Oxford Studies in Ancient Philosophy: Volume XXIX: Winter 2005

Created on 4 July 2005 at 14.33 hours page iii

OXFORD STUDIES I N A NC I E NT PHI LOSOPHY EDITOR: DAVID SEDLEY

VOLUME XXI X winter 2005




Created on 4 July 2005 at 14.33 hours page i

OXF ORD STUDIE S IN ANCIE NT PHILOSOPHY

Created on 4 July 2005 at 14.33 hours page ii

page ii is blank

Created on 4 July 2005 at 14.33 hours page iv




Great Clarendon Street, Oxford ox2 6dp Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide in Oxford New York Auckland Cape Town Dar es Salaam Hong Kong Karachi Kuala Lumpur Madrid Melbourne Mexico City Nairobi New Delhi Shanghai Taipei Toronto With o¶ces 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 ã Except where otherwise stated, Oxford University Press, 2005 The moral rights of the authors have been asserted Database right Oxford University Press (maker) First published 2005 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 Oxford studies in ancient philosophy.— Vol. xxix (2005).—Oxford: Clarendon Press; New York: Oxford University Press, 1983– v.; 22 cm. Annual. 1. Philosophy, Ancient—Periodicals. B1.O9 180.{5—dc.19 84–645022 AACR 2 MARC-S Typeset by John Wa‹s, Oxford Printed in Great Britain on acid-free paper by Biddles Ltd, King’s Lynn, Norfolk ISBN 0–19–928744–9 ISBN 0–19–928745–7 (Pbk.)

978–0–19–928744–4 978–0–19–928745–1 (Pbk.)

1 3 5 7 9 10 8 6 4 2

Created on 4 July 2005 at 14.33 hours page v

ADVI S ORY BOARD Professor Jonathan Barnes, Universit‹e de Paris-Sorbonne, Paris IV Professor Michael Frede, University of Athens Professor A. A. Long, University of California, Berkeley Professor Martha Nussbaum, University of Chicago Professor Richard Sorabji, King’s College, University of London, and Wolfson College, Oxford Professor Gisela Striker, Harvard University Contributions and books for review should be sent to the Editor, Professor D. N. Sedley, Christ’s College, Cambridge, cb2 3bu, UK. He can be contacted by e-mail on dns1Äcam.ac.uk. Contributors are asked to observe the ‘Notes for Contributors to Oxford Studies in Ancient Philosophy’, printed at the end of this volume. Up-to-date contact details, the latest version of Notes to Contributors, and publication schedules can be checked on the Oxford Studies in Ancient Philosophy website: www.oup.co.uk/philosophy/series/osap

Created on 4 July 2005 at 14.33 hours page vi

page vi is blank

Created on 4 July 2005 at 14.33 hours page vii

CONTENTS Empedocles on Colour and Colour Vision

1

KA TE R INA IE R ODIA KONOU

Xenophon at his Most Socratic (Memorabilia 4. 2)

39

DA V ID M . JOHNSON

Socrates’ Deliberative Authoritarianism

75

A NTONY HA TZ ISTA V R OU

The Unity of Virtue in Plato’s Protagoras

115

B E R ND M A NUW A L D

Shame, Pleasure, and the Divided Soul

137

JE SSIC A M OSS

Animadversions on Burnyeat’s Theaetetus: On the Logic of the Exquisite Argument

171

M IC HA E L V . W E DIN

Authenticating Aristotle’s Protrepticus D. S. HUTC HINSON & M ONTE R A NSOM E

193 JOHNSON

Aristotle on Why Plants Cannot Perceive

295

DA M IA N M UR PHY

Aristotelian Teleology

341

L INDSA Y JUDSON

On the Use of Stoicheion in the Sense of ‘Element’

367

TIM OTHY J. C R OW L E Y

Index Locorum

395

Created on 4 July 2005 at 14.33 hours page viii

page viii is blank

Created on 4 July 2005 at 14.33 hours page 1

EMP EDOCLES ON COLOUR AND COLOUR VIS ION KATERINA IERODIAKONOU

    is among the first ancient philosophers, if not the very first, not only to use colour terms as adjectives characterizing the way things present themselves to us, such as ‘white’ of white milk (31 B 33 DK: γλα λευκν), but also to refer to colours as items whose explanation forms a crucial part of a proper explanation of the world of experience. One basic fact which one cannot overlook about the objects of experience is that they are coloured, and another crucial fact is that they conspicuously di·er from each other in particular in the way they are coloured. The surviving textual evidence about earlier philosophers who seem to have concerned themselves with colours is unfortunately so meagre that it proves extremely di¶cult to get a clear sense of what they think about this topic. But since Empedocles in many instances already seems to have relied on earlier thought about the matter, we shall occasionally have the chance, while trying to reconstruct his colour theory, also to comment on the little information available about the first recorded views on colour. The main source for Empedocles’ own view is, of course, his poem On Nature (Περ φσεως), which probably ran to about two thousand lines of hexameter verse, and of which approximately one-sixth has survived in fragmentary form. To begin our enquiry into Empedocles’ account of colour, we should first examine the context in which he talks about colours, and the reasons why he ã Katerina Ierodiakonou 2005 I would like to thank Istv‹an Bodn‹ar, Myles Burnyeat, Michael Frede, Geo·rey Lloyd, Maria Michela Sassi, and David Sedley for their helpful comments on an earlier version of this paper. I would also like to thank an anonymous referee for advice on art-historical aspects. I wrote this paper during my stay at the Institute for Advanced Study in Princeton (September–December 2003), where I greatly benefited from the warm encouragement of my colleagues and their stimulating conversation.

Created on 4 July 2005 at 14.33 hours page 2

2

Katerina Ierodiakonou

raises this topic. Like most Presocratic philosophers, he is interested in presenting a theory about the generation of the world as we know it, its constituents and its transformations. That is to say, he wants to identify the elements it is made up of, and to understand it as it presents itself to us, in terms of what happens to these elements in the course of a cosmic process. He thus constructs an elegant cosmological system founded on a small number of uncreated and indestructible basic elements, the four ‘roots’ (ιζµατα), namely earth, water, air, and fire (e.g. A 37, B 6), which are united and separated during di·erent stages of the cosmic cycle by two personified motive forces, Love and Strife (e.g. B 17 and 35). It is precisely in terms of these that Empedocles attempts to explain the processes by which the things surrounding us, both animate and inanimate, were at some point created and came to acquire their present form as the result of the combination of some or all of the four basic elements, though in di·erent proportions (e.g. B 21 and 98). And since all objects, animate and inanimate, are characterized by the fact that they are coloured, Empedocles also undertakes to explain what it is that makes them coloured and, moreover, how it happens that they have the di·erent particular colours they have. To put the matter in another way, it is a crucial part of his overall project of understanding the cosmos and its evolution that Empedocles should also deal in particular with the physical question concerning the nature of colour, in general, and the production of the various colours. Parmenides had claimed that it just appears that there are things which are coloured and which even change colour, and that if people believe that there are things which are coloured, this is another example of how people foolishly confuse appearance with reality (28 B 8. 41 DK). What Parmenides had failed to explain was how objects could appear to be coloured if in reality there is no colour. Against Parmenides, Empedocles took it upon himself to explain both how it comes about that objects in the world actually are coloured and how, in consequence, it is that they appear to us as coloured. Thus, very much in the same spirit of grasping the apparent workings of the universe in detail, Empedocles also sets out to understand how human beings manage to perceive the coloured things around them. For he considers human beings in their present form to be the result of survival of the fittest, a survival partly due to their being equipped with sense-organs (e.g. A 72, B 61). That is to

Created on 4 July 2005 at 14.33 hours page 3

Empedocles on Colour and Colour Vision

3

say, on Empedocles’ view human beings are creatures constructed in a way which enables them to perceive the world, understand it, and thus fit into it. The human sense-organs, when properly used, are supposed to fulfil exactly this task, namely to obtain reliable information about our world that we need for our survival (e.g. B 2 and 3). But Empedocles is more interested in grasping the physiology of perception as a biological function which gives us the ability to be aware of and distinguish the things around us than in asking the epistemological question whether the features we perceive objects to have are those that actually characterize reality. And since sight is essentially concerned with how objects appear to us, he tries in this context to figure out how by means of sight we perceive the colour of objects. Thus the reason why Empedocles discusses issues related to the nature of colour and to colour perception can be traced to his preoccupation with the physical question of the constitution of things, as well as with the question of how human beings, in particular, are biologically equipped to function and survive in this world. He pursues these questions against the background of Parmenides’ claim that reality is not the way we perceive it to be, that, for instance, things are not coloured and do not change colour in reality. But to present Empedocles as having di·erent interests from those sketched above, and to rephrase his questions in the light of more recent developments as questions about the metaphysical status of colour and about the epistemology of colour perception, would, I think, be grossly anachronistic. Let us then turn to what Empedocles actually has to say concerning the two distinct questions of why or how things are coloured and how we perceive their colour. The surviving Empedoclean fragments and ancient testimonies relevant to the issue of the nature of colour are frustratingly few in number. Empedocles’ central claim on this topic is presented in the following fragment, in which it is clearly stated that things are coloured because they originate from the combination of the four basic elements (B 71):1 1 The texts cited here are from H. Diels and W. Kranz, Die Fragmente der Vorsokratiker, 6th edn. [DK] (3 vols.; Berlin, 1951), unless otherwise stated. As for the translations, they very much depend on those by B. Inwood (The Poem of Empedocles: A Text and Translation and Introduction, rev. edn. (Toronto, 2001)) and by M. R. Wright (Empedocles: The Extant Fragments, Edited with Introduction, Com-

Created on 4 July 2005 at 14.33 hours page 4

4

Katerina Ierodiakonou And if, concerning these things, your conviction is in any way wanting, as to how from the combining of water, earth, aether, and sun the forms and colours of mortal things might come to be, which have now come to be, fitted together [συναρµοσθντα] by Aphrodite . . .

Note the prominent position given to forms and colours here; what need to be explained, in the first instance, are the shapes or forms and the colours of the objects we encounter in the world. To understand better how the combination of the four elements results in the generation of all the colours of this world, another much-quoted Empedoclean fragment should be taken into consideration (B 23): As when painters adorn votive o·erings, men well taught [ε δεδατε] by wisdom in their art, and so when they take in their hands pigments of various colours [πολχροα φρµακα], mixing them in harmony [ρµον !η µε ξαντε], more of some, less of others, from them they prepare forms resembling all things, making [κτ ζοντε] trees and men and women and beasts and birds and water-nourished fish and long-lived gods, first in their prerogatives. In this way let not deception overcome your mind that there is any other source for the countless mortal things that are seen, but know these things clearly, having heard the story from a god.

The comparison here is between what painters actually do when they mix their pigments of various colours in order to paint all the di·erent things in the world, and what happens when the struggle between Love and Strife results in the generation of everything simply out of the combination of the four basic elements. Just as the painters manage to represent everything in the world by using pigments of various colours, Love and Strife bring it about that everything in this world arises out of these four elements.2 This fragment, however, does not, and is not meant to, give us mentary, Concordance [Extant Fragments] (New Haven, 1981; repr. 1995)), with a few changes. 2 Note here the use of the duals ε δεδατε, µε ξαντε, and κτ ζοντε; in order for the comparison to work better, Empedocles introduces two painters whose art can be compared to the workings of Love and Strife. However, it is unclear how far the parallelism extends; if Love and Strife, each in its own way, make things arise out of the combination of the four elements, in what sense do two painters, each in his own way, bring about a representation of the world by using various colours? Cf. S.

Created on 4 July 2005 at 14.33 hours page 5

Empedocles on Colour and Colour Vision

5

direct information about how all the colours originated from the combination of the four elements; it does not, for instance, say that the mixture of a few colours produces all the colours in the world. It rather points out that just by using pigments of various colours, painters can represent all the diverse objects encountered in the world. Nevertheless, we shall see that the comparison proves to be instructive in clarifying how all the things in this world came to be from the combination of the four elements, and in this way also instructive in clarifying how all the colours originated. For the comparison to work, this fragment must refer to the way painters mix their pigments (φρµακα);3 ‘mix’ not in the sense of completely blending pigments of various colours in order to produce new hues, but in the sense of arranging pigments of various colours side by side in order to portray the world realistically.4 For it seems that this was exactly the practice followed by the painters of the fifth century bc. They drew an outline, filled it in with colour, and then on top of that colour juxtaposed washes of different colours; and if they wanted to produce a di·erent shade of colour, what they did was to superpose a layer of colour on top of another, rather than blending two colours in advance.5 Hence, πολχροα φρµακα in the Empedoclean fragment does not mean ‘many-coloured pigments’, i.e. pigments produced by mixing many colours; they are simply pigments of various colours.6 The use of the adjective πολχροα here leaves it completely open how many colours of pigments Empedocles has in mind, though the comparison strongly suggests that he must be talking of a rather limited Tr‹epanier, ‘Empedocles on the Ultimate Symmetry of the World’, Oxford Studies in Ancient Philosophy, 24 (2003), 1–57 at 35–6. 3 The term φρµακα refers in this context to the pigments used for painting; cf. also Hdt. 1. 98; Ar. Eccl. 735; Plato, Rep. 420 c; Crat. 434 b 1; Pol. 277 c. 4 Wright, Extant Fragments, 38 and 180; A. P. D. Mourelatos, ‘Quality, Structure, and Emergence in Later Pre-Socratic Philosophy’ [‘Quality’], in J. J. Cleary (ed.), Proceedings of the Boston Area Colloquium in Ancient Philosophy, 2 (Lanham, Md., 1987), 127–94 at 141 n. 15. 5 According to Plutarch (De glor. Ath. 2), Apollodorus was the first painter to use mixed colours (φθορ) at the end of the 5th cent. On the painters’ use of the term φθορ as equivalent to µε$ξις, i.e. as meaning ‘corruption’ in the sense of a colour losing its distinctive character by being mixed with other colours, cf. Porph. De abstin. 4. 20. 28; Plut. Mor. 393 c and 436 b. On the subsequent widespread use of mixed colours by ancient painters, cf. D.H. De Isaeo 4. 6 Similarly, when Aristotle uses the adjective πολχροα for the eyes of human beings (HA 492A5; GA 779B9), he is simply referring to the fact that the eyes of human beings are of di·erent colours.

Created on 4 July 2005 at 14.33 hours page 6

6

Katerina Ierodiakonou

number of kinds of pigments. It could be four, like the number of the basic elements, but it could also be more or fewer. It cannot be a large number, however, for the reference to the colours used by ancient painters is supposed to help us understand how just four elements can explain the seemingly endless variety of objects in this world. Therefore, if we take seriously the comparison thus understood against the background of the actual practice of painting in Empedocles’ time, it suggests that, when some or all of the four elements combine and produce something new, they do not fuse together in some kind of complete, genuine mixture in which the di·erent elements can no longer be distinguished. Rather, it seems that the di·erent elements are arranged side by side, as a result not losing their identity, just as the colours in an ancient painting are placed side by side or on top of each other. That this, most likely, is the way Empedocles also understands the combination of elements is confirmed by the fact that in other fragments he uses for the same purpose the verbs συναρµζεσθαι (B 71. 4) and ρµζεσθαι (B 107. 1), both of which standardly mean to ‘fit’ and ‘join things together’, rather than to blend di·erent things in, as it were, a chemical mixture. These verbs, for instance, are used to refer to the construction of a boat or a wall, while in medicine they are used to describe how the di·erent parts of the human body are joined together.7 And elsewhere, Empedocles talks of gluing the elements in harmony (B 96. 4: ρµον ης κλλ!ησιν %ρηρτα; cf. B 34), a metaphor which again refers to the juxtaposition of elements, rather than to their complete mixture. Besides, such an account of Empedocles’ theory of the combination of elements is strongly supported by the way the ancient philosophers comment on the subject. For instance, Aristotle says that, according to Empedocles, when the elements are combined, they are set beside one another, just as bricks and stones are placed when building a wall (GC 334A26–31 = A 43). And Galen compares Empedocles’ understanding of the combination of elements to a powder composed of di·erent metals, finely ground and not completely mixed with each other (In Hipp. De nat. hom. xv. 32 Kuhn • = CMG 5. 9. 1. 19. 7–12 = A 34). Hence, since the four elements are said to be made up of minute discrete, but in principle 7 e.g. Hom. Od. 5. 248 and 361; Hdt. 2. 96; Eur. Hel. 233; Tro. 111; D.S. 2. 8. 2; Hipp. O·. 25; Oss. 12.

Created on 4 July 2005 at 14.33 hours page 7

Empedocles on Colour and Colour Vision

7

divisible, particles,8 Empedocles seems to hold that everything in this world is a result of the aggregation of imperceptibly minute ingredients which are set side by side without thereby changing their nature. Furthermore, in the surviving fragments and testimonies there are certain examples of combinations which suggest that, when Empedocles uses the verbs ρµζεσθαι and συναρµζεσθαι, or when he uses the phrase ρµον ης κλλ!ησιν %ρηρτα, the ‘harmony’ (literally ‘fitting’) he talks about is not simply the fitting together of the elements; it refers rather to a combination in accordance with an appropriate mathematical ratio, so that the thing generated is stable and can survive. For instance, blood and flesh are said to have all the elements in equal proportions (B 98; A•etius 5. 22. 1 = A 78; Theophr. De sens. 10 = A 86), bones seem to consist of four parts of fire, two of earth, and two of water (B 96), while sinews originate from fire and earth mixed with double the amount of water (A•etius 5. 22. 1 = A 78); even our ability to perceive and to think and our special talents crucially depend on the proportion of the elements from whose combination each person is created (Theophr. De sens. 11 = A 86). In this regard Empedocles seems to be influenced by Pythagorean views on mathematical ratios and harmony as the principle of the order both in particular things and in the whole universe.9 Still, it is not clear whether on his view the principle of specific ratios can be generalized to all cases of combinations of elements. That is to say, there are doubts whether Empedocles really thinks that all combinations of elements can be expressed in such ratios, or whether this principle should be confined either to organic compounds or even just to the specific examples mentioned above.10 Besides, in the comparison with painting in B 23, the harmonious juxtaposition (ρµον !η µε ξαντε) of colours does not necessarily imply a combination of colours in specific mathematical ratios. And it is interesting to note here that the noun ρµογ&, when used as a technical term of Greek painting, simply means the 8 e.g. Galen, In Hipp. De nat. hom. xv. 49 Kuhn • = CMG 5. 9. 1. 27. 22–7 = A 34; xv. 49 Kuhn • = CGM 5. 9. 1. 27. 24 = A 43; A•etius 1. 13. 1 and 1. 17. 3 = A 43. 9 On the Pythagorean influence on Empedocles in connection with mathematical ratios and harmony, cf. Porph. V. Pyth. 30; Simpl. In DA 68. 5–8 Hayduck; Philop. In DA 176. 32–177. 4 Hayduck; Sophon. In DA 32. 18–32 Hayduck. 10 E. Bignone, Empedocle: studio critico (Turin, 1916), 417 and 469; F. Solmsen, ‘Tissues and the Soul: Philosophical Contributions to Physiology’, Philosophical Review, 59 (1950), 435–68 at 436–41; Mourelatos, ‘Quality’, 167–71.

Created on 4 July 2005 at 14.33 hours page 8

8

Katerina Ierodiakonou

juxtaposition of, or transition from, one colour to another (cf. Pliny, Nat. 35. 29). But however this may be, what still remains to be explained is how the combination of the four elements, as this is presented in the comparison with painting as well as in the other relevant texts, throws light on the way in which all the colours in this world are produced. We still need to explain what exactly happens, according to Empedocles, when the elements are harmoniously fitted together so as to generate not just something, but something with a specific colour. The Empedoclean fragment about the mathematical ratio which explains how bones are generated is particularly illuminating; for in this case the mathematical ratio seems also to explain why bones have the colour they have, namely why bones are white (B 96):11 And the kindly earth into her broad hollows received two parts of gleaming Nestis out of the eight and four of Hephaistos; and they came to be white bones marvellously fitted together with the glues of harmony [ρµον ης κλλ!ησιν %ρηρτα].

Empedocles claims in this fragment that bones are generated from the harmonious combination of earth, fire and water, more specifically from the combination of four parts of fire, two of earth, and two of water.12 And presumably the underlying assumption is that fire is white, and that it is the excess of fire here that gives 11 For discussion of the alternative readings of this fragment, cf. D. Sider, ‘Empedocles B 96 (462 Bollack) and the Poetry of Adhesion’, Mnemosyne, 37 (1984), 14–24. G. E. R. Lloyd, Polarity and Analogy [Polarity] (Cambridge, 1966), 274 n. 3, suggests that the second sentence may refer to the harmonious arrangement of bones in the body, rather than to the generation of bones from the harmonious combination of elements. 12 These proportions are found in A•etius (5. 22. 1 = A 78). The ancient commentators (Simpl. In DA 68. 10–14 Hayduck; Philop. In DA 176. 30–2 Hayduck; Sophon. In DA 32. 18–23 Hayduck), on the other hand, interpret the phrase Ν&στιδος α(γλης as referring both to water and air, presumably following Theophrastus, who says that bones and hair are composed of all elements (De sens. 23 = A 86); and thus they all talk of four parts of fire, two of earth, one of water, and one of air. I agree with C. E. Millerd (On the Interpretation of Empedocles [Interpretation] (Chicago, 1908), 41) and J. Longrigg (‘The “Roots of All Things” ’ [‘Roots’], Isis, 67 (1976), 420–38 at 433) that there is no reason to suppose that all four elements need be constituents of everything. For the concept of everything in everything seems to be distinctive of Anaxagoras; and moreover, Empedocles himself does not mention air in the composition of sinews (A•etius 5. 22. 1 = A 78), while he seems to believe that fruit is composed of only water and fire (A•etius 5. 26. 4 = A 70).

Created on 4 July 2005 at 14.33 hours page 9

Empedocles on Colour and Colour Vision

9

bones their white colour; this is, at least, what the ancient commentators report (Simpl. In DA 68. 10–12 Hayduck; Philop. In DA 178. 6–8 Hayduck; Sophon. In DA 32. 19–20 Hayduck). Could we infer, though, just on the basis of the particular case, that this is Empedocles’ account of how everything acquires its colour? Does Empedocles really want to claim that something has the colour it has only because elements have a colour and one of its constituent elements exceeds the others in quantity, thus accounting for the colour of the object? Since everything, according to Empedocles, is generated from the harmonious combination of some or all of the elements, which are juxtaposed without losing their features, it is reasonable to suggest that the colour of something also depends on the combination of elements, and in particular on the combination of the colours of its constituent elements; and this combination should again be understood not as a complete mixture of the colours of the elements, but as a juxtaposition of those colours. However, if we suppose that the colour of an object depends only on the colour of the element which predominates, Empedocles would not be able to explain the colour of all things, but only of those which have exactly the same colour as one of the four elements, provided that the four elements all have di·erent colours. Besides, what happens when no element prevails, but they are all combined in equal proportion, as for instance in the case of blood and flesh, which obviously do have a colour? First, though, we have to examine the assumption that all four elements have a colour. And if they do, are they di·erent in colour? And if they are di·erent, precisely which are their colours? The fragment about the creation of bones (B 96) suggests that according to Empedocles the colour of fire is white. This is also confirmed by another Empedoclean fragment in which the sun, i.e. fire, is presented as white (or bright) and hot, while rain, i.e. water, is presented as dark and cold (B 21. 1–6): But come! Gaze on this witness to my previous words, if anything was in my previous [remarks] left wanting in form: sun, white [λευκν]13 to look on and hot in every respect, 13 λευκν might be better translated ‘bright’ rather than ‘white’. On the general issue about whether the Greek colour terms should be understood as denoting the qualitative or the quantitative di·erence between colours, i.e. whether the Greek colour terms have hues or luminosities as their primary connotation, cf. M. Platnauer, ‘Greek Colour-Perception’, Classical Quarterly, 15 (1921), 153–62; H. Os-

Created on 4 July 2005 at 14.33 hours page 10

10

Katerina Ierodiakonou heavenly bodies [)µβροτα]14 bathed in heat and shining light, rain everywhere dark and cold; and from earth issue firmly rooted solids.

It is important to note that in this fragment, although presumably all four elements are presented, only fire and water are characterized by their colours. Theophrastus, too, reports that on Empedocles’ view the colour of fire is by nature white (λευκν) and the colour of water is by nature black (µλαν), without saying anything about the colour of the other elements, namely earth and air (De sens. 59 = A 69a; cf. De sens. 7 = A 86). Moreover, when Theophrastus discusses Democritus’ theory of four primary colours, he explicitly contrasts it with the dominant view of the other philosophers of the period, who treated white and black as the only simple colours (De sens. 79 = 68 A 135 DK): First of all, his [i.e. Democritus’] increase of the number of primaries [%ρχς] is puzzling; for the other philosophers propose white and black as the only simple [πλν] colours.

But this remark would make no sense if Empedocles, with whose views Theophrastus is familiar, had already assumed four primary colours or even the very four primary colours Democritus came to postulate. So, even if the terms used here for primary and simple colours (%ρχς/πλν) are not Empedocles’ own, Theophrastus’ testimony, together with the surviving fragments, suggests that Empedocles talked only of the colour of fire and the colour of water, namely the colours white and black, as the basic colours. On the other hand, there is a text in A•etius which ascribes to Empedocles four colours corresponding to the four elements, namely white, black, red (+ρυθρν), and yellow (,χρν), but it does not specify the exact correspondence between these colours and the four elements (1. 15. 3 = A 92). On the basis mainly of this text scholars in the past have often assumed that Empedocles associated the four elements with four specific colours.15 It has rightly been borne, ‘Colour Concepts of the Ancient Greeks’, British Journal of Aesthetics, 8 (1968), 269–83; V. J. Bruno, Form and Color in Greek Painting [Form and Color] (New York, 1977), 47–51; J. J. Pollitt, ‘Περ χρωµτων: What Ancient Greek Painters Thought about Colors’, in M. A. Tiberios and D. S. Tsiafakis (eds.), Color in Ancient Greece (Thessaloniki, 2002), 1–8. On Empedocles’ view, see below. 14 )µβροτα probably refers here to the moon and the stars, which are understood as combinations of fire and air; cf. Wright, Extant Fragments, 178. • 15 K. Prantl, Aristoteles: Uber die Farben (Munich, 1849), 41–2; W. Kranz, ‘Die

Created on 4 July 2005 at 14.33 hours page 11

Empedocles on Colour and Colour Vision

11

suggested, though, that the doxographical tradition is most probably misleading, in that it ascribes to Empedocles the four colours of Democritus. For we learn from Theophrastus (De sens. 73–5 = 68 A 135 DK) that Democritus postulated four simple colours, namely white, black, red, and green (χλωρν). A•etius, however, presumably misreports this and presents the four Democritean colours as white, black, red, and yellow (1. 15. 8); and this is the list which the doxographical tradition then erroneously ascribes to Empedocles too, as well as to the Pythagoreans (1. 15. 7).16 But even assuming that this explanation of how part of the later tradition came to ascribe these four colours to Empedocles is correct, we can only speculate about the reason why A•etius or his source introduces yellow instead of green in all four-colour lists which he attributes to the ancient philosophers in question. For instance, it is tempting to connect A•etius’ mistake with the fact that from quite early on in classical antiquity both medical doctors and painters, each group for its own reasons, showed a particular interest in this same list of four colours, namely white, black, red, and yellow. In particular, the four humours which ancient doctors postulate and whose imbalance they consider to be the basic reason for lack of health, i.e. blood, phlegm, and black and yellow bile, are intrinsically characterized by the same four colours, red, white, black, and yellow, respectively. And it is interesting that a younger contemporary of Empedocles, Diogenes of Apollonia, claimed that all disease can be diagnosed on the basis of whether the patient’s outward appearance displays one of these four colours; for it is the patient’s colour, according to Diogenes, which unmistakably reveals the predominant humour in the patient’s body ([Galen], De humor. xix. 495 Kuhn • = 64 A 29a DK; Theophr. De sens. 43 = 64 A 19 DK). Could this mean that Empedocles, who also thought of himself as a healer (e.g. B 112), was influenced by the four-humour theory when he introduced the four elements and their colours, or could a• ltesten Farbenlehren der Griechen’ [‘Farbenlehren’], Hermes, 47 (1917), 126–40 at 127–8; H. Cherniss, Aristotle’s Criticism of Presocratic Philosophy [Criticism] (Baltimore, 1935), 217 n. 280; E. Siegel, ‘Theories of Vision and Colour Perception of Empedocles and Democritus; Some Similarities to the Modern Approach’ [‘Theories of Vision’], Bulletin of the History of Medicine, 33 (1959), 145–59 at 152–3. 16 For the erroneous attribution to Empedocles of a four-colour list, cf. H. Diels Doxographi Graeci (Berlin, 1879), 50 and 222; J. I. Beare, Greek Theories of Elementary Cognition [Greek Theories] (Oxford, 1906), 21 n. 6; Millerd, Interpretation, 83; Longrigg, ‘Roots’, 432–3.

Created on 4 July 2005 at 14.33 hours page 12

12

Katerina Ierodiakonou

it be the other way round, namely that the ancient doctors who accepted the four-humour theory were influenced by Empedocles’ four elements and their colours? To assume that the doctors influenced Empedocles, at least as far as this issue is concerned, is implausible. For the four-humour theory seems to have become a standard doctrine, if at all, only after Empedocles’ time. On the contrary, there is enough evidence to suggest that it is the Empedoclean theory of four elements that exercised considerable influence upon those doctors who postulated the four humours.17 For example, Empedocles’ influence can be traced in the Hippocratic treatise De natura hominis (chs. 4–7), in which the author closely links the four humours to the four opposites, namely hot and cold, dry and wet, which are said to characterize the four elements; that is to say, phlegm is cold and wet, blood is wet and hot, yellow bile is hot and dry, and black bile is dry and cold. And in the fourth century bc Diocles of Carystus (fr. 8 Wellmann) directly connects the four humours with the four elements and the four opposites; that is to say, phlegm with air and cold, blood with water and wet, yellow bile with fire and hot, and black bile with earth and dry. However, the fact that medical doctors after Empedocles are influenced by his theory of the four elements does not necessarily imply that they are also influenced, in their choice of humours, by the colours of Empedocles’ four elements. In other words, we cannot infer that Empedocles attributed to his elements the colours white, black, red and yellow, just because the ancient doctors after Empedocles talk about four humours with the very same four colours. Besides, fire is for Empedocles white, whereas the doctors associate it with yellow bile, and hence with yellow; furthermore, water for Empedocles is black, whereas the doctors associate it with blood, and hence with red. To sum up, it is not reasonable to suggest that Empedocles’ basic elements must have four di·erent colours, namely white, black, red, and yellow, and hence that A•etius was right, solely on the basis that in the ancient medical tradition there is the doctrine of the four humours with these particular colours. Nevertheless, nothing prevents us from thinking that this tradition may have been on 17 On the relationship between Empedocles’ four elements and the theory of four humours, cf. Kranz, ‘Farbenlehren’, 130; J. Longrigg, ‘Philosophy and Medicine: Some Early Interactions’ [‘Philosophy and Medicine’], Harvard Studies in Classical Philology, 67 (1963), 147–75 at 153.

Created on 4 July 2005 at 14.33 hours page 13

Empedocles on Colour and Colour Vision

13

A•etius’ mind, or on that of a careless source, when he ascribed the four-colour list to the philosophers, including Empedocles. Turning next to ancient painting, there is enough evidence, both archaeological and textual, that at a certain time in the classical period ancient painters used just four colours, namely white, black, red, and yellow; or as Pliny says (Nat. 35. 50), white from Melos, Attic yellow, red from Sinope, and lamp-black (atramentum).18 Plutarch confirms this (De def. or. 436 b–c), as does the pseudoAristotelian treatise De mundo (396B13), while Pliny, again, mentions Apelles’ painting Alexander Holding the Thunderbolt as an example of a painting in which only these four colours were used (Nat. 35. 92). Unfortunately, no original example of Greek pictorial art of the classical period survives. Historians of art manage to reconstruct the development of Greek painting largely depending, for instance, on the painted decorations of Greek pottery and on Roman copies of Greek murals and mosaics. Among such copies, the magnificent Alexander mosaic of the late second century bc from the House of the Faun in Pompeii, presumed to be a copy of a late fourth-century original by Philoxenus of Eretria, provides us with a typical example of a work executed in shades derived from white, black, red, and yellow.19 It has therefore been suggested that there must be a close connection between the ancient painters’ four colours and Empedocles’ colours of the four elements. And since there is no evidence that ancient painters were interested in Empedocles’ theory of the four elements, B 23 has often been used to support the claim that Empedocles was probably influenced by the painters’ practice, especially 18 On the production of these pigments in antiquity, cf. Pliny, Nat. 35. 30–49; Vitr. 7. 7–14. 19 The Alexander mosaic, known as The Battle of Issus, is now at the National Museum in Naples. A. Cohen (The Alexander Mosaic: Stories of Victory and Defeat (Cambridge, 1997), 167–9) remarks that green is occasionally used for some details, but it is insu¶cient to a·ect the general four-colour character of the mosaic. Sellers (K. Jex-Blake and E. Sellers, The Elder Pliny’s Chapters on the History of Art [Elder Pliny] (London, 1896), 97) mentions a modern example of the four-colour technique, The Crowning with Thorns by Titian in the Munich Pinakothek (c.1570– 6). G. Morelli (Italian Masters in German Galleries, trans. L. M. Richter (London, 1883), 43), who initially made the comparison between the Alexander mosaic and Titian’s painting, says that the aged Titian’s example was afterwards often followed by Rubens and Van Dyck, but most brilliantly by Frans Hals in the last years of his life; for instance, in his two celebrated paintings Regents of the Old Men’s Almhouse and Regentesses of the Old Women’s Almhouse, which are both dated 1664 and now belong to the Frans Halsmuseum in Haarlem.

Created on 4 July 2005 at 14.33 hours page 14

14

Katerina Ierodiakonou

in his supposed choice of the colours white, black, red, and yellow.20 But B 23 cannot really help us to settle this issue, for the comparison with painting in this fragment says nothing about the colour of the four elements. Rather, the comparison is supposed to work between the four elements used by Love and Strife for the creation of the world and the few colours painters used to paint the world; not between the four colours of the four elements and the four colours used by painters. In other words, the comparison would work perfectly even if some elements were not intrinsically coloured, or even if fewer than four colours characterized the four elements. Further, to assume that the four-colour technique of the ancient painters influenced Empedocles’ choice of colours for his four elements presupposes that this technique was standard at least during Empedocles’ life, if not beforehand. There is an ongoing discussion, however, as to when exactly the four-colour palette was introduced in ancient painting; historians of Greek art have interpreted ancient sources on this matter quite di·erently. The two relevant texts are Pliny, Nat. 35. 50, and Cicero, Brut. 18. 70: Pliny claims that the artists who used four colours were Apelles, Aetion, Melanthius, and Nicomachus, who all lived during the late fourth century bc. On the other hand, Cicero’s list of four-colour painters begins with Polygnotus, who was a very near contemporary of Empedocles, and continues with Zeuxis and Timanthes, who belong to the late fifth and early fourth centuries. Some art historians have suggested that Pliny’s and Cicero’s remarks do not have to be treated as contradictory. For it makes perfect sense to suppose that painters of the fifth century, e.g. Polygnotus, were the first to paint with just four colours, whereas the painters of the fourth century further developed this technique.21 More recently, though, others have insisted that Pliny’s remarks are more authoritative and that Cicero is plainly mistaken. They claim that ancient painters started using only four colours not earlier than the fourth century.22 Therefore, it would certainly be hazardous to assume that the painters of Empedocles’ time used the four colours white, black, red, and yellow, and 20 I. Scheibler, ‘Die “vier Farben” der griechischen Malerei’ [‘Farben’], Antike Kunst, 17 (1974), 92–102 at 101; Bruno, Form and Color, 56–7. 21 Jex-Blake and Sellers, Elder Pliny, 96–7. 22 J. J. Pollitt, The Ancient View of Greek Art [Ancient View] (New Haven, 1974), 110–11.

Created on 4 July 2005 at 14.33 hours page 15

Empedocles on Colour and Colour Vision

15

that Empedocles was influenced by them. But even if Polygnotus did restrict his colours to four, which in itself is extremely questionable, it seems clear that this was not the standard practice of the fifth-century painters, let alone in Empedocles’ day. However, even if it is an unfounded claim that Empedocles postulated for the four elements the same four colours which the painters of his time used, perhaps the painters’ practice does help us better to understand what the issue at stake was concerning colours at that particular period. Historians of Greek art have argued that, whenever ancient painters decided to limit the number of colours they used (and archaeological evidence suggests that they did not use four colours in all of their works), it was always a deliberate choice on their part. For even the earlier painters must have had a greater range of colours open to them, as becomes clear, for instance, from traces of blue and green paint which are found on buildings and sculptures; in fact, blue and green seem to have been available to ancient painters as early as the archaic period.23 But can we explain this deliberate restriction of colours? In the case of the fourth-century painters, who undoubtedly had achieved a high level of sophistication in producing new colours, their decision to use fewer colours, definitely including white and black, was motivated by their interest in naturalistic representation.24 For in order to simplify experiments with three-dimensional forms, ancient painters must have found it helpful, if not necessary, to restrict their colours. This trend may well go back to the fifth century, or even to Polygnotus, which would explain Cicero’s remarks. Thus, the palette with fewer colours seems to have been the invention of those ancient painters who were the major participants in the discovery of the technique known in European art as chiaroscuro, the method which, by using highlights and cast shadows, tries to simulate our normal optical experience of how light falls on objects. This is, in fact, the method the ancients called σκιαγραφ α, which literally means drawing or painting with shading.25 23 M. Robertson, Greek Painting [Painting] (Geneva, 1959), 13; A History of Greek Art [Art] (Cambridge, Mass., 1975), 260 and 500; Scheibler, ‘Farben’, 98–9. 24 Bruno, Form and Color, 66. 25 E. Pfuhl, ‘Apollodoros . Σκιαγρφος’, Jahrbuch des Arch•aologischen Instituts, 25 (1910), 12–28; ‘Skiagrafia’, ibid. 27 (1912), 227–31; Pollitt, Ancient View, 247–54. For a di·erent interpretation of σκιαγραφ α as an impressionistic technique which relies on the phenomenon of optical colour fusion, cf. E. Keuls, ‘Skiagraphia Once Again’, American Journal of Archaeology, 79 (1975), 1–16; E. G. Pemberton, ‘A Note

Created on 4 July 2005 at 14.33 hours page 16

16

Katerina Ierodiakonou

Hence, since painters of the fifth and fourth century were mainly interested in producing an atmosphere of chiaroscuro, they used the colour white as the equivalent of light, and the colour black as the equivalent of darkness, neither of which we would nowadays include among primary colours.26 But, again, there are two di·erent statements by ancient authors describing the invention of σκιαγραφ α. In Plutarch (De glor. Ath. 2) Apollodorus is supposed to be the inventor of this technique, an account also supported by Pliny’s discussion of Apollodorus’ contribution to art towards the end of the fifth century (Nat. 35. 60–1). On the other hand, Quintilian (Inst. 12. 10. 4) states that it is Apollodorus’ student Zeuxis who ‘invented the law of lights and shades’ (luminum umbrarumque invenisse rationem). Some art historians have suggested that Quintilian must be mistaken.27 Others have tried to reconcile the two apparently contradictory statements, by proposing that the two ancient sources may very well be speaking of two di·erent moments in the development of σκιαγραφ α: Apollodorus was probably the first to perfect the shading methods and impart to his paintings a more convincing three-dimensional appearance, whereas Zeuxis may have introduced a kind of chiaroscuro in which shading assumed a more dominant role and the nuances of colouring became more and more complex.28 Finally, in the fourth century Apelles, Protogenes, Pausias, and Nicias fully achieved in their works the naturalistic ideal, both by successful foreshortening and shading and by the use of mixed colours (Pliny, Nat. 35. 79 ·.). It seems, therefore, that it took the ancient painters some time to develop the technique of σκιαγραφ α. In fact, its history must have started even before Apollodorus and Zeuxis. For although it is true, generally speaking, that there is no depth in archaic Greek painting on Skiagraphia’, American Journal of Archaeology, 80 (1976), 82–4. In addition, A. Rouveret, Histoire et imaginaire de la peinture ancienne (Rome, 1989), 13–63, interprets it as a technique used in the classical period especially at the theatre, i.e. as painting en trompe-l’¥il. 26 Since the pigment which ancient painters actually used for black was really dark blue, they were able to produce a high degree of nuance both in chiaroscuro and in the variety of reds and yellows, browns and ochres obtainable. So, although blue was omitted as an active colour in ancient paintings of this period, this darkening agent together with white, red, and yellow made the four-colour palette a reasonable ‹ choice for painters of the classical period. Cf. E. Bertrand, Etudes sur la peinture et la critique d’art dans l’antiquit‹e (Paris, 1893), 132–44; Siegel, ‘Theories of Vision’, 153–4; Bruno, Form and Color, 58–9 and 79–87. 27 Pollitt, Ancient View, 252. 28 Bruno, Form and Color, 28–9.

Created on 4 July 2005 at 14.33 hours page 17

Empedocles on Colour and Colour Vision

17

up to the early fifth century, some kind of primitive shading does appear on late archaic works and in the early classical period. And certainly before the middle of the fifth century painters like Polygnotus seem to have deliberately restricted their colours, precisely because their main concern was to develop the chiaroscuro e·ect by using the colour white for highlights and the colour black for shades.29 That is to say, it may be perfectly true that both Apollodorus and Zeuxis played a crucial role in the development of σκιαγραφ α, but they presumably depended on previous generations of painters, systematizing and further developing their experiments. Thus it may well be that when Empedocles undertook to construct a cosmological theory in order to explain the constitution of the world, he was also interested in explaining light and darkness and thus associated the colour white with the brightness of the element fire and the colour black with the darkness of the element water. I do not, however, want to claim that there is a direct influence of painters’ practice in the fifth century on Empedocles’ discussion of the colours white and black. Instead, I want to suggest that it seems to have been a preoccupation of the time to understand how light falls on objects and how the contrast between light and darkness is produced. I therefore do not believe that, just because Empedocles did not provide colours for all elements and did not associate all four elements with other opposites, such as the hot and the cold, the dense and rare, or the bitter and the sweet, his theory was not fully worked out.30 Rather, one reason why he focused on only two colours, white and black, may have been that it is on the basis of these that light and darkness, which play such a crucial role in the way things present themselves to us, including the way they appear coloured, can adequately be accounted for. To sum up, scholars and historians of Greek art have usually stressed the connection between Empedocles and ancient painting with reference to the use of only four colours in a type or style of ancient painting. Nevertheless, the part of the doxographical tradition which claims that Empedocles associated four colours with the four elements is not reliable. The painters’ practice, just like the medical doctrine of the four humours, may have been in A•etius’ 29 On the main stages of the history of σκιαγραφ α, cf. Robertson, Painting, 14–15 and 153; Art, 489. 30 Cf. G. E. R. Lloyd, ‘The Hot and the Cold, the Dry and the Wet in Greek Philosophy’, Journal of Hellenic Studies, 84 (1964), 92–106 at 93 n. 4.

Created on 4 July 2005 at 14.33 hours page 18

18

Katerina Ierodiakonou

mind when he attributed to Empedocles a four-colour theory. But our other and more authoritative ancient sources all talk of Empedocles’ two colours, namely white and black, the colours of fire and water respectively. And there is no evidence to suggest that he provided di·erent colours for the other two elements, namely air and earth, or that he believed that air and earth themselves are intrinsically coloured. But if it really is the case that Empedocles postulated only the colours of fire and water, his claim that from the combination of elements all colours are generated remains a puzzle. For how are we to explain the production of all colours from the mixture of black and white alone?31 Interestingly enough, Empedocles is not the only ancient philosopher to defend such a counter-intuitive position. Aristotle, too, puts forward the very same view in De sensu (439B18 ·.), in which he claims that it is from the complete mingling in di·erent ratios of just black and white that we derive the rest of the colours. Moreover, before presenting his own explanation of how this is possible, Aristotle discusses another view on the production of colours, according to which all colours are produced by the juxtaposition (παρ0 )λληλα θσις) of very small white and black particles which are the constituents of all objects; although the white and black particles are not themselves apparent, because of their minute size, objects acquire certain colours as a result of the specific ratio in which the white and black particles are found in each object. It seems reasonable to attribute this view to Empedocles, since it fits in well with what we know from our other ancient sources about Empedocles’ account of the generation of things in this world. For, as we have said, everything is generated on Empedocles’ view from 31 No Empedoclean fragment talks about mixing other colours apart from white and black. The only fragment which has been interpreted in this way is B 93 βσσ1ω δ2 γλακοιο κρκου καταµ σγεται %κτ ς (Bennet’s text). It is not clear, though, what the term βσσ1ω means here: it has been understood as referring to linen, so that sa·ron mixes with linen in the sense of dyeing linen (Wright, Extant Fragments, 232–3); it has also been understood as referring to a purple colour (cf. Suda; Hesychius), so that sa·ron mixes with purple to make a better dye (J. Barnes, review of M. R. Wright, Empedocles: The Extant Fragments, Edited with Introduction, Commentary, Concordance (New Haven, 1981), in Classical Review, ns 32 (1982), 191–6 at 194). The context in which Plutarch quotes this fragment (De def. or. 433 b) seems to favour the latter interpretation, since the other two examples used by Plutarch in this text are both examples of mixing di·erent things in order to produce a better dye. But even if it is the case that Empedocles talks here of the mixture of two colours, i.e. purple and sa·ron, we still get no information about the production of a new hue.

Created on 4 July 2005 at 14.33 hours page 19

Empedocles on Colour and Colour Vision

19

the combination of the four elements, which are placed side by side and made up of minute discrete particles. So, if the elements of fire and water are respectively white and black, an object has a specific colour depending on the amount of the minute particles of these elements which constitute the particular object; or more specifically, depending on the mathematical ratio in which these elements are to be found in the particular object. But there is still the question of why Empedocles would think that the mere juxtaposition of white and black particles generates in some cases an object with a colour other than white or black, e.g. red, yellow, or blue. To deal with this problem, I think we first need to settle a more basic question, what Empedocles actually means when he says that fire is white and water is black. It would not make much sense to think of these colours the way we think of them today, when we say that milk is white or a crow is black; for placing side by side any amounts of these colours in whatever ratio would never give us the impression of colours like red, yellow, or blue. To be more precise, it would not make sense to think of these colours only in this way, for after all Empedocles does characterize milk as white and there is no reason to believe that he would not characterize a crow as black. At the same time, however, when Empedocles characterizes men as black (B 67) and bones as white (B 96), he must have something di·erent in mind; for he is certainly not referring only to Ethiopians when he talks of black men, and bones are not exactly white in the way milk is. Most importantly, when Empedocles talks of the sun as representative of the element of fire (B 21. 3) and characterizes it as ‘white’ (λευκν), it becomes clear that his notion of white is not limited to the white of the milk. Does this mean that the standard Greek terms for ‘white’ (λευκν) and ‘black’ (µλαν) cover a wide range of colours, that λευκν, for instance, refers not only to white but also to yellow, orange, and even red, although there are other terms in Greek to distinguish these colours, which Empedocles undoubtedly knows and uses? In following such a practice, it could be argued, Empedocles proves not to be very di·erent from ourselves, when we talk of white wine, or black grapes, or black and white men, in a sense which is not that found in ‘white’ milk and the ‘black’ crow. Nevertheless, this suggestion does not settle the issue. For, if that is the whole story, what exactly is the sense of ‘white’ and ‘black’ that Empedocles has in mind when he suggests that the combination of these colours

Created on 4 July 2005 at 14.33 hours page 20

20

Katerina Ierodiakonou

may produce all other colours? On this account all other colours would just be forms of white or black, rather than the result of a mixture of them. Another way to approach this issue is to think of the two main cases which Empedocles himself discusses, but which are also used as standard examples throughout antiquity: namely, the case of the sun, or fire, as being white, and the case of rain, or the sea, or water, as being black. There is a fragment quoted in one of Plutarch’s Quaestiones naturales (39) in which Empedocles claims that, when water is not illuminated by the sun’s rays, as in the depths of a river or in cavernous grottoes, it is black (B 94): And in the depths of the river a black [niger] colour is produced by the shade, and in the same way it is observed in cavernous grottoes.

This fragment, which is preserved only in a Renaissance Latin translation by Gilbert Longueil, is presented by Plutarch as an appropriate answer to the question ‘Why does the surface of the water look white and the depths black?’ It seems, therefore, that according to Empedocles’ account the colour of water, though black by nature, changes depending on whether it is illuminated, e.g. at the surface of the sea, where it looks light blue and sometimes even white, or is not illuminated, e.g. in the depths of the sea, where it really looks dark blue and black. The production of all the different colours in this case is thus due to the penetration of water by light, which on Empedocles’ view is fire emitted by a luminous body, i.e. the sun (e.g. Arist. DA 418B20–6; Philop. In DA 344. 33–7 Hayduck; Cod. Ath. 1249 = A 57). In other words, all the di·erent colours in this case are the result of the combination of watery with fiery particles in di·erent proportions. And what about the colour of the sun? In what sense is it white, and does it always remain white? Though at noon the sun looks white, at sunrise and around sunset it may look yellow, orange, pink, and red. The colours yellow, orange, pink, and red are thus all produced by the fact that the fiery particles of the white sun are combined in different proportions with the watery particles of the moisture which exists in the atmosphere (cf. Arist. De sensu 440A10–12; Meteor. 374A7–8). And if the amount of water were to increase, according to Empedocles, we would get all kinds of colour, until little or no fire is left, and then we would get dark blue or black.

Created on 4 July 2005 at 14.33 hours page 21

Empedocles on Colour and Colour Vision

21

Hence, when Empedocles talks of the colour of the sun, or the light of the sun, and of that of the sea, he refers to the two extreme colours white and black, but he also has in mind a whole range of colours, indeed all other colours, which are said to be produced by the combination of black and white. For the colour of the sea, being by nature black, gradually changes to the other extreme, depending on how much it is illuminated by the sun, i.e. penetrated by fire; and the colour of the light of the sun, being by nature white, changes to a whole range of colours, depending on how much moisture, i.e. water, there is in the atmosphere. That is to say, it seems that Empedocles understands the production of all the colours which characterize the things in this world as resulting from the combination of fiery and watery particles. For this is what he observes in nature, this is how the sun and the sea seem to him to acquire and constantly change their colours. And although the only Empedoclean fragment which mentions the rainbow does not say anything about how its colours arise (B 50), it was surely clear to Empedocles, as to the ancients in general, that the colours of the rainbow somehow must be due to the light of the sun and the water of the rain; for they will have noticed that we see rainbows only in places on which the sun shines and where water, e.g. in the form of rain, is dispersed. According to Empedocles, therefore, the colours white and black should be regarded as two extremes in a continuum, like hot and cold, or day and night. There is something that in absolute terms can be said to be the white and the black, i.e. the elements of fire and water, and everything else is characterized by colours which are understood as shades of black and white. For instance, something’s being yellow is understood as being more white than black, in the sense that it consists of more fire than water; and something’s being blue is understood as being more black than white, in the sense that it consists of more water than fire. Let us take two of Empedocles’ own examples. Blood and flesh, which are both red, are said by Empedocles to contain an equal proportion of all the elements (B 98; A•etius 5. 22. 1 = A 78; Theophr. De sens. 10 = A 86); thus, the colour red seems to occupy the middle point of the continuum, in the sense that it is no more white than black, since it is produced by the same amount of white fiery and black watery particles. Furthermore, Empedocles is reported to have said (Arist. GA 779B15–20 = A 91; [Arist.] Probl. 910A12–15) that eyes with more water than fire are black (µλανα), whereas eyes with less water than fire are grey-

Created on 4 July 2005 at 14.33 hours page 22

22

Katerina Ierodiakonou

blue (γλαυκ); so he clearly thinks that the di·erence between the colours grey-blue and black depends on the amount of fiery and watery particles contained in the objects which are characterized by these colours. To conclude, Empedocles claims that the colour of an object depends on the constituent elements of the particular object, and more specifically on the proportion of the black watery particles and the white fiery particles which the object contains; more watery particles produce a darker colour, while more fiery particles produce a brighter colour. Hence, it seems that Empedocles’ continuum between the white and the black should rather be understood as a continuum between the bright and the dark. This means, however, that Empedocles puts forward two views that sound counterintuitive to us: namely, that the other two elements, i.e. not only air but even earth, are colourless, and that all colours are produced by the combination of the white and the black, or the bright and the dark. Counter-intuitive though they may be, these are the views, it seems, which Empedocles puts forward, and I hope to have shown the rationale behind them. Instead of starting from our modern views about the nature of colour and the production of di·erent colours, we need to remember that Empedocles lives in a culture strongly inclined to think in terms of opposites, in this case the opposites white and black or bright and dark. It is not the case that Empedocles’ ‘imaginary vividness took hold of him with more persuasiveness than logical consistency’, so that ‘the important thing in understanding him is to stop thinking at the right moment’;32 the task rather is to try to think like him and his contemporaries, and to uncover the assumptions underlying this way of thinking. Let us now turn to Empedocles’ theory of colour perception. There are, fortunately, enough texts to provide us with a fairly clear picture of his views on perception, and in particular the perception of colour. To start with, there is the famous passage from Plato’s Meno (76 c 4–d 5 = A 92): ‘Do you want me to answer after the manner of Gorgias,33 which would enable you most easily to follow?’ 32 Millerd, Interpretation, 21. 33 On the relationship between Gorgias and Empedocles, cf. H. Diels, ‘Gorgias und Empedokles’ [‘Gorgias’], Sitzungsberichte der Preu¢ischen Akademie der Wissenschaften, 49 (1884), 343–68.

Created on 4 July 2005 at 14.33 hours page 23

Empedocles on Colour and Colour Vision

23

‘Of course I want to.’ ‘You say, then, following Empedocles, that there are certain e}uences [%πορρος] from things?’ ‘Certainly.’ ‘And pores [πρους] into which and through which the e}uences move?’ ‘Definitely.’ ‘And that some of the e}uences fit into some of the pores, and others are too small or too large?’ ‘That is so.’ ‘You also say, then, that there is such a thing as the organ of vision?’ ‘I do indeed.’ ‘ “Grasp what I tell you”, as Pindar said, on the basis of these points. For colour is an e}uence from things which is commensurate with the organ of vision [3ψει σµµετρος] and is perceptible.’

So, according to Plato, Empedocles claims that whatever human beings perceive in the world, they perceive it because of di·erent kinds of ‘e}uences’ which are emitted by every object and enter into the ‘pores’ of our sense-organs, in particular into those pores which are commensurate with them. In the case of visual perception, more specifically, we see colours because certain e}uences emitted by objects reach our eyes and, since they are commensurate with the pores of the eyes, they enter into our eyes and give us information about the colour of these objects. There are two points worth making concerning the way Plato in this passage presents Empedocles’ theory of colour perception. First, it seems that on Empedocles’ view what we obtain through our sense of sight are perceptions not of objects, but of colours. Since colour is the e}uence from objects which is commensurate with the pores of the visual organ, what we are really able to perceive with our eyes are colours; and we say that we see things, just because we see their colours. Second, the way the e}uences enter into and move through the pores of the eyes is described here by the verb ‘fit’, ρµττειν. This reminds us of the fitting together of the four elements on the basis of which everything is generated; but in this case the fitting of the e}uences into the pores is said to be harmonious in the sense that the e}uences and the pores are commensurate (σµµετρος), i.e. neither smaller nor larger, but of the same size.34 34 Examples of commensurability and of lack of commensurability are given in a fragment in which Empedocles talks of water being mixed with wine and of the

Created on 4 July 2005 at 14.33 hours page 24

24

Katerina Ierodiakonou

Theophrastus helps us to add some further details to our account of Empedocles’ theory of colour perception (De sens. 7 = A 86): Empedocles gives a similar account for all the senses and says that sense perception occurs by means of [things] fitting into [+ναρµττειν] the pores of each [sense-organ]. That is why they cannot discern each other’s objects, because some senses happen to have pores which somehow are too wide for the object of perception [i.e. of another sense], while others have pores which are too narrow, so that the objects which do not touch [ο5χ πτµενα] are able to pass through steadily and the others are completely unable even to get in. And he also tries to give an account of what the organ of vision is like. He says that the inside of it is fire and around this are water,35 earth, and air, through which fire passes, being fine like the light in lanterns. And the pores [of the organ of vision] are alternately of fire and water; we recognize white things with the pores of fire and black things with those of water (for each sort fits into the respective pores). And the colours are brought to the organ of vision by the e}uence.

Theophrastus thus confirms the two points made in Plato’s Meno, namely that (1) each sense has di·erent objects, because the pores of our di·erent sense-organs have di·erent sizes, and thus each sense-organ can obtain e}uences only of a certain size; and (2) the object of vision is colour, since we see when certain e}uences from the objects around us which correspond to their colour fit into (+ναρµττειν) the pores of our eyes. Furthermore, Theophrastus’ description of how on Empedocles’ view our senses successfully perceive objects helps to clarify why it is essential for the e}uences from the objects and the pores of our sense-organs to be of the same size. For, although it is clear why perception is impossible when the e}uences are larger than the pores, as in this case they cannot even enter them, it is at first puzzling why we cannot perceive when the pores of our senseorgans are larger than certain e}uences; one might expect that, inability of water to be mixed with oil (B 91). More generally, on the commensurability between the e}uences from objects and the pores of our sense-organs, cf. Arist. GC 324B26–35 = A 87; Theophr. De sens. 12 = A 86; A•etius 4. 9. 6 = A 90. 35 Some scholars have been reluctant to emend the text, as Diels does, by adding water to earth and air. They have claimed either that Theophrastus must have taken the presence of water for granted (Millerd, Interpretation, 83), or that at this particular point he is simply describing the composition, according to Empedocles, of the part inside the eye which consists only of the internal fire and the membranes separating it from the internal water (G. M. Stratton, Theophrastus and the Greek Physiological Psychology before Aristotle (London, 1917), 163–4 n. 25; Wright, Extant Fragments, 242).

Created on 4 July 2005 at 14.33 hours page 25

Empedocles on Colour and Colour Vision

25

since the e}uences are smaller than the pores, they could enter into the pores and move through them with no problem. But if we assume that certain e}uences from objects, e.g. those which correspond to how an object smells or tastes, are smaller than, for instance, the pores of our eyes, we could then see smells and tastes; and this is certainly counter-intuitive. So, Theophrastus makes it clear that on Empedocles’ view the important prerequisite for any kind of sense perception to take place is that the e}uences from objects exactly fit into the pores of our sense-organs, in the sense that they actually touch the pores of our sense-organs; and it is for this reason that Theophrastus uses in this passage the verb ‘touch’. That is to say, Empedocles seems to think, according to Theophrastus’ report, that any sense perception involves some kind of touching. Theophrastus in this passage also provides us with two further pieces of information about Empedocles, namely (1) his description of the structure of the human eye; and (2) his explanation of what happens when we see white or black. Concerning the anatomy of the eye, Theophrastus says that, according to Empedocles, it mainly consists of fire which is surrounded by water, earth, and air, and has pores which are alternately pores of fire and of water. In addition, he compares the way our eyes see to the way a lantern works; this simile, in fact, is also found in a much-discussed fragment in which Empedocles himself compares the way our eyes are built and function to the way a lantern is made and works (B 84):36 As when someone planning a journey prepared a lantern, A flame of burning fire through a wintry night, and fastened linen screens against all kinds of breezes, which scatter the wind of the blowing breezes, but the light leapt outwards, to the extent that it was finer, and shone across the threshold with unfailing beams; in this way [Aphrodite] gave birth to the rounded eye,37 primeval fire wrapped in membranes and in delicate tissues;

36 I do not follow Diels’s text, in which an extra line is inserted after line 8, <α6> χον!ησι δ αντα τετρ&ατο θεσπεσ !ησιν. This line, which has been pieced together from garbled words found in a single manuscript, has been discarded by many scholars for sound philological reasons; cf. J. Bollack, Emp‹edocle, vol. iii (Paris, 1969; repr. 1992), 327; Wright, Extant Fragments, 241. 37 I here translate the text as emended by F•oster and Ross; cf. Wright, Extant Fragments, 241.

Created on 4 July 2005 at 14.33 hours page 26

26

Katerina Ierodiakonou these kept back the surrounding deep water but let through fire, as much of it as was finer.

The structure of the human eye, as presented here, is relatively clear:38 the standard account o·ered is that the fiery part of the eye is the lens, surrounded by water and protected by membranes and tissues.39 It has also been argued, though, that in this fragment we are given a description of the externally visible features of the eye, rather than a description of the eye’s hidden internal structure; in this case, then, the fiery part of the eye is most likely the iris, the membranes and tissues protecting it are identified with the cornea, and the water which the membranes are said to keep out is the moisture on the surface of the cornea, i.e. the lachrymal fluid.40 As to the simile of the lantern, it has also been interpreted in many di·erent ways among both ancient and modern readers of Empedocles’ verses. Aristotle (De sens. 437B23–438A5), for instance, who cites this fragment, accuses Empedocles of being inconsistent; for if the simile of the lantern is taken seriously in all its details, and especially if the fact that the lantern emits light means that the eyes emit fire, then it seems that Empedocles explains vision both on the basis of incoming e}uences and on the basis of light issuing from the eyes. Modern interpreters, too, have been unable to agree on this matter, though many ingenious attempts have been made to find an appropriate solution.41 I am inclined to think that we 38 Perhaps it was Alcmaeon (24 A 5 and 10 DK), before Empedocles, who gave for the first time an account of the anatomy of the eye and its function. However, it is still disputed to what degree Empedocles was really influenced by Alcmaeon’s views on the structure of the eye; cf. Diels, ‘Gorgias’, 353–4; Beare, Greek Theories, 15; Longrigg, ‘Philosophy and Medicine’, 156–7; ‘Roots’, 437; Wright, Extant Fragments, 230 and 243. 39 There is a di·erence of opinion among scholars as to whether the membranes are membranes separating the internal fire from the internal water (Beare, Greek Theories, 16; Wright, Extant Fragments, 241–2) or, as I am inclined to assume, membranes separating the inside of the eye, namely the fire and the water, from the outside (Lloyd, Polarity, 326). 40 D. Sedley, ‘Empedocles’ Theory of Vision and Theophrastus’ De sensibus’ [‘Vision’], in W. W. Fortenbaugh and D. Gutas (eds.), Theophrastus: His Psychological, Doxographical and Scientific Writings (New Brunswick, 1992) 20–31 at 20–6; cf. also V. Caston, ‘Empedocles’ Theory of Vision’ (diss., University of Texas; Austin, 1985), 17–23. This interpretation is partly based on the assumption that dissection was not practised at the time of Empedocles, and thus he need not have had knowledge of the internal features of the eye; cf. G. E. R. Lloyd, ‘Alcmaeon and the Early History of Dissection’, Sudho·s Archiv, 59 (1975), 113–47, repr. in Methods and Problems in Greek Science (Cambridge, 1991), 164–93. 41 Beare (Greek Theories, 17), Millerd (Interpretation, 84–5), and Cherniss (Criti-

Created on 4 July 2005 at 14.33 hours page 27

Empedocles on Colour and Colour Vision

27

should not try to press the analogy between the way the lantern is made and works and the way the eye is built and functions in all its details. Rather, we should limit ourselves to understanding the simile the way Theophrastus does: namely, as an analogy between the function of the membranes which surround the internal fire of the eye and that of the linen screens which surround the lantern. For just as the linen screens of the lantern let the light pass through without allowing the wind to extinguish the light of the lantern, the membranes of the eye let the fire pass through without allowing the water to disperse. But whatever we think about the details of the simile of the lantern as well as about those of the structure of the eye, there is no doubt that Empedocles understands colours as e}uences from objects which are commensurate with the pores of the visual organ, and thereby capable of being seen. Moreover, according to Theophrastus’ account, Empedocles claims that human beings see colours because they are able to perceive the colour white with the pores of fire and the colour black with the pores of water. So, how are we able to perceive the colour white and the colour black? The fact that white objects are white means, according to Empedocles, that they cism, 317–18 n. 106) argue that the two accounts of vision presented by Empedocles cannot be reconciled, and the simile of the lantern should be carried no further than the adoption of a simple analogy between the interior of the eye and a lantern. G. R. T. Ross (Aristotle: De Sensu and De Memoria (Cambridge, 1906), 137– 8) claims that both accounts are needed, since the images of things entering by means of the pores have to be illuminated by the fire issuing from the pupil. W. J. Verdenius (‘Empedocles’ Doctrine of Sight’ [‘Doctrine’], in Studia Varia Carolo Gulielmo Vollgra· Oblata (Amsterdam, 1948), 55–64) assigns di·erent functions to both the incoming e}uences and the outgoing ocular fire: seeing is imagined to be at once something passive, i.e. receiving impressions, and something active, in the form of a projection by which we return our impressions to the object. A. A. Long (‘Thinking and Sense-Perception in Empedocles: Mysticism or Materialism?’ [‘Thinking’], Classical Quarterly, ns 16 (1966), 256–76) argues that on Empedocles’ view fire does not issue forth from the eye, although an intraocular fire is required for visual perception, since vision occurs only when there is the right correspondence between internal and external fire. D. O’Brien (‘The E·ects of a Simile: Empedocles’ Theories of Seeing and Breathing’ [‘Simile’], Journal of Hellenic Studies, 90 (1970), 140–79 at 159) says that if we were to synthesize the two accounts, then the simplest method would be to suppose that fire leaves the eye in order to make room for equivalent e}uences to enter the eye from the outside. Finally, Sedley (‘Vision’, 25–6) suggests that, since on Empedocles’ view it is the reflective surface of the eye which is responsible for vision, fire must come out of the eye to mix with the water on its outer surface and to receive in its pores the e}uences of things. For a more detailed list of earlier interpretations of Empedocles’ theory of vision, cf. O’Brien, ‘Simile’, 157–60.

Created on 4 July 2005 at 14.33 hours page 28

28

Katerina Ierodiakonou

emit an overwhelming proportion of minute particles of fire, which reach our organ of vision and enter into the pores commensurate with them. These are the pores which Theophrastus refers to when he talks of the pores ‘of fire’ (το7 πυρς). They are not pores in something which is fiery, since they are at the surface of our eyes, where there is no fire; rather, they are the pores of the membranes which surround the eye, and into which the fine particles of fire fit, but not the coarser particles of water, nor for that matter the particles of earth or air. That is to say, these pores are of fire in the sense that they are pores for fire, i.e. for receiving particles of fire which are commensurate with them and emitted by the objects around us. Thus, the fiery particles of the white objects pass through the membranes of the eyes and move inside the eye; when there is a high proportion of fiery particles, we see the colour white. Similarly in the case of black objects, the fact that they are black means, according to Empedocles, that they emit an overwhelming proportion of minute particles of water, which reach our organ of vision and enter into the pores commensurate with them. These are the pores which Theophrastus refers to when he talks of the pores ‘of water’ (το7 8δατος). They are the pores of the membranes which surround the eye, and into which the water particles fit, but not the finer particles of fire, nor for that matter the particles of earth or air. That is to say, these pores are of water in the sense that they are pores for water, i.e. for receiving particles of water which are commensurate with them and emitted by the objects around us. Thus, the watery particles of the black objects pass through the membranes of the eyes and move inside the eye; when there is a high proportion of watery particles, we see the colour black. In this way, Empedocles believes, human beings can see the colours white and black. The fiery and watery particles emitted from objects reach our eyes and enter into their pores, being attracted by the fire and the water of which human eyes are composed.42 But do we see only black and white? Of course not, unless we are completely colour-blind. So is Theophrastus right when he accuses Empedocles of suggesting a theory on which we cannot explain our perception of other colours apart from black and white? 42 It should be clarified here that the pores of our eyes receive fiery and watery particles of a certain size which is di·erent from that of the fiery and watery particles which another sense-organ may receive into its pores; for instance, the watery particles which fit into the pores of our tongue are not of the same size as those watery particles which fit into the pores of our eyes.

Created on 4 July 2005 at 14.33 hours page 29

Empedocles on Colour and Colour Vision

29

Let us look more closely at Theophrastus’ objection (De sens. 17 = A 86): Moreover, [there are problems] in his treatment of the individual senses. For it turns out that recognition occurs by what is like [τ1 .µο 1ω]. For as to the organ of vision, if it is composed of fire and its opposite, it might be able to recognize the white and the black by their likes, but how will it recognize grey and the other colours, namely the mixed ones? For he explains it neither with the pores of fire nor with the pores of water nor with other pores combined from both. Yet we see these no less than the simple colours.

To defend Empedocles’ theory of colour perception from such an objection, it has been suggested that we should perhaps assume that, apart from the pores of fire and of water, there must also be pores of earth and air which are crucially involved in our visual perception.43 This is, in fact, how some scholars have understood the following Empedoclean fragment (B 109): With earth we see [9ππαµεν] earth, with water water, with aether divine aether, with fire destructive fire, love with love, and strife with baneful strife.

The verb ‘see’ (9ππαµεν), however, need not refer literally to vision; for it can refer, as has rightly been pointed out, more generally to perception, grasping, and understanding.44 Moreover, it would not make sense to talk about literal seeing in the case of Love and Strife (cf. B 17. 21). That is probably why Aristotle (DA 404B8– 15), in paraphrasing this fragment, substitutes the verbs ‘perceive’ (α:σθνεσθαι) and ‘know’ (γινσκειν).45 We therefore do not have to assume that B 109 deals specifically with vision, rather than perception, quite generally, or some kind of awareness; and it thus cannot be used as evidence in support of the view that pores of earth and air also play a role in Empedocles’ theory of vision.46 43 Verdenius, ‘Doctrine’, 155. Long (‘Thinking’, 261 and 264) claims that it is possible, perhaps even probable, that there are pores of earth and air involved in the operation of sight, but he also recognizes that there is no evidence to support this view. 44 O’Brien, ‘Simile’, 164; Sedley, ‘Vision’, 28. 45 This also explains why Theophrastus claims (De sens. 10 = A 86) that what we have as fragment B 109 continued with the following two lines ( = B 107): For all are constructed and fitted together out of these, and it is with these that they think and feel pleasure and pain. 46 Galen (De plac. Hipp. et Plat. v. 627 Kuhn • = CMG 5. 4. 1. 2. 462. 1–19) refers to

Created on 4 July 2005 at 14.33 hours page 30

30

Katerina Ierodiakonou

But do we need to assume that pores of earth and air are crucially involved in vision if we want to explain how, according to Empedocles, human beings are able to see all the di·erent colours? The introduction of the pores of earth and air has been considered necessary because Theophrastus stresses that Empedocles’ theory of vision is based on the principle that perception is of like by like. Theophrastus says this in the passage quoted above (De sens. 17 = A 86). He also makes the same claim at the beginning of his treatise, at which point he tries to classify into di·erent groups the views on perception expressed by the ancient philosophers before his time (De sens. 1–2 = A 86): Parmenides, Empedocles, and Plato [make sense-perception] a result of the like (τ1 .µο 1ω), and the Anaxagoreans and Heracliteans of the opposite . . . The others more or less omit [any account] of each of the individual senses; but Empedocles tries to reduce these too to [a process involving] likeness.

The idea, therefore, is that on Empedocles’ view seeing is a particular case of the general principle that perception is of like by like. That is to say, we can see white objects because their fiery e}uences fit into the pores of fire, and we can see black objects because their watery e}uences fit into the pores of water. If there are also pores of earth and air involved in the process of seeing, then the particles of earth and air emitted from objects around us can also be perceived by our eyes; and if the elements of earth and air are characterized by colours other than black and white, then our eyes can perceive all four basic colours, and hence all mixed colours. In other words, this interpretation of Empedocles’ theory of vision assumes that (1) Empedocles’ views on colour perception are based on the principle that perception is of like by like; and (2) the elements earth and air are coloured, and their colours together with the colours of fire and water constitute the basic colours from which all other colours can be derived. Both of these assumptions have been seriously contested. Scholars have long remarked that Empedocles’ theory about the commensurability of pores and e}uences does not have to be understood as involving some kind of likeness; for it may be the case that the first two lines of B 109 in order to justify the view that each of the four elements can be associated with a particular sense: namely, fire with sight, air with hearing, water with taste, earth with touch, and ‘vapour’ (%τµοειδς) with smell. But this interpretation is not plausible, for it does not take into consideration that, according to Empedocles, sight involves both the element of fire and the element of water.

Created on 4 July 2005 at 14.33 hours page 31

Empedocles on Colour and Colour Vision

31

something is commensurate with something else but not similar to it. However, the initial reaction of scholars was to conclude that Empedocles was not aware of this distinction and that that is why in his theory of vision he applies the principle that perception is of like by like in this way. It is only recently that some scholars have been more critical of Theophrastus’ historiographical methods, and in particular of his way of classifying and schematizing earlier theories of perception. Thus, it has been argued that it is Theophrastus, rather than Empedocles himself, who correlates the commensurability between the e}uences and the pores in perception with the principle that perception is of like by like.47 It might be suggested, therefore, that e}uences of particles of earth and air are emitted from the objects and are perceived by the pores of fire and water of our organ of vision, where ‘pores of fire’ is now understood in the sense of ‘fiery pores’, and the same for water; for even if these e}uences are not like the pores of fire and water, they may be said to be somehow commensurate with them. But there is no evidence to support the view that the perception of particles of earth and air plays any role in our visual perception, and for that matter in our colour perception. Most importantly, though, if we understand the pores of fire and water as pores for exclusively receiving fiery and watery particles respectively, and not as pores in something fiery and in something watery, as suggested, this view does not make much sense. One would rather think that, according to Empedocles, the e}uences of particles of earth and air emitted from the objects around us do not fit into the pores of our eyes; in other words, they are not perceived by our eyes. The suggestion that the e}uences of particles of earth and air are not visible is after all in agreement with what we have previously claimed concerning the colours of the elements earth and air, even though this claim has been made mainly on the basis of an argument ex silentio. For our ancient sources, at least those we can rely on, do not talk of the colours of the elements earth and air. And to say that earth and air are colourless precisely means that the effluences of particles of earth and air emitted from objects are not commensurate with the pores of our eyes, and thus invisible. But the fact that earth and air have no colour does not cause a problem, as we have already seen, in our understanding of the production of 47 Sedley, ‘Vision’, 26–31.

Created on 4 July 2005 at 14.33 hours page 32

32

Katerina Ierodiakonou

all other colours; for according to Empedocles, just by combining the white of fire and the black of water we may derive all colours. Have we thus managed to rebut Theophrastus’ objection? Even if we do not have to correlate Empedocles’ theory of vision with the principle that perception is of like by like, and even if we do not have to assume that the elements of earth and air are coloured, there is still a problem: how does it happen that the fitting together of the white e}uences to the pores of fire and the black e}uences to the pores of water gives us the impression of di·erent colours? Empedocles does not provide us with any information on this matter. And Theophrastus may be understood as complaining exactly about this, namely that Empedocles has no account of how we perceive mixed colours. Nevertheless, the following can be said and is in tune with Empedocles’ views on visual perception: the fiery and watery particles which are emitted in a certain proportion from an object, e.g. a blue object, pass through the pores of fire and water in the membranes of our eyes, and enter into the internal fire and water of our eyes; thus, our visual organ as a whole registers the proportion of fiery and watery particles in it and, depending on the kind of proportion, it has the impression of, for instance, a blue object. This account may need some refinement to accommodate testimony according to which, on Empedocles’ view, eyes di·er in their constitution and, correspondingly, in their perceptual powers. Interestingly enough in the present context, the di·erence in constitution is correlated with their di·erence in colour. Aristotle says that, according to Empedocles, dark eyes have more water than fire, and thus see better in the daytime, whereas bright eyes have more fire than water, and thus see better at night (GA 779B15–20 = A 91; cf. Philop. In DA 217. 10–25 Hayduck): To suppose, then, that grey-blue [eyes] are fiery, as Empedocles says, and that dark eyes have more water than fire, and that this is why the former do not see sharply in the daytime [namely the grey-blue ones], because of their lack of water, and that the latter do not see sharply at night because of their lack of fire, this is not a good theory, if indeed one must assign vision in all animals to water, not to fire.

And Theophrastus, too, reports that on Empedocles’ view the eyes of animals which have less fire, because of their structure, see better

Created on 4 July 2005 at 14.33 hours page 33

Empedocles on Colour and Colour Vision

33

in the daytime, whereas those with less water see better at night (De sens. 8 = A 86):48 [Eyes] are not [all] constructed in like fashion, and others from the opposite things, and some have the fire in the middle, some on the outside. That is why animals have sharper vision in the daytime, others at night—the ones with less fire by day (for their internal light is equalized by the external light), the ones [with less] of the opposite [see better] at night (for they too have their deficiency supplemented). And each kind has the opposite [characteristic] in the opposite conditions. For those who have too much fire have dim vision (for being further increased in the daytime it covers over and blocks up the pores of water), while for those [with too much] water this same [problem] occurs at night (for the fire is blocked by the water). for the one group until the water is dissipated by the external fire, while for the other group until the fire is dissipated by the air. For the opposite is the cure for each group. The [organ of vision] which is constructed with an equal amount of both [fire and water] is optimally blended and best.

That is to say, Empedocles is here presented as saying that animals with too much fire in their eyes need some water from the outside, i.e. from the darkness of the night, in order to produce the balance of fire and water which helps them to see better at night; on the other hand, animals with too much water in their eyes need some fire from the outside, i.e. from the daylight, in order to make up the balance of water and fire which helps them to see better in the daytime. As soon as we start talking about di·erences in the construction of the eyes of di·erent animals, or of di·erent human beings, and a corresponding di·erence in perception, the problem of subjectivity in perception might seem to be raised. But we should note that Empedocles is discussing a more limited issue, namely the conditions under which di·erent kinds of animal, or even di·erent human beings, see better than others, i.e. see the same colours, but more clearly or more reliably. There is no evidence that he is interested in more general problems, such as whether di·erent 48 It is interesting to note that Anaxagoras, too, is reported to have claimed, for di·erent reasons, that some animals, in fact most animals, see better in the daytime and some animals see better at night (Theophr. De sens. 27). On the relationship between Empedocles’ and Anaxagoras’ theories of vision, cf. Beare, Greek Theories, 38; D. O’Brien, ‘The Relation of Anaxagoras and Empedocles’, Journal of Hellenic Studies, 88 (1968), 93–113 at 109–13.

Created on 4 July 2005 at 14.33 hours page 34

34

Katerina Ierodiakonou

animals or di·erent human beings see the same colours or di·erent colours, or whether the colours they see are the colours the objects actually have, or whether objects in reality have any colours at all. To conclude, Empedocles holds that, when human beings perceive the world with their eyes, they primarily see colours, because colours are e}uences from the objects commensurate with the pores of the eyes. And we see the colour white because an overwhelming proportion of fiery particles are emitted from a white object and are commensurate with the pores of fire in our eyes; and we see the colour black because an overwhelming proportion of watery particles are emitted from a black object and are commensurate with the pores of water in our eyes. Finally, the way we see all other colours depends on the proportion of watery and fiery particles emitted from the objects around us. It is therefore the elements of fire and water that play a crucial role both in the fact that objects have the colours they have and in the fact that human beings perceive the colours of objects in this world; for it is their di·erent combinations that are responsible for the generation of the visual characteristics of the world as well as for the way it appears to us. Earth and air, on the other hand, are colourless; we need our other senses to become aware, for instance, of the density of the particles of earth or the rarity of the particles of air emitted by everything around us. And the information we get through all our senses about the various combinations of the four elements helps us, according to Empedocles, to perceive every single characteristic of the objects in this world, and through them the objects themselves. National Technical University of Athens

B I B L I O GR A P HY Barnes, J., review of M. R. Wright, Empedocles: The Extant Fragments, Edited with Introduction, Commentary, Concordance (New Haven, 1981), in Classical Review, ns 32 (1982), 191–6. Beare, J. I., Greek Theories of Elementary Cognition [Greek Theories] (Oxford, 1906). ‹ Bertrand, E., Etudes sur la peinture et la critique d’art dans l’antiquit‹e (Paris, 1893). Bignone, E., Empedocle: studio critico (Turin, 1916). Bollack, J., Emp‹edocle, vol. iii [Emp‹edocle] (Paris, 1969; repr. 1992).

