquisitive. (Sext. Emp. Pyr. I 1–4)¹

In this opening paragraph of his Outlines of Pyrrhonism Sextus Empiricus suggests a general thesis about the probable progress of any investigation or inquiry; he connects the thesis to the particular case of philosophical inquiry; and he observes that there therefore seem to be three highest kinds or most general forms of philosophy—one of which is the sceptical philosophy. The general thesis about investigation intimates that any inquiry has one of three outcomes: you may find what you’re looking for, you may declare that it’s unfindable, or you may keep on looking. But that can’t be quite right—if only because the third of the items is not an outcome or result of inquiry. Well, no doubt Sextus means that at any given point after the beginning of an investigation, things will be in one or another of three states of play. What exactly are the states of play? In the second case, the inquirer denies that he has discovered anything and admits that the item he is after is inapprehensible: ¹ τοῖς ζητοῦσί τι πρᾶγμα ἢ εὕρεσιν ἐπακολουθεῖν εἰκὸς ἢ ἄρνησιν εὑρέσεως καὶ ἀκαταληψίας ὁμολογίαν ἢ ἐπιμονὴν ζητήσεως. διόπερ ἴσως καὶ ἐπὶ τῶν κατὰ φιλοσοφίαν ζητουμένων οἱ μὲν εὑρηκέναι τὸ ἀληθὲς ἔφασαν, οἱ δ᾿ ἀπεφήναντο μὴ δυνατὸν εἶναι τοῦτο καταληφθῆναι, οἱ δὲ ἔτι ζητοῦσιν. καὶ εὑρηκέναι μὲν δοκοῦσιν οἱ ἰδίως καλούμενοι δογματικοί, οἷον οἱ περὶ ᾿Αριστοτέλην καὶ ᾿Επίκουρον καὶ τοὺς Στωϊκοὺς καὶ ἄλλοι τινές, ὡς δὲ περὶ ἀκαταλήπτων ἀπεφήναντο οἱ περὶ Κλειτόμαχον καὶ Καρνεάδην καὶ ἄλλοι ᾿Ακαδημαϊκοί, ζητοῦσι δὲ οἱ σκεπτικοί. ὅθεν εὐλόγως δοκοῦσιν αἱ ἀνωτάτω φιλοσοφίαι τρεῖς εἶναι, δογματικὴ ᾿Ακαδημαϊκὴ σκεπτική.

Sextan Scepticism

323

the ‘and’ is epexegetical, and Sextus means that an inquirer may come to think that what he is after can’t be found. So the second state of play is characterized in terms of how the investigator sees things. The first state of play is called a discovery; and that is a matter of how things are, not of how the investigator takes them to be. Thus the first two states of play, as Sextus describes them, are heterogeneous. They need to be homogenized; and what Sextus says about the special case of philosophical inquiry makes it clear that the states of play are determined by how the investigator sees things rather than by how things actually are. So the general thesis with which Sextus starts the Outlines might be improved into the following claim: At any given point after the start of any investigation, the inquirer will find himself in one of the following states: (1) he thinks he has found what he was looking for; (2) he thinks that what he was looking for can’t be found; (3) he continues to look. The three states are plainly taken to be mutually exclusive. (And otherwise they could not be used to characterize three different philosophies.) But neither the first state nor the second seems to exclude the third: often enough, people go on looking for things after they think they’ve found them or although they think they can’t find them—and it is not evidently irrational so to behave. Frobisher was looking for the North-West Passage. ‘Have we found it, Cap’n?’, asked an excited midshipman. One possible reply: ‘Yes, my boy, I think we have, I think we have—but I’m taking one more sighting with the sextant to be sure.’ Another possible answer: ‘No, not a glimpse. What’s more, I doubt now that we’ll ever find it. Yet we must keep on looking—in any case, what else is there to do among these d****d ice-packs?’. If those replies are not incoherent, then you can look for something both when you think you’ve found it and also when you think it can’t be found. Belief and conviction, it is generally supposed, come in degrees, and perhaps in the sort of cases to which I have just gestured the belief is less than 100◦ proof. True, I believe that I’ve got it; but I’m not perfectly sure—and that’s why I keep on looking. True, the chase seems hopeless; but maybe there’s just a glimmer of a chance—so let’s carry on. If that is right, then the third of Sextus’ states of play is indeed compatible with the first and with the second—but only if the first and the second involve tentative beliefs. Or can you continue to look for something even when you’re perfectly certain it can’t be found? Convinced that his quest is hopeless, Frobisher may surely continue to inch westward, to box his compass, to splice his main brace—in short, to do everything he was doing when he was looking for the North-West Passage. But was he still looking for it? No—he was going through the motions. For if you are really looking for something, then you can’t believe firmly that it can’t be found.

324

Jonathan Barnes

Again, can you continue to look for something when you’re quite sure you’ve found it? Knowing that he’s actually found his Passage, Frobisher may carry on sailing: he may want to find a shorter route, or a more southerly route, or simply another route—or the same route a second time. Indeed he might. But in such cases, whether he’s looking for something else or for the same thing again, he’s not doing just what he was doing before: namely, trying to discover the North-West Passage. If such considerations hold good, then it becomes tempting to reformulate Sextus’ general thesis about investigation, thus: At any given point after the start of any investigation, the inquirer will find himself in one of the following states: (1) he is sure he has found what he was looking for; (2) he is sure that what he was looking for can’t be found; (3) he continues to look. Is the thesis true? Surely not. For there are many more than the three possibilities which it sets down. Having started an investigation, you might shelve it, or be distracted from it, or be bored by it, or forget it, … or die on the job. Sextus ignores such mundane happenings. True, his text says that one or other of the three states ‘probably’ comes about; and the ‘probably’ might in principle hint at possibilities unsaid. But it does not do so here. Rather, it is the first manifestation in the Outlines of Sextus’ pedantic determination to abstain from making, or from seeming to make, any positive and unqualified claim about how things actually are. Perhaps that does not matter. Suppose that the general thesis is modified to read ‘the inquirer will find himself in one of a number of states, among them the following: … ’. The thesis will then be true; and it will also perform its part in the economy of the Outlines —provided that the three states which it names are the only ones which bear upon the classification of philosophies. Sextus’ general thesis about investigation applies trivially to the particular case of philosophical investigation: ‘In every inquiry things are thus-and-so—hence in any philosophical inquiry things are thus-and-so.’ But that trifle is not what Sextus has on his mind: he doesn’t say that some philosophical inquiries have run one way and some another—he says that some philosophical inquirers have run one way and some another, and that the groups of philosophical inquirers then determine the three most general forms of philosophy. According to Sextus, you are a dogmatic philosopher if you claim to have reached the truth; and it is easy to think that he means that dogmatists are philosophers who claim to have successfully completed every inquiry they have undertaken. In the same way, Academics claim that none of the inquiries they have undertaken can reach a result, while the inquisitors are still carrying on any inquiries which they have got themselves into. But if that is so, then Galen, for example, will have no philosophy at all; for he thought that some philosophical

Sextan Scepticism

325

questions had been decided, that some were undecidable, and that others were still sub judice. Most reasonable philosophers are, in that respect, Galenic, so that most philosophers will do none of the three forms of philosophy which Sextus distinguishes. That consequence is something we might live with. But I doubt if Sextus would have stomached it—or had ever imagined it. He surely supposes that every philosopher must belong to one or other of the most general philosophies. Elsewhere Sextus remarks that someone who dogmatizes about a single item, either preferring any one impression to another in terms of reliability and unreliability or else making an affirmation about some unclear item—such a man is thereby of the dogmatic character. (Pyr. I 223)²

In other words, you are a dogmatic if you think you have hit upon even one philosophical truth. That suggests that Pyr. I 2 means to say this: you are dogmatic if you claim to have hit upon one or more philosophical truth—so that Galen was dogmatic. If the three forms of philosophy are to be mutually exclusive, then we had better add this: you are Academic if you do not claim that you have hit on any philosophical truths and do claim that in at least one case philosophical truth cannot be attained; and you are inquisitive if, making no claims at all, you are still pursuing all your philosophical inquiries. The general thesis on investigation does not entail that division of philosophy; but it makes it possible. Who are the inquisitors? The Greek word I have guilelessly translated as ‘inquisitive’ is ‘σκεπτικός’. The translation is correct—but the word is usually transliterated as ‘sceptical’. For the inquisitors are Sextus’ sect, they are the Pyrrhonian sceptics. So a sceptic—so far as the opening paragraph of the Outlines defines him—is first and foremost a philosopher: he is in the same line of business as a Stoic or an Aristotelian. A Stoic or an Aristotelian is defined by his views on certain philosophical issues. And although the domain of philosophy is broad, it is not universal: a philosopher will not pontificate, save incidentally, on the affairs of the special sciences—mathematics and medicine, astronomy and agriculture; nor will he opine, save incidentally, on the ten thousand humdrum questions which challenge the ordinary course of life. So too, then, with the sceptic: he exercises his talents in the philosophical domain, and his scepticism is an attitude toward philosophical questions. It follows that—for all the opening paragraph of the Outlines says—a sceptic may be stuffed with beliefs. He may be an opinionated bigot on every topic under the moon—provided only that it is not a philosophical topic. What’s ² ὁ γὰρ περὶ ἑνὸς δογματίζων ἢ προκρίνων φαντασίαν φαντασίας ὅλως κατὰ πίστιν ἢ ἀπιστίαν ἢ ἀποφαινόμενος περί τινος τῶν ἀδήλων τοῦ δογματικοῦ γίνεται χαρακτῆρος.

326

Jonathan Barnes

more, a sceptic may have philosophical beliefs. He continues to inquire only where he has once started to inquire; and there seems to be no reason why he should not have beliefs on matters into which he has never thought to inquire. Suppose that his Platonist father once assured him that numbers were substances, that he was enchanted by the notion, and that he has never thought to question that particular piece of paternal wisdom. In that case—for all that the opening paragraph of the Outlines says—he may be a sceptic while still believing, and fervently, that numbers are substances. Finally, a sceptic may have beliefs on the very subjects which he is still investigating. For investigation does not exclude belief. He may, for example, be pretty strongly inclined to think that his soul is mortal, even though he realizes that the discussion isn’t yet at an end. Such beliefs cannot be firm convictions of dogmatic strength; but they may be just as strong as most of our ordinary beliefs are or ought to be. There is a contemporary discussion, in which Myles Burnyeat has been a protagonist, on the question of whether Sextus was a radical and rustic sceptic or a moderate and urbane sceptic. The discussion concerns two distinct issues, which are sometimes run together. First, there is what may be called the depth of a scepticism. A sceptic abjures some ordinary epistemic attitudes. Perhaps he abjures sure scientific knowledge, or knowledge in general, or all stout convictions, or every belief, or any and every inclination to accept. The more he abjures, the deeper his scepticism—and the more rustic. Secondly, there is breadth. A sceptic abjures attitudes towards a certain domain of inquiry. The domain may be restricted—to philosophy, say, or to the theoretical sciences, or to any scientific proposition; and the domain may be universal and without any restriction. The larger the domain, the broader the scepticism—and the more rustic. Many—perhaps most—of his readers have taken Sextus to be radical or rustic in both dimensions: he renounces every inclination to accept, and he renounces it in every and any domain. But such a scepticism is often supposed to be philosophically absurd, or—worse—philosophically uninteresting; and in point of fact there are several passages in Sextus’ writings which appear to resist rustic interpretation. Hence different scholars have urged, each in his own way, that Sextus tends rather toward urbanity—that his scepticism restricts itself to the domain of the sciences, or that it renounces active assertion but permits passive belief. The matter cannot be settled by careful scrutiny of the pertinent passages. For the passages tell in different directions, and the differences cannot be explained away or airbrushed out: they are collaboratively incoherent. So there is a historical and philological question: how best is the incoherence accounted for? And there is also a historical and philosophical question: If a coherent scepticism is to be winkled out of the texts—a scepticism which Sextus does not embrace but which he could have embraced and would have been best advised to embrace—then what is it?

Sextan Scepticism

327

I call this notional scepticism, which is to be discovered or invented, ‘Sextan scepticism’. One question about Sextan scepticism is this: In what ways and to what extent is it rustic or urbane? The opening paragraph of the Outlines intimates an urbane scepticism—indeed, it intimates a scepticism more urbane than modern scholarship has dared to imagine. It is an intimation, not an announcement; and in any event opening paragraphs are not obliged to be clear and precise. But an opening paragraph is a starting point: Why not accept the suggestion made by the beginning of the Outlines? Why not use it as a criterion for the interpretation of the rest of the work—and of the ancient Pyrrhonism? The intimations of urbanity derive, in part, from Sextus’ observation that ‘the inquisitives are investigating’ (Pyr. I 3). The observation is unambiguously present in the text; and it is implicit in other passages of the Outlines —for example, when Sextus remarks that certain remarks are made ‘lest the inquisitives, seduced by the dogmatists, abandon the investigation’. (Pyr. I 205).³ Nonetheless, it can be shown that real Sextan sceptics do not investigate. In an early paragraph of the Outlines Sextus mentions and explains some of the names which scepticism carries: The inquisitive persuasion is also called investigative, from what it does when it investigates and inquires; and suspensive, from what happens to the inquirer after the investigation; and dubitative, either from the fact that it doubts and investigates everything (as some say) or else from the fact that it is incapable of assent or denial; and Pyrrhonian, from the fact that Pyrrho appears to us to have addressed himself to inquiry in a manner more solid and more conspicuous than his predecessors. (Pyr. I 7)⁴

The first two names, ‘investigative’ and ‘suspensive’, form a pair, the one alluding to the active aspect of scepticism (to an ἐνέργεια) and the other to a passive aspect (to a πάθος). But they are badly matched. For whereas the name ‘investigative’ recalls the observation that the sceptic’s researches are never over, the name ‘suspensive’ implies that his investigations are finished: the suspension of judgement which explains the name ‘suspensive’ comes about ‘after the investigation’, and ‘after the investigation’ does not mean ‘after a certain amount of investigating has been done’ but ‘after the investigation is over’. So there is a conflict between Pyr. I 7 and I 3—and a conflict within I 7 in so far as the name ‘investigative’ harks back to I 3. Some have sought to deny the conflict by invoking an ambiguity. The Greek verb ‘ζητεῖν’, they urge, sometimes means not ‘investigate’ but rather ‘doubt’ or ³ μή πως ὑπὸ τοῦ δογματικοῦ παρακρουσθεὶς ἀπείπῃ τὴν ζήτησιν. ⁴ ἡ σκεπτικὴ τοίνυν ἀγωγὴ καλεῖται μὲν καὶ ζητητικὴ ἀπὸ ἐνεργείας τῆς κατὰ τὸ ζητεῖν καὶ σκέπτεσθαι, καὶ ἐφεκτικὴ ἀπὸ τοῦ μετὰ τὴν ζήτητιν περὶ τὸν σκεπτόμενον γινομένου πάθους, καὶ ἀπορητικὴ ἤτοι ἀπὸ τοῦ περὶ παντὸς ἀπορεῖν καὶ ζητεῖν, ὡς ἔνιοί φασιν, ἢ ἀπὸ τοῦ ἀμηχανεῖν πρὸς συγκατάθεσιν ἢ ἄρνησιν, καὶ Πυρρώνειος ἀπὸ τοῦ φαίνεσθαι ἡμῖν τὸν Πύρρωνα σωματικώτερον καὶ ἐπιφανέστερον τῶν πρὸ αὐτοῦ προσεληλυθέναι τῇ σκέψει.

328

Jonathan Barnes

‘be perplexed’. Indeed, it has that sense on one of its occurrences in Pyr. I 7. For the third name for scepticism is ‘dubitative’, and Sextus explains that the sceptical persuasion is called dubitative ‘from the fact that it ἀπορεῖ and ζητεῖ everything’. It makes no sense to say that scepticism is called dubitative from the fact that it investigates—rather, the two verbs ‘ἀπορεῖν’ and ‘ζητεῖν’ are here used as synonyms, and they mean ‘be stymied’. So let us construe ‘ζητεῖν’ in Pyr. I 3 in that sense: when Sextus says that the sceptics ζητοῦσι, he does not mean that they are still investigating—he means that they are still perplexed. On the other hand, in Pyr. I 7 the phrase ‘after the ζήτησις’ does not mean ‘after the puzzlement is over’ but ‘after the investigation is over’. A Sextan sceptic is not perpetually investigating: he is perpetually puzzled, before and during and after his investigations. That seems to me to be a true account of Sextan scepticism. But it is a wholly implausible interpretation of Sextus’ text; for in the opening paragraph of the Outlines ‘ζητεῖν’ means ‘investigate’ rather than ‘be puzzled’. In that case, there is a conflict between Pyr. I 3 and I 7. The conflict must be resolved in favour of I 7, if only for the following reason. The starting point and cause of all philosophical investigation, according to Sextus, is a certain perturbation of mind: we are disturbed because we can’t decide between conflicting hypotheses or opinions; and so we undertake an investigation in order to allay the perturbation and reach a state of tranquillity. (So Pyr. I 12.) It is a presupposition of the investigation that once we have discovered the truth, we shall no longer be perturbed. But what if we don’t discover the truth? Surely we shall remain in the horrid state of perturbation? Not a bit of it: the sceptics found that ‘when they suspended their judgement, tranquillity followed as it were fortuitously, as a shadow follows a body’ (Pyr. I 29).⁵ A Sextan sceptic reaches Paradise by way of Kensal Green. How can that be? How can the very state of mind which once caused perturbation and excited research now cause tranquillity and the dolce far niente? The suggestion has seemed absurd. An inveterate valetudinarian, I am disturbed by the thought that I may have elephantiasis. I consult The Home Doctor. After prolonged research I still can’t say whether I’m dying or not. Do I sit back and murmur: ‘Thank goodness—that sets my mind at rest’? Not likely. Nonetheless, tranquillity does often follow suspension of judgement, at least on philosophical and semi-philosophical matters. That is to say—to put the point crudely—we often lose interest in a question after a few years or months or minutes of research. I used to worry about the comparative dating of Aristotle’s works. I got nowhere, and neither did anyone else. I don’t worry about the matter any more. But if tranquillity follows upon suspension of judgement, then one thing is plain: a Sextan sceptic has no reason at all for continuing his researches. ⁵ ἐπισχοῦσι δὲ αὐτοῖς οἷον τυχικῶς ἡ ἀταραξία παρηκολούθησεν ὡς σκιὰ σώματι.

Sextan Scepticism

329

Originally, he started investigating because he was perturbed. Now he is no longer perturbed. So why continue to inquire? If—as certain texts suggest—the perturbation which leads to research is nothing other than the desire to discover the truth, then once the perturbation is over, there will be no more research. If—as commentators tend to presuppose—the perturbation is a state of unease caused by a desire, then when the perturbation is over, there will be nothing to move the sceptic to inquire. Such considerations suggest that—pace Pyr. I 3—a Sextan sceptic does not continue his philosophical researches. In that case, the general thesis about investigation with which the Outlines opens cannot introduce that type of philosophy which forms the subject of the Outlines; for none of the three general forms of philosophy which it legitimates is or encompasses Sextan scepticism. The three possibilities which the thesis acknowledges are not exhaustive, and Sextan scepticism depends on one of the possibilities which the thesis ignores. Either there are more than three highest forms of philosophy, or else Sextan scepticism is not a form of philosophy at all. And since an emeritus professor is no longer a professor, surely a sceptical philosopher is no longer a philosopher. Sextan scepticism is not a philosophy: it is a retirement from philosophy. The retirement is characterized by a certain mental condition: namely, a suspension of judgement. Continuation of an inquiry is compatible with belief about the matter under inquiry: is suspension of judgement also compatible with belief? Sextus twice explains what he means by the term ‘suspension’: Suspension of judgement is a stand-still of the intellect on account of which we neither reject nor posit anything. (Pyr. I 10)⁶ Suspension of judgement is so-called from the fact that the intellect is suspended so as neither to posit nor to reject anything because of the equipollence of the items under investigation. (Pyr. I 196)⁷

Suspension is a negative disposition, a disposition not to posit or reject—a disposition not to assent. (Assent is a mental act, not a linguistic act—although it may be manifested in a linguistic act of assertion.) The disposition is caused by what Sextus calls ‘equipollence’. Elsewhere he identifies suspension as an equilibrium—a condition in which one is no more inclined to lean this way than that (Pyr. I 190). I undertake an inquiry. I come to find that all answers are equipollent. I thereby, or perhaps therefore, suspend my judgement. Is such a suspension compatible with belief of any sort or degree? No. If I believe—if I am halfinclined to think, if I rather suspect, … —that (say) my soul is mortal, then I ⁶ ἐποχὴ δέ ἐστι στάσις διανοίας δι᾿ ἣν οὔτε αἴρομέν τι οὔτε τίθεμεν. ⁷ καὶ ἡ ἐποχὴ δὲ εἴρηται ἀπὸ τοῦ ἐπέχεσθαι τὴν διάνοιαν ὡς μήτε τιθέναι τι μήτε ἀναιρεῖν διὰ τὴν ἰσοσθένειαν τῶν ζητουμένων.

330

Jonathan Barnes

thereby have a disposition to assent to the proposition that my soul is mortal and to dissent from the proposition that it is immortal; and if I have such a disposition, then I am not in a state of suspension of judgement. The disposition may be faint or feeble; but it will be enough to establish that I am not in a state of equilibrium—and hence not in a state of suspension. In other words, the fact that a Sextan sceptic does not continue to inquire but rather suspends his judgement requires him to be rustic in one dimension: he drops as deep as he can—wherever his scepticism takes him, there he can have no truck at all with belief. What about breadth? One pertinent consideration turns on the sort of philosophical questions which ancient scepticism addressed. Some philosophical questions are insulated from other areas of inquiry, in the sense that answers to them will not determine answers to questions outside the philosophical domain. Take, for example, the question which Aristotle addresses in the last books of his Metaphysics: What is the nature of numbers? Are they substances or accidents? In what mode or manner do they exist? That is a quintessentially philosophical question. Its resolution is a matter of vast philosophical importance. But it has no consequences at all for the rest of the world. It makes not the slightest difference to arithmeticians or to shop-keepers whether numbers are substances or accidents: two and two will still be four though lads may weep for it. For that reason, suspension of judgement on the philosophical question has no extra-philosophical implications. But not all philosophical questions are thus insulated. It is—or it was—a philosophical question whether anything is good or bad ‘by nature’. Any answer to that question has, or seems to have, implications outside philosophy. If, for example, you hold that nothing is good or bad by nature, then you must in consistency hold that it is not bad by nature to covet your neighbour’s wife. If you suspend judgement on the philosophical question, then in consistency you must suspend judgement on innumerable extra-philosophical questions. So scepticism will seep from philosophy into life. Such seepage is unsurprising: the dogmatic philosophies offered their ethics as something to live by; and if that is so for the dogmatists, it is so also—mutatis mutandis —for the sceptics. The ethical part of philosophy may be the first place to suspect seepage; but it is neither the only place nor the most important. Ancient philosophers asked themselves several sorts of question about causation, among them the question, ‘Are there any causes?’. Sextus raises the question; and a Sextan sceptic will suspend judgement on the philosophical question whether or not there are any causes. But in that case he cannot consistently make any of the thousands of causal suggestions with which we enliven our days. Why won’t the bloody chain-saw start?—Because the sparking-plug’s dirty. Why is the ceiling yellowish just there?—Because that’s where your pipe-smoke rises. Anyone who makes such remarks thereby supposes that there are causes of things.

Sextan Scepticism

331

The same is true for a large number of other philosophical issues which Sextus discusses: Are there any gods? Are there any bodies? Does space exist? Does time exist? Doubts about these and other philosophical issues require doubts about innumerable extra-philosophical issues—issues in other arts and sciences, and issues in daily life. It’s no good saying that the doubt in question is ‘philosophical doubt’, and that you can have philosophical doubt about ordinary matters without giving up your ordinary beliefs about them to any degree. For there is no such thing as philosophical doubt (unless it be ordinary doubt applied to philosophical matters), and Sextus never for a moment imagined that there might be. In short, many philosophical issues—and in particular most of the philosophical issues which a Sextan sceptic addresses—are not insulated, so that a Sextan sceptic will find himself, on pain of inconsistency, suspending his judgement all over the shop. On pain of inconsistency? Perhaps the penalty is higher. Suppose I say: ‘JeanMarc thinks that the tap’s leaking because of a faulty washer, though of course he actually suspends his judgement over the question whether anything is a cause of anything.’ Have I thereby described an inconsistent plumber—or stated an impossibility? Is it possible that someone accepts a singular proposition and does not accept its existential generalization, accepts that this item is thus-and-so and does not accept that some item or other is thus-and-so? Not if he grasps what existential generalization is. For if he understands what it is for there to be something which is thus-and-so, and if he believes that this particular item is thus-and-so, then he thereby believes that there is something which is thus-and-so. A Sextan sceptic suspends his judgement on all philosophical matters which he has investigated. He must therefore also suspend his judgement on a vast number of non-philosophical matters. But for all that has been said so far, he may still be awash with beliefs: he may have philosophical beliefs on the matters which he has not investigated, and non-philosophical beliefs on any matters which may be untouched by his philosophical investigations. But consider now the sorts of arguments with which a Sextan sceptic supports his scepticism—the sorts of arguments which he applies in philosophical investigations and which lead him, in those matters, to suspend his judgement. Most, if not all, of the arguments which Sextus deploys in books II and III of the Outlines against this or that piece of dogmatic philosophizing can, quite evidently, be applied outside the particular domains in which Sextus applies them. For example, in questioning the existence of causes, Sextus invokes the idea that pairs of correlatives must exist and be thought of at the same time. The relational aspect of the concept of a cause thus requires that causes and effects be simultaneous. But the causal aspect of the concept of a cause requires that causes precede their effects. Hence the conception of a cause is incoherent. If that

332

Jonathan Barnes

argument works for causes, it works for numerous other relational items—for signs and signifieds (to which Sextus does indeed apply it), and to parents and children (to which he does not). It will be said that the arguments in books II and III of the Outlines are essentially ad hominem, and that this fact limits their application. Thus at the end of his discussion of causes in book III of the Outlines, Sextus says this: ‘It is necessary to suspend judgement about the existence of causes, saying that there no more are than aren’t causes so far as what the dogmatists say goes’ (Pyr. III 29).⁸ The arguments against causation depend on ‘what the dogmatists say’. In particular, the principle about correlatives is a dogmatic principle, and it is as a dogmatic principle that Sextus appeals to it: ‘They themselves say that correlatives, insofar as they are correlatives, exist and are thought of together with one another’ (Pyr. III 27).⁹ And so often elsewhere. The destructive arguments hold ‘so far as what the dogmatists say goes’. The dogmatists say something about causes and effects, so that the sceptical arguments apply. The dogmatists say nothing about parents and children, and the sceptical arguments have no hold. Now Sextus’ conclusion at Pyr. III 29 is strange. The discussion which precedes it suggests quite a different inference, namely: ‘So there are causes so far as everyday life is concerned, and there are no causes so far as what the dogmatists say goes—and in the light of that disagreement we shall suspend judgement.’ And as a matter of fact, that is just how Sextus normally appeals to ‘what the dogmatists say’. For example: ‘As far as what is said by the dogmatists goes, truth is non-existent and truths do not subsist’ (Pyr. II 81).¹⁰ Suspension of judgement is not usually presented as something which a dogmatist’s own principles will oblige him reluctantly to indulge in. Rather, it is usually presented as the ineluctable consequence of an undecided disagreement—a disagreement one of the parties to which usually consists of the dogmatic philosophers. At Pyr. III 29 Sextus expresses himself carelessly. In any case, are Sextus’ arguments in books II and III ad hominem? An ad hominem argument has among its premisses or presuppositions certain propositions which, however odd they might seem to the likes of you and me, form part of the beliefs of the homo in question and are appealed to precisely because they form part of his beliefs. Sextus’ argument about causation is not like that. Ordinary life assures us that the world is full of causes. The clever philosophers have shown—implicitly and against their will—that there are no causes. So we shall suspend judgement. The philosophers are not invoked in order to be mocked or refuted—they are invoked in order that all reasonable men will come to a suspension of judgement. ⁸ ἐπέχειν ἀνάγκη καὶ περὶ τῆς ὑποστάσεως τοῦ αἰτίου, μὴ μᾶλλον εἶναι ἢ μὴ εἶναί τι αἴτιον λέγοντας ὅσον ἐπὶ τοῖς λεγομένοις ὑπὸ τῶν δογματικῶν. ⁹ τὰ δὲ πρός τι φασὶν αὐτοὶ καθὸ πρός τι ἐστὶν συνυπάρχειν καὶ συννοεῖσθαι ἀλλήλοις. ¹⁰ ὅσον ἐπὶ τοῖς λεγομένοις ὑπὸ τῶν δογματικῶν, ἀνύπαρκτος μέν ἐστιν ἡ ἀλήθεια, ἀνυπόστατον δὲ τὸ ἀληθές.

Sextan Scepticism

333

In the first book of the Outlines Sextus describes scepticism and its argumentative armoury. The most redoubtable weapons in that armoury, and the weapons of which Sextus makes most frequent use in books II and III, are the five modes of Agrippa. In principle, the modes may be deployed either singly or in groups. Sextus himself observes that ‘we shall show that every item under investigation may be referred to these modes’ (Pyr. I 169),¹¹ and he develops a schematic (and rather unsatisfactory) argument which puts all five modes to collaborative work. In many standard presentations of scepticism, three of the modes are put together to create the following schematic argument. Suppose you think to assent to a certain proposition, P1 . Then either you have nothing at all to say which favours P1 rather than its negation, or else you have something to say. If you have nothing at all to say, then you must suspend judgement (that is the ‘hypothetical’ mode). So suppose you have something to say, and call it P2 . If P2 is the same as P1 , then the reciprocal mode requires you to suspend judgement. So suppose it is a new proposition: then either you have nothing to say in its favour—in which case you must suspend judgement on it, and hence on P1 as well—or else you have something to say. Suppose you say that P3 . Then what has been said of P2 will be said, mutatis mutandis, of P3 , and so suspension of judgement will be defeated only by the introduction of another new proposition, P4 . And so on and on ad infinitum. But you can’t go on and on ad infinitum (that is the mode from infinite regression), so suspension is the only option. That is the roughest of presentations; but it is enough to establish the pertinent point: the modes of Agrippa are universal in their application—if they work anywhere, they work everywhere; if they work for philosophical propositions, they work for geometrical propositions and for geographical propositions and for any and every of the propositions which we stumble upon in our waking lives. That is beyond cavil. Sextus himself says that ‘we shall show that every item under investigation may be referred to these modes’. Any item which a sceptic investigates will fall to the modes, and a Sextan will be not only a sceptical philosopher but also a sceptical gardener—if he investigates any horticultural matters. And he will so investigate: Are the potatoes chitting nicely? Should I transplant the wallflowers? Every gardener asks himself questions of that sort. And even if Incapability Sextus doesn’t actually ask himself if the oak copse needs thinning, he need only say to himself, ‘The copse needs thinning’, in order to realize that were he to investigate the matter he would suspend judgement on it; and that, in turn, is enough to lead him to suspend judgement. That is enough—provided that he is minimally alert and minimally rational. It is logically possible that a Sextan sceptic should restrict his domain: he might not investigate, or even think of investigating, certain matters; or he might investigate ¹¹ ὅτι δὲ πᾶν τὸ ζητούμενον εἰς τούτους ἀνάγειν τοὺς τρόπους ἐνδέχεται, διὰ βραχέων ὑποδείξομεν.

334

Jonathan Barnes

them and yet fail to wheel on the Agrippan modes. But if a Sextan sceptic with a restricted domain is a logical possibility, he is not a rational possibility: no one who has exercised the modes and observed their character and their power could fail to exercise them across the board. Unless, of course, the board holds some items which are impervious to investigation. It is not easy to imagine that there could be any such items—but Sextus himself seems to suggest that there are: We do not overturn the items which lead us involuntarily and in accordance with a passive impression to assent … These items are the appearances: when we investigate whether the underlying object is such as it appears to be, we grant that it appears and our investigation is not about the appearance but about what is said about the appearance—and that is not the same as an investigation about the appearance itself. (Pyr. I 19)¹²

The passage doesn’t actually say that appearances can’t be investigated; but it suggests as much—and it certainly implies that a Sextan sceptic won’t investigate them. ‘Is that drill really straight?’, asks the sceptical gardener, and shrugs his sceptical shoulders. It’s enough, after all, that it looks straight to him now; and whether or not he says to himself ‘It looks straight enough’, he doesn’t stop to ask if it really looks straight. Why does—how can—a Sextan sceptic thus limit his investigations? Sextus doesn’t say. And the limitation is arbitrary. A gardener who has drunk his usual lunch may wonder if the drill really looks straight to him; and more sophisticated experts frequently pose similar questions: Does it really taste like a ’98 St Glinglin? Does it really feel like an antelope pulse or a gazelle pulse? Does the shadow under the elm really look dark purple? Does the third violin really sound slightly sharp? Appearances aren’t always what they seem to be—as the dogmatic philosophers were well aware. Sextus either ignores or disallows such questioning of the appearances. He is wrong to do so. And Sextan scepticism, despite the intimations of the first paragraph of the Outlines and notwithstanding numerous counter-indications in the Greek texts, is out-and-out rustic. ¹² τὰ γὰρ κατὰ φαντασίαν παθητικὴν ἀβουλήτως ἡμᾶς ἄγοντα εἰς συγκατάθεσιν οὐκ ἀνατρέπομεν … · ταῦτα δέ ἐστι τὰ φαινόμενα. ὅταν δὲ ζητῶμεν εἰ τοιοῦτον ἔστι τὸ ὑποκείμενον ὁποῖον φαίνεται, τὸ μὲν ὅτι φαίνεται δίδομεν, ζητοῦμεν δ’ οὐ περὶ τοῦ φαινομένου ἀλλὰ περὶ ἐκείνου ὃ λέγεται περὶ τοῦ φαινομένου· τοῦτο δὲ διαφέρει τοῦ ζητεῖν περὶ αὐτοῦ τοῦ φαινομένου.