Created on 4 July 2005 at 14.33 hours page 35

Empedocles on Colour and Colour Vision

35

Bruno, V. J., Form and Color in Greek Painting [Form and Color] (New York, 1977). Caston, V., ‘Empedocles’ Theory of Vision’ (diss., University of Texas; Austin 1985). Cherniss, H., Aristotle’s Criticism of Presocratic Philosophy [Criticism] (Baltimore, 1935). Cohen, A., The Alexander Mosaic: Stories of Victory and Defeat (Cambridge, 1997). Diels, H., Doxographi Graeci (Berlin, 1879). ‘Gorgias und Empedokles’ [‘Gorgias’], Sitzungsberichte der Preu¢ischen Akademie der Wissenschaften, 49 (1884), 343–68. and Kranz, W., Die Fragmente der Vorsokratiker, 6th edn. [DK] (3 vols.; Berlin, 1951). Inwood, B., The Poem of Empedocles: A Text and Translation and Introduction, rev. edn. (Toronto, 2001). Jex-Blake, K., and Sellers, E., The Elder Pliny’s Chapters on the History of Art [Elder Pliny] (London, 1896). Keuls, E., ‘Skiagraphia Once Again’, American Journal of Archaeology, 79 (1975), 1–16. Kirk, G. S., Raven, J. E., and Schofield, M., The Presocratic Philosophers, rev. edn. (Cambridge, 1983) [1957 edn. by Kirk and Raven]. Kranz, W., ‘Die a• ltesten Farbenlehren der Griechen’ [‘Farbenlehren’], Hermes, 47 (1917), 126–40. Lloyd, G. E. R., ‘The Hot and the Cold, the Dry and the Wet in Greek Philosophy’, Journal of Hellenic Studies, 84 (1964), 92–106. Polarity and Analogy [Polarity] (Cambridge, 1966). ‘Alcmaeon and the Early History of Dissection’, Sudho·s Archiv, 59 (1975), 113–47; repr. in Methods and Problems in Greek Science (Cambridge, 1991), 164–93. Long, A. A., ‘Thinking and Sense-Perception in Empedocles: Mysticism or Materialism?’ [‘Thinking’], Classical Quarterly, ns 16 (1966), 256–76. Longrigg, J., ‘Philosophy and Medicine: Some Early Interactions’ [‘Philosophy and Medicine’], Harvard Studies in Classical Philology, 67 (1963), 147–75. ‘The “Roots of All Things” ’ [‘Roots’], Isis, 67 (1976), 420–38. Millerd, C. E., On the Interpretation of Empedocles [Interpretation] (Chicago, 1908). Morelli, G., Italian Masters in German Galleries, trans. L. M. Richter (London, 1883). Mourelatos, A. P. D., ‘Quality, Structure, and Emergence in Later PreSocratic Philosophy’ [‘Quality’], in J. J. Cleary (ed.), Proceedings of the Boston Area Colloquium in Ancient Philosophy, 2 (Lanham, Md., 1987), 127–94.

Created on 4 July 2005 at 14.33 hours page 36

36

Katerina Ierodiakonou

O’Brien, D., ‘The Relation of Anaxagoras and Empedocles’, Journal of Hellenic Studies, 88 (1968), 93–113. ‘The E·ects of a Simile: Empedocles’ Theories of Seeing and Breathing’ [‘Simile’], Journal of Hellenic Studies, 90 (1970), 140–79. Osborne, H., ‘Colour Concepts of the Ancient Greeks’, British Journal of Aesthetics, 8 (1968), 269–83. Pemberton, E. G., ‘A Note on Skiagraphia’, American Journal of Archaeology, 80 (1976), 82–4. Pfuhl, E., ‘Apollodoros . Σκιαγρφος’, Jahrbuch des Arch•aologischen Instituts, 25 (1910), 12–28. ‘Skiagrafia’, Jahrbuch des Arch•aologischen Instituts, 27 (1912), 227– 31. Platnauer, M., ‘Greek Colour-Perception’, Classical Quarterly, 15 (1921), 153–62. Pollitt, J. J., The Ancient View of Greek Art [Ancient View] (New Haven, 1974). ‘Περ χρωµτων: What Ancient Greek Painters Thought about Colors’, in M. A. Tiberios and D. S. Tsiafakis (eds.), Color in Ancient Greece (Thessaloniki, 2002), 1–8. • Prantl, K., Aristoteles: Uber die Farben (Munich, 1849). Robertson, M., Greek Painting [Painting] (Geneva, 1959). A History of Greek Art [Art] (Cambridge, Mass., 1975). Ross, G. R. T., Aristotle: De Sensu and De Memoria (Cambridge, 1906). Rouveret, A., Histoire et imaginaire de la peinture ancienne (Rome, 1989). Scheibler, I., ‘Die “vier Farben” der griechischen Malerei’ [‘Farben’], Antike Kunst, 17 (1974), 92–102. Sedley, D., ‘Empedocles’ Theory of Vision and Theophrastus’ De sensibus’ [‘Vision’], in W. W. Fortenbaugh and D. Gutas (eds.), Theophrastus: His Psychological, Doxographical and Scientific Writings (New Brunswick, 1992) 20–31. Sider, D., ‘Empedocles B 96 (462 Bollack) and the Poetry of Adhesion’, Mnemosyne, 37 (1984), 14–24. Siegel, E., ‘Theories of Vision and Colour Perception of Empedocles and Democritus: Some Similarities to the Modern Approach’ [‘Theories of Vision’], Bulletin of the History of Medicine, 33 (1959), 145–59. Solmsen, F., ‘Tissues and the Soul: Philosophical Contributions to Physiology’, Philosophical Review, 59 (1950), 435–68. Stratton, G. M., Theophrastus and the Greek Physiological Psychology before Aristotle (London, 1917). Tr‹epanier, S., ‘Empedocles on the Ultimate Symmetry of the World’, Oxford Studies in Ancient Philosophy, 24 (2003), 1–57. Verdenius, W. J., ‘Empedocles’ Doctrine of Sight’ [‘Doctrine’], in Studia Varia Carolo Gulielmo Vollgra· Oblata (Amsterdam, 1948), 55–64.

Created on 4 July 2005 at 14.33 hours page 37

Empedocles on Colour and Colour Vision

37

Wellmann, M., Die Fragmente der sikelischen A•rzte (Berlin, 1901). Wright, M. R., Empedocles: The Extant Fragments, Edited with Introduction, Commentary, Concordance [Extant Fragments] (New Haven, 1981; repr. 1995).

Created on 4 July 2005 at 14.33 hours page 38

page 38 is blank

Created on 4 July 2005 at 14.33 hours page 39

XENOP HON AT HIS MOS T S OCRATI C (MEMORABILIA . ) DAVID M. JOHNSON

1. Introduction fl ’s Socratic writings are both more conventional and more derivative than those of Plato.1 But the di·erences are not due solely to Xenophon’s literary and philosophical shortcomings. Xenophon’s Socrates is more conventional than Plato’s because Xenophon’s intentions were di·erent. He was more consistently concerned with defending Socrates, and he appears to have written for a wider audience than Plato. The better to defend Socrates, Xenophon argues that he was not only innocent but beneficent, and as Xenophon’s Socrates benefits others by teaching, he must be more didactic than Plato’s Socrates.2 In the Memorabilia Xenophon therefore takes great pains to show Socrates conversing with and benefiting interlocutors of many di·erent types.3 Plato’s ã David M. Johnson 2005 1 Conventional: most famously in Irwin’s slashing review of L. Strauss, Xenophon’s Socrates (Ithaca, NY, 1972), in which Irwin’s faint praise of Xenophon is particularly damning: ‘Xenophon quite closely resembles a familiar British figure— the retired general, staunch Tory and Anglican, firm defender of the Establishment in Church and State, and at the same time a reflective man with ambitions to write edifying literature’ (T. Irwin, review of Strauss, in Philosophical Review, 83 (1974), 409–13 at 410). Derivative: most damagingly K. Jo•el, Der echte und der Xenophontische Sokrates [Echter Sokrates] (2 vols. in 3; Berlin, 1893–1901), and H. Maier, Sokrates (Tubingen, 1913). Both Jo•el and Maier see Antisthenes as the source • of anything worthwhile in Xenophon. For a rebuttal of their view, see J. Cooper, ‘Notes on Xenophon’s Socrates’ [‘Notes’], in Reason and Emotion: Essays on Ancient Moral Psychology and Ethical Theory (Princeton, 1999), 3–28 at 3–10. 2 For discussion of Xenophon’s Socrates as a teacher, see Cooper, ‘Notes’, 5, and D. Morrison, ‘Xenophon’s Socrates as a Teacher’ [‘Socrates as Teacher’], in. P. Vander Waerdt (ed.), The Socratic Movement [Socratic Movement] (Ithaca, NY, 1994), 181–208. 3 Like Plato, Xenophon sometimes chooses interlocutors who are ironically inappropriate exemplars of the virtue under discussion. In Plato’s case, consider Nicias on courage, Charmides and Critias on s»ophrosyn»e, and Meno on virtue. Xenophon

Created on 4 July 2005 at 14.33 hours page 40

40

David M. Johnson

Socrates, on the other hand, seems to have benefited rather few of his interlocutors. As Brickhouse and Smith must admit, ‘Students are often depicted as being confused by and annoyed at him; rarely if ever can they be said to have really learned anything’.4 This is not terribly surprising, given that neither of the two favourite tools of the Platonic Socrates, irony and the elenchus, is obviously and unambiguously beneficial. Socrates’ complex irony may have led some interlocutors to deeper reflection, and it has certainly had that e·ect on many readers, but it leaves others irritated and confused. The elenchus does e·ectively remove false beliefs, cleansing the soul and readying it for further education (Soph. 230 a–231 b; cf. Meno 84 b–c). But it also leaves the young, deprived of their lawful if ill-grounded traditional beliefs, open to corruption by flatterers who turn them towards a life of lawless pleasure-seeking (Rep. 7, 537 d–539 e). Xenophon therefore chose, more often than not, to steer clear of overt irony and of the elenchus, producing a Socrates who strikes readers of Plato as rather bland. Xenophon’s decision to present a less ironic and less elenctic Socrates shows that his Socratic works are not entirely derivative. But he evidently completed his Socratic works after many other such works were already in circulation, and he was certainly highly indebted to Plato and other Socratics.5 This debt, though, reflects a praiseworthy decision to respond to and make use of such prior work. Rather than insisting on absolute originality or absolute fidelity to the elusive histori-

has Socrates laud Callias’ relationship with Autolycus, an a·air that would provide comic material for Eupolis’ Autolycus (on Callias see D. Nails, The People of Plato (Indianapolis, 2002), 68–74). And Ischomachus and his wife, if she is indeed the Chrysilla who had a notorious a·air with the same Callias, are decidedly odd choices for an ideal couple in Oeconomicus (on which see now G. Danzig, ‘Why Xenophon’s Socrates Was Not a Farmer: Xenophon’s Oeconomicus as a Philosophical Dialogue’, Greece @ Rome, 50 (2003), 57–76). In such cases it is presumably the others present, especially the silent bystanders who function in large part as stand-ins for the reader, who are meant to benefit from the conversation. 4 Plato’s Socrates (Oxford, 1994), 4 (emphasis original). 5 The surest external reference in Xenophon’s Socratica comes at Mem. 3. 5. 4, which seems to allude to the strategic situation after the Spartan defeat at Leuctra in 371. Xenophon may have begun his Socratic works much earlier than this, of course, but recent scholarship stresses the unity even of his least unified Socratic work, the Memorabilia: V. Gray, The Framing of Socrates: The Literary Interpretation of Xenophon’s Memorabilia [Framing] (Stuttgart, 1998); L.-A. Dorion, ‘Introduction’, in M. Bandini and L.-A. Dorion (eds.), X‹enophon: M‹emorables, vol. i (Paris, 2000), vii–cclii at clxxxiii–ccxl.

Created on 4 July 2005 at 14.33 hours page 41

Xenophon at his Most Socratic

41

cal Socrates, we ought to count Xenophon’s complex responses to other writers as a strength. Of course, even if we can explain the di·erences between Xenophon’s Socrates and Plato’s Socrates without the assumption that Xenophon was too shallow to understand Socrates, this does not in itself show that Xenophon did understand Socrates—the Socrates, that is, of Plato’s early dialogues, whether or not we take him to be an accurate portrayal of the historical Socrates. Our best opportunity to test Xenophon’s understanding of Socrates comes on those occasions where Xenophon’s Socrates does sound more like the Socrates we are familiar with from Plato’s early dialogues. Xenophon’s Socrates is sometimes ironic, but irony is notoriously di¶cult to pin down.6 The Socratic elenchus, while itself complex enough, is a bit more straightforward, and so Xenophon’s use of it may provide us with a more practicable comparison. In an important programmatic passage (Mem. 1. 4), Xenophon comments on Socrates’ use of the elenchus. Unfortunately there is considerable debate about the meaning of the text: And if certain people believe, as some write and say about him on the basis of conjecture, that he was the best at turning people towards virtue [προτρψασθαι], but not capable of leading them to it [προαγαγε ν], let them consider not only what he said to those who thought they knew everything when, for the sake of correction [κολαστηρου νεκα], he refuted them by questioning [ρωτν λεγχεν], but also what he said as he spent his days with those who spent their time with him [λγων συνηµρευε το ς συνδιατρβουσι], and let them then judge whether he was capable of making those with him better.

The argument is about how many types of Socratic discourse Xenophon has in mind. Slings has argued that there are three types of discourse under consideration, a form of explicit protreptic, which the objectors have been considering but which is not exemplified in the Memorabilia, and two more constructive approaches, one elenchus of know-it-alls and one more didactic conversations with companions.7 But Dorion has vigorously reasserted the more 6 For irony in Xenophon, see D. Morrison, ‘On Professor Vlastos’ Xenophon’ [‘Vlastos’ Xenophon’], Ancient Philosophy, 7 (1987), 9–22 at 10–14, and my exchange with Vivienne Gray in Ancient Philosophy, 24 (2004). In my view, given his choice of characters, both Xenophon’s Symposium and his Oeconomicus are fundamentally ironic: see n. 3 above. 7 S. Slings, Plato: Clitophon [Clitophon] (Cambridge, 1999), 45 n. 86, 77–82. The explicit protreptic is of the sort parodied at Clit. 407 a–e.

Created on 4 July 2005 at 14.33 hours page 42

42

David M. Johnson

economical interpretation in which there are but two types of discourse, protreptic elenchus of know-it-alls, which the objectors have been considering, and daily conversations leading to virtue, which the objectors need to consider.8 While Slings’s view leaves open the possibility of a constructive role for the elenchus, Dorion argues that Xenophon here rejects any constructive role for the elenchus of the sort Vlastos posited for Plato’s early dialogues. Slings and Dorion agree that the passage is itself ambiguous, and that it must be read in the light of Xenophon’s practice in the rest of the Memorabilia. But they disagree on the nature of what follows, particularly the next few chapters, where we would expect Xenophon to illustrate what he has in mind here: Slings regards 1. 4 and 1. 6 as elenctic, while Dorion does not. This is understandable enough, as both 1. 4 and 1. 6 contain elements of both sorts of Socratic discourse. This is particularly clear in the nature of the interlocutors. Immediately after the programmatic passage, Xenophon introduces Socrates’ companion Aristodemus,9 who does not worship the gods and mocks those who do. Aristodemus seems to be neither simply a companion of Socrates nor simply a know-itall, but both, and Socrates’ conversation with him is both refutative and didactic. Socrates first argues that the gods have designed the world and then, when Aristodemus argues that the gods are too important to worry about humans, that they are in fact concerned about us. Socrates asks Aristodemus numerous questions, some but not all of which are clearly rhetorical. At the end of the conversation Xenophon says that by saying such things Socrates led his companions to abstain from wrongdoing out of respect for the gods, even when no human would notice them doing wrong. He does not say that Aristodemus learnt his lesson on this occasion, but that such talk, presumably on a number of occasions, made Socrates’ companions more pious. Then, after a didactic speech directed only at a vague audience (1. 5), Xenophon has Socrates refute the claim of ‘Antiphon the sophist’ that Socrates’ poverty, failure to take fees, and unwillingness to participate in politics shows that he is not wise (1. 6). Antiphon, whatever his precise identity or identities, 8 Dorion, ‘Introduction’, cxliv. So also Gray, Framing, 13–15, 74–82. 9 Xenophon does not explicitly tell us that Aristodemus was a companion of Socrates, but as he is the same small Aristodemus who reported the speeches of Plato’s Symposium and was obsessed with Socrates, he no doubt expected readers to regard him as such.

Created on 4 July 2005 at 14.33 hours page 43

Xenophon at his Most Socratic

43

is certainly not a companion of Socrates, and he is something of a know-it-all. Socrates’ conversation with him consists mainly of a string of rhetorical questions, which refute Antiphon’s charges while presenting Socrates’ own position. Xenophon tells us that the conversation benefited those who listened to it, rather than Antiphon himself. So are these conversations elenchus of know-italls or didactic conversations with companions? They are both. Xenophon says not that we are to forget about refutation of know-it-alls and consider only the daily conversations with companions, but that we must consider both together. There are, then, three types of Socratic discourse, but not the three that Slings has in mind. The first is elenchus by itself, which, as Dorion argues, is what the objectors evidently have considered and found wanting.10 The second is non-elenctic daily conversation, which does make up a large share of the Memorabilia, starting with 1. 5. But the third is a combination of the elenchus with daily conversation, or, more precisely, elenchus viewed not in isolation but as part of an overall educational strategy. It is not the case, then, that the elenchus is utterly useless, but rather that it does not, in isolation, lead interlocutors to virtue. Xenophon, unlike Plato, often provides an audience for Socrates’ conversations, and it is often these audience members who are said to benefit, as in 1. 4 and 1. 6.11 Xenophon’s point, then, is one in keeping with the practice of Plato, where Socrates’ interlocutors routinely fail to benefit, but the elenctic conversation is beneficial to those who mull over the results of multiple conversations, including Socrates himself and Plato’s readers. The elenchus can also benefit the individual who has been refuted, but only if he takes it to heart, as does Euthydemus in Memorabilia 4. 2, as we shall see. By providing many of his Socratic refutations with audiences, Xenophon makes explicit something left implicit in Plato, that the benefits of the elenchus are often clearest for others rather than for the one who undergoes it. If Xenophon recognizes that the elenchus can play a constructive 10 If I am correct, Slings too closely assimilates the error Xenophon has in mind with the objection at the outset of the Clitophon. In both cases the problem is that Socrates succeeds only in turning people towards virtue rather than in leading them to it, but for Clitophon, according to Slings, Socrates’ inadequate method consists of explicit protreptic. Xenophon has no objection to explicit exhortation towards virtue, but rather towards elenctic protreptic, which may turn some interlocutors towards virtue, but, taken in isolation, risks leading others astray. 11 Dorion, ‘Introduction’, cxlvii–cxlviii.

Created on 4 July 2005 at 14.33 hours page 44

44

David M. Johnson

role, this puts him closer to the constructive view of the elenchus that most scholars attribute to Plato. This is not to say that the elenchus is as central for Xenophon’s Socrates as it is for Plato’s; Xenophon’s statement here and his practice throughout the Memorabilia clearly show that he believed that the elenchus was insufficient, and that more didactic forms of instruction were valuable. And more precise comparison with Plato is complicated by the fact that Plato’s portrayal of the Socratic elenchus has become increasingly controversial in recent years, with Vlastos’s constructive understanding of it coming in for much criticism.12 Still, the least that can be said is that Socratic refutations ought to be distinct from refutations of the eristic sort performed by the brothers Dionysodorus and Euthydemus in Plato’s Euthydemus. There may be occasions on which Xenophon’s Socrates refutes interlocutors simply to humiliate them—as with the tyrants Critias and Charicles in 1. 2. 29–38, who, unlike the innocent Cleinias of the Euthydemus, fully deserve this treatment. With less reprehensible interlocutors, we expect more from Socrates. If Socratic refutation does not always lead to the discovery of new truths, it ought at least to be compatible with truths otherwise discovered. If the elenchus cannot bring interlocutors all the way to virtue, it ought to get them moving in the right direction, rather than simply showing them the error of their ways. When we turn to consider the two clearest cases of Socratic elenchus in the Memorabilia, we find, oddly enough, that Socrates plays no role in one of them, Alcibiades’ discussion about law with Pericles (1. 2. 40–6). A youthful and mischievous Alcibiades completely embarrasses Pericles, and this exchange is usually taken to be an example of the destructive elenchus at its worst.13 But if the conversation is entirely destructive, it is also singularly inept, as it functions to give Socrates’ adversaries more evidence for the ill e·ects of Socratic education.14 I have argued elsewhere that not only the elenctic form but also the philosophical substance of Alcibiades’ conversation with Pericles complements what Xenophon’s 12 G. Scott (ed.), Does Socrates Have a Method? Rethinking the Elenchus in Plato’s Dialogues and Beyond (University Park, Penn., 2002) presents a series of papers both inspired by and critical of G. Vlastos, ‘The Socratic Elenchus: Method is All’, in Socratic Studies (Cambridge, 1994), 1–37; see also Slings, Clitophon, 127–64. 13 Gray, Framing, 50–1; Dorion, ‘Introduction’, clx–clxix. 14 As Dorion must admit (‘Introduction’, clxviii).

Created on 4 July 2005 at 14.33 hours page 45

Xenophon at his Most Socratic

45

Socrates says about law in Memorabilia 4. 4.15 If this is correct, then Alcibiades’ encounter with Pericles, while certainly an example of the malicious use of an elenchus, intended to benefit neither the refuter nor the refutee, is not simply eristic, but may prove beneficial for Xenophon’s readers. But as Alcibiades’ refutation of Pericles is controversial, and does not directly involve Socrates, I turn here to what Xenophon clearly meant to be a positive use of the elenchus, Memorabilia 4. 2.16 Here Socrates humbles his young interlocutor, the beautiful Euthydemus, who had thought that his collection of books su¶ced to make him wise. (He is, of course, to be distinguished from the eristic Euthydemus of Plato’s dialogue of that name.) The discussion raises a number of matters that are of central concern to the Socrates of Plato’s early dialogues: justice, self-knowledge, wisdom, and happiness. My claim is that this most Socratic of Xenophon’s Socratic recollections shows considerable philosophical and literary sophistication. Xenophon does not merely have Socrates humiliate Euthydemus by any means necessary in order to pave the way for the more positive instruction promised at the end of this chapter. He also presents crucial elements of an introduction to Socratic philosophy. This claim will probably seem rather extreme, as on a first impression 4. 2 appears both highly derivative and slipshod, as it did long ago to Jo•el and Maier and still does to Kahn.17 We appear to have abbreviated and sloppy treatments of a great number of topics, jumbled together in no obvious order. It has long been recognized that the discussion of justice in 4. 2 is related to those in Dissoi Logoi 3, Republic 1, and the pseudo-Platonic On Justice, and that the discussion of voluntary wrongdoing is clearly related to that in the Hippias Minor. There are parallels between the discussion of happiness and wisdom and that in the Euthydemus. There are also broader structural parallels with the Alcibiades of Aeschines

15 ‘Xenophon’s Socrates on Law and Justice’ [‘Law and Justice’], Ancient Philosophy, 23 (2003), 255–81 at 277–9. 16 Other examples of the Xenophontic Socrates refuting his interlocutors include Sym. 5, Oec. 1. 5–15, and Mem. 2. 1. 1–20 (cf. 3. 8. 1). 17 Jo•el, Echter Sokrates, i. 383–424; Maier, Sokrates, 53–7; C. Kahn, Plato and the Socratic Dialogue: The Philosophical Use of a Literary Form [Socratic Dialogue] (Cambridge, 1996), 396–7. For more positive recent assessments, see Morrison, ‘Socrates as Teacher’, and Dorion, ‘Introduction’, clxix–clxxxii.

Created on 4 July 2005 at 14.33 hours page 46

46

David M. Johnson

of Sphettus and the Platonic Alcibiades I.18 While there are many complications here, including the uncertain authorship and dating of the Dissoi Logoi and the Alcibiades I, in most cases the influence probably flowed to Xenophon rather than from him.19 And a quick comparison between Memorabilia 4. 2 and the parallel passages is not favourable to Xenophon. His discussion of justice is far less complex than that of Republic 1. His discussion of voluntary wrongdoing fails to leave the escape clause which Plato cunningly leaves in the Hippias Minor, that one who is voluntarily unjust, if there is any such person, will be more just than one who is involuntarily unjust. Xenophon’s Socrates rejects even the fundamental principles that wisdom and happiness are unequivocal goods. If Xenophon has based his conversation in some part on the Alcibiades dialogues, his decision to replace the exciting Alcibiades with the cipher Euthydemus hardly seems to be an improvement. Thus our passage would appear to show that on the rare occasions when Xenophon attempted to show Socrates doing philosophy he made a mess of things, despite utilizing a number of more authentically Socratic sources. My reading of 4. 2 is meant to replace, or at least supplement, traditional source criticism with a sympathetic reading of the passage in its own right.

2. Euthydemus Let us begin with one aspect of 4. 2 that is not derivative, Xenophon’s decision to make Euthydemus Socrates’ interlocutor. Eu18 H. Dittmar, Aischines von Sphettos: Studien zur Literaturgeschichte der Sokratiker (Bern, 1912), 121–52. 19 Of the works mentioned, only On Justice probably post-dates Xenophon. On it see C. Muller, Die Kurzdialoge der Appendix Platonica [Kurzdialoge] (Munich, • 1975), 129–91. Muller (150–5) is rightly sceptical of previous attempts at source • criticism of Mem. 4. 2. There seems to be a trend towards a¶rming the authenticity of the Alcibiades I, although critics remain (D. Gribble, Alcibiades and Athens: A Study in Literary Presentation [Alcibiades] (Oxford, 1999), 260–2). I treat it as an authentic and relatively early work by Plato in ‘God as the True Self: Plato’s Alcibiades I’ [‘True Self’], Ancient Philosophy, 19 (1999), 1–19; so does J. Annas, ‘Self-Knowledge in Early Plato’ [‘Self-Knowledge’], in D. O’Meara (ed.), Platonic Investigations (Washington, 1985), 111–38. N. Denyer (Plato: Alcibiades (Cambridge, 2001), 11–26) would make it authentic but perhaps too late (the early 350s) to have influenced Xenophon. I follow T. Robinson (Contrasting Arguments: An Edition of the Dissoi Logoi (New York, 1979), 34–41) in accepting the traditional date for the Dissoi Logoi, c.400.

Created on 4 July 2005 at 14.33 hours page 47

Xenophon at his Most Socratic

47

thydemus is indeed no Alcibiades, but he is not a mere cipher lacking Alcibiades’ charms: he has, as we shall see, a character of his own, which well illustrates Xenophon’s chosen themes. At the beginning of book 4 of the Memorabilia Xenophon explains the different approaches Socrates used with those who had the potential for virtue and philosophy but were not interested in education.20 Some simply trust in their good nature. But those with good potential need a good education most of all, for, if poorly raised, they will turn out to be the worst, just as the most vigorous horses or dogs, which can become the most useful if properly trained, are the most useless if they are not properly trained. Others trust in their wealth, but wealth is useless if one does not know how to use it. Xenophon’s description of the good natures and the risks they pose closely resembles Plato’s description of similar types in the Republic (6, 491 a–495 c) and Alcibiades I (103 b–104 c), though the analogy with animals is a typically Xenophontic touch. It is often thought that Alcibiades is the paradigm of the good, philosophical nature; he could also serve as an embodiment of the rich type.21 The idea that Xenophon has Alcibiades in mind in 4. 1 is supported by Xenophon’s similar description of the corruption of Alcibiades in his defence of Socrates early in the Memorabilia (1. 2. 12–16, 24–5). But Xenophon barely mentions Alcibiades outside of his defence of Socrates in the first two chapters of the Memorabilia, and he seems to avoid mentioning him even when his sources did.22 Xeno20 With Morrison (‘Socrates as Teacher’, 185 n. 7) and against Strauss (Xenophon’s Socrates, 94, 100), I take Euthydemus (like all the other types mentioned in 4. 1) to be a young man with a good nature. Xenophon explicitly attributes good natures only to those with whom Socrates said, jokingly, that he was in love (4. 1. 2). There is nothing overtly erotic about Socrates’ conversation with Euthydemus, but the young man introduced as ‘Euthydemus the kalos’ was apparently one of Socrates’ so-called beloveds, as Xenophon’s readers would have known (Plato, Sym. 222 a 8–b 4; see n. 23 below). Euthydemus is also like the good natures in wanting to acquire the virtue that allows one to run cities and households and to benefit both others and himself (4. 2. 11Ü4. 1. 2). While it is true that Euthydemus, unlike the other troubled good natures, is not completely hostile to education, as he has collected a set of books, he is uninterested in furthering his education and in this resembles the good natures. It is important to Strauss that Euthydemus not be a good nature, for this allows book 4 to remain esoteric, or rather exoteric, i.e. addressed to the many rather than fully revealing Socrates’ views. 21 On Alcibiades and the philosophical nature, see Gribble, Alcibiades, 219. For Alcibiades’ wealth, see Alc. I 104 b–c, where Socrates’ claim that Alcibiades does not pride himself on his wealth is clearly enough ironic. 22 Alcibiades is mentioned outside of the direct defence of Mem. 1. 1–2 only in 1. 3, and there only as the father of a son too beautiful to be prudently kissed. Xenophon

Created on 4 July 2005 at 14.33 hours page 48

48

David M. Johnson

phon apparently considered it important to distance Socrates from Alcibiades. Euthydemus is a good choice in part because he is a less controversial version of Alcibiades. We learn elsewhere in Xenophon that Socrates sharply rebuked Critias for his lust for Euthydemus (Mem. 1. 2. 29–30). In Plato’s Symposium (222 a 8–b 4) Alcibiades tells us that Euthydemus was one of those who, like Alcibiades himself, Charmides, and many others, was duped by Socrates, who appeared to be their lover but ended up being their beloved. Euthydemus’ beauty, youth, and ambition, as well as his relationship with Socrates, make him resemble Alcibiades. He can thus play a similar role to that played by Alcibiades elsewhere, but without involving Socrates with such a controversial interlocutor. Xenophon, or at least many of his readers, would probably have considered it irresponsible had Socrates discussed justice in the aporetic way he does in 4. 2 with Alcibiades. But Euthydemus is not only a safer Alcibiades. Unlike those who scorn education and trust in their natures or their riches, Euthydemus does not entirely reject education, but rather thinks that he has already educated himself by amassing, and presumably reading, a collection of books written by the poets and sophists of the highest repute (4. 2. 1). Yet he is anxious to avoid associating with any teacher in public, presumably because he worries that publicly associating with a sophist, among whose number Socrates would be counted, could stand in the way of his political ambitions. His books are perhaps a private enough matter to escape censure, and study of the poets, at any rate, was traditional enough. We should also keep in mind that Socrates’ interest in the beautiful Euthydemus has erotic overtones, even if these are largely suppressed by Xenophon, and that Euthydemus would be expected to play hard to get.23 seems to have made use of conversations between Socrates and Alcibiades in Mem. 3. 6 (compare Aeschines’ Alcibiades, especially for the mention of Themistocles, and Alc. I 105 b–c and 123 d); in Mem. 3. 7 (compare Alc. I 114 b 6–115 a 1); and, as we have seen, in Mem. 4. 1. 2 (compare Alc. I 131 c and Sym. 210 b–c). One can go too far in postulating an ‘Alcibiades literature’ (Gribble, Alcibiades, 215 n. 6), and it is certainly true that here as elsewhere Xenophon has done much more than change the name of the character in his source. The least that can be said is that he chose to pass up on many an opportunity to make use of Alcibiades as an interlocutor. 23 Xenophon’s language at 4. 2. 2 may allude to the erotic situation. Someone asks whether Themistocles surpassed his fellow citizens thanks to intercourse (συνουσα)

Created on 4 July 2005 at 14.33 hours page 49

Xenophon at his Most Socratic

49

Euthydemus’ attitude thus makes dramatic sense. And it mirrors the attitude Xenophon may have expected from many of his readers at this point in the Memorabilia. Such readers would have been convinced that Socrates was innocent and even that he benefited those he dealt with. But many may have wished to leave it at that, rather than moving on to the more advanced Socratic teachings Xenophon would introduce in the fourth book. Such readers need to be goaded, as Euthydemus is here, to admit that they remain fundamentally ignorant of the most important matters, even if, like Euthydemus, they have done a fair amount of reading already. The refutation of Euthydemus, then, marks the transition between the part of the Memorabilia that is more strictly apologetic and the part that attempts to do more. If Euthydemus the reader is a stand-in for the reader of the Memorabilia, it is not surprising that he goes on to be the interlocutor for most of the rest of the work, nor that his character, once it has been developed enough to make the connection with the reader, becomes increasingly vague. Muller has argued that the interlocutors of the • short pseudo-Platonic dialogues are usually left anonymous so that readers may more readily identify with them, something which is rather di¶cult to do if Alcibiades is the interlocutor.24 But it is di¶cult to identify with complete non-entities like the nameless interlocutor of On Justice. Slings argues that Plato alone among the Socratics realized that the explicit protreptic of works like the Alcibiades I fails because readers do not identify the interlocutor’s failings with their own.25 Plato thus turned to the implicit and more constructive protreptic of the elenctic dialogues. But in our passage the argument consists largely of refutation rather than explicit exhortation. And because Euthydemus’ characterization is largely limited to traits which connect him to readers, readers cannot escape by assuming that they do not share Euthydemus’ faults or do not need to heed arguments directed solely at him. Xenophon also uses Socrates’ intellectual seduction of Euthydemus to show Socrates at work as a teacher. Socrates charmingly with some wise man; συνουσα can refer to sexual intercourse (LSJ s.v. I.4) as well as to the association of a student with a teacher. Socrates gives the answer he does because he wishes to stir up (κινε ν) Euthydemus; but κινε ν can also refer to sexual intercourse (LSJ s.v. II.4). 24 Muller, Kurzdialoge, 322–3. •

25 Slings, Clitophon, 164.

Created on 4 July 2005 at 14.33 hours page 50

50

David M. Johnson

mocks Euthydemus by comparing him to a candidate for surgeon general who o·ered this as his campaign speech (4. 2. 5): Gentlemen of Athens, I have never learnt the art of medicine from anyone, nor sought out any doctor for a teacher. For I have continually been on my guard not only against learning anything from the doctors, but against seeming to have learnt this art. Still, give me the doctor’s work: I’ll try to learn by taking my chances on you.

This is only one stage in an elaborate trial of Euthydemus, the most trying step in which is the refutation proper that makes up most of the chapter. By the end of the conversation Euthydemus is despondent, but, unlike many, he returns to Socrates, who then clearly reveals his views. Morrison has argued that Socrates’ encounter with Euthydemus shows that Xenophon’s Socrates was a responsible teacher who thoroughly vetted his would-be students. Only those whose moral and intellectual virtues were similar to those of Socrates himself would return to Socrates after such a humiliating encounter with him.26 But students with a strong desire to use Socratic tricks to their own evil ends may be just as willing to undergo Socratic refutation as those with a genuine desire for virtue. What Socrates selects for is ability and ambition. This limits potential damage by turning away those who would only be confused and thereby corrupted by him, and thus partially meets the objections to early acquaintance with dialectic that Plato has Socrates describe in the Republic. But Socrates cannot select for character, as it is character that he is to help form. What finally gets Euthydemus is flattery: Socrates praises his collection of books. Xenophon’s readers may have found a refutation of youthful scholarly pretension more pleasing than yet another proof that an aristocratic youth was not educated enough. But the specific nature of Euthydemus’ education, book learning, also provides his conversation with Socrates with an implicit theme, the limitations of writing. Plato’s discussion of Socrates’ views on the 26 ‘Socrates as Teacher’, 189. Dorion (‘Introduction’, clxxx–clxxxi) argues that the elenchus need only be employed with those who pride themselves on wisdom, not on those who pride themselves on their natures or their wealth. But, given the paradigmatic role Euthydemus plays in book 4 and Xenophon’s remarks about the many who underwent this process at 4. 2. 40, I would follow Morrison in making the elenchus part of Socrates’ regular procedure. Still less, as will become clear, would I follow Dorion (‘Introduction’, clxxvii–clxxviii) in arguing that the elenchus is only a tool of selection, not a part of Euthydemus’ education.

Created on 4 July 2005 at 14.33 hours page 51

Xenophon at his Most Socratic

51

written word are well known. Socrates himself wrote nothing,27 and in the Phaedrus Plato has Socrates explain the limitations of writing (274 b–277 a; cf. Epist. VII 341 b–344 c). The essential failing of writing is that it is not interactive. But this need not mean that written texts are worthless, so long as they are treated as playful goads to active reflection rather than as repositories of wisdom. And the practice of Plato’s Socrates shows that he is an acute if sometimes perverse interpreter of texts, as in his deconstruction of Simonides in the Protagoras and Republic 1, his use of Homer to humiliate Ion, and his close reading of the Iliad in the Hippias Minor. Xenophon also depicts Socrates’ interest in literature. Xenophon rejects the charge that Socrates abused poetry to make antidemocratic points (1. 2. 56–61) by presenting a less antidemocratic Socratic interpretation of those verses. He has Socrates quote poetry (as at Mem. 1. 3. 3, 2. 1. 20) and, at considerable length, Prodicus’ prose account of the choice of Heracles (Mem. 2. 1. 21–33). And in one passage Xenophon’s Socrates speaks highly of the value of literature (Mem. 1. 6. 14): And I go through the treasuries of the wise men of old, which they left behind written in their books, reading them together with my friends; and if we see anything good, we pick it out for ourselves.

Xenophon’s Socrates nowhere speaks explicitly of the limitations of writing. But it is important to note that in this passage Socrates speaks of reading texts, and making selections from them, in common with his friends, i.e. of reading together with conversation. Euthydemus apparently did not read his texts in this interactive way, but rather privately and passively on his own. Socrates does not explicitly criticize Euthydemus’ approach to his library, but his conversation with Euthydemus is shot through with references to writing. These references show that Xenophon understood the limitations of writing in much the same way that Plato did. The first hint of this comes at the end of Socrates’ list of possible careers for Euthydemus. He has not collected books to become a doctor, architect, geometer, or astronomer, but he has collected all the works of Homer. Does he perhaps want to be a rhapsode? No, Euthydemus replies, for the rhapsodes, while they may know Homer’s verses well enough, are fools (4. 2. 10). This is, however, the standard 27 Save, perhaps, for the verses Plato has him discuss at Phaedo 60 c–61 b.

Created on 4 July 2005 at 14.33 hours page 52

David M. Johnson

52

criticism of rhapsodes (as at Xen. Smp. 3. 6), and so in itself is hardly conclusive. But when the two come to discuss whether given actions are just or unjust, Socrates rather curiously suggests that they write a delta in one spot, perhaps drawing in the sand beneath them, and an alpha in another spot. They can then put whatever appears to be a work of justice (δικαιοσνη) by the one, and whatever a work of injustice (δικα) by the other (4. 2. 13). The actions under consideration initially seem unjust, resulting in something like Table 1. But Socrates quickly points out that in war it is just to do such things. So everything put with injustice must also be put with justice. Socrates suggests that they divide such actions up depending upon whether they are aimed at enemies or friends. We now get something like Table 2.   1 ∆

[ργα δικαιοσνης Just deeds]

Α

[ργα δικας Unjust deeds] ψεδεσθαι (to lie) ξαπατ!ν (to deceive) κακουργε ν (to harm) νδραποδζεσθαι (to enslave)

  2 ∆

Α

[ργα δικαιοσνης Just deeds]

[ργα δικας Unjust deeds]

ψεδεσθαι (to lie) ξαπατ!ν (to deceive) κλπτειν τε κα# πρ$ς πολεµους &ρπ'ζειν (to steal re enemies and seize)28 νδραποδζεσθαι (to enslave)

ψεδεσθαι (to lie) ξαπατ!ν (to deceive) κακουργε ν (to harm) πρ$ς φλους re friends νδραποδζεσθαι (to enslave)

But Socrates now introduces occasions on which it is just to lie 28 Here I assume that stealing and seizing property, which are listed instead of harming at 4. 2. 15, are merely examples of harming, but see below.

Created on 4 July 2005 at 14.33 hours page 53

Xenophon at his Most Socratic

53

even to friends, to deceive them, and even to steal their possessions. So these actions, even when directed at friends, must be put with justice, showing that even our second table cannot be right. Here, if not before, the attempt to produce a written table of just acts fails, and Euthydemus asks if he may be allowed to take back things that he has said (4. 2. 18). Socrates agrees, but also changes the topic, leaving us with the impression that it is impossible properly to categorize actions as just or unjust. We shall return to what this conversation tells us about justice, but for now we should consider why Xenophon had Socrates and Euthydemus make use of writing in this way. There is no obvious reason why they must write down their examples, nor does Socrates ask his interlocutor to make such a list anywhere else.29 Xenophon’s point is that the attempt to pin down concepts like justice in writing is a most difficult one. Individual circumstances can frustrate any attempt at a static written definition. The attempted chart cannot cover all eventualities, and this is a flaw endemic in the written word. Xenophon here subtly shows this flaw rather than labouring the point explicitly. This interpretation is confirmed by continued references to writing as we go on. The next argument, about voluntary wrongdoing, is often compared to that at the end of the Hippias Minor. But where Plato’s Socrates, in order to consider the meaning of intentional injustice, gives us many examples of those who intentionally perform badly, Xenophon provides but one, and one not included by Plato, that of the literate person who voluntarily writes or reads badly (4. 2. 20).30 In his central speech on self-knowledge, Socrates asks Euthydemus if he knows the saying written at Delphi, ‘Know Thyself’. Euthydemus of course knows it, but has learnt nothing more from it than he has from any of the books in his library.31 Finally, Socrates makes the point that such self-knowledge does not consist merely in knowing one’s name, but in knowing one’s capabilities (4. 2. 25). One of the things that Memorabilia 4. 2 is meant to show, then, is that written texts, at least if they are read as Euthydemus reads them, do not provide one with knowledge. Xenophon shows an 29 The geometric diagram of the Meno is a rather di·erent matter, as there the diagram helps the slave reach the correct answer. 30 Earlier in the Hippias Minor, at 366 c, the example of spelling Socrates’ name is considered in passing. 31 Strauss, Xenophon’s Socrates, 98.

Created on 4 July 2005 at 14.33 hours page 54

54

David M. Johnson

awareness of the limitations of writing elsewhere. At Hipparchicus 9. 1 he notes that the cavalry commander, after reading his treatise a few times, must always look to the particular circumstances in which he finds himself; one can no more write down everything that he will need to do than one can commit everything that will happen in the future to writing. At Oeconomicus 16. 1 Xenophon’s Ischomachus attacks those whose painstaking theoretical writings on the nature of soil are not supported by any practical experience; what is needed is observation of what a given soil can support, not technical writings. But of course Xenophon was a prolific writer who made use of the writings of others, never more so than in our Memorabilia passage. The point is not simply that writing is suspect, but that one must write and read in the right way, one that encourages active engagement of the sort Socrates suggests when he describes the value of reading books with his friends (Mem. 1. 6. 14). Xenophon hints as much near the close of our chapter. With Euthydemus committed to following Socrates, Socrates will change his tactics: And he [Socrates], when he understood his [Euthydemus’] attitude, would confuse him not at all, but would most simply and most clearly explain what he considered it necessary to know and what most important to do.

The simplicity and openness Socrates promises here is what we would expect from Xenophon’s Socrates, who is more didactic in his subsequent conversations with Euthydemus (4. 3, 4. 5, 4. 6).32 But the promise of openness in the future implies that Socrates had 32 One reason why Hippias replaces Euthydemus in 4. 4 is that the conversation there is less straightforward. Or so I argue in ‘Law and Justice’; it is more straightforward for D. Morrison, ‘Xenophon’s Socrates on the Just and the Lawful’ [‘Just and Lawful’], Ancient Philosophy, 15 (1995), 329–47. Xenophon’s promise to be straightforward in what follows seems to imply that, outside of conversations like 4. 2, his Socrates simply says what he thinks. At the end of the conversations with Euthydemus, Xenophon appears to confirm this: ‘That Socrates simply revealed his opinion to those who spent time with him, then, seems to me clear from what has been said’ (4. 7. 1). But this claim is considerably complicated by the passage that immediately precedes it, in which Xenophon outlines two di·erent Socratic methods. In the first Socrates leads those who disagree with him back to first principles, a method perhaps comparable to that of Plato’s Socrates (4. 6. 13–14); in the second, ‘whenever he went through something himself in a speech, he would make his way by means of the things that were most agreed upon, believing that this was the safest way of speaking. Therefore far more than anyone I know he produced listeners who agreed with him’ (4. 6. 15). This method, presumably that of Socrates in the bulk of the Memorabilia, seems straightforwardly banal (but see Morrison, ‘Vlastos’ Xe-

Created on 4 July 2005 at 14.33 hours page 55

Xenophon at his Most Socratic

55

not been simple and clear with Euthydemus in 4. 2 itself. We shall therefore have to read our chapter actively, with the assumption that it is not meant to be simple and straightforward, i.e. that Xenophon is not as na•§ve as he is normally made out to be. One of the points that Xenophon leaves implicit is the need to be on the lookout for implicit points. If we read Xenophon as he is normally read, we risk reading like Euthydemus.

3. Justice So the discussion of justice makes one important implicit point about the limitations of writing. But does it say anything substantive about justice? Once Euthydemus has consented to speak with Socrates, the two quickly enough reach agreement on the goal of Euthydemus’ ambition: he desires to acquire the virtue which will allow him to succeed in managing private and public a·airs in such a way as to benefit both himself and others. But this cannot be had without justice. Euthydemus says that he seems as just as anyone, and he agrees with Socrates’ suggestion that as carpenters, if asked, can show o· their deeds, so too the just should be able to explain theirs. But, as we have seen, Euthydemus is unable to suggest an example of unjust action which Socrates cannot show to be just under di·erent circumstances (4. 2. 11–18). Our argument is one of the passages in which Xenophon’s Socrates seems to condone harming enemies, a position that puts him at odds with the claim of Plato’s Socrates that it is never right to harm (κακουργε ν) anyone, i.e., to treat anyone in a way that would in normal circumstances be considered unjust, even if one has already been treated unjustly by the person in question (Crito 49 a–e; cf. Rep. 1, 331 e–336 a). This is a truly radical claim, as Socrates makes clear in the Crito, so radical that the many who do not share this view will have no common ground for deliberation with those few who endorse it.33 It is thus not surprising that Xenophon’s Socrates nophon’, 14–16, and ‘Socrates as Teacher’, 191–4). In the next sentence, however, Xenophon says that Socrates used to say that Homer referred to Odysseus as a safe orator because he did the same thing. Thus Socrates’ seemingly banal method is comparable to that of Odysseus, the master storyteller and inveterate liar. 33 For the radical nature of the claim, see G. Vlastos, Socrates: Ironist and Moral Philosopher (Ithaca, NY, 1991), 179–99; and M. W. Blundell, Helping Friends and

Created on 4 July 2005 at 14.33 hours page 56

56

David M. Johnson

never says that one must not harm anyone, even an enemy, and in fact even Plato’s Socrates sometimes appears, at least by the time of the Republic, to endorse the traditional idea that one should treat enemies very di·erently from the way one treats friends.34 And while Xenophon’s Socrates says much about acquiring and helping friends, he says little about dealing with enemies. Xenophon never depicts Socrates harming, or advising anyone to harm, an enemy. Now the fact that Xenophon’s Socrates does not put harming enemies on a par with helping friends does not prove that Xenophon’s Socrates di·ers from his peers in this respect, for harm to enemies is less often made an object of moral approval.35 But it is noteworthy that in summing up his case for Socrates at the end of the Memorabilia, Xenophon says that Socrates was so just that he harmed no one even slightly, but was of the greatest benefit to those who dealt with him (4. 8. 10–11). This would imply that Xenophon’s Socrates did not view the harming of enemies as just.36 Of course, if we are to understand what Xenophon’s Socrates means by helping friends and harming enemies, we had better understand what he means by friendship and what he means by harm.37 In Memorabilia 2. 6, his most thorough discussion of friendship, Xenophon’s Socrates argues that friendship between those who are bad, or between the good and the bad, is impossible, but that there is no reason in principle why the good cannot befriend one another (2. 6. 22–7). The good now have the bad as their enemies. Harming Enemies: A Study in Sophocles and Greek Ethics [Helping Friends] (Cambridge, 1989), 56. 34 In the Republic Plato has Socrates propose that the Guardians treat enemies, particularly barbarian enemies, di·erently from the way they treat friends, just as Xenophon’s Socrates says that the general must exhibit contrasting qualities to friend and foe (Rep. 2, 375 a–c; 5, 469 b–471 c; Mem. 3. 1. 6). In the Philebus (49 b) Plato has Socrates say that it is not unjust to rejoice at an enemy’s misfortune; Xenophon has Socrates imply as much at Mem. 3. 9. 8. Even Vlastos (Socrates, 197 n. 58) admits that ‘Plato’s recreation of Socratic thought may limp at this point [the claim that doing anyone evil means doing injustice], possibly because he had not been fully in sympathy with this particular aspect of Socrates’ teaching.’ 35 Blundell, Helping Friends, 53–9. 36 At Clit. 410 a–b, Clitophon recounts that he finally asked Socrates himself what justice is. ‘You told me that it was characteristic of justice to harm enemies and benefit friends. But later the just man appeared never to harm anyone, for everything he does to anyone is to their advantage.’ The movement from endorsing the traditional distinction between friends and enemies to rejecting harming anyone parallels that in the Memorabilia. But the Clitophon passage is admittedly a greatly compressed presentation by a figure not entirely sympathetic to Socrates. 37 Morrison, ‘Vlastos’ Xenophon’, 16–18.

Created on 4 July 2005 at 14.33 hours page 57

Xenophon at his Most Socratic

57

What does it mean to harm the bad? On the conception of harm that Plato’s Socrates articulates in the first book of the Republic, harm means reducing the characteristic excellence of something, its aret»e, be it dog, horse, or man. But as the bad already lack aret»e, one cannot really harm them.38 It is therefore possible that when Xenophon’s Socrates speaks of harming enemies, he may be speaking loosely of the sorts of things the many consider harmful, things like theft, bodily harm, and enslavement, rather than of the only unambiguous harm, loss of virtue. As Xenophon has Socrates note in the Oeconomicus, it often happens that in war noble enemies force those they defeat to become better by making them more moderate (πολλο(ς δ) βελτους *ν'γκασαν ε+ναι σωφρονσαντες 1. 23; cf. Rep. 5, 471 a). Deceit may serve to help defeat enemies and bring them to their senses; plundering them will rob them of the resources they might use to commit injustice against others and against themselves. Plato informs us that Socrates fought bravely for Athens against her enemies during three campaigns of the Peloponnesian War (Ap. 28 e; La. 181 a; Sym. 219 e, 220 e; Chrm. 153 a); while Socrates is praised mainly for his endurance and cool-headedness in retreat, to have served honourably he must also have been willing to wound or kill the enemy. Even Plato’s Socrates does not advocate a revolutionary change in one’s conduct towards enemies so much as in how one understands that conduct. A close reading of our passage will provide some evidence for this understanding of what Xenophon’s Socrates means by harming enemies. First, Xenophon does not simply say that one may justly enslave enemies, but requires that one be a general who is dealing with a city that is not only hostile but unjust (4. 2. 15). As Socrates later shows to Euthydemus, those who are unjust are those who do not understand justice and are therefore slavish. Is it really harmful to make slaves of the slavish? Enslaving was the last example of injustice in Socrates’ first list (lying, deceiving, harming, enslaving: 4. 2. 14). Socrates goes on to say that deceiving the enemy is just, and as deceiving entails lying, we may expect that Socrates, to be consistent, will complete the list by noting that it is also just to harm enemies (κακουργε ν). But instead he says that seiz38 At Gorg. 480 e–481 b Socrates says that, if it is right to harm anyone, rhetoric may be useful in harming the bad by helping them to enjoy their ill-gotten gains and to escape punishment. But at 481 b 4–5 Socrates reveals that this is really part of a reductio argument meant to show that rhetoric is useless.

Created on 4 July 2005 at 14.33 hours page 58

58

David M. Johnson

ing and stealing what belongs to them is just. He thus avoids ever explicitly saying that it is just to harm enemies. Now most readers will have assumed that seizing and stealing constitute harm, and Socrates does go on to sum up by asking whether ‘as many things as we placed with injustice, these things should also be placed with justice’ (4. 2. 15). This implies that harming, which had been put with injustice earlier, is now to be considered just. But in his next question Socrates asks whether they should posit that it is just to do ‘such things’ (τοια,τα) to enemies, but not to friends, with whom one must be as straightforward as can be. Here we expect that ‘such things’, inasmuch as they are opposed to honesty, will consist mainly of deceit, and so they do; the argument concludes by rejecting the idea that one must always be straightforward with friends (4. 2. 18). Harm is mentioned just once again, but only to introduce the paradoxical argument about whether it is just to harm a friend intentionally (4. 2. 19). We learn no more about harming enemies. Contrast On Justice, where harming enemies (βλ'πτειν) is not only explicitly said to be just, but made the justification for deceiving and lying to enemies, and where it is even said that it is unjust to benefit enemies (374 c–d). Thus by avoiding use of the general term ‘harm’ and restricting himself to the specific actions of enslaving, deceit, and theft, Xenophon does not commit his Socrates to the traditional understanding that harming enemies is just. But neither does Xenophon have Socrates explicitly stake out the revolutionary position that Plato’s Socrates adopts in the Crito. What can we say about Socrates’ view of justice in our passage? For Xenophon, Socrates’ justice consists above all in his benefiting of others, to which Xenophon devotes the bulk of the Memorabilia.39 The same idea seems implicit in our passage, but Xenophon does not have Socrates make this clear, the better to leave Euthydemus in a state of flux, and, perhaps, to force readers to think through what Socrates is about. What Socrates does is to show that Euthydemus’ notion that it is unjust to deceive friends is incorrect (4. 2. 16–18). For if a general lies to his troops and thus raises their spirits, or a father lies to his child and thus gets him to take the medicine that heals him, or if one who fears 39 This is clearest at Mem. 4. 8. 10–11. On Socrates’ justice, see P. Vander Waerdt, ‘Socratic Justice and Self-Su¶ciency: The Story of the Delphic Oracle in Xenophon’s Apology of Socrates’ [‘Justice’], Oxford Studies in Ancient Philosophy, 11 (1993), 1–48, esp. 46–7; Morrison, ‘Just and Lawful’; and Johnson, ‘Law and Justice’.

Created on 4 July 2005 at 14.33 hours page 59

Xenophon at his Most Socratic

59

that his depressed friend will kill himself and steals his sword, the deceit is just.40 Euthydemus says that he will therefore change his position, if this is possible, and Socrates, as we would expect, says that one must change one’s statement in such cases. Yet rather than giving Euthydemus the opportunity to revise his position, Socrates immediately confuses matters by introducing a further complication (4. 2. 19): Of those who are deceiving their friends so as to harm them, so we don’t leave even this uninvestigated, which is more unjust, the one who does so intentionally, or the one who does so unintentionally?

Socrates’ question shows that Euthydemus could have salvaged his account of injustice by adding the qualification that deceit is unjust when one deceives friends so as to harm them. As a definition of injustice this is too narrow, of course; at the very least, harming one’s friends in any way is unjust (leaving aside the question of harming enemies). But it is otherwise sound enough. Yet Socrates, rather than allowing Euthydemus to savour a moment of success, introduces the Hippias Minor paradox that makes intentional wrongdoing seem better than unintentional wrongdoing. The argument that follows is extremely compressed, and is therefore di¶cult to analyse, although at least some of the compression is well motivated, as we shall see. Though he is by now understandably hesitant, Euthydemus gives the commonsensical answer to Socrates’ question, only to see it refuted in short order (4. 2. 19–20): (1) Euthydemus’ hypothesis: one who willingly tells a falsehood [to friends so as to harm them?] is more unjust than one who does so unwillingly. (2) But one can learn and understand justice just as one can learn one’s letters. (3) He who intentionally writes or reads badly is more literate. (4) For he could write and read well, if he wished. (5) So one who intentionally writes badly is literate, and one who does so unintentionally is not. 40 The would-be suicide is found both in Dissoi Logoi 3. 4 and Rep. 1, 331 c, deception in administering medicine at Dissoi Logoi 3. 2. Xenophon adds the deceptive general, an example dear to his own heart (see also Cyr. 1. 6. 27–41; Mem. 3. 1. 6), but also a skilful segue from the discussion of just deceit of enemies to just deceit of friends. He skips the more convoluted examples of Dissoi Logoi 3 (such as tunnelling to rescue one’s captive father).

Created on 4 July 2005 at 14.33 hours page 60

60

David M. Johnson

(6) One who intentionally tells falsehoods and deceives [friends to their harm?] knows what is just. (7) One who knows one’s letters is more literate than one who does not. (8) One who knows what is just is more just than one who does not. [(9) Therefore one who intentionally deceives Éfriends to their harm?Ö is more just than one who does so unintentionally.] The first issue we had better consider is the status of my bracketed additions to the argument in (1), (6), and (9). If one reads the argument charitably from a logical point of view, some such addition seems necessary. Otherwise we begin with a sort of injustice, deceit, that has just been shown to be just as well as unjust. And the qualification about deceiving friends to their harm is included in the question which introduces this argument. We can then supply the same condition in the missing conclusion (9), which contradicts Euthydemus’ initial hypothesis. But it can also be argued, from what may be called a literary point of view, that it is more charitable to assume that the absence of the crucial qualifying phrase is significant. At the very least it serves to make the argument appear more palatable, as Socrates and Euthydemus need not explicitly say that intentional deceit which harms friends is somehow just. The absence of the qualifying phrase is then a euphemism, albeit one which is unfortunate from a logical point of view. But perhaps there is still more to it: the omission could be a Xenophontic hint about the weakness of the argument. For there is nothing which corresponds to harming friends in the example of the literate man. Justice is more complicated than literacy. If justice is benefaction, its object, the good, may be categorically di·erent from the objects of other sorts of knowledge. Xenophon elsewhere attributes the following claim to Socrates (3. 9. 5): For just actions and all actions which are done by virtue are fine and good; and neither do those who know these things choose to do something instead of them, nor are those who do not know them able to do them.41 41 For an argument to prove this, see 4. 6. 5–6. Mem. 4. 5. 6, where Xenophon comes perilously close to admitting the possibility of incontinence, complicates matters, but even Vlastos (Socrates, 99–102) is willing to acquit Xenophon of ‘rank confusion’ on this issue. For further complications about the good, see Mem. 3. 8, where Socrates refuses to name anything good which is not good for something.

Created on 4 July 2005 at 14.33 hours page 61

Xenophon at his Most Socratic

61

A fortiori, there is no reason to harm one’s friends intentionally. There is, on the other hand, nothing preventing one from reading or writing incorrectly on purpose. By forcing us to think about whether or not we are to include the material inserted in brackets above, Xenophon may put greater emphasis on the question of harm and its opposite, benefit. Benefaction becomes the key missing term in the whole first half of the conversation, missing not only from Euthydemus’ first attempts at defining justice but from the argument about intentional wrongdoing. And benefaction, in the form of what is to count as good, will be the explicit topic of the last section of the conversation, supporting the idea that Xenophon wanted readers to wonder about the role of the good in the earlier parts of the argument. At the end of the Hippias Minor Plato may have given us a similar hint, albeit more explicitly, by having Socrates wonder whether anyone does intentionally err when it comes to justice and the good. Xenophon almost certainly had the Hippias Minor in mind,42 and certainly we cannot but read him in the light of it. For our purposes the most important part of the Hippias Minor is the ending (376 b 4–c 6):  . So the one who intentionally errs and does what is shameful and unjust, Hippias, if there is any such person, would not be anyone other than the good person.  . I cannot agree with you here, Socrates.  . Nor can I with myself. But now, at any rate, this appears necessary to me, from the argument. Yet as I was saying a while back, I’m all over the place about these things, and they never seem the same to me. It’s nothing amazing if I’m confused, or if any other ordinary person is. But if even you wise ones are confused—this is awful for us, since we can’t put an end to our confusion even by coming to you.

In the Hippias Minor Socrates says that he is himself confused and disturbed by the argument, and he seems to provide a possible escape from its conclusion with the wa}ing ‘if there is any such person’. Xenophon’s Socrates does not confess to any such uncertainty. He does not admit to confusion because he is speaking to a youth 42 If Xenophon was inspired by the Hippias Minor, we can better explain why he combined the content of the first argument of that dialogue (the true and false man) with that of the second (intentional vs. unintentional wrongdoing), and why the language he used to describe Euthydemus’ confusion is very close to that which Plato had Socrates use of himself. For a comparison of the two passages, see J. Phillips, ‘Xenophon’s Memorabilia 4. 2’, Hermes, 117 (1989), 366–70.

Created on 4 July 2005 at 14.33 hours page 62

62

David M. Johnson

who is to become his student. Plato’s Socrates, on the other hand, ironically poses as someone eager to learn from Hippias, the better to highlight his humiliation of him. But neither does Xenophon’s Socrates endorse the conclusion that unintentional wrongdoing is better than intentional wrongdoing. For Euthydemus and Socrates, the argument serves to show a contradiction in Euthydemus’ thinking. Hence they do not bring the argument to a conclusion, but rather discuss Euthydemus’ inability to say the same thing about the same things, which Euthydemus concludes must be due to his belief that he knew something he did not know (4. 2. 21). Socrates shows Euthydemus that this condition is slavish, and Euthydemus despairs that, despite all his previous attempts at philosophizing, he cannot answer the questions that most need answering. It cannot be proven that Xenophon himself could answer these questions, though I have speculated above that he gave us a hint about how to go about doing so. But the same uncertainty complicates, and makes fascinating, the discussion of Plato’s intentions in the Hippias Minor.43

4. Self-knowledge Socrates now gives Euthydemus a brief respite from questions. More in keeping with his method elsewhere in the Memorabilia, Socrates outlines a positive position while only occasionally asking for his interlocutor’s assent. But while the mode of the argument resembles that seen elsewhere in the Memorabilia, the subject matter, self-knowledge, is one usually associated not with the Xenophontic but with the Platonic Socrates. It has often been remarked that an important di·erence between Xenophon’s and Plato’s portrayals of Socrates is Xenophon’s emphasis on self-control, γκρ'τεια, where Plato emphasizes self-knowledge.44 But in 4. 2 Xenophon clearly recognizes the value of self-knowledge. Our initial question should be whether his own account of self-knowledge is of 43 Most agree with Aristotle that the essential flaw in the argument is the failure to distinguish between a dunamis, the sort of ability that enables one intentionally to perform well or badly, and a hexis, an ability that includes a disposition that commits one to acting in a certain way (Kahn, Socratic Dialogue, 113–19; Vlastos, Socrates, 275–80). Contrast R. Weiss, ‘Ho agathos as ho dunatos in the Hippias Minor’, in H. Benson (ed.), Essays on the Philosophy of Socrates (Oxford, 1992), 242–62. 44 See e.g. Vander Waerdt, Socratic Movement, 12.

Created on 4 July 2005 at 14.33 hours page 63

Xenophon at his Most Socratic

63

any value. Xenophon’s Socrates argues that those who know themselves know what they can and cannot do. They therefore succeed in everything they undertake themselves, and hand over what they do not understand to those who have the relevant ability. While those with self-knowledge therefore flourish and are honoured by all, those without self-knowledge fail and meet with disgrace. And cities which know their own capability fare well, while those which do not know themselves end up enslaved or in civil war (4. 2. 26–9). The presentation of self-knowledge here is closest to that which Plato has Socrates outline in the Alcibiades I (116 e–118 b) and reject in the Charmides. In the latter Socrates outlines the possible benefits of such self-knowledge in terms similar to those employed here (171 d–172 a). But he rejects the possibility of such self-knowledge by arguing that while there may be a second-order knowledge that allows one to know that one knows something, one cannot know what one knows unless one has first-order knowledge of the relevant subject area (169 e–171 c). And even if this sort of second-order self-knowledge were possible, it would not be beneficial without one sort of first-order knowledge, knowledge of good and evil (172 c–175 a). Xenophon’s argument shows no inkling of the complicated problems with the possibility of secondorder knowledge that Plato discusses in the Charmides. This is not terribly surprising or disappointing, as these complications may be more Platonic than Socratic. But he may show some awareness of the second problem with Socratic self-knowledge, the worry that it is useless unless one has knowledge of the good. For Socrates and Euthydemus will turn to a discussion of the good following their discussion of self-knowledge, thus potentially meeting this problem. Certainly Xenophon’s self-knowledge is characteristically more concrete and practical than that presented by Plato in the Apology or Charmides. Xenophontic self-knowledge consists not in knowing that one knows nothing, or in knowing what or even that one knows, but in knowing what one knows how to do. If one can do all that one knows how to do, if knowledge is not only necessary but su¶cient to do anything worth doing, our passage is compatible with a highly intellectualist Socrates. But it is also compatible with a more commonsensical view in which knowledge is necessary but not su¶cient for success. This is at any rate the most natural reading of the final example, the power of the city in war,

Created on 4 July 2005 at 14.33 hours page 64

64

David M. Johnson

which presumably consists not only in the city’s self-knowledge, but also in its military resources, a topic Xenophon’s Socrates discusses elsewhere (Memorabilia 3. 5 and 3. 6). This more concrete understanding of self-knowledge is in keeping with that displayed in the central speech of the Alcibiades I, where Socrates would teach Alcibiades to know himself by showing that his riches and noble birth pale in comparison with those of his true rivals, the kings of Sparta and Persia. It is a sort of self-knowledge that can readily be connected with justice, the topic of the first part of Socrates’ discussions with Alcibiades and Euthydemus, inasmuch as knowing one’s own abilities is tantamount to knowing what is owed to others.45 As it is the politically ambitious Euthydemus to whom Socrates is trying to show the importance of self-knowledge, it is not surprising that he stresses the honour and esteem that self-knowledge can bring and the contempt that results from ignorant attempts to go beyond one’s capability. But in holding up the honours that those with self-knowledge gain from those without self-knowledge, Xenophon appears to get himself into trouble (4. 2. 28): And those who know what they are doing, succeeding in what they do, gain high repute and honour. And they gladly deal with those like themselves, while those who fail in what they do want them to provide counsel on their behalf, and to be in charge of them, and place all their hopes for success in them, and because of all of this are particularly fond of them.

This seems to grant those who are ignorant of themselves too much knowledge. Not only do we have the mystery of how self-knowledge can enable one to judge others, something familiar enough from the Charmides, we have the mystery of how those who lack selfknowledge can nevertheless be sensible enough to want the selfknowers to rule them. While avoiding many of the subtleties and problems of the Charmides, Xenophon appears to have introduced a new di¶culty of his own. But note that those lacking self-knowledge come to this realization only after failing themselves, and only after the self-knowers have met with success. It is apparently possible even for those who do not have full self-knowledge to judge the tree by its fruits. The unsuccessful experience of those ignorant of themselves may have given them a sort of self-knowledge, but if so it is of a limited and 45 Annas, ‘Self-Knowledge’, 118–26.

Created on 4 July 2005 at 14.33 hours page 65

Xenophon at his Most Socratic

65

sobering sort, the sort that makes them voluntarily place themselves under the leadership of others. This knowledge, if such it is, is clearly distinct from that of those with full self-knowledge, who recognize that they are in fact able to do something, and, while they gladly work with others, do not need to submit to others. As we shall see, it is precisely the sober sort of self-knowledge, knowledge of one’s ignorance, that Euthydemus will acquire in our conversation. Thus Xenophon’s added wrinkle, the ability of those without full self-knowledge to recognize their betters, works on two levels. In its immediate context it serves further to convince Euthydemus of the advantages of self-knowledge, and in the wider context of the whole conversation it helps to explain how Euthydemus can, despite his ignorance, realize that he needs Socrates’ help. The argument closes by attributing a sort of self-knowledge (or the lack thereof) to cities. Those with a mistaken (and inflated) notion of their own strength wage war on stronger cities, and are either enslaved or su·er revolution as a result. We are again in the concrete and practical world of Xenophon. Compare the long central speech of the Alcibiades I (120 e–124 b), in which Socrates details the material advantages of Athens’ rivals as a way of teaching Alcibiades self-knowledge. There is much more to that speech than this, as much of Socrates’ praise of Sparta and Persia is ironic, and the overall intent is to get Alcibiades not to consider material advantages but to look to the betterment of his soul. Alcibiades is to come to know that as he cannot rival Spartan or Persian kings in wealth or noble blood, he must excel them in wisdom (123 d). In the middle of Aeschines’ Alcibiades Socrates delivers a similar speech, detailing how Themistocles outwitted Xerxes despite Xerxes’ material advantages.46 Socrates’ speech about self-knowledge here plays the same structural role. And, like the speeches in the Alcibiades dialogues, it emphasizes the importance of intellectual virtue even when practical goals are at issue.

5. Goods It would appear that Euthydemus is now ready to be helped by Socrates. He has come to recognize that he does not understand 46 VIA.50 in Socratis et Socraticorum reliquiae, ed. G. Giannantoni (4 vols.; Naples, 1990).

Created on 4 July 2005 at 14.33 hours page 66

66

David M. Johnson

justice. He has been taught the great importance of self-knowledge, and now asks Socrates to show him how he ought to begin to investigate himself (4. 2. 30). Socrates, however, continues to ask Euthydemus questions rather than to reveal to him the path towards self-knowledge. The topic of self-knowledge is apparently dropped, save in the sense that Socrates’ further questions help Euthydemus to understand his own ignorance more fully. Surely, Socrates asks Euthydemus, you know what sort of thing is good and what sort is bad? Euthydemus agrees, saying that if he did not know such things, he would be worse o· than a slave. But of course Euthydemus cannot tell Socrates what is good and what not, so the elenchus continues. Compare, once again, the Alcibiades I (124 b), in which Socrates continues to refute Alcibiades even after the latter has asked him for help—although in the Alcibiades I, unlike our conversation, Socrates and Alcibiades explicitly return to the topic of self-knowledge. In our passage, the dramatic justification for the continued refutation is precisely Euthydemus’ failure to come to terms with the extent of his ignorance. After his failure to answer Socrates’ questions about justice, Euthydemus agreed that ignorance about what is fine, good, and just made one slavish, and realized he needed help in avoiding this fate (4. 2. 22–3); after being shown that he also does not understand the good, he will recognize that he truly is a slave (4. 2. 39). A philosophical justification must come in the topic of the continued elenchus: the good. As we have seen, the first part of the conversation suggested that justice was connected with benefaction. But it did not address what is to count as a benefit. And the central discussion of self-knowledge noted that those who do know themselves know what is useful for them (πιτ-δεια) or what they need, i.e. what is good for them, but it did not address what is to count as good. As the Charmides makes clear, self-knowledge is incomplete without knowledge of the good. Thus the last part of the conversation makes dramatic sense inasmuch as it completes the refutation of Euthydemus, and it makes philosophical sense as it addresses the most important concept missing from the first two parts of the conversation. Euthydemus’ first suggestion for a good is health, health itself and then, in a refinement in which one imagines he takes pride, the things that cause health. Socrates uses the refinement to undo Euthydemus, for he notes that health itself may be the cause of evils,

Created on 4 July 2005 at 14.33 hours page 67

Xenophon at his Most Socratic

67

as when good health allows one to take part in an unfortunate voyage or military campaign. Euthydemus intelligently replies that health can also be a cause of advantageous things, but Socrates argues that anything that is the cause of both good and evil is itself neither good or evil. The argument that only things that are invariably good are good in the strictest sense is eminently Socratic (Plato, Euthd. 278 e–282 b; Meno 87 d–89 a). So is the thesis that bodily goods are of far lesser value than those of the soul (Plato, Ap. 30 a– b; Gorg. 477 a–e, 511 c–512 b; Crito 47 c–48 a; Alc. I 129 b–131 b; Xen. Mem. 4. 1. 2, 4. 3. 14; Sym. 8. 10–36). Still, the treatment of health here is rather surprising, given the high praise Xenophon’s Socrates bestows on health elsewhere. In Memorabilia 3. 12 Socrates persuades his interlocutor to take better care of his body (cf. Mem. 4. 7. 9). And of course one reason why Xenophon’s Socrates does not prepare a defence speech is his awareness that old age brings with it ill health (Ap. 6; Mem. 4. 8. 1). But while Xenophon’s Socrates certainly spends more time praising health than does Plato’s, Plato’s Socrates also appears to believe that life is not worth living with an unhealthy body (Crito 47 e; cf. Gorg. 512 a–b). And Xenophon’s Socrates praises bodily health in large part because of its contribution to thought; the worst aspect of old age for him is the mental deterioration it so often brings (Mem. 3. 12. 6–7, 4. 8. 1; Ap. 6). Both Plato’s and Xenophon’s Socrates recognize that a healthy body is the most valuable thing after a healthy soul. Euthydemus’ next suggestion is even more promising, and thus Socrates’ quick rejection of it appears more problematic. Wisdom, Euthydemus suggests, is indisputably good (4. 2. 33). Socrates responds that Daedalus, because of his wisdom, was forced to be a slave in exile to Minos, and while attempting to escape he lost his son and ended up himself among barbarians, in slavery again. Palamedes perished because Odysseus begrudged him his wisdom. And many others, because of their wisdom, have been taken up to the king of Persia and are slaves there. Here we enter on ground untouched by Plato’s Socrates, the claim that wisdom itself is not always good. In the Meno Socrates goes as far as to say that justice, temperance, and courage are sometimes harmful, but it is precisely the presence of knowledge that makes them useful (87 d–89 a; see also Euthd. 281 a–e). Critics of Xenophon argue that he has garbled his sources, and failed to distin-

Created on 4 July 2005 at 14.33 hours page 68

68

David M. Johnson

guish between popular goods like health and the one true, Socratic good, wisdom.47 The problem with the argument Xenophon has Socrates advance here is that the sort of wisdom involved in the examples is not the moral knowledge which Socrates aims for, but rather some technical skill. Euthydemus’ initial formulation of his claim implies as much. ‘What sort of thing would a wise man not do better than an ignorant one?’ By ‘wise man’ Euthydemus presumably meant the man with knowledge of what is fine, good, and just, the sort of knowledge that distinguishes the slavish from the free, as Socrates and Euthydemus have just discussed (4. 2. 22). While Palamedes and Daedalus are credited with a great variety of inventions and skills in our sources, this by itself gives them no claim to understanding of this sort. Perhaps Xenophon simply failed to realize this. But Xenophon’s mythological examples show that he was up to something, for both Daedalus and Palamedes are connected to Socrates. Both Plato and Xenophon compare Socrates’ fate to that of Palamedes (Plato, Ap. 41 a–c; Xen. Ap. 27). Daedalus is Socrates’ mythological ancestor, and Socrates has the daedalic power to make arguments move about as Daedalus made statues move (Euthph. 11 b–d; Alc. I 121 a). This implies that if wisdom was bad for them, then it would be bad for Socrates. But it is hard to believe that Xenophon could have thought any such thing. Xenophon elsewhere in the Memorabilia (3. 9. 4–5; cf. 4. 5. 6) is clear on the ultimate importance of wisdom, and his Apology is largely dedicated to the goal of showing why death was no evil for Socrates. So it does not seem over-charitable to credit him with seeing the fallacy here.48 Probably, Euthydemus says, now less confident of himself, happiness is an utterly indisputable good (κινδυνεει . . . ναµφιλογ/τατον γαθ$ν ε+ναι τ$ ε0δαιµονε ν, 4. 2. 34). Unless one puts it together out of disputable goods, Socrates responds (ε1 γε µ- τις α0τ2, φη, 4 Ε0θδηµε, ξ µφιλ2γων γαθν συντιθεη). Euthydemus wonders what happy thing could be disputable (τ δ6 7ν, φη, τν ε0δαιµονικν µφλογον ε1η;). Nothing, Socrates replies, unless we add to happiness (προσθ-σοµεν α0τ8) beauty or strength or wealth or reputation 47 Kahn, Socratic Dialogue, 396–7; Jo•el, Echter Sokrates, i. 414–18. 48 At Sym. 3. 4 Antisthenes says that wisdom sometimes seems harmful to one’s friends and city; Jo•el (Echter Sokrates, i. 417) took this as confirmation that the treatment of wisdom in 4. 2 was Antisthenic. But even Antisthenes qualifies his remark with δοκε , and Xenophon’s Socrates need not share Antisthenes’ view.

Created on 4 July 2005 at 14.33 hours page 69

Xenophon at his Most Socratic

69

or something of that sort. Euthydemus cannot imagine happiness without such goods. But Socrates points out that they have caused great evils for many. In what sounds like the catalogue of the paradoxical obstacles Alcibiades faced, Socrates quickly notes that many have been corrupted by those moved by their beauty, many have been made profligate and been plotted against thanks to wealth, and many have su·ered great misfortune thanks to their fame and political power (4. 2. 35). Euthydemus is now still more despondent, and worries that he does not even know what to pray for. Here too, we can understand the argument in two ways. Perhaps Xenophon, eager to have Socrates throw Euthydemus into complete confusion but failing to understand what he was about, put together a hasty argument to show that even happiness is not always a good thing. But in this case Xenophon would not only be attacking a central Socratic premiss, but an idea central to Greek ethical thinking in general. For all believe that happiness is the end to be sought (Plato, Euthd. 278 e–282 b; Sym. 205 a; Arist. Rhet. 1360B4– 7; NE 1095A18). Of course there were many disagreements about just what happiness is, but happiness itself was not questioned in this way. Surely, then, Xenophon recognized that the impression left by the argument, that not even happiness is always good, was misleading. And the precise wording of the argument makes it clear that Xenophon knew what he was doing. Note that Socrates says that happiness becomes problematic only if one adds to it popular goods such as beauty and reputation. Euthydemus cannot but add such goods to happiness, but presumably Socrates would be able to resist this temptation. So long as we do not add disputable goods to happiness, then, happiness can remain an indisputable good. Euthydemus is now at a complete loss, but Socrates is not done with him yet. Perhaps Euthydemus does not understand these matters because his false belief that he understood them precluded any attempt to investigate them (cf. Alc. I 109 e–110 d). But he has been preparing himself to enter politics in a democratic city, so surely he knows what a democracy is. As it turns out, of course, Euthydemus does not even know this, for after he identifies the demos with the poor, he cannot correctly identify who is poor (4. 2. 36–9). Once again we have a transition which makes sense from a dramatic point of view, but seems less satisfactory as far as philosophical substance goes. From a general consideration of the most essential topics of ethics we turn to a cursory examination of one political form.

Created on 4 July 2005 at 14.33 hours page 70

70

David M. Johnson

But Socrates and Euthydemus do not in fact say much about democracy. Euthydemus’ definition of democracy as rule by the poor is commonplace enough, although rather than taking the demos as only the poor he could have more liberally defined it as including all the people, rich and poor alike.49 What Socrates and Euthydemus discuss is rather what it means to be poor or rich. Socrates has already argued that wealth as it is commonly understood is not an unambiguous good (4. 2. 35). But here Euthydemus suggests a more flexible understanding of wealth: the poor are those who do not have enough to spend on what is necessary, while the rich have more than enough. Socrates now points out that some can make do with little, while others cannot make do with much. Euthydemus, agreeing, volunteers that many tyrants are forced to commit injustice by their inability to make do even with their vast resources. Socrates responds that tyrants would then have to be put with the demos, and that those who possess little, but are economical with it, would have to be put with the rich. This is, at least in Euthydemus’ eyes, absurd, and he concludes that he must know nothing. He is no more able to classify the rich and poor than he was able to classify just and unjust deeds at the outset of their conversation. But here too we have not only a failure but the elements of a positive teaching. What matters is whether or not one is oikonomikos (4. 2. 39). Socrates does not here make it clear what it takes to be oikonomikos, but elsewhere he shows that management of a household is not di·erent in kind from other sorts of management, including rule of a city, and that the true ruler is the one who knows how to rule (Mem. 3. 4, 3. 6. 14, 3. 9. 10–11; Oec. 1. 1–15, 21. 2). Thus those with good natures are those who wish to learn all forms of knowledge that allow one to run a household and a city well, and allow one in general to deal well with people and human a·airs (4. 1. 2). We can see now what this final part of the conversation adds. Prior to it, Xenophon had Socrates reject all candidates for the good as being at best ambiguous. Wisdom seemed to fare no better than wealth. But the final argument makes it clear that knowledge is more important than any other good. 49 Pericles and Alcibiades similarly define the demos as the poor at 1. 2. 42–5, but Xenophon elsewhere has Socrates define democracy as a regime in which all can hold o¶ce (4. 6. 12). Such a definition would have allowed Euthydemus to avoid the refutation he su·ers from Socrates here.

Created on 4 July 2005 at 14.33 hours page 71

Xenophon at his Most Socratic

71

Euthydemus, however, sees only his own ignorance. He goes o· greatly dispirited, despising himself and considering himself truly slavish. It is now, o· stage, that the argument about self-knowledge bears fruit. For, as it turns out, Euthydemus is not one of the many who are too lazy and stupid (βλακ2τεροι) to return to Socrates after being refuted. Instead he realizes that he must spend as much time as possible with Socrates if he is ever to make anything of himself. He follows Socrates not only in word but in deed by adopting some of his habits (4. 2. 40). He resembles, then, those who turn themselves over to self-knowers after failing themselves. He has experienced Socrates’ success in argument and his own failure, and this is enough to give him, if not full knowledge of his own potential, then at least a sober realization of how much he needs to learn and how much he needs Socrates. It is not the case, then, that Socrates’ speech about self-knowledge contributes nothing to the rest of the conversation. Rather, it explains how the victim of a Socratic elenchus can benefit from his humiliation, as Euthydemus does here.

6. Conclusion Our passage, then, for all its surface aporia and its apparent lack of order, presents some of the essential teachings of Socrates. One cannot define justice simply by looking at the outward guise of an act; one must determine whether an act is beneficial or not. Beneficial acts are those that promote happiness, and the only unequivocal means to happiness is wisdom. Xenophon does not, however, simply have Socrates inform Euthydemus of these principles. Socrates rather forces Euthydemus, and Xenophon his readers, to deduce them from a careful reading of the conversation, a far more careful reading than any that Euthydemus has previously undertaken. Xenophon makes Euthydemus his interlocutor, in place of Alcibiades, precisely to illustrate this point about the limitations and possibilities of writing. The most important text to read is oneself, and Euthydemus gains a sort of self-knowledge through his humiliating experience with Socrates, enough self-knowledge to recognize that the only way forward is to submit himself to Socrates’ guidance. Socrates, for his part, will drop the elenctic mode with Euthydemus, and more clearly reveal his views.

Created on 4 July 2005 at 14.33 hours page 72

72

David M. Johnson

Our passage is based loosely on how other Socratics portrayed Socrates’ initial conversation with Alcibiades. In fact it resembles the Alcibiades I in providing a sort of survey of Socratic philosophy. The Alcibiades I is, rather perversely, sometimes considered unworthy of Plato precisely because it is a good introduction to Socratic philosophy. Xenophon’s reputation, on the other hand, would probably benefit from the realization that he can present such an introduction. Here Xenophon has crafted a philosophical and literary whole from his disparate source material. The result does not have the literary brilliance or philosophical insight of Plato at his best, but it does have a crabbed charm, subtlety, and wit of its own. Xenophon’s range is astounding: he was a mercenary, novelist, historian, biographer, economist, and horseman. But width need not preclude depth, and Xenophon was also a worthy follower of Socrates. Southern Illinois University Carbondale

B I B L I O GR A P HY Annas, J., ‘Self-Knowledge in Early Plato’ [‘Self-Knowledge’], in D. O’Meara (ed.), Platonic Investigations (Washington, 1985), 111–38. Blundell, M. W., Helping Friends and Harming Enemies: A Study in Sophocles and Greek Ethics [Helping Friends] (Cambridge, 1989). Brickhouse, T., and Smith, N., Plato’s Socrates (Oxford, 1994). Cooper, J., ‘Notes on Xenophon’s Socrates’ [‘Notes’], in Reason and Emotion: Essays on Ancient Moral Psychology and Ethical Theory (Princeton, 1999), 3–28. Danzig, G., ‘Why Xenophon’s Socrates Was Not a Farmer: Xenophon’s Oeconomicus as a Philosophical Dialogue’, Greece @ Rome, 50 (2003), 57–76. Denyer, N., Plato: Alcibiades (Cambridge, 2001). Dittmar, H., Aischines von Sphettos: Studien zur Literaturgeschichte der Sokratiker (Bern, 1912). Dorion, L.-A., ‘Introduction’, in M. Bandini and L.-A Dorion (eds.), X‹enophon: M‹emorables, vol. i (Paris, 2000), vii–cciii. Giannantoni, G. (ed.), Socratis et Socraticorum reliquiae (4 vols.; Naples, 1990). Gill, C., Personality in Greek Epic, Tragedy, and Philosophy (Oxford, 1996). Gray, V., The Framing of Socrates: The Literary Interpretation of Xenophon’s Memorabilia [Framing] (Stuttgart, 1998).

Created on 4 July 2005 at 14.33 hours page 73

Xenophon at his Most Socratic

73

Gribble, D., Alcibiades and Athens: A Study in Literary Presentation [Alcibiades] (Oxford, 1999). Irwin, T., 1974. review of L. Strauss, Xenophon’s Socrates (Ithaca, NY, 1972), in Philosophical Review, 83 (1974), 409–13. Jo•el, K., Der echte und der Xenophontische Sokrates [Echter Sokrates] (2 vols. in 3; Berlin, 1893–1901). Johnson, D., ‘God as the True Self: Plato’s Alcibiades I’ [‘True Self’], Ancient Philosophy, 19 (1999), 1–19. ‘Xenophon’s Socrates on Law and Justice’ [‘Law and Justice’], Ancient Philosophy, 23 (2003), 255–81. Kahn, C., Plato and the Socratic Dialogue: The Philosophical Use of a Literary Form [Socratic Dialogue] (Cambridge, 1996). Maier, H., Sokrates (Tubingen, 1913). • Morrison, D., ‘On Professor Vlastos’ Xenophon’ [‘Vlastos’ Xenophon’], Ancient Philosophy, 7 (1987), 9–22. ‘Xenophon’s Socrates as a Teacher’ [‘Socrates as Teacher’], in P. Vander Waerdt (ed.), The Socratic Movement (Ithaca, NY, 1994), 181– 208. ‘Xenophon’s Socrates on the Just and the Lawful’ [‘Just and Lawful’], Ancient Philosophy, 15 (1995), 329–47. Muller, C., Die Kurzdialoge der Appendix Platonica [Kurzdialoge] (Mu• nich, 1975). Nails, D., The People of Plato (Indianapolis, 2002). Phillips, J., ‘Xenophon’s Memorabilia 4. 2’, Hermes, 17 (1989), 366–70. Robinson, T., Contrasting Arguments: An Edition of the Dissoi Logoi (New York, 1979). Scott, G. (ed.), Does Socrates Have a Method? Rethinking the Elenchus in Plato’s Dialogues and Beyond (University Park, Penn., 2002). Slings, S., Plato: Clitophon [Clitophon] (Cambridge, 1999). Strauss, L., Xenophon’s Socrates (Ithaca, NY, 1972). Vander Waerdt, P., ‘Socratic Justice and Self-Su¶ciency: The Story of the Delphic Oracle in Xenophon’s Apology of Socrates’ [‘Justice’], Oxford Studies in Ancient Philosophy, 11 (1993), 1–48. (ed.), The Socratic Movement [Socratic Movement] (Ithaca, NY, 1994). Vlastos, G., Socrates: Ironist and Moral Philosopher (Ithaca, NY, 1991). ‘The Socratic Elenchus: Method is All’, in Socratic Studies (Cambridge, 1994), 1–37. Weiss, R., ‘Ho agathos as ho dunatos in the Hippias Minor’, in H. Benson (ed.), Essays on the Philosophy of Socrates (Oxford, 1992), 242–62.

Created on 4 July 2005 at 14.33 hours page 74

page 74 is blank

Created on 4 July 2005 at 14.33 hours page 75

S OCRATES ’ DELI BERATI VE AUTHORI TARI ANI S M ANTONY HATZISTAVROU

     to the current ‘orthodox’ interpretation of the moral philosophy of Plato’s early dialogues, Socrates1 is an archetypal anti-authoritarian philosopher. He is considered to advocate individual moral autonomy and rule out subordination to the authority of another person in moral decision-making. Popper claims, for example, that Socrates wanted individuals to ‘learn to make up their minds, and to rely, critically, on their decisions, and on their insight’ on moral matters.2 Klosko ascribes to Socrates a demand that ‘the individual become morally autonomous and rule himself’ based on his belief that ‘every individual has the capacity to develop his rational faculties and be governed by them’.3 Vlastos argues that for Socrates ‘everyone of us would have the royal art, each of us would pursue the “examined life”, we would all be ruled in our individual lives by our personal knowledge and vision of the good’.4 Finally, Gulley explains Socrates’ demand that ‘the individual must be free to realise his own good’ on the basis of the importance Socrates assigns to the moral worth of the individual.5 ã Antony Hatzistavrou 2005 I have greatly benefited from discussions and comments on earlier drafts by FritzGregor Herrmann, Terry Irwin, Rachana Kamtekar, Dory Scaltsas, Malcolm Schofield, and Michael Stokes. A special debt is owed to David Sedley for his comments on the penultimate draft. My research was supported by a Fellowship at the Institute for Advanced Studies in the Humanities of the University of Edinburgh. 1 Throughout the paper, ‘Socrates’ refers exclusively to the Socrates of Plato’s early dialogues. 2 K. Popper, The Open Society and its Enemies: The Spell of Plato [Society], vol. i (London, 1962), 129. 3 G. Klosko, The Development of Plato’s Political Theory [Development] (New York, 1986), 47. 4 G. Vlastos, ‘The Historical Socrates and Athenian Democracy’ [‘Democracy’], in Socratic Studies [Studies], ed. M. Burnyeat (Cambridge, 1993), 87–108 at 105. 5 N. Gulley, The Philosophy of Socrates (London, 1968), 177. Supporters of the anti-authoritarian reading also include A. Nehamas (see his ‘Socratic Intellectu-

Created on 4 July 2005 at 14.33 hours page 76

76

Antony Hatzistavrou

In this paper I challenge the anti-authoritarian interpretation. I try to show that a demand for moral autonomy is not a basic tenet of Socratic ethics, as the supporters of the anti-authoritarian interpretation assume. I argue that for Socrates it is legitimate to subordinate oneself to someone else’s judgement in moral decisionmaking. What is more, he also believes that, unless one subordinates oneself to divine or human moral authorities, one is unlikely to reach a stable state of virtue. I ascribe to Socrates a particular principle of moral deliberation according to which in deliberating about what it is just for the agent to do, the judgements of a third party should pre-empt the agent’s own judgements. I explain that the third party may be a divine or human moral authority. But granted certain conditions, it may also be a human non-moral authority or someone with whom the agent has special moral ties, such as a parent. This principle of moral deliberation is the kernel of what I call Socrates’ ‘deliberative authoritarianism’. The paper is divided into four parts. In the first, I explore basic presuppositions and important implications of the anti-authoritarian interpretation and contrast it with central claims of my interpretation. In the second, I argue for a basic tenet of my interpretation, namely, that Socrates believes that moral expertise is attainable by humans. In the third, I try to show that he accepts that in some cases one is morally and prudentially required to subordinate oneself to divine moral authorities, some human non-moral authorities, and people or institutions one has special ties with. In the fourth, I argue that he further holds that non-experts should subordinate themselves to the authority of a human moral expert if they are to achieve a stable state of virtue.

alism’ [‘Intellectualism’], in The Virtues of Authenticity [Authenticity] (Princeton, 1999), 27–58) and R. Weiss (Socrates Dissatisfied: An Analysis of Plato’s Crito [Dissatisfied] (Oxford, 1998), esp. chs. 2 and 4). I also take R. Kraut’s account of Socrates’ moral philosophy (Socrates and the State [State] (Princeton, 1984), ch. 7) ultimately to provide support for the antiauthoritarian reading. According to Kraut, Socrates believes that, if there were moral experts, one should obey them. But Kraut argues that Socrates also believes that moral expertise is not attainable by humans. On Kraut’s interpretation, Socrates opts for intellectual freedom and autonomous moral decision-making as the only viable conditions for advancement in virtue. But then the unattainability of moral expertise must be a conclusive reason for one to rely on one’s own judgement about moral issues. As I explain in the next section, this is precisely what the moralautonomy tenet requires.

Created on 4 July 2005 at 14.33 hours page 77

Socrates’ Deliberative Authoritarianism

77

I On the anti-authoritarian interpretation there is a sharp contrast between the Socrates of the early and the middle dialogues. The latter is considered to advocate the subordination of the people to the moral authority of the philosopher-kings in the perfect city and suggest an educational system which does not promote the development of critical thinking in the larger part of the population. By contrast, the former is considered to demonstrate, through his continuous elenctic examinations of any self-proclaimed moral authorities and of his fellow citizens, that all individuals need to develop their critical thinking and engage in autonomous moral decision-making.6 On the anti-authoritarian reading moral autonomy is a necessary condition of moral development for the Socrates of the early dialogues. Socrates rules out acquisition of virtue through indoctrination. He does not believe that one can become virtuous by being dogmatically taught by a moral authority what is right and wrong, as one may be dogmatically taught a certain craft.7 He believes that one should discover what is right and wrong by oneself, through the kind of critical examination provided by the elenchus. As Klosko puts it, ‘[Socrates] wanted people to pursue certain values and he wanted them to do so in a certain way. It would not have been enough for Socrates if a person were able to pursue the values of the soul, unless he did so as the result of a process of rational deliberation and choice.’8 Popper relates Socrates’ rejection of moral indoctrination to his general scepticism of all ‘professional learnedness’.9 In a similar vein, Klosko speaks of Socrates’ ‘provisionality’, which made him seek constant re-examination of every moral belief.10 Vlastos does not adopt this rather sceptical portrait of Socrates.11 He believes that according to Socrates a type of moral knowledge is attainable by humans. It is elenctic knowledge, i.e. knowledge of moral truths 6 The contrast between the liberal Socrates and the authoritarian Plato is a basic theme of two book-length analyses of the political philosophy of the Platonic dialogues: Popper, Society, and Klosko, Development. 7 Popper, Society, 129. 8 Klosko, Development, 31. 9 Popper, Society, 128. 10 Klosko, Development, 32–4. 11 See his ‘The Socratic Elenchus: Method is All’, in Vlastos, Studies, 1–37, and ‘Socrates’ Disavowal of Knowledge’ [‘Disavowal’], ibid. 38–66.

Created on 4 July 2005 at 14.33 hours page 78

78

Antony Hatzistavrou

justified through the elenchus. This elenctic knowledge is the ‘royal art’ which, as we have seen, Vlastos thinks that everyone can possess according to Socrates. It is achieved through the correct exercise of the elenchus, which requires the development of the agent’s critical thinking. As Vlastos puts it, ‘[Socrates is] a teacher who shuns didacticism, believing that moral truth has a dimension which eludes direct expression—a depth best revealed not by instruction . . . to teach is to trigger in a learner an autonomous learning process’.12 On the anti-authoritarian interpretation, Socrates’ rejection of moral indoctrination is intertwined with his rejection of subordination to someone else’s authority on moral issues. As Klosko puts it, ‘Socrates accepts beliefs only if he is able to defend them. He demands the rigorous examination of all convictions, while convictions based solely on authority are not worthy of consideration as such.’13 And of course it rules out the use of force as a means of making people virtuous.14 It is clear from the above that on the anti-authoritarian reading the value Socrates ascribes to moral autonomy is proportionate to the value he ascribes to his elenchus. Socrates appears to recognize no reliable method for advancement in virtue other than his elenctic examination, from which no one is excluded and in which everyone argues freely. He sums up the supreme value of the elenchus in his well-known maxim that the unexamined life is not worth living (Ap. 38 a 5–6). Thus, we can say that on the anti-authoritarian reading a specific tenet is ascribed to Socrates which I shall call the Moral Autonomy Tenet: (MAT) There are conclusive normative reasons (prudential and/or moral) for everyone to enjoy a life of moral autonomy, i.e. a life in which everyone’s moral decisions are autonomous. The notion of moral autonomy which the supporters of the antiauthoritarian reading use should be understood in contradistinction to the notions of power and authority: An agent is morally autonomous i· her moral decisions and actions are the result of neither the exercise of force by a third 12 Vlastos, ‘Disavowal’, 65. 14 Ibid. 48.

13 Klosko, Development, 32–3.

Created on 4 July 2005 at 14.33 hours page 79

Socrates’ Deliberative Authoritarianism

79

party nor the subordination of the agent’s judgement to the judgement of a third party. The contradistinction between autonomy and authority deserves further consideration. We may distinguish two types of subordination to someone else’s judgement. In strong subordination, one forms an intention to act in accordance with someone else’s judgement without considering for oneself the pros and cons of the prescribed course of action. This is what happens, for example, when I follow the medical prescription of my GP without having also made my own medical diagnosis. Strong subordination both requires the formation of an intention to act on the basis of someone else’s judgement and excludes the agent’s balancing of the reasons for and against the prescribed course of action. By contrast, in weak subordination the agent may form a judgement about the value of a certain course of action but form an intention to act not on the basis of her judgement but on the basis of the judgement of someone else. For example, a medical student may make her own medical diagnosis about a certain case but make a prescription on the basis of the judgement of the medical doctor. Weak subordination requires only that the agent forms an intention to act on the basis of someone else’s judgement; it does not preclude the agent’s forming her own judgement about the balance of reasons for and against the prescribed course of action. It prescribes obedience when the agent disagrees with what she is ordered to do. But it does not require that the agent should blindly follow the person in authority and totally abdicate her own reasoning.15 The supporters of the anti-authoritarian interpretation do not distinguish between strong and weak subordination. There is evidence that what they take to be Socrates’ demand for autonomy is incompatible even with weak subordination. Consider, for example, the emphasis they put on Socrates’ rejection of moral indoctrination, on his sceptical attitude towards any learned authorities, and on his general denial of the value of relying on someone else’s authority on moral issues; and, most importantly, their understanding of the elenchus as a method of teaching people how to make up their 15 For a distinction similar to that between strong and weak subordination, see J. Raz, The Morality of Freedom [Freedom] (Oxford, 1986), 38–42; cf. the notion of surrendering private judgement developed by R. B. Friedman in ‘On the Concept of Authority in Political Philosophy’, in J. Raz (ed.), Authority (Oxford, 1990), 56–91 at 63–8.

Created on 4 July 2005 at 14.33 hours page 80

80

Antony Hatzistavrou

own minds about moral issues. They do not take Socrates’ message to his fellow citizens to be simply that they should try to enhance their critical understanding of virtue. They believe that Socrates further claims that his fellow citizens should never follow a certain course of action unless they have been convinced of its justness, relying on nothing else than the authority of their own reasoning. So, we may say that on the anti-authoritarian reading MAT entails a particular Principle of Moral Deliberation which is incompatible with weak subordination: (PMD) In her deliberations about what is just for her to do the agent should rely exclusively on her own judgement. In this paper I shall try to show that Socrates espouses neither MAT nor PMD. My main disagreement with the anti-authoritarian interpretation concerns the thesis that Socrates denies weak subordination to someone else’s authority in the moral realm. Respect for moral authority is a central feature of Socrates’ moral philosophy and the anti-authoritarian interpretation unduly neglects it. It over-interprets the emphasis Socrates puts on the need for a critical understanding of virtue. True, Socrates takes constant elenctic examinations of one’s moral beliefs to be of supreme value and to contribute to one’s moral development. But he does not take the extra step of postulating that only moral beliefs which have been elenctically tested to the satisfaction of the agent should enter into her deliberations about how to act. He accepts that, for example, the moral beliefs of someone morally superior to the agent should not only be taken into account but also pre-empt the agent’s relevant judgements. In so far as the moral development of his fellow citizens is concerned, Socrates’ message is ‘try critically to understand virtue but obey your moral superior’. Or so I shall argue. Central to my argument is the thesis that Socrates distinguishes a level of moral expertise which is higher than the level of Socrates’ elenctic virtue and which is in principle attainable by humans. It is primarily to the authority of this moral expert that the non-experts should subordinate themselves, if they are to achieve a stable state of virtue. In the second section of this paper I provide an account of the basic features of moral expertise and argue that Socrates believes it to be attainable by humans. Before proceeding, let me state one thing I shall not do in this paper. I intend to distinguish the issue of Socrates’ authoritarianism

Created on 4 July 2005 at 14.33 hours page 81

Socrates’ Deliberative Authoritarianism

81

from the issue of Socrates’ attitude to democratic values in general or his attitude to the Athenian democracy in particular. I take the issue of Socrates’ authoritarianism to be the issue of whether he holds MAT and PMD. Further arguments are needed, however, to settle the issue of whether Socrates adheres to democratic ideals. For example, he could hold MAT and PMD, but support anarchy instead of democracy. On the other hand, Socrates could reject MAT and PMD but still consider the Athenian democracy to be the best political institution. So, although I shall argue in this paper that Socrates is an authoritarian philosopher, I shall not assume that his authoritarianism is conclusive evidence that he is an antidemocrat in general or an opponent of the Athenian democracy in particular. Nor shall I address these issues here.16

II In the early dialogues Socrates refers to the persona of the moral expert and describes some of his basic characteristics. There are explicit references to this persona in Laches, Crito, Protagoras, Gorgias, and Republic book 1. In the Laches Socrates refers to the only person who can really educate men in virtue and to whom one should entrust the moral training of one’s children (184 e 11–185 a 3, 201 a 2–6). In the Crito he invokes the persona of the man who possesses knowledge of what is just and what is unjust (48 a 5–6) and whose directives one should follow if one’s soul is to be saved (47 d 4–7). In the Protagoras he speaks of the ‘doctor’ (
ατρικς) of the soul who knows what lessons are good and harmful for one’s soul (313 d 5–e 2). In the Gorgias he refers to the person who possesses the real political art which should rule over all other crafts that care for the soul (464 b 1–465 d 6 and 517 c 7–518 c 1) and aims at making the citizens as virtuous as possible (504 d 5–e 4). In Republic book 1 he similarly speaks of the true ruler who acts out 16 Popper, Klosko, and Vlastos (in the works cited in nn. 2–4) take Socrates’ alleged anti-authoritarianism as evidence that he is a theoretical and/or practical supporter of democracy. For an interpretation of Socrates as a theoretical and practical opponent of democracy, see A. E. Taylor, Socrates (Edinburgh, 1932), 149–52; E. M. Wood and N. Wood, ‘Socrates and Democracy: A Reply to Gregory Vlastos’, Political Theory, 14 (1986), 143–53; I. F. Stone, The Trial of Socrates (London, 1988). For an illuminating account of the current debate about Socrates’ political stance, see M. Schofield, ‘Socrates on Trial in the USA’, in T. P. Wiseman (ed.), Classics in Progress (Oxford, 2002), 263–83.

Created on 4 July 2005 at 14.33 hours page 82

82

Antony Hatzistavrou

of concern for the well-being of the citizens (347 d 4–6). The moral expert has the following characteristics: (1) He possesses moral knowledge. He is called by Socrates παων (Crito 47 d 2 and 48 a 6; cf. Gorg. 508 c 2 πιστµονα τν δικαων), and τεχνικς (La. 185 a 1, b 11, and e 4; cf. Gorg. 504 d 6). He does not simply have true moral beliefs; he is able to o·er a rational account of them (Gorg. 465 a 2–6). (2) The moral expert has roughly three types of moral knowledge: (a) knowledge of what virtue is, (b) knowledge of what actions and persons are virtuous, and (c) knowledge of what properties virtue has (e.g. whether it is teachable or not). At La. 189 e 3–190 c 6 Socrates argues that the moral expert, if he is actually to make people virtuous, must possess knowledge of what virtue is. His argument rests on an analogy with the ophthalmologist: as one cannot cure the eye unless one knows what sight is, one cannot make men virtuous unless one knows what virtue is.17 Possession of knowledge of what virtue is appears to be the defining feature of moral expertise. For, according to Socrates, if one possesses knowledge of what virtue is, then one also possesses knowledge of both what actions and persons are virtuous and what properties virtue has. In the Euthyphro Socrates claims that, if he knows what piety is, he can use it as a standard to determine what is pious and what is not (6 e 4–7). Similarly, at Prot. 360 e 8–361 a 3 he claims that, if it comes to light what virtue is, it will become perfectly clear (κατδηλον) whether it is teachable or not. (3) Only the moral expert is capable of teaching virtue. Socrates believes (a) that the possession of knowledge of what virtue 17 At La. 189 e 3–190 c 6 Socrates puts forward the following two principles: (P1) If A (1) knows that, if x is joined to y, y becomes better, and (2) knows how to join x to y, then A (3) clearly knows what x is and (4) can give valuable advice about how B can get x in the best and easiest way. (P2) If A does not know what x is, then A cannot give valuable advice about how B can get x in the best and easiest way. He illustrates both principles by o·ering ‘sight’ for x and ‘eye’ for y. It is clear that Socrates considers (1) and (2) in (P1) to be evidence for (3). (P2) also makes it clear that he considers (3) to be a necessary condition for (4): that is, knowledge of what x is is necessary for teaching someone how to get x. Both principles apply to the case of moral training (we simply need to substitute ‘virtue’ for x and ‘soul’ for y in (P1) and (P2)). So, only if one knows what virtue is will one be able to teach virtue.

Created on 4 July 2005 at 14.33 hours page 83

Socrates’ Deliberative Authoritarianism

83

is is a necessary condition for teaching virtue; (b) that the defining feature of the moral expert is knowledge of what virtue is; and (c) that the moral expert can actually teach virtue. This is the reason why he urges his interlocutors in the Laches to entrust the moral training of themselves and their sons to the moral expert (201 a 2–b 5; cf. 184 e 12–185 a 3). Socrates must therefore believe that the moral expert is the only person who can teach virtue. (4) The moral expert possesses a superordinate craft which should rule over all other crafts which aim at providing what is good for the soul. In the Gorgias Socrates utilizes an analogy with medicine and gymnastics. As the latter should rule over the crafts which care for the body, so the true political art should rule over the crafts which care for the soul (517 c 5–518 c 1). What endows these crafts with the right to rule over the subordinate crafts is the fact that only they have knowledge of what is good for the body and the soul respectively and are able to provide an explanatory account of their subject matter (465 a 2–6 and 500 e 4–501 c 5). (5) The moral expert can successfully pass the test of Socrates’ elenctic examination. In the Apology Socrates claims that the main objective of the elenchus is to test other people’s claims to wisdom (23 b 4–c 1). In the Laches he employs the elenchus to test whether he and his interlocutors can actually be teachers of virtue. As we have already seen, Socrates believes that to be a teacher of virtue one needs to have knowledge of what virtue is, possession of which is a defining feature of moral expertise. What the elenchus will test in particular is whether he and his interlocutors have knowledge of what a part of virtue, courage, is (190 b 7–d 1). Socrates suggests that, if the elenchus showed them to know what courage is, they could be teachers of virtue (189 e 1–3) and thus moral experts. Does Socrates believe that he is a moral expert? The following considerations suggest that he does not. First, he denies that he has taught virtue. For example, in the Apology he claims that he has never taught anyone (33 a 5–6). Secondly, he denies that he could legitimately attempt to teach virtue. In the Laches he does not agree with Nicias’ suggestion that he is most qualified to teach

Created on 4 July 2005 at 14.33 hours page 84

84

Antony Hatzistavrou

virtue (200 c 7–d 4). He believes that he fails to meet either of the two conditions, either one of which is su¶cient for teaching virtue. These conditions are: (1) having been taught virtue by someone else: Socrates denies that he ever had a teacher of virtue (La. 186 c 1 and e 2); (2) having found virtue by oneself: Socrates denies that he has found virtue by himself (La. 186 e 1). Socrates equally denies that he possesses knowledge of what virtue is, which, as we have seen, he considers to be a necessary condition for teaching virtue. He disavows knowledge of what virtue as a whole is at Prot. 360 e 6–361 d 6. The sincerity of his disavowal of knowledge there is corroborated by the fact that he also disavows knowledge of what courage is at La. 199 e 11, of what friendship is at Lys. 223 b 5–7, of what justice is at Rep. 1, 354 a 12–c 3, and of what temperance is at Chrm. 175 c 9–d 6. Thus, Socrates does not think that he is a moral expert. He acknowledges that he is knowledgeable in the elenctic craft (τχνης, Euthph. 11 d 6–7; cf. Ap. 30 e 1–31 b 5) which allows him critically to examine moral issues. But being knowledgeable in the elenchus di·ers from having moral expertise in at least one important respect. Although Socrates possesses the elenctic craft, he cannot teach virtue. In this respect the level of virtue Socrates has attained through the elenchus is inferior to the level of virtue a moral expert would have. In Socrates and the State Kraut claimed that Socrates does not consider moral expertise to be attainable by humans. His interpretation rests on the false assumption that for Socrates the only type of knowledge human beings may achieve is knowledge of the limited value of human wisdom.18 Kraut relies on the evidence of Ap. 23 a 5–c 1: It appears, gentlemen, that the god is truly wise and this is what he means in his oracle, that human wisdom has little or no value at all. He further seems to be referring to ‘this Socrates’, using my name in addition, in order to make me an example, as if he was saying ‘This one of you, men, is wisest who like Socrates knows that he is truly worthless in respect of wisdom.’ 18 Kraut writes: ‘the only wisdom a human being can attain by his natural abilities is the recognition of how small his wisdom is’ (State, 291). Cf. T. Penner, ‘Socrates’, in C. Rowe and M. Schofield (eds.), The Cambridge History of Greek and Roman Political Thought (Cambridge, 2000), 164–89 at 172–3.

Created on 4 July 2005 at 14.33 hours page 85

Socrates’ Deliberative Authoritarianism

85

The main point Socrates puts forward here is that the wisest man is someone who, like himself, knows that, in comparison to the god,19 he is truly worthless in respect of wisdom. Kraut takes the realization that one’s wisdom is worthless in comparison to the god’s to be the sole content of human wisdom. But this interpretation is not supported by the text. Socrates does not take the god to reveal the conditions one should satisfy to be wise but the conditions one should satisfy to belong to the class of the wisest (σοφτατος, 23 b 2). This leaves open the possibility that there may be di·erent types of human wisdom. Awareness of the comparative humbleness of human cognitive achievements may be the most important of them but not the only one. And indeed Socrates allows that there are other forms of human wisdom. For example, he accepts that craftsmen possess a type of wisdom and are in some respects wiser than he is (22 c 4–d 3). And, as we have seen, he considers himself to be wise in the elenchus (Euthph. 11 d 5–6). So, contra Kraut, Socrates in the Apology allows that humans can attain types of wisdom other than the awareness of how unimportant their wisdom is in comparison to the god’s knowledge. Moral expertise may be one such type of wisdom. But if moral expertise is attainable by humans and thus may constitute part of human wisdom, will it still be true that awareness of the comparative humbleness of human cognitive achievements is the highest type of human wisdom? And if a moral expert is found, will Socrates continue to count as the wisest man? I think that we should understand the god’s message to be valid only for the Athens of Socrates’ day, in which there is no moral expert. On this interpretation the Socratic wisdom that the god praises is the highest form of wisdom which has actually been attained by any human in Socrates’ time; but it may not be the highest form of knowledge that humans are by nature capable of achieving. Moral expertise may be higher than Socratic wisdom and a moral expert may be wiser than Socrates.20 19 Clearly Socrates’ claim that human wisdom has little or no value is not to be understood in absolute terms but comparatively to divine wisdom. Socrates cannot have meant that his wisdom in elenchus (Euthph. 11 d 5–6) has little or no value, given that he considers the unexamined life to be not worth living (Ap. 38 a 5–6). 20 A corollary of this interpretation is that by νθρωπνη σοφα Socrates does not mean the type of wisdom humans are by nature capable of achieving, but rather the type of wisdom the ordinary contemporary of Socrates can achieve. Support for this reading of νθρωπνη is found at Ap. 31 a 3–7. Socrates claims that it will

Created on 4 July 2005 at 14.33 hours page 86

86

Antony Hatzistavrou

There are further problems with Kraut’s thesis that for Socrates moral expertise belongs exclusively to the domain of divine knowledge.21 This is hard to square with the advice Socrates gives to his interlocutors in the Laches to continue their search for a moral expert and his suggestion at the end of the Protagoras (361 c 1–d 6) not be easy for the Athenians to acquire someone like Socrates in the future. For his devotion to the task of making others care about virtue at the expense of all his private and family a·airs appears to be non-human (ο γ ρ νθρωπν!ω, 31 a 10–b 1). Presumably, Socrates’ point is not that such a gross neglect of one’s own a·airs is an attitude which falls beyond the psychological limits of human nature. His own case is evidence to the contrary. It is important to note that Socrates does not claim for himself the status of a divine superhuman creature; he claims only to be a gift of the god, and this does not make him anything other than human. And he does not say that it is impossible for someone else to adopt a similar attitude; he claims that it is not easy (ο "#αδως, 31 a 3) to do so. Thus, Socrates’ point must be that it is not typical of the average man to neglect his own a·airs as Socrates does (and in fact his experience has shown that none of his contemporaries has exhibited such an extreme neglect of their own a·airs). Similarly, when he rules out moral expertise from the sphere of human wisdom, he does not mean that moral expertise is beyond what humans are by nature capable of achieving. He means that moral expertise is not typical of what the ordinary person can achieve (and in fact his experience has shown that none of his contemporaries has achieved it). 21 Kraut’s evidence is Socrates’ claim at Ap. 20 d 8–e 2 that the wisdom advocated by the sophists is greater than human. Socrates claims only human wisdom for himself. Kraut seems to assume that the type of knowledge Socrates takes to be greater than human is knowledge of ‘human and political virtue’ (20 b 5) and that by ‘greater than human’ wisdom he means divine wisdom. But it does not seem to be the case that Socrates wants to deny here that humans may attain knowledge of ‘human and political virtue’ as such. He rather wants to ridicule the sophists’ ‘recipe-like’ method of teaching virtue. He finds bizarre their claims that they can go into a city and pass on their knowledge of virtue to young men with whom they engage for a short period of time and whom they persuade to give them money for their lessons (19 e 4–20 a 2). The clear implication is that people who behave like that are most likely to be fraudulent. The true moral expert does not teach virtue in this manner, nor does he have profit as his primary motivation. He keeps the citizens under constant supervision, trying to change their appetites by various means and to turn them into good citizens (Gorg. 517 b 5–c 2). He also acts out of concern for the well-being of his fellow citizens and does not exercise a profit-making craft (Rep. 1, 345 b 7–e 4, 347 b 5–e 7). Socrates’ point at 20 d 8–e 2 does not seem to be that the sophists are not reliable because the knowledge they claim to teach is beyond the knowledge humans can by nature achieve. His point is rather that it is not possible for one to teach virtue in the way and for the reasons they claim to do it (thus implying that they do not actually teach virtue). Similarly, when he speaks of ‘greater than human wisdom’, we should not take him to mean that the type of knowledge the sophists claim to have belongs only to the gods. His ironic point is rather that, if the sophists were truly able to impart true knowledge of virtue to their pupils in the way and for the reasons they claim to do it, their ability would have been beyond the sphere of the abilities of the average person; whereas Socrates claims for himself the type of wisdom the average person can achieve (see the analysis of human wisdom in the previous note).

Created on 4 July 2005 at 14.33 hours page 87

Socrates’ Deliberative Authoritarianism

87

that the quest for a definition should not stop. At La. 184 e 11–185 a 3 he suggests that, if neither he nor his interlocutors were proved to be moral experts, they should try to find someone else who is ($λλον τιν ζητε&ν). And he gives the same advice at the end of the Laches after he has shown that neither he nor his interlocutors were moral experts (201 a 3–4). He does not suggest that his interlocutors should decide for themselves about how to educate the young or that they can hope for nothing better than the Socratic elenchus. On the contrary, he advises against both policies (200 e 3–6) and urges them to intensify their search for the moral expert without sparing money or any other goods (201 a 5–6).22 If Socrates valued nothing higher than moral education and virtue, then why would he set his interlocutors on a futile search for a moral expert whom he does not believe to exist among humans?23 Furthermore, if the function of the god’s message at Ap. 23 a 5–c 1 were to reveal that moral expertise is beyond human limits and is instead the god’s monopoly, attempts to find moral definitions would be instances of hubris. Socrates, by encouraging the quest for moral definitions, would encourage people to trespass in the divine sphere of knowledge and promote intellectual vanity. It is safer, I think, to assume that Socrates genuinely believes that 22 It could be argued that the Laches represents a stage at which Socrates was still testing the oracle and had not concluded that moral expertise belongs only to god. But Socrates could not have given advice to others to look for a moral expert if he was still testing the oracle. As I explain in the next section, Socrates never doubted the truth of the oracle: he was not assuming that it was false. He was exceedingly perplexed about its meaning; but if he were in such a state of perplexity, he would know neither whether there were any moral experts among his contemporaries nor whether moral expertise was within human capacity. And if he lacked such knowledge, he would have abstained from giving any relevant advice to others. I therefore believe that the fact that he feels confident to give advice to others is evidence that he has solved the puzzle of the oracle. 23 Kraut also relates Socrates’ alleged belief in the impossibility of moral expertise with Socrates’ arguments for the unteachability of virtue at Prot. 319 a 8–320 c 1 and Meno 96 c 1–100 c 2. But in the Protagoras the argument for the unteachability of virtue rests on the ironic premiss that the Athenians are wise. Further, Socrates’ claim that virtue is unteachable is retracted at the end of the dialogue. In the Meno Socrates adduces as evidence for his claim that virtue is not teachable the examples of famous politicians of the past to whom he explicitly denies possession of knowledge of what virtue is (see Meno 99 b 11–100 b 4; cf. Gorg. 515 c 4–517 a 6). This makes it di¶cult to assume with Kraut that Socrates considers the failure of those who do not know what virtue is to teach virtue to be evidence for the fact that virtue is not teachable. For accounts of the argument in the Meno see D. Devereux,‘Nature and Teaching in Plato’s Meno’, Phronesis, 23 (1978), 118–26; K. Wilkes,‘Conclusions in the Meno’, Archiv f•ur Geschichte der Philosophie, 61 (1979), 143–53; T. Irwin, Plato’s Ethics (Oxford, 1995), 140–1.

Created on 4 July 2005 at 14.33 hours page 88

88

Antony Hatzistavrou

moral expertise is within human capacity, sincerely recommends the quest for it, and does not understand the god’s message to imply that this quest is futile. He has not found any moral expert among his contemporaries and does not think of himself as a moral expert, but he believes that this type of wisdom is not beyond the grasp of humans.

III I shall now turn to the issue of whether Socrates believes that it is legitimate to follow someone else’s judgements on moral issues. I shall argue that Socrates finds the following types of subordination to someone else’s judgement on moral issues legitimate (with some qualifications, which I shall explain later): (1) subordination to divine authorities; (2) subordination to non-moral human authorities; (3) subordination to people or institutions one has special ties with; (4) subordination to the moral expert. Let me point first to the evidence for (1) and (2). At Ap. 29 b 5–7 Socrates claims that disobedience to one’s superior (βελτονι), whether divine or human, is (prudentially) bad for one24 and shameful. Socrates’ claim is intended as a refutation of a potential charge that he should feel shame because by insisting on continuing his elenchus (despite the public outrage against it) he puts his life at risk (28 b 4–6). Socrates’ immediate response is that this charge rests on the false assumption that avoidance of deadly risks should be the primary concern of the agent. He claims that instead one should always consider only one thing when one acts, namely, whether one acts justly and in the way a good person would act, without fearing for one’s life (28 b 8–9). He illustrates this normative principle with the example of Achilles.25 Achilles, although he knew that his own death would inevitably follow Hector’s, decided to follow the 24 The context of Socrates’ argument makes it clear that κακν at 29 b 7 is used in a prudential rather than a moral sense (cf. M. C. Stokes, Plato: Apology [Apology] (Warminster, 1997), 146). 25 The fact that he considers Achilles’ example as an illustration of his normative principle is evident from the way he introduces the example: ‘For this is indeed, Athenians, the truth of the matter’ (28 d 5–6).

Created on 4 July 2005 at 14.33 hours page 89

Socrates’ Deliberative Authoritarianism

89

order of justice and scorn death (28 c 6–d 5). Socrates then distinguishes two possible ways in which his normative principle should be observed in a military context: (1) A commander like Achilles should not out of fear for his life desert the post he has decided for himself to hold in a battle. (2) A soldier should not leave the post he has been commanded to hold out of fear for his life (28 d 5–9).26 Socrates applies the normative principle to his own case. He claims that it would have been terrible if he had obeyed the Athenian commanders in the battle and held his post, but failed to obey the god’s oracle to continue philosophizing out of fear for his life (28 d 10–29 a 2). What is shameful is not to put his life at risk, but to disobey the god or indeed any divine or human ‘superior’ (βελτονι, 29 b 6). Socrates takes his elenctic mission to be a primary example of subordination to divine authorities. He considers this mission to be a service (λατρεαν) to the god (23 b 6–c 1; cf. (πηρεσαν, 30 a 6–7). He understands the god to have ordered him (τττοντος, 28 e 5; cf. κελε)ει, 30 a 5, προσττακται, 33 c 5) to exercise his elenchus and considers himself to be the god’s soldier, who will not desert his post (28 e 4–29 a 2). He insists that he will obey (πεσοµαι) the god rather than the Athenians and will not cease his philosophical activity (29 d 2–4). He further considers his obedience to the god’s oracle to be evidence which corroborates his case that he believes in gods: if he disobeyed, the Athenians would have every right to bring him to trial for atheism (29 a 2–4). What the god has ordered Socrates to do is of supreme moral value. The aim of his elenctic mission is to make his fellow citizens care about the most valuable things (30 a 1–2): namely, all those things related to the virtue of one’s soul (30 a 7–b 5). Thus, we may with confidence say that Socrates finds subordination to a divine authority concerning moral issues of the highest importance perfectly legitimate.27 In the Apology Socrates’ compliance with the god’s oracle is not 26 I therefore disagree with A. D. Woozley, Law and Obedience (London 1979), 47–9, and Weiss, Dissatisfied, 8–11, who think that the soldier too keeps the post he thinks it best to hold. This misrepresents the logical sequence of Socrates’ argument. For another criticism of a position like theirs, see C. D. C. Reeve, Socrates in the Apology [Apology] (Indianapolis, 1989), 100. 27 The oracles are not the only means through which divine authorities express

Created on 4 July 2005 at 14.33 hours page 90

90

Antony Hatzistavrou

the only behaviour which he describes as an instance of obedience to a divine authority. He also claims to have obeyed his daimonion, which he considers to be something divine (θε&ον, 31 c 8), a sign (σηµε&ον, 40 b 2) from the god, when it opposed his involvement in politics. His daimonion is an internal voice (φων, 31 d 2) which objects to some actions without ever recommending to Socrates any specific course of action (31 d 3–4). This internal voice has been operating since his childhood (31 d 2) and has quite often shown its opposition to possible actions even in matters of small importance (40 a 5–6). Socrates has significant epistemic trust in his daimonion. He considers the absence of opposition by it to any of his actions or arguments during his trial as strong evidence (µγα τεκµριον, 40 c 2) that he has been doing the right thing during his trial (40 b 7–c 4). Socrates’ attitude to both the god’s oracle and his daimonion is that of weak subordination. He epistemically trusts the god and his daimonion and does not believe that they try to deceive him. He therefore obeys them as a soldier obeys his commander and acts on their judgement. He acts on the god’s judgement when he continues to philosophize despite the menace of the Athenians, consistently describing his persistence in his elenctic mission as an act of obedience to the god. And he refuses to engage in politics following his inner daimonic voice. But this does not mean that he follows their authority blindly and does not rationalize their commands. He can produce arguments to the e·ect that his elenctic mission is the greatest good that has come to the city (30 a 5–7). And he can infer that one who wants to be just will not last long if he engages in politics (32 a 1–3). My interpretation of Socrates’ views about the legitimacy of weak subordination to divine authorities may face the following objection. On this objection, which is based on Nehamas’s interpretation,28 Socrates’ following the oracle is not an act of obedience. Nehamas claims that Socrates follows the oracle after he has been convinced of its truth.29 Consequently, Socrates behaves as a their will. Socrates claims to have been ordered to philosophize via dreams as well (Ap. 33 c 4–8). He finds subordination to a divine authority legitimate irrespectively of the means by which the god’s message has reached him. 28 Nehamas, ‘Intellectuallism’, 43–5. 29 And so Nehamas concludes that ‘[Socrates] only does, as he always has done, what he thinks is, on independent grounds, the best thing’ (‘Intellectualism’, 44).

Created on 4 July 2005 at 14.33 hours page 91

Socrates’ Deliberative Authoritarianism

91

morally autonomous human being and his own example suggests that everyone should behave in the same manner. Nehamas seems to make two assumptions. The first is that Socrates began his elenctic mission in order to test the truth of the oracle, to find out whether the god was telling the truth or not. This assumption is apparently supported by what Socrates says at 21 b 9–c 2. Socrates claims that he first decided to test someone famous for his wisdom. He thought that this was his best chance to put the oracle to the test ( λγξων), by showing that there was a person wiser than Socrates whereas the oracle had claimed that Socrates was the wisest of all men. The way Socrates describes the outcome of the relevant elenchus might create the impression that he wants to prove the god’s oracle wrong: ‘I will show the oracle: “This man is wiser than me, you said that I am”’ (21 c 1–2). But a couple of lines earlier Socrates has said that the god presumably (δπου) does not speak falsely (ψε)δεται) and that it is not lawful or permissible for him to speak falsely (21 b 6–7). ψε)δεται is ambiguous here between ‘lying’ and ‘saying something false due to ignorance’, but this ambiguity is irrelevant to the main point Socrates wants to convey. The way Socrates conducts his elenchus of the oracle shows that he does not sincerely doubt but rather presumes its truth or truthfulness. He has the opportunity of claiming an easy victory over it. His elenchus reveals that craftsmen are in some respects wiser than he (22 c 9–d 4). But Socrates does not treat this as evidence that the god lies, or at least is mistaken. He does not end his elenchus at this point and claim to have produced a counter-example to the oracle along the lines of the passage at 21 c 1–2. Instead, Socrates continues his elenchus looking for a more charitable interpretation of the oracle. And he is not always charitable to his interlocutors’ views in his elenchi.30 It is plausible to assume that what moves Socrates to apply the principle of charity here is his belief that the god does not speak falsely (either because he lies or because he does not know the truth) since it is not permissible for him to speak falsely. So it is better to think of Socrates’ elenchus of the oracle not as intending to refute but as intending to interpret the oracle. Socrates believes that the god speaks the truth (i.e. the god neither lies nor is mistaken). But what the god says is not clear. So Socrates tries to come up with a charitable interpretation of the oracle, and he uses 30 See e.g. the uncharitable account of Thrasymachus’ definition of justice at Rep. 1, 338 c 4–d 2.

Created on 4 July 2005 at 14.33 hours page 92

92

Antony Hatzistavrou

his elenchus to this end not with the intention of demonstrating the falsity of the oracle but with a desire to understand its meaning.31 The second assumption Nehamas makes is that Socrates does not, strictly speaking, ‘obey’ the oracle since he has been convinced of its truth. To support his interpretation he suggests that πειθε&ν at 29 b 6–7 is not properly translated as ‘disobeying’. He exploits the relation between πεθω and its cognates with the notion of persuasion and argues that it should be translated as ‘not having been convinced’. He takes Socrates to claim at 29 b 6–7 that he knows that it is bad and shameful to remain unconvinced by a divine or human moral expert. But it is unlikely that this is the proper meaning of πειθε&ν at 29 b 6–7. As we have already seen, this passage has been introduced with the analogy of the soldier’s obedience to the general, which suggests that Socrates understands himself to be the soldier of the god. Further, Socrates consistently describes the god as having given an order (28 e 5, 30 a 5, and 33 c 5) and not as having produced an argument that persuades Socrates.32 What is important is the conceptual point that underlies Nehamas’s interpretation. Nehamas seems to think that since Socrates has been convinced of the truth of the propositional content of the oracle, he cannot obey the oracle. But why not? There is no incompatibility, for example, between my obeying the law that prescribes that one should not use drugs and my being convinced that taking drugs is a bad thing. I mean that I may still obey the law because it is a law and may not take drugs because the law forbids it. Of course, I would still not have taken drugs even if there had been no law against it, since I believe that taking drugs is bad. But the existence of a law does make a di·erence to my deliberations, and I may abstain from taking drugs because the law says so and not because I believe that it is bad. In other words, to determine whether I recognize the authority of the relevant law instead of acting on my own judgement, one has to look at what causes or sustains my intention to abstain from drugs. Recognizing the authority of and obeying the relevant law do not 31 For similar arguments in defence of the claim that Socrates intends to interpret and not refute the oracle, see Reeve, Apology, 21–8, and M. L. McPherran, ‘Elenctic Interpretation and the Delphic Oracle’ [‘Oracle’], in G. A. Scott (ed.), Does Socrates Have a Method? (University Park, Penn., 2002), 114–44. 32 Nehamas in a subsequent publication (‘What Did Socrates Teach and to Whom Did he Teach it?’, in Authenticity, 59–82 at 63–4) appears to retract this interpretation.

Created on 4 July 2005 at 14.33 hours page 93

Socrates’ Deliberative Authoritarianism

93

mean that I treat the existence of that law as an extra reason, perhaps the most important one, which tips the balance of reasons against taking drugs. It means that my consideration that there is a relevant law pre-empts all other independent reasons for and against taking drugs. Obeying the law requires that I accept its pre-emptive force and thus the consideration of the existence of the relevant law is in this way causally responsible for the formation of an intention to abstain from drugs. If I do not recognize this pre-emptive force of the law, then even though I may do what the law prescribes, I do not obey the law.33 Thus, the fact that Socrates is convinced of the truth of the oracle is not evidence that he does not obey it. In order to decide whether Socrates obeys the oracle or not, we should examine what moves him to act in accordance with it. Is it respect for the god or his independent conviction that what the oracle ordains is true? In the latter case, he would not be obeying the oracle. He would simply be acting in accordance with it but for a reason other than that it is the god’s will. Socrates seems to be aware of the distinction between obeying someone else and acting on the balance of reasons. At 37 e 3–38 a 3 he distinguishes two di·erent responses to the imagined question of why he cannot abandon his elenctic mission in exile (37 e 3– 4). The first is that it is impossible for him to live a quiet life because that would constitute disobedience to the god (37 e 5–38 a 8). The second is that he believes that elenctic examination is of such supreme value that without it a life is not worth living (38 a 1–8). The first response suggests that what would sustain Socrates’ decision to continue his elenctic mission in exile is the consideration that the god ordered him to philosophize. The second suggests that it could be the consideration that an examined life is a better life. What, then, has sustained Socrates’ actual intention to engage in elenctic examination in Athens? Before he received and deciphered the god’s message, Socrates must have already been involved in elenctic examinations which earned him a reputation for wisdom and probably prompted Chaerephon to consult the Pythia about the status of Socrates’ knowledge.34 In the Apology Socrates does 33 For the pre-emptive character of political and legal authority in general see Raz, Freedom, 57–62. 34 I follow Stokes, who argues that Lysimachus’ and Nicias’ exchange at La. 187 d 1–188 a 5 shows that Socrates started his elenctic examinations soon after the age of

Created on 4 July 2005 at 14.33 hours page 94

94

Antony Hatzistavrou

not explain what motivated him to engage in his first elenchi. But it is plausible to assume that intellectual curiosity and possibly a sense that the elenchi were somehow of benefit to himself must have been his primary motivating reasons. In addition, the absence of any protestation from his daimonion must have fostered his confidence that he was not doing something bad. On receipt of the oracle a desire to understand the god’s message moved him to engage in an elenctic examination of its meaning (21 b 2–c 2). After conducting elenchi of some allegedly wise politicians, he gained the impression that he had found the right way to interpret the oracle’s meaning. He could see in what sense he was wiser than they: by not believing as they did that he was wise (21 d 6–8). Once Socrates realized that he was on the right track, a sense of religious duty to do what the god ordered seemed to sustain his intention to continue his elenctic examinations despite his fear that he was making many enemies (21 e 2–4; cf. 22 a 4). And this sense of religious duty seems to have sustained his intention to continue his elenctic mission once the meaning of the oracle became clear to him (23 a 5–7) and he began to regard himself as providing a service to the god (23 c 1). This religious duty is described by Socrates as a duty of obedience (29 b 5–7) when he is called to explain why he is not ashamed to put his life in danger by continuing his elenctic mission (28 b 4–6). And it is by reference to it that he explains his intention to refuse to save his life even if the Athenians decide to spare it on condition that he no longer engages in his elenchi (29 c 7–d 5). After the reception of the oracle, then, Socrates appears to consistently treat the duty to obey the god as his predominant motivating reason for engaging in philosophy. This suggests that he obeys the god and does not simply act in accordance with the god’s will for an independent reason. Of course, through the use of the elenchus he comes to realize the value of his mission for the city (30 a 6–c 2; 30 e 1–31 a 2) and thus acquires some understanding of why the god ordered him to engage in philosophy. And once he can rationalize the value of the elenchus, there can be no doubt that he acquires another strong motivating reason, di·erent from his desire to fulfil his duty to the god, to continue his elenctic mission. But a full understanding of the value of the elenchus must be a later deve18 (see M. C. Stokes, ‘Socrates’ Mission’, in B. S. Gower and M. C. Stokes (eds.), Socratic Questions (London, 1992), 26–81 at 53–4; cf. McPherran’s distinction of three periods in Socrates’ philosophical life (‘Oracle’, 120)).

Created on 4 July 2005 at 14.33 hours page 95

Socrates’ Deliberative Authoritarianism

95

lopment, the result of the systematic application of the elenchus as an act of obedience to the god. It is reasonable to assume that, until he reaches this understanding, what predominantly motivates him to engage in the elenchus is his desire to obey the god, as he repeatedly asserts. Further, even after he has understood the value of the elenchus, he continues to describe his decision to carry on philosophizing as an act of obedience to the god. Failure to distinguish between strong and weak subordination seems also to blur the scholarly debate concerning Socrates’ obedience to his daimonion. The whole debate seems to be based on the premiss that Socrates cannot, strictly speaking, obey his daimonion if he is convinced of the truth of what it prescribes. Thus Vlastos, on the one hand, tries to show that it is never the case that Socrates abstains from an action before he has come up with an interpretation of the daimonion’s message which satisfies his own reasoning.35 On the other hand, McPherran and Brickhouse and Smith try to show that Socrates sometimes abstains from an action even though he has not established the correct interpretation of the daimonion’s message.36 My discussion above shows that there is a third way. Socrates may still abstain from an action because the daimonion so prescribed, thus obeying his daimonion, even though he has rationalized the daimonion’s command and he himself believes that the relevant action is bad. For example, Socrates may abstain from politics because the daimonion so prescribes, while he may also be convinced that engagement in politics is bad for him. In the Apology Socrates thinks that weak subordination to some human authorities is also legitimate. As we have seen, he claims to know that disobedience of a ‘superior’ (βελτονι), whether divine or human, is bad and shameful and an act of injustice (29 b 5–7). The question is who this human superior is. Kraut claims that Socrates refers exclusively to the moral expert.37 Brickhouse and Smith argue that Socrates takes any duly constituted authority, military or legal, to be superior in the required sense.38 It is problematic to accept Kraut’s interpretation for the following reasons. First, Socrates does not make any reference to the moral 35 See G. Vlastos, Socrates: Ironist and Moral Philosopher (Cambridge, 1991), 166–78, 280–7. 36 See, M. L. McPherran, The Religion of Socrates (University Park, Penn., 1996), 184–208; T. C. Brickhouse and N. D. Smith, Plato’s Socrates [Socrates] (Oxford, 1994), 189–95. 37 Kraut, State, 234. 38 T. C. Brickhouse and N. D. Smith, Socrates on Trial (Oxford, 1989), 141.

Created on 4 July 2005 at 14.33 hours page 96

Antony Hatzistavrou

96

expert in the course of his argument at 28 b 4–29 c 1. Secondly, he mentions only military examples in this argument. Thirdly, at 28 d 6–9 he implies that it is shameful (α
σχρο-, 28 d 9) to disobey a non-moral authority, namely, a military commander. But it is equally problematic to accept Brickhouse and Smith’s interpretation. Although one could allow that Socrates acknowledged the existence of a prima facie obligation to obey any duly constituted legal or military authority, it is hard to understand why he would claim that it is prudentially bad for one to disobey any such authority.39 One can understand on the evidence of the Crito (47 a 2–48 a 10) how Socrates could defend the thesis that it is prudentially bad for one to disobey a moral authority. Socrates would claim that by doing so one harms one’s own soul. But it is unclear why Socrates would believe that it is harmful to disobey a duly constituted legal or military authority. This question becomes more pressing if we consider that on Brickhouse and Smith’s interpretation Socrates should not disobey even when he is ordered to commit a gross injustice.40 Would he not harm his soul by committing a gross injustice? We therefore need an interpretation of ‘superior’ which allows that some non-moral practical authorities may be validly considered ‘superior’ and o·ers a plausible prudential reason for obeying them. I suggest that Socrates uses ‘superior’ in reference to the relation between superordinate and subordinate crafts. He accepts that there is a hierarchy of crafts and that some crafts should rule over others and use their work (.ργον). For example, he claims in the Gorgias that gymnastics and medicine should rule over other crafts such as shoemaking, weaving, etc. because they know how to use what the latter crafts produce for the good of the body (517 d 6–518 a 1). He further claims that this hierarchical order according to which gymnastics and medicine are mistresses (δσποινας) and the subordinate crafts slavish (δουλοπρεπε&ς), servile (διακονικς), and unfree (νελευθρους) is an order of justice (κατ τ/ δκαιον) (518 a 1–5). We may understand in a similar manner the relation between a general and a hoplite or the god and Socrates. The general should rule over the hoplite and use his .ργον for the good of the whole army. The hoplite by virtue of possessing a subordinate craft and in accordance with the demands of justice should subordinate himself to the general. The general may be considered a ‘superior’ to the 39 See n. 24.

40 Brickhouse and Smith, Socrates, 151–5.

Created on 4 July 2005 at 14.33 hours page 97

Socrates’ Deliberative Authoritarianism

97

hoplite because he possesses a superordinate craft which can use the .ργον of the hoplite for the overall good of the army. Similarly, the god may be considered a ‘superior’ to Socrates because he possesses a superordinate craft which can use the .ργον of Socrates, his elenchus, for the overall good of the city. I therefore suggest that for Socrates a human ‘superior’ is someone who possesses a superordinate craft. Why is it prudentially bad for one to disobey someone possessing a superordinate craft? One obvious answer is that by disobeying one will fail to enjoy the product of the proper exercise of the superordinate craft. For example, a cook who does not prepare food in the manner the medical doctor prescribes will destroy his own health by consuming his own food. Or a hoplite disobeying his general may miss out on the benefits of victory and carry the blame in a defeat. Of course this holds only if the possessor of a superordinate craft does not misuse his craft. As the Hippias Minor makes clear (373 c 6–375 c 3), Socrates is well aware that craftsmen may misuse their craft. However, in the Apology, in the context of the argument of 28 b 4–29 c 1 Socrates assumes that the tasks of the imaginary general and of Apollo are just. He does not consider, for instance, the case of a general who aspires to become a tyrant and moves his army against his own city. Thus, Socrates must be understood to claim that one should obey not any ‘superior’ but a ‘superior’ who does not misuse his craft and has a just task. Why is it unjust to disobey the superordinate craftsman who does not misuse his craft? I think that Socrates has in mind cases in which one may wrong others by failing to perform one’s task successfully. For example, a doctor may behave unjustly if he misuses his medical knowledge and harms a patient. In order for the possessor of a subordinate craft to perform his task e¶ciently and thus act justly, he needs to rely on the judgement of the superordinate craftsman. If one disobeys, then one is responsible for one’s failure to perform efficiently one’s own subordinate task and thus may validly be accused of acting unjustly. For example, in order for a hoplite to act justly in performing his task as a hoplite, he needs to rely on the general’s judgement about which position to hold, which formation to fight in, or when to attack the enemy. Since, ex hypothesi, the hoplite lacks the superordinate special knowledge of the general, unless he relies on the relevant judgements of the general he will fail properly to perform his task as a hoplite and thus to act justly.

Created on 4 July 2005 at 14.33 hours page 98

98

Antony Hatzistavrou

If by ‘superior’ Socrates means the person who possesses a relevant superordinate craft, then the moral expert is a ‘superior’ in this sense. As we have already seen, he has the superordinate task of making the citizens’ souls good and of ruling over the crafts which provide for the soul (Gorg. 464 b 1–465 d 6 and 517 c 7–518 c 1). And he is the person on whose judgement one should rely if one is to be just and save one’s soul (Crito 47 a 8–d 7). I shall provide a detailed examination of Socrates’ arguments for the legitimacy of subordination to a human moral authority in the following section. Thus far, I have argued that Socrates finds subordination to divine moral authorities and to some human non-moral authorities legitimate. I shall now briefly try to show that he also believes that in some cases one should subordinate oneself to the judgement of people or institutions with which one has special ties. In the Crito Socrates claims that one should obey one’s parents and the laws of one’s country even when one disagrees with what they prescribe (50 d 1–51 c 5). When one disagrees with one’s parents or the laws of one’s country, one has only two options: either to persuade them to change their verdicts or, if one fails in this, to obey them (51 b 2–4). In the Crito Socrates focuses on cases in which the laws or the parents demand from the agent something which is unjust for the agent. For example, the laws demand that the execution of Socrates should take place, which is obviously an unjust punishment for Socrates. Socrates rules out that one who has been wronged by one’s parents or the laws of one’s country has a right to return the injustice. He claims that one’s relation to one’s country or one’s parents is similar to the relation of a slave to a master in the following respect: it is not the case that the same standards of justice apply to both parties (50 e 1–51 a 2). One has specific ties with one’s parents and the laws of one’s country. For example, they are responsible for one’s life, upbringing, and education (50 d 1–e 1). It is because of these specific ties that one should not behave as an equal to one’s parents or to the laws. Socrates clearly advocates subordination to parents and the laws. He thinks that although the agent may have formed her own judgement about the justness of a certain course of action, she should not act on her judgement but act instead on the demands of her parents or the laws of her country. The scope of this subordination, however, is not clear. In the Crito Socrates focuses exclusively on cases in which the parents or the laws demand from the agent a course

Created on 4 July 2005 at 14.33 hours page 99

Socrates’ Deliberative Authoritarianism

99

of action which is unjust for the agent. But given Socrates’ strongly held thesis that the agent should never act unjustly towards someone else (repeated at Crito 49 a 4–e 3), it is doubtful whether he would accept that it is just to obey one’s parents or the laws when they demand from the agent the performance of an action which is unjust for others. It is safer to assume that Socrates believes that subordination to the parents and the laws is legitimate when the actions demanded are just41 or unjust for the agent but not if they are unjust for others.42 Since there is strong evidence that Socrates considers as legitimate subordination to divine authorities, to some human non-moral authorities, and to people or institutions one has special ties with, we may ascribe to him the following Principle of Moral Deliberation: (PMD*) In some deliberations about what it is just for the agent to do the judgements of a third party should pre-empt the judgements of the agent. The set of moral deliberations in which pre-emption is legitimate is to be determined by considerations of (1) the epistemic status of the third party and (2) the justness of the actions prescribed. The higher the level of the epistemic status of the third party, the larger the scope of the relevant pre-emption should be. For example, Socrates seems to have a high level of epistemic trust in the moral authority of the god and his daimonion. He knows that the god does not speak falsely and that his daimonion has never misled him. He obeys them because he trusts them. He similarly trusts the nonmoral authority of a superior craftsman (who does not misuse his craft), i.e. the general, and believes that he should obey his commands, thinking perhaps that by so doing he will behave justly in fulfilling his own subordinate task. But given Socrates’ insistence that one should never harm others and since he does not believe that the possessors of a superordinate craft are moral authorities (on the contrary, the elenchus has shown that they are not), he cannot advo41 As I have already argued, even in the case in which the agent believes that the action is just, she can, strictly speaking, obey her parents or the laws. She may do what is right because her parents demanded it and not because she thinks it is right. 42 There might, of course, be hard cases. What should one do if obedience to one’s parents entails grossly unfair treatment of others while disobedience may substantially harm the parents? It is beyond the scope of the present paper to deal with this issue.

Created on 4 July 2005 at 14.33 hours page 100

100

Antony Hatzistavrou

cate subordination when the superordinate task is misused and he believes that it results in gross injustice. His considerations of the justness of the superordinate task may therefore limit the scope of his subordination to non-moral authorities. Finally, subordination to parents, the city, and its laws is not justified on the basis of their superior epistemic status, whether moral or non-moral.43 Socrates obeys them because of the special moral ties he has with them. But again his insistence that one should never wrong others suggests that he may not subordinate himself to them if he considers their commands to be grossly unjust for others. Thus, considerations about the justice of the commands restrict the scope of one’s subordination to people or institutions with which one has special ties. But Socrates allows that their commands should be obeyed if they are unjust only for the agent. A possible objection to ascribing PMD* to Socrates is that it conflicts with the import of his normative principle that one should always examine, when acting, whether one acts justly or not. On this objection, Socrates’ normative principle should be understood as a¶rming the value of autonomous moral decision-making to the exclusion of subordination to someone else’s judgement on moral issues.44 Socrates’ normative principle entails the following Principle of Practical Deliberation: (PPD) In practical deliberations (that is, in deliberations about what one should do, all things considered),45 considerations 43 Socrates’ views on the moral standing of the laws unquestionably deserve a much fuller treatment than the one I have the space to provide here. But let me briefly state some evidence for my claim that Socrates does not ascribe to the laws a superior epistemic status. First, he explicitly accepts the possibility of bad laws at H.Ma. 284 d 1–e 9. Secondly, the ‘persuade or obey’ doctrine in the Crito implies that there might be moral errors in the laws (cf. Brickhouse and Smith, Socrates, 151). Thirdly, a basic Socratic doctrine, the rejection of retaliation is in direct conflict with a basic ancient law, the lex talionis. As for the moral status of the parents, given Socrates’ pessimistic views about the moral failings of the many (which I shall later expound), it is unlikely that he would consider the parents to be sources of significant moral authority. 44 This is apparently how Weiss, Dissatisfied, 8–12, understands the import of Socrates’ normative principle. 45 In the rest of this paper I shall employ a distinction between practical and moral deliberations. By ‘practical deliberation’ I refer to a deliberation in which the agent has to weigh up di·erent types of consideration (moral, prudential, factual) and decide what she should do, all things considered. By ‘moral deliberation’ I refer to a deliberation in which the agent weighs up primarily moral considerations and

Created on 4 July 2005 at 14.33 hours page 101

Socrates’ Deliberative Authoritarianism

101

about the justness of an action should pre-empt considerations about the survival of the agent. PPD entails that concern for justice should be the primary concern of the agent. But it does not entail a particular principle an individual should observe in deciding what is just. In particular, it does not entail PMD, namely, that in her deliberations about what is just for her to do the agent should rely exclusively on her own judgement. Suppose, for example, that a hoplite decides to follow the judgement of his ‘superior’, his general, because he believes that by so doing he will behave justly in fulfilling his own task and performing in the best way his duty to his country. This hoplite both is governed by considerations of justice in his practical deliberations, as Socrates’ normative principle requires, and subordinates himself to the authority of his ‘superior’. Further, even in cases in which the agent may disagree with the command given, Socrates’ normative principle may be compatible with subordination to someone else’s judgement. Consider, for example, subordination to a divine authority. Socrates may as a result of an initial deliberation have come to the conclusion that he should follow a certain course of action, but his daimonion objects to it. Since Socrates has supreme epistemic trust in the moral authority of his daimonion and believes that it is more likely than he to get things right, he may decide to follow the daimonion’s sign as opposed to his own judgement. In this case, he would have both acted out of a desire to be just and subordinated himself to a divine authority. Or suppose that during a battle Socrates disagrees with the command issued by his general that his division form a column. If he epistemically trusts the general and believes that by following his commands he will fulfil his own task and thus behave justly, he may decide to subordinate himself to his command and not try to persuade his fellow soldiers to remain in a line formation. Again, he will have both acted out of a concern for justice and obeyed his decides what is just, or morally good, for her to do. Thus, one may distinguish practical from moral deliberations by their conclusions and the types of consideration which are primarily involved in them. If one weighs up primarily moral considerations and comes up with a decision about what is just for one to do, one performs a moral deliberation. If one weighs up other types of consideration as well and comes up with a decision about what one should do, all things considered, one performs a practical deliberation. On this distinction, I take Socrates to state at Ap. 28 b 6–c 1 a principle of practical deliberation and at Crito 47 c 8–d 3 a principle of moral deliberation.

Created on 4 July 2005 at 14.33 hours page 102

102

Antony Hatzistavrou

superior.46 A similar attitude is exhibited when Socrates obeys the laws of Athens and refuses to escape. Although he thinks that his execution is unjust (for him), he both obeys them and acts out of a concern for justice, since he also believes that it is unjust to harm the laws. I conclude that PPD is compatible with PMD*. Of course, PPD may be used to justify disobedience to a ‘superior’ craftsman when he misuses his craftsmanship, or disobedience to parents, the laws, and the country when they order actions which are grossly unjust to people other than the agent. But, as I have tried to explain, PMD* does not entail unlimited subordination to human non-moral authorities, or to parents and institutions one has special ties with. On the contrary, PMD* is compatible with extensive disobedience. But the fact that PPD may justify extensive disobedience does not mean that it rules out any instance of subordination to the judgement of a third party, as the objection we considered claimed. Thus far I have argued that Socrates recommends subordination of a limited scope to some human non-moral authorities and to people and institutions one has special ties with. He also recommends subordination of a large scope to divine authorities. In all cases he recommends weak and not strong subordination. In the following section I shall argue that he also recommends weak subordination of a large scope to human moral authorities as the only realistic means for the attainment of a stable state of virtue.

IV One of the arguments Crito puts forward to persuade Socrates to escape from prison is that the many would falsely believe about Crito that he valued money more than his friends and saved it instead of helping Socrates escape (44 b 8–c 5 and 45 d 8–46 a 1). Socrates’ immediate response is that one should not worry about the views of 46 In other words, Socrates’ normative principle does not exclude content-independent reasons (which are a distinguishing mark of authoritative utterances) from the weighing-up the agent should undertake in order to determine which course of action to follow. If content-independent reasons are taken into account and are assigned at least the same value as content-dependent reasons, then nothing in principle precludes that the agent should act on the balance of the contentindependent reasons. For the notion of content-independent reasons, see, H. L. A. Hart, ‘Commands and Authoritative Legal Reasons’, in J. Raz (ed.), Authority (Oxford, 1990), 92–114.

Created on 4 July 2005 at 14.33 hours page 103

Socrates’ Deliberative Authoritarianism

103

the many (44 c 6–9). He later provides a long argument (46 b 1–48 a 10) in support of his thesis. In this argument it becomes clear that he does not believe the main issue to be simply about the ‘public reputation’ of Crito. He reformulates Crito’s thesis as follows: ‘one should value (φροντζειν, 48 a 9; cf. περ0 πολλο- ποιε&σθαι, 46 e 1, and τιµ1ν, 47 a 3) the views of the many about moral issues’ (48 a 8–10). What Socrates means by ‘valuing’ one’s view becomes apparent in his discussion of the examples of the doctor, the gymnast, and the moral expert. To value one’s view is to obey47 (πειθµενοι, 47 d 10), to act on (πρακτον, 47 b 9; 2πεσθαι 47 d 1), to be in awe of (φοβε&σθαι, 47 d 1; cf. 47 b 5), to be humble towards (α
σχ)νεσθαι, 47 d 2) one’s views. So Socrates seems to ascribe to Crito a particular Principle of Moral Deliberation: (PMD*

many

) In deliberations about what is just for the agent to do, the judgements of the many should pre-empt the judgements of the agent.

Socrates wants to refute Crito’s principle of moral deliberation. He does not disagree with the pre-emption thesis per se, as one would expect if the anti-authoritarian reading were right, and claim that the agent’s own judgements should never be pre-empted by the judgements of someone else. He does not reject the pre-emption thesis per se since, as I have argued, he himself espouses PMD*. His disagreement with Crito revolves around the epistemic status of the many on moral issues. He does not epistemically trust the many’s moral competence.48 He wants to replace Crito’s PMD* many with the following principle of moral deliberation: (PMD*

expert

) In deliberations about what is just for the agent to do, the judgements of the moral expert should pre-empt the judgements of the agent.

47 The clearest link between ‘valuing’ (τιµ1ν) and ‘obeying’ (πεθεσθαι) the view of someone is provided at 47 c 1–3: ‘If one disobeys [πειθσας] the one [expert] and does not value [τιµσας] his view and his praises, but values instead the views of the many who have no expertise, will one not su·er something bad?’ The association of πεθεσθαι with φοβε&σθαι (47 b 5 and 47 d 1) and α
σχ)νεσθαι (47 d 2) makes it clear that it means ‘obeying’ and not ‘being persuaded’ in the context of Socrates’ argument. 48 For example, the many will never agree with Socrates that one should never commit an injustice (Crito 49 d 2), they cannot provide appropriate moral education for the young (Ap. 24 e 1–25 c 3), they do not know what is good and lawful (H.Ma. 284 d 1–e 9).

Created on 4 July 2005 at 14.33 hours page 104

104

Antony Hatzistavrou

He supports PMD* with an argument that comes in two expert stages. The first stage has the following structure: (1) One should value only the views of some and not of all individuals (47 a 4). (2) One should value only the beneficial and not the harmful views (47 a 7). (3) The beneficial views are the views of those who have knowledge, while the harmful ones are the views of those who lack knowledge (47 a 9–10). [(C1) Therefore, one should value the views of those who have knowledge.] This short argument does not immediately refute Crito’s view that one should value the views of the many on moral issues. For it is possible that the many may possess moral knowledge and thus that their views may be beneficial. In the second part of the argument Socrates shows that the moral views of the many are not beneficial, since they lack the required knowledge. He also explains why the views of the expert on moral issues are beneficial. The second stage of the argument comes in two parts: [(1) Only the gymnast and the doctor have knowledge about what is good for the body, and not the many.] (2) If one does not obey the views of the gymnast and the doctor and does not follow what they say, one will destroy one’s own body (47 c 1–7). (C2) Therefore, one should value and follow the views of the gymnast and the doctor on issues concerning the good of the body, and not the views of the many (47 a 13–b 12). [(3) Similarly, only the moral expert has knowledge about moral issues (he knows what is just, what is fine, etc.), while the many lack such knowledge.] (4) If one does not obey the moral expert [but follows instead the views of the many], one will destroy one’s best part (i.e. one’s own soul) (47 d 1–5). (5) Life is less worth living if one’s best part [one’s soul] is destroyed than if one’s body is destroyed (47 d 6–48 a 4). (C3) One should obey the moral expert and follow his views as

Created on 4 July 2005 at 14.33 hours page 105

Socrates’ Deliberative Authoritarianism

105

opposed to the views of the many on moral issues (48 a 5–7).49 This argument shows that, contra the anti-authoritarian reading, Socrates does not take moral autonomy to be a necessary condition of one’s moral development. It clearly indicates that for Socrates it is not a precondition of virtue that one act exclusively on one’s elenctically tested judgement. Instead, Socrates seems to believe that there is a di·erent manner in which one can acquire virtue and save one’s soul: by subordinating oneself to the judgement of the moral expert (see premiss (4) of my reconstruction of the argument). Note that Socrates is not referring in his argument to what should guide one’s moral reasoning when one receives some basic moral training. As the analogy with subordination to the doctor and the gymnast on issues concerning bodily health makes clear, the argument is about what should guide one’s moral reasoning throughout one’s life, or at least as long as one lacks moral expertise. Thus, Socrates accepts that it is legitimate for non-experts to subordinate themselves to the moral expert. The analogy with subordination to the doctor and the gymnast has another important implication. It invites us to think that for Socrates, unless a moral expert is found and one subordinates oneself to his moral authority, one is highly unlikely to achieve virtue and save one’s soul. Socrates seems to think that one could not remain healthy without being guided by the gymnast and the doctor (unless, of course, one is extraordinarily fortunate). In the Gorgias he provides a vivid description of the disastrous consequences for one’s health if one follows a diet prescribed not by a doctor and a gymnast but by servants of people’s appetites such as bakers and cooks (518 c 1–e 1). These servile crafts or knacks have no know49 Crito’s claim supported, as we have seen, an argument for the justness of Socrates’ escaping from prison. Socrates’ aim is to undermine Crito’s conclusion that Socrates should escape by showing that it is based on a false premiss. The validity of Socrates’ refutation is not threatened by the fact that Socrates does not identify a moral expert and has argued in the Apology that he has not found one throughout his life. Socrates wants (C2) to be a valid universal principle which indicates what should guide individual moral action. If (C2) is a valid universal principle, then any true moral belief should be consistent with it. If a moral belief p (Socrates should escape) is shown to follow from a principle q (one should act on the beliefs of the many on moral matters) which contradicts (C2), then one should not hold p on the basis of q and the truth of p has not been established. Further arguments are needed to show whether p or ¬ p is true. And Socrates adduces further arguments in support of ¬ p later in the dialogue.

Created on 4 July 2005 at 14.33 hours page 106

106

Antony Hatzistavrou

ledge of what is good for the body and cannot distinguish between beneficial and harmful pleasures (500 a 8–b 5). This is why they harm the body unless they are supervised by the crafts of gymnastics and medicine. Only the gymnasts and the doctors may benefit the body by preserving its good order and structure, which is health (504 a 2–c 10). In a similar manner, only the possessor of the true political art may benefit the soul and bring order and structure to it (504 d 1–e 5). This requires the ability to distinguish the good from the bad pleasures of the soul, and only the possessor of the true political art has the required knowledge (500 a 2–6). In a similar manner, Socrates warns the young Hippocrates in the Protagoras about the dangers of receiving lessons from the sophists, unless one is an expert regarding the soul and can distinguish the good lessons from the harmful ones (313 e 5–d 5). If, however, without a moral expert people are unable to acquire virtue, it follows that neither any of Socrates’ contemporaries nor probably Socrates himself has virtue, since he explicitly denies that he is a moral expert or that he has found anyone who is.50 To avoid this undesired conclusion, one may suggest that for Socrates subordination to the moral expert is not the only path to virtue. One may attain virtue through the proper exercise of the elenchus. This is Socrates’ own way to achieve virtue and it is the route he recommends to his fellow citizens. If there were a moral expert, one could as well attain virtue by subordinating oneself to his judgement. But this is not necessary. The elenchus provides another path to virtue. This interpretation assumes that (1) subordination to the moral expert is not necessary for virtue since (2) proper exercise of the elenchus is su¶cient for virtue. But there is uneasiness with both (1) and (2). If (1) is the case, then the existence of a moral expert is nothing more than a bonus, an extra way of attaining virtue. But to treat the moral expert simply as a bonus does not do justice to the importance Socrates seems to ascribe to the moral expert in the Crito, the Protagoras, and the Gorgias. The tone of Socrates is pretty dramatic. He speaks of the danger of the destruction of one’s soul and of living a meaningless life. He does not appear to recognize a third way between subordination to the many or to the sophists and subordination to the moral expert. The possibility of the elenctic life does not enter into the picture. And it does not seem 50 This objection and the possibility of treating the moral expert as a bonus were suggested to me by David Sedley.

Created on 4 July 2005 at 14.33 hours page 107

Socrates’ Deliberative Authoritarianism

107

to fit with some features of the dramatic story Socrates is telling. For example, if the elenchus is a way to achieve virtue, then it is simply not true that only a (moral) craftsman (τεχνικς) may distinguish between good and harmful pleasures of the soul (Gorg. 500 a 4–6). The possessor of the elenchus can do this as well. And for the same reason it is not true that one cannot distinguish which lessons are good for the soul unless one is a doctor (
ατρικς) of the soul (Prot. 313 d 5–e 1). Further, Socrates makes similar dramatic claims when he describes his own predicament of lacking moral expertise. For example, he claims in the Hippias Major that if he does not know what the fine is, he is no better alive than dead (304 d 5–e 3).51 On the other hand, if (2) is the case, then it is di¶cult to understand why Socrates claims to know nothing great or small and why he advises that the search for the moral expert should not end. If Socrates can through the elenchus lead people to virtue, then he knows how to save their souls, the most important thing of all. And a moral expert, if found, would not have made a great di·erence to one’s moral development. Further, if the elenchus can lead people to virtue, then Socrates has to explain why it has actually failed to do so. For example, it has failed to convince the majority of his fellow citizens that committing injustice is bad, and Socrates himself doubts whether it ever will (Crito 49 c 10–d 5). Our puzzlement is due to two basic elements in Socratic ethics which are di¶cult to square. On the one hand, Socrates tends to treat virtue as analogous to a craft. As a result we get an account of the real ‘craftsman’ of virtue, the moral expert, and the impression that moral knowledge may be imparted like any other craft knowledge, through authoritative guidance. On the other, Socrates stresses the importance of the elenchus, which he presents nevertheless as something di·erent from the craft of virtue. He distinguishes his elenctic skill from moral expertise and makes it clear that the elenchus is not a traditional method of moral indoctrination, although it contributes to making the people better. The supporters of the anti-authoritarian reading focus on the importance of the elenchus and downplay the importance of moral expertise and subordination to the authority of the moral expert. I have already tried 51 As I shall later suggest, Socrates can claim that despite his lack of knowledge of what virtue is, his soul may not be destroyed and his life is still worth living. Socrates obtains a more stable state of virtue than his fellow citizens through his subordination to divine authorities, such as his daimonion.