17 The Wife of Philinus, or the Doctors’ Dilemma: Medical Signs and Cases and Non-deductive Inference G. E. R. Lloyd

In a pioneering paper in Science and Speculation Myles Burnyeat opened up the hitherto neglected topic of ‘the origins of non-deductive inference’ (Burnyeat 1982). In it, and in the companion piece on ‘enthymeme’ contributed to the Symposium Aristotelicum volume on the Rhetoric (Burnyeat 1994), he examined especially the Aristotelian texts dealing with s¯emeion, tekm¯erion, enthum¯em¯e, and related terms and traced their influence on later, Hellenistic theories of signs. Aristotle comes out of the analysis with considerable credit for having seen that there are good arguments that are not formally valid, though that insight tended to be lost, or at least was not developed, in subsequent philosophy. From outside philosophy, Burnyeat drew on materials from Greek oratory especially and had some perceptive incidental comments to make on signs in the medical literature. For my contribution to this collection of essays in his honour I shall focus on a body of material that still has not figured as prominently as it might in the ongoing debate: namely, the fifth- and early fourth-century medical texts. The Hippocratic Corpus provides, indeed, immensely rich materials for the study of non-deductive inference—as also do the extant remains of Hellenistic medicine, although I shall defer consideration of them to another occasion. The interest of this material lies partly in the obvious fact that the doctors were working under considerable pressure to arrive at judgements in cases where their patients’ lives—as well as their own reputations—were at stake. My aim is not to go over their use of signs in order to attempt retrospective diagnoses in the light of modern medical knowledge, in the way that Major (1957), Siegel (1964), Grmek (1989), and others have done. Rather, I shall put myself as far as possible in their shoes, to examine what they thought they could infer from the signs they cite, given how they understood disease. Given that, from the perspective of modern scientific knowledge, their understanding of the causal factors in diseases was extremely limited, that makes the situation all the more interesting from the

336

G. E. R. Lloyd

point of view of the study of what were perceived to be reasonable or reputable arguments. Aristotle, it will be recalled, distinguished tekm¯eria from s¯emeia for example in An. Pr. 70b 1 ff. and Rh. 1357b 4 ff. The former, based on universal connections, can yield deductive demonstrations. While s¯emeion is sometimes a generic term, where it includes tekm¯eria as one of its species, it is also used of signs that do not relate to universal connections and that can always be refuted. The two questions I wish to press are, first, whether medical signs in practice conform, or should have conformed, to the analysis that Aristotle gives, and secondly, what the relationship is between medical signs and the individual case histories with which they are often associated. In the former case I am not interested in nomenclature or divergences in terminology between different Hippocratic writers and Aristotle, but with the doctors’ actual use of signs. My second question follows from the first, but goes beyond it, in so far as the individual case histories throw light on the bases on which the doctors constructed the signs they used. Let me begin with two Aristotelian preliminaries that will serve to give my questions bite. First, certain medical or quasi-medical examples figure in Aristotle’s discussions. One of the standard illustrations of a demonstrative sign that appears in Aristotle—and is then repeated over and over again in later literature—is the milk in the breasts of a woman, from which it can be inferred that she has given birth (tetoken), as Rh. 1357b 15 f. puts it, or, as An. Pr. 70a 13 f. does, that she is pregnant (kuousan). This is supposed to be a necessary connection, though as Burnyeat (1982: 204 n. 30), pointed out, in the latter case certain further conditions have to be met for this to hold. Nursing mothers and wet-nurses are not to be considered counter-examples. Nor should Aristotle’s remark, at Hist. an. 522a 1 f., that it is only ‘for the most part’ that milk is produced following pregnancy.¹ Rather, as Burnyeat put it, ‘Aristotle is thinking of a context in which the woman has not yet given birth to the infant she is carrying, and she or the doctor infers that she is carrying one from the signs given by a milky secretion in the breasts’. The problem that Burnyeat then craftily left ‘as an exercise for the knowledgeable reader’ is to formulate conditions C such that ‘all human females who have a milky secretion in the breasts and who meet conditions C are pregnant’ comes out true. It is notable that for the sequence of reasoning to be useful to the doctor, those conditions C had better not be such that by themselves they allow the inference that the woman is pregnant even without the observation of the milky secretion. The question is under what circumstances that observation plays a role in making the inference more reputable. ¹ The Hippocratic Aphorism V 39, L IV 544 14 ff. recognizes that milk can be produced by women who are neither pregnant nor have given birth, and in such cases diagnoses suppression of the menses.

The Wife of Philinus

337

Some of Aristotle’s other medical examples are not exactly impressive either. There may be an impeccable necessary connection between ‘having a fever’ and ‘being ill’, as Rh. 1357b 15 suggests, which may be useful in constructing syllogisms. But that is unlikely to be of any practical import to the doctor, for if he knows that the patient’s high temperature is due to a fever, he already knows that the patient is ill. The second medical example at Rh. 1357b 18 ff. also reveals where Aristotle’s interests in these discussions differ from those of the physician. Heavy or frequent (puknon) breathing as a sign of fever can be refuted, even if it happens to be true in a particular case that a patient who exhibited that symptom was suffering from a fever. Aristotle is concerned with the necessity of the consequence, not the truth of the conclusion—and the same may be said of the other non-necessary connection he refers to in An. Pr. 70a 36 ff., between a sallow complexion and pregnancy. It is worth recalling that in his discussion of the fallacy of the consequent in the Sophistical Refutations (167b 1 ff.) he remarks generally that in rhetoric proofs from signs commit this. My second Aristotelian preliminary relates to remarks concerning physiognomics at the end of An. Pr. II 27, 70b 7 ff., to which Burnyeat devoted some other acute remarks (1982: 203). The final chapter of the Prior Analytics discusses physiognomics in a purely hypothetical vein. If one grants that body and soul change together, and if one concedes that for each change or affection there is a single sign, then certain inferences are possible. Aristotle’s example is the connection between the lion’s courage and ‘possessing large extremities’ (70b 16 f.). The affection, courage, might then be revealed by that sign in other species too (such as humans), though there will be complications that Aristotle notices if the condition one affection, one sign, is not met. He does not here endorse physiognomic inferences, even though there are passages in Historia Animalium I especially (and cf. Hist. an. IX, though its authenticity has been doubted) which show that he was prepared to take some such connections seriously (cf. Lloyd 1983: 22). Again, the interest in the logical treatises is in what follows from what in a hypothetical situation, rather than in the truth of the conclusions. Such texts as Epidemics II 5 1 (L V 128 1 ff.) and II 6 1 (L V 132 15 ff.), which try their hand at physiognomy, are, then, closer to Hist. an. I, e.g. 491b 15 ff., than to the Prior Analytics. Before I turn to the Hippocratic texts themselves, it is as well to begin by recalling just how extensive early Greek ambitions to infer the unseen from the seen were. Even before Aristotle’s analyses, the dictum opsis adelon ta phainomena, ascribed to Anaxagoras and echoed in many other writers, covered sign inferences as well as other relations (Diller 1932; Lloyd 1966: 338–41; Allen 2001: 2 ff.). I outlined the main varieties of early Greek prediction in The Ambitions of Curiosity (Lloyd 2002: ch. 2), underlining the well-known ambivalence of the figures of the seer and others who claimed to have special skills in that domain. Already in Homer Calchas is reviled by Agamemnon, and Polydamas by Hector, although on both occasions the seer is correct in his analysis of the situation. The

338

G. E. R. Lloyd

portrayals of Cassandra, Teiresias, and others in tragedy are similarly ambivalent. Yet the fact that many actual seers were distrusted—and some were mercilessly lampooned by Aristophanes—should not be equated with any general scepticism about divination as a whole. We have only to think of the continued prestige of the Delphic Oracle throughout the classical period to realize that suspicion could be combined with respect. A pair of Hippocratic texts neatly exemplifies this ambivalence. In On Regimen I (CMG I 2 4, 136 6 ff.) divination, mantik¯e, is cited alongside other respectable, even honoured, human skills, technai, such as architecture, sculpture, pottery, and writing. But the writer of On Regimen in Acute Diseases ch. 3, L II 238 8 ff., 242 3 ff., first notes that doctors disagree among themselves about treatments (some saying that unstrained barley-gruel is good for patients suffering from acute diseases, while others claim it is very bad for them) and then goes on to deplore that fact. It is enough to give medicine a bad name among lay persons. Ordinary people might well compare medicine to divination, the writer says, where diviners dispute the interpretation of bird flight and the signs in the entrails, and even offer opposite judgements about the same sign. Some medical writers were evidently aware that in some quarters the physicians’ interpretations of their signs might appear arbitrary. But of course medical practice always and everywhere depended on such. The whole of the treatise Prognostic, and texts such as Epidemics I ch. 10, L II 668 14 ff., and On Regimen in Acute Diseases (Appendix) ch. 9, L II 436 5–442 6, set out lists of the signs that the doctor should pay attention to. But the writers responsible do not just stay at the level of general recommendations, telling the doctor to examine the excreta, the temperature of the extremities, whether the patient has slept or not. We have in addition detailed interpretations of what different types of urine, vomit, stools, and so on signify. It is this material that provides me with my opportunity to investigate the match or mismatch between the doctors’ modes of reasoning, in their actual medical practice, and Aristotle’s abstract analysis. Two points stand out immediately. First, the doctors’ concern focuses most frequently on forecasting the outcome of a complaint. They examine which signs may be taken to indicate that the patient will recover, which that he or she will die—where it is obviously more difficult to be certain of the former result than of the latter. Foretelling the outcome of a disease is, indeed, mentioned as one of the aims of prognosis in the treatise of that name, as it is also in Epidemics I ch. 11, L II 676 12 ff. In Prognostic ch. 1, L II 112 6 ff., the extra points are made that getting the prognosis right will increase the doctor’s reputation, and that if he has foretold a fatal outcome, he will not be blamed if that happens: that, at least, is the claim. All the individual case histories in Epidemics I and III, and the vast majority of those in the other books of the Epidemics record whether the patient died or had a crisis, where it is recognized that some crises may be only temporary, and there is a relapse. So one has to distinguish intermediate, from complete, crises.

The Wife of Philinus

339

My second general point is that the descriptions and interpretations of the signs already presuppose a good deal of medical expertise. The reader is expected to know what urine ‘of a good colour’ (euchros) is like, or what sputum or vomit evacuated ‘opportunely’ (epikairos) means. The famous description of the Hippocratic facies in Prognostic ch. 2, L II 112 12 ff., starts off by instructing the doctor to see whether the face of the patient is ‘like the faces of healthy people’. That does not seem very helpful, but there then follows a very detailed account of the signs that indicate an unhealthy face (‘nose sharp, eyes hollow, temples sunken, ears cold … .’). By contrast, when the doctor is told, in the surgical treatises—for example, On Joints ch. 10, L IV 102 9 ff.—to judge luxations by comparing the healthy and the unhealthy, it is specified that this means comparing the sound and unsound limbs of the same individual, not those of different ones, since they may differ. In many cases we do not have direct access to the evidential basis for the generalizations that are proposed. But there are certain exceptions to this which provide valuable insight into that question, and we can come closer to seeing how the doctors proceeded in their modes of reasoning. Thus it has always been recognized that the third Constitution in Epidemics I is connected with the fourteen case histories that follow. Although there are first-person remarks recording what the writer in question says he has seen or noted, it would be rash to assume that it is the same writer on each occasion, let alone that one man was responsible for the whole of the treatise (cf. Lloyd 1987: 62 f.). There are, as we shall see, certain minor discrepancies that suggest either multiple authorship, or editorial interference, or scribal error, or some combination of these factors. Nevertheless, the cross-references suggest that case histories and Constitution are to be read together. After the usual description of the weather conditions at Thasos (ch. 8, L II 642 4 ff.) the Constitution turns to the kausoi, ardent fevers, and says that those who had a good and copious nose-bleed were most likely to survive. ‘Indeed I know of no one who died in this Constitution if there was proper bleeding’, adding that in the case of three patients who died, there was only a slight nose-bleed on the fourth and fifth days. In the case of those who suffered from ‘jaundice’ (ikteros) on the sixth day, what was beneficial was either purging through the bladder, or a disturbance of the bowels, or a copious haemorrhage ‘as was the case with Heracleides’, who showed all three symptoms and had a crisis on the twentieth day. The servant of Phanagoras, however, had none of these symptoms and died. Indeed ‘the great majority of those who had no haemorrhage died’ (644 8 f.). Later nose-bleeding and menstruation are described as signs of recovery in women patients (646 9 ff.). The next chapter (9,656 7 ff.) remarks on the great variety of diseases, but sums up the signs of recovery as follows: ‘In this constitution there were four signs especially that signified recovery: a proper haemorrhage through the nose, copious urine with an abundant and good sediment from the bladder, disordered bowels

340

G. E. R. Lloyd

with bilious evacuations at the right time, and the onset of dysentery.’ Conversely, the deadly signs had been identified (650 12 ff.) as acute fever with slight rigors, sleeplessness, thirst, nausea, sweats about the forehead and collarbones (but not all over), delirium, fears, depression, very cold extremities—that is, the toes and hands and especially the latter. We are told too (652 4 f.) that the exacerbations were on the even days. But when we turn to the case histories that follow, certain anomalies come to light.² Three patients, named Philiscus, Epaminon, and Silenus, are mentioned in the Constitution, 642 8, and Philiscus and Silenus, along with a certain Epaminondas, reappear at 664 11 f. Neither an Epaminon nor an Epaminondas figures among the case histories, but there is a Philiscus in case 1, and a Silenus in case 2. Philiscus’ case history matches what is said about a person of that name in the Constitution, in two respects: a nose-bleed, but only a slight one on the fifth day, and death on the sixth. Where ‘Silenus’ is concerned, however, the Silenus in case 2 does not have a slight nose-bleed on either day four or day five (as was reported at 642 8), and he dies on the eleventh day, not the sixth (cf. 664 11 f.). So either we are dealing with a different Silenus (there is a ‘Silenus’ in whose house Bias lodges in the Constitution at 644 11, 650 1), or the text has been interfered with and is corrupt, or the compiler was none too careful in writing up the Constitution on the basis of the case histories. The lack of an exact match between the personnel of the case histories and those named in the Constitution should also be noted. Although there is one further case of a possible match, in that case 10 deals with a Clazomenian whose complaint tallies with what is reported concerning a certain Hermippus of Clazomenae at 660 5,³ the other eleven individuals whose case histories are given do not figure in the various chapters of the Constitution, while, conversely, the Constitution names or otherwise identifies some twenty individuals for whom we have no case histories. There are, then, certain complications or difficulties in tracing how the generalizations made in the Constitution may be derived from the individual cases that follow or that are mentioned in the Constitution itself. But there is a further stage of generalization that may prove even more problematic: namely, when the medical writers offer general dicta that purport to interpret a given sign unconditionally. Thus at one level the individual case histories deal with the particular patients whose cases they describe, but they rarely offer any general interpretative comments on the connections between the symptoms observed in the particular case and their general significance. At the next level up, the Constitutions describe what happened over a period of months at a particular location in a particular year, and tie many of their generalizations to that Constitution. But then at the highest level, we have generalizations about what ² The fullest discussion of these issues is Deichgräber 1933. ³ The main common feature is swellings by the ears that subsided after discharge of ‘thick’ urine.

The Wife of Philinus

341

some sign (whether on its own or taken in conjunction with others) can be taken to indicate, wherever and whenever it occurs. Prognostic ch. 25, L II 188 14 ff., explicitly claims that bad signs are bad in every year and everywhere in the world, whether in Libya, Delos, or Scythia, and correspondingly, good signs always indicate good. ‘So one must realize that in the same districts it is not strange if one is right in the majority of instances.’ Thus at the second level, each of the Constitutions in Epidemics I and III endeavours to describe and comment on the specific, complex medical situation it deals with. All three of the Constitutions in Epidemics I, for instance, allow for considerable variation in the experience of different sufferers from the ‘same’ illness, as well as between those diagnosed as suffering from different complaints. Evidently the remarks about the character of the excreta in any given Constitution are meant to apply only to that Constitution, sometimes indeed just to a majority of patients or to particular groups of them. We should interpret the observations about the periodicities of diseases similarly. Thus the Constitution in Epidemics III (ch. 6, L III 82 1 f.) remarks, as Epidemics I ch. 9, L II 652 4 f., had done, that the exacerbations were on the even days, and it adds that loss of memory, exhaustion, and speechlessness accompanied those exacerbations. If we may assume a correlation between that Constitution and the cases that follow (even though there are no names in this Constitution to aid identification and so support that assumption), we find exacerbations, together with speechlessness, on the even days, in case 5, for instance, and exacerbations on even days with delirium (though no mention of speechlessness) in case 10. Yet elsewhere in this set of case histories there are plenty of exceptions, with exacerbations noted on odd days. For example, in case 1 there are exacerbations on day 7 and day 17; in case 4 there is speechlessness on the second day but an exacerbation on the third, though the patient dies the next day. The sequences and the connections with other symptoms were clearly meant and stated to hold in most cases, not all.⁴ Several Hippocratic writers were, however, obviously keen to present general theories of critical days, and some of these are presented in unqualified terms. Indeed, one such apparently general theory is included in the last chapter of Epidemics I (ch 12, L II 678 5 ff.). There it is suggested that when the exacerbations are on even days, so too are the crises, and similarly and conversely, when the former are on odd days, so too are the latter. Yet it is well known that there is no orthodoxy in the Hippocratic Corpus on this question (cf. Lloyd 1991: 214 ff.). Prognostic ch. 20 (L II 168 6 ff.) is one text where the theory is not based on even- or odd-day correlations, but rather on complex sequences of days of both types. That chapter continues, moreover, with a warning that the periods cannot be calculated exactly in whole numbers of days, in that, resembling the ⁴ Contrast the unqualified remark about exacerbations on even days in the kausoi cases at Epidemics I ch. 9, L II 652 4 f., even though the later periodic sequences are said to hold only for the most part. The sequences documented at ch. 9, 660 6–668 12, exhibit very considerable variety.

342

G. E. R. Lloyd

solar year and the lunar month, they cannot be expressed exactly as a whole number of days either. Epidemics I ch. 10, L II 668 14 ff., remarks that in arriving at a ‘diagnosis’, the doctor must attend both to the ‘common nature’ (of all humans) and to the peculiar character of the individual, but the ambition to generalize, indeed to arrive at unqualified general dicta, on the basis of the individual case histories that are so carefully recorded, is apparent. Two pairs of chapters from Epidemics V and VII, to which Littré already drew attention and which were later studied by Hanzawa (1987), illustrate this.⁵ Epidemics V ch. 71, L V 244 20 ff., describes the case of Bias the boxer, a big eater, who suffered from ‘cholera’ and intermittent fevers after eating meat, especially rare pork, drinking aromatic wine, eating cakes and honey-cakes, ripe cucumber, and fresh polenta, and also drinking milk. In Epidemics VII ch. 82, L V 436 22 ff., the causes of ‘cholera’ are discussed in general. That text lists most of the items that had been identified in Bias’ diet, meat eating, especially undercooked pork, the aromatic wine, the cakes and honeycakes, ripe cucumber, and fresh polenta. But further items are added (chickpeas, certain types of shellfish, green vegetables, cooked lettuce), and that chapter also suggests sunbathing as a cause, which does not figure in the chapter on Bias. Again, though slightly less clearly, Epidemics V ch. 102, L V 258 6, refers to individual cases (the plural verbs show that several people are in mind) of patients suffering from ‘catarrh’ on one side of the head, where liquid ran from the nostrils and the patients had a fever that abated properly on the fifth day. Epidemics VII ch. 56, L V 422 14 ff. deals in general with certain conditions of headaches and fluxions. If there is pain on one side, and a thin or ‘concocted’ liquid flows from the nose, ear, or pharynx, that is not so dangerous as when there is more intense pain and dry emissions. Those are clearly different symptoms from the Epidemics V ch. 102 ones, but the chapter ends (422 19 ff.): where there is ‘catarrh’ on one side of the head, and pain, and liquid runs from the nostrils, the patients will suffer from fever but cool down properly on the fifth or sixth day. The extra points made in Epidemics VII ch. 56 clearly have some other origin, not Epidemics V ch. 102, but that chapter could well be the chief source for that final remark. Given the limited techniques of diagnosis available,⁶ the desire to extract the maximum prognostic information from those that were is entirely understandable. That ambition is in evidence in many texts. But so too are signs of hesitation, a recognition of the difficulties of generalization, and indeed of the problems of interpreting the significance of the symptoms described, whether on their own or in combination. ⁵ In other instances chapters in Epidemics V and VII are more or less doublets of one another: cf. Deichgräber 1933: 127 ff. ⁶ As is well known, the Hippocratic authors make no use of the pulse in diagnosis, a discovery attributed to Praxagoras around 300 .

The Wife of Philinus

343

The interpretation of the signs in urine exemplifies the difficulties the doctors were in. Prognostic ch. 12, L II 138 15 ff., 142 1, says that the ‘more fatal’ (thanatodestera) kinds are fetid, watery, melana (which covers all dark shades, though Grmek prefers black in this context⁷) and thick. When we turn to the individual case histories, we find many that note that the patient’s urine was ‘dark’ (melana) or ‘darkish’ (hupomelana) where the case ends fatally. In Epidemics III, for instance, that is true of cases 8 (L III 56 4) and 11 (62 8 f.) in the first series, and of case 3 in the second, which further remarks that there were suspended substances (enaiorema) in it that did not form sediment (114 1 and 10). Case 2 of that series observes that throughout that case (which also ends in death) the urine was dark, thin, and watery (112 10). There may, of course, have been additional symbolic factors that influenced the interpretation of melana urine as a sign of death, but certainly there are individual case histories that support the generalization in Prognostic ch. 12. But there are also plenty of exceptions to that general rule. The wife of Epicrates, whose history is given in case 5 of Epidemics I (L II 694 4 ff.), originally presents with (among other signs) urine that is thin and darkish (hupomelana), but so far from there being a fatal outcome, in her case, she has a complete crisis on the eightieth day. However, it is noted that the urine was ‘of a better colour’ (euchrootera) on the eleventh day, and it has a ‘light’ (or ‘white’, leuk¯e) sediment (noted as a good sign in Prognostic ch. 12, although with reservations about its needing to continue throughout the sickness). In Epidemics III case 9 in the second series (L III 128 2 ff.) similarly, the urine is thin and dark (sometimes with, but sometimes without, an enaiorema) at the start of the disease, but from the eightieth day it was ‘of good colour’ and had more deposit. By day 120 the patient reaches a complete crisis. The next case, 10 (130 4 ff.), similarly starts off with urine that is ‘thin and dark’, but by the twentieth day it is light or white and thick—though with no sediment—and on the twenty-fourth, when the fever is resolved in a crisis, it is abundant, light/white, and now with abundant sediment. One complication is, then, that changes in the colour of the urine frequently occur, and unfavourable ‘dark’ urine may give way to urine of a better colour—or vice versa. But even without any such change, ‘dark’ urine does not necessarily betoken death. Case 11 of the second series in Epidemics III (L III 134 2 ff.) is one that is resolved with a crisis on the third day, though on that very day her urine is noted as ‘dark’ and thin with a round enaiorema in it which did not form sediment. But this patient manifested the ‘good’ sign of copious menstruation, remarked on elsewhere as a sign of recovery in women. If we follow the fortunes of patients described as having urine that resembled that of draught animals (hupozugia)⁸—I shall call this draught animal or DA ⁷ cf. Grmek 1989: 285 ff., 295 f. ⁸ This has sometimes been taken to be asses’ urine, but Xenophon says that hupozugion covers oxen, mules, and horses (Oec. 18 4).

344

G. E. R. Lloyd

urine for short—we find a similarly complex picture. There are references to urine of this type in a number of places in the Corpus, of which four present a particularly striking set of comments. Three relate to different individual patients, and the fourth to a generalization, in the Aphorisms, that tallies with two of the individual cases, but not so well with the third. In the case of the wife of Philinus, case 4 of Epidemics I, the doctor who records the case first notes that the colour and consistency of the urine was like that of a hupozugion and adds: ‘such was the urine that I myself saw’ (L II 692 13 ff.). In Epidemics VII ch. 112 (L V 460 6 ff.) we are told that Polyphantus also passed DA urine, and that case has a number of features in common with the wife of Philinus. Both these patients suffered from pains in the head, both were feverish, both suffered from intermittent delirium, both had spasms, both died. Aphorisms IV 70 (L IV 526 11 ff.) also correlates turbid DA urine in cases of fever with headaches. Thus far it seems as if DA urine is being regularly connected with other phenomena in a syndrome that might provide a reliable guide to a complex affection. It might look as if we are dealing with what the physicians could claim was a necessary universal connection—in Aristotle’s terms a tekm¯erion. But the case of Parmeniscus in Epidemics VII ch. 89 (L V 446 7 ff.) tells against that. This is another instance of DA urine, and he exhibits at least some of the same symptoms as the wife of Philinus: sporadic thirst, pain in the hypochondrium, speechlessness, and (on one reading) continuous fever (the fever was intermittent in the case of the wife of Philinus, severe in that of Polyphantus). But Parmeniscus has no pain in the head, and his illness ends on the fourteenth day in recovery, not death. What, then, did any of these doctors think that draught animal urine indicated? The writer of Aphorisms IV 70 offers one, limited generalization. But even a small number of case histories is enough to suggest the difficulties of using DA urine on its own, or even in conjunction with other phenomena, as the basis for inferences concerning what patients were suffering from, or even whether they would recover. The final complication that must be taken into account in relation to prognostications from the urine is a point that is made implicitly in the Aphorisms and explicitly in Prognostic. This is that the urine may just indicate an affection of the bladder or of the kidneys, and not relate to the general condition of the patient. Aphorisms IV 75–81, L IV 530 4–532 4, gives seven examples where the urine indicates such an affection, whether of the bladder (75, 77, 79, 80, 81) or of the kidneys (75, 76, 78)—with IV 75 said to be a sign of one or the other. Thus thick urine with small pieces of flesh like hairs is here said to signify a secretion of the kidneys (IV 76); we may compare On the Nature of Man ch. 14, CMG I 1 3, 202 6 ff., where a very similar symptom is said to result from either kidney conditions or arthritic complaints. Where Prognostic ch. 12, L II 140 6 ff., notes that ‘bran-like’ (piturodees) urine is a particularly bad sign, both Aphorisms IV 77 and On the Nature of Man ch. 14 associate ‘bran’ appearing in

The Wife of Philinus

345

the urine, or a ‘bran-like’ character, with psoriasis of the bladder. While in that regard Prognostic ch. 12 offers a less specific inference, that chapter ends (142 12 ff.) with the warning that the doctor should not be deceived if the bad signs in the urine arise just from a disease of the bladder, not of the whole body. It issues this warning, but does not go into any detail as to how to differentiate the two. No doubt other signs of the general condition of the patient would be the chief way of doing so. Let me turn back now to the wife of Philinus to point out the hermeneutic problem the doctors were faced with. She had given birth to a baby girl, and there was nothing abnormal (so we are told) in the lochial discharge. But on the fourteenth day after the delivery she was seized with fever and rigor, and the case history begins, ending on the twentieth day with her death. But what was the significance of the stomach pain and the pain in the right hypochondrium that she suffered from at the beginning? Pain in the right hypochondrium is mentioned many times elsewhere in the Hippocratic Corpus, and pains on the right side are noted as especially dangerous.⁹ Then what about the pains in the genitalia, or those in the head, neck, and the lower part of the back? She clearly did not suffer just from a localized pain in the right hypochondrium. Then what about the bilious stools? And the cold extremities and thirst? Then there was her urine, thin and at first achros (either colourless or of a bad colour), then white and thick, that was passed with spasms, and that took a long time to settle, eventually urine the author tells us with such emphasis that he saw to have the colour and consistency of that of draught animals. Was it the combination of that urine, with intermittent delirium, that was significant, even though in some other cases that combination did not occur?¹⁰ Anything and everything in the case histories may be significant. Indeed, it is unlikely that the doctors would have bothered to record details that they could see at the time to be irrelevant—although in many instances we may assume that they were keeping an open mind about just what was essential in the course of the disease, and what merely coincidental. What we are given in this and many other cases are careful and detailed accounts of individual patients. But then the problem is, how is this information to be interpreted and used? No two patients are ever exactly alike.¹¹ This does not mean that no two diseases are the same, whether ‘disease’ is defined in ancient or in modern terms (as a ‘quartan’, for example, or as malaria). The individuality of a particular person’s ⁹ e.g. Prognostic ch. 7, L II 126 7 ff.; Epidemics I, cases 11–13; Epidemics III, cases 2, 3, 4, and 8 of the first series, cases 13 and 16 of the second. Jones (1923: i. 17 n. on the Prognostic passage) was confident that he could diagnose appendicitis, but Grmek countered that the descriptions were too indeterminate to do so with any confidence (1989: 132). ¹⁰ It does not in the case of the other individuals mentioned alongside Polyphantus in Epidemics VII ch. 112, discussed above. ¹¹ It is interesting, however, that the Epidemics sometimes tell us about brothers who both suffer from similar diseases, even though their conditions are noted to differ: e.g. Epidemics I ch. 9, L II 660 7 ff.

346

G. E. R. Lloyd

condition may reflect, rather or in addition, their age, sex, build, and life-style, among other factors. The possibility that each patient’s condition should be treated as a distinct disease is, indeed, one that is mentioned and criticized in On Regimen in Acute Diseases ch. 1 (L II 228 2 ff.), where the author attacks the positions of the authors and revisers of the (lost) work Cnidian Sentences, both on the grounds of their proliferation of what they counted as diseases and on those of the paucity of the remedies they were prepared to use. The proliferation of diseases, with many different species of diseases of the spleen, several varieties of ‘typhus’, and so on, is certainly a feature of such treatises as On Internal Affections and On Diseases I–III. That may be agreed, without endorsing the—to my mind, speculative—reconstructions of the theories and practices of a supposedly well-defined ‘Cnidian’ school of medicine that have been attempted, by Jouanna (1974), Grensemann (1975, 1987), and others (Thivel 1981; contrast Smith 1973, Lonie 1978). We noted that there are plenty of signs in different Hippocratic texts of their authors’ ambitions to arrive at sound generalizations, correlating signs and outcomes. That is a feature of the Constitutions in the Epidemics, where, however, the generalizations suggested are generally limited to the particular Constitution in question. But in such treatises as Aphorisms and Prognostic especially, there are many apparently quite unqualified general dicta. Yet the evidential basis for any such generalizations had ultimately to be the individual cases that the doctors encountered or that others had reported to them.¹² However, in the individual case histories as we have them, we have not just singletons, single prognostic or diagnostic signs, but whole batteries of them.¹³ The dilemma is obvious. If you keep all the complexity of these individual cases in play, they become impossible, or at least extremely difficult, to apply. But if for the sake of the application to other cases, generalizations are attempted, this will always involve selection as well as the loss of some of the information the cases contained. The proposed signs in the Constitutions are, in most cases, bound to be subject to reservations. That is no big problem if we take it that the correlations proposed are general, not universal, exceptionless rules—in other words as s¯emeia indeed, in Aristotelian terminology, rather than as tekm¯eria. ¹² In such treatises as On Internal Affections, which identify many different varieties of each general type of disease, those varieties are derived from speculative aetiologies, and those works do not generally reveal what empirical basis their taxonomies have. ¹³ It is not surprising that the later Empiricist doctors concentrated heavily on tracing the similarities between individual cases and on what they called historia (including research in the medical literature). Nor does their invocation of Hippocrates as authority lack some justification, at least in so far as a broadly similar methodology and similar modes of reasoning can indeed be found in the Hippocratic Corpus. We do not have sufficient primary evidence to evaluate the practice of taking case histories in the post-Hippocratic period (though we know of some observed by Erasistratus, e.g.). By the time we come to Galen, however, there is a striking decline in the meticulousness and systematicity of the cases he refers to in his Prognosis compared with the Hippocratic Epidemics, and that despite Galen’s professed great admiration for ‘Hippocrates’ and his deep involvement as commentator on those treatises.

The Wife of Philinus

347

But if one doubt may concern the match of those generalizations with their evidential basis, downstream as it were, there is the second problem, upstream, of how applicable the lessons of any particular Constitution will be to other Constitutions or other individual cases the doctor may face. The Hippocratics certainly collected an impressive mass of particular observations. The problem was how to extract useful information from them. In a later age, such data could be submitted to statistical analysis to determine the probabilities of various correlations or connections.¹⁴ But that was not within the intellectual horizon of anyone before the seventeenth century (Hacking 1975). The chief notion of likelihood in play in both Aristotle and the Hippocratics was a matter of what held ‘for the most part’, and as is now generally agreed, the ‘probabilities’ they were concerned with under the rubric of the pithanon were not the statistical notion, but what was persuasive. Yet even if, nowadays, medical examinations can draw on the information obtained from batteries of tests, the two main movements of thought involved in arriving at a diagnosis remain, on the one hand to work forwards from the signs the patient presents to the identification of the disease he or she is suffering from, and on the other to work back from the known or supposedly known characteristics of diseases to ascertain which the patient’s symptoms match. The first is still dealt with extensively in textbooks of differential diagnosis (such as that of Herbert French, ed. Bouchier et al. 1996), the second in handbooks of medicine (such as that of Wheeler and Jack, the twelfth edition of which appeared in 1963). Thus French’s discussion of haematuria subdivides what it may signify under three main headings: I, infections of some part of the urinary organs; II, diseases of the neighbouring viscera involving the urinary organs; and III, general diseases.¹⁵ Conversely Wheeler and Jack deal with specific infectious diseases, such as typhus and malaria, by setting out the etiology, morbid anatomy, symptoms, complications, and treatment. On the basis of very different knowledge, Prognostic, as we saw, sets out the signs in the urine and what they indicate, while treatises such as On Internal Affections attempt a general classification and description of the characters of diseases. From the point of view of the Aristotelian analysis, the Hippocratic texts present an extraordinary amalgam, and we have to draw on materials outside Aristotle’s remarks on tekm¯eria and s¯emeia to do justice to that. With regard to the particular medical case histories, one is reminded of Aristotle’s offhand remark, in the Poetics (1451b 2–4), that poetry is more philosophical than history, since ¹⁴ Yet in Major’s handbook on Physical Diagnosis (Delp and Manning 1968: 31) probability statements are graded not only quantitatively (from 0 to 1) but also qualitatively as ‘very probable’, ‘probable’, ‘possible’, and ‘unlikely’. ¹⁵ At a more popular level, the Physician’s Handbook ed. Krupp et al. (1968) discusses the colour of urine in nine categories: (a) quite colourless; (b) milky; (c) orange; (d) red; (e) greenish; (f ) dirty blue or green; (g) dark brown, brown-red, or yellow; (h) brown-yellow or brown-red (if acid) or bright red (if alkaline); and finally (i) brown, brown-black, or black (Krupp et al. 1968: 209).