Created on 4 July 2005 at 14.33 hours page 108

108

Antony Hatzistavrou

to show that this is the wrong interpretative path to take and have argued for the central role of the latter notions in Socratic ethics. The challenge for my interpretation is to explain (or explain away) the unappealing consequence that no one may attain virtue and save her soul unless she subordinates herself to the moral expert. My suggestion is in a nutshell the following. Socrates takes subordination to divine or human moral authorities to be necessary for the acquisition and preservation of a stable state of virtue. When he speaks of saving the soul, he means ‘acquiring a stable state of virtue’. The elenchus may lead to a more fragile state of virtue which may be easily lost. But Socrates is quite pessimistic about whether the many may ever attain even this fragile state. I believe that the dramatic tone of the Gorgias, the Protagoras, and the Crito is not hyperbolic. Socrates wants to convey the idea that, unless the many subordinate themselves to the moral expert, it is highly unlikely that they will save their souls. I believe that he is drawn to this dramatic conclusion by two main considerations. The first concerns his conception of moral expertise as a superordinate craft which can successfully provide authoritative guidance. The second concerns his pessimism about the possibility of the elenchus’ changing the moral outlook of the many. In the Crito he admits that one of the most important moral theses he has repeatedly tried to establish, namely, that one should never retaliate for an injustice done to oneself, will never be accepted by the many (49 c 10–d 2). And he adds that there is no common ground for deliberation between those who deny and those who accept this basic moral thesis (49 d 3–5). In the Apology he refuses to take any responsibility for the moral development of his companions on the grounds that he never claimed to provide moral training to anyone (33 b 3–5). In the same dialogue he suggests an explanation of why the elenchus has failed to convince his fellow citizens: they find it annoying (βαρ)τεραι) and grudging ( πιφθοντεραι) (37 c 5–d 2). This despising (πχθεια) of the elenchus has led to widespread slander against Socrates, which in turn has caused his indictment (28 a 5–b 2). But even if there were a way to overcome the annoyance and distrust generated by the elenchus, would it su¶ce to lead people to virtue? The elenchus certainly helps one’s moral development. It makes people focus on what really matters, their soul, and not on material goods (30 a 7–b 4). It also helps people advance their self-

Created on 4 July 2005 at 14.33 hours page 109

Socrates’ Deliberative Authoritarianism

109

understanding by revealing to them the depth of their ignorance concerning moral issues.52 By so doing it also prepares people to give their consent to the moral expert if they find one and entrust their moral education to him (La. 200 e 1–201 b 5). But it appears that the elenchus fails to provide the type of knowledge which is, according to Socrates, a defining feature of moral expertise. This is knowledge of what virtue is. Can one be virtuous without having this type of knowledge? The dramatic tone of the Hippias Major passage I have mentioned suggests that it is probably not possible. The Meno might help us acquire a more balanced understanding of Socrates’ stance on this issue. At 97 e 2–98 a 8 Socrates argues that what distinguishes truly believing that p from knowing that p is the stability of the relevant belief. In knowledge the relevant beliefs are tied by an explanatory account (α
τας λογισµ!, 98 a 4–5), while in the state of true belief the beliefs are unstable and likely to be lost. Although in this passage of the Meno the explanatory account is to be understood in the context of the theory of recollection which is evidently Platonic and not Socratic, the distinction between stable and unstable beliefs may help us understand the inadequacies of the elenchus. The elenchus may not provide the stability of one’s moral beliefs which the explanatory account of what virtue is can provide.53 The elenchus could therefore lead to a very fragile state of virtue, making loss of virtue very likely. This leaves us with a question about Socrates’ state of elenctic virtue. On the basis of the above it would seem that his virtue is not stable but very likely to be lost. But what is special about Socrates is the fact that he is aided by divine signs and especially his daimonion, which prevents him from doing bad things. His subordination to divine authority counterbalances the fragility of his elenctically acquired virtue. Thus, we can now understand Socrates’ dramatic tone when he urges subordination to the moral expert. He is pessimistic about whether the many will ever be able to overcome their pusillanimity and be convinced by the elenchus. And even if they did, they would 52 Thus, the elenchus can help people attain ‘human wisdom’, which as I explain in n. 20, is a type of knowledge which Socrates believes that the average person can achieve. Note that this Socratic belief is consistent with his pessimism that the many may not after all overcome their state of self-deception and achieve human wisdom. 53 See Gorg. 501 a 1–b 1, where Socrates suggests that the notion of a superordinate craft is able to provide an explanatory account of its subject matter.

Created on 4 July 2005 at 14.33 hours page 110

110

Antony Hatzistavrou

attain a very fragile state of virtue and require constant guidance by a human or divine moral authority. Of course, in the absence of a moral expert the elenchus is all they have and, unless they receive divine aid, as Socrates does, the fragile state of elenctic virtue is the most they can hope to attain. This is one feature of the elenchus which makes it so important and explains the value Socrates attaches to it when he declares that the unexamined life is not worth living. Unless one engages in the elenchus, one will fail to achieve even this fragile state of elenctic virtue. But, as I have explained, this Socratic belief in the value of the elenchus goes hand in hand with his pessimism about the possibility of the many’s being convinced by the elenchus. The other important feature of the elenchus is the fact that it helps the agent enhance her understanding of moral issues. As I have tried to make clear, this does not conflict with the demand that the many subordinate themselves to the moral expert. Socrates supports weak and not strong subordination. The many can elenctically examine their moral beliefs freely, provided that they obey the moral expert. Elenctic examination will help them to appreciate what the moral expert tells them to do and eventually to become moral experts themselves. Thus, the elenchus is valuable not only as the second-best path to virtue in a society in which there are no moral experts: it is also valuable even when one subordinates oneself to the moral expert as a vehicle to moral expertise and a stable state of virtue. Thus, since Socrates believes that non-experts should subordinate themselves to the moral expert as long as they remain nonexperts, he should not be credited with MAT. Instead, we may ascribe to him the Subordination to Authority Tenet: (SAT) There are conclusive normative reasons, both prudential and moral, for all individuals to subordinate themselves to the authority of the moral expert throughout their lives (or at least as long as they remain non-experts). I take SAT and PMD* to be the kernel of Socrates’ authoritarianism in the Apology and the Crito. I call this type of authoritarianism ‘deliberative authoritarianism’, since it prescribes that in deliberating about what is just for one to do, one’s judgements may be pre-empted by the judgements of someone else (a divine authority, a moral expert, a superior craftsman, or a person or institution one

Created on 4 July 2005 at 14.33 hours page 111

Socrates’ Deliberative Authoritarianism

111

has special ties with). It thus contradicts the principle of moral deliberation that the anti-authoritarian reading ascribes to Socrates. Deliberative authoritarianism does not conflict with Socrates’ claim in the Crito that ‘not only now, for the first time, but continuously, I have been that sort of person who obeys nothing of his own [τν µν] other than the argument [λγ!ω] which appears best to my reasoning [λογιζοµν!ω]’ (46 b 4–6). In this passage Socrates states two principles which guide his deliberations. The first is a second-order principle about how one should select and evaluate relevant reasons for action. It is not, like PPD and PMD*, a firstorder principle about which reasons for action the agent should act upon. It states that Socrates and indeed every agent should rely only on rational arguments. The second is a principle of epistemic trust indicating that Socrates’ reasoning capacities are epistemically reliable. We may formulate them as follows: Second-order principle: In one’s practical deliberations, one should consider only rational arguments. Principle of epistemic trust: Socrates is epistemically justified in accepting the argument which appears best to his reasoning. Both principles are compatible with Socrates’ first-order PMD*. The argument that in deliberations about what is just the judgements of human moral experts should pre-empt all other judgements appears the best (rational) argument to Socrates in the Crito. Further, he has epistemic trust in his reasoning. He possesses the elenctic craft and is aided by divine authorities. He is therefore epistemically justified in considering this argument to be true. Weiss takes Crito 46 b 4–6 to show that Socrates believes that, in the absence of a moral expert, one should follow one’s own judgement about moral issues.54 This means that, on Weiss’s interpretation, Socrates considers the principle of epistemic trust to be valid not only for himself, but for his fellow citizens as well. As we have seen, he does not trust the many’s moral reasoning. To give an example, in the Crito he argues that one should not value their opinions (δξαι) about moral issues (presumably their δξαι are what appears best to their reasoning). He further claims to know (ο3δα) that very few will ever agree with him that one should not return injustice or mistreat anyone (49 c 10–d 2). On Socrates’ view there is no common ground of arguments (οκ .στι κοιν4 βουλ) to 54 Weiss, Dissatisfied, 58–61.

Created on 4 July 2005 at 14.33 hours page 112

112

Antony Hatzistavrou

which both the few who agree with him and the many can refer so that they may rationally settle their dispute; instead, they must treat each other’s arguments with contempt (49 d 3–5). It therefore seems plausible to assume that in the Crito Socrates is quite sceptical about the reliability of the moral reasoning of his fellow citizens while being confident about the reliability of his own moral reasoning. There remains a final question about whether Socrates believes that he should subordinate himself to the authority of the moral expert as PMD* requires. I believe that, in so far as Socrates expert considers himself to lack moral expertise, he should take PMD* expert to be valid for himself as well. And it is also possible that he believes that, given the fact that he has reached a high level of moral understanding by virtue of the elenchus, he could quickly reach the level of moral expertise with the aid of the moral expert and thus need not rely on someone else’s judgement on moral issues throughout his life. To conclude, I have tried to show that, contra the anti-authoritarian reading, Socrates finds subordination to the judgement of someone else on moral issues legitimate. And he further holds that subordination to the judgement of a moral expert is necessary for the many’s moral development. He does not have absolute confidence in the elenchus as a vehicle for virtue, nor is he a champion of moral autonomy. His deliberative authoritarianism brings his ethical theory closer to the moral theory of the Republic than is normally supposed. University of Cyprus

B I B L I O GR A P HY Brickhouse, T. C., and Smith, N. D., Socrates on Trial (Oxford, 1989). Plato’s Socrates [Socrates] (Oxford, 1994) Devereux, D., ‘Nature and Teaching in Plato’s Meno’, Phronesis, 23 (1978), 118–26. Friedman, R. B., ‘On the Concept of Authority in Political Philosophy’, in J. Raz (ed.), Authority (Oxford 1990), 56–91. Gulley, N., The Philosophy of Socrates (London, 1968).

Created on 4 July 2005 at 14.33 hours page 113

Socrates’ Deliberative Authoritarianism

113

Hart, H. L. A., ‘Commands and Authoritative Legal Reasons’, in J. Raz (ed.), Authority (Oxford, 1990), 92–114. Irwin, T., Plato’s Ethics (Oxford, 1995). Klosko, G., The Development of Plato’s Political Theory [Development] (New York, 1986). Kraut, R., Socrates and the State [State] (Princeton, 1984). McPherran, M. L., The Religion of Socrates (University Park, Penn., 1996). ‘Elenctic Interpretation and the Delphic Oracle’ [‘Oracle’], in G. A. Scott (ed.), Does Socrates Have a Method? (University Park, Penn., 2002), 114–44. Nehamas, A., The Virtues of Authenticity [Authenticity] (Princeton, 1999). ‘Socratic Intellectualism’ [‘Intellectualism’], in Authenticity, 27–58. ‘What Did Socrates Teach and to Whom Did he Teach it?’, in Authenticity, 59–82. Penner, T., ‘Socrates’, in C. Rowe and M. Schofield (eds.), The Cambridge History of Greek and Roman Political Thought (Cambridge, 2000), 164–89. Popper, K., The Open Society and its Enemies: The Spell of Plato [Society] vol. i (London, 1962). Raz, J., The Morality of Freedom [Freedom] (Oxford, 1986). Reeve, C. D. C., Socrates in the Apology [Apology] (Indianapolis, 1989). Schofield, M., ‘Socrates on Trial in the USA’, in T. P. Wiseman (ed.), Classics in Progress (Oxford, 2002), 263–83. Stokes, M. C., Plato: Apology (Warminster, 1997). ‘Socrates’ Mission’, in B. S. Gower and M. C. Stokes (eds.), Socratic Questions (London, 1992), 26–81. Stone, I. F., The Trial of Socrates (London, 1988). Taylor, A. E., Socrates (Edinburgh, 1932). Vlastos, G., Socrates: Ironist and Moral Philosopher (Cambridge, 1991). Socratic Studies [Studies], ed. M. Burnyeat (Cambridge, 1993). ‘The Historical Socrates and the Athenian Democracy’ [‘Democracy’], in Studies, 87–108. ‘The Socratic Elenchus: Method is All’, in Studies, 1–37. ‘Socrates’ Disavowal of Knowledge’ [‘Disavowal’], in Studies, 38–66. Weiss, R., Socrates Dissatisfied: An Analysis of Plato’s Crito [Dissatisfied] (Oxford, 1998). Wilkes, K., ‘Conclusions in the Meno’, Archiv f•ur Geschichte der Philosophie, 61 (1979), 143–53. Wood, E. M., and Wood, N., ‘Socrates and Democracy: A Reply to Gregory Vlastos’, Political Theory, 14 (1986), 55–83. Woozley, A. D., Law and Obedience (London, 1979).

Created on 4 July 2005 at 14.33 hours page 114

page 114 is blank

Created on 4 July 2005 at 14.33 hours page 115

THE UNITY OF VI RTUE IN P LATO’ S PROTAGORAS BERND MANUWALD

  Protagoras’ great speech in the Platonic dialogue named after him, Socrates has just one ‘small point’ he would like to have explained: since Protagoras has talked about justice, temperance, and piety and called them all ‘virtue’, Socrates now enquires whether virtue is one thing and justice, temperance, and piety (etc.) are parts of it or whether all these qualities are (merely di·erent) names of one and the same thing (329 b 5–d 1).1 This question (together with two further ones developed from it) forms the starting-point for a long discussion by the two interlocutors of the unity of virtue; in this part of the dialogue the disjunctive pair just outlined, as well as some others, play a significant role, but no explicitly stated result is achieved. Interpreters of the Protagoras, however, have long been trying to attribute one of the positions sketched in the dialogue to the Platonic Socrates. The reason why this much-discussed problem is taken up once again here is the detailed and perceptive analysis of the problem of the unity of virtue in the Protagoras recently presented by Denis O’Brien.2 O’Brien comes up with the following conclusion: In the Protagoras, we are clearly meant to understand that Protagoras is ã Bernd Manuwald 2005 I would like to thank David Sedley for perceptive comments on an earlier version and my daughter Gesine Manuwald for the translation of this paper into English. 1 I adopt J. Burnet’s text and lineation (Platonis Opera, iii, 2nd edn. (Oxford, 1909)). All translations are mine unless otherwise indicated; some translations of passages from the Protagoras are inspired by the version of C. C. W. Taylor, Plato: Protagoras, trans. and comm., rev. edn. (Oxford, 1991). 2 D. O’Brien, ‘Socrates and Protagoras on Virtue’ [‘Virtue’], OSAP 24 (2003), 59–131. O’Brien’s article goes back to a paper read at a conference on the Protagoras in Naples (September 2002); it has now been published in Italian as well: ‘Socrate e Protagora sulla virtu’, › in G. Casertano (ed.), Il Protagora di Platone: struttura e problematiche [Protagora] (2 vols.; Naples, 2004), i. 173–250. All references are to the English version.

Created on 4 July 2005 at 14.33 hours page 116

116

Bernd Manuwald

wrong to suppose that a man can be brave and not be wise. But precisely how are we meant to understand the relation of bravery and wisdom? Are those two names both ‘names of one and the same thing’ (cf. 1b)? Or is wisdom the gold of which courage is a part (cf. 2b)?3 That question has to remain a question . . .4

Although O’Brien has highlighted a number of important points, his overall conclusion is in my view di¶cult to accept. I would therefore like to argue for a di·erent reading. It seems to me that the discussion in the Protagoras cannot be reduced to the possibilities set up by O’Brien; both positions are not meant to be regarded by the reader as potential solutions. Rather, the dialogue with its aporetic ending does not only consist of a refutation of Protagoras’ view and other theoretically possible ones, but also presents elements of Platonic doctrine whose interrelation cannot be sufficiently explained on the basis of the Protagoras alone. They may, however, be combined to form a plausible whole from the perspective of later dialogues (particularly the Republic) and thus o·er a potential Platonic solution to the problem of the unity and diversity of the virtues. I shall therefore contribute to the discussion of the unity of virtue by following a di·erent approach: I shall try to arrive at a plausible interpretation by looking at the Protagoras within the framework of Plato’s thinking. This paper will focus on a few essential points and some particular passages where the Greek text seems to suggest an interpretation other than that o·ered by O’Brien.

3 For the sake of clarity and simplicity the numbering employed by O’Brien is kept here. O’Brien (‘Virtue’, 60) included the numbers in his free and interpretative paraphrase of Socrates’ fundamental questions. For the convenience of the reader it is reproduced here (somewhat abrigded): Are justice, temperance, piety, and the like parts of one thing, which is virtue [1a], or are their names ‘all names of one and the same thing’ . . . [1b]? (329 c 6–d 1) Are the individual virtues like the parts of a face (mouth, nose, eyes, and ears) . . . [2a], or are they like ‘parts’ or pieces of gold, which di·er from one another and from the whole only ‘in largeness and smallness’ . . . [2b]? (329 d 4–8; cf. 349 c 4–5) Can people have only one part of virtue and not have another part [3a], or, if they have one part of virtue, do they necessarily have all the others as well [3b]? (329 e 2–4) 4 O’Brien, ‘Virtue’, 129.

Created on 4 July 2005 at 14.33 hours page 117

The Unity of Virtue in Plato’s Protagoras

117

1. Socrates’ questions (329 c–330 b) O’Brien analyses this paragraph (and the ‘recapitulation’, 349 a–c) with a detailed discussion of the positions of Terry Penner5 and Gregory Vlastos;6 these two scholars tried to extrapolate from the Protagoras a view of the unity of virtue to which Socrates subscribes in the dialogue (with divergent results). O’Brien correctly comments: ‘the error [sc. of Penner and Vlastos] lies . . . in the assumption that Socrates must be supposed to endorse one or other of the two theses that Protagoras has discarded.’7 Indeed, Socrates’ remarks are questions,8 addressed to Protagoras, in order to find out how his interlocutor understands ‘unity in diversity’ (329 c 5–d 1). O’Brien could simply have referred to an important methodological remark made by Socrates in this context: Socrates explicitly warns against mixing up his questions with opinions held by him (330 e 7–331 a 1). In so far as O’Brien states that one does not have to assume that Socrates endorses one of the hypotheses introduced into the discussion, his analysis of Socrates’ questions need not be taken up again; in the course of his argument, however, he nevertheless considers the questions as possible ‘theses’ of Socrates. For O’Brien believes that the dialogue formulates two rival theses Protagoras is confronted with: ‘a unity that has no parts (1b), and a unity that has parts (2b), but “parts” which, as virtues, cannot be had separately one from the other (3b)’.9 Such a disjunctive pair cannot be inferred from the text; the discussion of the question of unity takes a slightly di·erent startingpoint: Socrates wants to know what—according to Protagoras—the relationship between the individual qualities mentioned and virtue (uniting them) is like (cf. 325 a 2). For this purpose he asks Protagoras three questions (developed in the course of the dialogue), each 5 T. Penner, ‘The Unity of Virtue’, Philosophical Review, 82 (1973), 35–68 (several reprints). 6 G. Vlastos, ‘The Unity of the Virtues in the Protagoras’ [‘Unity’], Review of Metaphysics, 25 (1972), 415–58; reprinted ‘with substantial additions and corrections in both text and notes’ in id., Platonic Studies (Princeton, 1973), 221–69, and with further additions in the 2nd edn. (Princeton, 1981), 427–45. 7 O’Brien, ‘Virtue’, 82. 8 Ibid. 82, 85. In an earlier part of his article (p. 74, subheading), however, O’Brien himself speaks of ‘Socrates’ twin theses’. The recapitulation in the dialogue (349 b 1) explicitly uses the word
ρτηµα. 9 O’Brien, ‘Virtue’, 128.

Created on 4 July 2005 at 14.33 hours page 118

118

Bernd Manuwald

of them cast in the form of a choice between two distinct concepts: are justice, temperance, and other qualities parts of (one) virtue (1a) or are these all names of one and the same thing (1b) (329 c 2–d 2)? When Protagoras has chosen the position that they are parts of virtue, itself viewed as a unity (1a) (329 d 3–4),10 Socrates wants to define the character of these parts more clearly and asks: are they parts like those of a face, such as mouth, nose, etc. (2a), or like those of gold, which di·er from the whole in largeness and smallness only (2b) (329 d 4–8)? Protagoras again chooses the first alternative (2a): they are related to each other as parts of the face are related to the face as a whole (329 d 8–e 2), by which he probably means the whole face as a unifying factor in contrast to its parts. Socrates is somewhat astonished at this answer and asks: ‘Do some human beings then possess some of these parts of virtue and some another (3a), or does someone necessarily have them all if he has got just one (3b)?’ (329 e 2–4). The content of the sentence might be paraphrased as follows: ‘Do you then [ον11] really [κα, 329 e 2] allow, Protagoras, for the possibility of having only one virtue, or is it not rather necessary to have them all if one has just one?’ Socrates’ questions presuppose the unity of virtue ( ν µ ν τ
στιν  ρετ, 329 c 7); thus the only thing at issue is the character of this unity. Hence it is somewhat surprising that Socrates comes up with a question that considers a possible separation of the virtues (in Protagoras’ view). But the comparison with the face Protagoras has chosen exhibits a certain ambivalence: on the one hand, the face represents an organic whole and thus does not suggest the possibility expressed in Socrates’ question. For—in the language of these similes—one may well have a part of gold without having the whole lot;12 the idea, however, of having, for instance, a nose and no mouth sounds somewhat absurd.13 On the other hand, the parts 10 νς ντος τς ρετς µρια. 11 In line with his interpretation, O’Brien (‘Virtue’, 94 n. 51) regards ον as pointing to ‘a consequence’; that is, he assumes the common inferential use of ον (cf. J. D. Denniston, The Greek Particles, 2nd edn. (Oxford, 1954), 426). But ον can as well mean ‘Proceeding to a new point, or a new stage in the march of thought’ (Denniston, loc. cit.). 12 Cf. also O’Brien, ‘Virtue’, 94. 13 One can certainly assume that Plato’s use of the image is based on the normal situation and not on the idea of an impaired face. The possibility that, for instance, one can be blind without being deaf (cf. O’Brien, ‘Virtue’, 95) is of no relevance in this context, since what is at issue is having or not having the parts (which are assumed to be functioning) and not the question whether they function at all. The

Created on 4 July 2005 at 14.33 hours page 119

The Unity of Virtue in Plato’s Protagoras

119

of the face represent fairly autonomous and self-reliant unities, and this is what Socrates takes into account in his question (by straining the simile). Since the author wants Protagoras to voice specific views in the dialogue, the possibility of virtues existing separately and in particular being di·erent in quality has to be presented. That this point is important is demonstrated by the course of the argument: the di·erence in quality turns into the sole topic of the discussion after Protagoras has eagerly jumped at the idea that one could have individual virtues separately, and after wisdom (σοφα) and bravery (νδρεα) have been introduced in this context (329 e 5–330 a 2): Protagoras agrees that each part is di·erent from the others and has its own characteristic power (δναµις) (330 a 3–4). Socrates is not satisfied with having made Protagoras agree to this statement in its positive form, he also asks for his agreement to the negative version (which suggests a possible contrast between the parts), namely that one part of virtue is not like any other one, neither in itself nor as regards its power (330 a 4–b 1).14 It is these di·erences in quality that Socrates’ ensuing rhetorical question refers to:  δλα δ τι ο!τως #χει, ε%περ τ'( παραδεγµατ γε #οικε; ‘It clearly must be like this if it corresponds to the example?’ (330 b 1–2). Socrates then draws the conclusion, in agreement with Protagoras: ‘None of the other parts (di·erent from each other) of virtue is like knowledge, none like justice, none like bravery, none like temperance, and none like piety’ (330 b 3–6).15 O’Brien, however, does not refer  δλα δ . . . (330 b 1–2) to the remark about the di·erences in quality, but interprets as follows: fact that according to the image of the face ‘each “part” possesses a function, or dynamis, that is specifically its own’ is not a su¶cient basis for the conclusion that ‘the ability to exercise any one virtue does not presuppose the ability to exercise any other virtue’ (thus O’Brien, ‘Virtue’, 103). For normally one either has a complete face with all its concomitant functions or none at all; consequently the ability to see normally implies the ability to hear (in the case of human beings). 14 Cf. Protagoras’ answers at 330 a 3 and b 2–3. Typically, O’Brien (‘Virtue’, 60) focuses on the negative side right from the beginning of his argument: when Socrates introduces the image of the face (329 d 5–6), O’Brien already adds ‘where each part is di·erent from every other part and di·erent from the whole’ (although the text does not say so). At any rate, that makes understanding the image of the face as an organic unity more di¶cult. 15 The main di·erence between the two models (gold and face) is the fact that in one case (gold) there are no di·erences in quality between the individual parts, but in the other case (face) such di·erences are present. This interpretation is corroborated by the recapitulation of the argument (349 c 2–5).

Created on 4 July 2005 at 14.33 hours page 120

120

Bernd Manuwald

The thesis that we can have one part of virtue and not have another part (3a) is explicitly said to follow ‘clearly’ (330 b 1) from the image of parts of a face (2a). The implication is plainly that the image of pieces of gold (2b) will lead to the opposed thesis, that we cannot have one virtue without having all the others (3b).16

But if the concluding sentence just quoted is interpreted differently, there is no evidence for the implication assumed by O’Brien (which is not required even in his interpretation of the text). A concept which necessarily connects the ideas ‘parts like those of gold’ (2b) and ‘having one virtue implies having all the others’ (3b) cannot be inferred from the passage discussed. The character of the similes rather points to a thesis (based on the positive statements and the organic unity suggested by the image of the face) according to which the virtues are like parts of a face (2a) and must be had in their entirety (3b). This combination implied in the text (sc. 2a+ 3b), and not only the alternatives (1b) and (2b+ 3b) discussed by O’Brien, may be regarded as another possible position that contrasts with the opinion attributed to Protagoras that one could have one virtue without the others (3a).

2. Answers in the dialogue? Even if Socrates’ questions point to a preference neither for the hypothesis that the names of individual virtues are names of one and the same thing (1b) nor for the combination (2b+ 3b), it might appear from the continuation of the dialogue that both hypotheses play a prominent role or at least are more important than the combination (2a+ 3b) just inferred as a further possibility. While discussing the relationship between justice and piety Socrates says: ‘For my own part I should say both that justice is holy and that holiness is just’ (331 b 1–3).17 And on behalf of Protagoras he would give ‘the same answer, that justice is either exactly the same thing 16 O’Brien, ‘Virtue’, 94. The concept that the individual virtues are di·erent from each other does not necessarily imply the view that one may have one virtue without the others. Diversity and interconnection of the virtues are reconcilable concepts: the completely good republic is necessarily endowed with all cardinal virtues (Rep. 427 e), clearly di·erent from each other. It is also obvious that the virtues of the other parts of the soul do not exist without the virtue of the logistikon (Rep. 442 b 11–c 8). 17 Piety may be just if it is understood as correct, just behaviour particularly towards the gods (cf. Euthph. 11 e 7 ·.). But not everything just is pious in that sense

Created on 4 July 2005 at 14.33 hours page 121

The Unity of Virtue in Plato’s Protagoras

121

as piety (1b) or as like as possible (2b),18 and that, most definitely, justice is like piety and piety like justice’ (331 b 4–6). Socrates’ remark, however, is not meant to set up a precise thesis of his own, but is obviously intended to refute Protagoras’ position by all possible means. That can be concluded from the structure of Socrates’ argument: the sentence presenting the two alternatives is followed by the exact reversal of Protagoras’ view, and that is called a most definitely correct statement (µ*λιστα π*ντων).19 Protagoras committed himself to the position that, for instance, justice is ‘not like’ piety (330 b 3–6; cf. 330 e 7–331 a 5). For the purposes of this argument, it is of no relevance whether justice is called the same as piety or as like as can be; Protagoras’ position is refuted in either case.20 Similarly designed to deconstruct Protagoras’ position is Socrates’ attempt to identify temperance and wisdom (332 a 2 ·.). Even more clearly than the preceding argument, this one is based on exploiting di·erent meanings of the same word, an eristic method familiar to Plato (cf. Euthd. 277 e–278 b). While Protagoras had used s»ophrosun»e as an equivalent of aid»os (cf. Prot. 322 c 2 and 323 a 1–2), i.e. roughly in the sense of temperance, Socrates highlights the possible intellectual component of the term so that s»ophrosun»e in this interpretation is identical with sophia, understood as knowledge concerning actions. The ‘result’—‘Thus, s»ophrosun»e and sophia are (Euthph. 12 a 1), although it may be called thus in the sense of ‘holy’, ‘determined by divine law’ (cf. Theogn. 132 +ση . . . δκη; Euthph. 6 e 10, where Euthyphro defines σιον as τ . . . το-ς θεο-ς προσφιλ ς). The predication is due to the exploitation of di·erent meanings of σιος and thus terminologically not exact. Cf. B. Manuwald, • Platon: Protagoras. Ubersetzung und Kommentar [Kommentar] (G•ottingen, 1999), 258–9. I sometimes refer to my own previous research, especially to my commentary on the Protagoras, and repeat the results in part. Apart from the question at issue, that may be necessary for another reason as well, since in the words of an Italian colleague, ‘germanica non leguntur!’ (see W. Leszl, ‘Le funzioni della tesi edonistica nel Protagora e oltre’ [‘Le funzioni’], in Casertano, Protagora, ii. 574–638 at 580 n. 4); cf. also F. Trabattoni, ‘Unit›a della virtu› e autopredicazione in Protagora 329e–332a’ [‘Unit›a’], ibid. i. 267–91 at 279 n. 14. 18 O’Brien’s attributions are reproduced here (cf. ‘Virtue’, 63–4). Whether τι +µοιτατον really refers to the image of gold can hardly be proved or refuted. 19 The continuation of the presentation of the two alternatives is ignored by O’Brien. 20 Later Socrates summarizes the argument of this passage by saying that it showed justice and holiness to be ‘virtually the same’ (σχεδν τι τα0τν, 333 b 6); he thereby points out the purpose of the earlier argument.

Created on 4 July 2005 at 14.33 hours page 122

122

Bernd Manuwald

one and the same thing, aren’t they?’ (333 b 4–5)—is already used as a premiss and basically yields the ‘identity thesis’ (1b). But the way in which this result is achieved does not suggest that Plato wanted to present this hypothesis as a serious possibility. When Socrates says immediately afterwards ‘And previously, justice and piety as well [α]21 had appeared to us to be virtually the same too’ (333 b 5–6), that seems to be an attempt to bring the results of the arguments conducted so far close together rather than to sum up the two alternative models (names of one and the same thing (1b); parts of gold (2b)).22 For Socrates does not choose the weakest (and also the most certain) possibility put forward in 331 b 4–6 (µ*λιστα π*ντων), but another one, which vaguely asserts some kind of identity (σχεδν τι τα0τν, ‘virtually the same’). The precise aim of the next part of the argument, in which Socrates talks about the relationship between justice and temperance, cannot be determined as it is not brought to a close (333 b– 334 c).23 When finally (after the ensuing crisis of the dialogue (334 c –338 e), the interpretation of Simonides’ poem (338 e–347 a), and a further crisis of the dialogue (347 a–348 c)) the topic of the unity of virtue is taken up again, Socrates wants Protagoras to agree to an equation of wisdom with bravery (350 c 4–5). But Protagoras refuses because of a logical error in Socrates’ reasoning (350 c 6 ·.). Hence, this line of argument is not continued either. If one looks at all the ‘proofs’ attempted by Socrates, they can be understood collectively as steps to the ‘unity thesis’, i.e. that the names of the individual virtues are merely di·erent names of one and the same thing (1b). A hint at alternative models (1b (‘names’) or 2b (‘gold’)) may be inferred from the first argument at most; with a view to Socrates’ procedure as a whole, a clear aim to keep these two alternatives apart cannot be observed. At the same time the ‘proofs’ presented turn out to be faulty. In the case of wisdom and bravery that feature is explicitly mentioned.24 If one accepts that Plato has Protagoras uncover a mistake made by Socrates, he has thereby provided the reader with a methodologi21 Cf. F. Ast, Lexicon Platonicum sive vocum Platonicarum index (3 vols.; Leipzig, 1835–6; repr. in 2 vols., Darmstadt, 1956), s.v. 22 But cf. O’Brien, ‘Virtue’, 64–5. 23 The structure of this section suggests that no proposition important for the interpretation of the whole dialogue is to be developed there. 24 Cf. Manuwald, Kommentar, 364–9, 372–3, 373–4. See now E. Spinelli, ‘Un Socrate inconcludente? (Prot. 348c–351b)’, in Casertano, Protagora, ii. 496–512.

Created on 4 July 2005 at 14.33 hours page 123

The Unity of Virtue in Plato’s Protagoras

123

cal clue, suggesting that there might be logical errors in Socrates’ reasoning and that there is good reason to regard logical errors in other proofs as implying that wrong conclusions have been consciously drawn. At the same time that means that one cannot simply believe in the results arrived at by this kind of reasoning.25 This situation invites the assumption that Socrates does not introduce the extreme ‘unity thesis’ (1b) into the discussion as his sincere belief or as a serious possibility, but because he would like to test Protagoras by confronting him with a ‘thesis’ diametrically opposed to his own position ( = 3a).26 And there is an additional point: in the very passage in which Socrates makes methodologically important remarks and objects to the mixing up of his questions with views held by him (330 e 7–331 a 1), he does give answers to a hypothetical interlocutor. This person states that Protagoras and Socrates have used the term justice for ‘a thing’ (το1το τ πρ2γµα 3 4νοµ*σατε 5ρτι) and goes on to ask about piety: ‘Do you say that this too [piety] is “a thing”?’ According to his answers, Socrates regards justice and piety each as a πρ2γµα, ‘a thing’ (330 c 2–e 2). There is no indication in the text that justice and piety are di·erent terms for the same ‘thing’. The question ‘Do you say that this very thing [το1το α0τ τ πρ2γµα] is by its nature such as to be impious or such as to be pious?’ is answered by Socrates with special emphasis: ‘I should be annoyed at such a question . . . and answer: “Watch what you say, sir! How could anything else be pious if piety itself is not to be pious!”’ (330 d 5–e 1). This discussion of the quality of ‘a thing’ presupposes that the name of virtue refers to ‘a thing’ (cf. also 349 b 3–4). And the specific phrasing used virtually characterizes each virtue as a (Platonic) Idea.27 Socrates’ speech is solemn and 25 O’Brien thinks that the problem of whether Socrates’ arguments are successful is a di·erent question, which need not be dealt with in this context (‘Virtue’, 86 n. 39). But the problem becomes essential if there is good reason to suppose that there are deliberate logical errors which the reader is meant to recognize. 26 It is certainly true that Protagoras’ position is similarly refuted by both (1b) and (2b) (O’Brien, ‘Virtue’, 64–5); it would have been refuted already by an opposing theory not that extreme (µ*λιστα π*ντων . . ., 331 b 5–6), but Protagoras’ claim to knowledge is deconstructed in a particularly e·ective manner if he is helpless even when confronted with the most extreme contrary thesis. 27 Cf. Manuwald, Kommentar, 250–3; also Trabattoni, ‘Unit›a’, 285–7. There are significant parallels between the concepts expressed in Prot. 330 c–e and texts evidently based on the doctrine of Ideas. Phaedo 100 c 4–6 reads: ‘It seems to me that if anything else is beautiful besides the beautiful itself, it is beautiful for no reason at all other than that it participates in that beautiful; and the same goes for

Created on 4 July 2005 at 14.33 hours page 124

124

Bernd Manuwald

endowed with a kind of religious emphasis; that suggests that he is putting forward a statement of fundamental importance rather than voicing a truism. Therefore, this passage strictly tells against the view that Socrates wants to regard the names of the virtues as mere names of one and the same thing.28 Finally, Socrates himself talks of ‘parts of virtue’ in a passage all of them’ (trans. D. Gallop). In structure, this concept corresponds to Prot. 330 d 8–e 1: ‘How could anything else be pious if piety itself is not to be pious!’ According to both passages, beauty and piety respectively are beautiful or pious, and in each case this fact is the precondition for each individual thing to be beautiful or pious. And it is a constituent feature of the Idea (e.g. the beautiful itself) that it does not take in anything contrasting with the quality that is transmitted by participating in the Idea (cf. Phaedo 102 d; Rep. 479 a). Phrases such as ‘justice is just’ are likely to be interpreted as positive formulations of this fact and as expressions of the concept that the Idea is endowed with the respective quality in an unrestricted sense, whereas an entity created as an image participates in it to a limited extent only. F. Ferrari (‘I Pragmata e il problema dell’unit›a della virtu’, › in Casertano, Protagora, i. 292–300) thinks that this passage does not o·er su¶cient evidence for the assumption of Ideas (299–300). But his thesis (to mention only that) that πρ2γµα means ‘azione’ (297–300) seems implausible to me. R. M. Dancy (Plato’s Introduction of Forms (Cambridge, 2004), 68–75) convincingly views the introduction of πρ2γµα as a means to distinguish between ‘“name” and “thing (named)” ’ (73). The fact that ‘Protagoras allows unhesitatingly that justice and piety are “things” ’ is regarded by Dancy (69–70) as compatible with earlier remarks of Protagoras (324 d 7–325 a 4). He concludes: ‘When he [sc. Protagoras] concedes that there is such a thing as justice, he does not concede anything about Forms, since he knows nothing about them. So Forms are irrelevant to the argument’ (71). But that does not really hit the mark. What is at issue here is not that justice or piety are talked about as πρ*γµατα, but the way in which that is done. Doubtless, Protagoras is able to understand statements such as ‘justice is a πρ2γµα’ on the basis of his existing knowledge. And confronted with the disjunctive pair (whether piety is pious or impious), he seems to accept that one of the two predicates must be correct; since piety cannot be impious for him, the only choice left is that piety itself is pious too. But that does not rule out the possibility that Socrates’ remark ‘How could anything else be pious if piety itself is not to be pious!’ (330 d 8–e 1) might be viewed against the background of the doctrine of Ideas and of its specific form of participation of each individual in the Idea. Socrates’ emphasis on this statement could not be explained if its possible interpretation was limited to the level assumed for Protagoras. That the doctrine of Ideas is not developed and outlined further in the Protagoras is not a counter-argument either. The same is true for the barely sketched distinction between the relative and the absolute form of the art of measurement; a more detailed investigation of this matter is postponed to a later point (356 c 4–357 c 1; cf. below, pp. 128–9). 28 It may be of some importance in this context that the word χρµατα is used as a collective term for the individual virtues which Socrates will prove to be knowledge (361 b 1–2). If a possible misunderstanding of the virtues as things were to be avoided or it were to be left undecided whether Socrates possibly sees the names of the virtues as mere names of one and the same thing, Plato could have chosen a phrase other than π*ντα χρµατ*
στιν (e.g. π2σαι ρετα ε6σιν).

Created on 4 July 2005 at 14.33 hours page 125

The Unity of Virtue in Plato’s Protagoras

125

often overlooked by interpreters (353 b 2). Again, this term comes up in an answer with which Socrates reacts to Protagoras’ indignant question why one should have to deal with the opinion of the masses. One could object that Socrates is making concessions to Protagoras’ level of understanding without adopting his opinion. This possibility cannot be completely ruled out, but one would have to assume that Plato had his Socrates choose his words carelessly. Plato could have his Socrates talk about the relationship (for example) between bravery and the other virtues if he wanted to avoid the impression that his Socrates believed in parts of virtue as well. These observations do not point to the conclusion that the hypothesis that the names of the individual virtues are merely di·erent names of one and the same thing (1b) is a potential final result of the dialogue. O’Brien explicitly warns against the assumption ‘that Socrates can necessarily be taken as endorsing one of the two rival theories to the exclusion of the other’.29 But if one of them is not an option, as has been shown, and the dialogue is open only to these two solutions (as O’Brien seems to think), the other one must necessarily be the correct one (2b), if one grants O’Brien’s premisses. Hence the question arises whether there are positive or negative clues to such a solution. The overview of the arguments looked at so far has demonstrated that no consistent tendency towards hypothesis (2b) can be observed. Even if one includes Socrates’ later attempt to define the relationship between bravery and wisdom (359 a 2 ·.), there is no evidence for the model connected with the image of gold being the preferred one. If one takes the result (phrased as a question) literally (‘The knowledge of what is dangerous and what is not dangerous is bravery . . .?’, 360 d 4–5), a part of knowledge characterized by what it refers to is defined as bravery.30 29 O’Brien, ‘Virtue’, 126; cf. also 82. 30 O’Brien believes that just this result is open to his two theses (1b) and (2b). But this ‘definiton’ of bravery and the other definition, rejected by Protagoras as logically faulty (identification of σοφα and νδρεα, 350 c 4–5), cannot be valid at the same time, and only the latter agrees with the idea that the names of the individual virtues are merely di·erent names of one and the same thing (1b). The fact that the definition rejected by Protagoras is not defended by Socrates, but replaced by another one, may also be interpreted as a hint given by the author that thesis (1b) will not be maintained as a possible solution in the Protagoras. Perhaps in O’Brien’s view the second definition (360 d 4–5) does not mean that a part of knowledge determined by what it refers to is separated, but that knowledge of all good and evil is adapted

Created on 4 July 2005 at 14.33 hours page 126

126

Bernd Manuwald

If that proposition were transferred to other virtues, various areas of knowledge linked to di·erent reference points would result.31 If one realizes that the ‘definition’ of bravery concluding the second attempt is reached in a logically unacceptable manner32 and is basically equivalent to the definition in the Laches, which contributes to the aporetic ending of this dialogue,33 one will become cautious about drawing positive conclusions from this result. One also has to bear in mind that one passage in the Protagoras points to understanding the virtues as Ideas (330 c 2–e 2).34 This evidence argues clearly for qualitatively distinct virtues,35 which can hardly be reconciled with the parts of gold di·erent only in quantity. In order to show the importance of the image of gold for a solution to the problem of the unity of the virtues posed in the Protagoras, O’Brien makes use of another argument: he takes as a starting-point the fact that at the end of the dialogue the personified #ξοδος τ(ν λγων (‘conclusion of the dialogue’, 361 a 4) reproaches Socrates for aiming to show the whole of virtue to be knowledge.36 In O’Brien’s view that does not agree with the way in which Socrates comes back to the image of gold later in the dialogue, when wisdom is regarded as one of the parts of virtue (349 b–c); because of its phrasing, this passage does not represent Socrates’ real opinion, but that opinion might perhaps be found in Socrates’ first reference to the image, where the individual virtues are compared to parts of gold (before wisdom is mentioned) (329 c–d).37 In O’Brien’s view it is important that knowledge itself is not one to a specific range of action (cf. O’Brien, ‘Virtue’, 95; quoted below p. 127). But the wording of the text o·ers no basis for such an interpretation (cf. also La. 199 c 3–4). 31 O’Brien (‘Virtue’, 99) correctly states: ‘The pieces of gold are all parts of gold. They di·er, but not because one part has some quality or dynamis that the other lacks.’ The parts di·er in size only (cf. 329 d 6–8; O’Brien, loc. cit.). 32 Cf. Manuwald, Kommentar, 427–8, 437 (on Prot. 360 a 4–5), 438 (on 360 b 2–3), 440 (on 360 b 7). 33 On the relationship between the Protagoras and the Laches cf. Manuwald, Kommentar, 429–30, and on the interpretation of the aporia in the Laches cf. B. Manuwald, ‘Die Schlu¢aporie in Platons Laches’ [‘Laches’], Rheinisches Museum, 143 (2000), 179–91. 34 Cf. above, pp. 123–4 and n. 27. 35 According to the Phaedrus, one does not see only δικαιοσνη and σωφροσνη in themselves, i.e. as Ideas, in the 8περουρ*νιος τπος, but particularly
πιστµη (Phdr. 247 d 5–e 2). As regards their way of being, these three entities stand beside each other on equal terms. 36 ν1ν δ: ε6 φανσεται [sc.  ρετ]
πιστµη λον, <ς σ= σπεδεις, > Σκρατες (361 b 5–6). 37 O’Brien, ‘Virtue’, 86.

Created on 4 July 2005 at 14.33 hours page 127

The Unity of Virtue in Plato’s Protagoras

127

part of virtue;38 his detailed interpretation of this model runs as follows: ‘Gold’ is the knowledge of good and evil. Pieces of gold are the individual virtues. Each piece has all the qualities of every other piece and of the whole, because we cannot know good and evil only in part and because each virtue therefore requires knowledge of all good and all evil. In so far as the individual virtues circumscribe our knowledge, it is not by any intrinsic limitation to or diminution of our knowledge, but because each of the virtues requires our knowledge of the whole to be adapted to some specific range of action or behaviour.39

There are di·erences between Socrates’ first questions about the unity of the virtues (329 c ·.) and the later recapitulation (349 b–c) caused by the course of the discussion,40 but it is rather doubtful whether the di·erences concerning the role of wisdom are to be interpreted as O’Brien assumes.41 That wisdom and bravery are introduced in 329 e–330 a only after justice, temperance, and piety (329 c 7–8) is caused by the structure of the dialogue, because only these three virtues are explicitly mentioned in Protagoras’ great speech, to which Socrates’ questions refer. Thus, in employing the image of gold Socrates has not made a selection of virtues adapted to a specific hypothesis, but taken up what he has heard from Protagoras.42 The separate later mention of wisdom and bravery (first by Protagoras) is due to the development of the discussion the author has chosen; it can be seen as a compositional means of bringing out those two virtues, whose relationship to each other will become decisive in the debate with Protagoras. That wisdom is introduced later can easily be explained without regarding it as a hint that Socrates does not want to see wisdom as a part of virtue. Further, O’Brien’s interpretation does not explain what the di·erence in size between the parts of gold means. At any rate, in O’Brien’s reading all virtues would be knowledge (and nothing else) and endowed with the full knowledge of good and evil, though limited in quantity according to the image of gold. Despite all its inherent problems, such a theory cannot be excluded 38 Ibid. 95 n. 52. 39 Ibid. 95. 40 Cf. also Manuwald, Kommentar, 361–2. 41 Cf. also O’Brien, ‘Virtue’, 74 ·. 42 Incidentally, if O’Brien’s method were followed consistently, the fact that σοφα and νδρεα are mentioned only later would have to be referred to the image of the face as well, since it too is introduced before these two virtues are mentioned. What conclusion should be drawn from this?

Created on 4 July 2005 at 14.33 hours page 128

128

Bernd Manuwald

as a possible model. But there is nothing in the development of the dialogue to suggest that Plato is o·ering as a possible solution either the model based on the image of gold (2b) or the idea that the names of the individual virtues are merely di·erent names of one and the same thing (1b). Does the aporetic ending of the dialogue then mean that no Platonic doctrine may be derived from it?

3. What solutions does the dialogue Protagoras allow? A promising method for finding an answer to this question might be to look for statements which are marked as fixed points in the dialogue and then to consider in what way they may go together. In some instances, however, their coherence may be recognized only if one takes into account doctrines advanced in later dialogues. But if one considers the situation of contemporary readers, particularly readers from Plato’s circle, who knew his concepts from oral discourse, one may legitimately adopt this method. For contemporary controversies or oral discussions in Plato’s circle, or simply his thinking, could have been more comprehensive than the written testimony explicitly indicates. For instance, this method of interpretation can be shown to yield plausible results in the case of the Laches.43 In this dialogue the interlocutors agree on some statements that seem irreconcilable, which in the context of the Laches leads to aporia; these statements do, however, make sense in their specific combination if one interprets them on the basis of the doctrine of the soul and virtue outlined in the Republic. That is hardly a coincidence; it rather suggests that right from the start the Laches was composed so as to be open to notions spelt out in the Republic, or, in other words, that these notions were already present in the background when the Laches was written. This instructive method may be employed with even greater justification in the case of the Protagoras. For in this work one passage clearly shows that the author is relying on a wider-ranging concept than that spelt out in the dialogue: more detailed discussion of the ‘art of measurement’ is postponed to another occasion (357 b 5–6). But a closer analysis of the complete section (356 c 4–357 c 1) 43 No detailed account can be given here; for a more extended discussion see Manuwald, ‘Laches’.

Created on 4 July 2005 at 14.33 hours page 129

The Unity of Virtue in Plato’s Protagoras

129

reveals that two forms of the art of measurement are already being distinguished, an absolute one (described with expressions recalling the characterization of the world of Ideas)44 and a relative one; that distinction looks forward to a comparable one in the Politicus (283 c ·.).45 O’Brien, and perhaps other scholars, would probably object to employing later dialogues such as the Republic in interpreting the Protagoras and would relegate this method to the category of ‘syncretistic’ approaches, since O’Brien posits that Plato’s thinking has radically changed by the time of the Republic, at least with respect to certain aspects of the doctrine of virtue.46 This problem cannot be followed up here in detail, but for simple methodological reasons it is essential first to look in an unbiased way at what the texts say. Plato never provides any hint that his thinking might have radically changed and that he may be distancing himself from what he had said earlier. And if one remains aware of Plato’s specific view of written doctrine (cf. Phdr. 274 c ·.), it is not improbable to assume that a dialogue laid down in writing may be based on more elements of Platonic doctrine than it explicitly says. Socrates and Protagoras are agreed that knowledge of good and evil (i.e. the complete knowledge in which virtue resides) is a sufficient condition for acting correctly,47 and Socrates explicitly calls this statement true (ληθ, 352 d 4). Hence acting correctly in each area (and thus each virtue) depends on knowledge. That corresponds with the concept that the whole of virtue is knowledge, which the ‘conclusion of the dialogue’ (#ξοδος τ(ν λγων) attributes to Socrates (361 b 5–6), but it seems to contradict Socrates’ remark mentioned above, implying that he regards wisdom as one of the 44 See also above, pp. 123–4 and n. 27. 45 For further details see Manuwald, Kommentar, 396–9. M. Wesoly (‘La salvezza della vita: “un’arte e una scienza del misurare” (Protagora, 356c–357a)’, in Casertano, Protagora, ii. 513–27) seems not to discuss the distinction in the Protagoras mentioned above; likewise Leszl (‘Le funzioni’, 624) notices only the relativity of the art of measurement in the Protagoras. O’Brien (‘Virtue’, 129) focuses only on the negative component of the postponement, and he holds that the dialogue does not explain ‘the nature of the “expertise” (τ χνη) and the “knowledge” (
πιστµη) that would allow us to choose between pleasures and pains, between good and bad’. In fact, on the one hand the dialogue makes it clear that Socrates (in contrast to the masses) knows of a τ λος other than pleasure, namely the good (cf. Manuwald, Kommentar, 393–4); on the other hand, the postponement indicates that Socrates already has some concept of the art of measurement, but does not develop it in greater detail at this point. 46 O’Brien, ‘Virtue’, 103–4. 47 This might be a convenient summary of paragraph 352 c 2–d 4.

Created on 4 July 2005 at 14.33 hours page 130

130

Bernd Manuwald

parts of virtue as well.48 This consequence follows from the fact that among the other parts whose relationship with bravery is to be defined the most important one is wisdom, since it is the relationship between these two virtues alone that is dealt with in the last part of the argument (359 a 2 ·.).49 The whole, one is tempted to conclude, cannot be a part at the same time. But there is also the question of whether in Plato’s view what seems at first glance to be a contradiction is meant to be a proper contradiction (where only one of the statements can be correct) or whether the statements might actually agree with each other. Looking forward to Platonic doctrine as developed in the Republic may point to a solution to the paradox of whole and part in the Protagoras. As Socrates explains in the Republic, the smallest and superior part of the soul, the logistikon, which is wise (σοφν) and has forethought (προµθεια) for the whole soul (Rep. 441 e 4–5), has the knowledge of what is useful to each of the three parts of the soul and to the whole soul (Rep. 442 c 5–8). That means that the logistikon (at least that of the philosopher) has the complete knowledge of good and evil, which includes the knowledge of what is dangerous and what is not dangerous. Thus it can inform the spirited part of the soul (θυµοειδ ς) of what is dangerous and what is not, and what it is to follow in good and in bad times (δι* τε λυπ(ν κα@ δον(ν) (Rep. 442 b 11 ·.). Temperance (σωφροσνη) exists when the ruling parts as well as the ruled parts of the soul are agreed that the logistikon must rule (Rep. 442 c 10–d 1). That means that temperance too presupposes the rule of wisdom (σοφα) and thus requires it in an ordered cosmos of the soul. The same is true for justice: wisdom is the knowledge (
πιστµη) that governsa just action (Rep. 443 c 9–444 a 2, esp. 443 e 6–7). A person endowed with these qualities is good (γαθς, Rep. 449 a 1–2); that means he has the one and only (complete) virtue (ρετ), whereas the number of forms of evilness is unlimited (Rep. 445 c 5–6). Similarly, the completely good polis (Rep. 427 e 6–11) 48 Cf. 353 b 1–3 πρς τ
ξευρε-ν περ@ νδρεας, πρς τAλλα µρια τB τς ρετς π(ς ποτC #χει. See above, pp. 124–5. 49 In a detailed discussion of the opinion of G. Vlastos (‘Socrates on “the Parts of Virtue”’, in id., Platonic Studies, 2nd edn. (Princeton, 1981), 418–23), who regards the concept that the individual virtues are parts of virtue as a ‘standard Socratic doctrine’, O’Brien (‘Virtue’, 103–10) arrives at the conclusion that neither in the Protagoras nor in the Meno is Socrates of the opinion that knowledge is a part of virtue, but rather believes that knowledge is the whole of virtue.

Created on 4 July 2005 at 14.33 hours page 131

The Unity of Virtue in Plato’s Protagoras

131

is endowed with all cardinal virtues; thus despite their di·erences, they form a unity and cannnot occur separately, at least in their perfect shape.50 The logistikon and its superior knowledge can be seen as an equivalent to the knowledge of good and evil in the Protagoras (352 c 2–d 4) and the virtue that consists in knowing all good and evil in the Laches (199 d 4–e 4). In the Republic too wisdom (in contrast to other virtues—cf. Rep. 536 a 2–3) is not explicitly called a part of virtue; implicitly, however, it appears as such,51 and, most importantly, it is connected exclusively with one part of the soul and at the same time embodies the complete knowledge in which virtue resides. The paradox of whole and part is resolved here, of course not in the sense of genus and species (in which case the contradiction would remain unresolved), but within the specific framework of the Platonic doctrine of the soul. The aporetic ending of the Protagoras indicates that a proper solution to the problems has not yet been found within the dialogue itself. The search is not continued immediately; however, it is not Socrates who is responsible for that, but Protagoras, whose refusal to continue the analysis there and then proves him to be entirely unphilosophical. If one assumes from the indications given by the discussion between Socrates and Protagoras and the ending of the dialogue that Socrates would be able to develop a solution to the question of unity (to be endorsed by him), the situation is parallel to the Laches, which may also be interpreted on the basis of the Republic. To return to the models given by Socrates’ questions: if neither hypothesis (1b) (‘names’) nor the combination (2b+ 3b) (‘parts’ of gold which cannot be had separately) can be proved to be the opinion of the Platonic Socrates on the unity of virtue, it will not 50 In Plato’s view, this concept of virtue does not exclude simpler, ‘more popular’ forms of virtue (δηµοτικ κα@ πολιτικ ρετ), which are created by habit and practice ‘without philosophia’ (cf. Phaedo 82 a 11–b 3; Rep. 500 d 7–8; 518 d 9 ·.). But this kind of virtue is not ‘true virtue’ (ληθς ρετ), which requires understanding (φρνησις) (cf. Phaedo 69 a–c). The comparison of the Protagoras (and the Laches) to the Republic refers to virtue in its full sense. 51 Cf. Rep. 484 d 7. The philosopher-guardians certainly have σοφα (they have an
ναργ:ς
ν τD ψυχD . . . παρ*δειγµα, cf. Rep. 484 c 7–8) and are not to be inferior in any other part of virtue (µηδC
ν 5λλ'ω µηδεν@ µ ρει ρετς, d 7). In the Laches Socrates says, when discussing the parts of virtue: ‘Apart from bravery I [
γ] mention temperance [σωφροσνη] and justice [δικαιοσνη] and other qualities of this kind’ (La. 198 a 7–9). This statement does not explicitly include σοφα, but it does not explicitly exclude it either.

Created on 4 July 2005 at 14.33 hours page 132

132

Bernd Manuwald

necessarily follow that the combination (2a+ 3b) (parts of the face which cannot be had separately) represents his opinion. Nevertheless, a concept which posits parts of virtue, not identical in quality (though not contrasting with each other either) and depending on knowledge, and at the same time insists that the individual virtues cannot be had separately and that knowledge represents the whole of virtue, may emerge as Socrates’ opinion in the Protagoras, one which can be reconciled with his assertions in the Republic. To use the phrases of the image: the knowledge which gives unity (σοφα) can be envisaged as the face; the other virtues (parts of the face such as nose, eyes, etc.) cannot exist without the face as a whole (σοφα) and they cannot be had each on its own, but they each have their own power (δναµις) and do not consist solely of wisdom; in this respect σοφα is also a ‘part’. Various objections might be raised against this interpretation. For instance, one could argue that the Protagoras deals merely with virtue understood as knowledge and that other elements of virtue are not being discussed, in contrast to the Republic, where, for instance, bravery is also defined as a ‘power which saves’.52 This objection could be answered by pointing out that in the Republic too the most important thing for virtue as a whole is knowledge (Rep. 441 e 4–6; 442 c 5–8) and that thus bravery too depends on the knowledge of the logistikon (Rep. 442 b 11–c 3).53 On the other hand, the ‘intellectualist’ early dialogues may include the nonintellectualist component already, as the Laches indicates: Socrates’ remark at La. 194 a 1–5 shows that in his opinion too bravery is connected with ‘endurance’ (καρτ ρησις). This phenomenon may be interpreted as the result of di·erent accentuations of a common underlying doctrine. At any rate, Plato’s thinking has not radically changed by the time of the Republic. One can only observe with 52 Rep. 430 b 2–3 δναµιν κα@ σωτηραν διB παντς δξης Fρθς τε κα@ νοµµου δειν(ν τε π ρι κα@ µ, cf. Rep. 429 b 9–c 1 δναµιν . . . G διB παντς σσει τν περ@ τ(ν δειν(ν δξαν, 442 c 1 διασ'ζDη (of bravery in the soul). 53 Of course, the (ordinary) guardians have only to preserve the belief (δξα), i.e. the right belief (δξα Fρθ), about what is to be feared (Rep. 429 c 7–8; 430 b 2–3). But this belief is pronounced by the lawgiver (+ νοµοθ της παργγελλεν, Rep. 429 c 2), and who should the lawgiver be if not one of the citizens who possess knowledge (Rep. 428 c 11–d 3)? The relationship between knowledge and bravery becomes even more explicit in the case of the soul: ‘we call each individual courageous . . . when the part of him that is spirited in kind [τ θυµοειδ ς] preserves . . . the pronouncements [τ . . . παραγγελθ ν] of reason about what should inspire terror and what should not’ (Rep. 442 b 11–c 3, trans. C. D. C. Reeve).

Created on 4 July 2005 at 14.33 hours page 133

The Unity of Virtue in Plato’s Protagoras

133

respect to the doctrine of virtue that in its written version it has become more di·erentiated. It may seem odd that Plato’s Socrates does not say that he now o·ers (at least provisional)54 solutions to problems that had not, or not explicitly, been solved in earlier aporetic dialogues. But would Socrates’ silence not be even stranger if Plato had changed his concept fundamentally? Why did Plato not let Socrates explicitly refute opinions that may be inferred from earlier dialogues and have now been abandoned?55 One might also sceptically enquire after the function of the various hypotheses on the unity of the virtues Socrates mentions to Protagoras, if the very alternative (1b) argued for most strongly as a contrast to Protagoras’ position is not an o·er of a positive solution to the problem. That objection might be countered by an answer from the internal structure of the dialogue, which is designed to establish clearly what exactly Protagoras means and to check whether he can defend his position even against the extreme unity model (1b): that is, in the context of the dialogue, defeat of the sophist is the aim. But perhaps something else is at issue as well, and it is in this very context that O’Brien has provided an important, albeit not completely new, impetus: he draws parallels (with appropriate caution) between the hypotheses mentioned in the Protagoras and general ideas which are known doxographically, even if not all of them are attested for Plato’s time.56 Perhaps, under cover of a discussion dating back to the time of Protagoras, Plato is surveying some theories that he was aware of—either because they were generally known, such as that attested for the Megarians,57 or because they were be54 That the remarks in the Republic have to be supplemented as regards the doctrine of virtue as well is clear from the outline of the text (Rep. 430 c 3–5; 504 a–e; 611 b 10–c 5). Cf. also C. H. Kahn, Plato and the Socratic Dialogue: The Philosophical Use of a Literary Form (Cambridge, 1996), 222. He remarks that ‘Even in the Republic . . . Plato insists upon the inadequacy of the account he has given.’ As regards the Protagoras, Kahn (loc. cit.) suggests ‘that Plato in the Protagoras has deliberately left the thesis of unity indeterminate and open for further discussion’. 55 Rep. 438–9 is no exception; it is compatible with e.g. Gorg. 468 a ·., if one follows J. Adam’s reading of Rep. 438–9 (The Republic of Plato (ed. and comm.), i. Books I–IV (Cambridge, 1905), 251, on Rep. 438 a 3). 56 O’Brien, ‘Virtue’, 111 ·. 57 Cf. test. 25 D•oring = II A 32 Giannantoni = D.L. 7. 161: ρετ*ς τC οIτε πολλBς ε6σγεν [sc. Jρστων, SVF i. 351], <ς + Ζνων [SVF i 200], οIτε µα ν πολλο-ς Fνµασι καλουµ νην, <ς οL Μεγ α ρικ ο, λλB κατB τ πρς τ πως #χειν. Cf. O’Brien, ‘Virtue’, 117–18.

Created on 4 July 2005 at 14.33 hours page 134

134

Bernd Manuwald

ing discussed in Plato’s circle—or that he regarded as constituting possible objections; indirectly, they are shown not to yield a satisfying solution. But that does not necessarily mean that Plato did not have a possible solution in mind at the time the Protagoras was composed. It is only that it cannot be found simply on the basis of the means provided by the Protagoras alone. University of Cologne

B I B L I O GR A P HY Adam, J., The Republic of Plato (ed. and comm.), i. Books I–IV (Cambridge, 1905). Ast, F., Lexicon Platonicum sive vocum Platonicarum index (3 vols.; Leipzig, 1835–6; repr. in 2 vols., Darmstadt, 1956). Burnet, J., Platonis Opera, vol. iii, 2nd edn. (Oxford, 1909). Casertano, G. (ed.), Il Protagora di Platone: struttura e problematiche [Protagora] (2 vols.; Naples, 2004). Dancy, R. M., Plato’s Introduction of Forms (Cambridge, 2004). Denniston, J. D., The Greek Particles, 2nd edn. (Oxford, 1954). Kahn, C. H., Plato and the Socratic Dialogue: The Philosophical Use of a Literary Form (Cambridge, 1996). Ferrari, F., ‘I Pragmata e il problema dell’unit›a della virtu’, › in Casertano, Protagora, i. 292–300. Leszl, W., ‘Le funzioni della tesi edonistica nel Protagora e oltre’ [‘Le funzioni’], in Casertano, Protagora, ii. 574–638. • Manuwald, B., Platon: Protagoras. Ubersetzung und Kommentar [Kommentar] (G•ottingen, 1999). ‘Die Schlu¢aporie in Platons Laches’ [‘Laches’], Rheinisches Museum, 143 (2000), 179–91. O’Brien, D., ‘Socrates and Protagoras on Virtue’ [‘Virtue’], OSAP 24 (2003), 59–131 = ‘Socrate e Protagora sulla virtu’, › in Casertano, Protagora, i. 173–250. Penner, T., ‘The Unity of Virtue’, Philosophical Review, 82 (1973), 35–68 (several reprints). Spinelli, E., ‘Un Socrate inconcludente? (Prot. 348c–351b)’, in Casertano, Protagora, ii. 496–512. Taylor, C. C. W., Plato: Protagoras, trans. and comm., rev. edn. (Oxford, 1991). Trabattoni, F., ‘Unit›a della virtu› e autopredicazione in Protagora 329e– 332a’ [‘Unit›a’], in Casertano, Protagora, i. 267–91.

Created on 4 July 2005 at 14.33 hours page 135

The Unity of Virtue in Plato’s Protagoras

135

Vlastos, G., ‘The Unity of the Virtues in the Protagoras’ [‘Unity’], Review of Metaphysics, 25 (1972), 415–58; repr. ‘with substantial additions and corrections in both text and notes’ in id., Platonic Studies (Princeton, 1973), 221–69, and with further additions in the 2nd edn. (Princeton, 1981), 427–45. ‘Socrates on “the Parts of Virtue” ’, in id., Platonic Studies, 2nd edn. (Princeton, 1981), 418–23. Wesoly, M., ‘La salvezza della vita: “un’arte e una scienza del misurare” (Protagora, 356c–357a)’, in Casertano, Protagora, ii. 513–27.

Created on 4 July 2005 at 14.33 hours page 136

page 136 is blank

Created on 4 July 2005 at 14.33 hours page 137

S HAME, P LEAS URE, AND T HE DI VI DED S O UL JESSICA MOSS

Thumos—spirit, the middle part of the soul in Plato’s Republic— has a bad reputation. In one of the most sustained attacks on its integrity, Terry Penner writes: Plato’s true view [in the Republic] is . . . that there are only two parts of the soul, a rational part and an irrational part, and he allows himself thumos (spirit) for irrelevant political or moral reasons only . . . Plato had no logical or psychological arguments for going beyond two parts of the soul.1

Many scholars share Penner’s suspicions. Spirit is at best poorly unified and ill-defined, and at worst an ad hoc addition to Plato’s underlying view of the soul as the battleground between reason and non-rational desire, an afterthought which he introduces in book 4 to preserve his dubious analogy between city and soul, and then soon forgets, reverting in book 10 to his original view.2 Some have defended spirit by arguing that in describing it Plato ã Jessica Moss 2005 For all their help and good advice I thank John Cooper, Helen Cullyer, Cian Dorr, Alexander Nehamas, David Sedley, and Allan Silverman. 1 T. Penner, ‘Thought and Desire in Plato’, in G. Vlastos (ed.), Plato: A Collection of Critical Essays, ii. Ethics, Politics, and Philosophy of Art and Religion (New York, 1971), 96–118 at 96 and 113. 2 Allegedly at Rep. 602 c ·., where he distinguishes between the rational part of the soul, the part that calculates and measures, and an irrational part that falls prey to illusions. Hardie shares Penner’s view: ‘No adequate reasons are given by the discussion in the Republic for thinking [that spirit is a distinct third part of the soul] . . . or for modifying the opinion, so strongly supported both by plain men and philosophers, that the fundamental division here is into two and not three’ (W. F. R. Hardie, A Study in Plato (Oxford, 1936), 142); cf. T. M. Robinson, Plato’s Psychology, 2nd edn. (Toronto, 1995), 44–6. A more moderate version of this view is expressed by those who think that Plato posited spirit only because he was interested in the spirited type of character, or the spirited (honour-loving, military) life. B. Williams, ‘The Analogy of City and Soul in Plato’s Republic’, in E. N. Lee, A. P. D. Mourelatos, and R. M. Rorty (eds.), Exegesis and Argument: Studies in Greek Philosophy Presented to Gregory Vlastos (Phronesis suppl. 1; Assen, 1973),

Created on 4 July 2005 at 14.33 hours page 138

138

Jessica Moss

has captured a set of psychological phenomena that are importantly similar to one another, and importantly dissimilar from those he attributes to reason or appetite.3 But such defences, even when compelling, leave us with the unfortunate impression that spirit is the result merely of some insightful psychological speculation on Plato’s part, a flourish that does no vital philosophical work. Against these charges I argue that spirit is central to Plato’s ethical and psychological thought, because it is indispensable to his conception of virtue. Given that people have appetitive desires, and that these tend to lead us astray, without spirit there would be no systematic way to ensure that we overcome these desires and act in accord with reason instead. I identify the foundation for this view of spirit in the Gorgias’ treatment of shame, arguing that Socrates uses shame in this dialogue as a tool for undermining the attraction of ethically harmful pleasures. I then show that the Republic develops this opposition between pleasure and shame in its characterization of spirit as the ally of reason against the appetitive part of the soul. Lastly, I argue that spirit is indispensable because it forms a crucial link between reason and the non-rational: unlike reason, it is a powerful motivational force in those who are not virtuous; unlike appetite, it can be shaped and guided to lead those who are not virtuous towards virtue. And therefore Plato’s view of the soul in the Republic is fundamentally tripartite. It is crucial to his view that 196–206 at 205–6 is typical. Others countenance spirit but doubt that Plato has any clear view about what it is (see e.g. J. Annas, An Introduction to Plato’s Republic (Oxford, 1981), 126–8; P. Shorey, What Plato Said (1933) (repr. Chicago, 1979), 224; and Reeve’s comment (in defending spirit against such criticism) that spirit is ‘the dark horse of the psychic parts’ (C. D. C. Reeve, Philosopher-Kings: The Argument of Plato’s Republic [Philosopher-Kings] (Princeton, 1988), 136)). Underlying the charge that Plato had not worked out his conception of this part of the soul may be the view that it has no legitimate place in his psychological theory. 3 ‘[T]he motivations that Plato classifies under the heading of spirit are to be understood as having their root in competitiveness and the desire for self-esteem and . . . esteem by others’ (J. M. Cooper, ‘Plato’s Theory of Human Motivation’ [‘Motivation’], History of Philosophy Quarterly, 1 (1984) 3–21, repr. in Cooper, Reason and Emotion: Essays on Ancient Moral Psychology and Ethical Theory (Princeton, 1999), 118–37 at 133–4); each spirited desire ‘involves a conception . . . of the good, and a desire to keep us in line with that conception when our own behaviour, or that of others, disturbs the match between reality and ideals’ (Reeve, Philosopher-Kings, 138); ‘the essence of the human thumos is the need to believe that one counts for something, and that central to that need will be a tendency to form an ideal image of oneself in accordance with one’s conception of the fine and noble’ (A. Hobbs, Plato and the Hero: Courage, Manliness and the Impersonal Good [Hero] (Cambridge, 2000), 30).

Created on 4 July 2005 at 14.33 hours page 139

Shame, Pleasure, and the Divided Soul

139

reason is opposed, not merely by non-rational desire generically understood, but in particular by desires for pleasure (and aversions to pain). It is also crucial to his view that the soul can be redeemed by the workings of a third force, spirit, one able to dispel the lure of pleasure where reason on its own will fail.

1. Polus There are three objects of choice and three of avoidance: the kalon, the advantageous, and the pleasant, and their opposites, the aischron, the harmful, and the painful. (Arist. NE 1104b30–2)4

In the Gorgias, Socrates uses or is accused of using appeals to shame to refute each of his three interlocutors. Shame (it is said) makes Gorgias concede that he could and should teach his students about justice; it makes Polus concede that justice is better than injustice; and it makes Callicles concede that hedonism is false. Why does Socrates rely on shame in these refutations, instead of sticking to purely rational arguments? Recent literature o·ers us a compelling explanation. First, Socrates thinks that shame reveals what a person really believes, as opposed to what he says or even thinks he believes: ‘[Socrates’ method is] not to argue [his interlocutors] into believing [what he thinks true] but to maneuver them into acknowledging that deep down they have believed it all along . . . His chief weapon in this psychological warfare is not logic but shame.’5 Second, Socrates thinks that the deep beliefs revealed by shame are true beliefs about what is good and bad (although many of our surface beliefs are dangerously false): ‘Shame reflects a Platonic conception corresponding to our own notion of an innate moral sense . . . Shame operates in this dialogue as an obscure intuition of the good on the part of Socrates’ interlocutors.’6 I am in broad sympathy with these claims as interpretations 4 All translations are mine except where otherwise noted. 5 R. McKim, ‘Shame and Truth in Plato’s Gorgias’ [‘Shame’], in C. Griswold (ed.), Platonic Writings/Platonic Readings (New York and London, 1988), 34–48 at 37. 6 C. Kahn, Plato and the Socratic Dialogue: The Philosophical Use of a Literary Form [Dialogue] (Cambridge, 1996), 138. Cf. Kahn, ‘Drama and Dialectic in Plato’s Gorgias’ [‘Drama’], Oxford Studies in Ancient Philosophy, 1 (1983), 75–121.

Created on 4 July 2005 at 14.33 hours page 140

140

Jessica Moss

of Socrates’ implicit view of shame in the Gorgias (although they ignore the important criticisms of that view o·ered by Callicles, which I discuss in Section 4 below). But there is a crucial point that has been overlooked in the literature: Socrates uses shame in the Gorgias particularly as a force against the lure of pleasure. The Gorgias shows that appeals to a person’s feelings of shame and admiration may be able to succeed, when rational arguments have failed, in bringing him to see that a harmful pleasure is to be avoided, or that a beneficial pain is to be pursued. For on the account I shall o·er, Socrates’ refutations of both Polus and Callicles show that feelings of shame can separate a person’s judgements about what is pleasant from his judgements about what is good. And it is this feature of shame, I believe, that explains Plato’s interest in it as a tool of persuasion.7 The relations between pleasure, goodness, and shame are first brought to the fore in the dialogue in Socrates’ refutation of Polus’ defence of injustice. Polus has been claiming that injustice in various forms is better than justice.8 Socrates has insisted not only that justice is in fact better than injustice, but that Polus himself (along with ‘all other people’) really believes that justice is better than injustice (474 b 2–5; cf. 475 e 3–5). The argument he uses to make Polus concede these points begins with an appeal to Polus’ sense of what is shameful: Socrates asks if injustice is not more shameful than justice, and Polus readily admits that it is so. Then Socrates o·ers an analysis of judgements about the aischron (shameful or ugly) and its opposite, the kalon (admirable, beautiful, noble, or fine).9 When one judges something to be kalon, one does so either because it is pleasant or because it is good (beneficial) or both; when one judges something to be shameful, one does so because it is painful or bad (harmful) or both (474 d–475 b).10 Polus accepts the analysis, and declares when questioned that being unjust is not 7 There are other important aspects of Plato’s treatment of shame in the Gorgias that I do not discuss here. My aim is to examine one very significant and hitherto unexplored aspect of shame in the dialogue. 8 That it is worse to su·er injustice than to commit it (469 b), that unjust people are happy (470 d), and that the unjust person who escapes punishment is happiest, while the one who is punished is made miserable (472 d–e). 9 I shall chiefly use the Greek kalon, because the meaning of the word as Plato uses it in the dialogue is too broad to be conveyed by any single translation. I shall translate aischron as ‘shameful’, for convenience, but the more aesthetic sense should be kept in mind. 10
δι δονν τινα
δι φελαν (474 e 2–3; cf. 474 d 6–8, τοι κατ τν χρεαν . . .

Created on 4 July 2005 at 14.33 hours page 141

Shame, Pleasure, and the Divided Soul

141

more painful than being just, but in fact more pleasant; thus he is forced to concede that it must be worse. Socrates’ argument is a strange one. Polus leaves the conversation feeling duped,11 Callicles later charges Socrates with cheating and equivocation,12 recent literature has found the argument fallacious,13 and certainly it is hard to read the passage without feeling that Socrates argues in a less than straightforward way. But underlying Socrates’ dubious logical argument we can detect an implicit psychological theory about the moral e·ects of shame—a theory that, if true, justifies the refutation. On its surface, Socrates’ argument is about the role that pleasure and benefit (goodness) play in our judgements about what is kalon. On the interpretation I shall o·er, he is primarily interested in the role of pleasure and the kalon in our judgements about what is beneficial (good). He wants to show that these two criteria, pleasure and the kalon, yield conflicting judgements about what things are good: Polus, I shall argue, thinks injustice good because it is pleasant, but comes to see it as bad when he attends to the fact that it is shameful. Socrates also wants to show, more contentiously, that when these two criteria conflict, the second wins out: if we think something good because it is pleasant, but then come to see that it is shameful, we can no longer think it good. To see that this is the underlying rationale of the argument, let us begin by investigating the psychological state Socrates attributes to Polus at the beginning of the argument. First, deep down Polus believes that justice is better than injustice, and this belief is true. In fact everyone is in this condition: ‘I do believe that both I and you and all other people think committing injustice worse than suffering it, and not paying the penalty worse than paying it’ (474 b 2–5; cf. 475 e 3–5). Socrates o·ers no explanation for why we all have true moral beliefs; let us treat this claim simply as an axiom of his psychological theory.14 Second, this true belief is somehow
κατ δονν τινα . . .). Socrates allows Polus’ substitution of ‘good’ for ‘benefit’ at 475 a 3, and goes on to speak of good and bad in place of benefit and harm. 11 See 480 e 1–2. 12 482 c ·. I discuss his specific charges in sect. 4. 13 Beginning with G. Vlastos, ‘Was Polus Refuted?’ [‘Polus’], American Journal of Philology, 88 (1967), 454–60, and G. Santas, Socrates: Philosophy in Plato’s Early Dialogues (London, 1979), 230–46. 14 Vlastos identifies this claim as a presupposition of the elenchus (G. Vlastos, ‘The Socratic Elenchus’, Oxford Studies in Ancient Philosophy, 1 (1983), 27–58; Socrates: Ironist and Moral Philosopher (Cambridge and Ithaca, NY, 1991), 15).

Created on 4 July 2005 at 14.33 hours page 142

142

Jessica Moss

hidden from Polus himself: so far as he is aware, he believes that injustice is better. Let us grant Socrates, for the sake of argument, this distinction between what I shall call ‘deep’ unconscious and ‘superficial’ conscious value beliefs. Socrates thus sees his task as that of clearing away Polus’ false, superficial belief that injustice is good, and revealing his underlying true belief that injustice is bad. How should Socrates proceed? Consider an analogous case: I suspect that you believe deep down that capital punishment is wrong, despite your protestations that it is right. I believe, that is, that deep down your moral intuitions are reliable, but that your judgement has somehow been distorted. Perhaps you have been brought up in a religion or culture that supports capital punishment, or perhaps you have been swayed by your own emotional reaction to the description of some crime. If I want to dispel these distorting influences, I need first to identify them: I need to solicit your reasons for thinking capital punishment right—reasons that may well be obscure to you yourself. Next, I need to show you that these are bad reasons, ones you yourself on reflection do not endorse. On my reading, Socrates’ argument employs precisely this kind of strategy. Polus’ preference for injustice is clearly bound up with his admiration for power. What he means by ‘power’, Socrates reveals, is the ability to take whatever one wants from others without repercussions—that is, the licence to be as unjust as possible without ever submitting to justice. But what are his reasons for thinking this kind of power good? If we look at how Polus has presented and defended his view, what we notice is that he sees little need to advance any reasons at all. He considers his positions to be utterly self-evident, truths of common sense that only a perverse debater like Socrates would deny.15 When he wants to prove that many unjust people are happy, he simply describes the crimes of the tyrant Archelaus, not arguing that he is happy, but saying merely ‘You can picture [ρς] this man Archelaus’—as if Archelaus’ happiness were simply manifest, as things are visually manifest (470 d 5). When he wants to prove that it is bad to be punished for one’s crimes, he 15 He refers to Socrates’ objections as shocking, monstrous, and bizarre (σχτλια, περφυ (467 b 10), and τοπα (473 a 1)), ‘the sort of thing which no human being would maintain’ (473 e 4–5), and asks, ‘Couldn’t even a child refute you, and show that what you say isn’t true?’ (470 c 4–5). He considers his position so manifestly true that he even refuses to believe that Socrates genuinely rejects it: when Socrates claims that he does not agree with him, Polus says ‘Because you don’t want to—since you really do think that it is as I say’ (471 e 1).

Created on 4 July 2005 at 14.33 hours page 143

Shame, Pleasure, and the Divided Soul

143

gives a graphic description of the punishments, not arguing that these are bad but assuming that anyone in his right mind will simply see that this is so (473 b 12–c 5). Polus’ value beliefs are not based on theory or any other rational grounds; they take the form more of gut reactions or intuitions. Are we to conclude that Polus has no reason at all for valuing the power to be unjust, that he thinks its desirability a brute, inexplicable fact? If we look carefully at what he has said, we see instead that there is an underlying explanation for his preference, one that he is not fully aware of and cannot clearly articulate. For Polus’ praises of power and injustice, and denigrations of justice, are in fact unified by two main claims. First, power is desirable in so far as the powerful unjust person can do whatever he pleases.16 Second, power is desirable in so far as the powerful unjust person avoids punishment for his wrongs.17 When we recall that the monstrously unjust tyrant Polus admires becomes, in Republic 9, Plato’s model of a person ruled by his appetites, Polus’ implicit standard of value becomes clear. The power to act unjustly is desirable because it allows one to gratify all one’s appetites, to enjoy pleasures of every kind without undergoing any painful punishment. Justice is undesirable because it requires one to forgo such pleasures or else accept the pains of punishment. We have reason to suspect, then, that Polus says that injustice is better than justice because he finds it more pleasant. He confuses pleasure with benefit and goodness, and advocates injustice on simple hedonistic grounds. When he envies Archelaus’ life, he is thinking of the pleasures Archelaus enjoys; when he says that it is better to commit injustice than to su·er it, he is thinking of the pains of being wronged and the pleasures one can secure by doing wrong; when he says that it is better to escape the penalty than to pay it, he is thinking of the pains of punishment. Socrates makes this last point nearly explicit when he says that the orators, tyrants, and potentates whom Polus envies avoid discipline because they 16 What he sees fit, what seems best to him: δοκε! α"τ#$ βλτιστα ε&ναι (467 b 3–4 and throughout). Polus argues that the tyrant Archelaus, a paradigm of injustice, is enviable because he has the power to kill, banish, or rob whomever he wants, and more generally to do exactly as he likes (π'ντα πρ'ττειν κατ τν α"το( δ)ξαν, 469 c 7). 17 He claims that injustice is bad only if one is caught and punished (470 a), and he reaches his peak of eloquence in describing the physical and psychological horrors of punishment in graphic detail (473 b 12–c 5).

Created on 4 July 2005 at 14.33 hours page 144

144

Jessica Moss

‘see its painfulness but are blind to its benefit’ (479 b 6–7). It is the lure of pleasure and fear of pain that lead these people astray, and (Socrates implies) it is considerations of pain and pleasure that make Polus want to follow them. The strongest argument, however, for thinking that Socrates interprets Polus in this way is that doing so explains his otherwise mysterious analysis of the kalon as what is either good and beneficial or pleasant. Elsewhere in the Gorgias, and in other dialogues, Socrates consistently associates the kalon with the good and beneficial, but frequently opposes all three to the pleasant. What is pleasant but not beneficial is downright shameful, aischron.18 It should thus puzzle us to see him include pleasure here: he can only be referring to aesthetic pleasure,19 and distinguishing an aesthetic sense of kalon (beautiful) from a more ethical or utilitarian sense; but it is hard to see how aesthetic pleasure in the beautiful is relevant to Polus’ preference for injustice over justice.20 Including pleasure in the analysis, however, gives Socrates a pretext to question Polus as to whether justice, being more kalon than injustice, is also more pleasant, and Polus’ answers are very revealing. Socrates begins by asking Polus if committing injustice is more painful than su·ering it; the answer is an adamant no (ο"δαµ$ς, 475 c 3). The same question is put, and the same answer given, about avoiding a penalty when one has done wrong: it is more shameful but more pleasant (less painful) than paying the penalty 18 In the Gorgias this is most clear in Socrates’ contrast between di·erent kinds of rhetoric at 503 a: while flattering rhetoric, concerned only with pleasure, is shameful, ‘true’ rhetoric would be kalon because it would benefit people, and would say ‘the best things, whether they are more pleasant or more unpleasant to the hearers’ (503 a 7–9). What is good and beneficial is kalon even when it is not pleasant; what is pleasant but not good and beneficial—like flattery—is shameful. Indeed, whatever is bad is shameful: ‘I call bad things shameful [τ γρ κακ α-σχρ καλ$]’ (463 d 4). This is consistent with Socrates’ association of the good or beneficial with the kalon throughout the dialogues, which is perhaps most evident in connection with his frequent assumption that virtue is kalon. (La. 192 c–d is typical: courage, being a virtue, must be kalon; furthermore, what is kalon must be good and beneficial.) Meno 77 b ·. and Smp. 204 e ·. treat the good and the kalon as interchangeable; at Rep. 452 e 1–2 Socrates says that it is ridiculous and foolish ‘to take seriously any standard of the kalon other than the good’; at 508 e he claims that the Form of the Good is very kalon, more so even than knowledge or truth, and at 517 c he claims that this Form is the cause of all that is kalon in the world. 19 Compare H.Ma. 299 a: visual and auditory pleasures, but never sensual pleasures like eating or having sex, are called kala. 20 Pace Vlastos, ‘Polus’. In fact, as I mention below, we can see Polus’ reactions to injustice as analogous to aesthetic reactions, but only analogous.

Created on 4 July 2005 at 14.33 hours page 145

Shame, Pleasure, and the Divided Soul

145

(477 a 1–4). Polus then agrees that injustice itself is more shameful, but not more painful, than sickness or poverty (477 c–e). Polus initially preferred committing injustice to su·ering it, avoiding the penalty to paying it, and living unjustly to living in other ways. Now Socrates’ questions show that, according to Polus’ own, unhesitating judgements, each of these things is more pleasant than its alternative. It is an easy step from here to the conclusion that Polus preferred injustice just because it is more pleasant than justice—that he conflated pleasure with goodness, without being quite aware that he did so.21 Socrates thus thinks that he has identified Polus’ reasons for superficially valuing injustice. But how can he challenge and unseat them? One might of course be a rational, theory-based hedonist: one might have reasons and arguments for the view that pleasure is good. If Polus’ hedonistic preferences were of this form, rational arguments would be the proper tool for refuting them. But, I have argued, Polus’ preference for injustice is paradigmatically non-rational: he does not reason that injustice is better, but seems to see that it is so; his belief that injustice is good is not a rational belief but something more akin to a perception or a·ective reaction. Moreover, he is not even clearly aware that pleasantness is the property of injustice to which he is responding in seeing it as good: Socrates has to reconstruct and make explicit that this is so. A belief not based in reasoning is not easily challenged or corrected by means of reasoning: witness Polus’ incredulous responses to Socrates. Polus can see that injustice is good; why should he listen to arguments that challenge what is so clear? Socrates’ best hope is not to give him arguments that injustice is bad, but somehow to make him see it as bad—to counter one impression with another.22 If Polus’ preference is a response to one feature of injustice—its pleasantness—Socrates can try to unseat this preference by draw21 Polus’ hedonism must have been unwitting, for he happily consents to an analysis of kalon that distinguishes pleasure from benefit, and himself rephrases this as a distinction between pleasure and goodness (475 a 3). (If Polus is indeed thinking of aesthetic pleasure at this stage, this makes it easy for him to accept the contrast between pleasure and benefit. Had Socrates asked Polus to agree to a sharp distinction of benefit from pleasure of any kind, he might well have baulked.) 22 Compare Polus’ own rhetorical strategy in trying to persuade Socrates that punishment is bad by detailing the horrors of punishment: Socrates accuses Polus of trying to ‘spook’ him (µορµολ/ττ0η, 473 d 3), rather than give him arguments. The strategy fails to work on Socrates, whose own bases for preferring justice to injustice are deeply rational.

Created on 4 July 2005 at 14.33 hours page 146

146

Jessica Moss

ing his attention to another feature of injustice that will produce a countervailing response. It is in precisely this spirit, I maintain, that Socrates appeals to Polus’ sense of shame. Socrates asks which is more aischron, committing injustice or suffering it; committing injustice, Polus admits unhesitatingly. Why does he do so? Consider the following explanation: he simply feels or sees that it is so. His judgement is immediate and forceful, an emotional reaction based on his perception of a moral fact. Here the moral sense of aischron is closely analogous to the aesthetic sense, ugly: we might say that Polus reacts to the moral ugliness of wrongdoing with a feeling of shame, just as he might react to the ugliness of a human face with a feeling of distaste. This, I suggest, is Socrates’ view of the matter. By drawing Polus’ attention to considerations about what is shameful and what kalon, he believes that he has made him see justice and injustice in a new and unfavourable light. Plato is thus emphasizing the connection between the judgement that a thing is shameful (aischron) and the a·ective experience of shame (α-σχ/νη), and suggesting that the latter motivates the former.23 On the account I have o·ered, Polus initially thought injustice good because he was attracted by its pleasures. Now, Socrates hopes, he will realize that injustice is bad because he is repelled by its shamefulness. Responses to the pleasant and the painful on the one hand, and the kalon and the aischron on the other, are two bases for non-rational, quasi-perceptual value judgements. What one takes pleasure in strikes one as good; what one feels shame at strikes one as bad. In proceeding as he does with Polus, Socrates is emphasizing that these two types of response tend to conflict: many kalon things are painful, while many pleasant things are aischron, and thus we are pulled in di·erent directions. Moreover, Socrates is making the point that pleasure pulls us in the wrong direction, towards false value judgements, while shame pulls us in the right direction, towards the truth. Why does shame pull us towards the truth? Socrates’ explanation is that shame judgements reflect our deep-down moral beliefs, which are on his view true. Prejudices and confusions may hide 23 Compare my discussion of Callicles’ reaction to the catamite (494 e) in sect. 2. It is worth noting that even on Callicles’ cynical account of Polus’ refutation (see sect. 4 below), Polus’ judgement that committing injustice is aischron is motivated by a feeling of shame.

Created on 4 July 2005 at 14.33 hours page 147

Shame, Pleasure, and the Divided Soul

147

our deepest value beliefs from our own view, but in our feelings of shame they shine through. Deep down Polus truthfully believed all along that injustice was worse than justice (474 b 2–5, 475 e 3–5); finally his feelings of shame bring that belief to light. That this is Socrates’ understanding comes out in his analysis of judgements that a thing is shameful or kalon. He argues that we do not judge a thing shameful ‘looking towards nothing’, but rather in virtue of some badness (or harm), or of some pain; it becomes clear as he goes along that in the relevant cases it is always on the basis of badness.24 Socrates is arguing that Polus’ recognition that injustice is more shameful than justice must be based on a prior recognition that injustice is worse. He cannot be construing this prior recognition as an explicit thought, for Polus denies that he thinks injustice bad, and Socrates accuses him not of lying but rather of being unaware of his own deepest beliefs. Certainly there are cases of moral judgement in which we can analyse our immediate feeling of attraction or repulsion, admiration or disgust, as having rested on a ‘deep-down’ belief of which we were unaware. Someone thinks that she believes that all sex is fine so long as all parties consent; she contemplates a case of consensual sister–brother incest and reacts with outrage and disgust; she thereby realizes that deep down she did not believe that all consensual sex is fine but instead believed that incest, consensual or not, is wrong.25 If shame works to reveal our deep value beliefs in this way, then it clearly has potential as a tool of moral persuasion. In particular, appeals to shame can challenge false beliefs based on the appeal of pleasure more e·ectively than can rational argument. But as we read on, it becomes clear that Socrates thinks shame’s power extends beyond merely conflicting with pleasure: when shame and 24 These relevant cases are, roughly speaking, cases in which we call something aischron in the moral sense. We call things aischron ‘looking towards pain’ when we use aischron in the aesthetic sense, ‘ugly’. 25 Of course, there are many other ways to interpret the clash between an initial judgement and a gut reaction. In the paper from which I take the incest example, Jonathan Haidt argues that the non-rational, quasi-perceptual intuition that incest is wrong is the basis for the moral judgement (J. Haidt, ‘The Emotional Dog and its Rational Tail: A Social Intuitionist Approach to Moral Judgment’, Psychological Review, 108 (2001), 814–34). Callicles o·ers another type of explanation: the judgement that something is shameful sometimes amounts to no more than the recognition that society regards it as shameful. There is no shortage of ways in which one’s a·ective responses can fail to line up with one’s dispassionate judgements, but we can give Socrates the credit for having hit on one such way, and finding in it a positive use for feelings of shame.

Created on 4 July 2005 at 14.33 hours page 148

148

Jessica Moss

pleasure conflict, shame will win. This is the view that underlies the refutation: Socrates wants to show Polus not merely that in finding injustice shameful he must recognize it as prima facie bad, but that he must reject it out of hand. But surely Polus might react by thinking, ‘In so far as injustice is shameful I do think it bad, but my attraction to it qua pleasant outweighs that consideration, and overall I still prefer it.’ In fact, this seems to be precisely Polus’ condition at the end of the conversation (although he does not articulate it clearly): he leaves the conversation obviously unpersuaded. Does Socrates none the less have some reason for thinking that shame should tend to trump pleasure, or is this mere wishful thinking? Socrates’ confidence in the authority of shame may be stronger than warranted, but the structure of the argument indicates that he does have a particular reason for thinking that shame will trump pleasure. When Polus admits that injustice is more shameful than justice, he does not yet concede that it is worse; he is brought to this point only after Socrates first introduces the disjunctive analysis of the kalon and then uses it (as I have argued) as a pretext for questioning Polus about the relative pleasures and pains of justice and injustice. In answering these questions, Polus is faced with the fact that he takes pleasure in and thus pursues something aischron, and that he shirks the kalon through fear of pain. And this is a realization that would fill a man like him with shame, for two reasons. First, as many have recognized, shame is a selfregarding emotion: one feels shame about one’s own relation to things that are aischra, shame at doing or wanting them. Second, for men of Polus’ culture, taking pleasure in something aischron is particularly shameful—more than doing something aischron for other reasons; this is a fact that Plato will bring out in Socrates’ shaming of Callicles (see below). And these two facts give us reason to think that when shame and pleasure conflict, shame may win out. Ordinarily the fact that a particular pleasure is shameful does not enhance the pleasure one takes in it; for most men of Polus’ culture, however, it does enhance the shame. Socrates’ refutation of Polus thus rests on a psychological theory about the role of shame and pleasure in our judgements about what is good. Feelings of shame and desires for pleasure both act as non-rational, a·ective, emotional, quasi-perceptual bases for judgements of value. They tend to conflict. Shame has a signifi-

Created on 4 July 2005 at 14.33 hours page 149

Shame, Pleasure, and the Divided Soul

149

cant moral advantage over pleasure: it taps into our deepest moral intuitions, which are, on this theory, true. Shame also has a strategic advantage over pleasure: when a person realizes that his feelings of shame conflict with his desires for pleasure, the conflict tends to enhance his feelings of shame and make them more likely to dominate. Shame, then, can in principle divorce a person’s judgements about pleasure from his judgements about the good, and thereby show him that virtue is better than vice. In Polus’ case, the strategy fails. He leaves the conversation unpersuaded that justice is better than injustice, saying ‘I think these things are absurd’ when Socrates elaborates the consequences of that view (480 e 1). His feelings of shame do conflict with his unwitting hedonism, but they fail in the end to undermine it. In the next episode of the dialogue, however, we see Socrates use an appeal to shame with more success, and we learn more about the relation between pleasure, reason, and shame.

2. Callicles Callicles begins by defending a strong form of hedonism: the best life is the life devoted to strengthening and gratifying one’s appetites without discrimination or restraint (491 e–492 a). Socrates advances several attacks against him: the myth and fable of the jars, meant to show that the temperate person is happier than the intemperate (493 a–494 a); the examples of the itch-scratcher and of the catamite26 (494 c–e); the argument that pleasure and pain are experienced simultaneously, unlike good and bad (495 e–497 d); and finally the argument that the coward and the foolish person experience as much pleasure as or more pleasure than the brave or intelligent person (497 e–499 b). Callicles ends by withdrawing his claim, and conceding that not all pleasures are good. Is Callicles refuted by the cumulative weight of Socrates’ arguments, or is one of the arguments decisive? Plato does not give us an explicit answer to this question, but he does take care to draw attention to the importance of shame in the conversation. Socrates 26 The κναιδος: ‘roughly speaking, one who enjoys being the passive partner in anal intercourse. Attic law apparently treats this as equivalent to male prostitution, and su¶cient to deprive the guilty party of his citizenship rights’ (Kahn, Dialogue, 136).

Created on 4 July 2005 at 14.33 hours page 150

150

Jessica Moss

flatters Callicles by apparently conceding that shame led Gorgias and Polus astray, and adding that because Callicles himself has the important quality of frankness (παρρησα, 487 a 3) he will not say anything he does not believe through ‘a surplus of shame’ (487 e 4–5). Callicles is flattered: he clearly thinks that succumbing to shame is a sign of weakness and social conformity. Socrates returns to the point several times, in e·ect taunting Callicles by reminding him that his pride is staked on his immunity to shame: ‘take care not to be ashamed [2πως µ 4παισχυν0]’ (494 c 5); ‘I shocked Polus and Gorgias and made them ashamed, but you won’t be shocked or ashamed, for you are andreios [manly or brave]’ (494 d 2–4). Now, having aroused Callicles’ pride in his manly freedom from shame, Socrates sets out to show him that he is subject to shame. Callicles has claimed that the good and happy life is the life devoted to gratifying one’s appetites, whatever they may be. Socrates shows him first an embarrassing consequence of this view—that a man who spends his life scratching an itch will be happy—and then, following from this, a shameful one: ‘Isn’t the climax of such things, the life of the catamite, frightful and shameful and miserable?’ (494 e 3–5).27 Callicles responds violently: ‘Aren’t you ashamed to lead the discussion to such things, Socrates?’ (494 e 7–8). Why does he think that Socrates should be ashamed to refer to the pleasures of the catamite? Clearly because, as Socrates intends him to, he finds such things shameful: the thought of taking pleasure in something so unmanly makes him recoil in disgust, so much so that he thinks that even mentioning such pleasures should fill a man with shame. Having been struck with the shamefulness of certain pleasures, Callicles now ceases to defend hedonism with his former conviction. When Socrates dares him to maintain ‘still even now’ that pleasure and goodness are the same (495 a 2), he rises to the challenge, but says that he does so merely ‘in order that my argument will not be inconsistent if I say that they are di·erent’ (495 a 5–6). Hedonism no longer seems true to him; he defends it merely for the sake of argument. Callicles sticks to his line through Socrates’ next argument (that pleasure and pain, unlike good and bad, can exist simultaneously). The last argument, however—that cowards enjoy as much or more pleasure than brave men (andreioi)—is his undoing. Socrates has 27 Cf. 495 b 4–5: ‘both the many shameful things hinted at just now clearly follow, if [the good is pleasure] . . . and also many others’.

Created on 4 July 2005 at 14.33 hours page 151

Shame, Pleasure, and the Divided Soul

151

earlier flattered Callicles by calling him andreios (494 d 4); now he argues that Callicles’ hedonism would lead him to embrace the life of the coward instead of the manly person he prides himself on being. It is now that Callicles reneges on his earlier position, claiming to have recognized a di·erence between pleasure and goodness all along, and to have maintained the identity only ‘in jest [παζων]’ (499 b 5). What has happened to Callicles? How was he persuaded to renounce the view he had defended so vigorously? Socrates’ emphasis on shame in the discussion that leads up to the refutation, the explicit references to shamefulness in the case of the catamite and ‘many other’ consequences of hedonism (495 b 5), and the appeal to Callicles’ pride in the example of the coward all point to the same answer. Callicles initially held that all pleasures were good, and (thus) that the life of unlimited gratification of the appetites was the best life, because he did not notice that some pleasures are bad and some appetites better left ungratified. Socrates showed him that some pleasures, and some aspects of the life of unlimited gratification of the appetites, are shameful. In order to make this manifest to Callicles, he had first to make him sensitive to considerations of shame and admiration; hence his emphasis on Callicles’ παρρησα, his manly freedom from the social conventions that shackle the weak. Socrates was arousing Callicles’ feelings of pride, shame, and admiration.28 This done, he chose pleasures that would strike Callicles as manifestly shameful: those of the catamite and the coward, types whom Callicles in particular, with his pride in manliness, would find repugnant. Once Callicles is struck with the shamefulness of these pleasures, he simply sees that they are bad: for him, as for Polus, to recognize that something is shameful is to be repelled by it. Now he recognizes that all along, deep down, he thought certain pleasures bad (‘As though you really think that I or any other man doesn’t believe some pleasures better and others worse’, 499 b 6–8).29 28 This is not the only point of the emphasis on Callicles’ freedom from shame. The contrast between Callicles’ vaunted frankness and his eventual surrender to shame serves to refute him on the level of attitude as well as on the level of explicit statement (see Socrates’ claim that Callicles is ‘out of harmony’ with himself (482 b 5–6). As I discuss below, it also serves to undermine Callicles’ own theory of shame. 29 Note the similarity to Socrates’ claim against Polus: Polus and all other people really believe that committing injustice is worse than su·ering it (474 b 2–5, 475 e 3–5). Shame is able to uncover beliefs that are common to all. Cf. Kahn, Dialogue,

Created on 4 July 2005 at 14.33 hours page 152

152

Jessica Moss

The strategy is more e·ective on Callicles than on Polus: Socrates really does bring him to renounce what he thought he believed good, and succeeds in using shame to reveal what Callicles believed deep down all along.30 The strategy is more e·ective because Callicles is so driven by shame and related motivations—ambition, desire for honour, competitiveness, pride in manliness (all of which the Republic will assign to spirit, thumos). Because he cares so much about being andreios, he cannot countenance the thought of taking pleasure in the aischron. Where Polus would say ‘Injustice is shameful, but it is pleasant so I want it anyway’, Callicles thinks ‘The life of unlimited gratification of the appetites is shameful, so its pleasures are disgusting; therefore despite its pleasures I no longer want it.’ In the figures of Polus and Callicles, the Gorgias emphasizes the power of pleasure to lead people astray. (This is in fact emphasized throughout the dialogue, particularly in the characterization of rhetoric as appetite-gratifying flattery with enormous influence over ordinary people.) It also suggests that rational arguments, based on considerations of harm and benefit, will often fail to persuade someone that a bad pleasure is in fact bad, or a good pain in fact good: Socrates’ arguments for the superiority of justice over injustice leave Polus incredulous and Callicles indi·erent. The dialogue shows, however, that shame can sometimes neutralize the appetites’ destructive force where reason on its own has failed: it can make the agent recoil from the pleasures of vice and aspire even to the pains of virtue. In the next sections I argue that the Republic develops this insight about the opposition between pleasure and shame into a full characterization of virtue and a full programme of moral education. It modifies the idea that shame taps into our ‘deep-down’, true value beliefs, while developing and making use of other ideas from the Gorgias: that shame can be a more e·ective tool of persuasion than reason, that shame is more e·ective against pleasure in a person who has a strong sense of honour, ambition, and pride in manliness, and that shame is connected to how an agent sees himself. It does all this by assigning feelings of shame 138: although Socrates’ appeals to shame are ‘highly personalized’, the conclusions they result in are universally true. 30 Notice that the appeal to shame is su¶cient to persuade Callicles to reject the strong form of hedonism he first defended, but not su¶cient for Socrates’ ultimate goal: to persuade him that the temperate, just, philosophical life is best. One clear reason Callicles has for resisting this conclusion is that he finds the philosopher’s life shameful: see sect. 4 for discussion of this point.

Created on 4 July 2005 at 14.33 hours page 153

Shame, Pleasure, and the Divided Soul

153

and admiration to a part of the soul distinct from both appetite and reason, which has a crucial role in mediating between the two—in helping reason to overcome appetite and keep it under control.

3. Shame and pleasure in the Republic Plato introduces spirit (thumos) in Republic 4, with the example of Leontius, who wants to look at corpses by the roadside but is angry with himself for having this desire. Plato presents this as a conflict between appetite (τ6 7πιθυµητικ)ν) on the one hand, and spirit on the other. Why does spirit object to appetite’s desire? Clearly because it feels disgust: looking at corpses for pleasure is something shameful. (As Leontius gives in to his appetites and rushes towards the corpses, he tells his eyes ‘take your fill of the kalon sight’ (440 a 3). This is straight irony: the sight is not kalon, but the very opposite—aischron, ugly and shameful.) Leontius’ psychological conflict is thus of the kind that Socrates brings about in Polus and Callicles in the Gorgias: he is drawn towards something because it is pleasant, but shame motivates him to reject it. In Leontius’ case shame’s promptings are too weak, and spirit is overpowered by appetite. In a virtuous, well-ordered soul, however, spirit is stronger than appetite, and its task is to aid reason in ‘governing’ appetite and keeping it under control.31 Spirit is well suited to the task, because even in a soul that is not virtuous, it tends to side with reason against appetite: Whenever the appetites force someone against rational calculation, he rails at himself and is angry at the forcing thing within him, and just as when two factions war, the spirit of such a man becomes an ally to his reason. (440 a 8–b 4)

Why does spirit tend to side with reason, when reason and appetite conflict? To answer this question, let us examine courage, the virtue distinctive of spirit. ‘We call an individual brave in virtue of his spirited part, when through both pains and pleasures it preserves reason’s commands about what is to be feared or not’ (442 b 11–c 4). Reason decides 31 When properly trained, spirit and reason together ‘will govern the appetitive part . . . they will watch it lest, growing large and strong by filling itself with the body’s so-called pleasures, it no longer sticks to its own a·airs [τ ατο( πρ'ττ0η], but tries to enslave and rule’ the other parts of the soul (442 a 5–b 2).

Created on 4 July 2005 at 14.33 hours page 154

154

Jessica Moss

what is to be feared; this is part of its general task of determining what is good and what is bad.32 But, Plato says, strong pains and pleasures can ‘wash out’ reason’s dictates, as soap can wash out dye.33 A person who rightly believes that bodily harm is not to be feared, because it is not truly bad, may waver in this belief when faced with imminent powerful pain. Reason, recognizing the courageous action as best, pulls the agent towards that action. Which part of the soul pulls the agent away, dreading the pains of the courageous action? It must be appetite, which Plato hereby presents as the part of the soul motivated by considerations of pleasure and pain.34 The outcome of the battle is decided when the third party, spirit, joins its forces with reason in pulling the agent towards the courageous action.35 Why does it do this? It is not motivated by considerations of pain and pleasure, like appetite, nor of benefit and harm, like reason. The most compelling explanation is that it finds this action kalon, and finds the cowardly, pleasurable alternative shameful. The same principle holds for spirit’s role in another virtue: temperance (s»ophrosun»e ). If courage is the disposition to pursue what is good even when it is painful, temperance is the disposition to avoid what is bad even when it is pleasant—even when the appetites might desire it.36 Plato says that temperance involves all three parts of the 32 Reason is the part that calculates about what is better and what worse (441 c 1–2). 33 The guardians’ ‘belief about what they should fear and all the rest’ must become ‘so fast that even such extremely e·ective detergents as pleasure, pain, fear, and desire [epithumia—appetite] wouldn’t wash it out—and pleasure is much more potent than any powder, washing soda, or soap’ (430 a 3–b 2, trans. G. M. A. Grube/C. D. C. Reeve). 34 We learn in Republic 9 that each part of the soul experiences its own pleasures, but as we see here and throughout the dialogue, appetite alone is motivated by the pursuit of pleasure and avoidance of pain; it is the part that pursues pleasures qua pleasures (rather than qua beneficial or honourable). Plato even introduces the appetitive part of the soul in book 4 as the ‘companion of certain indulgences and pleasures’ (439 d 8), while pleasure plays no role in the definition of the other parts. We can thus take the Republic to be fundamentally in agreement with the Charmides in regarding appetite (epithumia) as the desire for pleasure (Chrm. 167 e 1–2; cf. Arist. NE 1111b17). 35 Republic 4’s description of psychological conflict in terms of dragging (9λκειν), leading (γειν), and hindering (κωλ/ειν) (439 c 6–d 1) encourages us to understand the situation in these dynamic terms. 36 Compare Aristotle: ‘On account of pleasure we do base things [τ φα(λα], while on account of pain we abstain from kala things’ (NE 1104b9–11); temperance and courage work to counteract these tendencies. In Plato as in Aristotle temperance and courage are very closely connected: the Republic’s discussion of courage includes

Created on 4 July 2005 at 14.33 hours page 155

Shame, Pleasure, and the Divided Soul

155

soul, and although he does not spell out spirit’s role in this virtue, we can easily infer it by analogy with courage. A person who rightly believes that over-indulgence is not to be pursued, because it is not truly good, may lose this belief when faced with imminent powerful pleasures. Just as in the case of courage, here too spirit can aid reason by ‘preserving its commands’, this time against the temptation of pleasures. Reason, calculating that it is best to abstain from some pleasant action, pulls the agent away from the action; appetite drags the agent towards it; spirit joins forces with reason against appetite because it finds the action shameful, and abstinence from it kalon. Spirit feels shame and anger at the thought of indulging base appetites, as well as at the thought of shirking painful duties; when these feelings are strong enough, they override the appetitive part’s desire for pleasures and its fear of pain, and the person behaves with temperance and courage.37 Of course, Plato never explicitly tells us in the Republic that spirit is motivated to pursue the kalon and avoid the shameful. He introduces spirit as the seat of anger in book 4, and tells us in books 8 and 9 that it strives for honour and victory. But the desire for what is kalon can explain the desire for honour and victory, these being paradigmatically kalon, paradigmatic objects of admiration and praise. Anger can be understood as a natural reaction to the aischron in others or to the judgement that something aischron has been done to oneself; the natural reaction to the judgement that one has oneself done something aischron is of course shame. A detailed reference to appetite and pleasure, as well as fear and pain; in the Laches, Socrates describes the courageous as ‘not only those who are courageous against pains or fears, but also those who are clever at fighting against appetites and pleasures’ (La. 191 d 6–e 1); in the Laws, the Athenian defines courage as ‘a fight against fears and pains’ but also ‘desires [π)θοι] and pleasures’ (Laws 633 c 9–d 1, trans. Saunders in Plato: Complete Works [Plato], ed. J. M. Cooper (Indianapolis, 1997)). This broad definition shows that ‘courage’ is in fact too narrow a translation of the Greek andreia; ‘manliness’ may better capture what Plato has in mind. 37 Note the association of temperance with α-δ:ς (shame, reverence, or modesty) at Phdr. 253 d 6 and Chrm. 160 e 4–5. Perhaps the object of appetite need not in itself be aischron: Hendrik Lorenz points out that in some cases reason may oppose an appetite that ‘as such is indi·erent to spirit’ (as when considerations of health motivate the thirsty person not to drink), but adds that in such cases spirit may still ally with reason ‘because it would be dishonourable or disgraceful for reason to be defeated by desire (440 a 8–b 4)’. The concern with disgrace shows that ultimately here too what motivates spirit are considerations about what is kalon and aischron. (H. Lorenz, ‘Desire and Reason in Plato’s Republic’, Oxford Studies in Ancient Philosophy, 27 (2004), 83–116 at 100).

Created on 4 July 2005 at 14.33 hours page 156

156

Jessica Moss

defence of these claims lies outside the scope of this paper, but I hope to lend the idea some intuitive appeal by showing the parallels between spirit in the Republic and judgements about the shameful and kalon in the Gorgias.38 Moreover, this interpretation of spirit’s role and motivations is bolstered by a comparison with the allegorical psychology of the Phaedrus. Here again we have a soul composed of three parts: a charioteer, corresponding to reason as described in the Republic, and his two horses. One horse is base and unruly, always dragging the soul towards low pleasures; it corresponds to the 7πιθυµητικ)ν, the appetitive part of the soul. The other is kalos, fine and noble (246 b 2, cf. 253 d 4), and ‘a lover of honour’ (253 d 6), clearly corresponding to thumos. The Phaedrus describes a case of psychic conflict between reason and appetite: the base horse desires the pleasures of sex, but the charioteer wishes to refrain. When this happens, the noble horse sides with the charioteer against the base horse, and it does so because it is motivated by shame: ‘controlled, then as always, by its sense of shame [α-δο!] . . . [it] prevents itself from jumping on the boy’; in the thick of the struggle it ‘drenches the whole soul with sweat out of shame [α-σχ/νης] and awe’.39 Returning now to the Republic, it is important to note that spirit’s contribution to the agent’s thought is not merely ‘this pleasant action is shameful’, but rather ‘for me to do the pleasant action would be shameful’ (or ‘for me to shirk this kalon but painful action would be shameful’). Spirit is the part of the soul in virtue of which, as Plato says, a person ‘rails at himself and is angry with the forcing thing within him [appetite]’ (440 b 1–2): its anger and shame are directed not simply towards an action or object but rather towards a person (oneself or another) qua agent.40 This is the basis for 38 Gabriel Richardson Lear makes a compelling case for an interpretation of Plato on which spirit desires the kalon, partly by way of connection with Aristotle’s characterization of spirited desire, in G. Richardson Lear, Happy Lives and the Highest Good [Happy Lives] (Princeton, 2004), 139–44. 39 Phdr. 254 a 2–3 and 254 c 4–5, trans. A. Nehamas and P. Woodru· (in Plato, ed. Cooper). Cf. 256 a 6: when the boy is in bed with the lover, his base horse wants to capitulate to the lover’s demands, but his noble horse holds back out of α-δ:ς. 40 Compare Cooper: spirit ‘is di·erent from appetite because appetites lack the self-reference which is essential to esteem and self-esteem’ (‘Motivation’, 135). Hobbs describes spirit as having second-order attitudes (Hero, 58–9). I would say that spirit’s attitudes are in part second-order, in part first-order: a person may find an act shameful or kalon in itself, and for that reason feel shame or pride in performing the action. The broad scope of the Greek terms kalon and aischron is important here: while it is trivial to say that performing a shameful action is shameful, it is

Created on 4 July 2005 at 14.33 hours page 157

Shame, Pleasure, and the Divided Soul

157

the reaction Socrates tried to bring about in Polus and Callicles: because one feels shame at one’s own relation to shameful things, awareness of the fact that one desires or takes pleasure in something shameful enhances one’s shame. I argued that in the Gorgias Socrates uses appeals to shame to make his interlocutors recognize the distinction between the pleasant and the good. Polus—and by implication the conventional morality he represents—unconsciously equates pleasure with the good; Callicles does so explicitly. Shame makes them each see certain pleasures as bad, and thereby undermines their conscious or unconscious hedonism. Now we see that shame plays the same role in courage and temperance as the Republic describes them: it helps the agent keep firm the distinction between goodness and pleasure. Without spirit’s aid, reason would be in danger of losing the battle to appetite. The pleasure that appetite craves would become the standard of value that determines the agent’s actions. The person who believes that injury in battle is not to be feared, because it is not truly bad, but then loses that belief through fear when the moment comes, has confused pleasure with benefit, pain with harm. Because injury is painful, he has come to believe that it is bad. The e·ect of spirit, when it holds firm against fear or temptation, is to prevent the agent from lapsing into this confusion. In virtue of his spirited part, the agent continues to distinguish what is pleasant from what is good, ‘preserving the commands of reason’ and keeping appetite under control. The question remains, of course, why reason needs spirit’s help in this way. We have seen the beginnings of an answer in the Polus episode of the Gorgias. If someone is going for something qua pleasant not because he reasons that it is good, but rather because the thing simply strikes him as good (as I have argued that Polus goes for injustice), he will have no patience for reasoned arguments that it is in fact bad. He can see that it is good; why should he listen to arguments to the contrary?41 But finding something shameful is perhaps less so to say that performing an aischron action is aischron. Here is an attempt to disambiguate: if an action strikes one as kalon or aischron in what we might call a quasi-aesthetic sense (something close to beautiful or ugly, but with moral overtones), performing it will be kalon or aischron in the ethical sense (admirable or shameful). 41 Compare the children who shout down the doctor, in the trial allegory at Gorg. 522 a. He argues that his treatments are beneficial, but the children, finding the

Created on 4 July 2005 at 14.33 hours page 158

158

Jessica Moss

not like considering a rational argument that it is bad; it is instead seeing the thing as bad, being struck by its badness. Now the person must question, and perhaps reject, his conviction that the thing was good. The Republic adds to this picture by assigning di·erent desires, beliefs, and motivations to di·erent parts of the soul. When we form value beliefs and desires on the basis of sophisticated reasoning, we are using the rational part of the soul; but this part of the soul develops only after childhood, and in many people (most, Plato implies) it remains weak or never really develops at all (441 a 9–b 1). For most people, the non-rational parts of the soul rule, and non-rational, a·ective, quasi-perceptual responses hold sway; in such people, reason simply is not developed enough or strong enough to combat appetite. Fortunately, spirit is present from the start (and, as we shall see in Section 4, can be strengthened and shaped by education well before reason develops). For people in whom reason is not ripe for arguments and rational persuasion— people who go on appearances, on how things strike them and feel to them—shame and other thumoeidic emotions are vital bulwarks against the destructive lure of appetitive pleasure. Thanks to shame, that is, even those people stuck at the non-rational level of quasiperceptual or emotional reactions to the world can be guided by a moral sense. We have seen that the Republic and Phaedrus expand and develop the role of shame as we saw it in the Gorgias. The power of shame and the other thumoeidic emotions helps to keep appetite from ruling in the soul—from usurping reason’s place and dictating what is good and bad, what the agent should pursue or avoid. Spirit systematically protects a person from the lure of pleasures, helps him to withstand the fear of pain, and preserves a system of values in which benefit is clearly distinguished from gratification of the appetites. It is worth noting that this interpretation reveals something quite strange about book 4’s account of the virtues. An agent whose appetites constantly threaten to lead her astray may find ways to control them, and may thus always perform the virtuous action, but she will possess only what Aristotle calls continence, in contrast to genuine virtue. In a truly virtuous soul the parts are harmoniously treatments painful, can simply see that they are bad; why should they listen to arguments here at all?

Created on 4 July 2005 at 14.33 hours page 159

Shame, Pleasure, and the Divided Soul

159

ordered, not pulling against one another like warring factions. The appetites desire only what reason countenances (see 586 d), and spirit has no need to fight against them.42 Perhaps the idea is that with the help of spirit, reason can eventually subdue the appetites so thoroughly that they become gentle and compliant. Perhaps, on the other hand, the hints of mere continence indicate that book 4 o·ers an account not of true virtue, but of a second-best state; perhaps true virtue is available only to genuine philosophers with knowledge of the Form of the Good.43

4. Shame, pleasure, and moral education Appetite and spirit are both non-rational forces in the soul. Neither exercises sophisticated reasoning, and neither is motivated by concern for the overall good of the soul and its parts. If either grows too strong and takes control of the soul, the agent will be unjust and unhappy. Why, then, is spirit privileged over appetite as the ally of reason? Why do feelings like shame tend not only to conflict with desires for pleasure, but also when they do so to lead us in the direction that reason would recommend? In the Gorgias, we saw that Socrates held an implicit theory of shame that would provide a straightforward explanation for its moral power. First, our feelings of shame are authoritative. Finding something shameful is not merely one factor among many in one’s judgement about the thing: one simply cannot find something shameful without thereby recognizing that it is bad, and ceasing to desire it. Second, everyone is disposed to recognize as shameful what is truly bad, and as kalon what is truly good. Feelings of shame are therefore reliable guides to the moral truth (unlike experiences of pleasure, which are so often deceptive). This strong view of shame’s power explains Socrates’ willingness to rely on his interlocutors’ shame judgements as revealing their sincere beliefs, and

42 Note the claim, at Soph. 228 b 2–4, that in disordered souls we find conflict between, among other things, thumos and pleasures. 43 Plato says that he has given an account only of civic (πολιτικ) courage at 430 c 3–5; while this may simply refer to the fact that he has thus far described only the courage of the city, many think that Plato is indicating that all the virtues defined in book 4 are in some sense inferior. Plotinus calls these virtues ‘civic’, and argues that they belong only to those whose appetites stand in need of control (Plot. 1. 2. 1–2).

Created on 4 July 2005 at 14.33 hours page 160

160

Jessica Moss

also his confidence that they will find injustice and intemperance shameful when asked. But Socrates’ view is only half the story about shame in the Gorgias: much of the dialogue’s discussion tends in a very di·erent direction. Polus charges that Gorgias contradicts himself because he is ashamed to say what he truly believes: that he does not and cannot teach his students to be just. Callicles charges that Polus in turn contradicts himself because he is ashamed to say what he truly believes: that committing injustice is more kalon than su·ering it. Far from revealing deep beliefs, they claim, shame often disguises them.44 The dramatic action of the dialogue also tends to undermine Socrates’ view of shame. The characters do not always feel shame at what they should, nor when they do is this always su¶cient to determine their desires and judgements of value. Callicles has a strong but misguided sense of honour and shame: he wrongly scorns the life of philosophy and justice, calling this life shameful and unmanly.45 Polus is ashamed of his desire for the pleasures of the unjust life, but desires them none the less. The dialogue thus points to a weaker theory than the one it attributes to Socrates. Being shameful is one way of appearing bad, and being admirable is one way of appearing good; these appearances can compete with but may not defeat rival appearances, in particular the appearances generated by pleasure. And the appearances of goodness or badness revealed to us through shame are not always true: sometimes our feelings of shame are misguided. These very same shortcomings of shame surface in the Republic as weaknesses of spirit, traits that can make spirit impotent or even dangerous. First, spirit’s tendency to oppose appetite—to find base pleasures shameful—is not enough to guarantee virtue, for spirit may be overpowered by appetite: Leontius, like Polus, desires and pursues a pleasure even though he recognizes it as aischron. Second, 44 More precisely, Callicles thinks that shame felt at something shameful ‘by nature’ reflects our sincere judgements, but shame felt at something shameful merely by nomos, convention, does not; moreover, he thinks, both Gorgias and Polus were shamed by things shameful merely by convention. For more on this distinction, see below. 45 Callicles says that the philosopher is not καλ6ς κ4γαθ)ς (484 d 1–2), that philosophy is shameful for adults (485 a 5, 489 b 8), that pleonexia and nurturing one’s own appetites are most kalon ‘by nature’ (483 a–b, 491 e 6 ·.), and that the most shameful thing is to be unable to protect oneself from su·ering injustice (483 a 7–b 4, 486 a 4 ·.; cf. 508 d 4, 522 c 4–6).

Created on 4 July 2005 at 14.33 hours page 161

Shame, Pleasure, and the Divided Soul

161

and worse still, spirit does not always feel shame at the bad and admiration for the good. Sometimes spirit opposes the dictates of reason: Plato uses Homer’s Odysseus as a paradigm of such a case, perhaps implying that the Homeric code of honour and shame generally fails to track the truth about good and bad.46 Indeed, the censorship of poetry in books 2 and 3 is in part motivated by this complaint: these poems teach spirit to admire much that is bad and feel shame at much that is good.47 Republic 8 o·ers examples of characters like Callicles in whom spirit’s values are dangerously wrong: the timocratic man, who values honour (reputation and glory) and victory above all else (548 e ·.), and the oligarch, who will not allow his spirited part ‘to esteem or honour anything other than wealth and wealthy people, or to aspire to anything other than the acquisition of money’ (553 d 4–6). If shame and the other thumoeidic emotions can go wrong in these ways—if Socrates’ view of shame in the Gorgias was too simple and too optimistic—then we are left with the questions posed at the beginning of this section. Why is shame a promising tool of moral persuasion, while appetites for pleasure lead us astray? Why in the Republic is spirit reason’s ally, and appetite the enemy? This asymmetry between spirit and appetite is more puzzling than it may at first appear. The explanation cannot be simply that what spirit admires, the kalon, is coextensive with the good, while what appetite desires, the pleasant, is not. First, Republic 9 argues that in an important sense the pleasant (the most pleasant, or the truly pleasant) is coextensive with the good.48 Second, even 46 Socrates quotes Od. 20. 17–18, at 441 b. 47 For citations, see n. 60 below. The censorship is also motivated by the complaint that such poems nurture and gratify unhealthy appetites. 48 580 d ·. I am here simplifying some complicated issues about pleasure. First, what is most truly pleasant, according to Republic 9, is not an object of desire for appetite at all, but rather for reason. None the less, Plato argues that the appetites are best satisfied—get ‘the truest pleasures possible’ for them, ‘and the ones that are most their own’ (586 d 8–e 1, trans. Grube/Reeve)—when they do as reason commands. Therefore the virtuous life is indeed the most appetitively pleasant, as well as o·ering loftier pleasures to reason. Second, perhaps it is not after all possible to be mistaken as to whether or not something is pleasant, although it is possible to be mistaken as to whether or not something is kalon. Although Republic 9 calls the base pleasures desired by vicious appetites ‘mere images and shadowpaintings of true pleasures’ (586 b 7–8, trans. Grube/Reeve), and the Philebus calls them ‘false’ (Phileb. 36 c ·.), it is arguably Plato’s claim that even these pleasures are pleasures: ontologically inferior pleasures, but pleasures none the less. What

Created on 4 July 2005 at 14.33 hours page 162

162

Jessica Moss

though what is truly kalon is good, one can be mistaken about what is kalon—witness again Callicles’ derision of philosophy, and the corrupt souls of Republic 8; in such cases spirit will be no more reliable a guide to the good than appetite. The explanation for the asymmetry must lie, then, not in any facts about the objects that the di·erent parts of the soul desire (the pleasant vs. the kalon), but rather in facts about the nature of the parts of the soul themselves. Appetite desires what it finds pleasant; although what most people find pleasant is worthless or harmful, what is really pleasant is precisely what is good; therefore in principle it should be possible to train the appetites to recognize and pursue the truly pleasant.49 The di¶culty seems to be that there is no way to alter what the appetites desire without first altering the appetites themselves. The Gorgias and Republic characterize vicious people’s appetites as different in nature from those of virtuous people: as the Gorgias puts it, vicious people’s appetites are unruly, insatiable, and shifty.50 This di·erence in constitution determines the respective pleasures of the vicious and the virtuous: it entails that harmful things please the vicious person, while the moderate, orderly pleasures of virtue do not. Thus there is simply no way to make a vicious person take pleasure in virtue. One must first devise a way to discipline and weaken her appetites; once they have been made temperate, they will come to take pleasure in things that are genuinely good. But this is to say that one must first make a person virtuous before she can take appetitive pleasure in virtue; and if this is so, then it is obvious that appeal to appetite cannot be a means of moral education.51 someone mistakenly thinks kalon, on the other hand, is in no way kalon at all, any more than what someone mistakenly thinks good is good. 49 We may compare the Protagoras’ art of measuring pleasures, which trains people to recognize what is most pleasant, and thereby makes them virtuous. Plato abandons any optimism about pleasure as a tool of moral education in the Gorgias and Republic. 50 See Gorg. 493 a ·.; cf. 505 a–b. 51 Aristotle holds that pleasure is instrumental in moral education, and attributes the same view to Plato: ‘One should be brought up right from childhood, as Plato says, to enjoy [χαρειν] and be pained by [λυπε!σθαι] the things one should; for this is what correct education is’ (NE 2. 3, 1104b12–14). But enjoyment and pain here do not refer solely or even chiefly to appetitive pleasure: we might paraphrase Aristotle as saying that, according to Plato, one must be brought up not to take appetitive pleasure in what is bad, and to take another, loftier sort of pleasure (thumoeidic or rational) in what is good. Burnyeat argues that for Aristotle it is precisely the pleasures of spirit that are at issue (M. Burnyeat, ‘Aristotle on Learning to Be Good’, in A. O. Rorty (ed.), Essays on Aristotle’s Ethics (Berkeley and London, 1980), 69–92 at 79).

Created on 4 July 2005 at 14.33 hours page 163

Shame, Pleasure, and the Divided Soul

163

Whether or not a thing appeals to the appetites depends on the constitution of the appetites, and therefore the appetites are not a promising target of moral education. In so far as moral education in the Republic is concerned with the appetites at all, its chief concern is to repress the dangerous ones.52 What about spirit? When we look at the details of the elaborate childhood education described in books 2 and 3, we see that spirit is in fact its primary target.53 Far from ignoring or suppressing spirit, moral education nurtures and trains it. Why is spirit educable where appetite is not? Precisely, I shall argue, because what appeals to spirit does not depend on facts about its brute constitution. Spirit, unlike appetite, is malleable. The roots of this idea, moreover, are articulated in the Gorgias— not by Socrates, but by Callicles. For in criticizing Socrates’ use of shame Callicles expresses his own view about it, and although this view is not put forth as one that Plato embraces, it does contain several crucial points that will surface in the Republic’s characterization of spirit. Callicles barges into the discussion at 481 b with the accusation that Socrates’ use of shame in his argument against Polus was deceitful and unfair. He distinguishes between what is shameful by nature and what is shameful merely by convention (nomos), and implies that most of the time when we call something shameful we are reporting on social convention rather than expressing our own deepest beliefs.54 Shame is mostly a matter of not wanting to look 52 Harmful, unnecessary appetite is something ‘that most people can get rid of, if it is disciplined and educated from childhood’ (559 b 9–10); in the well-ordered soul, the rational part ‘should take care of the many-headed beast’—the appetites—‘as a farmer does his animals, feeding and domesticating the gentle heads and preventing the savage ones from growing’ (589 b 1–3, trans. Grube/Reeve, emphasis added). A proper education and upbringing will kill o· the bad appetites, or even prevent them from arising. (As to the necessary appetites, the ‘gentle heads’, a virtuous person ‘feeds and domesticates’ them: presumably this entails giving the body enough to eat and drink, and providing body and soul with temperate pleasures. Plato does not say more on this topic, and it would be odd to consider the care of the appetites as a significant part of the positive programme of moral education he describes.) 53 This is explicit at Rep. 410 b–412 a, where Plato describes the goal of the education in music and physical training as the harmonizing of spirit with reason. I believe that spirit is the implied target throughout much of books 2 and 3: it is spirit that is a·ected when stories or music present certain actions or characters as kalon, encouraging the audience to admire and imitate them, and others as aischron, encouraging the audience to disdain them. See n. 60 below for citations. 54 Callicles says that Polus called committing injustice more shameful because he was ‘ashamed to say what he thought’ (482 e 2), and then says that Polus meant that committing injustice was more shameful by nomos only, not by nature (483 a 6). We

Created on 4 July 2005 at 14.33 hours page 164

164

Jessica Moss

bad to others. In judging something shameful, we are responding to thoughts about what others think of us, rather than to our own thoughts about what is good and bad. Moreover, while what is by nature shameful is bad and what is by nature kalon good, conventions about the kalon and shameful often fail to track the truth about good and bad (483 a 7 ·.). Certainly the dialogue does not present Callicles as having the final word on shame. He begins by bragging that he is above shame and ends by being caught in its snare himself: is it not likely that he underestimates the power of shame because he misunderstands its nature? Indeed, his distinction between the shameful by nature and the shameful by convention is too stark, and Plato indicates this by showing that Callicles’ own feelings of shame turn out to accord closely with social convention. Callicles finds the catamite and the coward repugnant not merely because society condemns them, but neither is it a coincidence that he finds shameful what society condemns.55 The most natural interpretation is that Callicles’ own view of what is shameful has been shaped by society—that, like most people, he has internalized social norms about what is shameful and what admirable, despite his radical views. As Bernard Williams argues, shame is a social emotion, not because we feel shame only when we consider what others think of us, but rather because through shame we internalize an external, societal viewpoint, and use it to guide our own judgements about ourselves.56 Callicles is right to point out that we feel shame when we think of how others will react to us, and Plato dramatizes the have to do a bit of work to see how the two charges are connected. Callicles means first, that one genuinely believes something shameful only if one thinks it shameful by nature; second, that if something is shameful by nomos, one will be ashamed to admit to others that one admires it. (According to Callicles, then, Polus does feel shame at the thought of injustice, as well as saying that he finds it shameful; the shame he feels, however, is not (as Socrates implies) a result of his finding injustice genuinely shameful.) 55 Condemnation of the catamite was a social institution: a prostitute was not allowed to speak in the assembly or hold public o¶ce (K. J. Dover, Greek Homosexuality (Harvard, 1978), 27). Aeschines refers to prostitution as the shaming of one’s body for hire (7'ν τις ;θηναων 7π< τ0 το( σ:µατος α-σχ/ν0η =κ>ν µισθαρν0 . . ., Against Timarchus 87 and throughout). 56 ‘Whatever it is working on, [shame] . . . requires an internalised other . . . whose reactions the agent can respect . . . [T]his figure . . . embodies intimations of a genuine social reality—in particular, of how it will be for one’s life with others if one acts in one way rather than another’ (B. Williams, Shame and Necessity [Necessity] (Berkeley and London, 1993), 102).

Created on 4 July 2005 at 14.33 hours page 165

Shame, Pleasure, and the Divided Soul

165

same point early in the dialogue when he shows that Gorgias feels shame about rhetoric’s shortcomings in front of people who admire him.57 But Callicles is wrong to think that this fact discredits shame—to think that the social nature of shame disqualifies it from being among the agent’s genuine beliefs and motivations. Williams might (although he does not) have had Callicles’ dismissive view of shame in mind when he argues that ‘the Greeks’ understanding of shame . . . was strong and complex enough to dispose of the familiar criticism that an ethical life shaped by it is unacceptably heteronomous, crudely dependent on public opinion’ (97). The Gorgias, then, presents two views of shame, Socrates’ and Callicles’, and suggests problems for both. When we look at the Republic’s characterization of spirit, the shame-feeling part of the soul, we see that Plato accepts some of Callicles’ criticisms of Socrates’ view, but does not think that they undermine the moral force of shame. For here we find a picture that has significant points in common with Callicles’ view, yet o·ers a more subtle understanding of the social nature of shame. On this refined view, the fact that shame judgements are socially inculcated turns out to increase, not detract from, shame’s potential as a tool of moral education. We can begin by noting a striking parallel between the speech in which Callicles denounces the social inculcation of shame, and the Republic’s description of the childhood education of the guardians. Callicles says that what is by nature kalon is to be powerful, strong, and ‘able to have more than one’s share [πλον ?χειν]’ (483 c 2). Those who make the rules (ο@ τιθµενοι τοAς ν)µους, 483 b 4–5), however, are the weak: unable to πλον ?χειν, they declare that it is just and kalon to have only an equal share. What is by nature 57 Throughout the conversation, Socrates reminds Gorgias that he is on display, and that his reputation as a teacher of rhetoric is at stake. Gorgias boasts of his rhetorical skill (449 a), his ability to speak succinctly (449 c), and the power of rhetoric (455 d ·.); Socrates gets him to engage in dialectic and to remain in the conversation by flattering him, by reminding him of his boasts, and by reminding him of his reputation (449 c–d, 455 c–d). As the conversation continues, and Gorgias fails to give good answers to Socrates’ questions, Gorgias’ responses stop displaying vanity and start displaying shame. He is trying to save face in front of the audience; at one point he even tries to escape the conversation by suggesting that his listeners are tired (458 b–c). When Callicles urges him to continue, he says, ‘It would be shameful after all this for me to be unwilling, having announced that I would answer whatever anyone wanted’ (458 d 7–e 1). Now shame has reared its head, and it is not long before Gorgias makes the fatal admission which Polus diagnoses as the product of shame (460 a 3–4).

Created on 4 July 2005 at 14.33 hours page 166

166

Jessica Moss

kalon they call shameful, while what is by nature shameful they call kalon.58 The rulers of society, then, wish to impose a morality upon their subjects—they wish to influence people’s judgements about what is shameful and what kalon. How do they do so, according to Callicles? By means of manipulative childhood education: Moulding [πλ'ττοντες] the best and most powerful among us, taking them from childhood . . . we enslave them with charms and sorcery [κατεπ'δοντς τε κα< γοητε/οντες], telling them that one must have an equal share, and that this is the kalon and the just. (483 e 4–484 a 2)

Plato makes similar use of the word ‘moulding’ in his description of the childhood education of the guardians in Republic 2. Here nurses and mothers tell prescribed stories to children, ‘and mould [πλ'ττειν] their souls with these stories’ (377 c 3–4; cf. 377 b 2).59 Moreover, the similarity goes far beyond this coincidence in vocabulary. For the educational programme of Republic 2–3 is an explicit attempt to carry out a version of the programme Callicles decries. Here too we have lawgivers who wish to impose a morality on their citizens: they wish to keep them free of the false morality that sees no intrinsic value in justice. They do so precisely by means of manipulative childhood education. Taking the citizens when they are young, they ‘mould’ them by means of music, poetry, stories, and other forms of non-rational influence which one might cynically describe as ‘charms and sorcery’ (Callicles’ phrase at 483 e 6). What is more, the Republic’s childhood education directly targets judgements about the shameful and the kalon.60 Reason, the part of the soul that has thoughts about good and bad, is undeveloped in children; but spirit, the part of the soul that feels shame and admiration, is present from the start (441 a 7–b 1). ‘Musical’ education thus targets not reason but spirit: it aims not to instil beliefs about the good and bad directly, but rather to cultivate children’s sense of shame about the very things that they should, when they come to reason, think bad, and their admiration for what they should think 58 ‘By nature what is worse is also wholly more shameful: su·ering injustice; but by convention committing injustice is more shameful’ (Callicles at 483 a 7–8). 59 Laws 671 c also implies that lawgivers educate children by moulding (πλ'ττειν) their souls. 60 Shame, the shameful, and the kalon are explicitly mentioned at 378 c 2, 388 d 6, 401 e 4 ·., and 403 c 6. Implicit throughout the discussion of music and poetry is the question of what behaviour and characters should be presented as admirable and what should be presented as shameful, worthy of disdain.

Created on 4 July 2005 at 14.33 hours page 167

Shame, Pleasure, and the Divided Soul

167

good. Without trading in talk of good and bad, it trains spirit’s reactions to track the truth about good and bad: [A]nyone who has been properly educated in music and poetry . . . [will] praise fine things [τ καλ'] . . . [and will] rightly object to what is shameful [τ α-σχρ'], hating it while he’s still young and unable to grasp the reason, but, having been educated in this way, he will welcome the reason when it comes and recognize it easily because of its kinship with himself. (401 e 4–402 a 4, trans. Grube/Reeve)61

How does the musical education succeed in educating spirit? A close investigation lies outside the scope of this paper.62 Details aside, however, we can recognize that in explaining this process Plato is pointing out and relying upon an empirical fact: the stories, songs, and images that we are raised on embed societal values, and in turn they shape our values, determining to a large extent what we admire and approve, as well as what we find distasteful, shameful, and wrong. Callicles’ sinister description of moral indoctrination, and his observation that judgements of shame are socially conditioned, thus reappear in the Republic as crucial parts of a positive programme of moral education. Moreover, the Republic’s characterization of spirit undermines Callicles’ dismissal of the shame a person feels at what is shameful ‘by convention’. Spirit is shaped and influenced by social forces: by the ways in which a person imitates others, admires others, competes with others, and sees herself as others see her. None the less, the judgements and desires of spirit are in no sense external to the agent; they reflect not merely what others think but also the agent’s own values. Precisely in virtue of the 61 Compare Richardson Lear: the aim of the musical education is ‘to mold our perception of the kalon, which we naturally esteem, to fit the form of what is good . . . to teach children to take aesthetic pleasure in—to perceive as fine—images of what is truly good and only them’ (Happy Lives, 142). 62 Book 3 implies that spirit will admire the characters that poems and myths present as worthy of admiration, and will motivate the agent to imitate them: see e.g. 395 c–d and 401 b–c. Gosling has a good discussion of the e·ect of myths and poetry on spirit: ‘The treatment of early education and the criticisms of poetry in the early books are concerned with the development of the spirited part. . . . gods and heroes are naturally to be taken as models . . . Children reading of their doings naturally aspire to do likewise, and take from them their standards of what is fine and what shabby behaviour . . . It is the normal function of legends of gods and heroes to arouse this capacity for admiration and fix, in youth, the general lines of an ideal which may stay with a man for life’ (J. C. B. Gosling, Plato (London, 1973), 44–5). A full study of the education of spirit must also take account of book 3’s discussion of music.

Created on 4 July 2005 at 14.33 hours page 168

168

Jessica Moss

fact that there is a thumoeidic part of the soul, society’s values can directly shape our own.63 Because they are open to social influence, shame judgements can be shaped and guided in the direction that society sees fit. In a society ruled by philosopher kings, what is shameful ‘by nature’, truly shameful, will also be shameful by convention, and therefore the citizens’ shame judgements will be shaped and guided to conform to moral truth.64 Appetite is not subject to positive moral education, and thus can play no positive role in virtue, because it is not possible to ‘redirect’ appetitive desire towards the good. Both Callicles’ speech and the Republic’s programme for musical education show that unlike desires for pleasure, feelings of shame and admiration are subject to morally useful redirection. It is possible to shape a person’s feelings of shame, in a way in which it is not possible to shape her appetites. If left to their own devices, or to the mores of a corrupt society, people may be mistaken about what is shameful and what kalon, just as they will be mistaken about what is truly pleasant.65 The right education, however, can train spirit to track the truth about good and bad. Hence the power of spirit, when properly educated, to combat appetite, ally itself with reason, and ensure virtue in the soul. The Gorgias raises a serious worry for moral education, by showing 63 Curiously, Williams overlooks the importance of spirit in the Republic, accusing Plato of finding no place for shame in his ethics. ‘The Guardians do not need [an internalised other] . . . because they have internalised something else, and carry in them a paradigm of justice gained from their intellectual formation’ (Necessity, 99). By Williams’s lights, this gives Plato an impoverished picture of morality similar to Kant’s. It seems to me that the Republic’s emphasis on the role of spirit in virtue and in moral education should protect Plato from this charge. (Although if there is indeed a distinction between civic and genuine courage (see n. 43), perhaps Williams’s criticism does apply to the latter.) 64 Plato does not reject Callicles’ distinction between what is truly shameful and what is shameful according to false conventions any more than he rejects Callicles’ claim that childhood education shapes our sense of shame. The Republic urges censorship of the poems that have so powerfully established false conventions in Plato’s own society, and seeks to replace them with art that teaches people to admire what is truly kalon and feel shame at what is truly shameful instead. 65 Outside of the ideal city a young person will be corrupted by the de facto moral education of the crowd, ‘carried by the flood wherever it goes, so that he’ll say that the same things are kala or aischra as the crowd does, follow the same way of life as they do, and be the same sort of person as they are’ (492 c 6–8, trans. Grube/Reeve, with my revisions).

Created on 4 July 2005 at 14.33 hours page 169

Shame, Pleasure, and the Divided Soul

169

that the appetites tend to lead people to vice and to undermine rational e·orts to bring them to virtue. It also begins to suggest a solution to this problem: shame and admiration have the power to succeed where reason fails, in revealing the value of virtue and silencing or overcoming the appeal of pleasure. The Republic develops this insight into an account of the soul on which spirit plays a crucial role alongside reason and appetite, showing us that spirit is no ad hoc afterthought but instead a crucial part of Plato’s picture of virtue and moral education. In doing so, the Republic o·ers us a tempered version of Socrates’ view of shame in the Gorgias: shame is a morally useful force in the soul not because it is authoritative and truth-tracking by nature, but because it can acquire those traits through careful education. University of Pittsburgh

B I B L I O GR A P HY Annas, J., An Introduction to Plato’s Republic (Oxford, 1981). Burnyeat, M., ‘Aristotle on Learning to Be Good’, in A. O. Rorty (ed.), Essays on Aristotle’s Ethics (Berkeley and London, 1980), 69–92. Cooper J. M., ‘Plato’s Theory of Human Motivation’ [‘Motivation’], History of Philosophy Quarterly, 1 (1984), 3–21, repr. in Cooper, Reason and Emotion: Essays on Ancient Moral Psychology and Ethical Theory (Princeton, 1999). 118–37. (ed.), Plato: Complete Works [Plato] (Indianapolis, 1997). Dodds, E. R., Plato: Gorgias (Oxford, 1959). Dover, K. J., Greek Homosexuality (Harvard, 1978). Gosling, J. C. B., Plato (London, 1973). Grube, G. M. A. (trans.), Plato: Republic, rev. C. D. C. Reeve (Indianapolis and Cambridge, 1992). Haidt, J., ‘The Emotional Dog and its Rational Tail: A Social Intuitionist Approach to Moral Judgment’, Psychological Review, 108 (2001), 814–34. Hardie, W. F. R., A Study in Plato (Oxford, 1936). Hobbs, A., Plato and the Hero: Courage, Manliness and the Impersonal Good [Hero] (Cambridge, 2000). Irwin, T., Plato’s Ethics (New York and Oxford, 1995). Kahn, C., ‘Drama and Dialectic in Plato’s Gorgias’ [‘Drama’], Oxford Studies in Ancient Philosophy, 1 (1983), 75–121.

Created on 4 July 2005 at 14.33 hours page 170

170

Jessica Moss

Plato and the Socratic Dialogue: The Philosophical Use of a Literary Form [Dialogue] (Cambridge, 1996). Lorenz, H., ‘Desire and Reason in Plato’s Republic’, Oxford Studies in Ancient Philosophy, 27 (2004), 83–116. McKim, R., ‘Shame and Truth in Plato’s Gorgias’ [‘Shame’], in C. Griswold (ed.), Platonic Writings/Platonic Readings (New York and London, 1988), 34–48. Penner, T., ‘Thought and Desire in Plato’, in G. Vlastos (ed.), Plato: A Collection of Critical Essays, ii. Ethics, Politics, and Philosophy of Art and Religion (New York, 1971), 96–118. Reeve, C. D. C., Philosopher-Kings: The Argument of Plato’s Republic [Philosopher-Kings] (Princeton, 1988). Richardson Lear, G., Happy Lives and the Highest Good [Happy Lives] (Princeton, 2004). Robinson, T. M., Plato’s Psychology, 2nd edn. (Toronto, 1995). Santas, G., Socrates: Philosophy in Plato’s Early Dialogues (London, 1979). Shorey, P., What Plato Said (1933) (repr. Chicago, 1979). Vlastos G., ‘Was Polus Refuted?’ [‘Polus’], American Journal of Philology, 88 (1967), 454–60. ‘The Socratic Elenchus’, Oxford Studies in Ancient Philosophy, 1 (1983), 27–58. Socrates: Ironist and Moral Philosopher (Cambridge and Ithaca, NY, 1991). Williams, B., ‘The Analogy of City and Soul in Plato’s Republic’, in E. N. Lee, A. P. D. Mourelatos, and R. M. Rorty (eds.), Exegesis and Argument: Studies in Greek Philosophy Presented to Gregory Vlastos (Phronesis suppl. 1; Assen, 1973), 196–206. Shame and Necessity [Necessity] (Berkeley and London, 1993).

Created on 4 July 2005 at 14.34 hours page 171

ANI MADVERS I ONS ON BURNYEAT’ S THEAETETUS: ON THE LOGI C OF THE EXQUIS ITE ARGUMENT MICHAEL V. WEDIN

 book Gamma of the Metaphysics Aristotle criticizes Protagoras’ doctrine that man is the measure of all things. This he understands as the thesis that what appears to be the case is the case. I shall refer to this as Protagoras’ Thesis. For the moment, we may formulate it as (PT) (x)(p)(x bel p → p), or, in a truth-added version, (PT ) (x)(p)(x bel p → p is true). T A number of Aristotle’s criticisms in the Metaphysics focus on absurd or unacceptable consequences of PT, without purporting to refute the thesis directly. In Γ 6, however, he turns to the thesis itself. But so far from providing a refutation, the chapter opens with the suggestion that the demand to prove PT false is itself mistaken, and indeed impossible to fulfil. So we should hardly expect the balance of Γ 6 to contain a proper refutation. Elsewhere I comment on Aristotle’s attitude towards Protagoras.1 At the moment, however, my main object of interest is Plato’s treatment of his illustrious ã Michael V. Wedin 2005 This paper has emerged from reflection over the past couple of years on Myles Burnyeat’s marvellous article on Protagoras and self-refutation, much of it carried on in seminars at Davis. In the first instance, I am indebted to the participants in those sessions, especially James Blackman. Burnyeat read, with characteristic grace and candour, a nearly finished version of the paper. In the final stage, I once again benefited from the acute comments of David Sedley, editor without peer. I am grateful to them all. It would please me most were the paper to serve as a mark of my esteem for Myles Burnyeat and his contributions to the study of the ancient philosophers. 1 In M. V. Wedin, ‘A Curious Turn in Metaphysics Gamma: Protagoras and

Created on 4 July 2005 at 14.34 hours page 172

172

Michael V. Wedin

predecessor. The interest stems, in part, from the fact, or likely fact, that a good deal of Metaphysics Gamma was composed with an eye to Plato’s representation of Protagoras in the Theaetetus. In Γ 5, for example, at 1010b11–14, Aristotle approves of Plato’s worry that according to PT the opinion of the doctor will carry no more weight than that of an ignorant man. This is an explicit reference to the Platonic dialogue (178 c–d), and there are other indicators. So an examination of Plato’s remarks might illuminate Aristotle’s account. But in this paper I am interested in Plato’s treatment mainly because of its boldness. In what has been called ‘the exquisite argument’, at Theaet. 171 a–b, Plato claims to refute Protagoras in his own terms—precisely what Aristotle declared impossible. We see nothing of this ‘self-refutation’ argument in Metaphysics Gamma, and Aristotle’s silence reflects a less sanguine attitude about the prospects for any such argument. Aristotle appears to have been followed by commentators young and old. But not by all. In a seminal paper on Plato’s refutation of Protagoras, Myles Burnyeat has argued that the ‘self-refutation’ argument in fact works.2 Thus, more than a few Plato scholars, and possibly Aristotle, are mistaken. Burnyeat’s account is surely the most able defence of Plato’s exquisite argument.3 Although the Strong Denial of the Principle of Non-Contradiction’, Archiv f•ur Geschichte der Philosophie, 85 (2003), 107–30, and ‘On the Use and Abuse of Non-Contradiction: Aristotle’s Critique of Protagoras and Heraclitus in Metaphysics Gamma 5’, Oxford Studies in Ancient Philosophy, 26 (2004), 213–39. 2 M. F. Burnyeat, ‘Protagoras and Self-Refutation in Plato’s Theaetetus’ [‘SelfRefutation’], Philosophical Review, 85 (1976), 172–95, repr. in S. Everson (ed.), Companions to Ancient Thought, i. Epistemology (Cambridge, 1990), 39–59; Burnyeat, The Theaetetus of Plato (Indianapolis, 1990), rea¶rms the view in slightly more cautious language. 3 Perhaps this ought to come with the caveat ‘. . . most able defence of a relativist interpretation of Plato’s exquisite argument’. Other interpretations have been advanced, perhaps most notably the ‘infallibilist’ interpretations of S. Waterlow, ‘Protagoras and Inconsistency’, Archiv f•ur Geschichte der Philosophie, 59 (1977), 19–36; G. Fine, ‘Protagorean Relativisms’ [‘Relativisms’], in J. Cleary and W. Wians (eds.), Proceedings of the Boston Area Colloquium in Ancient Philosophy, 10 (Lanham, Md., 1994), 211–43; ‘Relativism and Self-Refutation in Plato’s Theaetetus: Plato, Protagoras, and Burnyeat’ [‘Self-Refutation’], in J. Gentzler (ed.), Method in Ancient Philosophy (Oxford, 1998), 137–63; and ‘Plato’s Refutation of Protagoras in the Theaetetus’ [‘Plato’], Apeiron, 32 (1998), 201–34. Although there is much more to the view, infallibilists take Protagoras to claim that whatever anyone believes is true simpliciter, whereas relativists take him to claim that whatever is believed is true for whoever does the believing. I think Burnyeat is correct that relativism is the harder doctrine to combat. Thus, a relativist reading at once awards Protagoras a more invulnerable position and enhances the standing of any argument

Created on 4 July 2005 at 14.34 hours page 173

Animadversions on Burnyeat’s Theaetetus

173

account has received wide attention, his version of the exquisite argument itself deserves further scrutiny. In particular, it raises a number of interesting logical problems that need to be addressed. So it is Burnyeat’s version of Plato’s argument that I wish to examine. The following reflections, on what is arguably the best reconstruction available, show, I believe, that Plato’s fabled argument comes up short after all, but for very interesting logical reasons.

1. The context of the exquisite argument According to Burnyeat, the exquisite argument is the last of a set of three ‘linked arguments’ against Protagoras, constructed around the trivial truth that Protagoras believes PT or Protagoras does not believe it. In e·ect, Plato argues: (P1) if Protagoras does not believe PT, and no one else does, then PT is not true for anyone; (P2) if Protagoras does believe PT, and the majority of doxastic agents do not believe it, then (P2a) the more PT is not true than it is true, and (P2b) Protagoras must agree that PT is false. The conclusion of (P1) sets the target for the trio of arguments, namely, establishing that PT is true for no one. Burnyeat refers to this as the ‘crowning blow’. But it relies on the strong assumption that no one believes PT, and it is wildly implausible to attribute this to Protagoras. The reasoning in (P2) is meant to make this attribution stick. For (P2) grants that Protagoras believes his own doctrine, but then proceeds to argue that, on this very assumption, PT must be false—whether more false than true, case (P2a), or false by Protagoras’ admission, case (P2b). The majority argument, as one might call (P2a), can be set aside. It has no probative weight of its own and plays no role in establishing the crowning blow. This task is assigned to (P2b), for unlike (P2a), which merely allows the truth of others’ beliefs in the falsity of PT, (P2b) purports to show that Protagoras himself must join with the deniers of PT, precisely because he champions PT. Since this would hold for any advocate of PT, (P2b) promises to yield the crowning blow envisioned in (P1). What (P2b) describes is, of course, the exquisite argument. The argument is crisply stated at Theaet. 171 a–b: that defeats it. Burnyeat is advancing just such an argument, and so its success would be especially noteworthy. By the same token, assessment of the argument may yield especially noteworthy logical problems. And so, it seems to me, it does.

Created on 4 July 2005 at 14.34 hours page 174

174

Michael V. Wedin

Secondly, it [PT] has this most exquisite feature: on the subject of his own view, agreeing that everyone judges what is so, he for his part presumably concedes to be true the opinion of those who judge the opposite to him in that they think that he is mistaken . . . Accordingly, he would concede that his own view [PT] is false, if he agrees that the opinion of those who think him mistaken is true. (Burnyeat, ‘Self-Refutation’, 173)

The argument contained in these lines is the centrepiece of Plato’s self-refutation of Protagoras, and on its success or failure turns the course of Plato’s campaign against Protagorean relativism. In Burnyeat’s view the argument is successful. Let us take a closer look.

2. The logic of the exquisite argument According to Burnyeat, the function of the exquisite argument is ‘to show that Protagoras joins with everyone else in disputing his theory of truth’. Then, because Protagoras must join everyone else in denying PT, PT is not true for anyone at all. The latter, according to Burnyeat, gives ‘spectacular form’ to the refutation. But it rests on the exquisite argument, and that is the argument I wish to explore at the moment. The argument concerns the truth, or falsity, of PT, and so it will proceed by using this notion. Mindful of the fact that the measure doctrine is relativistic,4 Burnyeat proposes to capture this with the device of truth-in-a-subject’s-world. Thus, PT above is to be replaced by a worlds-version: T (PT ) (x)(p)(x bel p → in x’s world it is true that p).5 W With this, we may consider the form of the exquisite argument according to Burnyeat, beginning with the measure doctrine itself, which now says that whatever someone believes is true in his world: (1) (x)(p)(x bel p → in x’s world it is true that p). Proposition (1), which denies that anyone believes something that is not true in his world, is equivalent to (1a) ¬ (∃x)(∃p)(x bel p ∧ ¬ (in x’s world it is true that p)). 4 Not all parties agree: see Fine and Waterlow (n. 3 above) for a di·erent line. 5 I use ‘x bel p’ to express that x believes that p, or that it appears to x that p, or that x judges that p. At least for the moment, nothing turns on this semantic looseness.

Created on 4 July 2005 at 14.34 hours page 175

Animadversions on Burnyeat’s Theaetetus

175

Because (1a) a·ords a more perspicuous presentation of the argument, I shall usually operate with it. Now suppose that someone believes that PT is false, i.e. that (1a) is false: (2) (∃y)(y bel (∃x)(∃p)(x bel p ∧ ¬ (in x’s world it is true that p))).6 It appears to follow straightforwardly from (1) and (2) that (3) (∃x)(∃p)(x bel p ∧ ¬ (in x’s world it is true that p)), which is just the negation of PT, i.e. of (1a). So if someone believes that PT is false, then PT itself enjoins that PT is false. Thus, armed with the weak and, hence, plausible claim in (2), it appears that PT entails its own falsity, and that, in this sense, it is a self-defeating doctrine. However, according to (1) beliefs generate truths relative to believers. So (3) itself must be supposed established relative to a believer, e.g. Socrates. It would be in some believer’s world that PT is false. Rather than (3), then, at most Plato is entitled to (3a) (∃x)(∃p)(x bel p ∧ ¬ (in x’s world it is true that p)) is true in y’s world. Since (3a) follows from a belief held by bound y in (2), (3a) really should be replaced by (3b) (∃y)((∃x)(∃p)(x bel p ∧ ¬ (in x’s world it is true that p)) is true in y’s world). As I read Burnyeat, he would take (1), or (1a), to assert that everyone lives in a private world where belief entails truth and the conclusion, (3), to assert that the denier of PT does not live in such a world. Hence, PT asserts that everyone lives in a private world and entails that not everyone does. Clearly, this is an undesirable result for Protagoras. But this is not quite what the argument shows, for we can now see that (3b) does not rule out that y, the denier of PT, lives in a private world where belief entails truth. For what (3b) 6 Of course, one might worry that (2) requires that such a y actually cast his belief in quantifier logic. But I shall assume that the discourse is among logically informed persons and so set such concerns aside. And this seems appropriate if we are concerned with the logic of the argument, which surely is an argument among philosophers.

Created on 4 July 2005 at 14.34 hours page 176

176

Michael V. Wedin

says is that there is some world in which the entailment fails to hold. But this world need not be y’s. This alone is the negation of PT, and so this alone is what the explicit denier of PT denies. Thus, suppose that the world in which PT is denied is not y’s world. In this case, (3b) says that it is true in y’s world that it is true in some other world, say z’s, that something is believed but not true. But Burnyeat’s Protagorean worlds are private worlds and, hence, quantifiers cannot reach across them. In short, Protagorean semantics does not allow for cross-world reference. Yet, just such cross-world commerce is required by (3b). Of course, it could turn out that in (2) bound x and bound y are the same, but this is not something that can be obtained from simply denying PT. And the logical force of the argument calls just for simple denial: that is, it turns on someone’s believing the falsity of PT itself, not the falsity of PT in Socrates’ world, or in Callias’ world, etc. So, as far as the argument is concerned, Protagoras is free to insist that the denier of PT, the bound y of (2), lives in a private world. By the same token PT itself is saved from self-defeat. The exquisite argument promised to refute Protagoras’ measure doctrine by showing that it entails its own falsity, so long as it is granted that someone believes the doctrine false. So it is disappointing to find that the argument fails, even on Burnyeat’s elegant reconstruction. However, as we have seen Burnyeat emphasize, the exquisite argument is one of three related arguments, and perhaps it will be thought that this allows for a more favourable overall estimate of Plato’s attack on Protagoras. Recall that the crowning blow, as Burnyeat puts it, is the conclusion that PT is true for no one at all. The exquisite argument aims to establish that Protagoras’ own belief in PT ‘counts for naught’ and that he is ‘committed, despite himself, to agreeing with other people that it is false’.7 Presumably, the same can be said of any advocate of PT and, thus, the exquisite argument arguably functions as Burnyeat suggests. So it is no trivial matter to find, as we have, that the argument fails. Suppose we agree with the above animadversion on Burnyeat’s reading of the exquisite argument. Still, it might be thought that the ‘crowning blow’ would be established were everyone to believe PT false. Here the measure doctrine entails its own falsity, given the facts about what people believe, and this would be a significant result despite depending on the stronger and less plausible doxastic 7 Burnyeat, ‘Self-Refutation’, 186.