348

G. E. R. Lloyd

history just deals with the particular—a remark that singularly underestimates the ambitions in both Herodotus and Thucydides to derive general lessons from their narrative accounts. But at the other end of the spectrum of possibilities, the Hippocratics were some way away from being able to give syllogistic demonstrations based on middle terms, even if they had wanted to. Such universal necessary propositions as the doctor had to deal with (of the type: a wound penetrating the heart causes death) were unproblematic, but not very helpful, and medical practitioners would be even less concerned with analytic truths such as that which states that the patient with fever is ill. But faced with a paucity of Aristotelian tekm¯eria, did the Hippocratics not deploy plenty of s¯emeia in their medical reasoning? They certainly proposed many, indeed sometimes unqualified generalizations (as if they could rate as tekm¯eria), but more often qualified ones. But the recurrent problems were first to establish regularities and correlations, secondly to discriminate those that were merely coincidental from those that gave insight into a causal connection, and thirdly, when that was the case, to determine the direction of causality—that is, to distinguish causes from effects. These are the points at which the reputableness of the argument based on the sign—the quality of the inference—has to be judged. Everything depended on the skill and experience of the practitioner. But as often as not, the necessary knowledge of those causal connections was lacking, and they were reduced to mere conjecture. The Hippocratics wrestled not just with the possible connections between single signs and signifieds, but with groups of them, the syndromes in other words. The issue of when a sign or a set of them has a single significatum lurks in the background and sometimes surfaces in the texts,¹⁶ as too does the problem of whether a correlation is significant, that is reflects an underlying causal connection, or when it is merely coincidental.¹⁷ In Epidemics I ch. 9, L II 660 6 ff. the writer says: ‘concerning the crises, the factors from which we drew our diagnoses, were sometimes similar, sometimes dissimilar’. In Epidemics VI 8 26, L V 352 16 ff. the writer says: ‘for good physicians similarities produce vagueness and bafflement, but so too do opposites. What kind of explanation is there? It is difficult to reason things out, even for someone who knows the methods.’ Admittedly the context here is the practice of physiognomy: but the observation of the difficulty of interpretation has general applicability. There is thus a recurrent tension within the Hippocratic texts between the urge to generalize and the realization of its difficulty. The snappy dicta of the Aphorisms had the merit of memorability—but were often subject to exceptions and on ¹⁶ As we saw (above p. 344) Prognostic recognizes, e.g., that the character of the urine may reflect just a condition of the bladder or the kidneys, not that of the patient’s body as a whole. ¹⁷ On Ancient Medicine ch. 21, CMG I 1, 52 15 ff., and On Regimen III ch. 70, CMG I 2 4, 202 11 ff., both warn the doctor not to be taken in by coincidences when judging the causes of complaints; cf. also On Regimen in Acute Diseases ch. 11, L II 314 12 ff.

The Wife of Philinus

349

occasion may well have gone far beyond the available evidence. The price the practitioners were always paying for their generalizations (whether this was recognized or not) was the elision of some of the possibly relevant information in the original data. Giving the patients’ individual histories in full—itself a matter of the selection of relevant factors, to be sure—had the converse disadvantage that the doctor-reader would be left with the task of applying what was learnt to other cases, spotting the similarities, and making the necessary allowances for the differences. The doctor had no algorithm to tell him what was significant among all the observations he made. Aristotle’s analysis of the different types of signs allowed that there were some which were not formally valid, but that could be the basis of reasonable arguments nevertheless. But to decide when he had a probable connection, when a mere coincidence, on his hands, depended on the doctor using what the Greeks would have called his m¯etis, or cunning intelligence. On that issue, it is not what Aristotle has to say about the logic of signs that is most relevant, but rather his discussions of practical reasoning. It is well known that he frequently uses medical analogies to illustrate this, not least in connection with his analysis of what the phronimos or the spoudaios would decide. Those analogies more often relate to arriving at the correct decisions about treatment, rather than to diagnosis or prognosis, and of course to invoke the phronimos to resolve the issue is just to push the problem one stage further back, for we need to have a clear idea of how that skill in reasoning is to be acquired and applied. That is just another way of saying that we are dealing with areas where strict deduction is only of limited use. Even though Aristotle does not mention medical case histories as such, his repeated insistence on the doctor treating individual patients suggests that he saw that experience as more valuable in the training of the effective practitioner than exposure to the already processed, but often shaky, generalizations with which the Hippocratic texts abound. What the general dicta needed—and sometimes received—were heavy qualifications underlining their tentative nature or limited applicability. There was no alternative, to acquire the necessary medical skill, or m¯etis, but to concentrate on the individual patients in all their baffling complexity. It is striking that whereas the analysis of signs and of their role in medical methodology continued to be much discussed in the Hellenistic period and later, the actual practice of taking an individual patient’s case history never improved on those such as that of the wife of Philinus that we have in the Hippocratic Corpus.

REFERENCES Allen, James (2001), Inference from Signs: Ancient Debates about the Nature of Evidence. Oxford. Bouchier, A. D., Ellis, H., and Fleming, P. R. (1996), French’s Index of Differential Diagnosis, 13th edn. Oxford. (1st edn., Herbert French, An Index of Differential Diagnosis and Main Symptoms, 1912).

350

G. E. R. Lloyd

Burnyeat, M. F. (1982), ‘The Origins of Non-deductive Inference’, in J. Barnes, J. Brunschwig, M. Burnyeat, and M. Schofield, (eds.), Science and Speculation (Cambridge), 193–238. (1994) ‘Enthymeme: Aristotle on the Logic of Persuasion’, in D. J. Furley and A. Nehamas, (eds.), Essays on Aristotle’s Rhetoric (Princeton), 3–55. Deichgräber, K. (1933), Die Epidemien und das Corpus Hippocraticum, Abhandlungen der preussischen Akademie der Wissenschaften, Jahrgang 1933, 3, phil.-hist. Kl. Berlin. Delp, M. H., and Manning, R. Y. (1968) (eds.), Major’s Physical Diagnosis. Philadelphia. Diller, H. (1932), ‘opsis adelon ta phainomena’, Hermes, 67: 14–42. Grensemann, H. (1975) Knidische Medizin, i, Ars Medica Abt. II Bd 4,1. Berlin. (1987), Knidische Medizin, ii. Stuttgart. Grmek, M. D. (1989), Diseases in the Ancient Greek World, trans. M. Muellner and L. Muellner of Les Maladies à l’aube de la civilisation occidentale (Paris, 1983), Baltimore. Hacking, I. (1975), The Emergence of Probability. Cambridge. Hanzawa, C. (1987), ‘Epidemics VII: Translation and Commentary’, (M.Phil. dissertation, Cambridge). Jones, W. H. S. (1923–31), Hippocrates, Loeb edn. vols. i–iv. London and Cambridge, Mass. Jouanna, J. (1974), Hippocrate: Pour une archéologie de l’école de Cnide. Paris. Krupp, M. A., Sweet, N. J., Jawetz, E., and Biglieri, E. F. (1968) (eds.), Physician’s Handbook, 15th edn., Los Altos, Calif. (1st edn. by J. Warkentin and J. D. Lange, 1950.) Littré, E. (1839–61), Oeuvres complètes d’Hippocrate, 10 vols. Paris. Lloyd, G. E. R. (1966), Polarity and Analogy. Cambridge. (1983), Science, Folklore and Ideology. Cambridge. (1987), The Revolutions of Wisdom. Berkeley. (1991), Methods and Problems in Greek Science. Cambridge. (2002), The Ambitions of Curiosity. Cambridge. Lonie, I. M. (1978), ‘Cos versus Cnidus and the Historians’, History of Science, 16: 42–75, 77–92. Major, R. H. (1957), ‘How Hippocrates Made his Diagnoses’, International Record of Medicine, 170/9: 479–85. Siegel, R. E. (1964), ‘Clinical Observations in Hippocrates: An Essay on the Evolution of the Diagnostic Art’, Journal of the Mount Sinai Hospital, 31: 285–303. Smith, W. D. (1973), ‘Galen on Coans versus Cnidians’, Bulletin of the History of Medicine, 47: 569–85. Thivel, A. (1981), Cnide et Cos. Paris. Wheeler, A. and Jack, W. R. (1963), The Students’ Handbook of Medicine, 12th edn. Edinburgh. (1st edn. Edinburgh, 1894.)

18 Self-Refutation and the Sorites R. J. Hankinson

And it [sc. the sorites] also opposes him who speaks and him who argues with it. Galen, On Medical Experience, ix, p. 98 W¹ Why, then, do you persist with your slander, and say that experience is incoherent, claiming that ‘very many times’ is indefinite, and that it is unclear where it comes to a halt? In saying that you cannot discern where, do you think that you’re refuting us rather than yourself? It seems to me you rather refute yourself. There are two claims at issue here: (i) that things are discovered by observation alone,² I affirm, and you admit, albeit reluctantly; but (ii) how this occurs by observation alone, I say it is neither possible nor useful to discover, while your remaining task is to discover it. So in proving it puzzling by your sophism, you refute yourself but do no harm to us. Ibid. xv, p. 113 W

More than twenty years ago, Myles Burnyeat (1982) published an important article on the sorites;³ a few years earlier he had written two ground-breaking ¹ On Medical Experience, one of our most important ancient sources for the sorites, survives for the most part only in a ninth-century Arabic translation of an earlier Syriac version of Galen’s Greek, of which only a few fragments remain; this makes for some difficulties of interpretation, but fewer than one might antecedently suppose after such a multi-lingual centuries-long game of Chinese Whispers, which is itself a tribute to the skill and ability of the translators. Med. Exp. is edited by Walzer (1944) (= W); Walzer’s English version of the Arabic, along with an English translation of the surviving Greek sections (which Walzer left in Greek) by Michael Frede, is printed in Frede 1985. I should mention here that the epigraph is usually taken to refer to the sorites—but its ambit is rather greater than that: see below. Other texts of Galen are cited by way of the edition of Kühn (1821–33) (= K), even where later and better editions exist. ² This quotation is taken from one of the surviving Greek fragments; Frede translates ‘is found out by a single observation’, which is certainly possible, both as regards Greek and sense: much of the earlier discussion turned on how a single instance could make the difference between a generalization’s being technical or not. But that question is not precisely germane to the more general point at issue here: viz. whether a series of observations may ever on its own, and in default of theory, suffice to generate technical knowledge. ³ ‘Gods and Heaps’.

352

R. J. Hankinson

pieces on Protagorean self-refutation.⁴ So, if nothing else, at least the title of this article is an appropriate one for something written in honour of and out of admiration for my former graduate supervisor. My aim is relatively modest. I do not seek to discern a generalized ancient ‘solution’ to the ‘paradox’,⁵ or to offer one in my own voice (although I will make one or two tentative suggestions). My main purpose is to take seriously a passing hint of Cicero’s, one which has generally been either ignored or treated simply as a joke,⁶ that the sorites may be applied to itself, or to the more general notion of coherent argument, with potentially devastating results. I say ‘potentially devastating’, since it is unclear whether the threatened disaster must occur, and if it does, which of the opposing sides in the dispute it is potentially disastrous for. As such, the debate as I shall reconstruct it mirrors that between Sceptics and Dogmatists as to the question of whether sceptical arguments against argument are self-refuting, and if so, what the appropriate upshot of that ought to be.⁷

I First some general (and largely familiar) background. The sorites, the ‘paradox of the heap’,⁸ is one of a family of puzzles attributed to the eristic Eubulides by Diogenes Laertius (DL 2. 108).⁹ Chrysippus apparently treated of the sorites at length; Diogenes’ catalogue records a work ‘On Soritical Arguments applied to Words, three books’, as well as an ‘On the Little-by-Little Argument, to Stesagoras’¹⁰ ⁴ ‘Protagoras and Self-Refutation in Later Greek Philosophy’ (1976a); ‘Protagoras and SelfRefutation in Plato’ (1976b). ⁵ ‘Paradox’ in scare-quotes since the puzzle does not have the logical form of a true paradox: viz. of a sentence that is true if and only if it is false (rather, it has the form of an apparently sound argument with an apparently false conclusion: of course that’s bad enough); ‘solution’ similarly flagged because I am doubtful whether the puzzle—or puzzles—as such can yield to a single authoritative diagnosis. But these issues are largely irrelevant to the concerns of the present paper. ⁶ A partial exception to this general neglect is Myles himself (Burnyeat 1982: 327 n. 33). It was thinking about this footnote (and that of Barnes (1982: 57 n. 60)) that first started me down the road followed by this paper more than twenty years ago. ⁷ See Sext. Pyr. 2. 185–92; Math. 8. 465–80; and see McPherran 1987. ⁸ Or, more accurately, the ‘heaper’, or ‘accumulator’: as Barnes stresses, adding the ‘-it¯es’ suffix to the root word (here s¯oros, meaning heap) generally (although not invariably) signifies agency: Barnes 1982: 32–3 n. 18; see also Burnyeat 1982: 316 n. 3. ⁹ This attribution is not uncontested; and Diogenes’ testimony is vitiated by the fact that, a page or so after ascribing them to Eubulides, he reassigns two of the puzzles (the ‘Veiled Man’ and the ‘Horns’) to Diodorus Cronus (the misattribution of the ‘Horns’, along with several other sophisms, to Chrysippus at DL 7. 187 obviously rests on a misconstrual of the dialectical context in which Chrysippus discussed them). ¹⁰ Or ‘against Stesagoras’: the preposition ‘pros’ is ambiguous between the two meanings. Greek philosophical practice was notoriously disputatious (and nothing else is known about this Stesagoras); but the fact that a high proportion of Chrysippus’ huge output is recorded as being pros a large variety of different individuals, suggests that they are probably addressees, either pupils or associates. Two other works in Diogenes’ list are also addressed pros St¯esagoran.

Self-Refutation and the Sorites

353

in two,¹¹ and not all of that can be put down to Chrysippus’ legendary verbosity. If the length of the treatment he gave the sorites does not equal that accorded to the Liar,¹² it is still considerable. Whatever Eubulides’ intentions in originally formulating it,¹³ Chrysippus clearly took the sorites seriously as a potentially serious challenge to the coherence of his own philosophy, perhaps second only in importance to the Liar; and this assessment is corroborated, indirectly, by Cicero: in his presentation of the Academic attack on Stoic logic it is the sorites and the Liar that receive detailed treatment (Acad . II 91–8). From a contemporary standpoint, that concentration makes a great deal of sense; for if the Liar threatens semantics by undermining the Stoic commitment to the Principle of Bivalence, the sorites can be seen as posing a challenge to the orthodox logical syntax of modus ponens. Or so I intend to argue.¹⁴ The argument form of modus ponens¹⁵ is central to formal discussion of the sorites, although Galen in fact presents it informally, in question-and-answer form: I am going to tell you how, after due reflection and investigation, it has become evident to me that there is no criterion by which a thing may be judged as having been seen very many times. … For they say that a thing seen but once cannot be accepted or regarded as true, neither what was seen a few times only. They believe something can be accepted and considered true if it has been seen very many times, and in the same manner every time. I would ask them, therefore, if that which has been seen ten times is included in that which has been seen very many times, and their answer to this is ‘No’. Then I would say to them: ‘And what has been seen eleven times?’—and they say ‘No’. Then I would ask them further about a thing that has been seen twelve times—and they say ‘No’. … And so I never cease asking and adding another number to each until I reach a high number. Nothing remains for him thus questioned in any given case except either to deny that the ¹¹ Diogenes also records an ‘On Arguments directed against Judgements [hupol¯epseis] and on those who are Silent [h¯esuchazontes] to Onetor two books’, which may also have concerned the sorites (on the importance of the ‘methodological injunction’ to ‘keep silent’ in soritical contexts, see Barnes 1982: 49–56; Burnyeat 1982: 333–4; see also Sedley 1977: 91; and it is also possible that ‘Against those who Think that things may be both True and False’, in one book, may have touched on soritical issues, although it will presumably have principally concerned the Liar. It is reasonable to suppose that Chrysippus addressed the sorites in many places in his vast logical œuvre; a brief and fairly unilluminating mention of it survives in the papyrus of his Logical Investigations: SVF II 298a, p. 106, 7–12. ¹² Some seven texts in fifteen books, according to Diogenes: DL 7. 197. ¹³ Barnes (1982: 41 n. 47) rightly notes that ‘we know nothing at all about Eubulides’ motives’; and that Sextus for one presents him as exclusively interested in logic (Math. 7. 13); moreover, although Eubulides is said to have attacked Aristotle (DL 2. 109), nothing is known of the content of the assault; Diogenes simply says that ‘he slandered him [sc. Aristotle] greatly’. ¹⁴ Of course, the sorites can be seen to have predominantly semantic import; such is the general trend of the discussions of the issue that have been carried on in the analytic tradition over the past thirty years or so. Cf. e.g. Wright 1975 and 1976, the latter an article cited approvingly by Burnyeat (1982: 319 n. 13, 330 n. 34); Heck 2003; Dummett 1975; and equally the Liar can be seen as posing syntactical problems: see below, pp. 362 n. 45; 370–1. ¹⁵ Although not necessarily modus ponendo ponens: the argument can also be formulated by way of a sequence of arguments of the form modus ponendo tollens. See below, p. 356 ff.

354

R. J. Hankinson

number has reached the limit when one can say it constitutes very many times, or, should he admit that it has, to make himself a laughing-stock, since he would thus require people to allow him a number fixed solely by himself, and a decision made by him alone. For one might say to him: Why, for example, should anything that was seen fifty times be regarded as having been seen very many times, rather than that which was seen forty-nine times? In asserting this, you assert two mutually contradictory things. You previously acknowledged that what was seen once is not to be accepted … as true. But now we see you allow that it is acceptable and is to be considered true. For if something that was seen forty-nine times, and yet in all those times was not accepted or considered true, now by the addition of this one instance comes to be considered acceptable and true, it is evident that only having been seen one single time has it become acceptable and true. (Med. Exp. vii, pp. 96–7 W; cf. xv–xvii, pp. 112–19 W; xx, pp. 124–6 W; trans. Walzer, with alterations)

The speaker is a Dogmatic doctor (Galen claims to be reproducing an argument of the second-century  physician Asclepiades of Bithynia: ibid. ii, p. 97 W);¹⁶ ‘they’ are their Empiricist opponents (‘Menodotus if you like, or Serapion or Theodosius’: Galen’s piece is a school exercise).¹⁷ This informal, erotetic type of presentation is also to be found in non-philosophical contexts (e.g. Horace, Epistles 2. 1, 36–49; cf. Persius 7, 75–80); indeed, the appearance of the puzzle in such places without learned exposition is testimony to the extent to which it had penetrated (relatively) popular culture.¹⁸ It should be noted, however, that Galen’s Dogmatist is making a specific point against Empiricist methodology, and one that is not generalizable to all soritical cases. For the Empiricists, one swallow does not make a summer (or a spring, if you’re a Greek: Aristotle, Eth. Nic. I 7, 1098a 18–19); the Dogmatist then contends that if that is so, the addition of a single swallow to any collection can’t make one either. The point does not directly concern the alleged indeterminacy of ‘very many’, the supposedly soritical predicate in question.¹⁹ Galen’s interest in Med. Exp. is purely dialectical and expository (although this does not preclude him from making telling points in the voice of the Empiricist). ¹⁶ For Asclepiades’ medical stance, see Vallance 1990; for his dates see Rawson 1982. ¹⁷ That Dogmatist doctors should make use of the argument form is itself interesting, since Dogmatists are supposed to be interested in logic, even if the majority of them are no good at it (at least in Galen’s jaundiced view: On the Therapeutic Method X 28–30, 32, 109, 122 K; On Mixtures I 590–1 K); but, as we shall see (and as Galen’s Empiricist perhaps hints in the epigraph to this paper), the sorites is potentially devastating to logic. They seek to show that the Empiricists’ key concept of experience is infected with vagueness, and hence of no use in properly scientific contexts. But in order to do so, they must consider that, at least in this context, the argument pattern is sound: there is no possibility of a dialectical deployment of the argument against the Empiricists. ¹⁸ See Barnes’s invaluable appendix to his 1982 article, on texts referring to the sorites. ¹⁹ For all that, the argument certainly can be recast in canonical soritical form; and Galen himself refers to his treatment here as involving the sorites, both later in Med. Exp. (xvi, p. 115 W), and at Subf. Emp. 3, pp. 46–7 Deichgräber: ‘Each ‘‘experience’’ (empeiria) consists of many individual experiences; but the question on how many it rests does not allow of a definitive answer, and is subject to the puzzle some call the ‘‘sorites’’. This puzzle is discussed in more detail in another book called On Medical Experience.’

Self-Refutation and the Sorites

355

Elsewhere, in his own persona as representative of a sophisticated theoretical (and hence in a sense Dogmatic)²⁰ medicine, Galen adverts to the sorites in the course of discussing the vexed ancient question of when a deviation from the norm may properly be considered a disease (On Affected Parts VIII 25–30 K). He dismisses the question as being essentially terminological; but he also asserts that ‘the argument known as the sorites will not trouble us here’, since there are two sorts of generation, of uniform and of non-uniform things. In the case of the former, as soon as something has begun to be generated (i.e. as soon as some of it has come into being),²¹ it has been generated (for it is definitional of the uniform that any part of any uniform stuff is that same uniform stuff ); while for non-uniform things such as houses, generation of the substance cannot be said to have taken place until the last brick is in place. Thus, in the case of generation so defined, there will always be a determinate answer to the question: When do we have an F?, although when will differ from item to item according to its constitution; and equally there will always be a determinate answer to the question: When are we starting to get an F?²² Galen apparently thinks that this provides a general solution to soritical problems, at least as they are taken to affect serious questions in physics (although the text is a little unclear on this point).²³ He cites the example of the constant drip of water that imperceptibly hollows out the stone (one elsewhere cited in soritical contexts: the quotation is from Choerilus of Samos: Fr. 10), and argues that each drip must have (roughly) the same causal power: thus if it takes a thousand drips to produce a perceptible indentation, each drip must produce one thousandth of a perceptible indentation. The same is held to apply in the case of distemper; any deviation from the ideal is a form of disease, although it may not be perceivable as such.²⁴ ²⁰ Although Galen himself, the avowed enemy of all sects, would have resisted the characterization: On Affected Parts VIII 144 K; On the Order of My Own Books XIX 51–4 K; On the Diagnosis and Cure of the Errors of the Soul V 93 K. ²¹ This rider is necessary since there can be a mode of coming into being for a uniform stuff in which the source stuff changes gradually, and as a whole, into it—and in this case this condition is not satisfied (in fact this turns out to be an example of the second kind of generation). ²² We may ignore here problems created, for a metaphysics of continuity, by questions regarding the first moment of change: in the ancient context, see Sorabji 1979. ²³ Moreover, his Empiricist in Med. Exp. robustly asserts that ordinary language is incurably affected with vagueness, and hence, if the Dogmatist’s sorites works, nonsensical: ‘according to what is demanded by the argument [sc. the sorites], there cannot be any such thing in the world as a heap of grain, abundance or satiety, a mountain, strong love, a row, a strong wind, a city, nor anything else whose name and concept indicates it to have a measure of extension or plurality, such as a wave, the ocean, a flock of sheep, a herd of cattle, a nation or a crowd’ (xvi, p. 114 W; cf. xvii, pp. 75–8 W); as Burnyeat (1982: 315) points out, the ancients knew that the sorites was about more than just heaps. ²⁴ Thus Galen theoretically aligns himself with the doctrine of ‘perpetual suffering’ (On Mixtures I 676–7 K); and at On the Differences of Symptoms VII 50 K he characterizes disease as ‘a disposition of the body which is such as primarily to impair one of its functions’, although he will on occasion define disease as perceptible damage to some natural function (e.g. Ars Medica I 379 K).

356

R. J. Hankinson

This may be all very well, but it hardly counts as a general solution to all cases where the sorites threatens, as his own Empiricist would have pointed out. There will still be love, crowds, and baldness (see n. 23), and no obvious determinate, quantitative point at which any of them definitively come into being. None of the genuinely soritical predicates are uniform, obviously; but none of them are determinately non-uniform either. Galen’s distinction is thus of limited value.²⁵ II Diogenes Laertius, no doubt relying upon some handbook of Stoic logic, presents the argument in a more formal (or at least more readily formalizable) version: It is not the case that two are few and three are not; it is not the case that these [sc. three] are few and four are not; and so on up to ten [or: ten thousand]; but two are few; so then are ten [or: ten thousand].²⁶ (DL 7. 82)

That (with a little charity) yields the following logical structure (where ‘F ’ abbreviates ‘are few’ and ‘O’ names some particular type of object): [A] (1) F (2.O) (2) ¬(F (2.O) & ¬F (3.O)) ²⁵ Although it perhaps suggests a different sort of answer to the puzzle. Suppose we define ‘disease’ as a perceptible impairment of function (above, n. 24); then what counts as being a case of disease will presumably vary from individual to individual, partly as a result of straightforward differences in perceptual acuity, but partly also for reasons of interest: the point at which arthritis of the fingers meets this criterion will presumably be different for a concert pianist than for an interior decorator, since they use their hands for different, and differently grained, functions. But again, such a response seems to miss the point. My concern about my own increasing baldness might have been intensified, and my disposition to assent to the proposition ‘I am now bald’ triggered at a different (earlier) point, had I chosen the profession of male model; but even then I would still have been at a loss to quantify the precise amount of hair loss required to constitute the condition. The range of vagueness may shift in an interest-relative manner, but the vagueness itself is not thereby eliminated. ²⁶ Is it 10 or 10,000? The MSS have 10; Egli, mindful of the traditional length of soritical argument (see Barnes 1982: 27–8; cf. Sextus, Math. 7. 416–21), emended to 10,000, which Barnes accepts, as does Burnyeat (1982: 320 n. 15). Long and Sedley (1987: ii. 224, = 37A), retain the MSS reading, noting that ‘Estimates of the borderline between few and many vary alarmingly. … Ten is certainly many for children in a family or courses in a meal.’ As far as logic is concerned, it doesn’t really matter; but in favour of ‘10,000’ is the fact that in Sextus’s general presentation of the argument at Math. 7. 418, ‘fifty is few’—with no specified domain of objects—is given as an uncontroversial case of something cataleptically apprehended (for the importance of the notion of cataleptic impression in these contexts, see further below); moreover, sequences of relatively few (e.g. eight) members are unlikely to be obviously soritical for any candidate predicate—in other words, it will be less implausible to suppose that, for a such a sequence, there will be an agreed point (four, say) where the number of children—or courses for that matter—ceases to be few. And while one might reasonably contend that whether or not they were was dependent upon social context (four children are few for Mormons, Sultans, or Augustus the Strong of Poland; four courses a modest repast for a Georgian gentleman), that fact (or set of facts) on its own will not revive the soritical problem as such (since the answers may still be determinate for determinate social contexts: but cf. n. 25 above), although no doubt it may raise other puzzles.

Self-Refutation and the Sorites

357

(3) ¬(F (3.O) & ¬F (4.O)) (4) ¬(F (4.O) & ¬F (5.O)) (5) ¬(F (5.O) & ¬F (6.O)) (6) ¬(F (6.O) & ¬F (7.O)) (7) ¬(F (7.O) & ¬F (8.O)) (8) ¬(F (8.O) & ¬F (9.O)) (9) ¬(F (9.O) & ¬F (10.O)) so (10) F (10.O) (or … .. (10,000) F (10,000.O): see n. 27) The conclusion (‘10,000 grains of sand are few’) is supposed to be evidently absurd. Whether or not it is depends on the identity of the Os as well as the length (and the numerical intervals)²⁷ of the sequence, but that in itself does not matter. The successful paradoxographer is certainly under no obligation to show that all predicates are soritical for all sequences. We will be in quite enough trouble if he can show that any (or at least any useful and apparently ineliminable ones) are, for some. So let us allow that, in a particular case (why not take the traditional one of the heap?), it is absurd not to suppose that 10,000 (suitably aggregated: this condition is not of course trivial) grains of sand must constitute a heap. [A] is formulated in terms of negated conjunctions, which is also how Diogenes presents it, and hence it is reasonable at first sight to infer that the Stoics (whose logic he is reporting) adopted this form of presentation too. Why (indeed ²⁷ See Barnes 1982: 31 n. 14: ‘ ‘‘F( )’’ may satisfy the soritical conditions for some sequences of subjects but not for others. (‘‘Small’’ might be soritical for the sequence but not for the sequence )’ (cf. Math. 7. 416). It should also be pointed out that the schema of the particular form of [A] will not fit every argument the ancients labelled a sorites, most obviously Carneades’ notorious theological sorites (Math. 9. 182–90), discussed by Myles in his (1982); what links the various subspecies of ‘soritical’ argument considered in this more general sense (at any rate in principle, and so far as the more serious ones are concerned) is the initial plausibility of the supposition that if the predicate applies to the subject of the first proposition of the complex, then it ought to do so to the second as well—and so on throughout, although in the non-quantitative cases, the reasons why each particular premiss is held to be compelling will differ. See Burnyeat 1982: 326–33. Still, I am inclined to agree with Bobzien (2002: 227 n. 36), against Burnyeat (and explicitly Williamson (1994: 21, 25)) that such a usage is probably post-Chrysippean; I imagine it to derive from an Academic (presumably Carneadean) attempt to entrap the Stoics into admitting that some arguments which they themselves endorsed were in fact properly classified as sorites, and hence that their insistence that the sorites was a fallacy was incoherent. There is no evidence of any Stoic producing a canonical, numerical sorites in their own voice. See further below, n. 44. Note that while Burnyeat (1982: 327) writes that ‘the point I have been labouring is that the ancients’ conception of the sorites is in important respects indeterminate’, in such a way as to allow for ‘non-canonical’ sorites, he still allows it unlikely that Chrysippus would have labelled the sort of little-by-little arguments he approves ‘sorites’ ‘because he restricts this term to cases—above all, the quantitative cases—where the conclusion of a little by little argument is manifestly false. … ‘‘sorites’’ is indeed the name of something unsound’ (1982: 338).

358

R. J. Hankinson

whether) they would have done so is controversial;²⁸ but if this is the genuine Stoic formulation, then it will have been undertaken in order to make it clear that it is merely a contingent fact that the two propositions cannot be conjoined: in other words, there should be nothing in the nature of two’s being few (for a particular domain) that precludes (for some strong sense of ‘preclude’) three’s not being few, even though as a matter of fact if two are, three will be (Chrysippus made similar moves in the case of astrological ‘theorems’ in order to emphasize that the connection between antecedent and consequent was merely that of sign to signified: Cic. Fat. 12–15).²⁹ The important point is that evidently any such chain of premisses will, in the ordinary run of things,³⁰ validly entail its conclusion.³¹ For a soritical sequence to be unsound, at least one (and, on a material interpretation, at most one)³² of the conditional premisses must be false; but it seems reasonable to suppose that (in the troubling cases at least) this will not be because there is something in the nature of the appropriate number n which, in the case of a particular predicate F and type of object O, makes it the case that n Os are F , while n + 1 are not. Of course, if it could reasonably be made out that there was a conceptual, even a logical, connection between antecedent and consequent in ²⁸ See the dispute on which formulation counts as ‘weaker’, and which of the parties to the dispute might have proposed them, between Barnes (1982: 28–9 n. 12) and Sedley (1977: 91 n. 97); cf. Sedley 1982: 255 n. 41. ²⁹ For the technical issues in Stoic logic involved here, in particular regarding the interpretation of their strong conditional (that with which they standardly formulate their first indemonstrable, and which supplies the correct interpretation of the conditionals used in valid indicative sign inferences: Sext. Pyr. 2. 104–33, esp. 111), see Frede 1974: 80 ff.; Stopper 1983. See also Burnyeat (1982: 323–4), who finds the insistence on ‘Philonian’ (i.e. material) truth conditions for the conditional premisses surprising: ‘Not only does the Stoic logician apparently refuse to accept that there are conceptual or semantic pressures on us to accept the premisses of a sorites. He refuses this for the very cases where the idea is most compelling, because of the patently quantitative nature of the predicates involved’—but as he himself notes, the bare schema (‘two are few’, etc.) needs filling out in particular contexts (‘two are few for hairs on a head’); and there will be different intuitions concerning the truth of the various premisses so filled out in different circumstances. See further Williamson 1994: 22–6. ³⁰ i.e., if classical logic, involving bivalence and the transitivity of inference, holds: however, see further below, pp. 368–71. ³¹ For the Stoics, by way of the multiple application of one of their meta-logical themata, in this case the third: for the themata, see Bobzien 2003: 110–20 (for their application in this context, see Williamson 1994: 23); the alternative is to suppose that in some cases modus ponens fails, which is a very high price to pay for ‘solving’ the paradox, although it may not in some cases be intolerable: see further below, pp. 361, 369–70. ³² Since there will evidently be only one in which (for a conditional formulation) the antecedent is true and the consequent false; or (for the negated conjunction case) only one case in which it will be both true that F (n.O) and that ¬F (n + 1.O), and hence that the nth negated conjunction is false. Burnyeat (1982: 324 n. 25) writes: ‘if there is more than one false premiss, the truth conditions for material implication ensure that no two of them are adjacent’—which is true—but the material context of soritical sequences, at least those involving quantity, ensure that at most one premiss may be false. It couldn’t both be the case, e.g., that 12 are few but 13 aren’t and that 27 are few but 28 aren’t for the same sequence: in other words, as Williamson (1994: 13) put it, ‘the usual sorites sequences are monotonic’.

Self-Refutation and the Sorites

359

the soritical conditional premisses, then the argument would be (in one sense at least)³³ stronger: we would have less room for manœuvre in denying the (apparently absurd) conclusion. Yet it is precisely because there seems to be no obvious conflict³⁴ between affirming, say, that four are few and five are not, that the material formulation is appropriate; any set of four objects is clearly distinct from any five-membered set of the same type of objects, set identity being extensionally determined: and the same should hold for any n and n + 1; and hence no fact about the nature of the set should determine any equivalent fact about its successor.³⁵ On the other hand, it is reasonable to suppose that in some cases at least the Stoics would have held the connection between antecedent and consequent in the conditional premisses to be strong enough to support the Chrysippean formulation, at any rate for the Stoic sage: for the Stoics notoriously supposed that it was possible (for some sense of ‘possible’) for human beings to attain a level of perceptual and intellectual discrimination such that they could never be fooled by apparently identical sense reports: the criterial cataleptic impressions are always in principle discriminable from their false congeners;³⁶ and nor could they ever fail to undertake the correct course of action in any particular practical circumstance.³⁷ Sextus, indeed, presents a soritical argument in the context of attacking the notion of the cataleptic impression itself (Math. 7. 416–21): suppose that ‘fifty are few’ is the last cataleptically apprehended proposition of the soritical sequence which ends with the evidently false ‘10,000 are few’: then it is clear, presumably, that he [the sage] will assent to ‘fifty-one are few’ [the first ‘non-cataleptic impression’: note that Sextus does not say that it is false], since there is nothing between this one and ‘fifty are few’ (Math. 7. 419).³⁸

‘There is nothing between’ them presumably in the sense that the two propositions are indistinguishable from one another in respect of the predicate ‘few’ (perhaps as applied to some determinate class of objects: Sextus’ account is unclear on this point, and the argument is at least presented as involving the uninterpreted ³³ I am thinking again of the dispute regarding strength between Barnes and Sedley: see n. 28 above. ³⁴ Mach¯e: the criterion for the strong, ‘Chrysippean’, conditional: see Brunschwig 1980; Barnes 1980; Frede 1974; and Bobzien 1999; Bobzien (2003: 94–5) gives a concise summation. ³⁵ Thus if the argument is formulated in terms of strong conditionals, all of them may be false, since it is not because two are few that three are, and so on, which derails the argument at the outset; but we will still be left with the puzzle as it apparently applies to vague predicates in ordinary language. ³⁶ On the interpretation of the doctrine of the cataleptic impression, see Frede 1983; Annas 1990; Hankinson 1997, 2003. ³⁷ The literature on these issues is large: on sagehood and moral psychology, see Brennan 2003. ³⁸ It is worth noting that here we have an argument in a slightly different form from that which I have been calling the ‘canonical’ sorites; the latter deals in collective predicates (how many F s make a G?), while this apparently concerns manifold quantification (how many is many (for an F ?)); I am not sure whether this is significant.