Created on 4 July 2005 at 14.34 hours page 177

Animadversions on Burnyeat’s Theaetetus

177

assumption. Unfortunately, this argument fares little better than the exquisite argument and for similar reasons. Let us see how this is so. What must be shown is that in circumstances where everyone believes the negation of PT, PT is not merely false but self-defeating. In short, in such a world it must entail its own falsity. So we may begin, as before, with the measure doctrine itself: (1) (x)(p)(x bel p → in x’s world it is true that p), and its equivalent: (1a) ¬ (∃x)(∃p)(x bel p ∧ ¬ (in x’s world it is true that p)). Whereas the exquisite argument used the plausible assumption that someone believes the negation of (1a), the ‘unanimity argument’, as I shall call it, has everyone believing its negation. So rather than (2), we have (4) (y)(y bel (∃x)(∃p)(x bel p ∧ ¬ (in x’s world it is true that p))). As I understand the reasoning, (1a) and (4) are supposed to entail that PT is false in everyone’s world. But, of course, from (1a) and (4) what appears to follow is nothing more than: (3) (∃x)(∃p)(x bel p ∧ ¬ (in x’s world it is true that p)). So far, then, the unanimity argument gets us no further than the exquisite argument does—in using (4), it just accomplishes this by overkill. However, in looking at (3) above, in the context of the exquisite argument, I pointed out that what actually follows is a conclusion relativized to the denier of (1a). That gave us (3a) and its fully bound partner, (3b). Following that strategy, in place of (3b) with its particular quantifier we now need a conclusion relativized to all doxastic agents. So we get (5) (y)((∃x)(∃p)(x bel p ∧ ¬ (in x’s world it is true that p))) is true in y’s world), which seems to say that everyone is such that in their world someone believes something that is not true in that world. So read, proposition (5) certainly appears to say that everyone lives in a world where belief is not su¶cient for truth, i.e., in a

Created on 4 July 2005 at 14.34 hours page 178

178

Michael V. Wedin

world where PT is false. But, again, appearances are deceptive. All (5) says is that everyone’s world is such that the claim, that someone (x) believes something that is not true, is true in that world. As before, there is no guaranteeing that this x is identical to the y whose world is in question. And because worlds are private, (5)’s x cannot be distinct and a ‘member’ of the world determined by y’s beliefs. No more than existential quantifiers do universal quantifiers cross such worlds, and hence (5) does not establish that everyone lives in a world where belief fails to entail truth. Indeed, as far as (5) is concerned, it is not clear that anyone lives in such a world. Certainly, to believe that there is such a world is not to be in such a world, and, as we have twice seen, it is far from obvious that this sort of belief can be allowed when Protagorean worlds are interpreted as private worlds. If we both allow that PT is a doctrine about truth in a world and require that such worlds be private, then the argument that PT is self-defeating is itself defeated—both the exquisite argument and the unanimity argument.8

3. The strong measure doctrine: PT as an equivalence The above criticism notwithstanding, there is a further feature of Burnyeat’s account that deserves scrutiny. This concerns the claim that PT is formulated by Plato as a biconditional rather than as a conditional.9 Indeed, the biconditional formulation turns out to be crucial to the ‘crowning blow’ (that PT is true for no one, including Protagoras). To see the connection between the crowning blow and the strong measure doctrine, consider what the exquisite argument and the unanimity arguments aspire to establish, namely, 8 Upon reading a final version of the paper, Myles Burnyeat alerted me to the existence of an unpublished critical note by Gareth Matthews. Matthews was able to locate his ‘ancient’ remarks, and shortly thereafter several purple mimeographed pages appeared in my mailbox. I was pleased to see that Matthews’s reservations about Burnyeat’s reconstruction of the exquisite argument have much in common with my line of criticism. In particular, he (correctly) presses the point that from Socrates’ denial of PT it does not follow that Socrates’ world is not governed by PT. Matthews in turn alerted me to the use of this point in Fine, ‘Self-Refutation’. 9 Joining Burnyeat on this point, Fine, ‘Plato’, 205 n. 12, retracts her claim in ‘Relativisms’ that Protagoras holds only the conditional taking one from belief to truth. But no evidence is o·ered in favour of the biconditional formulation, at least not in Fine, ‘Plato’.

Created on 4 July 2005 at 14.34 hours page 179

Animadversions on Burnyeat’s Theaetetus

179

that everyone lives in a world where belief is not su¶cient for truth. Represent this as (6) ¬ (∃x)(in x’s world PT is true). Proposition (6) is equivalent to (6a) (x) ¬ (in x’s world PT is true), and this in turn may be taken as equivalent to (6b) (x)(in x’s world PT is false). So (6b) captures, in Burnyeat’s idiom, what Plato aimed to prove, namely, that PT is true for no one and so false for everyone. Suppose, now, that PT is formulated as (1) above: that is, suppose that it is formulated as a conditional licensing the move from belief to truth. Assume that everyone with any doxastic attitude towards PT believes it to be false. Then it follows that PT is false in all worlds determined by such believers. But it does not follow from this that PT is everywhere false, for in a world where it is not the object of belief PT might, for all we know, be true. It is clear that the conditional formulation allows this possibility, and just as clear that Plato aims for the strong result that PT is false for everyone, period. Now, Plato could just assert that everyone does believe the negation of PT, and this would be su¶cient for the strong result. We can, however, get the same result on the weaker doxastic assumption that anyone with a doxastic attitude towards PT believes it to be false. For, by upgrading (1) to a biconditional: (1*) (x)(p)(x bel p ↔ in x’s world it is true that p), we guarantee not only that anything believed is true, but also that anything true is believed. So there could not be a world that contained a truth that was not also an object of belief. In particular, there could be no world in which PT was true but not the object of belief. So it is not true in anyone’s world. Thus, the strong measure doctrine appears to deliver the crowning blow, namely, the result that PT is true for no one at all and so, as (6b) declares, false for everyone. Well, suppose that PT is false for everyone. Does it follow from this that PT itself is false? In approaching this, I shall assume that Protagoras intends PT to stand as a general thesis about belief and

Created on 4 July 2005 at 14.34 hours page 180

180

Michael V. Wedin

truth. In this I agree with Burnyeat. According to the thesis, truths are relativized to believers. That is the point of Burnyeat’s apt idiom, ‘true-in-so-and-so’s-world’. But nothing says that PT itself is relativized to anything. We may simply take it as a free-standing thesis about belief and truth. Of course, when someone believes PT, then it is subject to relativizing. But for purposes of refuting PT itself, it is the free-standing thesis that must be rejected. Once this point is granted, however, it becomes unclear what bearing the strong result, (6b), has on PT itself because it is unclear how the fact that PT is false in everyone’s world entails that it is false as a free-standing proposition. The biconditional, (1*), eliminated trouble from propositions that are not the object of someone’s belief. But if PT itself, as a free-standing thesis, lies outside the domain of relativized propositions, it is not clear why it should be false just because it is false in everyone’s world. Perhaps this situation can be headed o· by simply denying that there are any propositions that are not believed, period. We get this by replacing PT with what might be called the ‘absolutely strong measure T doctrine’, namely, (PT *) (x)(p)(x bel p ↔ p is true). T Because it holds for any proposition whatever, PT * e·ectively T foists an idealist thesis on Protagoras. Whatever is true and, so, whatever is the case, is something that is believed and vice versa. In short, there is no truth at all that is not an object of belief. By eschewing the ‘true for x’ idiom, PT * will strike most comT mentators as an unattractive encumbrance on Protagoras’ measure doctrine, an opinion not altered by the supposed fact that it makes Plato’s refutation ‘work’. Of course, on our view, even were PT * T held by Protagoras, it would not obviously improve Plato’s argument as developed by Burnyeat. For this is developed by using the notion of truth-in-a-world, in (1*), and it remains unclear how to connect this notion to PT *, which is free of such relativizing. The T most obvious way would be to add (7) (p)(p is true ↔ p is true in someone’s world), but this just assumes an answer to the question we have been worrying about, namely, whether truth and truth-in-a-world can be related. It is doubtful that this can be done with (7), since PT is meant to hold of all worlds, whereas truths in worlds are private to

Created on 4 July 2005 at 14.34 hours page 181

Animadversions on Burnyeat’s Theaetetus

181

those worlds. So PT at least could not go for p in (7). In short, the terms of argument do not appear to allow for something that could qualify as a candidate for PT in (6b). Suppose we set aside this last problem and suppose further that the upgrading envisioned in (1*) yields a version of the refutative argument that works ‘logically’. Is there any reason to think that this is Plato’s argument, i.e. that he actually attributes the strong measure doctrine to Protagoras? Burnyeat claims (i) that there is and (ii) that this in turn makes it likely that Protagoras himself maintains the strong measure doctrine. I think Burnyeat is correct on the latter point, (ii), but not the former. So I doubt that the strong measure doctrine is to be found in the Theaetetus. However, Burnyeat has an argument for his claim; so we had better take a look at it.

4. On finding the strong measure doctrine in the Theaetetus Burnyeat is emphatic about Protagoras’ commitment to the strong measure doctrine. Protagoras is held to claim not just that what appears (to someone) is true (for him) but also that what is true (for someone) appears (to him). And he is held to this claim, according to Burnyeat, because he holds that ‘man is the measure not only of what is, but also of what is not’.10 That is, PT is represented as the conjunction of (1) (x)(p)(x bel p → in x’s world it is true that p), which we have already introduced, and (8) (x)(p)(in x’s world it is true that p → x bel p). Together these give us the strong measure doctrine recorded in (1*). Our question is whether (8) is something that Plato attributes to Protagoras when he characterizes him as holding that man is the measure of what is not. Let us call this the negative measure doctrine. Burnyeat thinks that (8) captures the negative measure doctrine, indeed, that it is forthcoming from 10 Burnyeat, ‘Self-Refutation’, 178.

Created on 4 July 2005 at 14.34 hours page 182

182

Michael V. Wedin

(9) (x)(p)(it does not seem to x that p → it is not true for x that p), which I take to represent the doctrine. Now, of course, by contraposition (9) entails (10) (x)(p)(it is true for x that p → it seems to x that p), and (10) may be added to (1) to give us the strong measure doctrine, for it says that whatever is (or is true) for x is something that appears to x or is believed by x, etc. So it might appear that the negative measure doctrine justifies upgrading PT to a biconditional. However, the matter is not so straightforward. Proposition (9) has a narrow-scope and wide-scope reading, and heeding this distinction blocks the projected upgrading of PT. On the narrow-scope reading, x has a doxastic attitude towards p, namely, that of p’s seeming not to be the case. So the narrow-scope reading gives us (9a) (x)(p)(it seems to x that ¬ p → it is not true for x that p). Proposition (9a) would, I think, be favoured by Burnyeat, since he takes the negative measure doctrine to be a simple substitution instance of the positive measure doctrine, which we read as (1) above.11 Replacing ‘p’ with ‘not p’ simply converts (1) into (9a{) (x)(p)(it seems to x that ¬ p → it is true for x that ¬ p), and this can be taken as equivalent to (9a)—so long as ‘it is true for x that ¬ p’ and ‘it is not true for x that p’ are interchangeable. Because the rider, ‘for x’, is without doxastic force,12 the switch is unexceptionable. There is much to be said for the narrow-scope reading, (9a)/(9a{), because the very idea of someone’s being a measure requires that he display some doxastic attitude towards whatever he is the measure of. This is caught by (9a)/(9a{), and so the narrow-scope reading of the negative measure doctrine may well enjoy Protagorean credentials. On the wide-scope reading, the antecedent of the negative measure doctrine simply denies that something appears to someone. That is, on wide-scope the doctrine is (9b) (x)(p)( ¬ it seems to x that p → it is not true for x that p). 11 Bearing in mind n. 5’s liberal reading of ‘x bel p’ and reverting to the ‘for x’ idiom. 12 See n. 13.

Created on 4 July 2005 at 14.34 hours page 183

Animadversions on Burnyeat’s Theaetetus

183

As an interpretation of Protagoras’ measure doctrine, (9a) is considerably more attractive than (9b). And so it should be for Burnyeat, because, unlike (9a), (9b) is not a substitution instance of the positive measure doctrine, (1). The fact that (9a) in this way parallels (1) is precisely what enables it to serve as a negative measure. (9b), on the other hand, fails entirely to capture the notion of someone’s being a measure. For it is satisfied by something having no occurrent, or dispositional, doxastic attitude at all, but the measure doctrine just is a doctrine about agents’ doxastic attitudes. Unfortunately, however, (9a) is not su¶cient for the equivalence, (1*). For what (9a) yields is (9a*) (x)(p)(it is true for x that p → ¬ it seems to x that ¬ p), and this falls short. From the fact that it is not the case that it seems to Socrates that the cat is not on the mat, no one would conclude that it seems to him that the cat is on the mat. Socrates may have no attitude at all towards such a feline presence.13 So (9a*) does not enable passing from the truth of p to the appearance that p. Proposition (9b), on the other hand, is equivalent to (9b*) (x)(p)(it is true for x that p → it seems to x that p), and so it does pair with (1) to yield the strong measure doctrine. Unfortunately, as we have seen, (9b) is not a plausible reading of Protagoras’ negative measure doctrine.14 Thus, Burnyeat is left with a dilemma: the narrow-scope reading 13 One might try to circumvent this worry by insisting that ‘for x’ has quasidoxastic weight. Here for p to be true for x is for p to be held true by x. Then, the claim would be, (9a*) can be read smoothly as saying that it seems to x that p. But this makes a mockery of Protagoras’ doctrine, whose whole point is to insist that one can move from belief to truth, not from belief to being held true by the believer. (Here I am in complete agreement with Burnyeat, ‘Self-Refutation’, 180.) So ‘for x’ has an indexing function and, as such, for all we know there may be many things that are true for x without appearing (yet, or at all) to x (namely, those things that he is in a position to have appear to him). And, arguably, any tautology will be true for x regardless of what, if any, doxastic attitude he has towards it. 14 Someone might put in a plea for (9b) on the grounds that it contributes to the measure doctrine by eliminating from a subject’s world any truths of which the subject would fail to be the measure. The idea is that, without such support, the subject might after all be the measure of all things, i.e. of things that appear to others as well. This represents (9b), not as part of the measure doctrine proper, but as a kind of additional support. But it is support that commits Protagoras to the strong measure doctrine by an extra step, and this is not an outcome I would welcome. However, support of this kind is not needed. For Protagorean worlds are private, and this is enough to ensure that a subject is not the measure of all truths (which in

Created on 4 July 2005 at 14.34 hours page 184

184

Michael V. Wedin

of the negative measure doctrine may be plausible as an interpretation of Protagoras, but it is not su¶cient for the strong measure doctrine; the wide-scope reading does yield the strong measure doctrine, but it fails as a plausible interpretation of Protagoras.15 The dilemma gains additional bite from the fact that, if the strong measure doctrine goes, so goes Burnyeat’s final, and perhaps most potent, argument against Protagoras.

5. The exhaustion argument The final argument Burnyeat launches against Protagoras concerns the scope of the claim that all truths are relative. Now, we often allow a proposition to impose constraints on other propositions without being subject itself to the constraints, and such a proposition is often taken to express an absolute truth. Thus, for example, a thesis the present context must mean truths in his and others’ worlds) but only of those that populate, or could populate, his world. 15 M.-K. Lee, ‘The Secret Doctrine: Plato’s Defence of Protagoras in the Theaetetus’, Oxford Studies in Ancient Philosophy, 19 (2000), 47–86, finds the strong measure doctrine at work early in the Theaetetus. Immediately following introduction of the Protagorean measure doctrine, Plato embarks on an extended discussion of the socalled secret doctrine. According to Lee, this quasi-Heraclitean amalgam is meant to ‘support Protagoras’ claim that something is the case for one if and only if it is so to one’ (82). So, despite needing support, from the start Lee’s Protagoras champions the strong measure doctrine. And he is held to this on the basis of 152 b, which according to Lee establishes (8). Lee translates the crucial line: ‘Or shall we accept it from Protagoras that it’s cold for the one who feels cold, and not for the one who doesn’t?’ (56). The first alternative would give us (i): a seems cold to Socrates → a is cold for Socrates. The second alternative could be meant simply to deny this implication, in which case no support is forthcoming for (8) and so none for strong measure. But this reading, which Lee would no doubt refuse, is implausible. Still, the second alternative does not give us (8). For the two remaining readings, namely, (ii): a seems not cold to Socrates → a is not cold for Socrates, and (iii): ¬ (a seems cold to Socrates) → a is not cold for Socrates, fail for familiar reasons. Now (ii) is equivalent to (iia): a is cold for Socrates → ¬ (a seems not cold to Socrates), while (iii) transforms into (iiia): a is cold for Socrates → a seems cold to Socrates. So, as before, (iii) is su¶cient for (8) and thus supports strong measure, but it is implausible as an interpretation of Protagoras, or anyone for that matter. From the fact that it is not the case that something seems cold to Socrates it does not follow that that thing, or anything, is not cold for him. He may, for instance, be distracted by intense perusal of a hotly contested passage from Metaphysics Zeta. Reading (ii) enjoys Protagorean credentials, but it is not su¶cient for (8) and so does not give us the strong measure doctrine. In short, this case is subject to the same worry that undermined Burnyeat’s argument for finding the strong measure doctrine in the Theaetetus. I reserve comment on whether the secret doctrine must be read as supporting strong measure, as Lee proposes.

Created on 4 July 2005 at 14.34 hours page 185

Animadversions on Burnyeat’s Theaetetus

185

constraining propositions that express causal relations need not, itself, be subject to those constraints. The thesis, we might suppose, holds absolutely for such propositions and so gives us (part of) a theory of causal relations. According to Burnyeat, if PT is to be a significant general thesis, the same should hold of it. Thus, he maintains that Protagoras must hold PT as an absolute truth. Only in this way does his doctrine contain a theory of truth, more precisely, a theory relating belief and truth. Only as an absolute truth can PT provide a theory that holds for all beliefs and truths, as opposed to a theory that holds for all the beliefs and truths of Protagoras, say, or Gorgias, etc. Burnyeat is right about this and, in e·ect, we joined him on this point when we recommended above that PT be construed as a free-standing thesis. Burnyeat, however, takes this further and argues that it provides a final refutation of the measure doctrine. The strategy is to show that the project of relativizing truths to believers is undermined by what I shall call the exhaustion of belief. At the heart of this argument is a principle permitting translation of relative truths, i.e. instances of PT, into absolute truths. Thus, consider (10) (x)(x judges that a is F → ‘a is F’ is true for x), and (10a) (x)(‘a is F’ is true for x → ‘“a is F” is true for x’ is true). Taking (10) to represent an instance of PT, understood as in our (1), the idea is that the relative truth generated by PT is transformed into an absolute truth by the translation principle, (10a). It is an absolute truth that it is true for x that a is F. So while PT says that all truths generated from beliefs are relative, instances of PT turn out to yield absolute truths. In response, the Protagorean might try out the idea that PT ranges only over what might be called ‘firstorder’ propositions; hence, the more complex proposition in (10a)’s consequent is not allowed. Aside from invoking the notoriously difficult notion of a first-order proposition, by granting ‘second-order’ absolute truths the suggestion detracts from the doctrine’s appeal, which, according to Burnyeat, was to champion the thoroughgoing relative nature of truth, for to do this it must apply to all truths. Furthermore, (10a) is strictly a translation principle and, as such, is distinct from PT. It stands or falls on its own strength, and as

Created on 4 July 2005 at 14.34 hours page 186

186

Michael V. Wedin

a stand-alone principle it is not implausible. So this ‘Protagorean’ response is a non-starter. A more likely, and potentially more e·ective, response for the Protagorean would be to insist on reiterating the ‘for x’ rider and so on replacing (10a) with (10b) (x)(‘a is F’ is true for x → ‘“a is F” is true for x’ is true for x). The grounds for (10b) are straightforward: if truth can be iterated, so can truth for x. This seems entirely benign. Since iteration is recommended for the translation principle, PT itself remains an absolute truth. According to Burnyeat, however, the translation principle dooms Protagoras and company. For iteration of beliefs is not benign, yet the Protagorean faces just such iteration. He is commited to (10c) (x)(‘a is F’ is true for x → x judges that ‘a is F’), because he is committed to the claim that every truth is a truth for some judger. There are, in short, no unbelieved truths. Now there may be no problem with (10c), whose target proposition is easily borne in mind. But to avoid the e·ect of the translation principle, i.e. to avoid commitment to the absolute truth, the Protagorean must go on to countenance the likes of (10c{) (x)(‘“a is F” is true for x’ is true for x → x judges that ‘“a is F” is true for x’). Here the target proposition is more complex but, again, nothing that exceeds the Protagorean’s grasp. Very quickly, however, the iteration becomes intolerably expansive. As Burnyeat puts it, ‘he is not so clever that there is no limit to the complexity of the propositions he can understand and so judge to be true . . . At some point . . . Protagoras must stop and take a stand.’16 And when he does, he is committed to an absolute truth and can no longer maintain that all truth is relative. In short, at some point he su·ers doxastic exhaustion—he simply cannot muster the beliefs needed to keep pace with the endless parade of iterated propositions that are ‘true-for’ him. I wish to make two points about the exhaustion argument, as I 16 Burnyeat, ‘Self-Refutation’, 194.

Created on 4 July 2005 at 14.34 hours page 187

Animadversions on Burnyeat’s Theaetetus

187

have called it, both of which undercut its strength. First, commitment to (10c) requires the strong measure doctrine. As such, rather than (10) (x)(x judges that a is F → ‘a is F’ is true for x), what the argument calls for is the biconditional (10*) (x)(x judges that a is F ↔ ‘a is F’ is true for x). Unfortunately, as we have already seen, there is little reason to attribute (10*) either to Protagoras or to Plato on behalf of Protagoras. Second, even were the equivalence in (10*) adopted, the force of the exhaustion argument is unclear. One might, for example, suppose that the Protagorean is unable to sustain an infinite regress of beliefs to match the unending parade of truths generated by infinite iteration of the ‘for-x’ rider. Hence, contrary to (1*), there is not an equivalence between beliefs and truths, and so the strong measure doctrine is to be rejected. However, this worry will not faze the Protagorean precisely because, we are supposing for the moment, he holds the strong measure doctrine. Should he expire or should he lose capacity to track a sequence of beliefs, it does not follow that there is some true proposition that he does not believe. On the contrary, (1*) guarantees that there is no such truth. So the equivalence holds up. Burnyeat wisely eschews this option. Rather, for Burnyeat, the Protagorean’s evasive action requires him to grasp given propositions whose complexity outstrips his doxastic capacities (and here there need be no talk of infinite sequences of any kind). Let us assume that this is so. Still, the force of the argument is undercut by the fact that it is entirely pragmatic. Nothing in the logic of the strong measure doctrine requires the cessation of doxastic iteration. Rather, the argument appeals to supposed limitations on the complexity, or perhaps number, of propositions that can be grasped by a doxastic agent. Because they are entirely pragmatic, these considerations do not touch the logic of the view. By the same token, it is not the doctrine itself that is faulted but human abilities that are required by it. This will be disappointing to those who had hoped to find internal fault with the doctrine itself. In short, the exhaustion argument leaves in place the core of Protagoras’ doctrine even when that is read as the strong measure doctrine.

Created on 4 July 2005 at 14.34 hours page 188

188

Michael V. Wedin 6. On salvaging the exquisite argument

Three sections back I argued that the exquisite argument fails on its best reading, which I take to be that of Burnyeat. Simply put, the trouble was that believing that PT is false does not entail living in a world where belief is not su¶cient for truth. As long as Protagoras’ opponent has the negation of PT as the object of his belief, the self-refutation argument does not go through. For if this belief is to be coherent, it must be possible that it is true in the opponent’s world that something is the case in some other world. But this is impossible because the terms of argument call for Protagorean worlds and these are private. Is there a way out of this di¶culty? Well, suppose that right from the start the opponent believes, not the negation of PT, but something that keeps everything within the bounds of his own private world. In this case, there would be no worry about quantifiers in his world binding variables whose values are objects in another world. Cross-world reference would no longer be an issue. It might be thought that something as simple as (11) (∃x)(∃p)(x bel p ∧ ¬ (p is true)) can accomplish this. After all, (11) says that there is someone who believes something that is not true; so, if (11) is true, PT is surely false. Furthermore, nothing in (11) requires x to reach across Protagorean worlds. None the less, (11) is unhelpful because its doxastic content is exhausted by the claim that someone believes something, i.e. by detaching the lead conjunct of (11): (∃x)(∃p)(x bel p). So all that follows from (11) is that someone holds a false belief, and, by itself, this contributes nothing to the self-refutation of PT. To get the project of self-refutation under way, the target of the opponent’s belief must at least be a proposition that relates belief and truth but that does not take him out of his own world. One might suppose the following would work: (12) (∃x)(∃p)(x bel p ∧ x bel ¬ (p is true)). Like (11), proposition (12) does not ask a believer to delve into worlds distinct from his private domain. Moreover, (12) yields, by PT, a contradiction, and this sounds more in the nature of a

Created on 4 July 2005 at 14.34 hours page 189

Animadversions on Burnyeat’s Theaetetus

189

self-refutation. But (12) itself has the opponent himself believing a contradiction, and this can hardly serve as the doxastic assumption that pairs with PT to generate a refutation of PT. Moreover, (12) runs foul of the plausible principle that if someone believes p, then they believe that p is true. The trouble with (12) lies with the second occurence of ‘bel’. It needs to be moved outside the parentheses currently enclosing it. This gives us (13) (∃x)(∃p)(x bel (x bel p ∧ ¬ (p is true)), which might appear to be a better candidate for our salvage operation. Still, by locating the quantifier outside the doxastic context, (13) simply declares that there is a certain proposition such that x believes that he believes it and that it is not true. Some might object that this is not a plausible candidate for the doxastic assumption that leads to PT’s downfall because, the objector insists, the opponent himself ought to believe that there is such a proposition. This worry is accommodated by moving the relevant quantifier within the scope of the bel operator: (13{) (∃x)(x bel (∃p)(x bel p ∧ ¬ (p is true))). Proposition (13{) says that there is someone who believes that there is some proposition such that he believes it, and yet it is false. That is, he believes that at least one of his beliefs is false. Notice that (13{) is to be distinguished from something like (13]) (∃x)(x bel (x bel p ∧ ¬ (pn is true))), n which has x believing a particular proposition, pn , and which therefore clashes with the principle that believing a given proposition entails believing that the proposition is true. Unlike (13]), proposition (13{) is a plausible doxastic assumption to hand Protagoras’ opponent. Our question, then, is whether (13{) can be paired with PT to salvage the exquisite argument, i.e. whether it can produce a selfrefutation of PT. Putting (13{) into Burnyeat’s idiom, we get (13*) (∃x)(x bel (∃p)(x bel p ∧ ¬ (in x’s world it is true that p))). The idea behind (13*) is that someone believes there is a proposition such that he believes it and yet it is not true. The fact that the

Created on 4 July 2005 at 14.34 hours page 190

190

Michael V. Wedin

proposition is not true in his world means that (13*) does not require the believer to have access to worlds other than his own. Thus, the salvaged version of the exquisite argument concludes with (3a*) (∃x)(∃p)((x bel p ∧ ¬ (in x’s world it is true that p)) is true in x’s world). We may take (13*) to assert that belief does not entail truth and (3a*) to assert that this is a truth in the believer’s world. On the basis of (13*), then, we may conclude, in the style of Burnyeat, that not everyone lives in a world in which belief does entail truth. Note that we get (3a*) by adding an identity to (3b), the proper conclusion of the exquisite argument: (3b*) (∃y)((∃x)(∃p)(x bel p ∧ ¬ (in x’s world it is true that p)) is true in y’s world ∧ x = y). Proposition (3b*) simply decrees that the world in which p is believed and is false is the world determined by the believer of p. Since (3b*) yields (3b) by simplification, (3b*) entails the negation of PT. Thus, armed with (13*)’s reasonably sane doxastic assumption, we can count PT as false after all. However, as before, there is more to be said. For it is clear from the second section of this paper that the negation of PT is one thing that the opponent cannot claim to believe. If he does so, he is committed to quantifying across private worlds and, with this, the argument is doomed. So Protagoras’ foe, intent on salvaging the exquisite argument, maintains that he is in a doxastic situation of the kind described by (13*) and that to be in this situation, to have a belief of this sort, is not to believe the negation of PT itself. Rather, he believes a proposition that entails the negation of PT. In this way he fends o· the charge of meddling in worlds beyond his ken. But in so doing he does not as much salvage the exquisite argument as shift its force. For his brief must include the corrective point that Plato misdescribes his own argument. The detractors of Protagoras should not be represented as believing the negation of PT but as believing a stronger proposition that entails the negation of PT. Morever, they must acknowledge that despite this entailment, the terms of argumentation prevent them from holding the negation of PT as an object of belief. In short, they prove a result they cannot

Created on 4 July 2005 at 14.34 hours page 191

Animadversions on Burnyeat’s Theaetetus

191

believe.17 Protagoras would, I think, have been more than a little pleased. University of California, Davis

B I B L I O GR A P HY Burnyeat, M. F., ‘Protagoras and Self-Refutation in Plato’s Theaetetus’ [‘Self-Refutation’], Philosophical Review, 85 (1976), 172–95, repr. in S. Everson (ed.), Companions to Ancient Thought, i. Epistemology (Cambridge, 1990), 39–59. The Theaetetus of Plato (Indianapolis, 1990). Fine, G., ‘Protagorean Relativisms’ [‘Relativisms’], in J. Cleary and W. Wians (eds.), Proceedings of the Boston Area Colloquium in Ancient Philosophy, 10 (Lanham, Md., 1994), 211–43. ‘Relativism and Self-Refutation in Plato’s Theaetetus: Plato, Protagoras, and Burnyeat’ [‘Self-Refutation’], in J. Gentzler (ed.), Method in Ancient Philosophy (Oxford, 1998), 138–63. ‘Plato’s Refutation of Protagoras in the Theaetetus’ [‘Plato’], Apeiron, 32 (1998), 201–34. Lee, M.-K., ‘The Secret Doctrine: Plato’s Defence of Protagoras in the Theaetetus’, Oxford Studies in Ancient Philosophy, 19 (2000), 47–86. Matthews, G., ‘Burnyeat on Plato on Protagoras’ (unpublished). Waterlow, S., ‘Protagoras and Inconsistency’, Archiv f•ur Geschichte der Philosophie, 59 (1977), 19–36. Wedin, M. V., ‘A Curious Turn in Metaphysics Gamma: Protagoras and Strong Denial of the Principle of Non-Contradiction’, Archiv f•ur Geschichte der Philosophie, 85 (2003), 107–30. ‘On the Use and Abuse of Non-Contradiction: Aristotle’s Critique of Protagoras and Heraclitus in Metaphysics Gamma 5’, Oxford Studies in Ancient Philosophy, 26 (2004), 213–39. 17 Of course, if belief is closed under entailment, neither can they believe the stronger proposition contained in (13*).

Created on 4 July 2005 at 14.34 hours page 192

page 192 is blank

Created on 4 July 2005 at 14.34 hours page 193

AUTHENTI CATI NG ARI S TOTLE’ S PROTREPTI CUS D. S. HUTCHINSON @ MONTE RANSOME JOHNSON

 the gold rush of the 1860s, scholars went prospecting for fragments of the lost works of Aristotle, and in 1869 a crucial discovery was announced: Iamblichus of Chalcis, it seems, had quoted from Aristotle’s Protrepticus, using it and other books as sources when compiling his own Protrepticus. Scholars rushed in to mine that later text. ã D. S. Hutchinson @ Monte Ransome Johnson 2005 Finding the patterns in this textual labyrinth of Iamblichus has been a collaborative e·ort, in which many scholars in many places have given us assistance, criticism, and encouragement; to all these people we acknowledge their help and express our thanks. First in order of generation is our friend and colleague Brad Inwood, with whom we read chapters 5–12 of Iamblichus’ Protrepticus in 1999/2000, together with other members of the Greek philosophy reading group hosted by the University of Toronto’s Collaborative Program in Ancient and Medieval Philosophy: Donald Smith, Max Arnott, Joe Novak, David Hitchcock, Daniel de Montmollin, and Rob Butler. An earlier version of the reasoning in this paper was presented to the B Club at Cambridge University in January 2004, and we gratefully acknowledge the encouragement and guidance we received there from Nick Denyer, David Sedley, Geo·rey Lloyd, Malcolm Schofield, Robert Wardy, Catherine Osborne, Alex Long, and Mantas Adomenas. In Tuscany in June, we were hosted by our new friends Alessandro Linguiti and Francesco Ademollo at a seminar in the Department of Philosophy at the University of Florence, where we greatly benefited from their comments and those of Glenn Most and Walter Leszl. Antonio Carlini discussed Protrepticus matters with D.S.H. in Pisa, and introduced him to the Biblioteca Laurenziana in Florence, whose Director Franca Arduini facilitated research on Laur 86,3, the unique independent witness to the Protrepticus of Iamblichus. At the University of Victoria in October 2004, we presented a paper to a conference on the Aristotelian tradition, where useful scepticism and support came from many, especially Taneli Kukkonen, Miira Tuominen, Richard Bod‹eus, • Marguerite Deslauriers, and Victor Caston. Several scholars have given us the benefit of careful study of draft materials, including Phillip DeLacy, Jonathan Barnes, Stephen Menn, Rob Bradizza, Chris Young, Margaret Cameron, David Gallop, John Bussanich, Michele Alessandrelli, and four anonymous reviewers for Hackett Publishing Company. Support for M.R.J.’s research came from the Killam Foundation by way of a postdoctoral fellowship in philosophy at the University of British Columbia, and from Green College, which provided direct support as well as being a perfect place to bring us together to share ideas and friendship.

Created on 4 July 2005 at 14.34 hours page 194

194

D. S. Hutchinson @ Monte Ransome Johnson

The 136 years since then have seen dramatic rises and spectacular falls in the stock of Aristotle’s Protrepticus. The ore has proved to be more di¶cult to extract than expected, and currently the work languishes amidst widespread scholarly scepticism and sharp disagreement; indeed, some scholars consider the nuggets attributed to Aristotle’s Protrepticus to be nothing but fool’s gold. This is because no clear resolution has yet been found, despite several periods of intensive research and several ambitious publications, to a number of key hermeneutical questions: Where did Iamblichus start and stop using Aristotle text(s) as a source? Did Iamblichus simply quote the text of Aristotle, or did he take liberties? Did Iamblichus work from memory, or did he make an accurate copy? Did Iamblichus work sequentially, earlier passages quoted earlier, and later ones later? Did Iamblichus use the Protrepticus only, or other works of Aristotle as well? In this paper we venture answers to these questions, based on the relevant comparative structural analysis: we study Iamblichus’ method of quoting Plato, someone for whose surviving books we have excellent manuscript evidence. Our guiding initial hypothesis is that Iamblichus employed the same method of constructing his book when he was quoting Aristotle as when he was quoting Plato. In Section 1 we review the status of the question; in Section 2 we present the results of our study of the Plato quotations; and in Section 3 we apply these results to the chapters in which Iamblichus used Aristotle as a source, and find that the above initial hypothesis is amply confirmed. In Section 4 we summarize our conclusions. The method of literary construction used by Iamblichus for both the Plato and the Aristotle chapters is to assemble into each chapter one or a few pure blocks of quotation in a natural sequence, marking them o· from each other by thin comments of his own. The blocks of quotation are solid and pure, in that no passages are removed from them and no extraneous passages are added to them, though they are sometimes carefully modified. The sequence of quotation blocks is natural, in that Iamblichus never returns to a work he has earlier been quoting from, nor does he ever return to any earlier part of a work he has been quoting from. When we deconstruct chapters 6–12 of Iamblichus’ book in the light of this newly discovered construction method, a number of powerful conclusions emerge: the seven chapters contain fourteen separate blocks of quotation, each apparently a pure block of quo-

Created on 4 July 2005 at 14.34 hours page 195

Authenticating Aristotle’s Protrepticus

195

tation from Aristotle, marked o· by thin comments by Iamblichus; the blocks of quotation may or may not have been carefully modified, but Aristotle’s line of thought has apparently not been modified at all; several internal references and connected themes link the material in the seven chapters together and guarantee that they all come from the same work of Aristotle; the protreptic contents of several blocks guarantee that this one work was his Protrepticus; and, finally, the natural sequence of quotations mirrors and preserves the sequence of passages in Aristotle’s original work. Our scrutiny of the Plato quotations will also show that Iamblichus must have been making careful copies from good manuscripts, so his quotations of the Protrepticus, when used with appropriate caution, have an evidential status which is equivalent to that of a manuscript of Aristotle’s lost work. Newly authenticating these quotations is as good as finding a trustworthy new manuscript. Yet we must at the same time recognize that this newly trustworthy witness is also an incomplete one, whose testimony contains lacunae: when Iamblichus selected fourteen blocks to quote from, he skipped over the material in the thirteen gaps between the blocks, as well as the beginning of the book and (probably) the end. It is di¶cult to exaggerate the significance for Aristotle studies of what is at stake in these claims. Our knowledge of virtually every theme and topic in Aristotle’s thought is enhanced when we confidently take into consideration the Protrepticus evidence. For example, every recent study of Aristotle’s ergon argument that has not taken account of the Protrepticus evidence in chapter 9 rests on a defective, or at least incomplete, basis of evidence,1 and likewise for the discussions of pleasure that have not studied the subtle argument at the end of chapter 11. This lost work, even in this partial survival, is rich in stirring expressions of Aristotelian insight into central questions concerning reason and the human good. Unlike Plato, all of whose published dialogues survive, Aristotle had the misfortune to have all of his published works perish. And then, when these large blocks of quotation were rediscovered in 1869 by Bywater, the doubts and confusions of scholars condemned the work to a second death. But this second loss is reversible, and much will be gained when these newly authenticated fragments 1 This includes the analysis of the other two versions of the argument, in Eudemian Ethics and Nicomachean Ethics, in ch. 3 of D. S. Hutchinson, The Virtues of Aristotle (London, 1986).

Created on 4 July 2005 at 14.34 hours page 196

196

D. S. Hutchinson @ Monte Ransome Johnson

bring the twice-lost Protrepticus back into the classroom, where future generations of students and scholars are likely to continue to sharpen our focus and deepen our appreciation of the totality of Aristotle’s philosophy. This is probably our only chance to have any real access to any of Aristotle’s published works, barring future papyrus discoveries of substantial extent, and hence our only chance to assess his thought through both of his main channels of written expression: specialized treatises written originally for an inner circle of listeners, and lively works of philosophy written for a wider external or ‘exoteric’ public of readers. Readers whose primary interest is in Aristotle’s Protrepticus will want to read subsections 3.1 to 3.8, if only because these sections include new translations of all of the Aristotle material that might be recovered from this part of Iamblichus’ work; but they should also look at subsection 2.8, which summarizes the Plato chapters. Those sceptical about our reasoning will want to scrutinize the rest of Section 2 as well. Both sets of readers may want to orient themselves first with Section 1, a brief historical overview of the status of the scholarly question, and finish with Section 4, which summarizes the results of the whole study, draws general conclusions, and takes fresh stock of the status of the question. 1. Status of the question, p. 196; 2. The Plato chapters (13–19), p. 203; 2.1. Preliminary remarks, p. 203; 2.2. Chapter 14: quotations from Theaetetus, p. 206; 2.3. Chapters 15 and 16: quotations from Republic 7, p. 213; 2.4. Chapter 17: quotations from Gorgias, p. 218; 2.5. Chapter 18: quotations from Gorgias, p. 220; 2.6. Chapter 13: quotations from Phaedo, p. 224; 2.7. Chapter 19: quotations from Gorgias, Menexenus, Laws 2, p. 235; 2.8. Summary and conclusions about the Plato chapters, p. 240; 3. The Aristotle chapters (6–12), p. 242; 3.1. Preliminary remarks, p. 242; 3.2. Chapter 7, p. 244; 3.3. Chapter 8, p. 252; 3.4. Chapter 9, p. 258; 3.5. Chapter 10, p. 262; 3.6. Chapter 11, p. 265; 3.7. Chapter 6, p. 269; 3.8. Chapter 12, p. 278; 4. Summary, conclusions, and new status of the question, p. 281.

1. Status of the question In the ancient world, the Protrepticus of Aristotle was probably his most famous work, and perhaps one of the most famous works of

Created on 4 July 2005 at 14.34 hours page 197

Authenticating Aristotle’s Protrepticus

197

philosophy by any author.2 We are certain that Aristotle wrote a work with that title; the title appears in all three lists of Aristotle’s works that have survived from antiquity.3 It is also clear from reports naming the book that in it Aristotle set out arguments with this conclusion: φιλοσοφητον, ‘one should do philosophy’.4 Beyond that, nothing is certain. For example, how long was the Protrepticus? The lists disagree as to how many books it had: two of the lists say one, another three; perhaps it was sometimes written on two rolls. Was it a dialogue? It is listed among other dialogues, but none of the testimony or fragments tells us that it was, nor does any solid evidence seem to preserve any actual dialogue. The bulk of the evidence now available to us is lodged in the pages of Iamblichus, who seems to have quoted from the work at times, but without giving any indication of where and when he was quoting, or from what work; to make matters worse, he took various liberties with the works of Plato that he also exploited. These problems, compounded with confusions of their own making, have made scholars quite 2 We have a number of well-known reports about the book being read in antiquity. ‘Zeno said that Crates, while sitting in a shoemaker’s workshop, read the Protrepticus’ aloud to the shoemaker (Stob. 4. 32. 25). Cicero evidently looked to it in writing his own exhortation to philosophy: ‘Nor . . . are the arguments unknown which Cicero used in his Hortensius, which he modelled on a protreptic pattern’ (Hist. Aug. ii. 97. 20–2 Hohl). Cicero’s book had, in its turn, an influence on Augustine, who says in his Confessions: ‘I had already come across a book by a certain Cicero . . . That book of his contains an exhortation to study philosophy and is entitled Hortensius. The book changed my feelings. . . . I longed for the immortality of wisdom with an incredible ardour in my heart. . . . “Love of wisdom” is the meaning of the Greek word philosophia; this book kindled my love for it’ (3. 4. 7–8, trans. H. Chadwick (Oxford, 1991)). Another passage shows that Augustine was aware that some of the content of Cicero’s book was borrowed from Aristotle (see below, pp. 256–7). Boethius also exploited an idea in Aristotle’s work, either directly or indirectly (via Cicero’s Hortensius), at Cons. 3. 8 (see below, p. 255). 3 ‘Προτρεπτικς α’, D.L. 5. 22, title 12; ‘Προτρεπτικν α’, Vita Hesychii, title 14; ‘His book in which he exhorts to philosophy, three books. In Greek entitled προτρεπτικς φιλοσοφας’, notes Ptolemy-el-Garib, title 1–2, trans. in I. During, Aristotle in the • Ancient Biographical Tradition (G•oteborg, 1957), 222. 4 Alexander’s is the earliest report, and is possibly the source on which the subsequent reports are based: ‘if someone were to deny that you should do philosophy, then, since “to do philosophy” means not only to investigate this very thing—whether you should do philosophy or not (as he [sc. Aristotle] says in the Protrepticus)—but also to pursue philosophical contemplation, we will utterly refute the proposal by demonstrating that each of these is the proper business of human beings’ (Alex. Aphr. In Top. 149. 9–15 Wallies; cf. Suda, s.v. φιλοσοφεν, Φ 414 Adler). The other reports are: David, Prolegomena to Philosophy, 9. 2–12 Busse; Elias, Prolegomena to Philosophy, 3. 17–23 Busse; Olymp. In Alc. 144 Creuzer; Anonymous scholion to Aristotle’s Prior Analytics, found in the margin of fo. 263r of a manuscript in the Biblioth›eque Nationale in Paris, Cod. Par. gr. 2064.

Created on 4 July 2005 at 14.34 hours page 198

198

D. S. Hutchinson @ Monte Ransome Johnson

uncertain whether any particular passage in Iamblichus is a reliably accurate version of Aristotle’s work. Because of such uncertainties, the Protrepticus has fallen into neglect and, because it is the most important of the exoteric works (and its remains the most extensively preserved), its fall has taken with it most of the other exoteric works of Aristotle. But from the 1860s up to the 1950s there was much enthusiasm for the project of compiling the fragments of the exoteric works of Aristotle, and possibly even recovering and reconstructing them. A pioneer of this work was V. Rose, at first in his collection Aristoteles Pseudepigraphus (1863), and then in his collection of fragments for the fifth volume of the Berlin edition of Aristotle (1870).5 In 1869, when Ingram Bywater, in the second volume of the Journal of Philology,6 argued that large parts of Iamblichus’ Protrepticus contained extracts from Aristotle’s lost work, scholars became interested in that work specifically. By 1886 Rose had published eleven groups of fragments over seventeen pages of his Teubner edition of the fragments.7 This was the edition that Werner Jaeger used in expounding his hugely influential interpretation Aristoteles: Grundlegung einer Geschichte seiner Entwicklung,8 based in large part on a reading of the Protrepticus. Inspired by Jaeger’s highly productive use of this material, R. Walzer produced a collection of the fragments of Aristotle’s dialogues, adding many new fragments and documenting many persuasive parallels to the extant works of Aristotle.9 On the basis of Walzer’s edition, W. D. Ross published first a translation and then an edition of the Protrepticus (in his Select Fragments, vol. xii of the Oxford Translation of Aristotle).10 Things seemed to be going well for the Protrepticus and the fragments of Aristotle in the first half of the twentieth century. The 5 V. Rose (ed.), Aristoteles Pseudepigraphus (Leipzig, 1863); id., ‘Aristotelis qui ferebantur librorum fragmenta’, in Opera omnia Aristotelis, v (Berlin, 1870), 1463– 589. 6 I. Bywater, ‘On a Lost Dialogue of Aristotle’ [‘Lost Dialogue’], Journal of Philology, 2 (1869), 55–69. 7 V. Rose (ed.), Aristotelis qui ferebantur librorum fragmenta (Leipzig, 1886; repr. 1966). 8 W. Jaeger, Aristoteles: Grundlegung einer Geschichte seiner Entwicklung [Entwicklung] (Berlin, 1923); all references are to the English translation by R. Robinson (Oxford, 1948). 9 R. Walzer (ed.), Aristotelis dialogorum fragmenta (Florence, 1934). 10 W. D. Ross (trans.), Select Fragments = The Works of Aristotle, xii (Oxford, 1952); W. D. Ross (ed.), Aristotelis fragmenta selecta (Oxford, 1955).

Created on 4 July 2005 at 14.34 hours page 199

Authenticating Aristotle’s Protrepticus

199

acme was in 1957 at the Symposium Aristotelicum at Oxford devoted to discussing the connections between Aristotle’s early works and the Academy.11 Some members of the Symposium decided to collect all the evidence and fragments of Aristotle’s dialogues in order to produce completely new editions. Ingemar During, who • had already published articles on the Protrepticus,12 was charged with producing a comprehensive edition. Although During was • a staunch anti-developmentalist, he nevertheless saw tremendous value in what could be recovered, particularly in the Iamblichean material. It was at this point, mid-twentieth-century, that things went badly wrong for the e·ort to recover the Protrepticus. Sceptics had always been uneasy with Bywater’s use of the lost and reconstructed Hortensius of Cicero as evidence for the lost Protrepticus of Aristotle; and the material in Iamblichus seemed to be infected at times with so many Neoplatonic intrusions that it might all be just a production of his. W. G. Rabinowitz published a monograph,13 the first part of a projected two-part series, that aimed to expose as fool’s gold all of the flecks that had been panned by Walzer, Ross, and others, not to mention the earlier work of Bignone,14 who feverishly compiled and commented on what he perceived to be dozens of fragments from Aristotle’s Protrepticus, fragments whose links to solid reports that explicitly mention Aristotle and his work are sometimes rather tenuous. The publication, soon afterwards, of During’s Aristotle’s Pro• trepticus: An Attempt at Reconstruction15 was aimed in part at refuting Rabinowitz, but was ultimately counter-productive in the effort to answer growing scepticism. There are many reasons for this, but foremost was the fact that During, assuming that Iamblichus • 11 I. During and G. E. L. Owen (eds.), Aristotle and Plato in the Mid-Fourth • Century: Papers of the Symposium Aristotelicum Held at Oxford in August, 1957 [Mid-Fourth Century] (G•oteborg, 1960). 12 I. During, ‘Problems in Aristotle’s Protrepticus’ [‘Problems’], Eranos, 52 (1954), • 139–71; id., ‘Aristotle in the Protrepticus’, in Autour d’Aristote: recueil d’‹etudes de philosophie ancienne et m‹edi‹evale o·ert a› A. Mansion (Louvain, 1955), 81–97. 13 W. G. Rabinowitz, Aristotle’s Protrepticus and the Sources of its Reconstruction, vol. i [Sources] (Berkeley, 1957). 14 E. Bignone, L’Aristotele perduto e la formazione filosofica di Epicuro [Perduto] (2 vols.; Florence, 1936); id., ‘Nuove testimonianze e frammenti del Protrettico di Aristotele’, Rivista di filologia e d’istruzione classica, 64 (1936), 225–37. 15 I. During, Aristotle’s Protrepticus: An Attempt at Reconstruction [Attempt] • (Studia Graeca et Latina Gothoburgensia, 12; G•oteborg, 1961).

Created on 4 July 2005 at 14.34 hours page 200

200

D. S. Hutchinson @ Monte Ransome Johnson

had rearranged the material for his own purposes, himself further fragmented the text (by dividing it and labelling 110 fragments of unequal length), and arbitrarily reordered the fragments according to his own conception of its philosophical point and rhetorical genre. This rendered the work essentially unusable. Despite this, however, it is still being translated into several modern languages.16 During’s edition is much of the reason why the Protrepticus has not • been greatly used by Aristotle scholars in the ensuing decades, with some notable exceptions.17 The Munich dissertation of Gerhart Schneeweiss,18 which followed During’s method of reorganization • but went further by interspersing passages from extant works of Aristotle, was no more successful. Rabinowitz’s negative thesis was ruinously criticized by many eminent reviewers,19 so much so, apparently, that he never carried out his intention20 to assay the rest of the fragments attributed to Aristotle, which included almost all of what was contained in Iam16 First into English, by During himself, and also by A. H. Chroust, Aristotle: • Protrepticus. A Reconstruction (Notre Dame, 1964), and then by J. Barnes and G. Lawrence in a section entitled ‘Fragments: Dialogues’, in J. Barnes (ed.), The Complete Works of Aristotle: The Revised Oxford Translation [Revised Aristotle] (2 vols.; Princeton, 1984), ii. 2403–17. During also translated his rearranged Protrepticus • into German, in Der Protreptikos des Aristoteles (Frankfurt a.M., 1969); and it was translated into Italian by E. Berti, in Aristotele: Protreptico. Esortazione alla filosofia (Turin, 2000), into French by J. Follon, in Aristote: Invitation a› la philosophie (Paris, 2000), and into Romanian by B. Minca and C. Partenie, in Aristotel: Protrepticul (Bucharest, 2005). However, the most recent Italian translation, by M. Casaglia, rejects the rearrangement of During and returns to Ross’s arrangement, in Aristotele: • Protreptico. Esortazione alla filosofia (Florence, 2001). 17 G. E. R. Lloyd, Aristotle: The Growth and Structure of his Thought (Cambridge, 1968), 33–41; D. Sedley, ‘Is Aristotle’s Teleology Anthropocentric?’ [‘Teleology Anthropocentric?’], Phronesis, 36 (1991), 179–96; S. Menn, ‘The Origins of Aristotle’s Concept of νργεια: νργεια and δναµις’ [‘Origins’], Ancient Philosophy, 14 (1994), 73–114; A. W. Nightingale, Spectacles of Truth in Classical Greek Philosophy: Theoria in its Natural Context (Cambridge, 2004), ch. 5; M. Johnson, Aristotle on Teleology (Oxford, 2005). 18 G. Schneeweiss, ‘Der Protreptikos des Aristoteles’ [‘Der Protreptikos’] (diss. Munich, 1966). We have not yet been able to consult the freshly published translation of the Protrepticus by Schneeweiss (2005). 19 Including: W. Spoerri, Gnomon, 32 (1960) 18–25; P. Moraux, L’Antiquit‹e classique, 27 (1958), 460–4; S. Mansion, Revue philosophique de Louvain, 56 (1958), 316–20; D. Furley, Journal of Hellenic Studies, 89 (1959), 178–80; D. J. Allan, Classical Review, ns 9 (1959), 124–7; P. Wilpert, Archiv f•ur Geschichte der Philosophie, 42 (1960), 101–6; E. Berti, La filosofia del ‘primo’ Aristotele [‘Primo’ Aristotele] (Padua, 1962), 2nd edn. (Milan, 1997), 397–8, 407–8, 411–13. 20 ‘I therefore propose to review in chapter ii all the testimonia and the first five fragments listed in Walzer’s collection, as the first step of a study which will eventually bring all the fragments under review’ (Rabinowitz, Sources, 22).

Created on 4 July 2005 at 14.34 hours page 201

Authenticating Aristotle’s Protrepticus

201

blichus’ Protrepticus. In fact, Rabinowitz studied only a single set of Iamblichus’ excerpts of Plato (but not the relevant one), and he studied only one21 of the alleged fragments of Aristotle, and so his work is of limited importance to the hypothesis of this paper and to the passages on which we concentrate our argument: Iamblichus’ excerpts of Plato and Aristotle. In 1965 Helmut Flashar made some progress towards an understanding of Iamblichus’ method of quotation in an important article,22 in which he argued that the key to isolating the Aristotle material from its Iamblichean encasement was to study the way that Iamblichus had used Plato in the same work. Since we have a solid and independent manuscript tradition for the Plato material, we can judge Iamblichus’ fidelity to his source, and learn about his method of quotation, by studying how he quoted or modified Plato’s dialogues. Remarkably, no one, not even During or Rabi• nowitz, had done this systematically. Even more remarkably, Flashar himself did not do it completely; he gave only an outline account23 of the Plato excerpts in his article, and based his comments on the ‘exhaustive comparison’ that had been made in an essay written for a seminar by a graduate student of his named A. Schnaufer. Our present work is the first complete study of the Plato excerpts in Iamblichus, and the first complete analysis of the Aristotle section in the light of such a study. Our more complete study yields more solid results. Before presenting the results of this study, a word about what we do not touch on in this article. We shall not be expounding or evaluating any of the alleged fragments other than the material preserved in these seven chapters (6–12) of the Protrepticus of Iamblichus. We believe that this material can be recovered and authenticated independently of the other evidence which has sometimes been called into play, including the preceding chapter (5) of Iamblichus’ Protrepticus.24 Also excluded is the evidence derived 21 And he studied this single passage with arguments that were less than cogent; see below, pp. 274–5. 22 H. Flashar, ‘Platon und Aristoteles im Protrepticus des Jamblichos’ [‘Platon und Aristoteles’], Archiv f•ur Geschichte der Philosophie, 47 (1965), 53–79, repr. in P. Moraux (ed.), Fr•uhschriften des Aristoteles (Darmstadt, 1975), 247–69, and in Flashar, Eidola: Ausgew•ahlte Kleine Schriften (Amsterdam, 1989), 297–323. 23 ‘Auch hier kann eine solche Untersuchung nicht in allen Einzelheiten ausgebreitet werden’ [‘nor can such an investigation be laid out in all its details here’] (‘Platon und Aristoteles’, 57). 24 Whether it contains any material that can reliably be attributed to Aristotle

Created on 4 July 2005 at 14.34 hours page 202

202

D. S. Hutchinson @ Monte Ransome Johnson

from chapters 26–7 of the next work in the Pythagorean sequence of Iamblichus, the De communi mathematica scientia.25 Hard-won experience has taught us that these other two stretches of text were composed by Iamblichus using significantly di·erent techniques of literary construction. Nor do we here deal with any of the evidence not supplied by Iamblichus: the various papyrus fragments that have been attributed to the Protrepticus, the scattered evidence that has been variously used to reconstruct Cicero’s protreptic dialogue Hortensius, and the various and numerous alleged fragments or reflections or echoes of the work. We believe that the simultaneous treatment of all of these (extremely diverse) kinds of evidence has obscured the picture and confused the scene. By focusing exclusively on the quoted blocks of Iamblichus evidence, independently and analytically, we hope to stake out solid ground in what Glenn Most has called the swampiest of the ‘many swamps in Aristotelian studies’.26 Not only is it possible to establish chapters 6–12 as authentic quotations before attempting to authenticate the other alleged fragments, it is also necessary: in order to evaluate any alleged fragment of a work as authentic, one needs ideally a large and solid body of independent evidence about that work. Without chapters 6–12, the Protrepticus simply cannot be substantially resurrected except by a miraculous future papyrus discovery; with these chapters, we get not only the lion’s share of the currently available evidence, but also a large and solid body of evidence which will serve, in future is a controversial issue, well summarized by Berti: ‘It was already included among the Aristotelian extracts by Bywater, who made the series begin with it; then it was rejected by Hirzel and Rose, admitted by Hartlich, rejected again by Jaeger and thus by Walzer and Ross, admitted, lastly, by Nuyens, Festugi›ere, D‹ecarie, and During. Against attributing it to Aristotle’s Protrepticus are the fact that Iamblichus • presents it as a Pythagorean doctrine, and the occurrence of a few terms of clearly Stoic origin’ (‘Primo’ Aristotele, 447). D. J. Allan studied the chapter closely, and argues for attributing the chapter to Posidonius, to whom a Protrepticus is attributed in Diogenes Laertius (7. 91, 129): see ‘A Passage from Iamblichus in Praise of the Contemplative Life’, Archiv f•ur Geschichte der Philosophie, 57 (1975), 246–68. 25 N. Festa (ed.), Iamblichus: De communi mathematica scientia [DCM] (Leipzig, 1891; suppl. U. Klein, repr. Stuttgart, 1975). Nor do we deal here with the related parallels in Proclus’ Commentary on Euclid, which arguably derive from his reading of Iambl. DCM 26–7. 26 G. Most, ‘Some New Fragments of Aristotle’s Protrepticus?’ [‘New Fragments?’], in Studi su codici e papiri filosofici: Platone, Aristotele, Ierocle (Studi e testi per il Corpus dei papiri filosofici greci e latini, 6; Florence, 1992), 189–216 at 205.

Created on 4 July 2005 at 14.34 hours page 203

Authenticating Aristotle’s Protrepticus

203

research, to anchor arguments about the other evidence. These chapters are the key to settling the disputes. Of course, the recovery of Aristotle will be worthwhile only if the material recovered is philosophically interesting; but that is virtually the only thing about the Protrepticus that has never been in dispute.

2. The Plato chapters (13–19) 2.1 Preliminary remarks Our source27 is the second book of Iamblichus’ series of nine (or ten) books28 On Pythagoreanism. John Dillon has said that very little in these books is original to Iamblichus. They are chiefly centos of passages from earlier writers, including Plato and Aristotle, and they have been found useful mainly on that account (large parts of Aristotle’s Protrepticus have been discerned, with great probability, in Iamblichus’ work of the same title, and a nameless 5th century b.c. Sophist is discernible in the same work). All Iamblichus contributes is introductory and bridgematerial, useful chiefly from the stylistic and terminological standpoint.29

Iamblichus’ Προτρεπτικς π φιλοσοφαν (hereafter, Protrepticus) contains twenty-one chapters, designed and titled by Iamblichus himself. Iamblichus even composed his own chapter headings, possibly the first for an ancient book of philosophy.30 27 Jamblique, Protreptique, ed. E. Des Places (Paris, 1989); all page and line numbers refer to this edition. 28 There are various possible titles for this series of works; see D. O’Meara, Pythagoras Revived: Mathematics and Philosophy in Late Antiquity (Oxford, 1989), 32–3. Of the works in the series, the extant ones are: (1) Περ το Πυθαγορικο βου; (2) Προτρεπτικς π φιλοσοφαν; (3) Περ τς κοινς µαθηµατικς πιστµης; (4) Περ τς Νικοµ χου "ριθµητικς ε#σαγωγς; the remaining five or six books are lost. 29 J. Dillon, ‘Iamblichus of Chalcis (c. 240–325 a.d.)’ [‘Iamblichus’], Aufstieg und Niedergang der r•omischen Welt, 36.2 (1987), 862–909 at 876. 30 A. Segonds in Des Places’s edition of Iamblichus’ Protrepticus (see n. 27) informs us (155) that Iamblichus’ ‘K‹ephalaia sont peut-^etre le plus ancien exemple authentique, dans un texte philosophique, de ce genre litt‹eraire; cf. H. D. Sa·rey– L. G. Westerink, in Proclus, Th‹eologie platonicienne, i (Paris, 1968), p. 130 n. 2.’ Sa·rey–Westerink (129 n. 2) report that Syrianus treats the chapter headings in the DCM of Iamblichus as authentic (In Metaph., 101. 29–102. 25 Kroll). Chapter headings in ancient books were studied by H. Mutschmann, ‘Inhaltsangabe und Kapiteluberschrift im antiken Buch’, Hermes, 46 (1911), 93–107. For the headings • in Porphyry’s Isagoge and later commentaries on Aristotle’s Categories, see J. Barnes (trans.), Porphyry: Introduction, with a Commentary (Oxford, 2003), xvii n. 35.

Created on 4 July 2005 at 14.34 hours page 204

204

D. S. Hutchinson @ Monte Ransome Johnson

We can describe the contents of these chapters in three groups. Chapters 1–4, altogether about fifteen pages in the Bud‹e edition of Des Places, contain an introduction (including methodological remarks), a chapter of common sayings or γν%µαι, a chapter on specifically Pythagorean γν%µαι, and a chapter of protreptic arguments which Iamblichus calls ‘esoteric’ in the sense of being native to an inner Pythagorean tradition; he attributes these, probably incorrectly, to the Pythagorean philosopher Archytas, in his work On Wisdom, ‘right at the beginning’ (ch. 4, 48. 26–7). Chapters 5–20, about seventy-six Bud‹e pages, contain what Iamblichus calls ‘mixed protreptic’, an odd term which turns out to mean protreptic material borrowed from other writers, chiefly Plato and, it seems, Aristotle, organized by Iamblichus to fit in with his Pythagorean protreptic scheme. Iamblichus makes a point of saying (ch. 1, 41. 3–8) that his synthetic and syncretic work is very much open to protreptic arguments from every philosophical tradition: ‘For protreptic will incite to philosophy itself and to philosophizing in general, according to every system of thought, no particular school being expressly preferred, but all being accepted according to their peculiar merits, and ranked higher than mere human studies, by a common and popular mode of protreptic.’ Chapter 21, about twenty Bud‹e pages, contains thirty-nine sections, each consisting of one of the Pythagorean maxims or σµβολα, with brief commentary. Two things are immediately evident from this arrangement. First, the bulk of Iamblichus’ work is made up of quotations taken from other writers, especially Plato and Aristotle, it seems, but also the so-called Anonymus Iamblichi (in chapter 20).31 Secondly, Iamblichus has presented all of this material as if it was all already Pythagorean in content, and he has placed the least o¶cially Pythagorean material within a shell of Pythagorean scholasticism. Iamblichus himself describes in his first chapter his plan for the book and his methodology of using sources: After this we must use a sort of ‘middle’ method, neither thoroughly popular nor yet utterly Pythagorean, but not completely removed from each of these two modes either. In this way, we will organize the common and universal encouragements to philosophy, so as to present them separately from Pythagorean dogma, at least by way of this mode, but we will mix 31 For the Anonymus Iamblichi, see Anonimo di Giamblico: La pace e il benessere. Idee sull’economia, la societ›a, la morale, trans., intro., and comm. by M. Mari with a preface by D. Musti (Milan, 2003).

Created on 4 July 2005 at 14.34 hours page 205

Authenticating Aristotle’s Protrepticus

205

in, in the most opportune way, the most authoritative opinions of these men as well, so that it becomes familiar what the mode of the Pythagorean protreptic is, at least by way of this mediate mode of argument. (Iambl. Protr. ch. 1, 41. 8–17; cf. ch. 21, 131. 16–132. 15)

We see expressed his commitment to gathering protreptic arguments from various sources, compiling ‘common and universal’ encouragements to philosophy, ‘according to every system of thought, no particular school being expressly preferred’. Here is a synopsis of the Plato material used in the later chapters of Iamblichus’ book. The words ‘opening’ and ‘closing’ refer to short passages, not from Plato but written by Iamblichus, at the beginnings and endings of his chapters (generally of one sentence, but sometimes several); ‘bridge’ refers to a short transitional passage, not from Plato but written by Iamblichus. These will be carefully studied in what follows. Ch. 13. Opening—Phaedo 64 a–69 d—bridge—Phaedo 82 b–84 b—bridge —Phaedo 107 c–d—closing. Ch. 14. Opening—Theaet. 173 c–177 b—closing. Ch. 15. Opening—Rep. 7, 514 a–517 c—closing. Ch. 16. Opening—Rep. 7, 518 b–519 b—closing. Ch. 17. Opening—Gorg. 492 e–494 a—closing. Ch. 18. Opening—Gorg. 504 a–505 b—closing. Ch. 19. Opening—Gorg. 506 c–508 a—bridge—Menex. 246 d–247 b— bridge—Menex. 247 e–248 b—bridge—Laws 2, 660 e–661 c—closing.

Two general features are immediately evident from this catalogue: each chapter is designed as a vehicle to contain one or a few Plato blocks; the blocks of Plato are selected from several works, distributed unevenly across the chapters of Iamblichus. To go into a little more detail on chapter design: each chapter is designed to contain Plato material, within bookends consisting of thin opening and closing passages written by Iamblichus; the Plato material in any chapter is discontinuous from the material in other chapters; the chapters have either one block of Plato quotations, or up to four discontinuous blocks, in which case they are separated from every other by a bridge passage written by Iamblichus. In this chapter design, passages written by Iamblichus always indicate a gap in the continuity of the citation, and any two discontinuous citations are always separated by passages written by him. To go into a little more detail on selection of material: the twelve

Created on 4 July 2005 at 14.34 hours page 206

206

D. S. Hutchinson @ Monte Ransome Johnson

blocks of Plato quotations come from six works, placed unevenly into the seven chapters: 13 = Phaedo, 14 = Theaetetus, 15 + 16 = Republic 7, 17 + 18 = Gorgias, 19 = Gorgias+ Menexenus+ Laws 2. One chapter contains more than one work, and two works occupy more than one chapter. Two works contribute one block only (Theaetetus, Laws 2), but the other four contribute two or three blocks, either across two or three chapters (Republic 7, Gorgias) or within a single chapter (Phaedo, Menexenus). When moving from block to block, Iamblichus uses works one by one, not returning to a work once he has moved on to another; and, within a work, he follows the work in its natural sequence, not going back to earlier passages in a work after he has already quoted later passages. But here is the crucial question for us: how faithful to Plato are these twelve blocks of quotation? Are they literal quotations, or modified quotations, or paraphrases, or passing references, or some other kind of new tissue made of old patches and threads cut from the ancient book? To answer this crucial question we need to make a comparison of the two versions, Plato’s and Iamblichus’, collating them passage by passage, word by word, and letter by letter, and then noting and explaining all the di·erences. This is a complex and laborious job, but not an impossibly huge one, as the relevant part of the text is only about thirty pages long. In the comparisons below, we give the results of our study, beginning with chapter 14, described in excruciating detail, with every single di·erence between the two versions referred to and displayed, in both Greek and English. Other chapters will be described and explained more rapidly, but still with all the di·erences noted and referred to at least schematically. 2.2. Chapter 14: quotations from Theaetetus With those preliminaries o·ered, let us roll up our sleeves and report the results of the collations, starting with chapter 14. Conclusions about this chapter are to be found on p. 213, for those readers who have little interest in the details; and summary conclusions drawn from all seven Plato chapters are to be found in subsection 2.8, on p. 240. Consider the following abbreviated version of chapter 14. We shall use boldface to express our scholarly comments, and we shall

Created on 4 July 2005 at 14.34 hours page 207

Authenticating Aristotle’s Protrepticus

207

represent Iamblichus’ words in italic and Plato’s unmodified words in roman typeface. Chapter 14. It’s necessary also to construct the protreptic starting from the way of life of the men who are ‘chieftains’ in philosophy, following the precepts of Pythagoras. So, to begin with, the fellows who are like this, I suppose, grow up not knowing the market place <. . . here follows a long section in which Iamblichus continues to follow Theaet. 173 c 9–177 b 7 . . .> In the end, the things he says do not satisfy even himself; that famous eloquence of his somehow dries up, and he is left looking nothing more than a child. So if this is how things stand, and the most divine and most successful way of life appears to be that of those who lead their lives in philosophy, there’s nothing else to be done except to embrace philosophy with a noble spirit. (Iambl. Protr. ch. 14, 100. 21–105. 27)

Iamblichus has begun and ended this chapter in his own voice,32 but otherwise followed his source text closely. This ‘close following’ of Plato is, however, not exact quotation, but rather a particular sort of modified quotation, as there are certain alterations deliberately introduced into the text by Iamblichus to erase the dialogue frame and e·ace the speakers. Let us examine each of these types of difference carefully. The famous speech that Iamblichus has excerpted in this chapter is also a self-contained speech in its original context in Plato’s Theaetetus; in form it is an epideixis, in this case a display of the suitability of the philosopher’s stance towards the public political world, containing the unforgettable image of Thales observing the heavens above and falling down a well. It is preceded by an open choice of what topic to speak about (173 c) and concludes with a decisive putting aside of the topic; after his rhetorical climax ‘left looking nothing more than a child’, Socrates says (177 b 7–8), ‘But we had better leave it there; all this is really a digression.’ Into this self-contained epideixis Plato has written minimal interventions33 on the part of Theodorus, so as to keep the illusion of dialogue 32 Even when writing in his own voice, he borrowed a colourful word from the local environment, chamaeleon-wise (e.g. 173 c 8 κορυφαων ‘chieftains’). 33 The interventions are minimal in the sense that they consist either of pure unqualified question (‘what do you mean?’ at 174 a 3 and 176 e 2) or else pure unqualified assertion (‘very much so’, ‘what you say is true’, ‘that’s just like how it happens’ at 177 a 9, 174 b 8, 175 b 7), apart from this unexpected profession of faith (176 a 2–4): ‘Socrates, if your arguments convinced everyone else as much as they convince me, there would be more peace and less wickedness on earth.’ This is not to say that the dialogue frame is of minimal importance for Plato’s work; on the contrary, it is so important that Plato needs to keep reminding us of it. But it was

Created on 4 July 2005 at 14.34 hours page 208

208

D. S. Hutchinson @ Monte Ransome Johnson

ticking over while the speech takes place, and it is exactly these interventions that Iamblichus has removed in his modified quotation. Consider this first case, laid out in comparative fashion: first, a wider passage from Plato in translation, with a narrow passage highlighted in bold, showing the zone of alteration; next, this narrow passage in Greek; then, the corresponding narrow passage in Iamblichus in Greek; finally, the corresponding wider passage from Iamblichus in translation, showing the zone of alteration highlighted in bold. What is not highlighted has not been modified in any way. . . . , geometrizing upon earth, measuring its surfaces, astronomizing in the heavens above, tracking down by every path the whole nature of each entire existing thing, not condescending to anything that lies near to hand. What do you mean by that, Socrates? For instance, they say Thales was studying the stars, Theodorus, and looking upwards, when he fell into a well . . . π%ς τοτο λγεις, & Σ(κρατες; — )σπερ κα Θαλν "στρονοµοντα, & Θε+δωρε, κα -νω βλποντα (Theaet. 174 a 3–5 ≈ Protr. 101. 14–15) )σπερ δ. κα Θαλν "στρονοµοντα κα "ναβλποντα . . . , geometrizing upon earth, measuring its surfaces, astronomizing in the heavens above, tracking down by every path the whole nature of each entire existing thing, not condescending to anything that lies near to hand. For instance, they say Thales was studying the stars and looking up, when he fell into a well . . .

The dialogue exchange is excised with surgical precision, doing no harm to the surrounding organ of argumentation, leaving no scar tissue: nobody would be able to tell from this quotation, as modified by Iamblichus, that there was originally a quick exchange of words between speakers at the highlighted point; even an appropriate connecting particle δ is supplied, to stitch together the interrupted words of Socrates. A second example will further illustrate this surgical technique of dialogue excision: . . . is not what they suppose—corporal and not important for Iamblichus’ work, indeed counter-productive to it, in much the same way that modern interpreters of the speech prefer not to be distracted by the dialogue. For a detailed recent study of the speech, see D. Sedley, The Midwife of Platonism: Text and Subtext in Plato’s Theaetetus [Midwife] (Oxford, 2004), 62–86.

Created on 4 July 2005 at 14.34 hours page 209

Authenticating Aristotle’s Protrepticus

209

capital punishment—things which they might entirely escape, in spite of their crimes. It is a punishment from which there is no escape. And what is that? One of the patterns established in reality, my friend, is of what is godly and highly successful, and another one is of what is godless and utterly miserable . . . τνα δ. λγεις; — παραδειγµ των, & φλε, ν τ/% 0ντι 1στ(των (Theaet 176 e 2–3 ≈ Protr. 105. 6) παραδειγµ των γ2ρ ν τ/% 0ντι 1στ(των . . . is not what they suppose—corporal and capital punishment—things which they might entirely escape, in spite of their crimes. It is a punishment from which there is no escape. For one of the patterns established in reality is of what is godly and highly successful, and another one is of what is godless and utterly miserable . . .

Simple excisions of content-neutral speech exchanges such as the above also take place twice more (174 b 7–c 1, 175 b 7–8) in Iamblichus’ modified quotation of Theaetetus. But sometimes the surgery is a little more elaborate, not a simple deletion of the interlocutor’s words, but a partial exploitation, as here: . . . they will listen to us as clever men and rascals listen to certain fools. Indeed, very much so, Socrates. I know it, my friend; but there is a certain one thing . . . κα µ λα δ, & Σ(κρατες. — ο4δ τοι, & 1ταρε. 5ν µντοι τι Theaet. 177 a 9–b 1 ≈ Protr. 105. 17 κα µ λα δ. 6προγκον φρονοντες. 5ν µντοι τι . . . they will listen to us as clever men and rascals listen to certain fools, and think themselves very much superior; but there is a certain one thing . . .

A slightly more complex case is the following one, which exploits the whole of a significant comment by the interlocutor Theodorus, the only one of his utterances that was not a neutral question or neutral assertion: . . . to tune the strings of common speech to the fitting praise of the way of life of gods and of successful men. Socrates, if your words convinced everyone as they do me, there would be more peace and less evil among them. But it is not possible for evil to be destroyed, Theodorus; some opposite . . . ε# π ντας, & Σ(κρατες, πεθοις 7 λγεις )σπερ µ, πλεων 8ν ε#ρνη κα

Created on 4 July 2005 at 14.34 hours page 210

210

D. S. Hutchinson @ Monte Ransome Johnson

κακ2 λ ττω κατ9 "νθρ(πους ε:η. — "λλ9 ο<τ9 "πολσθαι τ2 κακ2 δυνατ+ν, & Θε+δωρε — 6πεναντον γ ρ τι Theaet. 176 a 2–6 ≈ Protr. 104. 3–7 ο4µαι δ. ο=ν, ε# πεθοιντο ο> -νθρωποι τοτοις, πλεονα 8ν ε#ρνην κα κακ2 λ ττω κατ9 α?το@ς Aσεσθαι. τ2 µBν ο=ν κακ2 ο<τ9 "πολσθαι δυνατ+ν· 6πεναντον γ ρ τι . . . . . . to tune the strings of common speech to the fitting praise of the way of life of gods and of successful men. And so I think, if people were convinced by this, there will be more peace and less evil among them. Now then, it is not possible for evil to be destroyed; some opposite . . .

Notice that in this case Iamblichus removes the address to Theodorus, as of course he needs to, to e·ace the speakers. He also does this in the other places where Socrates addresses his silent listener, simply excising ‘sir’ and ‘my good man’ from the text (177 b 4, 176 b 2), and e·ecting a simple replacement here: Of the two types, Theodorus, this is the man who has been brought up in real freedom and leisure, the one you call a philosopher, to whom it is no disgrace to seem na•§ve . . . οEτος δ. 1κατρου τρ+πος, & Θε+δωρε, F µBν τ/% 0ντι ν λευθερGα τε κα σχολH τεθραµµνου, Iν δ. φιλ+σοφον καλες, /J "νεµσητον ε?θει δοκεν Theaet. 175 d 7–e 2 ≈ Protr. 103. 22–4 οEτος δ. 1κατρου τρ+πος τς ζως· F µBν τ/% 0ντι ν λευθερGα τε κα σχολH τεθραµµνου, Iν δ. φιλ+σοφον καλεν, "νεµσητον ε?θει δοκεν Of the two types of life, this is the man who has been brought up in real freedom and leisure, the one to call34 a philosopher, to whom35 it is no disgrace to seem na•§ve . . .

This completes our inventory of the modifications implemented by Iamblichus to flatten to monologue by erasing the dialogue frame and e·acing the speakers, each of them executed with as little change to the text as possible. There is only one further modification in the chapter, of the opening words, so that they flow on more smoothly from the opening sentence written by Iamblichus: 34 As the translation indicates, the text of Iamblichus has su·ered corruption here: in H. Pistelli’s earlier edition (Iamblichi Protrepticus (Leipzig, 1888)), he had supplemented δ. <δε> (‘the one that should be called a philosopher’), whereas Des Places emended καλεν to καλες (‘the one you call a philosopher’), on the basis of the Plato text and a late Iamblichus manuscript. 35 Another corruption; the supplement is taken from the Plato text and a late Iamblichus manuscript.

Created on 4 July 2005 at 14.34 hours page 211

Authenticating Aristotle’s Protrepticus

211

And, to begin with, those fellows, I suppose, from childhood onwards . . . ο=τοι δ που κ νων πρ%τον Theaet. 173 c 9 ≈ Protr. 100. 23–4 ο> γ2ρ τοιοτο που κ νων πρ%τον So, to begin with, such fellows, I suppose, from childhood onwards . . .

And with this trivial modification we have exhausted all the changes that Iamblichus has brought to the text of this part of Plato’s Theaetetus. Each of them is of a sort that does not alter Plato’s line of thought, introduce any new ideas, or even introduce any significantly di·erent ways of expressing Plato’s original ideas. The ‘close following’ of one text by Iamblichus is a matter of pure quotation, except for strictly limited modifications, and these only in order to flatten to monologue. The block of quotation is solid and pure, in that nothing significant is taken away and nothing at all is added to it; and the sequence is natural, in that Iamblichus has naturally quoted it in order, earlier to later. Now we can go further: when Iamblichus does quote Plato’s Theaetetus, we can ascertain that he quotes with rigorous literal accuracy. We collated the version found in the Iamblichus manuscript36 to the versions found in the Plato manuscripts and other witnesses37 to see what manner of witness it is. Generally the Iamblichus version contains a fair quantity of normal erroneous38 variants, of the sort and at the rate one usually finds in manuscripts 36 We relied on the recent edition by Des Places for the readings of F, the only independent manuscript. The manuscript was accurately collated by its first editor Pistelli, as both During and Des Places have confirmed from their own collations. • D.S.H. visited the Biblioteca Medicea Laurenziana in May 2004 to consult the manuscript and do partial collations; the manuscript remains clearly legible and we reconfirm that it has evidently been collated to a high degree of accuracy. Many thanks to the Laurenziana’s Director Franca Arduini, to Sabina Magrini, and to our mutual friend Antonio Carlini. 37 We relied on the recent edition of Plato, Theaetetus, in Platonis opera, vol. i, ed. E. A. Duke et al. (Oxford, 1995) [1995 OCT]; translations from Theaetetus, here and hereafter, after Levett (in J. Cooper and D. S. Hutchinson (eds.), Plato: Complete Works [Complete Plato] (Indianapolis and Cambridge, 1997)). We relied on the 1995 OCT edition for reports of the main variants in the Plato witnesses. 38 For the sake of simplicity in the analysis that follows, let the OCT editors’ choice define as being ‘erroneous’ any reading not preferred by them; certainly it might be that the editors will have chosen the incorrect variant here and there, but the pattern would change little if we were to re-examine every decision made by them. (In subsect. 2.6, however, we do subject the variants in chapter 13 (Phaedo) to this extra degree of scrutiny, which confirms that it makes little di·erence; we have

Created on 4 July 2005 at 14.34 hours page 212

212

D. S. Hutchinson @ Monte Ransome Johnson

of ancient Greek texts. Many of these variants are unique to it,39 but some of these errors are shared by certain other witnesses, especially those in the family of MSS W and P; and these common variants are of a sort and at a rate to indicate common ancestry.40 The coincidence of W and P with Iamblichus in these readings proves that they were ancient readings, at least as ancient as Iamblichus himself; this means that they should all be examined on reconsidered every editorial decision, but the picture is the same, only with slightly more focus.) The information derived from the apparatus criticus of the 1995 OCT has been further simplified by suppressing reports of erasures, corrections, and later hands. As for accentuation, we dispense with this also, not only for simplicity, but also on the principle that it is usually pointless to accentuate variants in an apparatus criticus, since they cannot generally be traced back as far as an ancient manuscript, and are rarely collated carefully. Besides the Greek MSS B (with DT) and W (with P), the witnesses are I = Iamblichus (MS F); Stobaeus; Eusebius; Clement; Themistius; Porphyry. 39 All the variants present in Iamblichus alone, relative to other witnesses of the Plato text, are of the standard sorts that occur in books copied by hand repeatedly from one manuscript to another. Most of the corruptions involve a single Greek letter: 173 d 7 γεγονος] γεγονως 174 a 5 ανω βλεποντα] αναβλεποντα 174 b 4 δε] δη 174 b 9 συγγιγνοµενο<ι>ς 174 c 1 εκαστω] εκαστος 174 e 4 πληθει] πληθη 175 c 6 τε τινε] γε τινες 175 c 8 αυτοιν] αυτην 175 d 6 εναντιως] εναντιοις 175 e 1 [ω] 176 c 6 δοκουσ[α]ι 177 a 2 τ<ε>ινουσι 177 b 3 εθελησωσιν] εθελησουσιν 177 b 3 εθελησουσιν. Most of the rest of the corruptions are also on a small scale, and of normal types: 174 b 3 [ει] 174 c 7–8 ατ9 ουκ] αυτος ουκ 174 d 5 συβωτην] των συβωτων 175 d 6 ου<δε> γαρ 175 e 3 επισταµενου<ς> 175 e 6 [δε] 176 c 3 ανανδρια] ανδρεια. Sometimes a loss took place, due to a normal mechanical corruption: 175 a 7–b 1 Αµφιτρυωνος [ατοπα . . . Αµφιτρυωνος]—the eye of a scribe skipped forward from one ‘Amphitryon’ to another, with the loss of the intervening line, which also caused 175 b 1 [πεντεκαιεικοστος]; probably the cause of 176 e 4 [αθεου] αθλιωτατου is also mechanical. This leaves only one case, 175 e 5 τε] θατερου, that might possibly have been caused by intervention; but accidental corruption is here too the best explanation, as the intervention would have been pointless on the part of Iamblichus. 40 Shared erroneous variants indicate common ancestry when any of them is unlikely to arise independently, as in these two cases, with W: 174 a 7 εµπροσθεν] οπισθεν I+W+Euseb. 175 c 4 τ9 αυ BDT] πολυ I+Euseb.; τ9 αυ πολυ WP. Shared errors also indicate common ancestry when they are too numerous to be plausibly explained by coincidence, as here, nine times in a short passage, with W, including these seven cases: 173 d 7 τω] το I+W 175 c 5 <η> βασιλειας I+WP+Euseb. 175 e 3 επισταµενου BDT+Euseb.] επισταµενους I; επισταµενος WP 175 e 7 [γ9] I+WP+ Euseb. 176 b 7 λεγωµεν BDT] λεγοµεν I+WP+Euseb. 176 c 3 <η> ουδενια I+WP+ Euseb. 177 b 3 εθελησωσιν BDT+Euseb.] εθελησουσιν I: θελουσιν WP. The next two cases are probably coincidence in error, genetically insignificant: 176 b 3 [τι] I+ BD 177 b 4 φ<ε>υγειν I+BDT. Finally, for the sake of completeness, four cases of genetically insignificant agreement in error, in which the preferable reading is not o·ered by Iamblichus or any of the main manuscripts or witnesses, but rather by ancient or modern conjecture: 173 e 5 τα<ς> τε Campbell from Clement 175 d 4 βαρβαριζων] βατταριζων Pearson from Themistius 175 e 7 ελευθερ<ι>ως Athenaeus 176 a 1 βιον [αληθη] Athenaeus.