360

R. J. Hankinson

predicate ‘few’). This is a central feature of all soritical argument, the claim that the difference between any two steps on the ladder is in the appropriate sense negligible, or that the two items are, from the appropriate standpoint (i.e. relative to the predicate at issue in the sequence) indiscernible. Call this the ‘Indiscernibility Thesis’, or IT. But why should not Stoics simply deny this, at least in the case of the sage, whose perceptual apparatus is tuned to the finest possible pitch of discrimination? Indeed, the Stoics claimed that the sage could distinguish between objects that were, to others, indistinguishable: even two hairs or two grains of sand (Acad . II 85).³⁹ On the other hand, no Stoic text explicitly says that all distinctions will be discriminable to the sage, and as Sextus at least presents it, Chrysippus’ sage will stop assenting to the conditionals in a soritical sequence, and before he has reached the last clear case where the conditional is true: In the case of the sorites, when the last cataleptic impression lies right next to (parakeimen¯e) the first non-cataleptic impression, and is pretty much indistinguishable from it (dusdioristos schedon), Chrysippus and his associates say that in cases where the difference between impressions is so small, the sage will come to a halt and keep silent, but when he comes across cases in which the difference is greater, he will assent to the former as being true. (Math. 7. 416)⁴⁰

Most commentators take this to imply that there will be some distinctions beyond even the capacity of the sage to discriminate, and that in those cases the sage will suspend judgement; what the sage will not do is rashly slide down the slippery slope into evident falsity. That may be right; but the text is conditional in tone, and I am inclined to think that Stoics’ own conception of sagehood may have been too indeterminate to suggest a clear answer: perhaps sages will get all (perceptual) discrimination right—or perhaps they won’t, but even if they don’t, they will not be led into falsehood. At any rate, the upshot of this is that, for the ideal cognizer at least, there are no non-cataleptic impressions (or at least none that will incline him to assent). But this suggests that, in a soritical argument with an evidently false conclusion and an evidently true categorical first premiss, there will be one, and indeed determinately one, conditional premiss which is false. And, what is more, for the sage it will be evident that it is so; which further suggests that for the sage at least the conditionals in question can be given their strong Chrysippean truth conditions. Sextus, among other things (the argument is multi-faceted), mounts a soritical argument against the coherence of the crucial Stoic notion of the cataleptic impression itself: there can be no discernible difference between cataleptic and non-cataleptic impressions. The Stoics, if I am right, would reply ³⁹ As Myles notes (Burnyeat 1982: 335), these examples can hardly be chosen accidentally. ⁴⁰ This is a highly puzzling methodological prescription, and has been much discussed (Burnyeat 1982: 333–4; Barnes 1982: 49–56; Bobzien 2002: 228–30, 233–7). I shall have no more to say directly about it.

Self-Refutation and the Sorites

361

that the difference is discernible, even if it is very rarely (perhaps never: the issue turns on whether there are in fact any sages) in fact discerned. The truth is out there; it’s just that in its fine grain it is uncapturable for us mere mortals.⁴¹ All of this of course squares perfectly with the Stoics’ firm commitment to bivalence and classical logic. We will not squirm out of the sorites by claiming that some propositions are neither true nor false, or a little bit of both.⁴²

III But how are we to do it? Barnes (1982: 47 ff.) usefully distinguishes between ‘radical’ and ‘conservative’ opponents of the sorites: the conservative merely seek to show that some sorites are unsound (which, since the argument form is evidently valid, involves claiming that one of the premisses is false); radical opponents, by contrast, propose to revise the canons of sentential logic themselves, by abandoning modus ponens altogether, or at least by restricting its applicability. But that, even in its weaker form, is a very tough bullet to bite on; it risks undermining the most fundamental principle of rationality, that of conditional inference.⁴³ Chrysippus’ opposition will have been of the conservative sort (in so far as he is opposed to the sorites at all; for as Myles has emphasized, if the argument form is valid, we should expect to find some sound, if perhaps surprising, instances of it—some slopes are indeed slippery, and Chrysippus would probably have agreed).⁴⁴ ⁴¹ This suggests one possible resolution to a dispute regarding the interpretation of the cataleptic impression. Frede (1983) argued that the sceptical attack on the notion is an ignoratio elenchi, since the Stoics require only that the cataleptic impression cause the appropriate sort of belief appropriately, not that it be apprehended as doing so; against that, Annas (1990: 195 n. 25) and others have noted that such a grotesque ignoratio should surely not have persisted for several centuries, as it apparently did, and that the Stoics must have been committed to the transparency (to the cognizer) of the cataleptic impression; cf. Hankinson 2003: 73. Now perhaps we can see that both sides are in a sense right: the Stoics are committed to transparency, but only (invariably at least) in the case of the sage. For the rest of us, mere catalepsis (as one might say) will do. This deserves further investigation. ⁴² This is a strong claim to make on Chrysippus’ behalf, perhaps too strong: see Bobzien 2002; n. 45 below. ⁴³ Another alternative suggests itself: we might simply refuse to allow that soritical sequences are properly formulable as chains of conditional inference; the facts, e.g. about baldness, are not such as to be exhibited in sequences involving the addition or subtraction of single hairs, and the associated conditionals. But this seems very unpromising: after all, what is there to baldness other than a relative lack of hairs (perhaps their distribution matters too, as it does in the case of the heap: but that is irrelevant here)? And why can’t you lose them singly? ⁴⁴ He produced arguments in his own voice that others at least called sorites (Plut. Comm. not. 1084c–d, = SVF II 665; but see n. 27 above), and given his commitment to the canons of his own logic, must therefore have supposed their conclusions to be sound, if surprising. The example of a Chrysippean ‘sorites’ given by Plutarch concerns the physical question of what sorts of things are bodies: ‘it is not the case that the night is a body and that the evening, dawn and midday are not bodies; and it is not the case that the day is a body but that the first, tenth, fifteenth and thirtieth

362

R. J. Hankinson

Conservatism of this sort anticipates the modern reaction to the paradox known as epistemicism, championed most vigorously by Williamson; our inability to defuse the sorites in any particular case derives from our ignorance of just where the cut-off occurs: but for all that, it indubitably does somewhere.⁴⁵ Thus classical logic and semantics are after all perfectly appropriate for dealing with vague discourse: the fault lies with us, not with the logic, the facts of the matter or the predicates used to describe them.⁴⁶ Williamson (1997: 279) argues that the evident superiority of classical logic and semantics to their various alternatives ‘in terms of simplicity, power, past success, and integration with theories in other domains’ are such that the burden of proof lies heavily with those who would abandon them, so much so that if that means accepting the relatively unpalatable Chrysippean consequence that, all initial plausibility aside, there really is a single false complex proposition in the series, then so be it. of the month are not bodies; and so the month is a body, and summer, and autumn and winter’. It is worth noting (a) that presentation is fairly informal, but that in its more formal parts it invokes the pattern of the third indemonstrable (see nn. 28–9); and (b) that it exhibits the sort of elasticity noted by Burnyeat in regard to the Carneadean theological ‘sorites’ (Burnyeat 1982: 326–32); i.e. it is not in what I have called the ‘canonical’ form (see n. 27 above). The technical term used by Plutarch is ‘ho kata mikron logos’, the argument from small steps, or little-by-little argument, which is how Chrysippus himself refers to it, by the variant ‘para mikron logos’ in the Logical Investigations (SVF II 298a, p. 106, 9–10). For more examples of argument patterns endorsed by the Stoics which are described as sorites, see Cic. Fin. IV xviii 48–51; but these descriptions all occur in writers more or less hostile to the Stoic position. ⁴⁵ Williamson 1994; in ch. 1 he argues that Chrysippus did indeed pioneer the epistemic response to the sorites. Recently, however, Susanne Bobzien (2002) has challenged this interpretation. I cannot do justice to her extremely clear, acute, and scholarly paper. Suffice it to say that I think that she demonstrates that Williamson’s reasons for labelling Chrysippus an epistemicist are inconclusive. She is surely correct to point out that commitment to bivalence alone will not force Chrysippus into this position, since bivalence is a property of Stoic propositions, axi¯omata (indeed, axi¯omata are defined in part in terms of their satisfying bivalence: DL 7. 65; Math. 8. 74; Cic. Fat. 38; Acad. II 95), and the Stoics deny that all declarative sentences express propositions, hence it is open to them to deny that the crucial ‘indeterminate’ premisses in a sorites express propositions at all (this is thus a version of the conservative response—but a fairly radical one). She also demonstrates that ‘precipice’ before which Chrysippus (if indeed it is he: see Bobzien 2002: 225 n. 29) advises the sorites-responder to pull up his horses (Acad. II 94; cf. Sext. Pyr. 2. 253) is not that between truth and falsity, but rather that between sense and absurdity (2002: 225–6). She is also correct in noting that, in the case of the Liar, Chrysippus explicitly allows that bivalence fails for the problematic sentences (Bobzien 2002: 221–2; cf. Barnes 1997; Mignucci 1999). None the less, I am less convinced that Chrysippus would have treated the relevant sentences of the sorites as being on a par with Liar sentences; after all, the latter are problematic in virtue of their form, but there is no formal difference between the unproblematic ‘if two are few, so are three’ and the (supposedly) problematic ‘if fifty are few, so are fifty-one’. Moreover, there is no independent evidence that the Stoics would have regarded these sentences as ‘indefinite’ in the appropriate sense or senses. So I believe that the possibility that Chrysippus adopted an epistemicist stance remains open, and I am inclined provisionally to suppose that he did; I call this the ‘orthodox’ account (it is also adopted, without much discussion, by Myles (Burnyeat 1982: 335)). ⁴⁶ ‘Such ignorance … is just what independently justified epistemic principles would lead one to expect’ (1997: 215).

Self-Refutation and the Sorites

363

Against this, it may be retorted that the unpalatable consequence is simply too unpalatable to swallow: moreover, it simply misses the point about the nature of vagueness. Vagueness is a real semantic phenomenon, not simply a consequence of our poor faculties. As Schiffer has put it (2003: 183), what on earth could be the semantic facts of the matter that would determine that ‘[the word ‘‘rich’’] expresses the property of having at least $495,946.47 and not the property of having, say, at least $495,946.48?’⁴⁷ Furthermore, if the Williamson-type solution is correct, why isn’t it obviously so? If one (or more) of the complex premisses of a sorites is false, why are we disposed to think it (or them) true?⁴⁸ Properly dealing with the paradox involves accounting for its fallacious allure.⁴⁹ Sorensen (2001: 12) argues that ‘linguistic competence commits us to each premise of a sorites argument (while also committing us to the negation of their conjunction)’. This does not, however, demand a revision of logic as such: it merely means that we are psychologically compelled to hold at least one false belief, if we confront the premisses of a sorites argument as a whole. Bivalence still holds, as does transitivity.⁵⁰ The fault is psychological. But linguistic competence in the use of vague predicates demands that, for any suitably constructed⁵¹ soritical argument, a competent speaker will assent to each of the complex propositions when considered in isolation, although when led through them one by one will, at some point and for reasons of brute psychology, simply stop assessing them. This is designed to answer the so-called forced-march sorites introduced by Horgan (1994), which is nothing more than a modernized version of the sort of aggressive questioning that Chrysippus imagined himself responding to, and which Carneades reinvented with a vengeance (Cic. Acad . II 92–4; cf. Sext. Math. 7. 416, Pyr. 2. 253). A subject is presented with a suitably soritical sequence, and asked at every stage of the sequence whether they accept the conditional premiss. Language mastery suggests that the subject should accept each premiss, but not of course the conclusion which follows from their conjunction. Conversely, any subject deciding to reject any particular complex premiss stands accused of a lack of understanding of the semantics of the vague predicate itself.

⁴⁷ The same point is made with equal vigour in the case of the ancient argument by Barnes (1982: 56): ‘Chrysippus’ project is surely doomed to failure for very many soritical predicates. Can we really imagine that one grain, unknown to us, miraculously turns a mere collection into a genuine heap? or that one heart-beat, yet to be determined by science, makes an adult from a child? Such suppositions are patently ridiculous: there are no undiscovered facts of the sort which Chrysippus’ solution would postulate.’ ⁴⁸ This is what Delia Graff (2000: 50) has labelled the ‘Psychological Question’. ⁴⁹ See Mills 2004: 1 ff. ⁵⁰ The proposal thus differs from that canvassed very generally by Barnes (1982: 63–4 and n. 88), where we are to assess the generalized existential (∃x)(Fx) for a certain domain as being true, even though no individual member of the domain is (determinately) F . ⁵¹ i.e. one in which the steps really are of a suitably indiscriminable nature.

364

R. J. Hankinson

IV This brief excursus through some modern responses to the sorites serves to highlight both how much they owe to the ancient puzzle, but also how some distinctive contemporary concerns (crucially that of linguistic competence) are not seriously adumbrated in the ancient texts.⁵² The argument with which we began was designed to show not that the Empiricist had no understanding of the semantics of ‘experience’, but rather that there couldn’t be any such thing as experience at all. The Empiricist’s response, by pointing out that very many perfectly quotidian (and perhaps indispensable) predicates of ordinary language are subject to soritical objections as well, does indeed indicate the price that must be paid by anyone seriously wishing to root out vagueness from language; but even so, the emphasis is not on intuitions regarding what makes a speaker of a language competent in it. So far, we have been investigating the sorts of circumstance in which ancient philosophers of various colours deploy sorites, and what they (and some moderns) say about them. I want finally to follow up a different line of inquiry, the one to which the title of this article refers, and consider two questions: (1) did the ancients ever seek to develop a self-referential soritical argument, and if so, (2) what could such an argument (or arguments) show? First of all the evidence. Consider again the first epigraph to this article: ‘And it [sc. the sorites] also opposes him who speaks and him who argues with it’ (Med. Exp. ix, p. 98 W). Barnes (1982: 57 n. 60) remarks: ‘this is not the exciting claim that soritical arguments are self-refuting, but the ad hominem observation that even those Dogmatists who pretend to reject ‘‘experience’’ because of the sorites do in fact frequently rely upon ‘‘experience’’ ’. This assessment clearly applies to the second epigraphical passage (Med. Exp. xv, p. 113 W: the Dogmatists accept, reluctantly, that experience generates knowledge; moreover, they hold that all knowledge should be subject to the canons of logical assessment; yet they cannot apply those canons in this case), but matters are less clear in the case of the first: after all, it would be natural in that case to say that experience opposes the sorites, rather than that the sorites opposes the opponent of experience, and I am inclined to think that Galen at least hints at something more than simply that the Dogmatists are guilty of pragmatic inconsistency. But I do not rest very much on that. Much less equivocal, I think, is a section of the passage in which Cicero attacks (in his own, sceptical persona) the pretensions of logic to be able to make good the deficiencies of perceptual discrimination in Acad . II 91–8. Logic, it is said, ⁵² Sextus directs a sorites, or a series of them, against the grammarians (Math. 1. 68–71); but what is at issue here is not ordinary linguistic competence but rather the grammarians’ theoretical pretensions.

Self-Refutation and the Sorites

365

‘was discovered in order to be as it were the arbiter and judge of truth and falsehood’ (II 91); but Since you make so much of this art, do you not see that it is by nature such as to be totally against you, when at its first step it blithely imparts the elements of discourse, the resolution of ambiguities (ambiguorum intelligentiam) and the argument of logical consequence (concludendique rationem), but then, by a series of small additions (paucis additis), it comes to the sorites, surely a slippery and dangerous place (locus: = topos), which you recently said to be a vicious sort of argument (vitiosum interrogandi genus).⁵³ So: is this vice then our fault? The nature of things has given no knowledge of limits such that we are able to state in any case whatever where they are. And this is true not only in the case of the heap from which it derives its name, but in absolutely everything. If we are asked gradually (minutatim) to say with how much either added or subtracted (quanto aut addito aut dempto) something is rich or poor, famous or obscure, many or few, great or small, long or short, wide or narrow, we will not be able to answer. ‘But sorites are vicious’. Smash them then if you can, otherwise they will be a source of great trouble for you. (Acad . II 92–3)

The tone is jocular and playful (note the pointed ambiguity of ‘locus’: you arrive at a slippery place, but also at a slippery type of argument); but I suspect that it also hides a serious, or at least semi-serious, point. The phrase ‘paucis additis’ clearly refers to soritical moves, as the later ‘aut addito’ confirms (and see II 49, quoted in n. 53). The claim that is made is that we can find ourselves led by a series of imperceptible stages from something that seems logically impeccable to the sorites, which is somewhere we (as good believers in classical logic at least) will not want to be, since as a form of argument it is ‘vitiosum’, which must at the very least mean that it has the capacity for producing unhappy results (see n. 53). What is it that is logically impeccable? Cicero offers a conjunction of three conjuncts: ‘elements of discourse’, ‘resolution of ambiguities’, and what I translate, perhaps tendentiously, as ‘the argument from logical consequence’. Together, they may be supposed to characterize the fundamentals of logic, broadly construed in the ancient manner to include rhetoric and grammar, in which case my translation may reasonably seem over-restrictive, and some version like that of Rackham in the Loeb edition (‘the theory of the syllogism’) may seem more appropriate. But, or so I claim, the phrase is deliberately ambiguous (a fact which is of course doubly pointed coming as it does immediately after the claim that logic is a tool for resolving ambiguity). Cicero’s Latin is intended to suggest something like the Greek sunaktikos logos (‘ratio’ being his standard Latinization of logos), ⁵³ Here Cicero clearly refers back to Lucullus’ remark at II 49 (indeed the language of the latter passage recalls that of the former at many points): ‘Et primum quidem hoc reprehendendum quod captiosissimo genere interogationis utuntur, quod genus minime in philosophia probari solet, cum aliquid minutatim et gradatim additur aut demitur. Soritas hoc vocant, quia acervum efficient uno addito grano. Vitiosum sane et captiosum genus!’ For the Stoic characterization of the sorites as vitiosus, cf. Fin. IV xviii 50; the term presumably renders the semi-technical mochth¯eros (or the slightly less common equivalent phaulos), used in Stoic contexts (e.g. Pyr. 2. 150, 153, 154, 156) to indicate formal deficiency.

366

R. J. Hankinson

which could refer in general to theory of consequence (as it is doubtless partly, when considered with the other two conjuncts describing the ambit of logic, intended to do); but equally it might refer to a particular logos or argument form, one that is paradigmatically sunaktikos, or concludent: and the obvious candidate for that is modus ponens.⁵⁴ So the suggestion then is that we can move, insensibly by the addition of small steps, from modus ponens to a sorites. It is not hard, informally at least, to see how. Modus ponens is a valid conditional argument with one conditional premiss (and of course one categorical premiss); if we then add a second conditional (of the appropriate form of course) to the argument, then the new argument should still be valid (of course it will have a new conclusion—or a new categorical premiss—but that does not matter). So if (i) ‘P; P → Q  Q’ is valid, then so too should be (ii) ‘P; P → Q, Q → R  R’. And so on. Such a sequence seems clearly (in some sense) soritical. Can it be represented more clearly in the canonical form? I think it can, although not without a certain clumsiness. The base categorical premiss will be something along the lines of: [B] (1) A one-conditional conditional argument, given that certain material and formal conditions C are satisfied, is sound.⁵⁵ But then (2) If a one-conditional conditional argument, given that certain material and formal conditions C are satisfied, is sound, then a two-conditional conditional argument, given that certain material and formal conditions C are satisfied, is sound; and so on, yielding (10,000) If a 9,999-conditional conditional argument, given that certain material and formal conditions C are satisfied, is sound, then a 10,000conditional conditional argument, given that certain material and formal conditions C are satisfied, is sound. But (10,000) is of course the case of the canonical sorites—and far from being sound, it is, so it is said, vitiosum, and hence non-concludent, asunaktos.⁵⁶ Thus ⁵⁴ This is confirmed by the relevant passage of Pyr. (2. 145–51, esp. 149); see further Brunschwig 1980 and Barnes 1980; of course, nothing precludes the argument being formulated using the third indemonstrable, since exactly the same syntactic considerations will apply, and perhaps given earlier considerations it would be better if it were; but for ease of presentation I retain the conditional form. ⁵⁵ The ‘material conditions’ are of course the truth of both the conditional and the categorical premisses; the ‘formal conditions’ simply that the argument exhibits the form of modus ponens. The argument could have been represented in terms of mere validity, but for presentational reasons I prefer to exhibit it in terms of soundness. ⁵⁶ For this term, see Pyr. 2. 146, 151, 152, 155, etc.; for its sense, see Barnes 1980.

Self-Refutation and the Sorites

367

by a meta-sorites we have shown how a mere, harmless modus ponens can be transformed into a soritical monstrosity.⁵⁷ And that it can be so transformed is guaranteed by the Stoic logical reductive mechanisms of the themata, or more generally by way of the so-called dialectical theorem (Math. 8. 231), which states that if an argument contains premisses which potentially generate a conclusion, then we can use that conclusion along with the other premisses in order to arrive at the actual conclusion of the argument, and so transform complex arguments into a series of simple ones whose consequentiality is guaranteed by the indemonstrables.⁵⁸ Simple modus ponens leads, inexorably, to the sorites;⁵⁹ logic destroys itself. And, as I suggested at the outset, the pairing of the sorites with the Liar in this anti-logical context is suggestive: if I am right, the meta-sorites, generated itself soritically on the basis of unimpeachable (to the Stoics) logical foundations, shows that something is rotten at the heart of their logical syntax, while insoluble⁶⁰ problems like the Liar undermine the fundamental semantic commitment to bivalence. Logic devours itself from within in terms of both grammar and sense.⁶¹ That, I claim, is what lurks behind Cicero’s light-hearted text. That is, I think it at least possible, and perhaps even likely, that some such logical gymnastics were engaged in in the ancient world, and that our texts preserve their shadow. But what would be the point of such an exercise? That partly depends upon who it is mounted by and against what opposition. As Cicero presents it, this sorites is clearly a sceptical⁶² manœuvre conducted by sceptics against the Dogmatists. The Dogmatists in question claim that the sorites is a vitiosum genus of argument, yet they are trapped by the very application of their own logic into transforming patterns that they claim to be evidently valid, by evidently valid means, into patterns which are evidently (to them) invalid (or vice versa). In this case, the issue concerns the formal syntax of the logical system; and the upshot is that such a logic is inconsistent. This result is similar, although deeper, than the one arrived at by Galen’s Empiricist. The latter simply points out that the sorites pattern, as deployed by the Dogmatist, fatally undermines all sorts of everyday predicates which the Dogmatist, willy-nilly, employs, a result which is enough, says the Empiricist, to discredit any deployment of the argument form. What the Dogmatist should do about that is the Dogmatist’s own business, Empiricists having no time for logic in the first place. In the case of Cicero’s argument, an ⁵⁷ For the claim that the sorites creates monsters see Cic. Nat. D. III 44. ⁵⁸ e.g., in argument pattern (ii) above it allows us to infer ‘Q’ from ‘P; P → Q’, and then to use that to detach the consequent of ‘Q → R’. For further analysis, see Bobzien 2003: 112–15. ⁵⁹ Or rather, any sorites can be analysed, by means of the standard meta-rules, into indemonstrables: this is not quite the same, but for our purposes the difference is not important. ⁶⁰ Inexplicabilis: II 95 is Cicero’s Latin for aporos (it became standard: see Sen. Ben. V 9): see DL 7. 44, 82, both of which refer to the sorites (among others) and cf. the book titles at 7. 175, 198. ⁶¹ For a subtle discussion of Cicero’s treatment of the Liar, see Barnes 1997. ⁶² ‘Sceptical’ in a loose sense; it thus contrasts with the argument reported in Med. Exp., which is a sceptical argument aimed by Dogmatists as at a certain sort of scepticism (medical Empiricism).

368

R. J. Hankinson

internal challenge is mounted to the coherence of logic itself, one that is at most only obliquely hinted at in Galen (see above, p. 364): ‘What if this very art [i.e. logic] destroys at the end the steps which came before, like Penelope unweaving her web?’ (Acad . II 95). That question unequivocally makes the charge of self-refutation; and while the particular image of Penelope is an unusual one, it clearly recalls the illustrations that the sceptics themselves use when embracing the self-refuting nature of their own procedures (sceptical arguments are like purgative drugs that expel themselves along with the rest of what they purge: Pyr. 1. 206, 2. 188; Math. 8. 480; scepticism is like a ladder to be kicked away after ascending: Math. 8. 481), and resisting the claim that the self-refutation (in the particular case of the ‘proof against proof ’) can be turned against themselves.⁶³

V All this may have seemed somewhat speculative; but I think there is enough in Cicero’s arch presentation, as well as in the wider general background, to justify such a reconstruction. Let me embark on even more speculation. In our meta-sorites [B], it is not the sorites that self-refutes, as in the case of certain other celebrated logoi, but logic itself; or perhaps, rather, modus ponens (and other indemonstrables) reduce themselves to absurdity. However, mindful of the fact (referred to in this passage: ‘aut addito aut dempto’ ; see also II 49, quoted in n. 53) that sorites can be generated ‘downwards’ as well as ‘upwards’ (the classical ‘downwards’ sorites being of course the Bald Man), we can perhaps construct a self-refutation of the sorites itself. Here the appropriate predicate is ‘is a sorites’, and we start with the paradigm case: [C] (1) A 10,000-conditional conditional argument, given that certain formal conditions C are satisfied,⁶⁴ is a sorites. But (2) If a 10,000-conditional conditional argument, given that conditions C are satisfied, is a sorites, then a 9,999-conditional conditional argument, given that conditions C are satisfied, is a sorites. And so on, eventually yielding the patently absurd conclusion that a oneconditional conditional argument is a sorites. There is no hint, so far as I can see, in any of the relevant texts that any such argument was ever mounted in antiquity. But equally there is no reason why it should not have been—and if ⁶³ On the ‘Proof against proof ’, see McPherran 1987. ⁶⁴ Note that in this case no material conditions need be satisfied.

Self-Refutation and the Sorites

369

it were, it would give substance to the claim of Galen’s Empiricist in the first epigraph. What, if anything, do arguments of this sort show? [C] might be mounted by a Dogmatist against sceptical deployments of sorites, in effect to claim that the notion of a sorites is itself hopelessly vague. But that would be of limited logical power: after all, the arguments themselves could still be constructed, and would still have their power to puzzle, even if we are perplexed as to how to label them. Secondly, it might be taken to show that there is no ‘magic’ number of premisses in a sorites—but that in itself is hardly a radical conclusion (after all, the number 10,000 is merely conventional). So I conclude that it is argument [B], which is not strictly self-refutational as such, which is of the greater intrinsic interest. How might a supporter of logic respond to it? As we have seen, there are a variety of possible strategies, but none of them is cost-free (if they were, the paradox would be ultimately uninteresting). If the orthodox account is right, Chrysippus accepted that soritical patterns of argument were valid, and held that in every case where the conclusion was false, there was also a false conditional premiss, albeit one indiscernible to the ordinary mortal eye. He is constrained to do so in part (albeit only in part) by his commitments to classical logic, as are the modern epistemicists discussed in section III above. If Bobzien is correct, on the other hand, he supposes that some of the conditional sentences fail to express propositions, and hence are not subject to truth evaluation. Both of these manœuvres are semantic in nature. Are there any syntactic alternatives? One could argue that there might be some important logical fact about highly extended arguments which renders their conclusions insecure, at least in certain circumstances and if certain further (material) conditions are met. We might essay an account of ‘true’⁶⁵ conditionals such that the satisfaction of their antecedents does not allow us to detach their consequents simpliciter, but only with a high degree of probability; and let us accept as ‘true’ conditionals in which the satisfaction of the antecedent probabilifies the holding of the consequent to some suitably high, if arbitrary, degree (say 0.9). If a conditional is ‘true’ on this account, and the antecedent holds, then we may detach the consequent. But, in the case of chains of conditionals, ordinary transitivity may sometimes fail to hold. Consider a chain of conditionals each of which is such that the antecedent probabilifies the consequent to degree 0.99. If we were to detach each consequent in sequence, and then use it as the antecedent of the subsequent conditional to detach the latter’s consequent, and so on, then the soritical argument will go through in the ordinary way, and the final consequent will be detached. But supposing we block that procedure, effectively by rejecting the ‘dialectical theorem’, and insist that in chains of conditionals of this sort, ⁶⁵ The point of the scare-quotes will become apparently shortly.

370

R. J. Hankinson

we obtain the truth-value of ‘collapsed’ conditionals⁶⁶ by multiplying their probability assignments. Thus, if the first probabilifies the second to degree 0.99, and the second the third to the same degree, the first probabilifies the third to degree 0.9801 (and so, on our proposed truth conditions, ‘if the first then the third’ still comes out true); but generalizing this procedure it is easy to see that after a certain number of steps (ten, as it turns out), the ‘collapsed’ conditional falls below the relevant threshold—and that conditional will be ‘false’. This allows us to say that none of the conditionals in a problematic sorites is itself false, but that many (very many) of the ‘collapsed’ conditionals, including most importantly the last one (‘if the first, then the ten thousandth’), are false. And so the inference does not go through. Of course, this is nothing more than a sketch, and a pretty sketchy one at that; and I doubt whether anything much like it was ever seriously considered by ancient logicians.⁶⁷ I am not even sure if such a suggestion can be consistently worked out in such a way that it will not yield contradiction. Certainly the abandonment of the ordinary transitivity of inference is not something to be undertaken lightly. Yet there is one piece of evidence that suggests that Chrysippus may have leaned (in some cases) in this direction. It derives from a hostile source, and thus needs careful handling, but the same is of course true for the vast bulk of what we know, or think we know, about Stoic logic. It comes from the beginning of Plutarch’s On Common Conceptions, and refers to Chrysippus. Having remarked ironically that the Stoics must believe that it was providence that sent him ‘since it needed him to turn everything in life topsy-turvy’, the speaker continues: He seems to me to devote the utmost conscientiousness and cleverness in overturning (anatrepein) and striking down (kataballein)⁶⁸ common experience (sun¯etheia), as some at any rate even of those who revere the man testify when they take issue with him about the Liar. For to deny that a conjunction formed of indefinite contradictories is unrestrictedly false, and to say that some arguments with true premisses and valid consequences (ag¯ogai hugieis) will none the less still have the contradictories of their consequents true: what conception (ennoia) of demonstration and what preconception (prol¯epsis) of confirmation does this not overturn? They say that the octopus gnaws off its own tentacles in winter; but as Chrysippus’s logic destroys and cuts off its vital parts, namely its first principles, what of the remainder of its conceptions will it leave unscathed? (Plut., Comm. not. 1059d–e) ⁶⁶ A ‘collapsed’ conditional is one which is obtained by applying the dialectical theorem, vel sim., to two or more sequential conditionals (thus ‘P → R’ is a collapse of ‘P → Q’ and ‘Q → R’). ⁶⁷ On related issues, see Barnes 1982: 64; at the minimum, the ancients had no conception of probability assignments of the sort required to articulate the suggestion. But this does not of course mean that they could not have come up with some less formal analogue to it. ⁶⁸ The language here is significant: both verbs have a long history in the context of refutation and self-refutation: anatrepein is used of refutation by Aristophanes in the Clouds (901); kataballein (originally a wrestling metaphor) was used by Democritus (68 B 125 DK); hoi kataballontes was the title of a work of Protagoras (80 B 1 DK).

Self-Refutation and the Sorites

371

The particular focus of the passage is the Liar; but it neither says nor implies that such considerations are restricted to that context; and the second of the two contested Chrysippean claims (‘some arguments with true premisses and valid consequences will none the less still have the contradictories of their consequents true’) seems at least as at home in the discussion of the sorites as it is of the Liar (although it can be effective there too: Acad . II 96). Each of the premisses, considered on its own,⁶⁹ does indeed appear to be true. The argument is evidently (by classical canons) valid. The conclusion is absurd. The claim is here explicitly made that Chrysippus’ logic self-refutes, and does so by destroying its own first principles; and if bivalence is a first principle of Stoic logic, so too surely is modus ponens —it is, after all, indemonstrable. Chrysippus may have countered this by at least considering restricting the applicability of the reduction rules (which in turn explains the sceptical retort uttered by Cicero, that he has no criterion for determining where they are applicable: Acad . II 95–8).

VI The development of the sorites against the concepts of classical logic, even if only deployed destructively, points the way (or at least a way) forward (if it is forward) towards a re-examination, and perhaps a radical overhaul, of the foundations of logic itself. I myself think (for reasons hinted at above and in the notes) that that is an unnecessarily high price to pay. Whether or not it is may in the end turn out to rest on empirical as much as on conceptual analysis.⁷⁰ But whatever the truth of the matter, the sorites still retains its powerful capacity to puzzle and disconcert; and as perhaps some of the ancients saw, its discombobulating power ranges beyond the semantics of vague predicates and into the heart of logic itself. REFERENCES Algra, K., Barnes, J., Mansfeld, J., and Schofield, M. (1999) (eds.), The Cambridge History of Hellenistic Philosophy. Cambridge. Annas, J. (1990), ‘Stoic Epistemology’, in Everson (1990), 184–203. Barnes, J. (1980), ‘Proof Destroyed’, in Barnes et al. (1980), 161–81. (1982), ‘Medicine, Experience and Logic’, in Barnes et al. (1982), 24–68. ⁶⁹ This way of putting it coheres with my sketched ‘solution’; it also recalls a rather different modern response to the sorites, that of Delia Graff: whenever we inspect any particular conditional in a soritical sequence, we will be constrained to accept it, and for Sextan reasons (cf. Math. 7. 419: §II): ‘if two things are saliently similar … then it cannot be that one is in the extension of a vague predicate … while the other is not’ (Graff 2000: 57). It is also a feature of her view that vague expressions are subject to different criteria of proper use on different occasions, which has some affinities with the Empiricist position; but I cannot pursue these issues any further on this occasion. ⁷⁰ See Barnes 1982: 64–5 and, in a modern context, Dummett 1976.