Created on 4 July 2005 at 14.34 hours page 213

Authenticating Aristotle’s Protrepticus

213

their own merits.41 It also proves that Iamblichus quoted from an ancient copy with perfect literal accuracy, so literal that his version preserves distinctive genetic features that characterize the readings of the family of that ancient copy; this cannot be done from memory. This means that the Iamblichus manuscript counts in evidence just as if it were a Plato manuscript, for these sections. And it provides valuable evidence, to be sure, as it preserves the correct reading in several cases where important Plato manuscripts contain an error, and one case (176 c 2) where all the important Plato manuscripts go wrong.42 In certain circumstances, when all the Plato manuscripts happen to go wrong simultaneously, it might uniquely preserve the right reading.43 2.3. Chapters 15 and 16: quotations from Republic 7 The picture of Iamblichus’ methods that we derive from the above examination of chapter 14 holds good, more or less, for the rest of his Plato chapters. In what follows, we analyse each of these chapters to obtain more precision regarding this ‘more or less’; we shall quickly note the types of modification familiar to us from chapter 14 (without, however, going into as much detail), slowing down to study in detail anything new or di·erent. In this way we can study every chapter and each type of modification in the most e¶cient and clearest way. Let us start again with chapters 15 and 16, in which Iamblichus gives us a close version of the Myth of the Cave, from Plato’s Republic book 7: Chapter 15. Next, then, it is necessary to compare the e·ect of education 41 This was done by the OCT editors, very consistently as far as indicated by our study of this part of their apparatus criticus; they also quite properly suppress most of the unique variants o·ered by the Iamblichus version when they are evident errors. 42 Seven cases where the reading transmitted by Iamblichus is preferable to that of Plato manuscripts: 173 d 6 τις I+WP+Clement+ Porphyry] τι BDT+Euseb.+ Themistius 173 e 4 πετεται I+WP+Porphyry] πεταται Clement+Euseb.; φερεται BDT 175 c 7 κτησασθαι I+WP+Euseb.] κτησεσθαι BDT 175 e 4 F δ9 I+P] M δ9 W; ουδε BDT; του δ9 Euseb. 176 b 5 δη I+WT] δε Euseb.; om. BDP 177 b 2 οταν I+ WP+Euseb.] οτι αν BDT 176 c 2 τουτο I+Euseb.+Stob] τουτου BDTWP—in this last case Iamblichus’ reading is preferable to that of all of the Plato manuscripts. 43 This does seem to have happened, not only at Theaet. 176 c 2, but also at Phaedo 69 a 8, 83 b 7–8, 83 c 8, and 84 b 1 (below, n. 76), Rep. 7, 515 b 4–5 (below, p. 217), and Gorg. 492 e 7 and 506 e 1 (below, p. 237).

Created on 4 July 2005 at 14.34 hours page 214

214

D. S. Hutchinson @ Monte Ransome Johnson

<. . . excerpt from Rep. 7, 514 a 1–517 c 4 . . .> it is necessary to see this for the one who is going to act sensibly either in private or public. If this is the function of education, and if the di·erence between it and lack of education is that great, what else would it befit us to do than to embrace education and philosophy, and renounce those things considered to be greatly desirable by the majority, since those things have no influence on success worthy of the name? (Iambl. Protr. ch. 15, 105. 28–110. 2)

This and the next chapter are notable in that the opening sentence is not even an addition to the Plato passage but a transformation of its beginning. The closing sentence, which is similar to that of chapter 14, is of a sort that will become familiar to us. The Myth of the Cave is presented by Plato in the form of a onesided dialogue between Socrates and Glaucon, which causes Iamblichus to modify the quotation in various ways so as to suppress the speakers and the context of the discussion. So, for example, Iamblichus combs out from this passage every single ‘I said’ and ‘he said’ that had originally punctuated the dialogue between Socrates and Glaucon in Plato’s text.44 Also suppressed are an address to Glaucon, a rhetorical question to him, and a reference to what he sees (517 a 8, 515 e 7, 514 b 5), as well as references to the ‘present’ discussion and to what has ‘just’ been said.45 Sometimes Iamblichus suppresses dialogue by simply excising the words of Glaucon (514 b 8, 515 e 4, 516 b 3, 516 b 7), but at other times by transforming the words of Glaucon into an assertion that expresses their content.46 The most significant alteration is the deletion of two lines at 517 b 5–7, in which Socrates expresses a rather particular speaker’s attitude to Glaucon; in the exposition of doctrine designed by Iamblichus, this would have struck a jarring and disorienting note. The original version in Plato’s text reads: ‘you won’t mistake my intention—since that’s what you want to hear about. Maybe God knows whether they are in fact true, but the appearances appear like this to me; in the intelligible realm, etc.’ Here is the version after the deletion and transformation of Iamblichus: he won’t mistake the truth. And this is surely how it appears; in the intelligible realm, etc. 44 ‘I said’ 514 a 1, 515 a 5, 515 e 5, 515 c 1, 515 c 4, 516 e 3, 517 a 8; ‘he said’ 515 a 4, 515 a 9, 515 c 3, 515 d 8, 516 a 4, 516 c 3, 516 e 1, 516 e 6, 517 a 7. 45 516 a 3 and 517 b 1; both changes are reported by both of the OCT editors (J. Burnet (ed.), Platonis opera, vol. iv (Oxford, 1902), and S. R. Slings (ed.), Platonis Respublica (Oxford, 2003)) as if they were variant readings, whereas they are more likely to be the product of deliberate intervention on the part of Iamblichus. 46 515 a 4, 515 a 9–b 1, 515 d 8, and 516 a 4, causing textual problems at 107. 24–5.

Created on 4 July 2005 at 14.34 hours page 215

Authenticating Aristotle’s Protrepticus

215

Notice that even in this relatively large intervention, Iamblichus takes care to make use of as much of Plato’s thought and language as possible, stripping away only the words that express a personal attitude. But for some reason Iamblichus does not feel any strict need to eliminate all the dialogue. At 515 b 2–c 3 we find preserved four pairs of comments, four by each speaker, in what is, evidently and suddenly, a dialogue between two speakers. Beyond the deletions of ‘he said’ (noted above), the only change (515 b 10) is from ‘By Zeus, not I!’ to ‘Nothing else’, a less enthusiastic and personal response than that given by Glaucon. In a later and longer section (516 c 7–517 c 4) we have even more fidelity to the text, with no changes introduced by Iamblichus beyond the deletions of ‘I said’ and ‘he said’. The result has a flowing modern feel to it: Do you think he would . . . go through any su·erings, rather than share their beliefs and live as they do?—Yes, I think he would rather su·er anything than live like that.—Consider this too, then. If this man went back down into the cave and sat down in his same seat, wouldn’t his eyes be filled with darkness, coming suddenly out of the sun like that?—Certainly.—Now, if he had to compete once again with the perpetual prisoners in recognizing the shadows . . . (516 d 4–e 8, trans. Reeve,47 unmodified)

It would be hard to say that this should even count as an alteration, still less a degradation of the text. In fact, the above translation of Plato’s text by Reeve serves equally precisely as a translation of Iamblichus 108. 22–109. 4, since Reeve has systematically done to Plato’s text what Iamblichus does occasionally—remove the ‘he said’ and ‘I said’ but preserve everything else. Both ancient and modern scholars can look to Plato himself for a justification of this procedure: One conscious deviation from strict accuracy, however, will be obvious at a glance. The Republic is largely in reported speech. Socrates is relating a conversation he had in the past. But I have cast his report as an explicit dialogue in direct speech, with identified speakers. In the Theaetetus, Plato has Euclides adopt a similar stratagem. ‘This is the book,’ he says to Terpsion; ‘You see, I have written it out like this: I have not made Socrates relate the conversation as he related it to me, but I represent him as speaking directly to the persons with whom he said he had this conversation.’ (Reeve, viii) 47 Plato, Republic , trans. C. D. C. Reeve (Indianapolis and Cambridge, 2004).

Created on 4 July 2005 at 14.34 hours page 216

216

D. S. Hutchinson @ Monte Ransome Johnson

Iamblichus generally goes further than Reeve in modifying the text of Plato, but his intention was similar: in order to sharpen the focus on the essential protreptic ideas, it can be useful to purify the central texts of Platonism and free them from the bonds of their sometimes distracting narrative embodiment.48 However great the degree of intervention that Iamblichus makes in this Myth of the Cave material, his version is strictly structured: he quotes a pure block of text, nothing added, nothing significant taken away; and he follows a natural sequence, moving from Theaetetus to Republic 7, and from earlier to later in both texts. Just as in chapter 15, Iamblichus opens chapter 16 not by crafting his own sentence, but by barely modifying the first line of the passage he is excerpting. He suppresses the ‘I said’ at 518 b 7 and brings forward the backward-looking comment ‘if this is true’: Chapter 16. Now then, again, if this is true, it is necessary to think about these things in some such way as this. Education is not what some people advertise it as being <. . . an excerpt from Rep. 7, 518 b 8–519 b 5 . . .> the things towards which it <sc. the vision of the soul> is presently turned. So now then, now that we have reached this point, and it is manifestly evident what kind of function philosophy has, it is also immediately clear that it is valuable. For to strip away the becoming from the soul, and to purify its activity which is capable of reasoning, are very appropriate to it. So ‘this is the best mode of life, to practise justice and the rest of virtue in living and also in dying’. So we would follow this, if we wanted to be successful in reality. (Iambl. Protr. ch. 16, 110. 3–111. 17)

The closing in this case is not a single sentence but a little paragraph of concluding remarks by Iamblichus that show many typical traits; they are meta-textual, enthusiastic, and often adorned with memorable words or phrases from the author, as in the case of the quotation in the penultimate sentence (from Gorg. 527 e 3–5). Iamblichus continues quoting Republic 7 in chapter 16, but from 48 Reeve’s claim, that this enhances readability and intelligibility, has been enthusiastically received by some early readers whose opinions are printed on the book’s back cover. ‘Reeve’s brilliant decision to cast the dialogue into direct speech produces a compelling impression of immediacy unmatched by other English translations currently available’ (Lloyd Gerson). ‘In this version, Socrates’ narration of his conversation with Glaucon and Adeimantus et al. is converted so far as possible into dramatically gripping and e·ective direct speech. Its increased accessibility promises to make it the number-one choice for undergraduate courses’ (John Cooper). Iamblichus may well have thought that he was doing a similar service for the students for whom he compiled his book, by o·ering them enhanced versions of the great passages.

Created on 4 July 2005 at 14.34 hours page 217

Authenticating Aristotle’s Protrepticus

217

a little later, the passage that indicates that teaching philosophy is not like putting sight into blind eyes, but turning the student around, converting the student, whose mind then gazes in the right direction to gain understanding; this passage is evidently highly appropriate to Iamblichus’ collection of protreptics. The only changes are intended to remove the dialogue (all of it this time): excision of ‘I said’; excision of Glaucon’s short comments; and excision of an insignificant speech exchange.49 Some stretches of dialogue in other chapters required more transformation on the part of Iamblichus, especially when Socrates’ interlocutor contributed important ideas to the discussion, but in this case it su¶ced to make several simple excisions only. Again, the block of quotation is solid and pure, and the sequence is natural: Iamblichus moves from an earlier passage in book 7 to a later one, and within the passage preserves the natural order. So much for the deliberate modifications. Now let us ask about the accidental di·erences, the variants found in the various manuscripts. After pooling together the results from both chapters 15 and 16, in which Iamblichus quotes from Republic 7, we can investigate the family a¶liation of the Plato manuscript used by Iamblichus. In addition to a possible correction50 and a few dozen errors not shared by others,51 there are a few that are shared by others, perhaps suggesting a¶liation with the family of F, but these are too few and too weak for this to be a confirmed result.52 Yet the Iamblichus version does some service to readers and editors of Republic 7 at a few places, especially at 515 b 4–5, a serious crux where the preferred text in the two key variants is that of Iamblichus, not that of the main manuscripts. Here too, it seems that Iamblichus (or his scribe) was working carefully from a decent copy of Plato’s text of the Republic, as he did with the Theaetetus. 49 518 c 4, 518 d 3, 519 a 7; 518 c 3, 518 d 8, 519 a 6; and 518 d 1–2 [N γ ρ; να]. 50 518 d 7 διαµηχανσασθαι] δε µηχανσασθαι. 51 Most of these are recorded in Slings’s OCT apparatus (at 514 b 5, 514 b 9, 515 c 5, 515 d 6, 516 b 1, 516 c 8, 516 e 8, 517 a 2, 517 a 4, 517 b 8, 517 c 3, 518 b 7, 518 c 1, 518 c 5, 518 c 9, 518 d 5, 519 a 2, 519 b 2), but other variants are recorded only in the Iamblichus apparatus (106. 24, 107. 8, 107. 13, 108. 21, as well as spelling di·erences), as well as one variant that ought to have been recorded in the apparatus criticus of Plato, as it is a candidate reading: 516 b 6 οOος <τ> στι 108. 8. 52 516 e 6 [και] µαλα I+F (but this could be deliberate on Iamblichus’ part) 515 c 4 των <τε> δεσµων I+FD (but this may be agreement in truth, not error) 514 b 7 δεικνυασιν A] δεικνυουσιν I+D; δεικνυσιν F 516 d 7 <ε>κεινα I+F 518 c 5 /J] I I+ F (but these are trivial and non-probative).

Created on 4 July 2005 at 14.34 hours page 218

218

D. S. Hutchinson @ Monte Ransome Johnson

2.4. Chapter 17: quotations from Gorgias Now let us turn to chapter 17, in which Iamblichus quotes from Gorgias. The speech in question is the crucial speech analysing hedonism by means of the metaphors of sieves (insatiability of desire) and leaky jars (recurrence of need). It is a medium-sized speech, broken in the middle by a brief exchange between Callicles and Socrates (another fruitless stand-o·), and surrounded fore and aft by several lines of significant dialogue. This is how it ends up, nested in the context of Iamblichus: Chapter 17. But if it’s necessary to remind the listeners of this invocation from the ancient arguments, stories of the priests and others, and the Pythagoreans, then let’s start this too from this point. It is right to say that those who have no need of anything are successful, and that those whose desires are unlimited have a terrible way of life. <. . . Iamblichus continues with this modified quotation of Gorgias 492 e 3–494 b 6 . . .> The one has his success in his being filled up as much as possible, for which it is necessary for there to be a lot going out and for there to be certain large holes for the e}uent, which is nothing else than living the way of life of some stone curlew, but the other, having been filled up once for all with his own proper goods, remains always independent for all time. Something of this sort is also a time-honoured invocation to the discipline of virtue. (Iambl. Protr. ch. 17, 111. 18–113. 21)

Here again, at beginning and end we see the voice of Iamblichus, providing a sentence or two of opening and closing, sentences that make comments about the text quoted, meta-textual notations. Note that Iamblichus has also stitched a half-sentence on to the end of the version, to smooth the transition to the conclusion, and to end on a positive hortatory note. What remains is entirely a close following of Plato’s text, especially the middle section 492 e 7–494 a 2. When we look closely, we also find two sentences (translated above), the one after the opening and the one before the closing, that paraphrase the opening and closing dialogues that bookend the speech. These examples are instructive of how Iamblichus transforms into continuous prose a section of lively Platonic dialogue. The first of the above paraphrasing sentences corresponds to this in the original version by Plato (492 e 3–7): ‘So then, is it not right to say that those who have no need of anything are successful?—No, for in that case stones and corpses would be the most successful.—

Created on 4 July 2005 at 14.34 hours page 219

Authenticating Aristotle’s Protrepticus

219

But then the ones that you say are have53 a terrible way of life, too.’ When we compare this with the version in Iamblichus (above), we see that he has e·aced the dialogue, as well as dispensed with the reference to corpses and stones; he has also made more explicit the reference of ‘the ones that you say are’, which the reader of Plato can fill in by recent context, but Iamblichus’ reader cannot. Still, little of the idea or of the language is lost in Iamblichus’ careful paraphrase. The same holds for the later paraphrasing sentence, which corresponds to these sentences in the original (494 a 3–b 7): When I say this, do I persuade you at all into conceding that the orderly way of life is better than the one lacking in self-control, or do I not?— You do not persuade me, Socrates: the man who has filled himself up has no pleasure any more, but, as I was saying just now, that’s living like a stone when he’s been filled up and experiences neither joy nor pain; no, living pleasantly consists in this: having as much as possible flowing in.— Therefore isn’t it surely necessary that if there’s a lot flowing in, there should also be a lot going out, and that the holes for the e}uent should be big ones?—Certainly.—Now you’re talking about the way of life of some stone curlew, not a corpse or stone.

As we can see by comparing with his version (above), Iamblichus has deleted the initial exchange about whether Callicles is persuaded, and carefully paraphrased the rest of the dialogue, keeping the same words as much as possible, preserving them in unmodified forms whenever possible. But he also evidently felt it neater or in some other way necessary to eliminate here again the objection that Callicles has raised to Socrates, couched in terms of ‘stones and corpses’. So here is a loss of information, relative to the Plato text, albeit a small one. Generally there is more modification in these lines than we have seen before, since the original was in the form of a conversation, which does need more modification to render the prose that Iamblichus’ book requires. And yet we find no alteration of the line of thought, except in a detail deleted in order for the line of thought to be simplified. The rest of chapter 17, the bulk of it, consists in a slightly modified quotation of the speech by Socrates to (or at) Callicles on the 53 Reading at 492 e 7 κα Jν γε, suggested by Badham on the basis of the Iamblichus version, and adopted in Dodds’s edition, rather than κα )ς γε (all manuscripts), the reading adopted in Burnet’s edition. See E. R. Dodds (ed.), Plato: Gorgias. A Revised Text, with Introduction and Commentary [Gorgias] (Oxford, 1959), ad loc.

Created on 4 July 2005 at 14.34 hours page 220

220

D. S. Hutchinson @ Monte Ransome Johnson

topic of pleasure (sieves and jars), with just the same kinds of modification to erase the dialogue frame as we saw in the Theaetetus material of chapter 14 above: deletion of a content-neutral speech exchange (493 d 2–4) and deletions of formulae of address to Callicles (493 b 4, c 4–5, c 7–d 1, d 5–6). No other deliberate modifications take place in the chapter. The block is solid and pure, and the sequence is natural; Iamblichus moves on to Gorgias after Theaetetus and Republic 7 (not retracing his steps), and he quotes each block of text in order. As for accidental variants and the quality of Iamblichus’ quotations as a witness to the text of Plato, we need to postpone the analysis until we can pool the data from the Gorgias quotations in chapters 18 and 19 (subsections 2.5 and 2.7 below), since evidence of manuscript a¶liation becomes clearer over larger stretches of text. For now, the following comment by Dodds, in his landmark edition of Plato’s Gorgias, su¶ces to indicate that Iamblichus is a valuable witness: Iamblichus, the Neoplatonist philosopher, pillaged the Gorgias, along with other classical works, to provide material for his Protrepticus, an exhortation to the philosophical life. Sometimes he paraphrases and abbreviates; and even where he transcribes he does not present his quotations as quotations, but adapts them to the form of a ‘protreptic’, eliminating whatever might betray their origin in a dialogue. As a rule, however, he makes only such changes as are necessary for this purpose, so that the value of his evidence is not greatly impaired, though he has a bad name for negligence and lack of scruple.54

Dodds values Iamblichus as a witness to the text of Gorgias for all the same reasons as we found above for concluding that he is a valuable witness to the text of Theaetetus and Republic 7. 2.5. Chapter 18: quotations from Gorgias But for now let us carry on with our analysis, and consider what happens in chapter 18: 54 Dodds, Gorgias, 64, referring to H. Alline, Histoire du texte de Platon (Paris, 1915), 152, for the ‘bad name for negligence and lack of scruple’. It would appear that Alline was relying on C. G. Cobet, ‘Platonica’, Mnemosyne, ns 2 (1874), 241– 82, in which an alarming number of variants in the text of Iamblichus were listed unsystematically at 261–78; the methodological error in Cobet’s work was not to separate the accidental variants from the deliberate alterations of Iamblichus, which had the e·ect of exaggerating the impression of unreliability.

Created on 4 July 2005 at 14.34 hours page 221

Authenticating Aristotle’s Protrepticus

221

Chapter 18. But akin to this is another approach, which makes an inference by analogy from what appears clearly in the body to what is both good and bad for the soul, and separates us from the bad things and alienates our choice from the shameful, and calls on us to embrace the good things with all our strength, and gets our thought accustomed to the fine in every way. For example, if the a·ections and the diseases of the body are to be avoided, much more so are those of the soul, and if the disease of the body produces the kind of life not worth living for us, then much more so does injustice, being a disease of the soul, make life unendurable for the one who has it, and if the unbalanced condition of the body destroys in us the perfection of the soul, when that by which we live is in a faulty condition and at war with itself, it will not be possible to live out one’s life correctly. So let’s proceed to that point in the following way. Just as being organized and orderly makes our bodies good, and being disorganized makes them bad, similarly with the soul, too <. . . here Iamblichus continues with this modified quotation of Gorg. 504 a 8–505 b 12 . . .> So getting disciplined is better for the soul than lacking self-control, as most people think,55 and to be orderly and organized is superior to being disorderly and disorganized, hence in every way one should train for justice and self-control as opposed to the opposite states. And this must be our state as well. (Iambl. Protr. ch. 18, 113. 22– 115. 21)

The first comment that needs to be made is about the opening and closing sections in the voice of Iamblichus: whereas the closing is very similar to the closing of chapter 17 (a half-sentence of transition plus a short sentence of explicit protreptic), the opening section shows a phenomenon that we have not seen before, an extended opening discussion occupying several sentences. Typical of the voice of Iamblichus here are the following features: it is meta-textual, in that it points to a stretch of text while giving a reason to take it seriously (‘So let’s proceed to that point in the following way’, 114. 6–7); it is also repetitious, in that it paraphrases the argument which is just about to follow; it also uses long parallel constructions with distinctive vocabulary, referring to a line of argument as an ‘approach’ (Aφοδος) that alienates, embraces, and invites ("λλοτριοσα, "ντιλαµβ νεσθαι, παρακελευοµνη, 113. 22–7); and it expresses its points in hyperbolic and enthusiastic ways that are not strictly supported by the argument, e.g. ‘with all our strength’ 55 The text seems to be corrupt here, as this is precisely not the sort of thing that most people (ο> πολλο) believe. Des Places says it might be preferable, following Plato, to insert a ‘not’: ‘as most people <don’t> think’. Alternatively, we suggest the conjecture ο> παλαιο: ‘as the ancients think’.

Created on 4 July 2005 at 14.34 hours page 222

222

D. S. Hutchinson @ Monte Ransome Johnson

(113. 27), ‘in every way’ (113. 28), ‘destroys in us the perfection of the soul’ (114. 4). In other respects this chapter resembles chapter 17, in that it consists of a central target section, in which Iamblichus has slightly intervened to flatten to monologue, surrounded by sections before and after, which he has turned into monologue by a larger degree of modification. The main di·erence between these two chapters is that here Iamblichus has been obliged to modify more heavily overall, as the section he chooses to excerpt from Plato has no real speeches, only active dialogue between Socrates and Callicles (a rather intractable interlocutor), into which he needs to intervene more extensively. Whereas chapter 14 reworks material that was a real speech in Plato’s book with no real dialogue, chapter 18 reworks a real dialogue with no real speech, and chapter 17 reworks a real speech surrounded on both sides by dialogue. We can best appreciate the types and degree of intervention by making the following fully explicit comparison (504 a 8–d 7, trans. after Zeyl in Complete Plato): So if a house gets to be organized and orderly it would be a good one, and if it gets to be disorganized it would be a bad one?—I agree.—This holds true for a boat, too?—Yes.—And we surely take it to hold for our bodies, too?—Yes, we do.—What about the soul? Will it be a good one if it gets to be disorganized, or if it gets to have a certain organization and order?—Given what we said before, we must agree that this is so, too.—What name do we give to what comes into being in the body as a result of organization and order?—You mean ‘health’ and ‘strength’, presumably.—Yes, I do. And which one do we give to what comes into being in the soul as a result of organization and order? Try to find and tell me its name, as in the case of the body.—Why don’t you say it yourself, Socrates?—All right, if that pleases you more, I’ll do so—and if you think I’m right, give your assent; if not, refute me and don’t give way. I think that the name for the states of organization of the body is ‘healthy’, from which health and the other bodily excellences come to be in it. Is this so, or not?—It is.—And the name for the states of organization and order of the soul is ‘lawful’ and ‘law’, from which people become law-abiding and orderly, and these states are justice and self-control?—do you assent to this or not?—Let’s say so.— Hence that orator, the good and skilful one, will be looking at these when he applies to people’s souls whatever speeches he makes . . .

Next the version in Iamblichus, with the dramatic conversation stripped away (114. 7–22):

Created on 4 July 2005 at 14.34 hours page 223

Authenticating Aristotle’s Protrepticus

223

So just as if our bodies get to be organized and orderly they would be good ones, and if they get to be disorganized they would be bad ones, so too, in the same way, a soul which gets to be disorderly will be bad, and one which obtains a certain organization and order will be good. Now then, in the body the name of what results from organization and order is ‘health’, whereas, on the other hand, to what results in the soul from organization and order is given the name ‘lawful’ and ‘law’. For just as in the body the name for its states of organization is ‘health’, from which health and other excellences of the body come to be in it, so in the same way the states of orderliness of the soul are called ‘law’, from which we become organized and orderly people; and these states are justice and self-control. Hence those who take any care for their own souls will be looking at these when they apply to people’s souls whatever speeches they make . . .

This passage in the version of Iamblichus is a condensed paraphrase which lacks certain elements, not only the empty responses of Callicles, but also the banter of Socrates directed towards Callicles, as well as the analogical reasoning with which Socrates begins his line of argument (houses, boats, therefore similarly for bodies), and the back-reference to the earlier part of the dialogue. But all the essential elements, and almost all the words, that are to be found in the Plato passage are preserved in the version of Iamblichus, despite this significant truncation and compression. In the paraphrase that finishes the modified quotation we find the same combination of truncation and compression together with accuracy to Plato’s argument, thought, and words (505 b 1–12): And about the soul, my good friend, isn’t it the same way? As long as it’s corrupt, being foolish, lacking in self-control, unjust, and irreligious, it should be kept away from its appetites and not be permitted to do anything other than what will make it better. Do you agree or not?—I agree.—For this is no doubt better for the soul itself?—Yes, it is.—Now isn’t keeping it away from what it has an appetite for, disciplining it?—Yes.—So getting disciplined is better for the soul than lacking self-control . . .

Next, the modified version in Iamblichus (115. 11–17): And yet about the soul, too, it’s the same way: as long as it’s corrupt, being foolish, lacking in self-control, unjust, and irreligious, it should be kept away from its appetites and not be permitted to do anything other than what will make it better. For this is no doubt better for the soul itself. Hence keeping it away from what it has an appetite for is to discipline it. So getting disciplined is better for the soul than lacking self-control . . .

Created on 4 July 2005 at 14.34 hours page 224

224

D. S. Hutchinson @ Monte Ransome Johnson

This example needs no further comment, except to reinforce the point that Iamblichus does not modify the text in any way that is not necessary for his purpose of flattening to monologue. The more unlike a speech the Plato section is, the more modification he finds necessary; but he never makes more modifications than necessary. The centre of the chapter is a group of four comments by Socrates (504 d 5–505 a 9) united into one block of continuous prose mostly by the removal of one address to Callicles (504 e 6) and of four pairs56 of comments, in which Socrates demands that Callicles answer and he gives one-word answers (504 e 4–5, 504 e 9–505 a 1, 505 a 3–5, 505 a 9–11). And Iamblichus needed to bring about one more change, from singular to plural, in the verbs of 504 d 5–9, in order to generalize the point from what ‘that orator, the good and skilful one’ would say (in Plato’s version) to what ‘those who take any care for their own souls’ would say (in Iamblichus’ version); otherwise this comment would reveal the wider narrative frame of the Gorgias. So we see that chapter 18 shows a relatively high degree of modification in the quoted blocks; but here again, the blocks are pure and the sequence is natural, as Iamblichus continues to quote from Gorgias in the natural order. 2.6. Chapter 13: quotations from Phaedo This chapter displays a rather more complex mixture of elements and construction techniques. The most important new phenomenon is the use of two bridge passages to connect disparate excerpts from Plato’s Phaedo; and we shall see this phenomenon again in chapter 19, where Iamblichus uses three bridge passages to connect four excerpts from three disparate works. Another important new phenomenon is a page at the end consisting of a pastiche, with scraps of several Plato texts woven in, a new kind of literary texture which di·ers from the block quotations that we have seen so far; but this odd literary texture is confined to one page. Here is an outline of the greater part of the chapter, with indications of where Iamblichus is using block quotations, as well as the pastiche page at the end: 56 Not quite four pairs: all four answers are removed, and three of the four demands, but Iamblichus has retained, perhaps by oversight, the demand at 504 e 9–10 Aστι τατα;

Created on 4 July 2005 at 14.34 hours page 225

Authenticating Aristotle’s Protrepticus

225

Chapter 13. But if, with the conceptions about philosophy, we need to call in some other assistance as well as this one,57 let’s get going and let’s proceed in that way, starting with these. Other people may well be unaware that all who actually engage in philosophy correctly are practising nothing other than dying and being dead <. . . a version of Phaedo 64 a 4–69 d 2 . . .> as they say about the initiation rites: ‘many wave the wand of Bacchus, but true believers are scarce’, and these are, in my opinion, none other than those who have practised philosophy. If therefore philosophy alone, because of its nature, causes perfect virtue and purification of the soul, that alone is worthy of being embraced. But those who have not practised philosophy and passed away in absolute purity may not lawfully come to join the family of gods <. . . a version of Phaedo 82 b 10–84 b 3 . . .> the soul, coming to join its kin and its kind, will be released from the evils of the human condition. From this approach it is clear that philosophy brings release from human and bodily bonds and a liberation from the accidents of birth in time, and leads to that which truly is, and to a knowledge of truth itself, and purification of souls. But if in this above all things there is true happiness, we must cultivate philosophy most seriously, if we wish to be truly happy. And yet this point is also worth keeping in mind, that if indeed the soul is immortal, then it needs to be taken care of, not only over the course of the only time, this time, in which we call ourselves ‘living’ <. . . a version of Phaedo 107 c 1–d 5 . . .> at the very outset of the journey yonder. For the better soul dwells with the gods, traverses the heavenly sphere, and receives a better allotment; but the soul which has been in contact with unjust acts and is full of impiety and sacrilege goes to subterranean dungeons to receive the fitting punishment. For these reasons, one should do everything so as to take part in virtue and wisdom in this life, for fair is the prize and great is the hope <. . . a closing pastiche woven from Phaedo, Apology, and Menexenus . . .> Let’s say that such approaches, too, are ones that lead from these things to the exhortation to philosophy. (Iambl. Protr. ch. 13, 90. 16–100. 20)

The opening sentence needs no comment, and the closing pastiche page needs so much comment that it is best postponed until later (below, pp. 233–5). Instead, let us focus on the bridge passages, starting with the first, a single sentence at 96. 10–13: ‘If therefore philosophy alone, because of its nature, causes perfect virtue and purification of the soul, that alone is worthy of being embraced.’ It is a back-referring summarizing version of the intervening text, and it contains enthusiastic 57 Reading π τοτου with the manuscript, not Scaliger’s conjecture π τοτο, accepted by editors. It seems to refer to the last Aristotle excerpt that was cited by Iamblichus, just as "π τοτων seems to refer to the Plato excerpts that are about to come.

Created on 4 July 2005 at 14.34 hours page 226

226

D. S. Hutchinson @ Monte Ransome Johnson

language typical of Iamblichus: ‘perfect virtue’; ‘purification’; ‘embraced’. The second bridge passage (98. 25–30) is similar: this too is a back-referring summary with some neo-Pythagorean terminology adapted from Plato; but the most important thing to notice about it is that it is a succinct and e·ective expression of certain ideas that are common ground between Plato and the Pythagorean tradition of Iamblichus. Now let us look at the quoted blocks in this chapter, starting with the shortest, the third one. The only change is at the very beginning of the excerpt, where Iamblichus suppresses the dialogue frame. In Plato’s dialogue Cebes had fully accepted the proof o·ered by Socrates of the immortality of the soul, whereas Simmias still harbours some lingering sceptical doubts, and Socrates acknowledges these doubts in his next comment (107 c 1–2): ‘But this much anyway, gentlemen, is right to keep in mind, if the soul is immortal, etc.’ This is a comment which Iamblichus minimally transforms (99. 2): ‘And yet this point is also worth keeping in mind: that if indeed the soul is immortal, etc.’ We can use this short quotation to check the accuracy of Iamblichus’ quotation against other witnesses, including Stobaeus, who quotes it as well. Whereas we see in Stobaeus three corruptions,58 and each of the main manuscripts has a few,59 Iamblichus has only one trivial one: 107 c 3 τουτου µονον] µονου τουτου. On the basis of this little snapshot, it would appear that the version in Iamblichus is just as carefully preserved as the version in the main Plato manuscripts, and better preserved than some other ancient witnesses. This appearance is confirmed below (pp. 230–3), on the basis of the whole data set. Let us carry on and proceed to scrutinize the second excerpt: the passage from 82 b 10 to 84 b 3, the conclusion of Socrates’ first series 58 Stob. 1. 49. 58 ( = i. 434. 6–18 Wachsmuth–Hense): 107 c 1 το [δε] γ9 c 6 [αν] (with T) c 6 κακοις] κακως d 7 οσπερ] ωσπερ. 59 Corruption in T—107 c 6 [αν] (with Stob); corruptions in W—107 c 1 [γ9] c 3 εν ω<περ>; errors in W, apparently attempted improvements—107 c 2 αθανατος <εστιν> c 5 αµεληει<εν>; error in B—107 d 4 και λεγεται µεγιστα] και µεγιστα λεγεται. (In this latter case, we prefer the reading supported by all the other witnesses but rejected by the two OCT editions. When the reading και λεγεται µεγιστα is adopted, the passage construes as follows: ‘For the soul arrives in Hades having nothing else but its education and upbringing, which are indeed said greatly to help or harm the deceased, right at the start of his journey yonder’ (107 d 2–5). The di·ering functions of κα in the two versions bring about a perceptible di·erence in the speaker’s attitude, and the version of B is stemmatically isolated and not preferable.)

Created on 4 July 2005 at 14.34 hours page 227

Authenticating Aristotle’s Protrepticus

227

of arguments for the immortality of the soul, finishing with the unforgettable image of the soul repeatedly unpicking, like Penelope, the threads whose tissues attach us to the body. Iamblichus has brought about no unfamiliar kinds of change to this passage; he suppresses all seven of the speech exchanges between Socrates and Cebes, sometimes with slight concomitant modification to the first sentence of the speech of Socrates,60 and he suppresses all five personal addresses.61 The only other modification e·ected by Iamblichus is, typically, at the beginning of the excerpt (82 b 10), where he modifies the particles for a smoother transition: δε] γαρ. As for the variants or accidental changes, let us combine these with those from the remaining excerpt of 64 a 4–69 d 2, to which we now turn. The first of three Phaedo excerpts is Iamblichus’ version of the ‘second apology’ of Socrates. Iamblichus gives literal quotations of both the beginning and the end of this passage, together with Socrates’ three main speechlets (65 e–67 b, 67 e–68 b, 68 e–69 d), but the rest of the ‘second apology’ required him to modify his quotations, as Plato has carefully written this up as primarily a private discussion, a dialogue between friends, unlike the ‘first apology’, which was an extended public speech addressed to people who did not know him, in a situation that did not permit discussion. So, just as in the Gorgias excerpts, we see here a relatively high degree of intervention on the part of Iamblichus. And these interventions are all of a familiar sort, except for two cases of a new phenomenon: a special kind of bridge passage which gives a quick paraphrase of an extended section in Plato’s text. They have the e·ect of moving the reader onwards quickly to the next quotable passage, for which reason we can think of them as ‘fast-forward’ paraphrases. Let us focus now on these ‘fast-forward’ paraphrases, to see how they are related to the underlying text: Other people may well be unaware that all who actually engage in philo60 Simple suppression of speech exchange at 82 c 9–d 2, 82 d 8–9, 83 c 8–10, 83 e 4, 83 e 6–84 a 2; with slight modification at 83 c 4–5 οτι <δη> and 83 d 3–4 οτι] επει. 61 To Socrates (82 c 9, 82 d 8); to Cebes (82 a 7, 83 e 5); to Cebes and Simmias (82 c 2). A curious oversight in Iamblichus’ programme to transform the stretch of Phaedo to monologue is the accidental retention of Mπερ ο=ν λγω (at 82 e 7), ‘what I mean is . . .’, an oversight which leaves a clue that the original text was a dialogue. The imperfection of Iamblichus’ transformation means that we might find surviving traces of dialogue in Iamblichus’ quotations of the Protrepticus, if Aristotle’s work was a dialogue.

Created on 4 July 2005 at 14.34 hours page 228

228

D. S. Hutchinson @ Monte Ransome Johnson

sophy correctly are practising nothing other than dying and being dead <. . . a version of Phaedo 64 a 4–65 d 2 . . .> the soul of the philosopher utterly disdains the body and flees from it, seeking rather to become itself, alone by itself. But this will be especially clear from a contemplation of the Forms; for justice itself, and beauty, and good, and all things on which we set the seal of ‘what exists’, no one has ever seen with the eyes, nor by any of the other senses that act through the body, but he who is prepared to focus his mind most, and most accurately, on each topic of enquiry—he will come the closest to knowing each of them <. . . a version of Phaedo 65 e 2–67 d 10 . . .> the practice of philosophers is just this: a release and a separation of a soul from a body. Therefore philosophy, since it provides us with the greatest good, namely a liberation from the bonds in which the soul is bound since its birth in time, should be pursued with the utmost seriousness. This so-called ‘bravery’ is to be expected most of all in those who are in that condition <. . . a version of Phaedo 68 c 5–69 d 2> as they say about the initiation rites: ‘many wave the wand of Bacchus, but true believers are scarce’, and these are, in my opinion, none other than those who have practised philosophy. (Iambl. Protr. ch. 13, 90. 18–96. 10)

The first of these fast-forward paraphrases boils down the ideas in 65 d 4–e 1, where Socrates opens a new phase of his ‘second apology’ by referring to Plato’s theory of Forms. In the original it comprised seventy-two words, but in the boiled-down version only forty. The paraphrase is accurate to the thought, but not to the expression of it. Iamblichus not only thoroughly suppresses the dialogue, but he also adds a meta-textual comment: ‘But this will be especially clear from a contemplation of the Forms.’ When we consider this fast-forward bridge paraphrase we must recognize that, without the survival of Plato’s full text, we would probably not have recognized it as partly paraphrase, and not entirely quotation. Now let us look at the second ‘fast-forward’ paraphrase. This time, Iamblichus chooses to condense a larger section of text, including a longish speech by Socrates (67 e 5–68 b 6) that Iamblichus could have quoted without modification. But he carefully steps around the whole section (67 d 12–68 c 4), as it contains both a backward reference to an earlier claim in the frame narrative and also something else that is awkward for Iamblichus: here it is not so much an exhortation for people to become philosophers as an exhortation for philosophical people to embrace death cheerfully and confidently. In all, Iamblichus suppresses this 158-word speech as well as the preceding exchange of speeches (thirty words) and the succeeding two speech exchanges (seventy words). This ends

Created on 4 July 2005 at 14.34 hours page 229

Authenticating Aristotle’s Protrepticus

229

up being just the following single-sentence fast-forward paraphrase comprising twenty-one words (94. 17–19): ‘Therefore philosophy, since it provides us with the greatest good, namely a liberation from the bonds in which the soul is bound since its birth in time, should be pursued with the utmost seriousness.’ The compression ratio here is more than 12 : 1, whereas for the first fast-forward paraphrase it was less than 2 : 1. This should teach us the lesson that if we find a fast-forward paraphrase in the Aristotle section, we shall probably not be able to make any reliable estimate of how much of Aristotle’s text Iamblichus chose to speed through with condensed paraphrase. When we isolate and remove these two fast-forward paraphrases, what remains is a set of quotations modified in ways that are all familiar to us by now, modified quotations copied carefully from an ancient manuscript of Plato. The ‘second apology’ in Plato’s Phaedo is a very complicated and artful construction, in which Plato transforms this apology in genre from a rhetorical monologue delivered to strangers into a persuasive conversation carried on between intimate friends and associates. The resulting dialogue and banter in Plato requires Iamblichus to engage in a high degree of transformation, of several types, severely in certain passages. Readers who begin their study of the Platonic excerpts in chapters 13–19 with this passage (the first in chapter 13) would naturally get an impression of a bewildering quantity of alterations, an impression which leads to the feeling that Iamblichus’ version is very free. But this impression would be misleading, and the feeling mistaken: Iamblichus’ version does not take liberties, and is in fact a highly disciplined exploitation of Plato. The many (sometimes complex) modifications in Iamblichus’ version are executed with surgical precision for precisely one purpose—to reduce a literary dialogue to dogmatic monologue while preserving all the thought and as much as possible of the language of Plato. The simplest kinds of modification here are simple excisions of Socrates’ addressing of Simmias by name (65 a 4, 66 a 7, 67 b 3, 67 e 5, 68 c 5, 69 a 6) and simple excisions of the bare agreements uttered by Simmias.62 Iamblichus suppresses every remark by Simmias, and most of these cases require modification of the beginning of the sentence of Socrates that follows the suppressed remark of 62 ‘That’s right’, at 65 c 10; and ‘it seems so’, at 65 a 3 and 67 d 11, this latter stimulated by Socrates’ ‘isn’t that right?’, a question also excised by Iamblichus.

Created on 4 July 2005 at 14.34 hours page 230

230

D. S. Hutchinson @ Monte Ransome Johnson

Simmias.63 Sometimes Iamblichus suppresses both of two adjacent remarks by Simmias, but he usually takes care to transform and preserve the contents of the intervening comment by Socrates,64 with one exception, in which the intervening comment by Socrates is unusable for Iamblichus, as it refers back beyond the beginning of the ‘second apology’ and recalls the narrative frame of Plato’s work. The main remaining type of modification in chapter 13 is the transformation of the many rhetorical questions into assertions. An instructive and simple example: ‘Do sight and hearing in human things have any truth? Aren’t even the poets always harping on such themes anyway, that . . .’ (65 b 1–3). With surgically minimal changes,65 the translated result is this: ‘Sight and hearing in human beings do not have any truth, for the poets are always harping on such themes, that . . .’ (91. 19–21). The only remaining significant change brought about by Iamblichus to Plato’s ‘second apology’ is at the beginning, where he suppresses a wry discussion between Simmias and Socrates about how most people would agree that death is a deserved fate for philosophers (64 a 6–c 3), a discussion that Iamblichus cannot use, as it refers backwards to Plato’s narrative frame. The only remaining detail to note is tiny: at the very beginning of his exposition (90. 18), Plato’s connecting particle γ ρ has no function, so Iamblichus omits it. The foregoing analysis has accounted for every one of the deliberate changes that Iamblichus made to the text of Plato’s Phaedo. Despite the great complexity of Plato’s ‘rhetorical dialogue’ construction, which forces a correspondingly complex set of transformations on Iamblichus, the underlying principles and purposes of all these changes are simple and familiar to us from other chapters we have studied so far. The blocks are pure and the sequence is natural, however, despite this relatively high degree of intervention. If the version of Plato’s Phaedo that we find in Iamblichus was derived from a fair copying of an ancient manuscript of Plato, as was 63 The simplest case is 65 e 6–7, in which Iamblichus deletes Simmias’ ‘very much so’ and alters the next sentence of Socrates from a question to a statement: ‘then would that be done . . .’ to ‘then that would be done . . .’. Many other cases, of slightly more complexity, are to be observed: 64 c 8–d 2, 64 e 2–4, 64 e 7–8, 65 a 3–4, 65 a 8–10, 65 b 6–9, 68 c 7–8, 68 d 1–2, 68 d 4–5, 69 a 5–6. 64 64 d 5–9, 65 c 1–4, 66 a 7–b 1, 67 d 3–6; the exception is 67 b 3–c 4. 65 65 b 1 αρα] ουκ b 2 η τα γε] τα γαρ. Similar deft transformations of rhetorical questions to assertions can be observed at 64 c 4–8, 64 d 2–4, 65 b 9, 64 d 11–12, 65 e 1–2, 67 c 5–d 2, 68 c 5–6, 68 c 10–12.

Created on 4 July 2005 at 14.34 hours page 231

Authenticating Aristotle’s Protrepticus

231

Iamblichus’ version of other texts of Plato, then we should expect it to behave like a manuscript and show typical patterns of erroneous readings; and this it does. First, in respect of reliability, the version of Iamblichus, with eighteen errors,66 is comparable to those of the main manuscripts: B with fifteen errors,67 T with eight,68 and W with twenty-two.69 By contrast, an ancient manuscript, the Petrie papyrus (P. Lit. Lond. 145),70 is much denser in errors, with at least twenty showing up in a much shorter stretch of text (82 d–84 a). So the manuscript of Plato’s Phaedo that Iamblichus used seems to have been of a good quality, not as erroneous as the Petrie papyrus, and certainly no less reliable than it. Yet the true value of a witness to a text is not its reliability, but its ability to preserve good ancient readings lost in other witnesses; and this ability depends on its relative genetic position. Can we find evidence of this? Yes, with care we can discern a complex relationship. We find ten cases of B agreeing with W against T agreeing with Iamblichus, and seven cases of TW against B agreeing with Iamblichus; but virtually no cases71 of BT vs. W and Iambli66 The errors of the Iamblichus manuscript are two reorderings of short groups of words (83 c 7, 107 c 3), five corruptions involving articles or particles (66 b 3, 67 a 8, 69 a 3, 83 a 4, 83 c 6), and eleven other cases of small-scale and normal corruptions (65 c 8, 66 b 2, 66 b 5, 66 c 4, 66 e 3, 66 e 5, 69 a 3, 82 c 4, 82 c 4, 82 d 3, 83 e 1); not counted are several cases where the change may have been deliberate at 65 b 9, 68 d 8, 69 b 7, 69 b 8; not counted is the case which is erroneously reported as a variant in the 1995 OCT, whereas it is actually a deliberate change (82 d 2 [/& Κεβες]), as well as various orthographical di·erences, especially iotacism and initial ξ for initial σ. 67 The errors of B are five errors with particles (64 d 5, 65 a 4, 67 d 9, 68 d 2, 84 b 3), seven other normal corruptions (65 c 5, 67 a 6, 67 d 5, 69 a 10, 69 b 6, 69 b 8, 82 c 3), and three other cases where B’s erroneous reading was selected by Burnet in his OCT edition (the first two were reversed in the 1995 OCT): 69 b 6 [και] B, Burnet 69 c 2 [η] ανδρεια B, Burnet 107 d 4 λεγεται µεγιστα] µεγιστα λεγεται B, Burnet, and 1995 OCT; see above, n. 59. 68 The errors of T are two homoioteleuta and six small-scale corruptions: 64 b 9 θανατου [και οιου θανατου] 83 c 6 ησθηναι [σφοδρα η λυπηθηναι]; and 66 b 3, 66 c 2, 66 d 5, 67 b 4, 69 a 6, 107 c 6. 69 The errors of W are eight cases of attempted improvement or correction (65 e 8, 66 c 8–9, 68 d 2, 83 a 4, 83 b 4, 83 e 6, 107 c 2, 107 c 3), seven cases of reorderings of groups of two words (66 b 4) or three words (65 c 5–6, 68 e 7, 69 b 2–3, 83 b 6, 83 d 8, 83 d 9), and seven cases of other small-scale corruptions (68 c 8, 83 a 6, 83 c 2, 83 d 2, 84 a 5, 107 c 1, 107 c 5). 70 A. Carlini and C. Strachan, Studi dell’Academia Toscana di Scienze e Lettere, 129 (1992), 147–67. 71 69 b 8 εχη BT+Stob.] εχει I+W and C (Tub. gr. Mb 14), an easy mistake that evidently happened more than once 67 a 1 η ψυχη εσται BT] εσται η ψυχη I+W+ Plutarch; both are possibly correct, and the corruption (in either direction) could easily have happened twice.

Created on 4 July 2005 at 14.34 hours page 232

232

D. S. Hutchinson @ Monte Ransome Johnson

chus. When we look at the seven cases of Iamblichus+ B vs. TW, we find that these all seem to be mistakes of TW,72 so this gives no evidence of genetic a¶liation between B and the manuscript used by Iamblichus, only an a¶liation between T and W. The ten cases of Iamblichus and T vs. BW are of a di·erent nature: whereas two of these errors are shared by B and W,73 the others are Iamblichus+ T in probable or possible error, often of a subtle nature, so that they are all such as to appear to be candidate readings to a competent editor (and in several cases the 1995 OCT has chosen the alternative that Burnet rejected).74 The likeliest explanation for this pattern of agreement of Iamblichus with B at times and with T at times (but not with W) is that T is related by contamination to the Plato manuscript used by Iamblichus; an ancestor of T had two ancestors: a descendant of the common ancestor of BTW, and a manuscript in the family of the one used by Iamblichus. This would explain why there are no egregious errors shared in common between them, only undecidable or disputable or subtle errors that might slip through a process of correction. The oldest Plato manuscript B seems also to have benefited from contamination in its descent, which explains why its agreements with I seem always to be correct, and with TW showing an error that was removed in the B family by contamination and also absent from the manuscript used by Iamblichus. All the data are easily explained on this genetic hypothesis, including three cases where B alone seems to preserve the correct reading (66 b 4, 82 c 1, 83 d 10); all the variants are explained, down to the last iota.75 72 65 c 7 µηδε] µητε TW 66 b 6 <του> τοιουτου TW 68 d 6 µεγαλων] µεγιστων TW 68 d 12 αλογον] ατοπον TW 68 e 3 γε <που> TW 68 e 4 οµιον <ειναι> TW. The only apparent case of B agreeing with Iamblichus in error seems to us to be incorrect selectio on the part of the OCT editors: rather than accepting τινες into the text and reporting the variant as 69 c 4 ου φαυλοι [τινες] ειναι I+B+Stob., as Burnet and the 1995 editors do, it would be better to regard the τινες as an insertion: ου φαυλοι <τινες> ειναι TW. 73 82 c 3 οι φιλοσοφοι I+T+ Π2] οι φιλοσοφουντες BW 83 c 2 κακων I+T] κακον BW. 74 65 e 9 τινα BW+1995 OCT] om. I+T, del. Burnet 66 a 7 και BW] om. I+ T 67 d 1 δεσµων BW+1995 OCT] <εκ> δεσµων T+I; del. εκ Burnet 69 c 8 [ως] I+T 83 b 7 και λυπων και φοβων BW+1995 OCT] και λυπων T+I, del. και φοβων Burnet; three cases where both OCT editions seem to us to have made the wrong choice: 64 d 4 [τε] και ποτων BW, both OCT edd. 64 c 7 χωρις δε την ψυχην <απο> του σωµατος BW, both OCT edd. 65 b 3 αει ηµιν I+T] ηµιν αει BW, both OCT edd. 75 64 d 4 σιτιων T seems to have been a (successful or unsuccessful) correction for σιτων I+BW, or else the σιτιων in T and Iamblichus’ Plato manuscript lost its iota again en route to being copied to the Iamblichus MS F.

Created on 4 July 2005 at 14.34 hours page 233

Authenticating Aristotle’s Protrepticus

233

Given the genetic position of the manuscript that Iamblichus used, we can expect it to play an occasionally valuable role in establishing the text of Plato; and this it does.76 This has been the most painstaking study of any chapter’s genetic a¶liations, in that we used both OCT editions and we independently judged which variant to select, not necessarily following the judgement of the editor. The result of this study indicates a definite place in the ancient stemma of the manuscripts of Plato’s Phaedo for the manuscript used by Iamblichus. It is the same place in the stemma that had been discovered for it in A. Carlini’s study of the manuscript tradition of Plato’s Phaedo.77 We could not have reached this result twice, independently, unless Iamblichus’ version had been a literal manuscript-to-manuscript fair copy. Now it is time to tackle the page-long concluding section of chapter 13, in which we see a new phenomenon, a sort of pastiche of passages from Plato, woven together in a new way. In the first of three paragraphs (below) there are two quotations from later in Plato’s Phaedo, perfectly literal, but then in the second paragraph Iamblichus appears to quote from memory parts of a longer section from Plato’s Apology, and finally in the concluding paragraph strings together two short pithy comments, one from later in the Apology, the other from the Menexenus. This is how it looks in translation, with the words not borrowed from Plato in bold for increased contrast: 76 The most significant case is probably 83 b 7–8, where the original text πειδ ν τις σφ+δρα SσθH T λυπηθH T φοβηθH T πιστηµσHη is vulnerable, because of the homoioteleuta, to accidental loss or transposition; this text is preserved in Iamblichus and Stobeus and Π2, but corrupted in three separate ways in BWT, as the 1995 OCT editors correctly have it: [η λυπηθη] B; και λυπηθη W; η φοβηθη η λυπηθη T (Burnet had chosen the transposed word order of T, and then deleted the words missing in B). Similarly at 83 c 6 Iamblichus and Π2 preserve the original reading ησθηναι σφοδρα η λυπηθηναι in the face of a transposition in BW and an omission in T. At 84 b 1 Iamblichus and B preserve the cluster of words ο:εται οUτω δεν in the original order, it seems, against two or three independent transpositions of order, in W, T, and Π2. At 69 a 8, where Burnet had printed [και] µειζω with the note και om. I+Stob., the 1995 editors print, more perspicuously, µειζω, with this apparatus entry: µειζω I+Stob.] και µειζω BTW, another case where the version of Plato’s text consulted by Iamblichus was free, it seems, from a corruption common to the three main manuscripts of Phaedo. Finally, we may note two cases involving a single letter, where Iamblichus preserves the correct reading against BTW: 69 c 3 κινδυνευουσι I+ Stob.] κινδυνευωσι TW and probably B 65 a 5 µετεχει I] µετεχει<ν> BTW. 77 A. Carlini, Studi sulla tradizione antica e medievale del Fedone [Tradizione del Fedone], (Rome, 1972), 98; the two results are independent, because we did not consult Carlini’s work until after we had finished our own analysis.

Created on 4 July 2005 at 14.34 hours page 234

234

D. S. Hutchinson @ Monte Ransome Johnson

For these reasons, one should do everything so as to take part in virtue and wisdom in this life, for fair is the prize and great is the hope. And for these reasons any man should be confident for his own soul who in his life said goodbye to the other pleasures, the ones concerning the body and its ornamentation, thinking of them as being both alien and bringing about rather the other result, but devoted himself to the ones concerning learning, ornamenting his soul not with an alien thing but with its own ornaments: self-control, and justice, and courage and liberality and truth, thus awaiting his journey into Hades, to make the journey whenever Fate calls. Now then, these things being so, one should not take care of as much wealth as possible, nor honour and glory, but understanding and truth, and the soul, so it can be the best possible. For nor should one be setting the lowest value upon the most precious things, or rating inferior ones more highly. And nor should one care for one’s body or for one’s wealth so intensely as for the greatest possible well-being of one’s soul. For it is not wealth that produces goodness; rather, it is from goodness that wealth and all other benefits for human beings accrue to them in their private and public life. One should fix one’s mind on this one single truth: nothing will be able to harm a good man, either in life or in death, nor are his a·airs neglected by the gods, so that all the good things that conduce to success are dependent on himself, and the one whose preparation makes him closest to this would be the one who lives out his life in the greatest possible happiness. Let’s say that such approaches, too, are ones that lead from these things to the exhortation to philosophy. (Iambl. Protr. ch. 13, 99. 20–100. 20)

While the middle paragraph is a decent paraphrase of the main ideas of Socrates’ self-reported ‘usual sort of message’, the first paragraph juxtaposes against each other two (exact) quotations that were separate in Plato’s Phaedo (‘one should do everything, etc.’ 114 c 6–9; ‘for these reasons, etc.’ 114 d 8–115 a 3), and the third paragraph juxtaposes against each other two quotations that were originally from separate works (‘fix your minds, etc.’ Ap. 41 c 9–d 3; ‘all the good things etc.’ Menex. 247 e 7–8). In other words, these are not pure blocks of quotation, and we must admit that Flashar is correct to say that ‘the reader of Iamblichus who doesn’t know the Phaedo could hardly be in a position to notice a gap here’.78 If this sort of phenomenon were to take place in the Aristotle section, then we would be unable to disentangle the words of Iam78 Flashar, ‘Platon und Aristoteles’, 61.

Created on 4 July 2005 at 14.34 hours page 235

Authenticating Aristotle’s Protrepticus

235

blichus from those of Aristotle in any reliable way. We cannot rule out the possibility of this sort of passage cropping up in the Aristotle chapters, even though it happens only once in all the Plato chapters. This is the only page in the Plato chapters which justifies the scepticism and uncertainty which has surrounded the recovery of Aristotle’s work out of Iamblichus’; the general rule of Iamblichus’ approach, however, is quite di·erent, consisting as it does of pure quotations or quotations modified with care and discipline. 2.7. Chapter 19: quotations from Gorgias, Menexenus, and Laws 2 Finally, we turn to the last chapter, in which Iamblichus quotes a third passage from Plato’s Gorgias and then two from Menexenus and another from Laws 2, before his closing remarks. This is how it looks in outline: Chapter 19. Someone might also make distinct progress on the subject on the basis of goods in the soul, by indicating the one of them that is perfect and in charge of happiness. For to the degree that the soul surpasses the body, to a still higher degree are its goods superior, and the bodily ones are despicable, while those of the soul will be honourable and dignified. And so the good is not the same as what the pleasant is <. . . a version of Gorg. 506 c 6–508 a 8 . . .> for you neglect geometry. So in general this is how it is with the happy life. And what follows next is that people choose to die a noble death instead, if they have the power to live on ignobly, rather than subject children and descendents to reproach <. . . a version of Menex. 246 d 1–247 b 7 . . .> and public recognition on one’s own part. For it’s from one’s own resources that one should acquire79 all the things that are good and bring one fame and happiness. ‘Nothing too much’ has long been thought an excellent adage <. . . a version of Menex. 247 e 5–248 b 4 . . .> neither too much distressed nor too much afraid if dying should be in the present, or su·ering any other human frailty. For it is indeed necessary to hold this opinion earnestly, that the good man, because he is self-controlled and just is successful and happy <. . . a version of Laws 2, 660 e 2–661 c 5 . . .> but if he survives, for only the briefest possible time, the evil is less . . . Then, if the goods are most attached to the life of virtue, and it’s then that they are goods in reality, and enjoyment accompanies doing philosophy only, then for the sake of all these things this life should be chosen by those who wish to be successful in reality. (Iambl. Protr. ch. 19, 115. 22–121. 6)

By now, the voice of Iamblichus will seem familiar to us, in both the opening and the closing sentences, as well as in the three in79 Conjecturing χρσθαι for -ρχεσθαι.

Created on 4 July 2005 at 14.34 hours page 236

236

D. S. Hutchinson @ Monte Ransome Johnson

tervening bridge passages. To di·ering degrees, each of these five interventions preserves some material from Plato, but their main function is to provide connections between passages, often by heavy use of thin inferences. When we consider the deliberate transformations brought about to the Gorgias passage by Iamblichus, we must keep in mind that it was already transformed by Plato. This is an odd passage in which Socrates speaks both parts of a sham dialogue, after Callicles baulks and drops out. Callicles’ only contribution is ‘Say on, sir’ (507 a 4). This is removed by Iamblichus, but he had to invest far more e·ort than this to transform the dialogue-reported-in-monologue into plain one-dimensional monologue. This e·ort involved him in a moderate amount of the usual sorts of transformation: he changes a rhetorical question and excises the direct addresses to ‘my friend’ or ‘Callicles’ (506 c 7–8 and 507 a 3, 507 c 1, 507 e 6, 508 a 3), and he depersonalizes and smoothes out the beginning and the end of the passage (506 c 6–8, 508 a 7–8), as well as the beginnings of a few sentences (506 e 4, 507 a 5). Iamblichus suppresses all the unproductive contributions80 made by the imaginary interlocutor (Socrates) who is being taken by the hand of the imaginary philosopher (Callicles) in the dialogue-reported-in-monologue of Socrates. The remaining modifications are all clearly motivated by the need felt by Iamblichus to depersonalize the discussion: he suppresses three speech exchanges, i.e. three pairs of comments,81 and he excises two comments whose only function in Plato’s text is to remind readers that Socrates and Callicles are locked in a stalemate of opposing convictions.82 On the other hand, he apparently felt no need to excise the backward reference at 507 c 1 (‘as we have explained’), as this would not necessarily jar with the readers of Iamblichus’ book. Where there is no clear need to modify the quotation, Iamblichus leaves it unmodified. With this analysis, we have finished with all three of the Gorgias 80 506 c 9, 506 d 2, 506 d 4, 506 e 4, 506 e 5, 507 a 1, 507 a 7, 507 b 4; the only odd case is at 507 a 7, where Iamblichus suppresses π νυ γε, but replaces it with πολλ. "ν γκη, the previous comment of the interlocutor at 507 a 1; this would appear to be a mistake made in the course of a deliberate modification. 81 506 d 8, 507 a 2–4, and 506 e 2, with the content of this latter exchange replaced by an empty confirmation invented by Iamblichus: ‘as someone who argued decently would say’. 82 507 c 5–7 ‘the one you were praising’, 507 c 8–9 ‘so this is how I set down the matter and I say that this is true’; 508 a 4–6: ‘I believe you don’t pay attention to these facts, etc.’

Created on 4 July 2005 at 14.34 hours page 237

Authenticating Aristotle’s Protrepticus

237

quotations, in chapters 17, 18, and 19, and we can now pool together the variants brought to light by our collations to determine the genetic a¶liation of the Gorgias manuscript used by Iamblichus. Our results turn out to be extensions of the ones noted by Dodds in his edition. The text used by Iamblichus ‘seems closer to the second family than the first; it shows striking agreement in error with F at 493 b 5 and at 505 b 4 and b 7, beside which a minor agreement in error with BTW at 492 e 8 is less probative’.83 Adding to this is more evidence in Dodds’s apparatus criticus of significant agreement in error between F and Iamblichus,84 buttressed by a more complete account of the insignificant agreements in error between Iamblichus and other witnesses.85 So the text used by Iamblichus does indeed belong to the second family, and could have been represented in Dodds’s stemma (Gorgias, p. 67) as finding its place there. The consequence of this location in the stemma means that we should expect that Iamblichus is capable, at times, of being the only variant carrier with a correct reading, when F has the same (or a di·erent) erroneous reading as BTW. And this does happen; not only does the witness of Iamblichus often support the correct reading against two or more manuscripts,86 but also at one place (492 e 7), as Dodds pointed out, ‘his paraphrase points to what I believe to be the true reading, lost in the direct tradition’, and at another place he o·ers a reading which Dodds finds preferable to that of all four main manuscripts: 506 e 1 [τι] BTW F. This indefinite τι is definitely not insignificant. Moving on to the Menexenus citations, we notice two chunks of earnest rhetoric taken from the inner speech that we have to imagine the dead soldiers declaiming from beyond the grave, in that absurdly patriotic funeral oration that Socrates says (236 b – 83 Dodds, Gorgias, 64. 84 493 a 6 σικελος BTW] σικελικος I+F+Stob.+Olympiod. 505 b 11 [η] I+F 507 e 5 δε BTW] γαρ I+Stob.+ Π2: om. F. Also adding some weight to the evidence of a¶liation between F and I are several cases of agreement between them in a correct or preferred reading, against BTW in error, either between the two of them (493 e 7 and e 8) or with Stobaeus (493 a 1 and b 1) or with Π1 (504 e 7 and 505 a 6). 85 The three cases of agreement in error between I and BTW could well be the result of correction in F (492 e 8, 493 a 7, 506 d 6); and the other three cases of agreement in error with other witnesses are scattered and insignificant: 507 c 9 δε] δη I+Stob. 507 e 2 ταυτα[ς] I+B λυσιτελει<ν> I+ Π2. 86 493 a 1 I+F+Stob] BTP 493 b 1 I+F+Stob] BTP 493 c 4 I+Tf] F; BP; Wb 493 e 7 I+F] BTW 493 e 8 I+F] BTW 504 e 7 I+FΠ] BTW 504 e 8 I+ FΠ] BTW 504 e 8 I+FBΠ] TW 505 a 3 I+FΠWP] BT 505 a 6 I+FΠ] BTW.

Created on 4 July 2005 at 14.34 hours page 238

238

D. S. Hutchinson @ Monte Ransome Johnson

c, 249 d–e) he got from the real author of Pericles’ funeral oration, namely his mistress Aspasia. At the end of this inner speech, Socrates says (247 e), ‘Children and parents, the dead commanded me to report those words, and I report them with all my heart.’ Of those words, Iamblichus reported only some, namely the address (246 d 1–247 b 7) from the fallen fathers to their sons, exhorting them to shun wickedness, to vie with them for glory, and not to waste the treasure of honour accumulated by their ancestors, as well as a small part (only) of the address from the fallen sons to their parents, namely, about a dozen lines of preaching (247 e 5– 248 b 4) that apply the Delphic apophthegm ‘nothing too much’ to the grief of their parents. Nothing of the material skipped over from 247 c 1 to e 5 is visible through the bridge passage, which draws its repetitive content rather from the quotation it introduces; it is only a bridge, not a ‘fast-forward’. Even though this inner speech is already a monologue, it still requires some transformation on the part of Iamblichus so that it is no longer a speech from beyond the grave from dead men hectoring their orphan sons (and bereaved parents), but an impersonal exhortation to humans in general. Iamblichus suppresses references to ‘we’ and ‘you’ and changes from second-person plural ‘you’ to an indefinite third-person τις or ‘someone’.87 He also suppresses a rhetorical claim about the fathers of the deceased and an assertion of the deceased speakers’ attitude.88 But when all is said and done, the most significant and the most typical modification was the one made at the beginning (above, p. 235) to depersonalize the start of the inner speech (246 d 1–3); compare the version above with the original version in Plato: ‘Sons, that you are sons of good fathers is revealed by the present state of a·airs: we who can live on ignobly choose to die a noble death instead, rather than subject you and descendants to reproach.’ Here again we see a di·erent set of techniques of modification, but always limited to a limited goal: flattening dialogue to monologue and personal discussion to impersonal statement. We are closing in on the end, the last quotation in the last Plato chapter, but before we get there, let us first study the way the chapter comes to its final conclusion in the voice of Iamblichus: 87 246 d 4, 246 d 8, 247 a 3, 247 a 4, 247 a 7, 247 b 1 (twice). 88 248 b 1–2 and 246 d 5–6, modifying Plato’s ‘we consider it to be’ to ‘for it is’.

Created on 4 July 2005 at 14.34 hours page 239

Authenticating Aristotle’s Protrepticus

239

And he who lives in an unjust and o·ensive way will of necessity live in a disgraceful way, and if disgracefully, then badly; so that he will also live in a way that’s unpleasant and not beneficial to himself, for when we declare that the most pleasant life is the same as the life said by the gods to be best, what we say is most true. Then, if the goods are most attached to the life of virtue, and it’s then that they are goods in reality, and enjoyment accompanies doing philosophy only, then for the sake of all these things this life should be chosen by those who wish to be successful in reality. (Iambl. Protr. ch. 19, 121. 1–6)

This last concluding passage is composed of two sentences of which the first half of the first one is an accurate condensed paraphrase of the line of thought to which the Athenian Stranger is struggling to persuade the others at Laws 2, 661 e 6–662 a 7. It is an old Socratic truth whose implausibility Plato has been commenting on since before he wrote the Gorgias. This paraphrase is completed by an earnest asseveration introduced by a particle only apparently inferential; and the last sentence of the chapter is a typical closing sentence: it looks like an argument, but it is in fact mostly a repetition of very general themes from the preceding passages, organized to deliver an explicit protreptic message: ‘this life should be chosen by all who wish to be successful in reality’. Before this, the last Plato quotation was introduced with only the briefest of bridge passages, ‘For it is indeed necessary to hold this opinion earnestly’ (119. 21). Iamblichus then proceeds to quote his last quotation, from Laws 2, in which the Athenian Stranger approvingly attributes to Crete and Sparta an active but unacknowledged policy of forcing their poets to represent the virtuous as happy and the wicked as wretched, no matter how wealthy or good-looking or healthy they may happen to be. Iamblichus has an interest in the thesis, but not in the poets who are obliged to preach it, so he removes the two references to the poets, at the beginning and near the end of his quotation. This is the modified version: [Iamblichus excises 660 e 2: You oblige your poets to say] that the good man, because he has self-control and justice, is successful and happy . . . but if such a man survives for only the briefest time, the evil is less. [Iamblichus excises 661 c 5–8: I imagine you will persuade or compel the authors in your states to embody this doctrine of mine in the words, rhythms, and ‘harmonies’ they produce for the education of their youth. Isn’t that right? Look here, now,] my clear meaning is that so-called ‘bad’ things are good for those who lack justice but bad for those

Created on 4 July 2005 at 14.34 hours page 240

240

D. S. Hutchinson @ Monte Ransome Johnson

who have it, whereas so-called ‘good’ things really are good for the good, and bad for the bad. (Plato, Laws 2, 660 e 2–661 d 3, trans. after Saunders)

The only other change is at 661 b 4–5, where Iamblichus takes Plato’s meta-conversational comment and turns it into a metatextual note: ‘and you and I, presumably, argue this, that . . .’ becomes: ‘and my argument argues this, that . . .’. Apart from the banishing of the poets from Plato’s text (which has a certain ironic appropriateness, given that the general e·ect of Iamblichus’ modifications is to remove the drama and poetry from Plato’s original version), and the suppression of what ‘you and I’ argue, there are no modifications to report or explain. The quotation is otherwise perfectly literal. The block is pure, and the sequence is natural. The selections from Menexenus and Laws 2 are too brief to provide a solid enough foundation for any reasoning about the genetic a¶liation of the manuscripts that were used by Iamblichus. This brings to an end our analysis of this chapter, the last in this group of seven Plato chapters, and, with it we have finished our self-imposed labours and have identified, classified, and explained every single possible di·erence that could be revealed by the most scrupulous possible collation between the two versions, Plato’s original and Iamblichus’ modified one. It is now a complete study. 2.8. Summary and conclusions about the Plato chapters Now that we have studied every variant in every Plato chapter, we can draw some general conclusions about the techniques used by Iamblichus to produce them. At the most general level, Iamblichus simply selected and arranged into a certain order the most protreptic passages that he could find. He modifies the quotation to varying degrees, but these can all be explained in the way we have, as being various responses to the challenges presented by Plato’s dramatic dialogue. It is Plato whose works vary widely in their di·erence from dogmatic monologue, and this wide variety of degrees of di·erence forces on Iamblichus a wide variety of different degrees of intervention. But here is the important negative result: this is the only mode of intervention that Iamblichus engages in, and there seems to be no other reason that would motivate any other kind of intervention. Iamblichus gives an essentially faithful version of the Plato he quotes, even when he modifies it. His selection of protreptic pas-

Created on 4 July 2005 at 14.34 hours page 241

Authenticating Aristotle’s Protrepticus

241

sages from Plato is well informed and judicious, and he refrains from making any changes to the text of Plato. Even when he makes modifications to his quotation, he preserves all the thought and as much of the language of Plato as is consistent with his formal objective of flattening to monologue. No reader can be misled about Plato’s ideas by reading them in the essentially faithful version of Iamblichus. We found in each of the chapters except chapter 13 that the texture of Iamblichus’ work is pure blocks of quotations in each chapter, bookended by thin comments; this describes the situation for all but one page in these chapters (99. 20–110. 20), where a set of closing comments is composed of a pastiche made up of various small bits of several works of Plato, with threads from Iamblichus as well. We found in each of the chapters (again with the above exception) that the sequence of quoted blocks is natural, from one book to another without returning, and from earlier to later within books without returning. We studied not only the deliberate changes but also the accidental events that caused variants to crop up between Iamblichus’ version and Plato’s text. This has resulted in the unintended benefit of shedding some new light on the early history of the Plato texts. We saw that for the Gorgias, the manuscript used by Iamblichus was related to the F family against BTW; for the Theaetetus to W against BT; for the Phaedo to T, but by contamination, against BW; for Republic 7 there was insu¶cient information to confirm the suggestion of a¶liation with F. It stands to reason that the various manuscripts used by Iamblichus would have various genetic relationships to the various manuscripts used by those ancient editors who compiled the missing-link hyparchetypes that spawned all our medieval manuscript tradition. This new evidence needs to be put in the wider context of other ancient witnesses to Plato’s text,89 but that would take us too far afield from the present subject. For our purposes, the main lesson to draw from our study of variants is that we must regard the manuscript of Iamblichus (where he makes no intervention) as a manuscript copy of a manuscript of 89 Important work in this area was done by G. Jachmann, ‘Der Platontext’, Nachrichten von der Akademie der Wissenschaften in G•ottingen, phil.-hist. Kl. 11 (1941), 225–389, repr. in id., Textgeschichtliche Studien (K•onigstein/Ts., 1982), 581– 745, by E. R. Dodds in his Gorgias edition, and by A. Carlini, Tradizione del Fedone. The more than eighty papyrus fragments of Plato’s works have been recently studied in the Corpus dei papiri filosofici series.

Created on 4 July 2005 at 14.34 hours page 242

242

D. S. Hutchinson @ Monte Ransome Johnson

Plato. If this is how he worked in his Aristotle chapters, then we must trust the Iamblichus version as a manuscript of Aristotle’s lost work, unless we are in a zone where he has modified the quotation.