372

R. J. Hankinson

Barnes, J. (1997), ‘Logic in Academica I and the Lucullus’, in Inwood and Mansfeld (1997), 140–60. Schofield, M., and Sorabji, R. R. K. (1979) (eds.), Articles on Aristotle, iii: Metaphysics. London. Burnyeat, M. F., and Schofield, M. (1980) (eds.), Doubt and Dogmatism. Oxford. Brunschwig, J., Burnyeat, M. F., and Schofield, M. (1982) (eds.), Science and Speculation: Studies in Hellenistic Theory and Practice. Cambridge. Beall, J. C. (2003) (ed.), Liars and Heaps: New Essays on the Sorites Paradox. Oxford. Bobzien, S. (1999), ‘Logic: The Stoics’, in Algra et al. (1999), 92–157. (2002), ‘Chrysippus and the Epistemic Theory of Vagueness’, Proceedings of the Aristotelian Society, 102: 217–38. (2003), ‘Logic’, in Inwood (2003), 85–123. Brennan, T. (2003), ‘Stoic Moral Psychology’, in Inwood (2003), 257–94. Brunschwig, J. (1980), ‘Proof Defined’, in Barnes et al. (1980), 125–60. Burnyeat, M. F. (1976a), Protagoras and Self-Refutation in Later Greek Philosophy’, Philosophical Review, 85: 44–69. (1976b), ‘Protagoras and Self-Refutation in Plato’s Theaetetus’, Philosophical Review, 85: 172–95; repr. in Everson (1990), 39–59. (1982), ‘Gods and Heaps’, in Schofield and Nussbaum (1982), 315–38. (1983) (ed.), The Skeptical Tradition. Berkeley. Dummett, M. (1975), ‘Wang’s Paradox’, Synthèse, 30: 301–24; repr. in Dummett (1978) 248–68, and Keefe and Smith (1997), 99–118. (1976), ‘Is Logic Empirical?’, in Lewis (1976), 45–68; repr. in Dummett (1978), 269–89. (1978), Truth and Other Enigmas. London. Evans, G., and MacDowell, J. (1976) (eds.), Truth and Meaning. Oxford. Everson, S. (1990) (ed.), Epistemology, Companions to Ancient Thought, i. Cambridge. Frede, M. (1974), Die stoische Logik. Göttingen. (1983), ‘Stoics and Sceptics on Clear and Distinct Impressions’, in Burnyeat (1983), 65–93. (1985), Galen: Three Treatises on the Nature of Science. Indianapolis. Giannantoni, G. (1977) (ed.), Scuole socratiche minori e filosofia ellenistica. Bologna. Graff, D. (2000), ‘Shifting Sands: An Interest-Relative Theory of Vagueness’, Philosophical Topics, 28: 45–81. Hankinson, R. J. (1997), ‘Natural Criteria and the Transparency of Judgement: Philo, Antiochus and Galen on Epistemological Justification’, in Inwood and Mansfeld (1997), 161–216. (2003), ‘Epistemology’, in Inwood (2003), 59–84. Heck, R. (2003), ‘Semantic Accounts of Vagueness’, in Beall (2003), 106–27. Horgan, T. (1994), ‘Robust Vagueness and the Forced-March Sorites’, Philosophical Perspectives, 8: 159–88. Ierodiakonou, K. (1999) (ed.), Topics in Stoic Philosophy. Oxford. Inwood, B. (2003) (ed.), The Cambridge Companion to the Stoics. Cambridge. and Mansfeld, J. (1997) (eds.), Assent and Argument: Studies in Cicero’s Academic Books. Leiden. Keefe, R., and Smith, P. (1997) (eds.), Vagueness: A Reader. Cambridge, Mass.

Self-Refutation and the Sorites

373

Kühn, C. G. (1821–33), Galeni Opera Omnia, 20 vols. in 22. Leipzig. Lewis, H. D. (1976), Contemporary British Philosophy, 4th series (London). Long, A. A. (1996), Stoic Studies. Berkeley. and Sedley, D. N. (1987), The Hellenistic Philosophers, 2 vols. Cambridge. McPherran, M. L. (1987), ‘Skeptical Homeopathy and Self-Refutation’, Phronesis, 32: 290–328. Mignucci, M. (1999), ‘The Liar Paradox and the Stoics’, in Ierodiakonou (1999), 54–70. Mills, E. (2004), ‘Competence and Contradictory Commitments’, Philosophical Books, 45: 1–11. Rawson, E. (1982), ‘The Life and Death of Asclepiades of Bithynia’, Classical Quarterly, 32: 358–70. Rist, J. M. (1978) (ed.), The Stoics. Berkeley. Schiffer, S. (2003), The Things We Mean. Oxford. Schofield, M., and Nussbaum, M. C. (1982) (eds.), Language and Logos. Cambridge. Sedley, D. N. (1977), ‘Diodorus Cronus and Hellenistic Philosophy’, Proceedings of the Cambridge Philological Society, 23: 74–120. (1982), ‘On Signs’, in Barnes et al. (1982), 239–72. Sorabji, R. K. K. (1979), ‘Aristotle on the Instant of Change’, in Barnes et al. (1979), 159–77. Sorensen, T. (2001), Vagueness and Contradiction. Oxford. Stopper, M. R. (1983), ‘Schizzi Pirroniani’, Phronesis, 28: 265–97. Vallance, J. (1990), The Lost Theory of Asclepiades of Bithynia. Oxford. Walzer, R. (1944), Galen on Medical Experience. Oxford. Williamson, T. (1992), ‘Vagueness and Ignorance’, Proceedings of the Aristotelian Society, suppl. vol. 65: 145–62; repr. in Keefe and Smith (1997), 265–80. (1994), Vagueness. London. Wright, C. (1975), ‘On the Coherence of Vague Predicates’, Synthèse, 30: 325–65. (1976), ‘Language Mastery and the Sorites Paradox’, in Evans and McDowell (1976), 223–47; repr. in Keefe and Smith (1977), 151–73.

19 Ideas Leap Barriers: The Value of Historical Studies to Philosophy Richard Sorabji

One of the greatest pleasures of my philosophical career resulted from Myles Burnyeat’s invitation to me to give lectures together to classes of 100 students in London University in the years 1970 up to 1976. We were both present in the room together, one lecturing and one offering objections from the front row, and we had a discussion hour after the lecture. Myles—I find it hard to call him Burnyeat—was always electric both inside and outside the classroom. I shall here discuss some of our intersecting views about the history of philosophy.

QU E S T I O N S Do ideas leap across historical and cultural barriers, or are they prisoners of the milieu in which they were thought up? Are any past ideas so obsolete that they could not be revived, or even taken seriously? Are any present ideas so entrenched that they could not be given up without some vast disaster? Do ideas belong to the context of a certain time, so that they could not occur before or after a certain date? Such views have been taken by some of the very best philosophers of our time. Yet, I should like to present a view of ideas as able to leap much more freely than this, and this is also relevant to the value to practitioners of philosophy of studying its history. Whichever view we take, it may be very important to be aware of the history of our ideas. But on one view, studying that history may show us why we are trapped, why we have to think the way we do. This is sobering, and may raise questions about the objectivity of our ideas. On another view, studying the history of ideas is liberating. The history of ideas may be a reservoir from which we can replenish our own ways of thinking. Philosophy depends on a philosophical imagination. The last fifty years have seen many good ideas, but the last 2,500 years of Western philosophy is a far greater resource, especially if it is taken to include Islamic and Jewish philosophy, even before we consider

Ideas Leap Barriers

375

philosophies from further east. On the other hand, the history of our own ideas may be salutary in a different way, that it leads us to discard some of them as ill-founded.

A N T I C I PAT I O N A N D R EV I VA L Let me start with an example of an idea which did not have to wait until modern times for its birth, because it was anticipated 1,400 years earlier. I am thinking of Bishop Berkeley’s theory known as idealism. Berkeley in the eighteenth century solved a problem about knowledge made acute by Descartes: if we perceive only the ideas in our minds, how do we ever get to know about the tables and chairs that give us those ideas? Berkeley’s reply was that tables and chairs are just bundles of ideas existing in minds, always in the mind of God and intermittently in our minds, so we know them directly. It has been said that this view could not have occurred before Descartes, and I shall return to this. But the idea that tables and chairs are just bundles of the ideas existing in God’s mind is fully formulated in the fourth century  by Saint Gregory of Nyssa, only in relation to an entirely different problem posed by the pagan philosopher Porphyry: since cause must supposedly be like effect, how can an immaterial God create a material world? Gregory’s answer: the world is not material in quite the way you think. Material objects are just bundles of God’s ideas. Even the idea of bundles is probably taken from Porphyry, although he had been speaking of individuals as bundles of qualities and quantities, not of ideas. Here the same theory is devised in relation to an entirely different problem, although the relation between Gregory and Berkeley is closer, because Berkeley claims it as an extra merit of the theory that it also solves the problem of how an immaterial God can create a material world. Did Berkeley know Gregory’s theory, or reinvent it? I don’t know. But what I can say is that a theory does not necessarily change out of recognition when it is used for a completely different purpose, although it may sometimes. In the case of Gregory’s theory and Berkeley’s, they correspond even in many fine details. This possibility of detachment from the original context makes it easier for ideas to recur, and I shall come back to it. If Gregory’s theory is an anticipation, Berkeley’s is a revival. If one speaks of anticipation, this may be merely because Berkeley’s version is currently better known. Of course there are differences too between Gregory and Berkeley, and I have elsewhere mentioned three.¹ For Gregory the creation takes the form of God bundling the ideas together, for Berkeley the form of his making them perceptible ¹ Sorabji 1983: ch. 18 on Gregory is developed further in Sorabji 1988: ch. 4.

376

Richard Sorabji

to others. Again, Gregory understands the formless void in the Genesis account of creation as a power for accommodating qualities, although it is not, like the matter which Berkeley attacks as unknowable, a subject that possesses qualities. Thirdly, Gregory does not express Berkeley’s interest in the dependence of the bundles on their continuing to be in God’s mind. This third difference flows from another, that since Gregory’s interest is not in scepticism about current perceptions, but only in the original creation, he does not, like Berkeley, have to consider the status of the bundles after the Creation.

A M O R E R E G U L AT I V E V I EW O F H I S TO RY: ‘ E A R L I E S T D AT E ’ F O R A N I D E A Four of the best philosophers of the present time have expressed a more regulative view of history than mine, and one of them, of course, is Myles Burnyeat, from whom I have taken the example of idealism.² He pays the closest attention to history, when he argues, convincingly, that the ancients, even the ancient sceptics, did not raise the doubt, ‘Is there a world external to the mind?’ It was only when Descartes raised this doubt that Berkeley had any reason to postulate that the world of bodies was not external after all, but only a set of bundles of ideas in the mind of God, and fleetingly of ourselves. Although Augustine, before Descartes, used the Cogito argument to answer sceptical doubt, the doubt he was answering was not this radical doubt; so he did not depend so exclusively on the Cogito as an answer. Berkeley was wrong to think that his idealism was anticipated in any form by the ancients, except possibly the Neoplatonists, who, as Burnyeat recognizes, do come much closer to having some form of idealism. On idealism Burnyeat says that it was not reached before Berkeley, and that Berkeley’s version, according to which for perceptibles to be is to be perceived could not be reached until Descartes: Idealism, whether we mean by that Berkeley’s own doctrine that esse est percipi, or a more vaguely conceived thesis to the effect that everything is in some substantial sense mental or spiritual, is one of the very few major philosophical positions which did not receive its first formulation in antiquity.

Descartes had reached a decisive shift of perspective without which no one, not even Berkeley, could have entertained the thought that esse est percipi. But I believe it would be hard not to call Gregory’s view a case of idealism in the wider sense allowed, and one reason why this anticipation could occur is that an idea thus widely conceived can arise in philosophy for more than one reason, and can be detached from any one particular reason. Ideas can be reused for different purposes. Certainly, Gregory’s immediate problem did not require him to offer ² Burnyeat 1982. The quotations below are from pp. 3 and 40.

Ideas Leap Barriers

377

Berkeley’s particular emphasis on the mind-dependence of ideas, but Burnyeat’s idea of that having been impossible is stronger. Gregory had taken the crucial step. He spoke of the ideas in God as logoi, borrowing the word from Platonists who had in turn exploited the Stoic idea of ‘seminal’ logoi, and who viewed the logoi in God’s mind as creative causes of physical objects. Gregory’s new step is to see the logoi not as causes merely, but as constitutive of the physical objects. Burnyeat’s view that an idea could not be reached before a certain date can be parallelled in other philosophers, and there is also the view that certain ideas cannot be taken seriously after a certain date, and again the view that other ideas cannot be surrendered over a certain period at least until some vast disaster upsets them. Let me illustrate these views with another quotation from Burnyeat, and one each from Derek Parfit and Charles Taylor and two from Bernard Williams. Derek Parfit, like Burnyeat, has set an earliest date, but for something different: progress in ethics. He has suggested that ethics cannot progress until atheism has taken over, a process which began in the 1960s. Belief in God, or in many gods, prevented the free development of moral reasoning. Disbelief in God, openly admitted by a majority, is a recent event, not yet completed. Because this event is so recent, … it is not irrational to have high hopes.

Parfit’s book of 1984, Reasons and Persons,³ is one of the best philosophical works of our generation. He has, roughly speaking, sought to replace the idea of a continuing subject which has consciousness with the idea of an embodied stream of consciousness, and he claims that a rather different sort of ethics is required when there is no continuing subject of consciousness to take into consideration. But in a later paper of 1999, he raises a more radical question, whether certain ethical questions can arise at all. ‘Can one feel sorry for a series of experiences?’ he asks. He postpones addressing the question. But although I am confident that he will have a very interesting answer, I have nonetheless expressed my own perplexity as to how more than a proportion of our ethical ideas could be applied to an embodied stream of consciousness.⁴ As regards theism, I doubt whether it always makes a difference in philosophy, even ethical philosophy, whether an able philosopher is theist or atheist. Of course it sometimes does, but theism takes such diverse forms in, say Aristotle, the Stoics, and the Epicureans, none of whom believed in human immortality in any ordinary sense, and even the most insistently religious of these, the Stoics, nonetheless think it possible to expound their ethics up to a point without necessarily bringing in their theology.⁵ If we turn from ethics to Parfit’s very interesting arguments about our being embodied streams of consciousness rather than possessors of it, at least these arguments have their parallels in ³ Parfit 1984, quotations from pp. 453–4. ⁴ Sorabji 2006b: ch. 15. ⁵ Sext. Math. 7. 22–3; concerning Plutarch’s charge of Stoic inconsistency about this, Comm. not. 1035A–D, and the conflicting evidence in Diog. Laert. Lives 7. 40, see Brunschwig 1991.

378

Richard Sorabji

highly religious debates about the Christian hope of resurrection. Parfit imagines two scenarios in which the question might be raised whether the same person still exists. The first is teletransportation, as envisaged in the television series Star Trek. Suppose your body is destroyed, but all the information required to reconstitute it out of new matter is beamed electronically to a distant spot. Would the reconstituted person be you? The second is brain transplant. Suppose your body is ageing, but at least half your brain, carrying all your mental characteristics, is transplanted into a fitter body. Would the transplant be you? And would it still be you, Parfit asked, following Bernard Williams, if a second successful transplant was done with the other half of your brain?⁶ Although Parfit’s two scenarios are drawn from science fiction, yet they match the scenarios drawn from a religious context in the third century . Would the resurrection after death in which Christians believe be achieved by God creating an entirely new body for us, with photographic likeness, as suggested by the brilliant Christian thinker Origen? Or would such teletransportation by God be destruction? And were the majority of Christians right that some of your present body, like Parfit’s half-brain, would have to be transplanted by God into the resurrection body? I believe that antiquity even anticipated, in the different context of the Stoics on differentiating people, the question about whether a person’s survival could depend simply on the survival of another person sufficiently like the first. That, at any rate, is how I interpret what I might call the Shrinking Argument, which was posed in the third century  by the Stoic Chrysippus in reply to the Growing Argument. If the Growing Argument had been right that growing gives us a new person, then surgically induced shrinking should bring us back the original person. But it would not: the original person would be denied this resuscitation, though merely by the survival of the one who had shrunk.⁷ At least in metaphysics, whatever may be thought about ethics, Stoic theism did not prevent them raising the same issue.

‘ L AT E S T D AT E ’ B E F O R E B E C O M I N G O B S O L E T E A N D S T R A N G E , I R R E P L AC E A B I L I T Y O F W H AT S U PE R S E D E S In his important work of 1989, Sources of the Self, Charles Taylor has set a latest date for certain old ideas and regarded certain more recent ideas as, for the time being, irreplaceable:⁸ ⁶ Parfit 1984; Williams 1973: 20, discussed by Sorabji 2004: ch. 6 and 2006b: chs. 3–4. ⁷ This is how I interpret the report in the Jewish philosopher Philo of Alexandria, On the Eternity of the World 48, in Sorabji 2004: ch. 6 (h) and 2006b: ch. 4. ⁸ Taylor 1989: 393–6.

Ideas Leap Barriers

379

These two big and many-sided cultural transformations, the Enlightenment and Romanticism, with its accompanying expressive conception of man, have made us what we are. … We sense ourselves to have already departed from the orbit of Deism, Lockean or Hutchesonian, let alone such seventeenth century notions as the divine right of kings. … These earlier views have become strange to us; it is hard to recapture in imagination what they could ever have had going for them. Some watershed has been passed. … Certain moral ideas emerge from this crucial period which still form the horizon of our moral outlook. One thing the Enlightenment has bequeathed to us is a moral imperative to reduce suffering. … Another major idea we have seen developing is that of the free self-determining subject. … This, together with the ideal of universal benevolence, has generated another deeply entrenched moral imperative, to universal justice, which has found expression in our century in the various universal declarations of rights. … And of course these ideas of freedom and dignity, in association with the promotion of ordinary life, have steadily eroded hierarchy and promoted equality. … Most notably, it has helped to bring about the steady rise in democracy as a legitimate form of political rule, to the point where it has become in the late twentieth century the inescapable source of legitimacy. … However unsuccesful mankind has been in attaining ‘the blessings and security of self-government’, no other aspiration ultimately incompatible with this is now available.

Taylor’s views have justly been influential. I once asked Myles Burnyeat, by way of a challenge, to name a philosophical idea that was too obsolete to return, and his first suggestion (he had others) was Taylor’s, the divine right of kings. I wish I could be so sure, when I consider some current leaders in the West. Bernard Williams has strongly endorsed both of Taylor’s points in discussing the ancient Stoic view that philosophy can calm the emotions. He alleges both the strangeness of that older view and the almost deterministic barrier created by certain more recent movements of thought which prevent us, he thinks, from taking the older view seriously. He comes very close to Taylor in describing us as standing on the other side of Christianity and Romanticism:⁹ But can we really believe that philosophy, properly understood in terms of rigorous argument, could be so directly related to curing human misery, the kind of suffering that priests and doctors and—indeed—therapists address? How many people can really have believed it? … [The author under review] has left me, for one, feeling how strange it might be to see rigorous philosophy (Chrysippus’ logic, for instance) exclusively or mainly in this light, and also how great a distance separates these thinkers, inasmuch as they did believe this, from the modern world. … We are surely bound to find the Epicureans too rationalistic, the Sceptics too procedurally self-obsessed, the Stoics (at least in their Roman incarnation) too unyieldingly pompous for us to take entirely seriously, not just their therapies, but the idea of them as philosophical therapists. … Standing on the other side of so much history, above all of Christianity and Romanticism, we are bound to find these therapists very strange, in their aims, their tone and their methods. ⁹ Williams 1994b: 25–6.

380

Richard Sorabji

In a more recent article, he comes close to Taylor again.¹⁰ There are ethical views which for us are ‘simply there’, and have no alternative. Their replacement makes no ethical sense for us. Part of his argument is against Hegel’s belief that we can show that our ethical ideas are an improvement on what has gone before. His answer to Hegel is that we cannot justify these views at all. They were not arrived at by argument on an agreed subject matter; history made us and our ideas together, and we can imagine the views being changed only as a desolation. Moreover, there are some values which are a manifestly contingent cultural development, like authenticity. It comes close to sounding as if history has imprisoned us with our ideas: There are some such virtues, such as authenticity or integrity of a certain kind, which are as a whole a manifestly contingent cultural development: they would not have evolved at all if Western history had not taken a certain course. … This outlook is ours just because of the history that has made it ours; or, more precisely, has both made us and made the outlook as something that is ours. … We believe, for instance, that in some sense every citizen, indeed every human being—some people, more extravagantly, would say every sentient being—deserves equal consideration. … It can seem, at least in its most central and unspecific form, unhintergehbar: there is nothing more basic in terms of which to justify it. … For us, it is simply there. … What in this connexion seem to be simply there, to carry no alternative with them, are elements of our ethical and political outlook, and in those terms there are no alternatives for us. … [An alternative] makes no ethical sense for us, except as a sense of retrogression, or desolation, or loss.

Burnyeat has also ascribed to antiquity, in the person of Aristotle, a view so remote from ours that he says we can hardly imagine what it would be like to take it seriously. Aristotle’s conception of the material or physical side of the soul–body relation is one which … no modern philosopher, whatever his persuasions, could share. … Aristotle’s solution to the mind–body problem … is worked out in terms of, and cannot be understood apart from, various physical assumptions which we can no longer share; assumptions, indeed, of such a kind that we can scarcely even imagine what it would be like to take them seriously. Aristotle’s philosophy of mind is no longer credible because Aristotelian physics is no longer credible.¹¹

I shall not take up this last case, because I agree that the view ascribed to Aristotle is one we could hardly imagine taking seriously. Burnyeat and I have started a controversy in the publication cited as to whether Aristotle took this view or a very pedestrian one instead, and it would spoil the fun of a flourishing debate to wade in and take it up again here. But Bernard Williams’s view of the Stoics as emotional therapists I regard differently, because I think that the Stoic claims, so far from being unimaginably strange, are simply right. Their philosophy does ¹⁰ Williams 2000, quotations from pp. 488–94. ¹¹ Burnyeat 1992:16. Sorabji’s alternative view is at Sorabji 1992.

Ideas Leap Barriers

381

indeed help with the ups and downs of everyday life. It analyses the beliefs that all emotions have in common, distinguishes them from mere side-effects, and helps you to target mistaken beliefs, by looking at the situation differently. The Stoics invented cognitive therapy. I am not saying that this would deal with, say, bereavement, unless one took Stoic advice very much further in the way that Epictetus took it. But that involves a kind of detachment in family life, which I regard as the unacceptable face of Stoicism. Once again, I am detaching certain aspects of the Stoic view, but then the leading Stoic Chrysippus himself detached certain views, saying that he could calm the emotions even of those who rejected the Stoics’ most central ethical values.¹² What I am saying is that there is an acceptable part of Stoic philosophy which does indeed help with the ordinary ups and downs of daily life. But readers can easily make their own judgements, because Williams very obligingly agreed to publish his view that the Stoic claims cannot be taken seriously alongside my explanation of why I think them simply right.¹³ As regards our shared ethical outlook, I am not sure how uniformly it is shared, nor how incomparable it is with ancient values. Comparison might be profitable between our idea of authenticity and the late Stoic idea of being true to yourself, including to your very individual persona.¹⁴ The idea of universal justice is certainly Stoic, though I myself have urged that it is not connected with any idea of human rights.¹⁵ As regards uniformity, it is hard to find uniformity of view in any population as argumentative as the Greek philosophers. But even amongst ourselves, one can feel almost nostalgic, reading about agreement on the lofty principles enumerated above. It is sobering how quickly a change of government, or the need to apply the principles to inconvenient cases, can lead to principles that were apparently agreed being abandoned anyhow.

D E TAC H A B I L I T Y I do in fact agree that some ideas would be very hard for us to entertain again at least at present. But an important reason why ideas nonetheless can recur more freely than one might expect is the possibility of detaching them from their original context. Detachment may sometimes leave one with too little of the original idea, but sometimes it leaves a lot intact. I have mentioned detaching the concept of idealism from a sceptical context, of detaching the Stoic therapy of emotions from the Stoic condemnation of all emotions, and of detaching the aim of universal justice from the idea of human rights. I will give one more example. In the sixteenth century, the Spanish philosopher Vitoria urged that the ¹² Chrysippus ap. Origen Against Celsus 1. 64; 8. 51. ¹³ Williams 1997, in reply to Sorabji 1997b. ¹⁴ Sorabji 2006b: ch. 8. ¹⁵ Sorabji 1993: ch. 11, ‘Did the Greeks have the Idea of Human or Animal Rights?’

382

Richard Sorabji

Conquistadors could not justly invade the American Indians in order to rescue them from the cruel practice of human sacrifice, because one had to consider how many of them one would oneself kill in the process. His idea of justice in war was embedded in ideas of natural law as connected with divine law, ideas which many could not now share. But his point has a simple appeal even detached from that embedding, and so it has proved in recent times. Ideas about the conditions for just war have appealed to many people in the United Kingdom, even where their original grounding would—I agree—have made little sense.¹⁶

VA LU E O F S T U DY I N G T H E H I S TO RY O F PH I LO S O PH Y I have emphasized the reusability of ideas, because it supplies one reason, only one, why studying the history of philosophy is important for philosophers. On this I believe that Burnyeat and I would agree, and indeed I shall express indebtedness to him at various points. But not everyone, I think, would agree that the history of philosophy is a source of useful ideas. Those who do agree could each supply their own examples, and Burnyeat’s choice of examples would, I know, make very good reading. But some examples I think are needed, since the point will not be recognized by everyone. In the last century, drawing on the Middle Ages, Arthur Prior brought back to our attention tense logic, and Alvin Plantinga kinds of necessity that had been overlooked in a period when the focus was almost exclusively on necessity based on the meaning of words. It should be acknowledged that Kripke equally reintroduced other forgotten kinds of necessity without drawing on historical studies, but even that illustrates the danger of forgetting knowledge that was once fully understood.¹⁷ I have already mentioned how useful the ideas of just war have recently proved to clear thinking, though devised originally by twelfth-century canon lawyers building on St Augustine from around 400 . If we take antiquity as the model, then I benefited from Aristotle’s insight that coincidences have no causal explanation, a subject which I was working out in lectures when Burnyeat and I were first lecturing together. In Aristotle’s example, there may have been a perfectly good explanation of the thirsty man going to the well at 2 p.m. and a different perfectly good explanation of the murderous thugs going to the well at 2 p.m. But it does not follow that there was any explanation of the fact, which proved to be fatal, that the two parties went to the well at the same time. So it is not true that whatever happens has a causal explanation. Since then, David Owens has used Aristotle’s point against the attempt to reduce explanation in economics to facts about the behaviour of millions of individual participants in an economy. Economic inflation can be explained, but the behaviour of millions of people in ¹⁶ Sorabji 2006a.

¹⁷ Prior 1967, 1968; Plantinga 1974; Kripke 1972.

Ideas Leap Barriers

383

such an economy will be reduced to a giant and inexplicable coincidence, unless it is brought together under the heading of inflationary behaviour. The notion of inflation cannot be reduced to the behaviour.¹⁸ As a matter of fact, I think that Aristotle provides another salutary warning against misguided reductions, when he warns that explanation by his so-called material cause, a thing’s ingredients, is not enough on its own. Here is just one more example of the value of recovering ancient viewpoints. I believe that ancient ethics, almost without exception, thought that in order to make right decisions, not only morally right ones, but practical decisions in general, one would typically need to consider one’s life as a whole. Modern ethics, whether post-Kantian, utilitarian, or intuitionist, has too often treated decisions as isolated from the individual’s life as a whole. This is something I speak about elsewhere.¹⁹ A second value of studying the history of philosophy is the opposite of revival. When we see the ancient origins of our entrenched ideas, this can lead us to question and even reject them. A case in point for me was the very entrenched idea that there is no harm in killing animals. We eat them every day. When I wrote a book on ancient ideas about animal minds, I was struck not by any arguments that we should not kill animals, but by the atrociously weak character of the arguments that it was perfectly alright to do so.²⁰ This is a reason for studying the history of philosophy that Williams stresses in an article of 1994. In the article of 2000, cited above, the emphasis was, as here, on strangeness: the past was strange, our present ethical concepts apparently familiar, but in need of being understood by seeing historically how they came to take the forms they did. In the 1994 article, the point of writing historically about philosophy lies principally in its making things strange. But in this article, the recommendation is to extend the sense of strangeness even to our present philosophical materials, in order to question them. This, I think, is the same as the second reason I have given for studying the history of philosophy, and Burnyeat rightly describes it as liberating us in a certain way from our present.²¹ But when Williams suggests that the principal philosophical point is making things strange, this does not seem to allow for my first reason, which is also liberating: namely, finding insights to revive from earlier thought:²² What was called in the original distinction ‘the history of philosophy’ is essential to any activity that is going to give a philosophical point to writing historically about philosophy. That point is going principally to be found in the possibility of the past philosophy’s being untimely, and helping to make strange what is familiar in our own assumptions. … What we must do is to use the philosophical materials that we now have to hand, together with ¹⁸ ¹⁹ ²¹ ²²

Owens 1992. Aristotle’s point is expounded in Sorabji 1980: ch. 1. Sorabji 2006b: ch. 9. ²⁰ Sorabji 1993. Burnyeat 2006, which he kindly showed me in proof after this chapter was written. Williams 1994: 26–7.

384

Richard Sorabji

historical understanding, in order to find in, or make from, the philosophy of the past a philosophical structure that will be strange enough to help us to question our present situation and the received picture of the tradition, including those materials themselves.

The desire to revive a forgotten idea is only one of the benefits to be gained from studying the history of philosophy. Revision is a second, but here is a third. Philosophers do not on the whole agree with each other. The test of a great philosopher is not whether you agree with him or her, but whether he or she makes you rethink. We are more likely to agree with very minor philosophers who confine themselves to a few cautious points. What the great philosophers do is to make us rethink our ideas by presenting a powerful alternative, whether or not we end up agreeing with it. Philosophers with the stature to produce this effect do not all come in one century, for example in the latest century, but are scattered through the ages. With those whom we find congenial, we can learn most of all by wrestling with, though not necessarily accepting, their ideas. I think this third reason was what Burnyeat was referring to, when he said to me a few years ago that some philosophers are good to think with. It may be asked about these three reasons for studying the history of philosophy (and the list is not exhaustive): why not just tell us which were the good ideas from the past, instead of us having to study them? There is justice in this retort as regards the first reason. It is in fact part of the point of including historians of philosophy in the life of a philosophy department that useful ideas from the past can be learnt at second hand. But the retort does not address the second and third reasons. For you need to understand for yourself the history of the presupposition that there is nothing wrong with killing animals if you are to assess whether it is justified, and possibly revise your beliefs. The retort also does not address the third reason. You cannot think with a great philosopher without wrestling with that philosopher yourself. Sometimes philosophy departments compromise by taking their curriculum back a little way to Descartes or to Thomas Aquinas. But I think that Burnyeat’s article on Descartes discussed above shows that you cannot understand Descartes without understanding the ancient Greek sceptics. As for Aquinas, one cannot understand him if one does not recognize that his Aristotle was not the original Aristotle, but an Aristotle Platonized by the intervening Neoplatonists to believe in a Creator God and an immortal human soul, in ways which accidentally made him more acceptable to Christianity.

D I S C O N T I N U I T I ES : ( A ) LO S S O F WO R K S I have so far been stressing the revival of ideas, but it is time to stress the discontinuities and barriers. Not all ideas will be revived. Some are lost forever, some for centuries. In the ‘dark ages’ of the Latin-speaking world only a very

Ideas Leap Barriers

385

small part of Greek philosophy was known. Less was known than might have been, because Boethius, whose project it had been to translate into Latin Plato and Aristotle and to paraphrase their Greek commentators, was executed for political reasons at around the age of 45, c. 524 , with only some of the logical works completed. It was not for another six centuries in the late twelfth to thirteenth centuries that the works of Aristotle and his Greek commentators began to become available in profusion in Latin translations from both sides, from the Arabic versions in Spain and from the Greek versions in Constantinople.²³ This translation movement transformed Latin medieval philosophy. It may give some comfort that philosophical ideas can be recovered after 600 years. But a warning has been given by Carlos Steel²⁴ about how precarious revival can be. Translation on its own provides no guarantee of revival, unless the climate is right. Why did the works of Plato have no comparable impact on the Latin-speaking West until the time of Ficino in the Renaissance? It was not that there were no Latin translations of Plato, because four works had been translated. But except for the Timaeus, his other three works were comparatively little read. D I S C O N T I N U I T I E S : ( B ) S E PA R AT E D C U LT U R E S I spoke earlier of the revival of ideas. But is a threat to revival suggested by the case of Indian philosophy? With Indian philosophy, the accounts given to me by scholars have sometimes led me to think, ‘Great minds think alike’, but have never led me, except in the case of Buddhism, to think, ‘The Indian and Western ideas are so close that one must have read the other’. Does this mean that ideas are never really close, unless one party has heard, at least indirectly, of the ideas of the other? Rather, I think that Indian philosophy constitutes a special case. Wilhelm Halbfass in his book, India and Europe (1988), has argued that Indian philosophy, unlike technology and other disciplines, was the preserve of Brahmin priests, who even today are brought up speaking ancient Sanskrit and eating separately from others. Unlike the Buddhists, who did proselytize, the Brahmin priests considered that the West was only good at boring things like conquering. But when it came to actual thinking, the West couldn’t do it. One should not study a foreign philosophy until one had competely mastered one’s own, for which one incarnation would scarcely be sufficient. This tradition is already exemplified by the case of the conquest of north-west India by Aristotle’s pupil Alexander the Great in the fourth century . The story may be apocryphal, but it testifies to ²³ Sorabji, 1990. ²⁴ Steel 1990. I thank Carlos Steel for discussing this subject with me at a symposium in September 2001 at New York University’s Center for Ancient Studies, where he outlined some of his own very valuable views about the history of philosophy.

386

Richard Sorabji

an attitude that was real enough. Alexander sent for the Indian philosophers and said, ‘Come and tell me about your philosophy.’ ‘No,’ they replied, ‘if you are interested, come to us.’ On Halbfass’s view, there is a special reason, if Indian philosophical ideas, though often analogous to Western ones, are seldom, if ever, quite identical. There was a peculiarly strong separation of the traditions. A contrast is provided by the case of the ancient Buddhists. There is a work set in the second century , though compiled a little later, The Questions of Milinda. In this work, a Buddhist monk is represented as teaching Menander, a Greek king of Bactria, successor of Alexander’s generals, the Buddhist doctrines about there being no continuous self and the ethical consequences. These include the consequence that, when so many momentary selves have died already, we should not fear the final death. I believe I have found this Buddhist idea reflected in Greek and Latin texts of the first century  by Seneca and Plutarch.²⁵ The idea was certainly known to Avicenna, and a version of it has been reinvented independently in modern times by Parfit. Some minimal contact of traditions seems to be enough. But nonetheless, the Graeco-Roman echoes are superficial. They were not harmonized with the rest of what Seneca and Plutarch said. The ancient controversy in India between Buddhist and Hindu philosophers on the self went deeper and produced more radical rethinking, I believe, than anything in Western controversy, and I am inclined to think, contrary to Parfit, that this is because the issue was so central to their religious beliefs.²⁶ The real contrast is between the cases of Indian and Islamic philosophy, the first separated from our tradition, the second an inseparable part of it. It is widely acknowledged that from the ninth century , Islamic philosophy was in close dialogue with earlier Greek philosophy. But in 1931, it was suggested by the German scholar Otto Pretzl that, before that, there was an indigenous Islamic philosophy too irrationalistic to be influenced by Greek philosophy. The star specimen of such work had been translated into English. But because of the ambiguities created by the medieval Arabic convention of not writing in vowels, the translator was not sure whether the discussion was about atoms or ants. In fact, however, the Arabic was answering sentence by sentence and line by line an earlier Greek discussion about the effects of discontinuity in space, time, or motion. Atoms would have to move in the cinematographic manner of objects on a movie screen, disappearing from one spot and reappearing a little further on, without ever having been in between.²⁷ So much for the idea of a totally independent phase of philosophy, although I should stress again that dependence in no way precludes the most extreme originality. From the beginning, Islamic philosophy built on Greek philosophy with the utmost originality and took the subject forward in its own way. Below I shall take ²⁵ Milinda’s Questions, trans. Horner (1964); Sorabji 2000: ch. 16. ²⁶ For an initial attempt to consider the discussion at second hand, see Sorabji 2006: ch. 16. ²⁷ Sorabji 1983: ch. 25.

Ideas Leap Barriers

387

up the particular case of Avicenna, who crystallized the idea of intentional objects of awareness, which Brentano was to pick up from the later Latin scholastic tradition and make central to modern philosophy of mind. Avicenna did this against the background of Aristotle’s theory of what goes on in the sense-organs. He also, quite independently of Augustine, though possibly influenced by hints in Porphyry, anticipated aspects of Descartes’ treatment of the Cogito in his Flying Man argument for the independence of the self from the body.²⁸ One of the crucial premisses is explained in a recently translated letter by Avicenna to one of his students. One of a number of places where he produced the argument was in The Return [of the soul after death]. This work was translated into Latin by Alpago (died 1522) a century before Descartes, and was available in a Venice edition of 1546, although Descartes himself had Augustine to go on and did not need this further stimulus. D I S C O N T I N U I T I E S: ( C ) T R A N S M U TAT I O N S A third case of discontinuity is that of the transmutation, as opposed to the preservation, of ideas. I have been particularly interested in this. The Stoic devices for avoiding emotional agitation were transmuted into Christian devices for avoiding temptation, which in turn gave birth to the recognition of seven cardinal sins.²⁹ I have already mentioned the transmutation of Aristotle’s perceptual theory, which I have taken to be about the reception of colour patches in the eye jelly, into a theory of intentional objects. I have also mentioned the transmutation of Aristotle’s disagreements with Plato into a wholesale agreement which Christianized Aristotle almost as much as it Platonized him.³⁰ Funnily enough, the process of harmonizing disagreements among schools started off in someone like Antiochus in the first century  as the very opposite of finding things strange. It was a process of appropriating from other schools what had once been seen as alien. But it finished up eventually by transmuting Aristotle’s ideas. All these transmutations had good effects of their own. But does not this discontinuity, like the others, imply a barrier? The original ideas are not being passed on. In many cases, they are not being understood correctly, unless the adaptation is made with conscious knowledge that it is transmuting the original, or, as in the example of Christian temptation, it makes no claim to be passing on the original. May not we, like some of the transmuters, be unable to understand philosophical periods to which we are outsiders? I would like to suggest that, difficult as it is, the outsider may even have some advantages, alongside his or her disadvantages. It is hard to describe what is exactly contemporary with oneself, except by way of contrast. One can describe it by contrasting it with what went before, as Aristotle described his ²⁸ Sorabji 2006b: ch. 12.

²⁹ Sorabji 2000.

³⁰ Sorabji 1990.

388

Richard Sorabji

own philosophy. Or one can subsequently describe it by contrasting it with what came later. I believe that it is easier to see the force of what Plato and Aristotle were saying when one sees how other people later developed what they said in various different directions, or, on the other hand, disagreed with it and offered alternatives. To this extent, the historian, looking backwards and forwards, has an advantage in understanding, and this is a reason for studying the texts of more than one period. By the same token, one of the hardest things for the historian is to be conscious of the bias of his or her own time and discount for that. At the height of the ‘ordinary language’ way of doing philosophy in England in the 1950s, Plato and Aristotle were often seen as ‘ordinary language’ philosophers. This fitted Aristotle better than Plato, because he so often does rest his case on appeals to ordinary language. But even he was doing much more than ordinary language philosophy. Just one of those many things was what we would nowadays call science.³¹ Now, no doubt, the biases of which we are least aware are the biases of the present time. It is very easy, by contrast, to see the biases of the Neoplatonist age, which was interpreting Aristotle as being in agreement with his teacher Plato, partly in an effort to establish the inner consistency of pagan religion in face of the crushing triumph of Christianity. Moreover, the Neoplatonists made Aristotle the introduction to a long teaching curriculum which was intended to culminate in mystical religious experience. But have I not, then, given away the case to the sceptics, by admitting that the Neoplatonists in 250–600 , 600 years and more after Aristotle, were hopelessly far away from understanding him? My earlier point that their misunderstanding created a fruitful new philosophy is no defence against this charge. Even Alexander back in 200 , the greatest of all the defenders of the Aristotelian school, distorts Aristotle in an attempt to show that Aristotelians already anticipate, where they don’t outright reject, the theories of the rival Stoic school. Surely the biases of our own time are going to be no smaller. That is true, but even the extreme case of Neoplatonism reveals that there is truth inextricably mixed with bias. So difficult was the task of presenting Aristotle as the harmonious introduction to Platonist religious revelation, that every word of the chosen texts had to be known and scrutinized. Again and again, the Neoplatonists investigate why Aristotle has put his point in a certain way, where modern readers might have skimmed over the wording without noticing anything striking. Repeatedly the Neoplatonists are right that Aristotle is echoing Plato. We do not have to agree with their interpretations, but then we do not agree with the interpretations of our contemporaries. As philosophers, we are professionals at disagreement. What the Neoplatonists often reveal is that Aristotle’s wording is chosen for a reason. We may disagree with them as to what the reason is. But we must thank them for drawing our attention to significant ³¹ Sorabji 1969.

Ideas Leap Barriers

389

wording that we might have missed. At the same time, our distance from the Neoplatonists positively helps us to discount for their biases. Our own ability to understand partially the ideas of Aristotle is thus positively enhanced by some of the intervening attempts to understand.

D O E S PH I LO S O PH Y P RO G R E S S ? I have recognized barriers within and between philosophical traditions, as well as suggesting that philosophical ideas can revive more freely than one might expect. In so far as I have stressed that revival is always possible, am I not committed to agreeing that philosophy does not progress? We have seen Williams attacking Hegel’s account of the progress of philosophy, and I agree with Williams, even if not for his reasons, that philosophy does not progress. What I would add on the other side is that each individual can progress in his or her personal search for understanding, and I think this makes philosophy of great value to the individual. But the best account I know of the status of philosophy as a whole was given to me by Burnyeat in our early days together. His comparison was with art. We do not think of ancient Egyptian painting, Leonardo da Vinci, and the Impressionists as representing linear progress. The earlier are not superseded; they are different, and we want them all. This is not to deny that there is progress. For example, the discovery of perspective constitutes progress, if realism is what you are seeking. But then Picasso was not necessarily seeking realism. Such is the status of philosophy. REFERENCES Brunschwig, J. (1991), ‘On a Book-Title by Chrysippus’, Oxford Studies in Ancient Philosophy, suppl. vol.: 81–95. Burnyeat, M. F. (1982), ‘Idealism and Greek Philosophy: What Descartes Saw and Berkeley Missed’, Philosophical Review, 90: 3–40. (1992), ‘Is an Aristotelian Philosophy of Mind Still Credible? A Draft’, in M. Nussbaum and A. Rorty (eds.), Essays on Aristotle’s De Anima (Oxford), 15–26. (2006), Introduction to Bernard Williams, The Sense of the Past: Collected Essays in the History of Philosophy (Princeton), pp. xiii–xxii. Halbfass, W. (1988), India and Europe. Albany, NY. Horner, I. B. (1964), Milinda’s Questions, trans. in Sacred Books of the Buddhists, xxii–xxiii. London. Kripke, S. (1972), Naming and Necessity, three lectures in G. Harman and D. Davidson (eds.), Semantics of Natural Languages (Dordrecht), 253–355, 762–9. Owens, D. (1992), Causes and Coincidences. Cambridge. Parfit, D. (1984), Reasons and Persons. Oxford. Plantinga, A. (1974), The Nature of Necessity. Oxford. Prior, A. N. (1967), Past, Present and Future. Oxford.

390

Richard Sorabji

Prior, A. N. (1968), Papers on Time and Tense. Oxford. Sorabji, R. (1969), ‘Aristotle and Oxford Philosophy’, American Philosophical Quarterly, 6: 127–35. (1980), Necessity, Cause and Blame. London and Chicago. (1983), Time, Creation and the Continuum. London and Chicago. (1988), Matter, Space and Motion. London. (1990) (ed.), Aristotle Transformed. London. (1992), ‘Intentionality and Physiological Processes’, in M. Nussbaum and A. Rorty (eds.), Essays on Aristotle’s De Anima (Oxford), 195–225. (1993), Animal Minds and Human Morals. London. (1997a) (ed.), Aristotle and After, Bulletin of the Institute of Classical Studies, suppl. 68. (1997b), ‘Is Stoic Philosophy Helpful as Psychotherapy?’, in Sorabji (1997a), 197–209. (2000), Emotion and Peace of Mind. Oxford. (2004), Philosophy of the Commentators 200–600 AD: A Sourcebook, iii: Logic and Metaphysics. London. (2006a), ‘Just War from Ancient Origins to the Conquistadors Debate and its Modern Relevance’, in R. Sorabji and D. Rodin (eds.), The Ethics of War: Shared Problems in Different Traditions (Aldershot), 13–29. (2006b), Self: Ancient and Modern Insights about Individuality, Life and Death. Chicago and Oxford. Steel, C. (1990), ‘Plato Latinus (1939–1989)’, in Rencontres de cultures dans la philosophie medievale: Traductions et traducteurs de l’antiquite tardive du XIV si`ecle (Louvain-laNeuve), 301–16. Taylor, C. (1989), Sources of the Self. Cambridge. Williams, B. (1973), Problems of the Self. Cambridge. (1994a), ‘Descartes and the Historiography of Philosophy’, in J. Cottingham (ed.), Reason, Will and Sensation (Oxford), 477–96. (1994b), ‘Do Not Disturb’, review of Martha Nussbaum, The Therapy of Desire, London Review of Books, 20 Oct., 25–6. (1997), ‘Stoic Philosophy and the Emotions: Reply to Richard Sorabji’, in Sorabji (1997a), 211–13. (2000), ‘Philosophy as a Humanistic Discipline’, Philosophy, 75: 477–96.

Myles Burnyeat: Publications A RT I C L E S A1 ‘Time and Pythagorean Religion’, Classical Quarterly, n.s. 12 (1962), 248–51. A2 ‘Belief in Speech’, Proceedings of the Aristotelian Society, 68 (1967/8), 227–48. A3 ‘The Material and Sources of Plato’s Dream’, Phronesis, 15 (1970), 101–22. A4 ‘Play and Pleasure’, Proceedings of the Philosophy of Education Society of Great Britain, 5 (1971), 29–36 (co-author Nancy Gayer). A5 ‘Virtues in Action’, in Gregory Vlastos (ed.), The Philosophy of Socrates (New York: Doubleday, 1971), 209–34. A6 ‘Protagoras and Self-Refutation in Later Greek Philosophy’, Philosophical Review, 85 (1976), 44–69; repr. in Terence Irwin (ed.), Classical Philosophy: Collected Papers, viii: Hellenistic Philosophy (New York and London: Garland Publishing Inc., 1995), 66–91. A7 ‘Protagoras and Self-Refutation in Plato’s Theaetetus’, Philosophical Review, 85 (1976), 172–95; repr. in Stephen Everson (ed.), Epistemology, Companions to Ancient Thought 1 (Cambridge: Cambridge University Press, 1990), 39–59. A8 ‘Plato on the Grammar of Perceiving’, Classical Quarterly, n.s. 26 (1976), 29–51; repr. in Japanese translation, Shiso, no. 694 (1982), 75–112. A9 ‘Examples in Epistemology: Socrates, Theaetetus and G. E. Moore’, Philosophy, 52 (1977), 381–98; repr. in Terence Irwin (ed.), Classical Philosophy: Collected Papers, ii: Socrates and His Contemporaries (New York and London: Garland Publishing Inc., 1995), 321–38. A10 ‘Socratic Midwifery, Platonic Inspiration’, Bulletin of the Institute of Classical Studies, 24 (1977), 7–16; repr. in Hugh H. Benson (ed.), Essays on the Philosophy of Socrates (Oxford and New York: Oxford University Press, 1992), 53–65, and in Greek translation in Deucalion, 12 (1993), 75–93; Bosnian translation forthcoming. A11 ‘The Philosophical Sense of Theaetetus’ Mathematics’, Isis, 69 (1978), 489–511. A12 ‘The Upside-Down Back-to-Front Sceptic of Lucretius IV 472’, Philologus, 122 (1978), 197–206. A13 ‘Reply’ to Wilbur Knorr, Isis, 70 (1979), 569–70. A14 ‘Conflicting Appearances’ (The 1979 Annual Dawes Hicks Lecture), Proceedings of the British Academy, 65 (1979), 69–111; repr. in Terence

392

A15

A16

A17

A18 A19

A20 A21 A22

Myles Burnyeat Irwin (ed.), Classical Philosophy: Collected Papers, iv: Plato’s Metaphysics and Epistemology (New York and London: Garland Publishing Inc., 1995), 303–45. ‘Can the Sceptic Live his Scepticism?’, in E3 (1980), 20–53; repr. in E5 (1983), 117–48, and in E9 (1997), 25–57; then in Croatian trans. in Pavel Gregori´c, Filip Grgi´c, and M. Hudoletnjak Grgi´c (eds.), Helenistiˇc ka filozofija: epikurovci, stoici, skeptici (KruZak: Zagreb, 2005), 423–53. ‘Aristotle on Learning to be Good’, in Amélie Oksenberg Rorty (ed.), Essays on Aristotle’s Ethics (Berkeley and Los Angeles: University of California Press, 1980), 69–92; repr. in Ted Honderich (ed.), Philosophy Through its Past (Harmondsworth: Penguin, 1984), 51–77; in Japanese trans. in Tadashi Inoué and Takashi Yamamoto (eds.), Girisha Tetsugaku no Saizensen (= The Forefront of Greek Philosophy) (Tokyo: University of Tokyo Press, 1986), ii. 86–124; English version repr. in Nancy Sherman (ed.), Aristotle’s Ethics: Critical Essays (Lanham, Md., Boulder, Colo., New York, and Oxford: Rowman & Littlefield, 1999), 205–30; German translation, ‘Lernen, ein guter Mensch zu sein: Aristoteles über moralische Bildung und Charakterentwicklung’, in Christoph Rapp and Tim Wagner (eds.), Wissen und Bildung in der antiken Philosophie (Stuttgart: Metzler, 2006), 215–37; Portuguese translation in a volume edited by Marco Zingaro, 2008. ‘Socrates and the Jury: Paradoxes in Plato’s Distinction between Knowledge and True Belief ’, Aristotelian Society Supplementary Volume, 54 (1980), 173–91; repr. in Japanese trans. in Tadashi Inoué and Takashi Yamamoto (eds.), Girisha Tetsugaku no Saizensen (= The Forefront of Greek Philosophy) (Tokyo: University of Tokyo Press, 1986), i. 146–69, and in French translation in Les Paradoxes de la connaissance: essais sur le Ménon de Platon, recueillis et présentés par Monique Canto-Sperber (Paris: Odile Jacob, 1991), 237–55; English original repr. in Nicholas D. Smith (ed.), Plato: Critical Assessments (London: Routledge, 1998), iii. 71–86. ‘Tranquillity without a Stop: Timon frag. 68’, Classical Quarterly, n.s. 30 (1980), 86–93. ‘Aristotle on Understanding Knowledge’, in E. Berti (ed.), Aristotle on Science: ‘The Posterior Analytics’, Proceedings of the Eighth Symposium Aristotelicum (Padua: Antenore, 1981), 97–139; repr. in Italian trans. in Giuseppe Cambiano and Luciano Repici (eds.), Aristotele e la Conoscenza (Milan: Casa Editrice Ambrosiana, 1993), 221–61. Foreword to B. A. F. Hubbard and E. S. Karnofsky, Plato’s Protagoras: A Socratic Commentary (London: Duckworth, 1982), pp. vii–ix. ‘Gods and Heaps’, in Malcolm Schofield and Martha Nussbaum (eds.), Language and Logos (Cambridge: Cambridge University Press, 1982), 315–38. ‘Idealism and Greek Philosophy: What Descartes Saw and Berkeley Missed’, Philosophical Review, 90 (1982), 3–40; also published in G. Vesey (ed.),

Publications

A23 A24 A25

A26

A27 A28

A29

A30 A31 A32 A33

393

Idealism: Past and Present, Royal Institute of Philosophy Lectures, 13 (Cambridge: Cambridge University Press, 1982), 19–50; repr. in Michael Williams (ed.), Scepticism, International Research Library of Philosophy, 5 (Aldershot: Dartmouth Publishing Co., 1993), 3–40. ‘The Origins of Non-Deductive Inference’, in E4 (1982), 193–238. Introduction to E5 (1983), 1–8. ‘The Sceptic in his Place and Time’, in Richard Rorty, J. B. Schneewind and Quentin Skinner (eds.), Philosophy in History (Cambridge: Cambridge University Press, 1984), 225–54; expanded version in Richard H. Popkin and Charles B. Schmitt (eds.), Scepticism from the Renaissance to the Enlightenment, Wolfenbütteler Forschungen Band, 35 (Wiesbaden: O. Harrassowitz, 1987), 13–43; repr. in Spanish trans. in Anales del Seminario de Metafísica, 27 (1993), 273–306; English version repr. in E9 (1997), 92–126. ‘Platonism and Mathematics: A Prelude to Discussion’, in Andreas Graeser (ed.), Mathematics and Metaphysics in Aristotle, Proceedings of the Tenth Symposium Aristotelicum (Bern and Stuttgart: Paul Haupt, 1987), 213–40; repr. in Croatian trans. in Pavel Gregori´c and Filip Grgi´c (eds.), Aristotelova Metafizika: Zbirka rasprava (Zagreb: KruZak 2003), 411–40. ‘Plato’ (printed version of M2), in Bryan Magee (ed.), The Great Philosophers, after the television series of the same name (London: BBC Books, 1987; Oxford: Oxford University Press, 1988), 12–30. ‘Wittgenstein and Augustine De Magistro’, Aristotelian Society Suppl. Vol., 61 (1987), 1–24; repr. in Gareth B. Matthews (ed.), The Augustinian Tradition (Berkeley, Los Angeles, and London: University of California Press, 1999), 286–303. ‘The Practicability of Plato’s Ideally Just City’, in K. Boudouris (ed.), On Justice: Plato’s and Aristotle’s Conception of Justice in Relation to Modern and Contemporary Theories of Justice (Athens: Greek Philosophical Society, 1989), 95–104; revised version entitled ‘Utopia and Fantasy: The Practicability of Plato’s Ideally Just City’, in Jim Hopkins and Anthony Savile (eds.), Psychoanalysis, Mind and Art: Perspectives on Richard Wollheim (Oxford: Blackwell, 1992), 175–87; repr. in Gail Fine (ed.), Plato 2: Ethics, Politics, Religion and the Soul, Oxford Readings in Philosophy (Oxford: Oxford University Press, 1999), 297–308. Foreword to E7, Phronesis, 35 (1990), 21–2. ‘Gregory Vlastos’, Phronesis, 37 (1992), 137–40. ‘Is an Aristotelian Philosophy of Mind Still Credible? A Draft’, in Martha C. Nussbaum and Amélie Oksenberg Rorty (eds.), Essays on Aristotle’s De Anima (Oxford: Clarendon Press, 1992), 15–26. ‘Aristôte voit du rouge et entend un : Combien se passe-t-il de choses? Remarques sur , II, 7–8’, Revue philosophique de la France et de l’Étranger, 118 (1993), 263–80, and in Cristina Viano

394

A34 A35 A36

A37

A38

A39 A40 A41 A42 A43 A44 A45

Myles Burnyeat (ed.), Corps et Âme: sur le De Anima d’Aristote (Paris: Vrin, 1996), 149–67; an earlier version in Spanish translation in Revista Latinoamericana de Filosofía, 20 (1994), 3–20; final English version published as ‘How Much Happens When Aristotle Sees Red and Hears Middle C? Remarks on De Anima 2, 7–8’, in the paperback edn. of Martha C. Nussbaum and Amélie Oksenberg Rorty (eds.), Essays on Aristotle’s De Anima (Oxford: Clarendon Press, 1995), 421–34. ‘Did the Ancient Greeks Have the Concept of Human Rights?’, Polis, 13 (1994), 1–11; text of M3, with additional material; Russian version published in Hyperboreus, 1 (1994), 19–27. Editor’s Preface to E8 (1994), pp. ix–xi. ‘Enthymeme: Aristotle on the Logic of Persuasion’, in David J. Furley and Alexander Nehamas (eds.), Aristotle’s Rhetoric: Philosophical Essays, Proceedings of the XIIth Symposium Aristotelicum (Princeton: Princeton University Press, 1994), 3–55; cf. also A37. ‘Enthymeme: Aristotle on the Rationality of Rhetoric’, in Amélie Oksenberg Rorty (ed.), Essays on Aristotle’s Rhetoric (Berkeley and Los Angeles: University of California Press, 1996), 88–115; re–presentation of most of A36, with additional material. ‘The Impiety of Socrates’, in Aminodav Dykman and Wlad Godzich (eds.), Platon et les Poètes: hommage à George Steiner (Geneva: Université de Genève, Faculté des Lettres, 1996), 13–36; repr. with revisions in Ancient Philosophy, 17 (1997), 1–12; in Thomas C. Brickhouse and Nicholas D. Smith (eds.), The Trial and Execution of Socrates: Sources and Controversies (New York and Oxford: Oxford University Press, 2002), 133–44; and in French translation in Methodos: Savoirs et Textes, 1 (2001), 207–22; repr. again in Rachana Kamtekar (ed.), Plato’s Euthyphro, Apology and Crito: Critical Essays (Lanham, Md., and Oxford: Roman & Littlefield, 2005), 150–162. ‘ ’ (‘Anger and Revenge’), Hyperboreus, 2 (1996), 3–20; Bulgarian trans. published by St Kliment Okhridsky University of Sofia (1999). ‘Antipater and Self-Refutation: Elusive Arguments in Cicero’s Academica’, in Brad Inwood and Jaap Mansfeld (eds.), Assent and Argument: Studies in Cicero’s Academic Books (Leiden: Brill, 1997), 277–310. Editors’ Preface to E9 (1997), pp. ix–xi (with Michael Frede). ‘First Words: A Valedictory Lecture’, Proceedings of the Cambridge Philological Society, 43 (1997), 1–20. ‘Postscript on Silent Reading’, Classical Quarterly, 47 (1997), 74–6; attached to T3. ‘Dissoi Logoi’, in Edward Craig (ed.), The Routledge Encyclopedia of Philosophy (London: Routledge, 1998), iii. 106–7. ‘The Past in the Present: Plato as Educator of Nineteenth-Century Britain’,

Publications

A46 A47 A48 A49 A50 A51

A52 A53

A54 A55 A56 A57

A58 A59 A60 A61

395

in Amélie Oksenberg Rorty (ed.), Philosophers on Education: Historical Perspectives (London and New York: Routledge, 1998), 353–73. ‘Culture and Society in Plato’s Republic’, The Tanner Lectures on Human Values, 20 (1999), 215–324; from 2001 also at ‘Plato on Why Mathematics is Good for the Soul’, in Timothy Smiley (ed.), Mathematics and Necessity: Essays in the History of Philosophy, Proceedings of the British Academy, 103 (Oxford: Oxford University Press, 2000), 1–81. ‘Plato’, Master Mind Lecture, Proceedings of the British Academy, 111 (2000), 1–22; Internet Journal of the International Plato Society (), June–Dec. 2001. ‘Aquinas on ‘‘Spiritual Change’’ in Perception’, in Dominik Perler (ed.), Ancient and Medieval Theories of Intentionality (Leiden, Boston, and Cologne: Brill, 2001), 129–53. ‘James Mill on Thomas Taylor’s Plato: Introduction’, Apeiron, 34 (2001), 101–10 (introduction to E10). ‘What was ‘‘the Common Arrangement’’? An Inquiry into John Stuart Mill’s Boyhood Reading of Plato’, Utilitas, 13 (2001), 1–32; Philologus, 145 (2001), 158–86; Apeiron, 34 (2001), 51–90 (with the prior agreement of the editors of all three journals). ‘De Anima II 5’, Phronesis, 47 (2002), 1–64. ‘Plato on How Not to Speak of What is Not: Euthydemus 283a–288a’, in Monique Canto-Sperber and Pierre Pellegrin (eds.), Le Style de la Pensée: Recueil de textes en hommage à Jacques Brunschwig (Paris: Les Belles Lettres 2002), 40–66. ‘Σωκράτης και σοϕιστές’, in Βασίλης Καρασμάνης (ed.), Σωκράτης: Σοφός που δε Γνώριζε Τίποτα (Athens: Livani 2002), 61–77; trans. into Greek from the transcript of M7. ‘Apology 30b 2–4: Socrates, Money, and the Grammar of γίγνεσθαι’, Journal of Hellenic Studies, 123 (2003), 1–25. ‘Aristotelian Revisions : The Case of De Sensu’, Apeiron, 37 (2004), 177–80. ‘Aristotle on the Foundations of Sublunary Physics’, Introduction to Frans de Haas and Jaap Mansfeld (eds.), Aristotle: On Generatione et Corruptione, Book I, Proceedings of the XVth Symposium Aristotelicum, (Oxford: Clarendon Press, 2004), 7–24. ‘Fathers and Sons in Plato’s Republic and Philebus’, Classical Quarterly, 54 (2004), 80–7. ‘Ryle, Gilbert (1900–76)’, in Robert B. Todd (ed.), Dictionary of British Classicists (Bristol: Thoemmes Continuum, 2004), iii. 846–9. ‘Archytas and Optics’, in Reviel Netz and Serafina Cuomo (eds.), The History of Early Science, Science in Context, 18 (2005), 35–53. ‘On the Source of Burnet’s Construal of Apology 30b 2–4 : A Correction’, Journal of Hellenic Studies, 125 (2005), 139–42.

396

Myles Burnyeat

A62 ‘Platonism in the Bible: Numenius of Apamea on Exodus and Eternity’, in Ricardo Salles (ed.), Metaphysics, Soul, and Ethics in Ancient Thought: Themes from the Work of Richard Sorabji (Oxford: Oxford University Press, 2005), 43–169; repr., with mild editing, in George H. van Kooten (ed.), The Revelation of Yahweh’s Name to Moses: Perspectives from Judaism, the Pagan Graeco-Roman World, and Early Christianity, Themes in Biblical Narrative, 9, (Leiden and Boston: Brill, 2006), 139–68. A63 ‘The Truth of Tripartition’, Proceedings of the Aristotelian Society, 106 (2005–6), 1–23; forthcoming also in French trans. in a volume ed. Gwenaëlle Aubry (Paris: Vrin, 2008). A64 ‘Εἰκὼς Μῦθος’ Rhizai, 2 (2005), 143–65; to be repr. in Catalin Partenie (ed.), Plato’s Myths (Cambridge: Cambridge University Press) and in Romanian trans. in a bilingual edn. of the Timaeus and Critias, ed. Catalin Partenie (Humanitas Press). A65 Introduction to E12 (2006), pp. xiii–xxii.

BOOKS B1 The Theaetetus of Plato (Indianapolis: Hackett Publishing Co., 1990); repr. with corrections 1992; French trans. by Michel Narcy of the Introduction (pp. 1–241) published separately as Introduction au Théétète de Platon (Paris: Presses Universitaires de France, 1998); pp. 7–31 and 39–52 of the original repr. in Gail Fine (ed.), Plato 1: Metaphysics and Epistemology, Oxford Readings in Philosophy, (Oxford: Oxford University Press, 1999), 320–54. B2 A Map of Metaphysics Zeta (Pittsburgh: Mathesis Publications, 2001).

B O O K S A N D OT H E R I T E M S E D I T E D O R C O L L A B O R AT E D I N E1 Philosophy As It Is, ed. and introduced by Ted Honderich and Myles Burnyeat (London: Allen Lane, 1979). E2 Notes on Zeta: Notes on Book Zeta of Aristotle’s Metaphysics, being the record by Myles Burnyeat and others of a seminar held in London, 1975–79, Oxford Study Aids Monograph, 1 (Oxford: Subfaculty of Philosophy, 1979). E3 Doubt and Dogmatism: Studies in Hellenistic Epistemology, ed. Malcolm Schofield, Myles Burnyeat, and Jonathan Barnes (Oxford: Clarendon Press, 1980; repr. in paperback, 1989).

Publications

397

E4 Science and Speculation: Studies in Hellenistic Theory and Practice, ed. Jonathan Barnes, Jacques Brunschwig, Myles Burnyeat, and Malcolm Schofield (Cambridge: Cambridge University Press; Paris: Maison des Sciences de l’Homme, 1982; repr. in paperback, 2005). E5 The Skeptical Tradition, Major Thinkers Series, ed. by Myles Burnyeat (Berkeley, Los Angeles, and London:, University of California Press, 1983). E6 Notes on Eta and Theta of Aristotle’s Metaphysics, recorded by Myles Burnyeat and others, Oxford Study Aids Monograph, (Oxford: Subfaculty of Philosophy, 1984). E7 †G. Ryle, ‘Logical Atomism in Plato’s Theaetetus’, Phronesis, 35 (1990), 21–46. E8 †Gregory Vlastos, Socratic Studies, ed. Myles Burnyeat (Cambridge: Cambridge University Press, 1994). E9 The Original Sceptics: A Controversy, ed. by Myles Burnyeat and Michael Frede (Indianapolis: Hackett Publishing Co., 1997). E10 † James Mill, reviews of Thomas Taylor, The Works of Plato (1804), Apeiron, 34 (2001), 111–79. E11 †Heda Segvic, (i) ‘Aristotle on the Varieties of Goodness’, Apeiron, 37 (2004), 151–76; (ii) ‘Protagoras’ Political Art’, Rhizai, 1 (2004), 9–36; (iii) ‘Homer in Plato’s Protagoras’, Classical Philology, 101 (2006), 247–62; (iv) ‘Deliberation and Choice in Aristotle’, forthcoming in Giles Pearson and Michael Pakaluk (eds.), Psychology and Human Action in Aristotle; (v) forthcoming book From Protagoras to Aristotle: Essays in Ancient Moral Philosophy, containing (i)–(iv) plus other already published pieces, Princeton University Press. E12 †Bernard Williams, The Sense of the Past: Collected Essays in the History of Philosophy, ed. with an Introduction by Myles Burnyeat (Princeton and Oxford: Princeton University Press, 2006). T R A N S L AT I O N S T1 (French) with Jennifer Barnes: Pierre Couissin, ‘Le Stoïcisme de la Nouvelle Académie’, Revue d’histoire de la philosophie, 1929, for E5 (1983), 31–63. T2 (Ancient Greek) Revision of M. J. Levett, The Theaetetus of Plato (Glasgow: Glasgow University Press, 1928), incorporated in B1 and separately published in Plato: Theaetetus, ed. with introduction by Bernard Williams (Indianapolis: Hackett Publishing Co., 1992); repr. with revisions in Plato: Complete Works, ed. John M. Cooper (Indianapolis: Hackett Publishing Co., 1997), 157–234; regularly revised since then. T3 (Russian) A. K. Gavrilov, ‘ ’ (‘Techniques of Reading in Classical Antiquity’), Classical Quarterly, 47 (1997), 56–73.

398

Myles Burnyeat

T4 (Latin) Quantities of Baroque Latin in E10. ’ (‘Tiger Hunt’), with Ruth T5 (Russian) Il’ya Siel’vinsky, ‘ Padel, in Ruth Padel, Tigers in Red Weather (London: Little, Brown, 2005), 396. T6 (Russian) Irina Levinskaya and Sergei Tokhtas’ev, ‘Response Article: The Album of Bosporan Inscriptions on Trial: A Further Indictment’, Bryn Mawr Classical Review 2006.02.15, 1–8. R EV I EW S R1 Gilbert Ryle, Plato’s Progress, Listener, 23 Feb. 1967. R2 Alasdair MacIntyre, A Short History of Ethics, Listener, 17 Aug. 1967. R3 G. F. Parker, A Short Account of Greek Philosophy, Times Literary Supplement, 24 Aug. 1967. R4 Alvin Gouldner, Enter Plato, Times Literary Supplement, 26 Oct. 1967. R5 John Ferguson, Socrates, Listener, 6 May 1971. R6 K. T. Fann, Wittgenstein’s Conception of Philosophy; Rush Rhees, Discussions of Wittgenstein; David Pears, Wittgenstein; Ludwig Wittgenstein, Prototractatus, Philosophische Grammatik, and On Certainty, Times Literary Supplement, 17 Sept. 1971. R7 A. J. Ayer, Russell and Moore: The Analytical Heritage, Listener, 27 Jan. 1972. R8 John McDowell, Plato Theaetetus, Times Literary Supplement, 24 May 1974. R9 R. S. Bluck, Plato’s Sophist, Times Literary Supplement, 18 Apr. 1975. R10 K. J. Dover, Greek Popular Morality in the Time of Plato and Aristotle, Listener, 8 May 1975. R11 David Pears, Questions in the Philosophy of Mind, Times Literary Supplement, 22 Aug. 1975. R12 Leo Strauss, The Argument and the Action of Plato’s Laws; J. C. B. Gosling, Plato Philebus, Times Literary Supplement, 9 Apr. 1976. R13 David Gallop, Plato Phaedo, Times Literary Supplement, 7 May 1976. R14 C. C. W. Taylor, Plato Protagoras, Times Literary Supplement, 17 Sept. 1976. R15 R. S. Peters, ed., Nature and Conduct, Times Literary Supplement, 24 Sept. 1976. R16 Leonard Brandwood, A Word Index to Plato, Times Literary Supplement, 22 Apr. 1977. R17 Roderick Chisholm, Person and Object, Times Literary Supplement, 26 Aug. 1977. R18 Alice Swift Riginos, Platonica: The Anecdotes concerning the Life and Writings of Plato, Times Literary Supplement, 23 Dec. 1977.

Publications

399

R19 John M. Cooper, Reason and Human Good in Aristotle, Philosophical Review, 87 (Jan. 1978). R20 Nicholas P. White, Plato on Knowledge and Reality, Philosophical Review, 87 (Oct. 1978). R21 C. J. Classen, ed., Sophistik, Classical Review, n.s. 28 (1978). R22 E. N. Tigerstedt, Interpreting Plato, Classical Review, n.s. 29 (1979). R23 J. N. Findlay, Plato: The Written and Unwritten Doctrines and Plato and Platonism; Terence Irwin, Plato’s Moral Theory; New York Review of Books, 27 Sept. 1979. R24 Fritz Wehrli, Sotion, Classical Review, n.s. 30 (1980). R25 W. K. C. Guthrie, A History of Greek Philosophy, v: The Later Plato and the Academy, Philosophical Review, 90 (Jan. 1981). R26 Donald Regan, Utilitarianism and Cooperation, Times Literary Supplement, 20 Feb. 1981. R27 Martha Nussbaum, Aristotle’s De Motu Animalium, Archiv für Geschichte der Philosophie, 52 (1981). R28 Ronna Burger, Plato’s Phaedrus: A Defense of a Philosophic Art of Writing, Classical Review, 31 (1981). R29 Charles H. Kahn, The Art and Thought of Heraclitus, New York Review of Books, 13 May 1982. R30 Robert Nozick, Philosophical Explanations, Times Literary Supplement, 15 Oct. 1982. R31 Ian Mueller, Philosophy of Mathematics and Deductive Structure in Euclid’s Elements, Times Higher Education Supplement, 15 Oct. 1982. R32 R. E. Allen, Plato’s Parmenides: Translation and Analysis; Kenneth M. Sayre, Plato’s Late Ontology: A Riddle Resolved; Times Literary Supplement, 13 July 1984. R33 Leo Strauss, Studies in Platonic Political Philosophy, New York Review of Books, 30 May 1985; repr. in Chicago Maroon, 8 Oct. 1985, and in Nicholas D. Smith (ed.), Plato: Critical Assessments, (London: Routledge, 1998), i. 333–48; French trans. in Raisons politiques: Études de pensée politique, 16 (Nov. 2004), 61–81. R34 Jonathan Barnes, ed., The Complete Works of Aristotle: The Revised Oxford Translation, London Review of Books, 6 Nov. 1986. R35 I. F. Stone, The Trial of Socrates, New York Review of Books, 31 Mar. 1988. R36 Trevor J. Saunders, Plato’s Penal Code: Tradition, Controversy, and Reform in Greek Penology, The Times Saturday Review, 14 Dec. 1991. R37 Francis Fukuyama, The End of History and the Last Man, London Review of Books, 23 July 1992. R38 C. A. J. Coady, Testimony: A Philosophical Study, London Review of Books, 4 Nov. 1993. R39 William M. Calder III and Stephen Trzaskoma, eds., George Grote Reconsidered: A 200 th Birthday Celebration, Times Literary Supplement, 20 June 1997.

400

Myles Burnyeat

R40 Michael Frede and Gisela Striker, eds., Rationality in Greek Thought, Times Higher Educational Supplement, 10 Oct. 1997. R41 Alain Martin and Oliver Primavesi, L’Empédocle de Strasbourg, Times Literary Supplement, 28 May 1999. R42 J. E. Raven, Plants and Plant Lore in Ancient Greece, Times Literary Supplement, 8 June 2001. R43 Anthony Gottlieb, The Dream of Reason: A History of Western Philosophy from the Greeks to the Renaissance, New York Review of Books, 1 Nov. 2001. R44 Gerasimos Santas, Goodness and Justice: Plato, Aristotle and the Moderns, Times Literary Supplement, 14 June 2002. R45 William Harris, Restraining Rage: The Ideology of Anger Control in Classical Antiquity, London Review of Books, 17 Oct. 2002; repr. in Bosnian trans. in Odjek, Autumn–Winter 2002. R46 Geoffrey Elliott and Harold Shukman, Secret Classrooms: An Untold Story of the Cold War, Times Literary Supplement, 14 Mar. 2003. R47 A. A. Long, Epictetus: A Stoic and Socratic Guide to Life, Times Literary Supplement, 11 Apr. 2003. R48 Ruby Blondell, The Play of Character in Plato’s Dialogues, London Review of Books, 7 Aug. 2003. R49 Andrea Nightingale, Spectacles of Truth in Classical Greek Philosophy: Theoria in its Cultural Context, Times Literary Supplement, 24 Feb. 2006; to be repr. at . R50 Christof Riedweg, Pythagoras: His Life, Teaching and Influence; Charles H. Kahn, Pythagoras and the Pythagoreans: A Brief History, in London Review of Books, 22 Feb. 2007. JOURNALISM J1 ‘Homes before Roads’, Tribune, 24 Apr. 1970 (co-author David Wiggins). J2 ‘A question of origins’, Times Literary Supplement, 15–21 June 1990; complete version published as ‘The Importance of Classics’, Cambridge Review, 111 (1990), 120–4; later repr. in Greek trans. in Indiktos, 5 (June 1996), 25–42. J3 ‘Freelance’ (Ancient Philosophy in Leningrad), Times Literary Supplement, 26 July 1991. J4 ‘Professor Richard Sorabji’, The Classical Association News, Dec. 1991. J5 ‘Urbs Sancti Petri’, Omnibus, 24 (Sept. 1992), 24–5. J6 ‘Classics and Democracy’, Times Literary Supplement, 15 Oct. 1993. J7 ‘Urbs Sancti Petri: A Second Visit’, Omnibus, 28 (Sept. 1994), 13. J8 ‘Plato in St Petersburg’, Times Literary Supplement, 13 Sept. 1996. J9 ‘A Farewell to Robinson College’, Robinson College Record, Spring 1997. J10 Interview, Harvard University Gazette, 11 Dec. 1997.

Publications

401

J11 ‘Art and Mimesis in Plato’s Republic’, London Review of Books, 21 May 1998 (version of Part 2 of A46); repr. in Nigel Warburton (ed.), Philosophy: Basic Readings (London and New York: Routledge, 2006), 457–74. J12 ‘Letter from Budapest’, Times Literary Supplement, 4 Dec. 1998. J13 ‘Letter from Sofia’, Times Literary Supplement, 16 Apr. 1999. J14 Extract from ‘Plato’ (A48), British Academy Review, Jan.–July 2000. J15 ‘John Griffin’, Old Bryanstonian Yearbook, 65 (2000–1). J16 ‘Siberian Diary’, London Review of Books, 19 Feb. 2004. J17 Interview ‘Platona e¯n¯a’, Rigas Laiks (Riga, Latvia), Apr. 2005, 40–5. J18 ‘Reflections on Returning’, Bin Brook (Robinson College Magazine), Lent Term 2007, 7.

MEDIA M1 ‘Ancient Science’, conversation on BBC World Service, 1978. M2 ‘Plato’, conversation with Bryan Magee in the television series The Great Philosophers, BBC2, 6 Sept. 1987. Written up as A27. M3 ‘Ancient Politics and Ours’, talk on Radio 3, 9 Apr. 1992; text published as A34. M4 ‘Ad Lib. Robert Robinson meets philosophers’, discussion on Radio 4, 24 Apr. 1993; repeated 17 June 1993. M5 ‘Aristotle’, interview in the radio series Western Philosophers in a Nutshell, BBC World Service, 20 and 22 Sept. 1993. M6 Contributions to BBC World Service series ‘Thinking Big’: (1) ‘What is Philosophy?’, (2) ‘Plato’, Spring–Summer 2000. M7 ‘Socrates and the Sophistic Movement’, contribution to Greek television programme ‘Socrates, 2400 years from his death’, Summer 2001; written up as A54.

This page intentionally left blank

General Index This index is in three parts—a general index, a name index, and an index locorum. Figures and notes are indexed in bold.

Academic philosophy 322, 324–5, 353 affirmations 303–5, 307–9, 314, 316, 320, 322, 323 see also negations aggression 176, 178, 179, 180, 189, 190–3 alteration 276, 285, 290, 291 ambiguity 304, 310, 311, 314, 318, 327–8 of expressions 309–10, 319 see also disambiguity amphiboly 301, 302, 310 andreia, see courage animals 282, 283–5, 295–7, 298, 299 appearances 14–15, 103–4, 107, 334 appetites 137–9, 145, 152, 184, 191, 192, 193, 245, 246 artefacts 232–5, 237–8, 239, 240, 242, 243, 244, 245, 251, 278, 279 arts 176, 181, 182–3, 185, 187, 188, 190, 193, 220, 331 Athens 42, 46, 47, 57, 63, 102, 103, 159, 179, 185, 223 beauty 97, 102, 104–7, 114, 117–18, 122–6, 129, 131–3, 179, 230–1 of boys 120–1 Form of 108–12, 116, 118, 123–4 and goodness 117, 120, 130 of human body 108, 114, 129, 131, 143 and love 107, 112, 114, 132 pursuit of 118–19, 133 see also kalon belief 15–16, 39–40 bivalence 353, 358, 361, 363, 367, 371 bodies: changes in 283, 298–9 pleasures of 51, 53, 65 brain-in-a-vat hypothesis 203 Buddhism 385, 386 cataleptic impressions 359–61 Cave 147, 148, 151, 195, 213–14, 220–1, 229 as allegory 230

interpretation 217, 221–8, 223–6, 228–31 mathematics 217–19 prisoners in 195–6, 218, 209, 223, 228–9 conditions of 197, 199, 203, 204–9, 212–13, 226; and use of language 196–8, 199–202, 204, 205, 207–11, 213; and shadows 200–201, 204, 206–11, 213, 222–8; change and rest 277, 278, 284, 185, 294, 297 channel argument (Republic VI) 136, 147, 152 Christianity 377–8, 384, 387 classical logic 361, 362, 369, 371 see also logic Cogito 376, 387 cognitive states 3, 5, 7, 9, 16 common nouns 303, 306 community life 181–6 conversion 221–2, 224, 225–6, 228, 229 corruption 56–7, 58 cosmos 235, 236, 237, 238, 241, 246, 247, 249, 250 courage (andreia) 39, 40, 41, 42, 45, 52, 55, 59, 60, 70, 157, 169, 177–80, 221 of philosophers 64, 65 Socrates on 47, 51, 58, 59, 61 crafts 235, 237, 247 darkness 219, 224, 226, 255, 260, 271, 273 death 51–2, 64, 95, 150, 246 Demiurge 236–7, 238–9, 241, 245, 246, 249, 250 democracy 137, 138, 140, 141, 220 desires 3, 15, 100, 107, 116–8, 131, 138, 161, 190, 192, 245 bodily 50, 58, 60 and hydraulic effect 45, 50, 51, 54–7, 60, 61–5 dialectic 38–9, 202, 220, 270, 301, 307–8, 312, 319, 321 dialectical theorem 367, 369–70 digestion 249, 280, 282–3, 286, 287, 289

404

General Index

disambiguity 9, 309, 311, 312, 319, 322 see also ambiguity diseases 335–6, 338–48 diagnostics 342, 345–9, 355 dogmatist philosophy 322, 324, 325, 331, 332, 352, 354, 364, 367, 369 education 64, 66, 128, 130, 148, 178, 213, 218, 219, 220–1 of young warriors 189–90 Eleatics 254, 256, 257, 258, 260, 263, 264, 265, 266, 267, 269, 270, 271–3 emotion 3, 5, 7, 160, 171, 191, 245, 379, 380, 381, 387 Empiricists 354, 364, 367, 369 endurance (karteria) 59, 60, 61 Epicureans 26, 31, 32, 377 epistemicism 362, 369 equality 68–85, 91 Equals 82, 83, 84–5 er¯os, see love erotic relations 157, 160, 161 with boys 48, 100 ethics 69, 157, 377, 380, 383 eudaimonia, see happiness ex-prisoners 226–7 see also Cave, prisoners falsehood 79, 256, 257, 270, 360 five modes of Agrippa 333–4 Forms 68–9, 70, 72–5, 78, 81, 83–5, 88–91, 92–4, 123, 152, 179, 196, 250–1, 255, 258, 267 hypothesis of 87–9, 91 frame dialogue 9, 11, 15–17 God, as creator of world 375 gods 32, 177, 328 good 117, 120, 129, 131, 179, 189, 218, 219, 228, 230, 255 happiness (eudaimonia) 64, 107, 117, 126, 131, 133, 149 health 69, 70, 110 higher-order thought 7, 8, 17, 18 Hippocratic writers 336, 341 homonymy 306, 309, 310, 312, 320–1 in dialectic 308–9, 311, 321 of expressions 301–5 in questions 307, 311 honour 116, 130, 132, 137, 150, 152, 158, 178, 192 human beings 245–7, 249–50, 317 human life 51–2, 97–8 Socrates on 24, 28, 30, 32, 35, 39, 51

human nature 39, 41, 192 idealism 375–6, 381 ideas, see Forms immortality, desire for 108, 131, 150 impiety, charges against Socrates of 32, 222 Indian philosophy 385–6 inequality 76, 78, 79, 84–5 inquiring into oneself 8, 14 inquiry, see investigation investigation 322, 324, 325, 327–30, 333–5 Islamic philosophy 386–7 Joint Illumination (Sophist) 258–60, 262–3, 267, 268 just war 182, 382 justice 39, 40, 41, 42, 45, 52, 55, 58, 130, 148, 179, 180, 214, 222, 223, 225, 228, 229 kalon 98, 99, 100, 101, 102, 126–7, 129, 130, 133 see also beauty karteria, see endurance knowledge 1–2, 4, 5, 17, 31, 32, 35, 45, 51, 55, 59, 64, 66, 123, 130, 145, 213, 256, 290, 291, 294, 295, 326 pursuit of 65, 118 and Socrates 44, 46, 48–9, 58, 61 see also self-knowledge largeness 9, 69, 70, 71, 82, 83, 221 late-learners (Sophist) 268–70 Latin 29, 165, 166, 170n, 365–6, 384–7 lawless desires 138–40, 141, 152, 191–2 learning 221, 213, 291–2 see also love of learning life, way of 20–9, 30–2, 40, 42, 89 living beings 279–80, 282, 283, 284, 285, 294, 299 see also human beings locomotion 278, 280, 282, 283, 284, 289 logic 38, 361, 362, 364–8, 382 Aristotle on 309 of signs 349 Stoic 353, 356, 357, 370, 371 see also classical logic logos 261, 264, 265, 266–9, 270, 271, 272, 273, 366 love (er¯os) 97, 106, 107, 108, 109–12, 115, 116, 117, 118–19, 124, 131, 138, 140–2, 147, 149–52 of learning 45, 46, 50, 51, 130, 136 and madness 140, 141, 142, 143, 146 and obsession 136–7, 146

General Index philosophical 142, 148–9 lovers 119, 131, 143, 144, 145, 151, 152, 161, 173 of boys 112–14, 118–21, 123 philosophical 114–15, 116, 118, 121, 132, 136, 144, 146, 149, 150 of women 115–16 madness 140–3, 146 manliness, see courage mathematics 217–20, 221, 229, 231, 235, 237–8 metaphysics 243, 246 Aristotle on 252 Plato on 236 Xenocrates on 239 modus ponens 353, 361, 366, 367, 368, 371 monism 258, 264, 266, 269 mortals, see human beings motion 279, 282, 283, 298 and rest 258, 260, 262, 269 natural philosophy 50–4, 55–8, 63, 65–6, 276–7 natural substances 243–4, 277, 285, 287–8 natural virtue 45, 46, 55–7, 58, 64–5 in Republic 59, 60, 63 see also virtue nature 276, 277, 278 negations 304, 305, 308, 316 see also affirmations Neoplatonism 21, 376, 384, 388–9 numbers 21, 76, 186, 238, 257–8, 326, 330 nutriment 286–8, 289, 290, 295 nutrition 280–2, 286, 287, 289, 290, 292, 293, 294, 299 nutritive soul 244–5, 252, 285, 288, 290, 296–8 paederasty 108, 115, 159, 161, 162, 168–169, 171, 174 Parity Assumption (Sophist), see Joint Illumination peace 177–8, 182, 188, 190 perception 1, 2, 3, 5–6, 7–12, 14–16, 17, 72, 214, 230, 276, 291–2, 293, 294, 295–6, 297, 299, 387 see also self-perception philosopher-rulers 59, 147–51, 185, 189, 191–2, 218, 220 philosophy: Burnyeat on 154–5 history 374–5, 382–4, 387–9 phron¯esis 59, 60, 64 piety 40, 42, 152

405

Platonic dialogues 25, 36, 42, 158, 176, 226 Platonism 32, 74–5, 154, 168, 235, 237, 239–41, 244, 251, 326, 377, 388 poetry 101, 130, 131 pride 126–7, 128, 130, 133 prisoners, see Cave, prisoners propositions 358, 369 complex 362–3 psychological states 3, 5, 6, 7, 9, 12, 140, 363 questions 303, 308, 309, 310, 311–12, 363 reasoning 23, 25, 26, 28, 30, 31, 42, 130, 131–2 recollection 68, 70, 72, 74, 75, 77, 80, 144, 213 reflexivity 2, 3, 5, 6, 7, 8, 18 relations 2, 3, 5, 7, 9, 10, 13, 14, 15–17, 141, 142 religion 32, 152, 378 reproduction 246, 247, 248, 249–52, 296, 297, 298 rhetoric 51, 52, 166, 167, 223, 335, 337, 365 rulers, see philosopher-rulers sceptical philosophy 333, 352, 367, 376, 384, 388 Sextus on 322, 325–6, 328–34 sciences 188, 189, 220, 233, 235, 245 self-control, see s¯ophrosun¯e self-knowledge 1–2, 3, 5, 6–8, 9, 11–12, 14, 16–18 see also knowledge self-motion, see locomotion self-perception 1, 3, 11, 18 self-refutation 352, 368–9, 371 semantics 353, 362, 363, 367 of vagueness 363, 364, 369, 371 sense objects 284, 290, 292, 293, 294–6, 299 sense organs 297–9, 387 sexuality 113–15, 123 sight 3, 6, 9–10, 13, 14 see also perception signs: medical 336–49 theories of 335–6 social concerns 141, 144, 146–7, 151 society, war-free 185–8 Socratic dialogues (Plato) 33, 34, 35, 37, 41 sophists 33, 256–7, 220, 267, 270, 271–3 s¯ophrosun¯e (self-control, temperance) 1–2, 8, 11–14, 39, 40, 42, 46, 50–2, 54, 55, 60, 63, 65 Socrates on 47, 48, 58, 59, 60, 61, 100

406

General Index

sorites 351–71 soul 41, 58, 71, 87, 91, 92, 95, 101, 116, 130–1, 141, 143, 220–1, 245, 248, 276–7, 299 beauty of 114, 115, 121, 122, 129 education of 130, 213–14 as first-mover 282–3, 289 good of 32, 34, 40, 42 of philosopher 143–4, 145–6 pleasures of 51, 53, 65 reorientation of 217–19 Socrates on 15, 17, 25, 27, 28, 30, 31, 37–8, 88, 92–5 speaking 265, 266–8, 273, 318–19 SPEM (Semantic Principle of Excluded Middle) 304–6 statements 303–6, 307, 308, 309, 310, 312, 314, 315–16, 318, 320 Stoics 26, 29, 31, 32, 42, 322, 353, 356–61, 367, 370, 371, 377, 378, 380–1, 387, 388 suspension of judgement 328–33, 360 syllogism 77–82, 84, 85 temperance, see s¯ophrosun¯e Theban Sacred Band 158–9, 163, 168, 169, 173 thinking 278, 291, 296 timocracy 137, 138, 152, 177 transitivity 6, 13, 16, 363 Trojans 168, 169, 174

truth 40, 129, 136, 243, 254, 263, 266, 322, 369–71, 388 tyranny 136–8, 152, 177 and philosophy 141, 146, 151 Twin Earth 203–206 utterance 265, 266, 268 values 25, 28, 30–1, 224, 228, 229 human 30, 37–8 virtue 1, 2, 14, 24, 28, 39–40, 42, 44, 52, 53, 69, 100, 103, 117, 123, 124, 125, 129, 131, 133, 150–1, 156, 180 full 46, 52, 58, 64, 65 Socrates on 47, 48–9, 59, 59n, 60, 62, 63 true 64, 132 see also natural virtue war 179, 189 Hobbes on 182 origin of 180–8 Plato on 176–7, 179, 180–4 Socrates on 190–1 what is 254–71, 271f, 272–3 what is not 254–71, 271f, 272–3 wisdom 31, 32, 39, 40, 41, 52, 64, 108, 110, 112, 129, 130, 131, 248 divine 32 human 29, 38 women 173–4

Name Index Achilles 62, 98, 108n, 127, 160, 161, 171, 172n Ackrill, J. 304n, 307n, 312, 313, 315, 316, 317n, 318, 320, 321 Adam, J. 53n, 137n, 180n, 197, 198, 199, 200, 203 Adeimantus 138, 140, 180 Admetus 159, 160 Aeneas 166, 167, 168, 170, 170n, 171n, 172, 174 Aeschylus 171 Agamemnon 337 Agathon 99, 107, 111, 128, 157 Alcestis 108n, 159, 160 Alcibiades 46–7, 48–9, 55, 57, 58, 59, 60, 61, 62, 63, 64, 98, 102, 113, 114, 121, 122, 129n, 162n, 168 Alcmaeon of Croton 13n Alexander the Great, King of Macedonia 234n, 385–6, 388 Allen, J. 337 Anaxagoras 21n, 337 Anaximander 21n Anaximenes 21n Annas, J. 137n, 140n, 181n, 216, 359n, 361n Antiochus 387 Aphrodite 127 Apollonius of Rhodes 162, 261n Aquinas, Thomas 384 Archer-Hind, R. D. 94n Aristodemus 168 Aristophanes 24n, 48n, 62n, 102, 106, 158, 159n, 338, 370n Armstrong, D. M. 94n Arnott, W. G. 94n Ascanius 165, 167, 168, 173n, 174 Asclepiades 354 Athena 127 Athenaeus 183n Augustine 376, 382, 387 Austin, R. G. 170n Avicenna 386, 387 Bailey, D. 2n Baldry, H. C. 183n Barker, E. 186n Barnes, J. 183n, 232n, 240n, 352n, 353n, 354n, 356n, 357n, 358n, 359n, 360n,

361n, 362n, 363n, 364, 366n, 367n, 370n, 371n Barney, R. 180n, 185n, 188n, 192n Bastianini, G. 34n Beere, J. 116n Berkeley, George 375–7 Bestor, T. W. 195n Biglieri, E. F. 347n Blackburn, S. 216n Bleisch, P. R. 169n, 170n, 172 Bloom, A. 137n, 152n Bluck, R. S. 73n, 82n, 94n Bobzien, S. 357n, 358n, 359n, 360n, 361n, 362n, 367n, 369 Boethius 385 Bonitz, H. 234 Bostock, D. 51n, 73n, 75n, 94n Boter, G. 198n Bouchier, A. D. 347 Brandwood, L. 177n Brennan, T. 359n Brentano, F. 387 Brentlinger, J. T. 109 Brisson, L. 183n, 186n Broadie, S. 98 Brown, E. 148, 150n Brown, L. 259 Brunschwig, J. 13n, 197n, 224n, 359n, 366n, 377n Burge, T. 205, 206, 208, 210, 211 Burkert, W. 23n Burnet, J. 34n, 41n, 79n, 94n, 197, 198 Burnyeat, Myles 2n, 6n, 12n, 34, 41n, 68, 103n, 108n, 130n, 132n, 154, 181n, 183n, 189n, 195n, 197n, 214n, 217n, 219n, 220n, 228n, 230n, 233n, 234n, 242n, 248n, 256n, 276, 276n, 286n, 290, 291n, 292, 293n, 294n, 301n, 326, 335, 336, 337, 351, 352, 353n, 355n, 356n, 357n, 358n, 360, 361, 362n, 363, 364, 366n, 371n, 376–7, 380 Bury, R. G. 178n Calchas 337 Callias 302, 307 Cambiano, G. 186n Campbell, L. 260, 261, 264n Carneades 363 Cato of Utica 29, 30, 31

408

Name Index

Catullus 106 Cherniss, H. 234n, 235, 236n Chrysippus 352, 353, 358, 359, 360, 361, 362, 363, 369, 370, 371, 378, 380, 381n Church, F. J. 94n Cicero, Marcus Tullius 29n, 65, 352, 353, 358, 364, 365, 367, 367n, 371 Cleon 223 Clinias 186 Cohen, D. 115 Cooper, J. M. 22n 24n, 27n, 28n, 41n, 48, 49n, 52, 54n, 58, 94n, 136n, 139n Cornford, F. M. 136n, 247, 248n, 258, 261n, 262, 263n, 264n, 266n, 272 Courtney, E. 166n Craig, L. H. 178n, 179 Creon 223 Crombie, I. M. 183n Cross, R. C. 180n, 216 Crubellier, M. 243n Cunningham, M. R. 105n Dale, A. T. 82n Damascius 272, 273 Dancy, R. 73n, 79n Davidson, J. 158n Dawson, D. 176n, 182n Deichgr¨aber, K. 340n, 354n Delp, M. H. 347n Democritus 21n, 27n, 261, 370n Denyer, N. C. 13n Descartes, Ren´e 375, 376, 384, 387 Deslauriers, M. 192n Deucalion 186n Dicaerchus 183n Diderot, Denis 44n Dido 171n, 174 Diller, H. 337 Dillon, J. M. 183n, 188n, 273n Dimas, P. 72n, 82n Diodotus 223 Diogenes La¨ertius 22n, 255n, 352, 353n, 356, 357–8, 377n Diomedes 164n, 168 Dixsaut, M. 73n, 79n Dover, K. 99, 100n, 101, 103n, 111n, 115n, 156n, 158n, 161n Duff, T. 47n Duke, E. A. 79n, 260n Dummett, M. 353n, 371n Ebert, T. 73n, 79n Ellis, H. 347 Empedocles 21n, 24n, 285 Epicurus 21n, 322 Eubulides 352

Euripides 261 Euryalus 161, 162, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174 Eurydice 169, 170n Eusebius 255n Evans, G. 209 Fagles, R. 98n, 128n Feeney, D. C. 172n Ferrari, G. 136n, 144n, 146n, 198, 222n Ferrill, A. 182n, 191n Fine, G. 73n, 81n, 88n, 232n, 233n, 235n, 239n Fitzgerald, R. 162n, 164, 165, 166n, 167n Flaubert, G. 107 Fleming, P. R. 347 Foucault, M. 22 Fowler, H. N. 73n, 94n Frankfurt, H. G. 106n Frede, D. 95n, 351n, 358n, 359n, 361n Frege, G. 205, 211, 214 Freud, S. 45n Furley, D. 280n, 281n, 295, 296 Galen 324, 325, 346n, 351, 351n, 353, 354, 355–6, 364, 367, 368, 369 Gallop, D. 73n, 94n Geach, P. 83 Gifford, M. 7n Gill, M. L. 284n Gorgias 254, 255 Gow, A. S. F. 84n Graff, D. 363n, 371n Graham, D. 280n, 281, 282n, 283n Gregory of Nyssa 375–7 Grensemann, H. 346 Griffith, T. 136n, 198, 222n Grmek, M. D. 335, 343, 345n Grube, G. M. A. 73n, 94n, 136n, 139n Guthrie, W. K. C. 183n, 187n, 189n Hackforth, R. 73n, 82n 94n, 245, 245n Hacking, I. 346, 347 Hadot, P. 21, 23n, 26, 29n, 31n Halbfass, W. 385 Hankinson, R. J. 359n, 361n Hanzawa, C. 342 Hardie, P. 161, 162, 163, 164n, 165, 167n, 168n, 171n Harlfinger, D. 233n Harte, V. 195n Hatzistavrou, A. 53 Heck, R. 353n Hector 337 Hegel, G. W. F. 179, 380, 389 Hera 127, 146, 152

Name Index Heracles 162 Heraclides Ponticus 23 Heraclitus 18, 21n Herodotus 23n, 348 Hesiod 131, 133, 183 Hicken, W. 34n, 79n, 260n Hill, K. 105n Hinds, S. 156n Hippias 33n, 35 Hippocrates 24n, 336–49 Hobbes, Thomas 117n, 182, 193 Hobbs, A. 180n Homer 98n, 102n, 128, 128n, 131, 133, 157, 161, 163, 164, 167, 168, 172, 188n, 261n, 270n, 337 Horgan, T. 363 Hunter, R. 49n, 156n, 160 Hutchinson, D. S. 94n Hylas 162 Irwin, T. 146n, 149, 150, 232n Isocrates 24n, 29n Jack, W. R. 347 Jaeger, W. 23n Jawetz, E. 347n Joachim, H. H. 240n Johansen, T. K. 245n Jones, D. 105n Jones, H. S. 192n Jones, W. H. S. 24n, 345n Jouanna, J. 346 Jowett, B. 73n, 94n 136 Judson, L. 239n, 242n, 243n Julius Caesar 29 Kahn, C. H. 22n, 259, 259n Kalligas, P. 122n Kang, S.-H. 116n Keller, S. 109 Kenney, E. J. 154n Kirk, G. S. 255n Klein, J. 261n Koller, S. 70n Kolodny, N. 109n Kosman, L. A. 11n Kraut, R. 148n Kripke, S. 209, 382 Krupp, M. A. 347n K¨uhn, C. G. 351n Laches 168 Lane, M. S. 44n, 48n, 49n, 186n Langlois, J. H. 105n Lear, G. R. 128n, 130n, 132n

409

Lear, J. 45n Lee, D. 136n, 177n Lee, K. J. 105n Leitao, D. 158, 160n Lennox, J. G. 278n, 284n Lennox, P. G. 166, 167n Leszl, W. 235n Liddell, H. G. 98, 99, 114n, 127, 192 Lissarague, F. 102, 103 Little, A. 105n Littr´e, E. 342 Lloyd, G. E. R. 53n, 337, 339 Locke, John 16n Long, A. A. 356n Lonie, I. M. 346 Lycaon 127 Lycurgus 125, 131, 133, 149 Lynch, J. P. 21n Lyne, R. O. A. M. 171n Lyons, J. J. 2n McCabe, M. M. 12n, 186n McPherran, M. L. 352n, 368n Major, R. H. 335, 347 Makowski, J. F. 163n, 171n Manning, R. Y. 347n May, K. A. 105n Medea 261 Megillus 186, 188 Melissus of Samos 254–5 Menn, S. 236n, 246n, 251n Mignucci, M. 362n Mills, E. 363n Mills, K. W. 78n Montaigne, Michel de 22, 44n Moravcsik, J. M. E. 258 Morison, B. 280n Moss, J. 130n Mynors, R. A. B. 164n Nagel, T. 11n Nails, D. 1n, 45n Nehamas, A. 2n, 21, 22, 28n, 44, 46, 48, 49, 58, 98n, 104n, 105n, 106n, 127n, 129n Nestor 62 Nettleship, R. L. 136n, 139n, 140n Nicoll, W. S. M. 79n, 260n Nietzsche, Friedrich 22, 129 Nightingale, A. W. 47 Nisus 161, 161, 164, 165, 166, 167, 168, 169, 170, 170n, 171, 171n, 172, 172n, 173, 174 Notomi, N. 254n, 255n, 256n, 257n, 262n, 264n, 269n, 271n Nussbaum, M. C. 112n, 146n

410

Name Index

O’Brien, D. 262, 263n, 264n, 266n O’Connell, R. L. 182n Odysseus 168 Olympiodorus 255n Opheltes 165 Origen 378, 381n Orpheus 159, 161, 169, 170, 170n, 171, 174 Owen, G. E. L. 82n, 198, 258, 259, 260n, 261n, 263 Owens, D. 382, 383n Pallas 171n Parfit, D. 377, 378, 386 Paris 127 Patroclus 161, 161n, 172n Pausanius 170n Peacocke, C. 211, 212 Penner, T. 72n, 210n Pericles 62, 128 Perrett, D. 105n Pindar 101, 101n Plantinga, A. 382 Plotinus 21, 122, 123n Polycrates 22n Polydamas 337 Porphyry 255n, 375, 387 Pretzl, O. 386 Price, A. W. 112n, 113n, 115n, 118n, 120n, 123n Prior, A. 382 Proclus 234n Protagoras 255, 256, 352, 370n Putnam, H. 203, 204, 205, 206, 208, 210 Pyrrha 186n Pyrrhus, King of Hellenistic Epirus 132 Pythagoras 21n, 23n Quintilian (Marcus Fabius Quintilianus) 33n Raven, J. E. 255n Rawson, E. 354n Reed, J. D. 163n Reeve, C. D. C. 124n, 136n, 139n, 170n, 197n Richards, I. A. 136n Rivaud, A. 248n Roberts, A. R. 105n Robinson, D. 79n, 260n, 261n Roggman, L. A. 105n Ross, W. D. 232n, 235n, 242 Rousseau, Jean-Jacques 44n Rowe, C. J. 48n, 60n, 62n, 72n, 73n, 75, 75n, 94n, 98, 118n, 121n, 145n, 156n, 159n, 160n, 186n Rudebusch, G. 214n Ryle, G. 16n

Salius 166 Saylor, C. 165n Schiffer, S. 363 Schmid, W. T. 13n Schofield, M. 79n, 180n, 181n, 182n, 183n, 255n, 352n, 353n, 356n, 357n, 358n, 359n, 360n, 361, 363, 364, 371n, 372n, 373n Scott, D. 61, 69n, 75 Scott, R. 98, 99, 114n, 127, 192n Sedley, D. N. 34n, 68n, 69n, 70n, 73n, 89n, 251n, 353n, 356n, 358n, 359n Segal, G. M. A. 205n Seneca the Younger (Lucius Annaeus Seneca) 29, 162n Sextus Empiricus 322–33, 352, 354, 359, 360, 362n, 364n, 377n Shay, J. 193n Sheffield, F. C. C. 47n, 150n Shorey, P. 4n, 136n Siegel, R. E. 335 Sihvola, J. 146n Simplicius 235, 286, 287 Slings, S. R. 197, 198, 199 Smith, N. 82n, 225n Smith, R. 307n, 319n Smith, W. D. 346 Solon 23n, 125, 131, 133, 149 Sorabji, R. 6n, 236n, 291n, 352n, 353n, 355n, 356n, 357n, 358n, 359n, 360n, 361, 363, 364, 371n, 375n, 377n, 378n, 380n, 381n, 382n, 383n, 385n, 386n, 387n, 388n Sorenson, R. 363 Statius, Publius Papinius, 172n Steel, C. 385 Stendhal, Marie-Henri Beyle 126n Stopper, M. R. 358n Strachan, C. 79n Sweet, N. J. 347n Taylor, A. E. 245n Taylor, C. 377, 378, 379 Thales 21n Theocritus 162 Theognis 99, 103, 128, 130 Theophrastus 235n Themistocles 223–4, 302, 307 Thersites 128 Thivel, A. 346 Thompson, W. H. 145n Thrasyllus 255n Thucydides 23n, 129n, 223, 348 Tredennick, H. 94n Turnus 170n, 171n, 173 Tye, M. 16n

Name Index Valimigli, M. 73n Vallance, J. 354n Velleman, D. J. 109n Verdenius, W. J. 79n Vernezze, P. 149, 150 Vicaire, P. 79n Vidal-Naquet, P. 183n Virgil 154–6, 161–74 Vitoria, Francisco de 381, 382 Vlastos, G. 38, 39n, 48n, 83n, 93n, 109, 114, 115, 151n, 249n Volcens 173 Wagner, T. 260n Walzer, R. 351n, 354 Wardy, R. 255n Waterfield, R. 11n, 122n, 124n, 129n, 136n Wedin, M. V. 82n Weidemann, H. 303n, 313n Weiss, R. 64 West, D. 161, 162, 166n, 167n Wheeler, A. 347 Whitaker, C. W. A, 307, 308, 312, 313, 315, 318, 319, 320, 321

411

White, N. 181n, 257n, 264n Wilderbing, J. 223n Williams, B. 126n, 133n, 377, 378, 379, 380, 381n, 383, 389 Williamson, T. 357n, 358n, 362, 363 Wilson, J. R. S. 228n Woodruff, P. 2, 48, 49n, 98, 129n Woods, M. 101n Woolf , R. 51n, 61n, 214n Woozley, A. D. 180n, 216 Wright, C. 353n Xenocrates 233–4, 235, 236, 237, 239, 244, 250, 251 Xenophon 21, 22n, 33n, 37, 38, 99, 343n Yunis, H. 27n Zalmoxis (priest-king of Thrace) 17–18 Zeus 159n

Index Locorum Apollonius of Rhodes Argonautica I 131–2: 162 II 317–40: 261n II 549–97: 261n II 564–5: 261n Aristophanes Ecclesiazusae 571: 24n Aristotle Constitution of Athens 16, 7: 183n De anima I 1 402a 7: 276 I 1 402b 13: 292 I 3: 289, 295, 296n I 3 405b 31–406a 12: 282n I 3 406b 10–11: 284 II 1: 277 II 1 412a 19–21: 277 II 1 412b 15–17: 277 II 2 413a 25–7: 277 II 4: 244n, 286n, 288, 289, 290, 292, 293, 294 II 4 415a 16–22: 285 II 4 415a 18: 292 II 4 415b 8–15: 285 II 4 415b 21–7: 285 II 4 416a 30: 286n II 4 416a 34– b 3: 286n, 287n, 293 II 4 416b 3: 287 II 4 416b 3–7: 287 II 4 416b 11–17: 287 II 4 416b 13: 287 II 4 416b 15: 288 II 4 416b 17: 294 II 4 416b 17–19: 286, 288 II 4 416b 17–20: 294 II 4 416b 19: 294 II 4 416b 25: 286 II 4 416b 25–9: 288n II 5: 286n, 290, 293, 296 II 5 416b 33–5: 290n II 5 417a 3–6: 12n II 5 417a 9–12: 292 II 5 417a 21: 290

II 5 417a 31–3: 291n II 5 417a 33: 291n II 5 417b 2–7: 291 II 5 417b 3: 294 II 5 417b 8–9: 293 II 5 417b 15: 291n II 5 417b 15–16: 291n II 5 417b 16–21: 292 II 5 417b 17–20: 294, 298 II 5 417b 18: 292n II 5 417b 19–21: 290 II 5 417b 20: 278n II 5 418a 3–6: 295 III 2: 6n III 2 425b 12–20: 2n De Caelo I 10–12: 236n I 10 279b 32–280a 2: 235n I 10 280a 28–33: 251n De Interpretatione 1: 309n 5: 315, 317, 318n, 320 5 17a 15–17: 315 6: 320 7: 304, 305–6 7 16a 9–11: 308 8: 304, 305, 305n, 306, 308, 309n, 312, 315, 316, 317, 318, 318n, 319, 320–1 8 16b 33–17a 3: 308 8 17a 16: 306n 8 18a 8–9: 304 8 18a 13–14: 304 8 18a 18: 303, 314 8 18a 18–19: 304, 313, 314 8 18a 18–26: 306 8 18a19: 303 8 18a 19–25 314 8 18a 21: 303n 8 18a 22–3: 307 8 18a 23: 305n 8 18a 25: 305n, 314 8 18a 25–6: 314, 315 8 18a 26–7: 304 8 18b 6: 305n 9: 304, 305–6 11: 314, 315, 316, 317, 318n, 320 11 20b 12–22: 316, 317n

Index Locorum 11 20b 15–16: 317n 11 20b 18–22: 314n 11 20b 22–26: 317n 12: 316 De Motu Animalium 702a 17: 288n De Sensu 4, 441b 27–424a 3: 288n Eudemian Ethics I 6 1216b 32–9: 243n II 6 1223a 9–14: 102n III 1 229a 15: 22n VIII 3 1243b 18–30: 102 VIII 3 1249a : 103 Generation of Animals II 1 734b 36–735a 4: 298n II 1 735a 2–4: 297 II 1 735a 12–25: 296 II 3 737a 18–20: 296n II 4 740b 29–34: 297n II 6 742a 19–742b 17: 297n II 6 743b 33–744b 11: 297n V 1 778b 1: 278n Historia Animalium I: 337 I 1 488b 24: 130n III 20 522a 1 f: 336 IX: 337 Metaphysics I 6 987b 1–4: 22 I 6 988a 7–14: 251n I 9 990b 8–11: 232 I 9 990b 13: 233 I 9 991b 3–9: 232, 242 I 9 992a 24–6: 241n I 10 993a 15–16: 241n II l 993b 23–31: 124n III 4 999b 17–20: 232 III 4 1000a 1–1001a 2: 238n IV 3 1005b 20–2: 4n VII 2 1028b 8–13: 235n VII 2 1028b 24–7: 235n VII 4–6: 242n VII 7–9 1032a 12–1034b 19: 241n, 242n, 243n VII 7 1032b 1–2: 242n VII 7 1032b 11–14: 252n VII 7 1032b 12–14: 242n

413

VII 8 1033b 19–1034a 5: 242n VII 8 1034a 3–5: 242n VII 17 1041b 28–30: 235 VIII 1 1042a 6–11: 235 VIII 3 1043b 18–23: 232n, 235n, 240n IX 7: 297 IX 7 1048b 37–1049a 18: 298n XII 3–4: 243n XII 3 1070a 13–30: 242n XII 3 1070a 14–19: 234 XII 3 1070a 16–17: 252n XII 3 1070a 18–23: 232n XII 3 1070b 14: 238n XII 7 1072b 14: 238n XIII 4 1078b 17–19: 22n XIII 4 1078b 27–31: 22n Nicomachean Ethics I 6 1096a 11–17: 240n I 7 1097b 11: 144n I 7 1098a 18–19: 354 I 8 1099a 13: 130n II 3: 99n III 1 1109b 31–2: 102n IV 4 1125b 12: 130n VI 1 1138b 25–6: 243n VI 1 1138b 32–4: 243n VI 3–4: 233 VI 13 1144b 3–4: 45n VI 13 1144b 8: 45n VII 2–3: 22n VII 5 1148b 19–25: 139n VIII 2: 99n X 9 1179b 9: 130n On Generation and Corruption I 3 318a 2–8: 279n I 5 321b 6–10: 280 I 7: 276n II 9 335b 18–20: 238n II 9, 335b 20–4: 240 II 9 335b 28–30: 287n On the Parts of Animals I 1: 297n I 1 641b 5–9: 278n On Breath 27 (21) 480a 17–12: 283n Peri Ide¯on 79.13: 233 79.22–80.7: 233

414

Index Locorum

Physics I 9, 191b 35–192a 24: 251n II 1 192b 13–15: 285n II 3: 297 II 3 194b 29–32: 278n II 4: 284n II 8 199a 20–30: 252n III 1: 290 III 1–3: 276n VIII: 283 VIII 1–2: 236n VIII 2 253a 8–21: 280n VIII 4: 295 VIII 4–6: 279 VIII 4 254b 30–1: 279 VIII 4 255a 28–30: 278 VIII 4 255a 30-b 13: 295 VIII 4 255b 17–31: 295n VIII 4 255b 22–3: 295n VIII 5: 283n, 289 VIII 5 256a 4–10: 281n VIII 5 256a 22–5: 281n VIII 5 257b 6–10: 279n VIII 6: 279, 281, 283, 284, 288n, 295, 296n VIII 6 259b 3–6: 279 VIII 6 259b 6–15 280 VIII 6 259b 9: 283n VIII 6 259b 12–13: 289n VIII 6 259b 15–20: 281 Poetics 1447b 11: 22n 1448b 9–12: 105 1451b 2–4: 347 Politics I 2 1253a 3: 144n V 4 1304a 18–29: 178n Posterior Analytics I 23 84b 7–8: 278n Prior Analytics II 27 70a 13 f: 336 II 27 70a 36 ff: 337 II 27 70b 1 ff: 336 II 27 70b 7 ff: 337 II 27 70b 16 f: 337 Rhetoric I 2 1357b 4 ff: 336 I 2 1357b 15: 337 I 2 1357b 18 ff: 337 III 5 1407a 33–9: 302n

Sophistici Elenchi 4 166a 4–5: 312n 4 166a 8: 312n 4 166a 12–14: 311, 311n 4 166a 14: 312n 4 166a 20–1: 312n 5 167b 1 ff: 337 17: 306n, 308, 310, 311, 316n 17 175a 39–41: 309n 17 175b 15–24: 312n 17 175b 39–176a 5: 301, 30, 306, 307 17 175b 39–176a 18: 310 17 175b 41–176a 1: 307, 317n 17 176a 2: 302 17 176a 6–14: 317n 17 176a 14–16: 311 17 176a 15: 302n 18 177a 6–8: 302n, 308 19: 310, 311n 19 177a 9–15: 302n, 310 19 177a 18–23: 311 19 177a 21–2: 310n 19 177a 23–4: 311n 22 178a 27–8: 309n Topics VIII 7: 310 VIII 7 160a 23–9: 308n, 311 VIII 7 160a 25: 308n VIII 7 160a 26: 310n Athenaeus Deipnosophists VI 267e–270a: 183n Chrysippus Logical Investigations SVF II 298a, p.106, 7–12: 353n SVF II 298a, p.106, 9–10: 362n Cicero, Marcus Tullius Academica II 49: 365 II 85: 360 II 91–8: 353, 364 II 92–3: 365 II 92–4: 363 II 95: 368 II 95–8: 371 II 96: 371 De Natura Deorum II 95: 367n III 44: 367n De Fato 12–15: 358

Index Locorum De Finibus Bonorum et Malorum IV 48–51: 362n IV 50: 365n IV 61: 29n Demosthenes 21. 219–24: 27n Diogenes La¨ertius Lives VII 40: 377n Euripides Medea 1–2: 261n 432–3: 261n 1263–4: 261n Galen Ars Medica I 379 K: 355n On Affected Parts VIII 25–30 K: 355n VIII 144 K: 355n On Mixtures 1676–7 K: 355n On the Diagnosis and Cure of the Errors of the Soul V 93 K: 355n On the Differences of Symptoms VII 50 K: 355n On the Order of My Own Books X1X 51–4 K: 355n Hippocrates Aphorisms IV 70, L IV 526 11 ff: 344 IV 75: 344 IV 75–81, L IV 530 4–532 4: 344 IV 76: 344 IV 77: 344 V 39, L IV 544 14 ff: 336n Epidemics I: 338, 339, 341, 344, 345 I 8, L II 42 4 ff: 339 642 8: 340 I 8, L II 644 8: 339 644 11 ff: 340 I 8, L II 646 9 ff: 339 650 1: 340

415

650 12 ff: 340 I 9, L II 652 4 f: 340–1, 341n 9, 656 7 ff: 339–40 I 9, L II 660 5: 340 I 9, L II 660 6 ff: 348 I 9, L II 660 6–668 12: 341n 664 11 f: 340 I 10, L II 668 14 ff: 338, 342 I 11, L II 676 12 ff: 338 I 12, L II 678 5 ff: 341 Case 4, L II 692 13 ff: 344 Case 5, L II 694 4 ff: 343 II 5 1, L V 128 1 ff: 337 II 6 1, L V 132 15 ff: 337 III: 338, 341, 343, 345n Case 8 of 1st series L III 56 4: 343 Case 11 of 1st series L III 62 8 f: 343 III 6, L III 82 1 f: 341 Case 2 of 2nd series L III 112 10: 343 Case 3 of 2nd series L III 114 1: 343 III 114 10: 343 Case 9 of 2nd series L III 128 2 ff: 343 Case 10 of 2nd series L III 130 4 ff: 343 Case 11 of 2nd series L III 134 2 ff: 343 V 71, L V 244 20 ff: 342 V 102, L V 258 6: 342 VI 8 26, L V 352 16 ff: 348 VII 56, L V 422 14 ff: 342 VII 56, L V 422 19 ff: 342 VII 82, L V 436 22 ff: 342 VII 89, L V 446 7 ff: 344 VII 112, L V 460 6 ff: 344, 345n On Ancient Medicine 21, CMG I 1 52 15 ff: 348n On Diseases I–III: 346 On Internal Affections: 346 On Joints 10, L IV 102 9 ff: 339 On Regimen I, CMG I 2 4, 136 6 ff: 338 III 70, CMG I 2 4, 202 11 ff: 348n On Regimen in Acute Diseases 1, L II 228 2 ff: 346 3, L II 238 8 ff: 338 3, L II 242 3 ff: 338 11, L II 314 12 ff: 348n

416

Index Locorum

On Regimen in Acute Diseases, Appendix 9, L II 436 5–442 6: 338 On the Nature of Man 14, CMG I 1 3, 202 6 ff: 344–5 Prognostic 1, L II 112 6 ff: 338 2, L II 112 12 ff: 339 7, L II 126 7 ff: 345n 12, L II 138 15 ff: 343 12, L II 140 6 ff: 344 12, L II 142 1: 343 12, L II 142 12 ff: 345 20, L II 168 6 ff: 341 25, L II 188 14 ff: 341 Homer Iliad II 106–9: 128n II 212–19: 128 II 611–12: 98n III 8: 179 IV 431: 179 XXIV 348: 162 Odyssey IV 712: 164n IX 105–566: 188n XII 59–72: 261n XII 73–126: 261n XII 86–7: 270n XII 222–62: 261n XII 428: 261n Isocrates Antidosis 261–9: 29n Lysias 1.34–6: 27n Pausanius Description of Greece X 26 1: 170n Plato Apology 19d: 24n 20c3: 49 21a: 27 21b: 49 21b ff: 35

21b–23b: 31 23a: 36 23a–b: 32 29b6–7: 41 29d: 27 29e1–3: 40 30a–b: 24n 30a8–b4: 40 30b: 103 37e–38a: 27 38a: 25 Charmides 154b: 100 154e: 12 155c: 100 155c–d: 13 156a9: 100n 156d1–2: 100n 156e: 18 156e6–157a1: 122 157a5: 100n 157d: 11 157e ff: 11 158c–d: 11 158e–159a: 12 159b–160d: 100 159c3–160b5: 101 160d–e: 12 160e1: 14 164c–d: 1 165b4: 1 165c7: 1 166c2–3: 2 166c–d: 8 166d1: 18 166e5–6: 2 167–9: 14 167a: 8 167b11–c2: 2 167c–168a: 2 167c4: 9n 168a3–4: 3 168a6–8: 3 168b2–3: 4 168b10–c2: 4n 168b–d: 2 168c4–7: 4n 168d–e: 9 168d–169a: 2 168d1–2: 5 168d3: 6 168d10: 6 168e3: 9n 168e4: 14

Index Locorum 168e5–6: 6 168e9–169a1: 6 169d–171c: 7n 169e: 17 173a ff: 13n Cratylus 389a ff: 234n 389b: 237n 389b5–6: 72 391d: 248n 429d: 256 439c7–d1: 74 Critias 109d–e: 186n 111a–b: 186n 111e3: 130n 112a: 186n 112c–d: 186n Crito 46b3–4: 39n 46c2–3: 39n 48d8–e1: 39n 49c–d: 39n Euthydemus 278e–281e: 128n 283a–e: 33n 283e–284c: 256 287d–288a: 33n 293b–e: 33n 295b–298a: 33n 300e–303a: 33n Euthyphro 5d–e: 121n 7c3–d8: 69 Gorgias 462a1–b3: 33n 463a–466a: 33n 466e: 33n 467c5–468e5: 81n 467e–468c: 110 481d: 114 500b6–d4: 39n 507e3–508a4: 139n Hippias Major 294a8–b4: 71 295c–e: 99n

Laches 181a–b: 178 190d7–191c5: 132 190e: 121n 192a1–b4: 69 192c–d: 100 Laws 628c–e: 177 643b–d: 186n 677–80: 186–8 681a: 187n 706b–d: 178n 707c: 178 707c: 179 713a–714b: 186, 188 721b–c: 247n 793e–794b: 186n 797a–798d: 186n 803d: 177 895a–986b: 93 Meno 72d4–e9: 69 80a: 34n 80a–b: 226 82a8–86c3: 69 Parmenides 129a8–b1: 84n 129b1: 83 129d5: 83 130b7–8: 74 130d: 238n 132a–133a: 90 132b3 ff: 251n 134b14: 73 150c6–e1: 71 151b5–e2: 70n Phaedo 65d4–5: 74 65d4–8: 69 65d4–e6: 69 65d12–13: 70 65d12–e11: 69 66c: 177 68c1–2: 64 69b1–3: 64 70e–71d: 93 73c5–d5: 75n 73c7–dl: 77 73c9: 77 74 ff: 15n

417

418 Plato (cont.) 74a10: 76 74a9–10: 73, 74 74a9–b3: 68 74a9–c6: 68 74a9–I2: 68 74b–c: 91n 74b1: 68 74b2: 68, 73, 74, 76 74b3: 68, 76 74b4: 75, 76 74b6–c6: 76 74b8: 79 74d6: 73 74b8–9: 80 74b10: 76 74c1: 82 74c1–2: 76 74c3: 76 74c4–5: 77, 82, 85 74c6: 77 74d4–75b9: 80 74d6: 73 75b1–2: 73 75b5–6: 73 75c7–d5: 75 75c9: 70 75d2: 72 76a6: 75 76b4–7: 75 76b4–12: 75 76d7–8: 74 76d8–9: 74 85c–d: 261n 92d9–e1: 72 95e: 87 96e–97b: 94 99c–d: 261n 100a: 87 100b: 87 100b1–101d3: 83 100b6: 74 100d: 89, 93 100d–e: 89 100e: 89 102b–c: 91 102b–103b: 92n 102b8–103a3: 82 102b–c: 91 102b8–d2: 71 102c1–2: 71 102d–e: 92, 114 103b: 92 103c: 93 103c–d: 94 104d: 93 105b–c: 93

Index Locorum 105c: 93, 94 106d: 92 106e: 95 106e–107a: 87 118a15–18: 44 Phaedrus 237b–241d: 142 238a1–c4: 142 241a4: 141 244a–257b: 142 244a3–8: 142 245c–e: 93 246a6: 142 248a: 32n 248d2–3: 144 248d2–e3: 145 248d3: 130n 248e2: 145n 249c4–e1: 144, 145, 146 249c8–d3: 147 249c9–d8: 146 249d3–5: 129n 251a1–252b1: 143 252a1–6: 143, 144 252a2–3: 54 252a3–4: 54n, 143 252a5: 143n 252c3–253c6: 144 252e3: 144 253a3–4: 147 253b1–2: 145 262b7–8: 73 265a6–11: 143 265a9–10: 141 272b2–5: 243n 278d: 32n Politicus 268d–e: 186n 270b–d: 186n 271–2: 185–6 271a: 186n 277b: 186n 297d–e: 145n 302e: 145n 306a–308b: 45n 308a: 177n Protagoras 309a–b: 114 331c4–dl: 39n 333c: 39n 344b ff: 32n 349e: 100 352b4: 145

Index Locorum Republic 331c–332a: 227 336b–337a: 33n 337a8–b4: 214n 346a4: 39n 360c: 151n 369–73: 184, 184n 369–75: 182 369b–372d: 180 372–5: 189 372b: 181 372b–c:184 372d: 181 373:184 373a–b: 181 373d: 179, 189, 190 373d–e: 189 373d10: 181 373e: 177, 178n 375: 184 375a–c: 53n 375e: 189n 375e9–376a5: 130 375e10: 189 376c–412a: 189n 376e2–4: 130 377b–391e: 53n 378b: 179n 378b–c: 177 379–399: 190n 389e: 179 395a4: 53 399: 183n 399e: 190 401a1–8: 122 402a3: 130 402c3–403a7: 141 402d1–4: 129, 130 402d10–e4: 129 403a10: 146 403a7: 141 403a8–b3: 142, 146 403d2–7: 122 409e5–410a4: 129 412b: 54n, 189n 412d–414a: 220 413b9–c3: 141 414d–e: 179, 190 415d–e: 179, 190 421–3: 179, 190 422–3: 179 424e–425a: 186n 429e–430c: 178 430a1–b5: 141 431a–b: 190, 192 432d8–433a4: 214 440b3: 130

442e–443a: 227n 442e4–443a10: 54n 457a: 178n 457b–462a: 116n 458c–d: 116n 460a: 179n 466e–467e: 178 467a: 178 468b: 178 468c: 178 473c–e: 189n, 191 473e–474a: 183n 474c–475e: 112 476–480: 2n 476a5–7: 132 476b4–d3: 83n 476d9: 74 476e4–479d9: 80 478c: 255 479b2: 76n 479d: 227 479d7: 111 484e4: 50 484e5: 50 485–7: 46 485a4–487a8: 45, 50 485a–487a: 53n, 136, 147 485b1: 50, 136 485c4: 51 485d3: 50 485d6–8, d10–el: 50 485d7: 61 485d12: 50, 152 485e: 50, 52 485e3: 59n 485e3–5: 147 486a8–10: 51 486a8–b2: 147n 486b1: 51 486b3–4: 51 486b6: 59n 486b6–8: 52 486c3–d3: 53n 486d4–10: 53n 487a7–8: 53 487c4–d5: 56 489e4–490a1: 56 489e4–490a3: 129n 490a5–6: 56 490b1: 136 490c: 221 490c9–10: 56n 490e–497a: 230n 491b8–9: 57 494b: 221 494c–d: 57 495a4–8: 128n

419

420 Plato (cont.) 495c3–4: 56 496a7–9: 56 496c: 46 4991b9–10: 57 500b–c: 149, 152 500b8–c1: 147 500c9–dl: 151n 500d4: 148 500d4–8: 149 501d2: 136 502c: 218 503a–e: 220 503b–3: 53n 503b3–4: 246n 503b6–7: 45n 503c1–308b: 45n 503c4: 53n 503c5: 53n 503c8: 53n 505d: 132n 505d6: 132 506a4: 132 507a–509c: 218 507a2: 72 507b7: 72 509d–510b: 91 509d–511e: 218 510d–511a: 220n 510d–e: 228n 511b–c: 220 511d–e: 217 514a: 218, 229 514a–516e: 222 514a–521b: 218 514a1–515a3: 195 514a1–517a7: 195 514a2: 213 514a4: 195 514c–515a: 91, 225 515a: 224n 515a2: 199 515a2–3: 199 515a5: 214 515a5–b1: 195 515a5–c3: 195 515a6: 195 515a6–7: 197 515a9: 199 515b2: 197 515b2–3: 195 515b4: 201 515b4–5: 197, 200, 201, 203 515b4–6: 195, 196, 199 515b5: 199, 200, 201 515b7–10: 199 515b8: 199

Index Locorum 515b9: 199 515c1–2: 199 515c1–3: 195 515c4–d2: 206 515d: 217n, 225 515d3: 207 515d4–6: 207 515d5: 199 515d5–7: 202 515d6: 202 515d6–7: 207 516b–c: 218n 516c: 223 516c–d: 223 516c8–d2: 201 517a–b: 217n 517a–c: 230 517b–c: 218n 517d: 225 517d–e: 222 517d4–e1: 196n 518b–d: 218 518b7–d7: 214 518c–d: 219, 231 518d: 218 518e: 221 518e2: 151n 519a: 220 519d: 222n 519d–521b: 217n 519d8–521b11: 149 519d8–9: 147 519e4: 148 520a8: 148, 151n 520a–c: 222n 520c: 225, 227 520c–d: 224, 224 520d8: 149n 520e2: 148 520e4–521a8: 148 521b7: 148 521c: 219, 229 521c–541b: 189n 521c–d: 218, 231 521d: 179, 220 521d–522b: 220 522b–531d: 219 523 ff: 12n 523–5: 15n 523b–525b: 230 524e–525a: 229 529a–c: 219n 532b: 189n 532b–c: 217, 219 532d–e: 220 533a10–b2: 73 533b–c: 220, 220n

Index Locorum 533–534b: 217n 534b–d: 220 535a: 53n, 54n 535a–536b: 53n 535a6–7: 53n 535a10–b1: 53n 535b5–c4: 53n 538c3: 121n 539c6: 146n 539e3: 148 540a: 189n 540b: 189 540b5: 148 540b5–6: 149 543a: 190 545c–562a: 137 546b: 190n 546d: 190n 547e–548a: 177, 189 549a9–b4: 137 549c8–550d7: 140n 550a4–b4: 137 554b7–e1: 137 559d7–561a1: 140n 561c6–d7: 137 562a–566d: 138 562b6–c5: 137n 566d–569c: 138 566e–567a: 177 568b: 151n 571–2: 192 571a–2b: 191 571a–573c: 138 571b: 191 571c3–d4: 139n 572d1: 137 572d2: 139 572d5: 140 572e4–573b4: 140 572e5: 136 573b6–7: 140 573c3: 140 573c3–5: 141 573c–576b: 138 576b–588a: 116n 580a3–4: 139 580d–581b: 192 580d–583a: 116 580d8: 147n 581a9–b3: 130 581b9: 116 586b6–7: 147n 587e: 116n 588c: 261n, 261n 590b6–9: 137n 596b ff: 234n 601b–602a: 128n

601d4–602a10: 132 602e4–5: 70 604b–c: 149 604b12–c1: 147 607a: 179 611b–d: 192 Sophist 227b: 266 236d–251a: 266 236d–264b: 257 236e3–237a1: 264n 237a3: 264n 237a3–4: 264 237a6: 264n 237a8: 264n 237a8–9: 264 237b: 263, 265 237b–241b: 257, 264n 237b4: 264n 237b–e: 265 237c7: 264 237d–e: 265 237e: 265 237e7: 264n 238a–239b: 265 238a7–9: 264 238b: 265 238b5: 264n 238c: 265, 266, 268, 272 238c–239a: 262 238d9–e2: 264 239a6: 264n 239d: 273n 241c: 267 241d–242a: 256, 247, 264n, 267 242b ff: 257, 267 243b: 265 243b–c: 257 244a: 267 244b–d: 264 245e–246a: 258 246e–247c: 267 248e: 267 249c–250d: 258, 267 249c–d: 262, 269 250a: 260 250a–b: 269 250c: 267 250d–e: 258, 265 250e–251a: 258 251a: 264 251a1–3: 260 251a–c: 268 252b–d: 268 252c: 266

421

422 Plato (cont.) 252d: 268 252d–253c: 268 253b–e: 269, 270 253e–254b: 255, 271 254c: 262 255a–b: 260 255c–d: 266n 256d–257a: 260 257b: 260, 272 257b6–7: 70 257b–c: 95 258e–259a: 272 259b–d: 270 259d–e: 269 260a: 269 261e–262d: 264n 264c: 271 265c: 243n, 249n Symposium 178d1–179b3: 156 178d6–7: 157 178e3: 158 178e3–179a2: 158 178e4: 159 178e5: 159 179a7–b3: 157 179b5: 159 179b8–c2: 157 179c1–2: 160 179d5: 170n 179d5–6: 161 181e5: 151n 183a: 123 189c–193d: 106 193a2 ff: 158 198d ff: 156 200a ff: 117 201a: 107 201e–202b: 95 202d: 107 203c: 106 204a1–b5: 111n 204d–205a: 131 204d3: 107 204e6–205a7: 107 205b7 ff: 117 206b5–8: 107 206a5–12: 107 206c2–3: 108 206e: 247n 206e5: 149, 107 206e8–207a4: 108 207a: 247n

Index Locorum 207a5–208b6: 108 207d: 247n 207d7–8: 108n 208c1–2: 150 208c3: 116 208c5: 116 208c5–6: 151 208d2–6: 108n 208d5: 116 208d8: 116 208e2–209a4: 149 208e2–3: 150 208e3–5: 116n 208e4: 131 208e5–209e4: 108, 150 209a ff: 115 209a1–2: 150 209a9: 120 209b4–7: 122n 209b8–e4: 108 209d–e: 133 209d1–e4: 150 209d3: 108n, 116 209d3–4: 133 209d4–7: 149 209e1–2: 125 209e4: 131 209e5–210a2: 111n 209e5–210a2: 150 210a4–9: 120 210a7–8: 113, 115 210a8–b2: 132 210a8–b3: 120 210b–c: 121 210b1: 120n 210b3–6: 109 210b5: 121 210b8–c1: 115n, 121n, 122n 210c3–6: 113n, 114 210c5–6: 114 210c6–d3: 123 210d1–3: 123 210e2 ff: 111 210e4–5: 123 210e5: 108n 210e5–6: 110, 112n 211a1–5: 111 211c2: 108n, 110, 110n 211d1–2: 124 211d1–3: 97n, 131 211e–212a: 124 211e4–212a7: 117 212a2–5: 125 212a3–6: 149 212a3–7: 131

Index Locorum 212a4: 132 212a4–5: 150 213b–d: 114 215a4–d1: 129n 215a5–222b7: 48, 60n 215a5–b6: 122 215d6–e4: 57 216a6–7: 57 216d2–3: 113 216d7: 47 216e2–5: 49 216e4–5: 48 216e7: 49 216e7–217a1: 47 217a4–5: 49n 218d–219d: 113n 218d6–7: 48n 219a–221c: 178 219c2–6: 59, 60 219d5: 59 219d6–7: 59 219e: 59n 219e8–220a1: 60 220a5–6: 60n 220a6–c1: 60 220c1: 59n 220c1–d4: 60 220d5–6: 60n 221a8–b1: 62n 221b3–4: 62n 221b6: 60 221c4–5: 62 221d7–222a6: 122 Theaetetus 142b: 178 148e: 34n 150c4–7: 34 151e: 35n 152a–186e: 35n 155a2–5: 70n 167a–b: 256 173d: 223 174c: 223 183c–184b: 256 188c–189b: 256 190b2–d2: 79 Timaeus 19b–c: 174 22–3: 186n 22a–23a: 186n 23c–d: 177n 25c–d: 186n 28a: 236n, 237n 29a: 236n, 237n

29c3: 217n 30b: 236n 41a–c: 246n 41a–d: 246n 41d: 246n 41d1–3: 246n 41e–42e: 248n 42a ff: 248n 42d–e: 239n 42e: 236n 42e–43a: 246n 44a8–b1: 248n 48a: 246n 50b–53b: 251n 51b8: 74 52–3: 237n 69c–71e: 245n 69c–d: 192n 70d–e: 246n 70e: 245n 73c–d: 248n 76d–e: 248n 77a–c: 245n, 246n 79a–e: 245n 80d–81e: 245n 91a: 248n 91a2–3: 246n 91a3: 246n 91b2: 246n 91b6: 249n 91c–d: 247 91d–92c: 248n 92a3: 248n 92b3: 248n 92b5: 248n Plotinus Enneads I. 6. 7. 29–30: 122n Plutarch On Common Conceptions 1035A–D: 377n 1059d–e: 370n 1084c–d,—SVF II 665: 362n Alcibiades I 3: 162n Quintilian Institutio Oratoria II 15.26: 33n

Seneca the Younger De Beneficiis 5.9: 367n 2.1–3: 29n

423

424 De Constantia Sapientis 2.1–3: 29n 7.1: 29n De Providentia 2.7: 29n Sextus Empiricus Against the Mathematicians I 68–71: 364n VII 22–3: 377n VII 415–21: 356n VII 416: 360, 363 VII 416–21: 359 VII 418: 356n VII 419: 371n VIII 231: 367 VIII 465–80: 352 VIII 480: 368 VIII 481: 368 IX 182–90: 357n Outlines of Pyrrhonism I 1–4: 322 I 2: 325 I 3: 327, 328 I 7: 327, 328 I 10: 329 I 12: 328, 333 I 19: 334 I 29: 328 I 169: 333 I 190: 329 I 196: 329, 333 I 205: 327 I 206: 368 II: 331–2 II 81: 332 II 104–33: 358n II 145–51: 365n, 366n II 185–92: 352 II 188: 368 II 253: 363 III: 331–2, 333 III 27: 332 III 29: 332 Sophocles Antigone 71–2: 98n Statius Thebaid X 441: 172n

Index Locorum 445–8: 172n Theocritus XIII 6: 162 XIII 8–9: 162 XIII 10: 162 XIII 70–3: 162 Theognis Elegies 2. 1377–9: 128n Theophrastus Metaphysics III 6a 23-b 9: 235 Thucydides History of the Peloponnesian War I 138: 223 III 37–48: 223 III 38: 223 Virgil Aeneid IV 677–9: 171n Aeneid V 331: 166 344: 161 355–6: 166 358: 166 Aeneid IX 169–70: 162 176: 161 177: 174n 179: 161, 165 179–80: 162 181: 161 182: 163, 165 183: 163 184–5: 164, 165 194–5: 165 197–8: 165 199–204: 165 205: 165 208: 166 217–18: 174n 223: 161 240–3: 166, 167 257: 168

Index Locorum 260–2: 168 263–72: 165 263–74: 167 275–6: 173n 275–80: 167, 173n 284: 165 285–6: 174n 297–8: 174n 300: 174n 303–7: 165 355: 165 378: 165 389: 170n 428–9: 170 444–5: 170 446–9: 171 465: 173n 590–663: 174 Aeneid X 147–50: 172 160–2: 171n 433–6: 171n 507–9: 171n 791–3: 171n

Aeneid XII 880–1: 171n Georgics I 37: 164n IV 490–1: 170n IV 525–7: 170n Xenophon Anabasis IV 8.26: 98n Memorabilia I 1.16: 24n I 2.9 22n, 36 33n I 5: 33n I 7: 33n II 4: 33n III 8: 33n III 9: 33n Symposium V 2–8: 99n

425