3. The Aristotle chapters (6–12) 3.1. Preliminary remarks Having completed the study of the Plato section in Iamblichus’ Protrepticus, we are now in a position to exploit our understanding by applying it to the Aristotle section. We operate with the hypothesis that Iamblichus used essentially the same technique of quotation in the Aristotle section as he did in the Plato section, and this hypothesis yields some predictions to be tested. The general pattern is for Iamblichus to quote pure blocks of text assembled in a natural sequence, separated by thin bookend comments in his own voice. This voice has become rather familiar to us from studying the Plato chapters, and we shall recognize it all the more easily once we know where to look for it, in the openings and closings of each chapter. We expect that these openings and closings will consist of a single sentence, but on occasion they may be less than a sentence (as in the openings of chapters 15 and 16), or several sentences instead of one (as in the closings of chapters 17 and 19). Whereas the only di¶culty with the openings and closings is to know how extensive they are, i.e. one sentence or several, it is a more subtle business to detect a bridge passage or a fast-forward passage, as these could occur anywhere within a chapter, and can be discerned only by internal considerations. Our familiarity with the voice of Iamblichus will grow every time we identify another pair of sentences as Iamblichus opening or closing his own chapters. It greatly simplifies matters to focus our initial scrutiny on the contribution of Iamblichus, not on the contribution of Aristotle. We can resolve each chapter into one or a few pure blocks of quotation if we succeed in stripping away the voice of Iamblichus, which leaves us with only one or a few questions to ask, as to the author and work from which the block might have been drawn. If we had not first stripped away the comments to reveal the blocks, we would have needed to assess each sentence or paragraph in turn (Aristotle? Protrepticus?); and since there are many times as many

Created on 4 July 2005 at 14.34 hours page 243

Authenticating Aristotle’s Protrepticus

243

Aristotle sentences as Iamblichus sentences, this would have been many times more laborious. And it would also have been less reliable, since we are able to become very familiar with the rather distinctive types of comments that Iamblichus makes in this book. And although Aristotle’s ideas and ways of thinking are rather familiar to us, we have little direct knowledge of the style and voice(s) of any of his published works, and we have no prior knowledge of which work(s) any of Iamblichus’ quotations came from. By focusing on Iamblichus first, we turn the traditional problem inside out. Once we remove the voice of Iamblichus, we are left with blocks which are predicted to have come from Aristotle. This prediction is then independently confirmed, when possible, sometimes with multiple arguments. Though a few blocks of text (considered each on their own) could perhaps have come from another author, most blocks could have come only from Aristotle; and each chapter, without exception, contains material that could have come only from him. It is important to appreciate that our attribution of a passage to Aristotle is founded generally on two independent considerations: formally, what is not Iamblichus must always be from the single author he is consistently quoting; materially, independent testimony generally corroborates the block in question as Aristotle’s in origin. This approach is possible only because we have found out that Iamblichus quoted pure blocks of Plato text in a natural sequence, and can be expected to have done so for Aristotle. (Yet the pattern of pure blocks quoted in natural sequence is not quite strictly observed; as we saw in chapter 13 (above, pp. 233– 5) Iamblichus departed, for a page or so, from sequential block quotation in favour of a looser pastiche. We return to this question in Section 4, when we take stock of what we do and do not know, and of the degree of reliability of our conclusions.) One thing we definitely do not know is whether Aristotle’s work(s) contained dialogue, as we expect that Iamblichus will have removed all, or virtually all, trace of it. We shall not generally know the degree of modification of any modified quotation, as this seems to depend, to judge from the Plato chapters, on the amount of dialogue there was in the original, the very thing we do not know. As we work through the Aristotle chapters, we refrain from claiming which of Aristotle’s works are in question, until Section 4. A beneficial by-product of our work of authentication is to provide the first modern English translation of this Aristotle material

Created on 4 July 2005 at 14.34 hours page 244

244

D. S. Hutchinson @ Monte Ransome Johnson

in its context. Since the main issue is to distinguish the voices, the judicious plan is to hear both voices, in other words to translate the chapters of Iamblichus with integrity, treating them as the units they were designed to be. Finally, a word by way of preface about what is old and what is new in our treatment of these seven Aristotle chapters. Generally, we have little new evidence to present which might establish individual sentences or passages as coming from Aristotle (or his Protrepticus). The improvements come about as a consequence of our structural analysis of Iamblichus’ book, and of the layout of the evidence. Our procedure will be to study first the inner five chapters, and then the outer pair, as we did with the Plato chapters, as this allows readers to understand the simpler chapters first before tackling the most complex chapters, 6 and 13, which stand at the beginnings of their respective sections. 3.2. Chapter 7 Let us begin this process by discussing the beginning of chapter 7, pausing at first to assess the first sentence: Chapter 7. Someone might see the same point more recognizably on the basis of the following. To have wisdom and cognition is in itself valuable for humans, for it is not possible to live as a human without these; and it is also useful for our way of life, for nothing good comes to us unless it is accomplished after we have reasoned and acted in accordance with wisdom. Moreover, whether living successfully consists in enjoyment or in having virtue or in wisdom, according to all these we should do philosophy, for these things come to us most of all, and in a pure way, through doing philosophy. (Iambl. Protr. ch. 7, 71. 13–22)

Chapter 7 has an opening of a single line: :δοι δ9 -ν τις τ α?τ γνωριµ(τερον "π τοτων. Iamblichus is remarkably uncreative in the construction of his bridge passages, and we see almost the same sentence as a bridge in chapter 8: γνοη δ9 -ν τις το α?τ κα "π τοτων ‘Someone might recognize the same thing on the basis of the following as well’ (77. 12–13). There are certain patterns of speech indicating that Iamblichus wrote these words, patterns which are especially obvious when compared with similar patterns in the Plato sections. In the present sentence, we notice the use of the potential optative with an indefinite pronoun, and a prepositional phrase ("π τοτων), in the context of a procedural comment. There are

Created on 4 July 2005 at 14.34 hours page 245

Authenticating Aristotle’s Protrepticus

245

several openings in the Plato section employing a very similar construction,90 and further similarities with closings and bridges in the Plato section.91 The same patterns also recur in openings,92 closings,93 and bridges94 in the Aristotle section. Our hypothesis is that the arguments following Iamblichus’ opening are Aristotle’s; and they will be immediately recognized as such. What we have is a protreptic version of the ‘three lives’ argument. We need look no further than Aristotle’s own works to find confirmation that this argument is his: ‘Whether living successfully consists in enjoyment or in having virtue or in wisdom, according to all these we should do philosophy’ (71. 18–21). In the Eudemian Ethics Aristotle mentions that the distinction between internal and external goods was made in arguments he labels ‘exoteric’, meaning popular or even published:95 ‘All goods are 90 The most closely parallel opening in the Plato section is in chapter 19: δναιτο δ9 -ν τις κα κατ9 #δαν "π τ%ν ν τH ψυχH "γαθ%ν πεξελθεν τ/% προκειµν/ω (chapter 19 opening, 115. 22–3). But see also: :θι δ. κα "π τοτων πλθωµεν ο6τωσ (chapter 13 opening, 90. 17–18); "ρχ(µεθα κα τοτου ντεθεν (chapter 17 opening, 111. 21; compare the chapter 9 opening, below, p. 258); µετωµεν δ. κα π θ τερα κατ2 τν α?τν τρ+πον (chapter 18 opening, 114. 6–7). 91 The closing of chapter 13: τοιαται Aστωσαν κα α> "π τοτων π τ.ν ε#ς φιλοσοφαν προτροπ.ν Aφοδοι (100. 19–20). 92 Throughout the Aristotle section, in many opening sentences, Iamblichus frequently uses the preposition "π+ in connection with the arguments that follow: "π τ%ν ναργ%ς πVσι φαινοµνων (chapter 8 opening, 75. 15–16); "π το τς φσεως βουλµατος (chapter 9, 79. 7–8); "π τ%ν τεχν%ν (chapter 10 opening, 84. 9); "π τς Mλης ε?δαιµονας (chapter 12, 89. 8). Two of the chapters in the Aristotle section open with the potential optative or subjunctive moods: Jδε ο=ν λγωµεν (chapter 6 opening, 67. 23); δλον 8ν γνοιτο ντεθεν (chapter 11, 86. 11–12). 93 The closing of chapter 8: οUτως -ν τις τ2ς "π τ%ν ννοι%ν φ+δους συγκεφαλαι(σαιτο δε+ντως ε#ς προτροπν (79. 3–4). 94 Two bridges: κ τ%νδε πεισθεη τις -ν (chapter 6, 70. 22); γνοη δ9 -ν τις τ α?τ κα "π τοτων (chapter 8, 77. 12). 95 The expression ν τος ξωτερικος λ+γοις refers to arguments in writings intended for a more popular audience, as the arguments in the Protrepticus selfevidently were. Philoponus and Simplicius both tell us that the expression refers to arguments that are neither strictly demonstrative nor intended for students, but rather for the majority of people out of a desire to persuade them by means of more commonly held beliefs (Philop. In Phys. 705. 22 Vitelli; Simpl. In Phys. 695. 34 Diels). Aristotle uses the expression in the following places: Phys. 4. 10, 217b31; Metaph. Μ 1, 1076a28; NE 1. 13, 1102a26; 6. 4, 1140a3; EE 1. 8, 1217b22; 2. 1, 1218b34; Pol. 1. 5, 1254a33; 3. 6, 1278b31; 7. 1, 1323a22. In some cases it is possible that he is referring to multiple works; this seems to be the case in the Metaphysics passage, where he refers to arguments against the Forms that we have reason to believe were made in multiple works. In the Physics passage, on the other hand, the reference to plural ξωτερικ%ν λ+γων is to arguments, not multiple works, about time. Similarly, the reference in Politics 3. 6 is obviously to a single treatment of

Created on 4 July 2005 at 14.34 hours page 246

246

D. S. Hutchinson @ Monte Ransome Johnson

either outside or inside the soul, and of these those in the soul are more desirable; this distinction we make even in the exoteric arguments. For wisdom, excellence, and pleasure are in the soul, and some or all of these seem to all to be the end’ (EE 2. 1, 1218b32–6). According to Eudemian Ethics, then, the ‘three lives’ argument is supported by the division of goods into those of the body and of the soul. In the section of the Rhetoric devoted to advice on the construction of a protreptic speech, Aristotle highlights this distinction between internal and external goods (1360b24–9). And so we are not surprised to find it in the continuation of the argument in chapter 7 of Iamblichus’ Protrepticus: Furthermore, part of us is soul, part body; and the one has authority, the other is under authority; the one uses, the other supports it as a tool. Further, it is always with reference to that which has authority and uses it that the use of that which is under authority, i.e. the tool, is co-ordinated. And of the soul one part is reason (which by nature has authority and judges our a·airs), the other part is a follower and is naturally under authority. And everything is well disposed when it is in accordance with its own proper virtue, for to obtain this is good. (Iambl. Protr. ch. 7, 71. 22–72. 3)

Still more external support, this time from the Nicomachean Ethics, is available to show that this exact distinction was drawn in popular arguments: ‘Some things are said about it [the nature of the soul], adequately enough, even in the exoteric arguments, and we must use these; I mean that one element in the soul is irrational and one possesses a rational principle’ (NE 1. 13, 1102a26–8). Now the argument in the Protrepticus continues with a discussion of the reason why it is the proper virtue and the goods of the soul that should be valued and chosen: Moreover, it’s when a thing’s most dominant and most honourable parts have their virtue that it is well disposed; therefore the natural virtue of that which is better is naturally better. And that which is by nature more authoritative and more commanding is better, as a human is over the other animals; thus soul is better than body (for it is more authoritative), as is the part of the soul which has reason and thought, for this kind of thing is the various kinds of rule, not to multiple works or books about kinds of rule. The remaining references in the ethical-political works may also be to arguments in individual works. For a di·erent account, which has the expression refer to ‘arguments • not peculiar to the Peripatetic school’, see H. Diels, ‘Uber die exoterischen Reden des Aristoteles’ [‘Exoterische Reden’], Sitzungsberichte der Berliner Akademie der Wissenschaften, 19 (1883), 477–94.

Created on 4 July 2005 at 14.34 hours page 247

Authenticating Aristotle’s Protrepticus

247

what prescribes and proscribes and says how we ought or ought not to act. Whatever, then, is the virtue of this part necessarily is the most valuable virtue of all, both for everything in general and for us; in fact, I think one might actually take this position, that we are this part, either alone or especially. (Iambl. Protr. ch. 7, 72. 3–72. 14)

Here we can o·er very solid confirmation that these are Aristotle’s ideas, from the Politics, where Aristotle gives an extended version of the whole argument that we see here, and refers to it having been made in the popular version of the arguments: Considering that much has been given an adequate treatment also in the exoteric arguments concerning the best life, we will now only repeat what is contained in them. Certainly no one will dispute the propriety of that partition of goods which separates them into three classes, viz., external goods, goods of the body, and goods of the soul, or deny that a successful man must have all three . . . whereas external goods have a limit, like some instrument, and all things useful are useful for something; where there is too much of them they must either do harm to those who have them, or at any rate be of no use, every good of the soul, the greater it is, is also of greater use, if the term useful as well as noble is appropriate to such subjects. No proof is required to show that the best state of one thing in relation to another corresponds in degree of excellence to the interval between the natures of which we say that these very states are states: so that, if the soul is more noble than our possessions or our bodies, both absolutely and in relation to us, it must be admitted that the best state of either has a similar ratio to the other. Again, it is for the sake of the soul that goods external and goods of the body are desirable at all, and all wise men ought to choose them for the sake of the soul, and not the soul for the sake of them. (Pol. 7. 1, 1323a21–b21, after Jowett in Revised Aristotle)

The continuation of the Protrepticus passage gives a more extensive version of the latter part of this argument: Furthermore, it’s when the natural function of each thing is achieved, not as a result but in itself, that it is called finest, and then it should also be called good, and one should take the most dominant virtue to be the one by which each thing naturally accomplishes this very thing. So that which is composite and divisible into parts has many di·erent activities, but that which is by nature simple and whose being is not relative to anything else necessarily has a single virtue in itself, in the strict sense. So if a human is a simple animal whose being is ordered according to reason and intellect, there is no other function for it than only the

Created on 4 July 2005 at 14.34 hours page 248

248

D. S. Hutchinson @ Monte Ransome Johnson

most precise truth, i.e. having the truth about existing things; but if it is naturally composed of several capacities, it is clear that, of the several things it can naturally achieve, the best of them is always their function, e.g. of the doctor health, and of the pilot safety. And we can name no function of thought, or of the thinking part of our soul, better than truth. Truth therefore is the most dominant function of this part of the soul. And it does this simply with knowledge, and it does this more with more knowledge; and the most dominant end for this is observation. For when of two things one is valuable because of the other, the one on account of which the other is valuable is better and more valuable; for example, pleasure is better than pleasant things, and health than healthy things, for the latter are said to be productive of the former. Thus nothing is more valuable than wisdom, which we say is a capacity of the most dominant thing in us, when disposition is judged against disposition; for the cognitive part, both apart and in combination, is better than all the rest of the soul, and its knowledge is a virtue. (Iambl. Protr. ch. 7, 72. 14–73. 17)

Notice that the version of the argument in Iamblichus’ Protrepticus is not only more extended but also has a more protreptic focus. This, of course, stands to reason if Iamblichus was quoting from Aristotle’s Protrepticus. But the main point in each version of the argument is that wisdom (and hence philosophy) is the proper virtue of the highest part of us, that among the three goals of life it is the most valuable, and that we should do philosophy, whether or not the ultimate conclusion is explicitly drawn in the Eudemian Ethics, Nicomachean Ethics, or Politics. Thus in no fewer than three extant works we have Aristotle’s own word that the argument contained in chapter 7 of Iamblichus’ Protrepticus was published in a more popular and accessible form. Iamblichus apparently quoted this argument directly from the work that contained it, and appears not to have altered it, beyond o·ering his customary one-line introduction. In the extant ethical works, Aristotle uses the ‘three lives’ argument as a framework for focusing on the kind of life that is better than others because of its independence and its appropriateness to the kind of things we humans are. But in the Rhetoric Aristotle advises the speech-writer to construct a protreptic speech with reference to pleasure and virtue, since all are agreed that these are things at which everyone aims (1360b14–18 and 1362b2–12). And so in the Protrepticus, whose audience is necessarily diverse and unlikely to be predisposed to the contemplative life, Aristotle pre-

Created on 4 July 2005 at 14.34 hours page 249

Authenticating Aristotle’s Protrepticus

249

fers to co-opt the goals of his various readers: philosophy conduces to each of the goals of life—pleasure, virtue, and wisdom. We have just seen him advance the argument about wisdom; now here is what he goes on to say about virtue: Therefore its function is none of what are called ‘parts of virtue’, for it is better than all of them and the end produced is always better than the knowledge that produces it. Nor is every virtue of the soul in that way a function, nor is success; for if it is to be productive, di·erent ones will produce di·erent things, as the skill of building (which is not part of any house) produces a house. However, wisdom is a part of virtue and of success, for we say that success either comes from it or is it. Thus according to this argument too, it is impossible for this to be productive knowledge; for the end must be better than the thing which comes to be, and nothing is better than wisdom, unless it is one of the things we have named and none of those is a function distinct from it. Therefore a certain observational knowledge is what one should name this kind, since it is surely impossible for production to be its end. Hence wisdom and observation are a function of the virtue, and this of all things is the most valuable for humans, comparable, I think, to seeing for the eyes, which one would choose to have even if there wasn’t anything di·erent that was going to result from it beyond the vision itself. (Iambl. Protr. ch. 7, 73. 17–74. 7)

Here it is interesting to observe that Aristotle in part eliminates virtue as the ultimate goal of life (since the soul’s highest function is none of the individual virtues), and in part co-opts the goal of virtue for the cause of wisdom (by making wisdom the highest virtue of the soul). He pursues a similar strategy with respect to pleasure in the sequel, which will take us to the end of the chapter: Again, if we like sight for its own sake, this gives su¶cient witness that everybody ultimately likes wisdom and cognition. Again, if someone likes a particular thing because something else coincides with it, it is clear that he will wish more for that which has more of it: for example, if someone happened to choose walking because it’s healthy, and it occurred to him that running is more healthy for him, and possible, he will choose this even more, and would choose it as soon as he recognized that. Further, if true opinion is similar to wisdom, since having true opinions is valuable in that and in so far as it is similar to wisdom on account of its truth, if this exists more in wisdom, then wisdom will be more valuable than having true opinion. But yet, living is distinguished from not living by sense perception, and living is defined by its presence and power, and if this is removed life is not worth living, as though life itself were removed along with sense percep-

Created on 4 July 2005 at 14.34 hours page 250

250

D. S. Hutchinson @ Monte Ransome Johnson

tion. But among the senses the capacity of sight is distinguished by being the most distinct, and for this reason as well we value it most; but every sense perception is a capacity for becoming familiar with things through a body, just as hearing perceives the sound through the ears. Therefore, if living is valuable because of the perception, and the perception is a kind of cognition, and we choose it because the soul is able to have familiarity by means of it, and we’ve always said that of two things the more valuable one is always the one which has that more, and of the senses vision is necessarily the most valuable and honourable, and wisdom is more valuable than it and all the others, and more valuable than living, then the wisdom of truth is more dominant; hence all humans pursue wisdom most. For in liking living we like having wisdom and familiarity; for we value life for no other reason than for the sake of perception, and above all for the sake of vision; for we evidently love this capacity exceedingly; for it is, compared with the other senses, virtually a sort of knowledge. (Iambl. Protr. ch. 7, 74. 7–75. 13)

It appears here that Iamblichus has truncated or even altered the argument at lines 74. 7–9, where we read: ‘again, if we like sight for its own sake, this gives su¶cient witness that everybody ultimately likes wisdom and cognition’. The connection between enjoyment and perception has not been explicitly drawn in a way that would complete the ‘three lives’ argument announced earlier in the chapter. Of course, the connection between perception and pleasure is not hard to draw. At any rate, it is clear that Aristotle has pursued a co-option strategy, arguing that pleasure is valued because of sense perception, that vision is the most valued of sense perceptions, and that vision is valued because it is akin to cognition and knowledge and wisdom. This conclusion is detectable in the final sentences of the chapter, but we regard these lines as Iamblichus’ words because they are insu¶ciently progressive and excessively enthusiastic: ‘exceedingly’, ‘virtually a sort of’, etc. The excessive repetition of the inferential particle γ ρ in a conclusion is a construction of Iamblichus’ that we see also in the Plato section.96 In 74. 20–1 the idea is expressed that life is to be identified with perception. This is a traditional view, and Aristotle discusses it in NE 9. 9 (1170a16 ·.) and EE 7. 12 (1244b26 ·.), adding in DA 1. 2 (403b25–7) that motion is also critical. Aristotle’s thought that ‘if this is removed life is not worth living, as though life itself were removed along with perception’ is echoed later in chapter 10: 96 Compare the endings of chapters 16 and 19, which consist of repetitive pseudoconclusions strung together by particles only apparently inferential.

Created on 4 July 2005 at 14.34 hours page 251

Authenticating Aristotle’s Protrepticus

251

‘for we should be almost entirely motionless if deprived of it [sc. perception]’ (86. 4–5). Aristotle appears to refer to one or both of these statements at the outset of a portion of the Eudemian Ethics that seems in some way directly related to the Protrepticus: The matter [about the relationship between self-su¶ciency and friendship] is clear if we grasp what life is, in the active sense and as an end. It is manifest that life is perception and knowledge, and that consequently social life is perception and knowledge in common. But self-perception and self-knowledge are highly desirable for each individually (and it is owing to this that the desire for life is implanted by nature in all, for living must be deemed a sort of knowing). If therefore one were to abstract the knowledge and make it itself by itself (though this is not evident in the argument as we have written it, but in practice it is evident), there would be no difference between this knowledge and another person’s knowing instead of oneself; but that is like another person’s living instead of oneself, whereas perceiving and knowing oneself is more desirable, as is reasonable. For two things must be connected together in the argument that life is desirable and that good is desirable, and as a consequence that it is desirable for ourselves to possess a nature of that sort. If, therefore, of the pair of corresponding series of this kind one is always in the class of the desirable, and the known and the perceived are generally speaking constituted by their sharing in the ‘determined’ nature, so that to wish to perceive oneself is to wish oneself to be of a certain sort—since, then, we are not each of these things in ourselves but only by participation in these faculties while perceiving or knowing (for when perceiving one becomes perceptible in that way and in that respect in which one had earlier perceived; and when knowing one becomes knowable)—hence owing to this one wishes always to live because one wishes always to know; and this is because one wishes to be oneself the object known. (EE 7. 12, 1244b23–1245a10, trans. after Solomon in Revised Aristotle)

If this parallel su¶ces to support the authenticity of the second block, then two blocks have been proven to be Aristotle’s. These two blocks appear to be one block but condensed by having content briefly alluded to in a ‘fast-forward’ paraphrase. If this is right, what we have in this chapter is one block of quotation, demonstrated by multiple proofs to be Aristotle’s words (perhaps modified, especially if they were originally contained in a dialogue), drawn from his ‘exoteric arguments’. As to which work contained these arguments, the best candidate is the Protrepticus; but this attribution has not yet been solidly proven, and to assert it would be premature.

Created on 4 July 2005 at 14.34 hours page 252

252

D. S. Hutchinson @ Monte Ransome Johnson

3.3. Chapter 8 So let us proceed to recover further quoted material, resuming with the beginning of chapter 8. Chapter 8. Still, it is no bad idea to call the subject to mind, reasoning on the basis of the common conceptions, too, from what manifestly appears to everyone. So then, this, at least, is quite clear to everyone, that nobody would choose to live in possession of the greatest wealth and power of all people if they nevertheless lacked wisdom and were raving mad, not even if they were going to live enjoying the wildest pleasures, in the way that some people out of their minds lead their lives. Thus everybody, it seems, avoids being unwise most of all. Now wisdom is the opposite of being unwise, and of these opposites the one is to be avoided, the other is valuable. So, just as being sick is to be avoided, so is being healthy valuable for us. (Iambl. Protr. ch. 8, 75. 14–24)

Here again it is clear that the first sentence was written by Iamblichus, as it contains a number of tell-tale traits: it is a procedural comment, designed to motivate a reading of the argument to come; no progressive argument is entered into, and we are left on a very general level, with no detailed or particular considerations; finally, it is organized around the idea of the basis of evidence, claiming to derive the arguments ‘from the common conceptions ["π τ%ν κοιν%ν ννοι%ν]’.97 We see none of these traits in the successive sentences, which contain a clear progressive argument on the basis of substantial particular premisses. Hence we should make the separation of voices after the first sentence (75. 14–16), and regard the next sentences as being a quotation (modified or not) from Aristotle. As for what appears to be the Aristotle content, we can see that the style of reasoning is consistent with Aristotle’s, but we lack evidence to authenticate him definitely as the source of this particular block. After these first eight lines of quotation, we read the following, starting with a bridge passage in the voice of Iamblichus: Wisdom, it seems, according to this argument too, is the most desirable of all things, and not for the sake of anything else that results from it, as the common conceptions give witness. For even if someone had everything, but was suffering from mental illness, that way of life would not be valuable, for none 97 The term ‘common conceptions’ is anachronistic, or at least does not occur in the extant works of Aristotle, though it is common currency from the Hellenistic period to the end of antiquity, thanks especially to the influence of the Stoics and Euclid.

Created on 4 July 2005 at 14.34 hours page 253

Authenticating Aristotle’s Protrepticus

253

of his other goods would be of any benefit. Hence everybody, in so far as they have some perception of being wise and can have a taste of this thing, think the other things to be nothing, and this explains why not a single one of us would put up with being drunk or childish up to the ends of our lives. So, on account of this too, though sleep is extremely pleasant, it is not valuable, even if we were to suppose that all of the pleasures were present to the sleeper, because the images during sleep are false, while those of the waking are true. For sleep and waking are no di·erent from each other except that then the soul often gets the truth, but when sleeping is always thoroughly deceived; for the phantom in dreams is actually entirely false. And the fact that most people avoid death also shows the soul’s love of learning; for it avoids what it does not recognize, what is dark and not clear, and naturally seeks what is evident and recognizable. This is the main reason why we say one should honour those who have caused us to see the sun and the light, and revere our fathers and mothers as causes of the greatest of goods; and causes they are, it seems, of our having any wisdom and sight. It is for the same reason that we also enjoy what we are acquainted with, both things and people, and call ‘friends’ those with whom we are familiar. (Iambl. Protr. ch. 8, 75. 25–76. 26)

Here at the beginning we have a typical bridge passage, indicated by the procedural agenda and the tag α> κοινα Aννοιαι (‘the common conceptions’). We predict that what follows, for twenty-four more lines, was written by Aristotle. To confirm this we may compare the following from the Eudemian Ethics: For there are many contingencies such as to make men throw away their life, such as sicknesses, overwhelming pain, and storms, so that it is clear that, even if it was desirable at first, if one were given the choice, one would choose not to be born, for these reasons anyway. Further, the life we lead as children is not desirable, for no one in his senses would put up with returning again to living like that. . . . We may say the same of the pleasure of sleeping. (EE 1. 5, 1215b18–1216a3, after Solomon in Revised Aristotle)

In the version found in Iamblichus’ Protrepticus, death, madness, sleep, and puerility were used to illustrate why one should pursue wisdom, since wisdom is somehow the opposite of what is avoided in each of these. These are treated together briefly in the context of the ‘three lives’ argument of Eudemian Ethics 1. 5 (and individually elsewhere in the corpus),98 but it appears that in Aristotle’s original book there was a more elaborate set of endoxic arguments for 98 Aristotle makes a similar argument about the prospect of lifelong childhood in NE 10. 2 (1174a1 ·.) and MM 1. 4, 1185a2; on the emptiness of sleep cf. NE 1. 13, 1102b2–11; 10. 8, 1178b18–20; EE 1. 5, 1216a2–10.

Created on 4 July 2005 at 14.34 hours page 254

254

D. S. Hutchinson @ Monte Ransome Johnson

this conclusion (what Iamblichus calls an argument ‘from common conceptions’). This kind of argument is obviously perfectly suited to a popular work, comprising, like the material from chapter 7, clearly ‘exoteric arguments’ meant for popular consumption.99 Chapter 8 continues with citations from Aristotle, punctuated by two more bridge passages and a closing sentence authored by Iamblichus. Let us look at the next block and bookend: These things, then, might show distinctly that we like what’s recognizable and evident and what’s clear; and if we like what’s recognizable and what’s distinct, it is clear that we also like recognizing and being wise, likewise. In addition to these, just as with property, it is not the same possession that is for the sake of living, and of living well, for humans; so too, with wisdom: we do not, I think, need the same wisdom for merely living and for living nobly. Now then, much allowance is made for the many who do this (they pray to be successful, but like it if they can just stay alive), but anyone who thinks that there is no need to endure living in every way already thinks it’s ridiculous not to bear every burden and exert every e·ort so as to possess this wisdom that will know the truth. (Iambl. Protr. 8, 76. 26–77. 11)

We do not have much to say about what appears to be the Aristotle quotation, other than pointing out the obvious consistency of this line of reasoning with his criticism of wealth-getting as the end of human life during his discussion of the ‘three lives’ argument in Nicomachean Ethics 1. 5 and elsewhere. As for Iamblichus’ bridge passage, it is typical in diction, grammar, and extraneousness, in contrast to the words it precedes. The same is true of the bridge passage which introduces the following fascinating block: One might recognize the same thing from the following as well, if one observed the human way of life in the clear light of day. For one will discover that all the things that seem great to people are an optical illusion. This makes it also right to say that the human creature is nothing and that nothing is secure in human a·airs. For strength, size, and beauty are laughable and of no worth—and beauty seems to be the sort of thing it is by our seeing nothing accurately. For if one were able to see as keenly as they say Lynceus did, who saw through walls and trees, how could such a sight seem bearable, seeing what bad things they are composed of? And honours and reputations, objects of more striving than the rest, are full of indescribable nonsense; for to those who behold anything eternal it is 99 In the section of the Rhetoric devoted to the construction of protreptic speech, Aristotle recommends the discussion of goods understood as opposites of bad things (Rhet. 1. 6, 1362b29 ·.).

Created on 4 July 2005 at 14.34 hours page 255

Authenticating Aristotle’s Protrepticus

255

silly to take seriously those things. What is great or what is long-lasting in human a·airs? No, it is owing to our weakness, I think, and the shortness of our life, that even this appears sizeable. (Iambl. Protr. 8, 77. 12–27)

Both bridge passages are easily identifiable by the potential optative construction, as well as the summary nature of their contents; and the second of these bridges is very similar to the opening of chapter 7, discussed above. The prediction that what remains when these are removed is Aristotelian can be confirmed through an external parallel. Boethius evidently wove the main idea of the second block into the fabric of his Consolation of Philosophy; he must have remembered it, either directly or mediated through another source,100 from Aristotle’s Protrepticus: If, as Aristotle says, men had had the eyes of Lynceus, so that their sight could pierce through obstacles, would not the body of Alcibiades, so fair on its surface, have seemed most foul when its inward parts were seen? So it is not your own nature, but the weakness of the eyes which see you, that makes you seem beautiful. (Boeth. Cons. 3. 8 = Arist. fr. 10a101 Ross)

Boethius gives no indication as to which work featured Lynceus and his keen eyes, but he does declare it to be a work of Aristotle. Continuing with Iamblichus, we read, at the end of chapter 8, a pure quotation from Aristotle up to the closing sentence supplied by Iamblichus: So who could look at all this and think themselves successful and happy, if, right from the start, we are naturally put together as if for punishment, all of us, as they say in the initiation rites? For the ancients have an inspired saying that says that the soul ‘pays penalties’, and we live for the atonement of certain great failings. For the conjunction of the soul with the body looks very much like a thing of this sort; for as the Tyrrhenians are said often to 100 In this text, the keen eyesight of Lynceus is applied to the beautiful young Alcibiades, a scenario that seems also to be derived from a passage in [Plato], Alcibiades, in which Socrates says to Alcibiades, ‘your beauty is just beginning to bloom; I shall never forsake you now, never, unless the Athenian people make you corrupt and ugly . . . the “people of great-hearted Erechtheus” [sc. Athenians] might look attractive on the outside, but you need to scrutinize them in their nakedness’ (132 a). It is not clear whether the Hortensius contained a mention of Alcibiades, since either Cicero or Boethius was capable of constructing this scenario, in which the keen eyesight of Lynceus sees through the beauty of Alcibiades; the latter person apparently did not make an appearance in Aristotle’s Protrepticus, at least not here. 101 This is not actually a ‘fragment’ of Aristotle’s book, but a bit of evidence about that book (obviously, as it is in Latin); the same goes for the ‘fragment’ from Augustine which we discuss below.

Created on 4 July 2005 at 14.34 hours page 256

256

D. S. Hutchinson @ Monte Ransome Johnson

torture their captives by chaining corpses right onto the living, fitting limb to limb, similarly the soul seems to be extended through and stuck onto all the sensitive members of the body. So nothing divine or happy belongs to humans apart from just that one thing worth taking seriously: as much intelligence and wisdom as is in us, for, of what’s ours, this alone seems to be immortal, and this alone divine. And by being able to share in such a capacity, our way of life, although naturally miserable and di¶cult, is yet so cleverly managed that, in comparison with everything else, a human seems to be a god. For ‘intelligence is the god in us’—whether it was Hermotimus or Anaxagoras who said so—and ‘the mortal phase has a portion of some god’. We ought, therefore, either to do philosophy [φιλοσοφητον] or say goodbye to life and depart from this world, since all of the other things anyway seem to be a lot of nonsense and foolishness. In this way one may get the requisite summary of the approaches from the conceptions102 by which people are exhorted to feel the need to do philosophy observationally, and to live as much as possible the way of life in accordance with knowledge and the intellect. (Iambl. Protr., ch. 8, 77. 27–79. 6)

Obviously we should make the separation of voices before the last sentence (79. 3–6), and regard the previous sentences as being a quotation (modified or not) from Aristotle. The abrupt descent, from the rhetorical high point ‘we ought, therefore, either to do philosophy or say goodbye to life and depart from this world’ down to the dry and earnest verbiage of Iamblichus, is palpable, especially if the passage is declaimed out loud. The comment of Iamblichus is meta-textual, with distinctive diction: ‘get the requisite summary’, ‘approaches’, ‘ conceptions’. The prediction that the remaining material is Aristotelian can be strongly confirmed through external evidence as well. It has long been known that Aristotle wrote it, and that Cicero paraphrased it, from the following comment by Augustine: How much better and nearer the truth than yours were the views about the generation of men held by those whom Cicero, as though led and compelled by the very evidence of the facts, commemorates in the last part of the dialogue Hortensius! After mentioning the many facts we see and lament with regard to the vanity and the unhappiness of men, he says: ‘From these errors and cares of human life it results that sometimes those ancients—whether they were prophets or interpreters of the divine mind 102 Returning to the reading of the manuscript and rejecting Kiessling’s conjectural addition κοιν%ν as unnecessary: Plutarch omits κοινα repeatedly in reference to α> κοινα Aννοιαι (e.g. Comm. not. 1060 d, 1061 d).

Created on 4 July 2005 at 14.34 hours page 257

Authenticating Aristotle’s Protrepticus

257

by the transmission of sacred rites—who said that we are born to expiate sins committed in a former life, seem to have had a glimmer of the truth, and that that is true which Aristotle says, that we are punished much as those were who once upon a time, when they had fallen into the hands of Etruscan robbers, were killed with studied cruelty; their bodies, the living with the dead, were bound as exactly as possible one against another; so our minds, bound together with our bodies, are like the living joined with the dead.’ (Contra Jul. 4. 15. 78, trans. Ross, Arist. fr. 10b)

Without entering into a discussion of Cicero’s Hortensius, which is lost and can therefore serve only very tangentially as a source of evidence for Aristotle’s book, we learn from this quotation by Augustine that Cicero paraphrased Aristotle in both the main ideas in the relevant paragraph (we are born to expiate sins, tying soul to body is torture), expressly crediting him with the truth of the matter about the second idea. Evidently this was a famous passage, and the scrupulous Cicero referenced it as Aristotle’s. We also have good reason to believe that the second paragraph (78. 12–79. 2) was written by Aristotle, for he wrote something rather similar, including a curious detail, in another work that we possess: When one man said that intelligence was present, as in animals, so throughout nature, as the cause of the world and all its order, he seemed like a sober man in contrast with the random talk of his predecessors. We know that Anaxagoras certainly adopted these views, but Hermotimus of Clazomenae is credited with expressing them earlier. (Arist. Metaph. Α 3, 984b15–20)

Intelligence is to be regarded as the ‘god in us’ because it is the source of order in us, as it is the source of order in the natural world at large. This is a view associated with Anaxagoras, but Aristotle mentions the possibility of Hermotimus, an earlier local forerunner of Anaxagoras, being the author of this idea, much as Leucippus of Abdera was thought to be a local forerunner of Democritus. Without entering into this question, we know that it is right to attribute to Aristotle precisely the uncertainty about Hermotimus that we find in the quotation by Iamblichus; so it is right to attribute the passage in this second paragraph also to Aristotle. Furthermore, it is clear in general that Aristotle, like Plato, was convinced that, through contemplation and intelligence, humans can become like

Created on 4 July 2005 at 14.34 hours page 258

258

D. S. Hutchinson @ Monte Ransome Johnson

god.103 So it must be regarded as settled that Aristotle is quoted here at the end of chapter 8. But what work is being quoted from? The best candidate is Protrepticus, because of the evident protreptic purpose of the quoted material, including the formulaic conclusion ‘one should do philosophy’ (which we shall encounter many times in this vicinity), and its energetic rhetorical design. Yet we have to admit the possibility, advanced by responsible scholars, that some other of Aristotle’s published works could have contained sections that amounted to exhortations to philosophy—his Eudemus, for instance, or On Philosophy—so the attribution to Protrepticus must be regarded for now as probable, not as definitely settled.104 3.4. Chapter 9 Chapter 9 is a relatively simple case. It is clear that Iamblichus introduces and concludes the chapter with his customary one-line opening and closing. Here is a translation of the entire chapter, which seems to consist of one huge block quotation. We do not detect in this any bridge passages or other interventions of Iamblichus: Chapter 9. Starting on a higher level, from the intention of nature, we proceed to the same exhortation in the following way. Some of the things that come to be come from a certain kind of thought and skill, e.g. a house or a ship (for a certain skill and thought is a cause of both of these), while others come to be not by means of any skill but through nature; for nature is a cause of animals and plants, and all such things come to be by nature. But then some other things come to be by luck as well, for of all the things that come to be neither through skill nor through nature nor by necessity, we say that most of these come into being through luck. Now then, of the things that come to be from luck, none comes to be for the sake of anything, nor do they have any end; but the things that come into being by skill have in them both the end and the purpose (for those who have a skill will always provide you with a reason why they wrote, i.e. 103 This is persuasively shown, though without reference to the evidence in the Protrepticus, by D. Sedley, ‘“Becoming like God” in the Timaeus and Aristotle’ [‘Becoming like God’], in T. Calvo Mart‹§nez and L. Brisson (eds.), Interpreting the Timaeus-Critias (Proceedings of the IV Symposium Platonicum; Sankt Augustin, 1997), 327–9. 104 See J. Brunschwig, ‘Aristote et les pirates tyrrh‹eniens (›a propos des fragments 60 Rose du Protreptique)’ [‘Pirates’], Revue philosophique de la France et de l’‹etranger, 88 (1963), 171–90; A. P. Bos, ‘Aristotle’s Eudemus and Protrepticus: Are they Really Two Di·erent Works?’ Dionysius, 8 (1984), 19–51.

Created on 4 July 2005 at 14.34 hours page 259

Authenticating Aristotle’s Protrepticus

259

for what purpose), and this is better than what comes to be because of it. (I mean all such things as skill is naturally a cause of, in virtue of itself and not coincidentally, for we should, strictly speaking, assume medicine to be the cause of health rather than of disease, and architecture to be the cause of houses, not of their demolition.) Therefore everything done with skill comes to be for the sake of something, and this its end is the best thing; however, that which is by luck does not come to be for the sake of anything, for something good might happen from luck indeed, but yet it is not in so far as it is from luck and in accordance with luck that it is good; and that which comes to be by luck is always indeterminate. But yet what is in accordance with nature does come to be for the sake of something, and is always constructed for the sake of something better than what comes to be through skill; for nature does not imitate the skill, but it nature, and it exists to help nature and to fill in what nature leaves out. For some things nature itself seems capable of completing by itself without actually needing any help, but others it completes with di¶culty or is completely unable to do, for example, to begin with, even with reproduction, some seeds presumably germinate without protection, whatever kind of land they fall onto, others also need the skill of farming, and, in a similar way, some animals also attain their full nature by themselves, but humans need many skills for their security, both at first in respect of their birth, and again later, in respect of their nurturing. Further, if skill imitates nature, from this it follows for the skills as well that everything that comes to be comes to be for the sake of something. For we should take the position that everything that comes into being correctly comes into being for the sake of something. And surely if nobly, then correctly; and everything that comes to be (or has come to be) in accordance with nature at any rate comes to be (or has come to be) nobly, since what is unnatural is foul, and a coming into being in accordance with nature comes to be for the sake of something. And someone could see this also from each of our parts; if, for example, one inspected the eyelid, one would see that it has come to be not in vain but in order to help the eyes, so as to provide them with rest and prevent things from falling into the eye. Thus it is the same thing, both that for the sake of which something has come to be and that for the sake of which it needs to have come to be; for example, if a ship needed to come to be to provide transport by sea, that’s why it actually has come to be. Moreover, the animals are surely things that have come to be by nature,105 either absolutely all of them or the best and most honourable of them; for it makes no di·erence if someone thinks that most of them have 105 Returning to the reading of the manuscript, and rejecting Vitelli’s conjectural supplement: φσει <τε κα κατ2 φσιν> (81. 5–6).

Created on 4 July 2005 at 14.34 hours page 260

260

D. S. Hutchinson @ Monte Ransome Johnson

come into being unnaturally because of some corruption or wickedness. But certainly a human is the most honourable of the animals down here; hence it’s clear that we have come to be both by nature and according to nature. This is the thing for the sake of which nature and the god have brought us into being. So what is this thing?106 When Pythagoras was asked, he said ‘to observe the heavens’, and he used to claim that he himself was an observer of nature, and it was for the sake of this that he had been released into this way of life. And they say that when somebody asked Anaxagoras for what reason anyone might choose to come to be and be alive, he replied to the question by saying, ‘To observe what’s in the heavens,107 the stars in the heavens, as well as the moon and sun’, because everything else is worth nothing. Further, if for everything the end is always better (for everything that comes to be comes to be for the sake of the end, and that for the sake of which is better, indeed the best of all), and an end in accordance with nature is that which is naturally the last to end up coming to be, when the coming to be reaches its limit without interruption, surely the first parts of a human being to reach their end are the bodily ones, and later on the parts of the soul, and somehow the end of the better part always comes later than its coming to be. Surely the soul is later than the body, and wisdom is the final stage of the soul, for we see that it is the last thing to come to be by nature in humans, and that is why old age lays claim to this alone of good things; therefore, some form of wisdom is by nature our end, and being wise is the ultimate thing for the sake of which we have come to be. Now surely if we have come to be, it’s also clear that we exist for the sake of some kind of wisdom and learning. Therefore Pythagoras was right, according to this argument anyway, in saying it’s for the sake of cognition and observation that every human person has been put together by the god. But whether the object of this cognition is the cosmos or some other nature is a question for us perhaps to consider later; what we have said is enough for us for now as a preliminary. For if wisdom is an end in accordance with nature, then to be wise would be best of all. Hence, the other things we do we ought to do for the sake of the goods that come about in one,108 and, of these goods, those in the body for the sake of those in the soul, and virtue for the sake of wisdom; for this is the highest of all. To seek from every kind of knowledge something other than itself and 106 Returning to the reading of the manuscript τ δ. τοτ+ στι·, and rejecting Zuntz’s conjectural deletions: [τ δ.] and [στι] (81. 12). 107 Returning to the reading of the manuscript το θε σασθαι τ2 περ τν ο?ρανν κα περ, and rejecting Pistelli’s conjectural emendations: [τ2 περ] and κα <τ2> περ (81. 18). 108 Returning to the reading of the manuscript α?τ/%, and rejecting Ross’s conjectural breathing mark α6τ/% (82. 17).

Created on 4 July 2005 at 14.34 hours page 261

Authenticating Aristotle’s Protrepticus

261

to require that it must be useful is the demand of someone utterly ignorant of how far apart in principle good things are from the necessities; they are totally di·erent. For among the things without which living is impossible, the ones which are liked on account of something else should be called necessities and joint causes, while all those that are liked for themselves, even if nothing else results from them, should be called goods in the strict sense; for this is not valuable because of that, and that for the sake of something else, and this goes on proceeding to infinity—rather, this comes to a stop somewhere. So it is absolutely ridiculous, then, to seek from everything a benefit beyond the thing itself, and to ask ‘So, what’s the benefit for us?’ and ‘What’s the use?’ For it’s true what we say: such a fellow doesn’t seem like someone who knows noble goodness, or who distinguishes between a cause and a joint cause. One might see that what we say is all the more true if someone conveyed us in thought, as it were, to the Isles of the Blest, for in that place there would come to be no use for anything, nor would anything benefit anything else, and only thinking and observation remains, which we say even now is an independent way of life. If what we say is true, would not any of us be rightly ashamed if when the right was granted us to settle in the Isles of the Blest, we were by our own fault unable to do so? Thus the payment that knowledge brings is not to be despised by humans, nor is the good that comes from it a slight good. For, as the expert poets say that we reap the rewards of justice in Hades, in the same way, it seems, we reap the rewards of wisdom in the Isles of the Blest. It is not weird at all, then, if it does not seem to be useful or beneficial; for we don’t claim it is beneficial but that it is itself good, and it makes sense to choose it not for the sake of something else but for itself. For just as we travel to Olympia for the sake of the spectacle itself, even if nothing more is going to accrue from it (for the observing itself is better than lots of money), and as we observe the Dionysia not in order to take something away from the actors (rather, we actually spend on them), and as there are many other spectacles we would choose instead of lots of money, so the observation of the universe, too, is to be honoured above all things that are thought to be useful. For surely we should not travel with great e·ort to see people imitating women and slaves, or fighting and running, and yet not think we should observe the nature of things, i.e. the truth, without payment. Now then this is how, proceeding from the intention of nature, we made an exhortation to wisdom as being inherently good and in its own right honourable, even if nothing useful for the human way of life comes about from it. (Iambl. Protr. 9, 79. 7–84 .6)

A quick inspection su¶ces to show that the opening and closing contributions of Iamblichus are limited to a single sentence each.

Created on 4 July 2005 at 14.34 hours page 262

262

D. S. Hutchinson @ Monte Ransome Johnson

Once Iamblichus’ programmatic remarks are removed, we are obviously left with a very substantial quotation of Aristotle. It is a teleological argument for philosophy, evidently an early version of the ‘ergon’ argument also found in both Eudemian Ethics 2. 1 and Nicomachean Ethics 1. 7. Here the argument begins by making background assumptions explicit, comparing luck with skill and nature. Things that come into being by nature come into being, like skill but unlike luck, for the sake of something (79. 7–80. 20); and humans come to be by nature, and thus for the sake of something (80. 20–81. 11). This is the thing for the sake of which nature and the god109 have brought us into being. Aristotle, or an interlocutor, asks rhetorically, ‘So what is this thing?’ The answer, of course, is that humans come to be for the sake of wisdom, with the protreptic consequence that they should do philosophy (81. 11–82. 20). A close analysis of the passage, and comparison with the parallels in the extant works, can shed light on interesting facets of both Aristotle’s teleology110 and its ethical applications. This is not the place to perform such an analysis. Our analysis is limited to the structure of the literary text, which, after being stripped of its Iamblichean opening and closing sentences, shows no signs whatever of being anything other than a continuous unbroken quotation, unmodified as far as we can tell, of a long passage, the largest continuous fragment we currently possess of the Protrepticus (as we shall conclude it is below, pp. 282–7). Since it is surely Aristotle who wrote this other version of his ‘ergon’ argument, and the quotation is continuous, we must also attribute to Aristotle the passages about Pythagoras, the rejection of the infinite regress of utility, the Isles of the Blest, and the glorious simile comparing being a research scientist to watching the Olympic Games. 3.5. Chapter 10 Chapter 10 appears to be a relatively straightforward case, consisting of an opening and closing by Iamblichus and a single block quotation, fairly substantial, from Aristotle. Here is a translation of the entire chapter: 109 The curious conjunction ‘nature and the god’ (which is echoed a little later, at 85. 20–1, when Aristotle refers to ‘looking to nature and at the divine’) has a parallel in On the Heavens, where Aristotle also says ‘god and nature create nothing that is without a point’ (De caelo 1. 5, 271a33). 110 Sedley, ‘Teleology Anthropocentric?’, 188–9, uses the argument c. 80. 5 ·.

Created on 4 July 2005 at 14.34 hours page 263

Authenticating Aristotle’s Protrepticus

263

Chapter 10. And yet surely the fact that the benefits that are greatest to us for the human way of life are provided by the observational wisdom, someone will easily discover from the skills. For just as all the sophisticated doctors and most sophisticated athletic trainers pretty much agree that those who are to be good doctors or trainers must be experienced about nature, so good lawmakers too must be experienced about nature—and indeed much more than the former. For some are producers of virtue only in the body, while others, being concerned with the virtues of the soul and claiming to teach about the success and failure of the state, need philosophy much more. For just as in the other skills the best of their tools were discovered by their producers from nature (for example, in the builder’s skill, the plumb line, the ruler, and the compass), for some are grasped with water, others with light and the rays of the sun, and it is by reference to these that we scrutinize what to our senses is su¶ciently straight and smooth—in the same way, the statesman must have certain norms taken from nature itself, i.e. from the truth, by reference to which to judge what is just and what is good and what is advantageous. For just as in building these tools surpass all, so too the finest law is the one that has been laid down most in accordance with nature. But this is not something which can be done by someone who hadn’t done philosophy and become familiar with the truth. And in the other skills people do not generally know their tools and their most accurate reasonings by taking from the primary things; they take them from what is second or third hand or at a distant remove, and get their reasonings from experience, whereas the imitation is of the precise things themselves only for the philosopher, for the philosopher’s vision is of these things themselves, not of imitations. So just as no one is a good builder who does not use a ruler or any other such tool, but approximates them to other buildings, so too presumably if someone either lays down laws for cities or performs actions by looking at and imitating other human actions or political systems, whether of Sparta or Crete or of any other such state, neither is he a good lawmaker nor is he an excellent man; for an imitation of what is not noble cannot be noble, nor can an imitation of what is not divine and secure in nature be immortal and secure. But it is clear that the philosopher is the only producer to have both laws that are secure and actions that are right and noble. For he alone lives looking at nature and at the divine, and, just like some good helmsman, ties the first principles of his life onto things which are eternal and steadfast, goes forth, and lives as his own master. Now then, this knowledge is observational, but it provides us with the ability to produce, in accordance with it, everything. For just as sight is a maker and producer of nothing (for its only function is to judge and to make clear each visible thing), but provides us with the ability to perform an action in accordance with it and gives the greatest help towards our actions (for we should be

Created on 4 July 2005 at 14.34 hours page 264

264

D. S. Hutchinson @ Monte Ransome Johnson

almost entirely motionless if deprived of it), so it’s clear that, though the knowledge is observational, we do thousands of things in accordance with it nevertheless, accept some things and avoid others, and generally gain through it everything good. (Iambl. Protr. ch. 10, 84. 7–86. 9)

What we find Iamblichus quoting is a suggestive elaboration of the familiar Aristotelian principle that art imitates nature (Phys. 194a21–2, 199a16; Meteor. 381b6; De mundo 396b12, cf. Protr. ch. 9, 80. 7–9, 18–20), applied to political science, in which the author explains the idea by means of stock examples from the arts, construction, and medicine. Doctors and trainers must know the nature of the body and its physiology in order to promote its virtues, especially health. The builder’s tools, such as the plumb line, ruler, and compass, utilize aspects of nature, e.g. water and sunlight, in order to guide them in how to construct a building properly.111 Similarly, the politician must draw from nature the principles and norms with reference to which what is good is to be done.112 Aristotle here advocates a more idealistic method of political science than a strictly empirical method involving surveying existing constitutions and selecting from these the best institutions and forms of government. Whether or not this provides support for a developmental thesis, as Jaeger contends,113 Aristotle also criticizes the merely empirical method of political science at the end of the Nicomachean Ethics, a parallel that su¶ces to attribute the non-Iamblichean material to Aristotle: Those sophists who advertise that they teach political science seem to be very far from doing so. In fact, they are absolutely ignorant of what sort of science it is and of what sorts of things it is about; otherwise they would not class it as identical with, or even inferior to, the art of rhetoric.114 Nor would they think that it is easy to frame a constitution by making a collection of such existing laws as are reputed to be good ones, selecting the ones that are best as if even this selection did not call for understanding, 111 Compare On the Heavens, where Aristotle says that the Pythagoreans took their emphasis on the number three ‘from nature as, so to speak, a law of it’ (De caelo 268a13–14). 112 Compare Nicomachean Ethics 5. 10, where Aristotle says that ‘the things that are just according to humans, and not nature, are not the same everywhere, since neither are political constitutions [the same everywhere]—but only one [constitution] is everywhere the best according to nature’ (1135a3–5). 113 Jaeger, Entwicklung (Eng. trans.), 260–2. 114 This appears to be a direct contradiction of Isocr. Ant. 80, which is also thought by many to be a target of Aristotle’s Protrepticus.

Created on 4 July 2005 at 14.34 hours page 265

Authenticating Aristotle’s Protrepticus

265

and judging correctly were not the main issue, as it is in music. (Arist. NE 10. 9, 1181a12–19, trans. after Ross in Revised Aristotle)

It would be excellent to learn more positively about this ideal knowledge of nature, but it would seem that much of the argumentation about the benefit of philosophy for the political life that we might expect to be contained in Aristotle’s Protrepticus does not appear here. Iamblichus seems to have substituted an uninspiring and not very relevant summary in his closing. Reasons for this, should one wish to speculate, are not hard to find.115 3.6. Chapter 11 Chapter 11 contains an accessible protreptic application of Aristotle’s method of distinguishing, analysing, and prioritizing homonyms. Structurally, the chapter is straightforward enough, containing a fairly obvious opening and closing, and at most a single bridge passage in the middle. Here is the whole chapter: Chapter 11. Further, that living pleasantly too belongs most of all to those who choose to live according to intellect might become clear from the following. The word ‘living’ seems to mean two things, one with reference to a capacity and the other with reference to an activity, for we call all those animals ‘seeing’ who have vision and are naturally capable of seeing (even if they happen to have their eyes shut), as well as those who are using the capacity and are applying their vision. And similarly with knowing and having cognition, 115 Part of the explanation is undoubtedly sociological: the political situation in 5th-cent. bce Athens and its ideal of the wise lawgiver was simply not applicable in 3rd cent. ce Syria, a Roman principality. Apparently, Iamblichus had acquaintances or even admirers in the imperial administration (Dillon, ‘Iamblichus’, 875 n. 46). What can be gleaned about Iamblichus’ ethics and politics indicates that he was inclined towards an austere Platonism, rather than by Peripateticism. As Dillon points out: ‘A treatise such as the “Protrepticus” might be thought to give evidence of his ethical doctrine, and it is in fact useful, but not as much as might be expected. For one thing, it is largely a cento of quotations, so that one can only deduce Iamblichus’ views from the authorities he uses. From these, however, one may conclude that, like Plotinus and Porphyry, he inclined to a more austere tradition of Platonism, influenced by Stoicism and Pythagoreanism rather than Peripateticism’ (‘Iamblichus’, 902). In general, Iamblichus was absorbed with otherworldly philosophy and shows little direct interest in civic virtue and politics. Where Plotinus distinguishes between ‘civic’ and ‘purificatory’ virtues in Ennead 1. 2, Iamblichus distinguishes eight kinds of virtue. The lowest three are natural, ethical, and civic, and these are either proper to animals and children, or to humans, but with respect to their irrational element only. The rest of the virtues (purificatory, theoretic, paradigmatic, and hieratic) are focused on the assimilation of the human being to the divine (Dillon, ‘Iamblichus’, 902–4). But see recently D. O’Meara Platonopolis (Oxford, 2003).

Created on 4 July 2005 at 14.34 hours page 266

266

D. S. Hutchinson @ Monte Ransome Johnson

we mean, in one case, using and observing, and in the other case, having acquired the capacity and having the knowledge. Further, if we distinguish living from not living by perceiving, and perceiving means two things—in the strict sense, for the using of the senses, but in the other sense, for the having the capacity to use them (that’s why we say, it seems, even of people who are sleeping that they are perceivers)—if so, it’s clear it will follow that ‘living’ also means two things: a waking person should be said to live in the true and strict sense, but sleeping people must be said to live because they are capable of making the transition into the process in virtue of which we say of someone that he is both waking and perceiving things. So when some one word means each of two things, and one of the two is so called either by acting or being acted on, we shall attribute the term as applying more to this one: for example, we attribute ‘knowing’ to the one who makes use of knowledge more than the one who has it, and ‘seeing’ to the one who is applying his vision more than the one who has the capacity. (For we use ‘more’ not only in respect of excess in things which fall under one definition, but also in respect of what is prior and posterior; for example, we say that health is more a good than the things that conduce to health, and that what is valuable in itself by its own nature is more a good than what produces it. And yet we see, surely, that it is not by the definition of ‘good’ being predicable of both that it applies to each of them, beneficial things as well as virtue.) Therefore the waking person should be called more ‘alive’ than the sleeping one, and the one who exercises his soul than the one who merely has it; for it is on account of this that we say that he is alive, that he is the sort who is such as to act or be acted upon in this way. Thus this is what it is to use anything: if the capacity is for a single thing, when someone is doing this very thing, and if the capacity is for a number of things, when he is doing the best of them, for example with flutes, one uses them either only when playing the flute, or especially then; for presumably this applies to the other cases. Thus we should say that someone who uses a thing correctly is using it more, for the natural objective and mode of use belong to someone who uses a thing nobly and accurately. Now the function of the soul, either alone or most of all, is thinking and reasoning. Therefore it is now simple and easy for anyone to reach the conclusion that he who thinks correctly is more alive, and he who most attains truth lives most, and this is the one who is wise and observes according to the most precise knowledge; and it is then and to those that living perfectly, surely, should be attributed, to those who are using wisdom, i.e. to the wise. Now if for any animal, at least, to live is the same as to exist, it is clear that a wise person would surely exist to the highest degree and in the strictest sense, and most of all at that time when he is being active and actually observing the most knowable of existing things. And yet, surely

Created on 4 July 2005 at 14.34 hours page 267

Authenticating Aristotle’s Protrepticus

267

the perfect and unobstructed activity has its enjoyment in itself; hence the activity of observation would be the most pleasant of all. Furthermore, there is a di·erence between enjoying oneself while drinking and enjoying drinking; for nothing prevents someone who is not thirsty, nor has been brought the drink he enjoys, from enjoying himself while drinking, not because he is drinking but because he happens at the same time to be seeing or being seen as he sits there. Thus we will say that this fellow enjoys himself, and enjoys himself while drinking, but not because he is drinking, and not because he enjoys drinking. Thus in the same way we will also call walking and sitting and learning and every process pleasant or painful, not in so far as we happen to feel pain or pleasure in their presence, but in so far as we all feel pain or pleasure by their presence. So, similarly, we will also say that they live pleasantly whose presence is pleasant to those who have it; and we will say that not all to whom it happens that they enjoy themselves while living are living pleasantly, only those to whom living itself is pleasant and who enjoy the pleasure that comes from life. Thus we attribute living more to the one who is awake rather than to the one who is asleep, to the one who is wise more than to the one who is foolish, and we say the pleasure that comes from life is the one that comes from the uses of the soul, for this is being truly alive. Further, even if there are many uses of the soul, still the most dominant one of all, certainly, is the use of wisdom to the highest degree. Further, it is clear that the pleasure that arises from wisdom and contemplation must be the pleasure that comes from living, either alone or most of all. Therefore living pleasantly and feeling true enjoyment belong only to philosophers, or to them most of all, for the activity of our truest ideas, filled up by the most real of things and preserving steadfastly for ever the perfection vouchsafed to us, that activity, of all of them, is also the one that is most e¶cacious of cheerfulness. Hence too on account of the enjoyment itself of the truths and good pleasures those of us with any sense should do philosophy. (Iambl. Protr. ch. 11, 86. 10–89. 6)

The opening of this chapter follows the pattern of forward-looking, summarizing tag line that Iamblichus so frequently uses to introduce a block quotation. The closing, while longer, involves many of the other kinds of trait we have come to expect from Iamblichus: otiose inferential particles, excessive enthusiasm, pretentious diction, and generally a lack of progressive argumentation and syllogism. The complete argument is extremely complex, but we shall o·er an outline of it so far as this furthers our goal of assaying the words as attributable to Aristotle: philosophers live most pleasurably and exist to the highest degree because they actively engage the highest

Created on 4 July 2005 at 14.34 hours page 268

268

D. S. Hutchinson @ Monte Ransome Johnson

and best part of their soul; the predicates ‘living’ and ‘pleasurable’ apply ‘more’ to certain things that, it turns out, philosophers do to the highest degree; homonyms are identified (that is, di·erent things to which the same name applies), and then a procedure is employed in order to determine to which of the things the comparative ‘more’ and superlative ‘most’ apply. The kind of analytical procedure on display here is well known from the extant works. Each of the terms is said to apply to two different things on the basis of a distinction between capacity (δναµις) and activity (νργεια),116 as well as possession (Aχειν, κεκτσθαι) and use (χρσθαι). In the Topics, a text possibly contemporaneous with the Protrepticus,117 Aristotle asserts that perception118 and knowledge119 are homonyms on the basis of the same distinction between use and possession. In the Rhetoric, he advises that protreptic speeches mentioning wealth should observe that true wealth consists in use and not mere possession.120 In chapter 11 of the Protrepticus the distinction is illustrated with reference to vision: a seeing animal is one that either has the capacity for sight (even if its eyes are closed) or one that is actively looking at something. But the same procedure is used to analyse living, perceiving, and knowing. Knowing is argued to be the highest of this series. The philosopher, through contemplation, knows to the highest degree, and so lives to the highest degree. And since contemplation is a continuous activity, and continuous activities are the most pleasant, it follows that the philosopher lives the most pleasantly. The focus on vision is not only typical of Aristotle’s views (one thinks immediately of the opening of the Metaphysics) but is also connected to the passage on vision quoted by Iamblichus from Aristotle in chapter 7. Let us briefly take stock of these inner five chapters before pro‹ de Strycker, ‘Pr‹edicats univoques et pr‹edicats analogiques dans le 116 See E. “Protreptique” d’Aristote’, Revue philosophique de Louvain, 66 (1968), 597–618; and Menn, ‘Origins’, 78–87. 117 Brunschwig puts the date for the Topics between 360 and 348 (J. Brunschwig, Aristot‹el›es: Topiques (Paris, 1967), xc), and this encompasses the consensus dating for the Protrepticus (353–351 according to During, Attempt, 174). • 118 τ α#σθ νεσθαι πλεω σηµανει, 5ν µBν τ α:σθησιν Aχειν, Xν δB τ α#σθσει χρσθαι (Top. 5. 2, 129b33–4). 119 τ πστασθαι . . . πολλ2 σηµανει, τ µBν γ2ρ πιστµην Aχειν . . ., τ δ9 πιστµHη χρσθαι (Top. 5. 2, 130a19–21). 120 Mλως δB τ πλουτεν στιν ν τ/% χρσθαι µVλλον T ν τ/% κεκτσθαι· κα γ2ρ S νργει στι τ%ν τοιοτων κα S χρσις πλοτος (Rhet. 1. 5, 1361a23–5).

Created on 4 July 2005 at 14.34 hours page 269

Authenticating Aristotle’s Protrepticus

269

ceeding to the outer two. Most of these blocks (seven out of nine) can separately be shown to be by Aristotle, and every chapter contains at least some evidence that proves this. The blocks in chapter 7 contain material which Aristotle says he elaborated in his ‘exoteric arguments’, and the last block of chapter 8 was known to Cicero from a published work. As for which published work of Aristotle the blocks come from, the best candidate for each chapter is his Protrepticus, with the possible exception of chapter 8, whose ending has also been attributed to his Eudemus. 3.7. Chapter 6 Now that we have examined these five chapters, the inner core of the section that has been attributed to Aristotle, we are in a position to examine the outer extremities, the two chapters that open and close the Aristotle section. Chapter 6 is a more complex chapter: the opening and closing are fairly straightforward and easy to identify as Iamblichean in grammar, diction, and purpose, but the stripping o· of the words of Iamblichus does not reveal a single block quotation; we believe we have identified three linked bridge passages which divide the chapter into four blocks. Here is the beginning: Chapter 6. But since ‘our conversation is with humans’, not with those who have easy access to the divine share in life, with these kinds of invocations we should mix in some exhortations to the political and practical way of life as well. Let’s put it this way: The things that are supports for our way of life, e.g. a body and what’s around it, support it in the manner of certain tools, the use of which is dangerous, and rather the contrary is accomplished by those who use them in ways they shouldn’t. Well then, we should desire both to acquire this knowledge and to use it appropriately, this knowledge through which we will put all these things to good use. Hence we should do philosophy, if we are going to engage in politics correctly and conduct our own way of life in a beneficial way. Furthermore, there is a di·erence between the kinds of knowledge that produce each of the things of which we want to have more and more in our way of life, and the kinds of knowledge that make use of these kinds of knowledge, and the ones that give service are di·erent from the others that issue orders; and in these as it were more commanding kinds of knowledge exists what is good in the strict sense. If, then, only that kind of knowledge which (a) has correctness of judgement, and (b) uses reason, and (c) observes the good as a whole—that is to say, philosophy—is naturally capable of using all of them and issuing orders, we ought, by all

Created on 4 July 2005 at 14.34 hours page 270

270

D. S. Hutchinson @ Monte Ransome Johnson

means, to do philosophy, since only philosophy includes within itself this correct judgement and this wisdom to issue orders without errors. (Iambl. Protr. ch. 6, 67. 20–68. 14)

The opening remark from Iamblichus is a textbook case of a navigational passage whose function is to indicate what comes next, and the decision where to draw the line between the voices is easy: after ‘Let’s put it this way’. As he sometimes likes to do, Iamblichus tosses in a stylish tag, this one from Plato’s Laws 5 (732 e 3), since ‘our conversation is with humans’, making this remark in his opening in order to announce a new start, saying that he will provide some ‘exhortations to the political and practical way of life121 as well’. The three linked bridge passages that we identify in this chapter have a common structure that gives them away as interpolations of the excerptor. Each of them briefly summarizes the previous argument, and then briefly anticipates the argument to come. The mood and content of the passages stand out from the quoted text, and it is instructive to read them together: Furthermore, since everyone chooses what is possible and what is beneficial, it must be indicated that both these features belong to philosophy, and also that the di¶culty of acquiring it is more than outweighed by the magnitude of its benefit; for we all work at the easier tasks with greater pleasure. Now then, that we are capable of acquiring the kinds of knowledge about the just and the expedient and also the ones about nature and the rest of truth, it is easy to demonstrate. (68. 14–21) Now then, that there is a kind of knowledge of the truth and of the virtue of the soul, and why we are capable of acquiring these, let us take these comments on those as given; and that it is the greatest of goods and the most beneficial of all will be clear from what follows. (69. 20–4) And as to the benefit and the greatness of the thing, I think this has been su¶ciently proven; but why it is much easier to possess it than other goods, one might be convinced by the following. (70. 20–2)

These are evidently markers in the context of a multi-phase ar121 Although what immediately follows does relate to the practical life—how philosophy enables one to acquire the knowledge necessary to make proper use of instruments—and the conclusion is drawn that philosophy will be useful for politics and in our own lives (68. 2–3), the rest of the chapter is decidedly not related to the practical and political life: it makes various observations about how it is possible and worthwhile to learn di·erent kinds of science ‘despite no reward coming to those who do philosophy’ (70. 23–4).

Created on 4 July 2005 at 14.34 hours page 271

Authenticating Aristotle’s Protrepticus

271

gument whose main purposes are to show regarding philosophy that it is a feasible subject (first bridge, 68. 18–21), that it is highly beneficial for us to acquire (second bridge, 69. 20–4), and that it is well worth the trouble (third bridge, 70. 20–2). This scheme is announced in advance in the first bridge (68. 14–18) and recollected at the end, in the closing (71. 9–12), both times in the same order. What this seems to indicate is that Iamblichus is working from a longer passage in Aristotle’s book and is careful to preserve the order and structure of the overall protreptic argument: philosophy is possible, beneficial, and worth the trouble. This seems a clear parallel to what we saw in Iamblichus’ version of Plato’s Phaedo (chapter 13): fast-forward passages are constructed by Iamblichus to skip quickly on to the next highlight, without his readers losing sight of the big picture. Having identified the interventions of Iamblichus, we predict that what remains is from Aristotle. This can be confirmed both by parallels in the extant works of Aristotle and by other kinds of evidence. The first block (67. 23–68. 14) contains an argument that we need knowledge of the proper use of instruments, including the organs of the body, in order to be successful both in politics and in our own lives, because the improper use of instruments is harmful to others and ourselves. There are striking parallels in this passage with ideas unique to Aristotle. That our bodies are tools (used by our souls) is a position explicitly argued for by Aristotle throughout his psychological and biological works.122 At the beginning of this block, Aristotle distinguishes between ‘the kind of knowledge that [enables] us to get more and more of what we want in life’ and another kind of knowledge, an ordering or prescriptive kind of knowledge, and at the end he refers to ‘prescriptive wisdom’ (68. 7, 11, 13). What he means here is made explicit in the Nicomachean Ethics when he says that wisdom (φρ+νησις) is a prescriptive kind of knowledge (πιτακτικ) because its end is the determination of what should and should not be done.123 He also makes the same point explicitly in the Eudemian Ethics: ‘Since the intellectual [virtues] involve reason, they belong to the [part of the soul] possessing 122 Especially DA 2. 1, 412a27–412b10; also 415b18–20, PA 645b14. For a recent searching account of this idea, see S. Menn, ‘Aristotle’s Definition of Soul and the Programme of the De anima’, Oxford Studies in Ancient Philosophy, 22 (2002), 83–139. 123 S µBν γ2ρ φρ+νησις πιτακτικ στιν· τ γ2ρ δε πρ ττειν T µ, τ τλος α?τς στν (NE 1143a8–9).

Created on 4 July 2005 at 14.34 hours page 272

272

D. S. Hutchinson @ Monte Ransome Johnson

reason, which prescribes to the soul by possessing reason, but the ethical [virtues belong to] the irrational [part of the soul], which follows the one possessing reason by nature.’124 For these reasons, we should take everything between the obvious opening (‘Let’s put it this way’ at 67. 23) and the conclusion of the next argument (‘we ought, by all means, to do philosophy’ at 68. 11–12) to be a quotation of Aristotle. Which Aristotle work is Iamblichus here exploiting? Simply by reflecting on the overall line of argument which Iamblichus sketches in his bridge passages we are driven to the inescapable answer: the only work which really comes into consideration as the home of a long multi-phase argument that philosophy is possible,125 beneficial, and worthwhile is his Protrepticus. It is di¶cult to see in what kind of work other than a protreptic this stretch of argument would be appropriate. When we look at the details of how these three conclusions are reached, in the last three blocks of the chapter, we shall find further corroboration. This argument, based on the overall argumentative structure of the chapter, can be buttressed by another consideration, based on the single distinctive word φιλοσοφητον (‘one should do philosophy’). As we pointed out at the beginning of this paper, this and the title are the only two things we know for sure about Aristotle’s Protrepticus: a work with that title was written by Aristotle, and it contained the conclusion ‘one should do philosophy’ (the reports of this fact name both Aristotle and the Protrepticus). This term occurs five times in the Aristotle part of Iamblichus’ book—in chapters 6, 7, 8, and 12—but not once in the Plato section.126 So let us hold on to this conclusion, that the chapter is derived 124 πε δ9 α> διανοητικα µετ2 λ+γου, α> µBν τοιαται το λ+γον Aχοντος, I πιτακτικ+ν στι τς ψυχς HN λ+γον Aχει, α> δ9 Yθικα το "λ+γου µν, "κολουθητικο δB κατ2 φσιν τ/% λ+γον Aχοντι (EE 1220a8–11). 125 In the section of the Rhetoric devoted to protreptic speech, Aristotle dwells on how one might establish that the good in question is possible (Rhet. 1. 6, 1363a21–4). 126 In the Aristotle section: 68. 2 and 68. 11–12, 71. 20–1, 78. 21, and 89. 22– 3. It occurs twice in chapter 5 (61. 11, 64. 28), a chapter with a very complex structure, which it would require another study to elucidate, and once in the closing to ch. 11 (89. 5). It occurs once in a bridge passage written by Iamblichus while he is excerpting from the anonymous author (‘Anonymus Iamblichi’) he quotes in chapter 20 (124. 16), and he uses it in two other places as a generic protreptic prescription in chapter 21, his commentary on the Pythagorean symbolae (142. 22 and 143. 25). But all six of these cases can be explained by supposing that Iamblichus found the formula repeatedly in Aristotle’s work, and then proceeded to use it himself in his own contributions to the excerpts in his book.

Created on 4 July 2005 at 14.34 hours page 273

Authenticating Aristotle’s Protrepticus

273

from Aristotle, and from his Protrepticus, and proceed to study the next block, the second one: Furthermore, since we all choose what is possible and what is beneficial, it must be indicated127 that both these features belong to philosophy, and also that the di¶culty of acquiring it is more than outweighed by the magnitude of its benefit; for we all work at the easier tasks with greater pleasure. Now then, that we are capable of acquiring the kinds of knowledge about the just and the expedient and also about both physical nature as well as the other kind of truth, it is easy to demonstrate. For prior things are always more familiar than posterior things, and what is better in nature than what is worse. For there is more knowledge of what is determinate and orderly than of their opposites, and again of the causes than of the e·ects. And good things are more determinate and organized than bad things, just as a fair person is more determinate and organized than a foul person; for they necessarily have the same mutual di·erence. And prior things are causes more than posterior things, for if they are taken away, then so are the things that take their being from them: if numbers <are taken away>, then so are lines, if lines then surfaces, and if surfaces then solids.128 Hence since soul is better than body (being more authoritative in nature), and the kinds of skill and wisdom concerned with the body are medical science and athletic training (for we regard these as being kinds of knowledge and say that some people possess them), clearly for the soul too and the virtues of the soul there is a certain discipline and skill, and we are capable of acquiring it, since surely we are also capable of acquiring knowledge of things of which our ignorance is greater and cognition is harder to come by. Similarly too for the natural sciences; for wisdom about the causes and the elements is a much higher priority than wisdom about what is secondary; for these are not among the highest, nor from them that the first principles naturally grow, rather it’s from those that all other things come into being and are evidently constituted. For whether it is fire or air or number or any other natures that are the causes and first principles of other things, it would be impossible to be ignorant of these things and to recognize any of the other things; for how could anyone either be familiar with speech who was ignorant of syllables, or have knowledge of these who understands nothing of the letters? (Iambl. Protr. ch. 6, 68. 14–69. 19)

In the first block of this chapter, Aristotle had concluded that phi127 Rejecting the conjectures "ποδεικτον (Pistelli) and παραδεικτον (During) and • returning to the manuscript’s παραδεικτον. Iamblichus uses the term παρ δειξις and its verbal forms repeatedly throughout the DCM. 128 Omitting 69. 2–3 στοιχεα . . . συλλαβ%ν as a confusing gloss, not present in the parallel passage in Iamblichus’ DCM.

Created on 4 July 2005 at 14.34 hours page 274

274

D. S. Hutchinson @ Monte Ransome Johnson

losophy is an ordering knowledge, because it has correctness of judgement, uses reason, and observes the good as a whole (i.e. the good in the strict sense, without qualification, the authoritative good); and further that one ought to do philosophy (φιλοσοφητον, 68. 11–14), since this is the most authoritative kind of wisdom. This next block of argument is part of a larger stretch which aims to show that such an authoritative science does in fact exist, despite arguments to the contrary (presumably these arguments to the contrary were contained in the unpreserved section of the Protrepticus that Iamblichus skipped over between the end of the first block at 68. 14 and the start of the second block at 68. 21). In Metaphysics Α Aristotle says that the most authoritative knowledge is based on the first principles and causes: The first principles and causes are the most knowable; for by reason of these, and from these, all other things are known, but these are not known by the things subordinate to them. And the science which knows to what end each thing must be done is the most authoritative of the sciences, and more authoritative than any ancillary science; and this end is the good in each class and in general the supreme good in the whole of nature. (Metaph. Α 2, 982b2–7; cf. NE 1. 1, 4. 1, and 8. 3)

In the block just quoted, Aristotle gives a detailed account of how it is in fact possible to come to know first causes and principles. Thus we have a definitively Aristotelian position—that the most authoritative knowledge is of primary principles and causes—but di·erently put and supported, with an explicitly protreptic purpose. At this point we need to turn aside and consider an objection to our attribution of the material in these chapters to the Protrepticus of Aristotle. In his polemic against the reconstruction of Aristotle’s book, W. G. Rabinowitz focused his attack on one idea in the first block of chapter 6 (68. 28–69. 2): ‘And prior things are causes more than posterior things, for if they are taken away, then so are the things that take their being from them [τ2 τ.ν ο?σαν ξ κενων Aχοντα]: if numbers <are taken away>, then so are lines, if lines then surfaces, and if surfaces then solids.’ According to Rabinowitz, It will be noted that at Protrepticus 38. 11–14 [Pistelli = 68. 28–69. 2 Des Places] the author writes as one committed to a doctrine according to which lines are generated from numbers, planes from lines, and solids from planes. Whoever may have been the originator of this doctrine, it is at least clear

Created on 4 July 2005 at 14.34 hours page 275

Authenticating Aristotle’s Protrepticus

275

that Aristotle did not hold it, for he knows and attacks the doctrine as a Platonist tenet again and again in the Metaphysics. (Sources, 85, emphasis added)

But this seems to be a bad mistake: Aristotle expresses no commitment to the ‘doctrine’ stated by Rabinowitz—that bodies are generated out of planes, planes out of lines, lines out of points, and points out of number. That is indeed a thesis of certain Platonists rejected by Aristotle in the passages cited by Rabinowitz; but of course this does not mean that Aristotle rejects the priority relation in other ways, such as in knowledge or account.129 In one of the passages cited by Rabinowitz himself130 against the idea of generation from the geometrically prior, Aristotle says, ‘We find similar di¶culties in the case of the kinds posterior to number—the line, plane, and solid’ (Metaph. Μ 8, 1085a7–9). This is all that Aristotle needs to be saying in the Protrepticus passage, and it is perfectly reasonable to refer to these geometrical concepts as prior, in accordance with the conventions of the geometrical sciences, and still reject the speculative idea that solid bodies are physically generated out of numbers, points, lines, and planes understood as separate substances and bodies. In the second Protrepticus block in chapter 6, we find other indications that the distinctions, technical terms, and arguments in the quoted text are the building blocks and stock in trade of Aristotle’s analytical philosophy. For example, Aristotle says in this block that causes are prior to e·ects; the orderly is prior to the disorderly; and the naturally good is prior to the bad; and this is a doctrine widely evidenced in several surviving works.131 Another indication: it is 129 In fact, it is evident that he follows conventional geometrical science on this point, as he says without ambiguity in the Categories: ‘we use the term “prior” in regard to any order, in both sciences and their arguments. In the sciences, using demonstration, we have what is prior and what is posterior in order. For the elements are prior in order in the geometrical sciences, and in grammar the letters are prior to the syllables’ (Cat. 12, 14a36–b2). 130 Metaph. 992b13–18, 1001b26–1002b12, 1028b15–27, 1060b6–19, 1076b11–39, 1077a24–36, 1080a12–b36, 1085a7–b34, 1090b5–32. 131 Aristotle often distinguishes between things better known to us and things better known by nature. ‘Things are prior and more familiar in two ways; for it is not the same to be prior by nature and prior in relation to us, nor to be more familiar and more familiar to us. I call prior and more familiar in relation to us what is nearer to perception, prior and more familiar absolutely what is further away. What is most universal is furthest away, and the particulars are nearest; and these are opposite to each other’ (Post. An. 1. 2, 71b33–72a5). See also Pr. An. 2. 23, 68b35–7; Top. 6. 4, passim; Phys. 1. 1; Metaph. Ζ 3, 1029b1–12; NE 1. 4, 1095b2–3.

Created on 4 July 2005 at 14.34 hours page 276

276

D. S. Hutchinson @ Monte Ransome Johnson

announced in the Protrepticus (by Iamblichus, in the bridge passage at 68. 18–21) that it is possible to acquire knowledge, not only about ‘the just and the expedient’, but also about ‘physical nature’, and about ‘the rest of truth’ as well, a division found in Aristotle’s Topics (1. 14, 105b20–1); and the arguments of this second quoted block make these three points, just as Iamblichus had announced, about logic and epistemology (68. 21–69. 2), about practical ethics (69. 3–10), and about natural science (69. 10–19). The next block of quotation, the third of four in this chapter, develops the notion of authoritative wisdom that this elaboration of causes and principles was designed to substantiate: Now then, that there is a kind of knowledge of the truth and of the virtue of the soul, and how we are capable of acquiring them, this is what we have said about those topics; and that it is the greatest of goods and the most beneficial of all will be clear from what follows. For we all agree that the most excellent person ought to have authority, i.e. the one who is the supreme in nature, and ought to be alone in a dominant position and to have authority over the law; and this is a sort of wisdom, i.e. a statement based on wisdom. And again, what norm do we have or what more precise standard of good things, than the wise man? For all things that this person will choose, if the choice is based on their knowledge, are good things and their contraries are bad. And since everybody chooses most of all what conforms to their own proper dispositions (a just man choosing to live justly, a man with bravery to live bravely, likewise a self-controlled man to live with self-control), it is clear that the wise man will choose most of all to practise wisdom; for this is the function of that capacity. Hence it’s clear that, according to the dominant judgement, wisdom is the supreme among goods. So one ought not to flee from philosophy, since philosophy is, as we think, both a possession and a use of expertise, and expertise is among the greatest goods; nor should we sail to the Pillars of Heracles and run many risks for the sake of property, while devoting neither e·ort nor expense for the sake of wisdom. It would surely be slave-like to crave living rather than living well, for one to follow the opinions of the majority rather than evaluating the worth of the majority in terms of one’s own opinions, and to seek out property but for what is noble to take no trouble whatsoever. (Iambl. Protr. ch. 6, 69. 20–70. 19)

The argument from 69. 27 onwards is that the wise person is authoritative and is the norm or standard (Mρος) of the good, because the wise person is in possession of the criterion (καν(ν) of the good, and so can be looked to as a model for the patterns of behaviour and choice that are good. The notion of a good man serving as the

Created on 4 July 2005 at 14.34 hours page 277

Authenticating Aristotle’s Protrepticus

277

norm and standard of the good is a leading idea in Aristotle’s philosophy,132 which is here brought into play as a motivation in favour of the study of philosophy. Once again, we see familiar Aristotelian ideas being deployed for protreptic purposes. The last block of chapter 6 is evidently a shorter selection from a longer argument that the rewards of philosophy easily outweigh the alleged di¶culty of the subject; this shorter section evidently focused on undercutting the impression that philosophy is di¶cult, a conclusion reinforced by Iamblichus’ typical one-sentence reiterating conclusion: And as to the benefit and the greatness of the thing, I think this has been sufficiently proven; but as to why it is much easier to possess it than other goods, one might be convinced by the following. For, despite no reward coming from people to those who do philosophy, which would make them keen to exert considerable e·ort in this way, and despite having given the other skills a big head start,133 nevertheless the fact that in running a short time it has surpassed them in precision seems to me to be a sign of the easiness of philosophy. And again, the fact that everybody feels at home with philosophy and wishes to occupy their leisure with it, renouncing everything else, is no slight evidence that the close attention comes with pleasure; for no one is willing to labour for a long time. In addition to these, its practice greatly di·ers from all others: philosophers need neither tools nor special places for their job; rather, wherever in the inhabited world anyone’s thought runs, one apprehends the truth everywhere equally as if it were present there. Thus it has been proven that philosophy is possible, and why it’s the greatest of goods and is easy to possess; hence on all counts it is worthwhile that we should take it to heart. (Iambl. Protr. ch. 6, 70. 20–71. 12)

Students of ours who have read this passage have expressed open scepticism about the truth of this conclusion, that philosophy is easy. Fair enough, but there is no reason to doubt that Aristotle did entertain this view;134 and he regards it as a consideration in favour of philosophy and the sort of scientific wisdom that comes from objective observation that it does not require its practitioner to be 132 For the φρ+νιµος as the standard of the good, see also Top. 3. 1, 116a13–17; NE 6. 5, passim; 6. 12, 1144a36–b7; Rhet. 1. 6, 1363a17. For the idea that the φρ+νιµος should rule, see Pol. 3. 4, 1277a14–16. On the καν(ν of the good, see NE 3. 7, 1113a33. For the Mρος of the good, see NE 6. 1, 1138b21–5, 32–4. 133 Omitting ε#ς in 70. 25, as in the parallel passage at DCM 82. 19. 134 In fact, Aristotle recommends discussing the ease of the activity one is encouraging in one of the chapters of the Rhetoric devoted to protreptic speech (1. 6, 1363a21–4).

Created on 4 July 2005 at 14.34 hours page 278

278

D. S. Hutchinson @ Monte Ransome Johnson

dependent on particular political allies, or on expensive support (NE 10. 7). Before leaving chapter 6, we should note that the promissory note issued by Iamblichus at its beginning (‘we should mix in some exhortations to the political and practical way of life as well’) is fulfilled in a later chapter: in Protrepticus chapter 10, Aristotle offers the argument that, as in the other skills, the politician ‘must possess a norm of the good with which he can judge what is just and noble and beneficial’ (84. 24–7), and so a real political scientist must have a grasp of reality as well as a knowledge of comparative politics. This connection indicates that chapter 6 and chapter 10 are taken from the same book of Aristotle. 3.8. Chapter 12 Next, let us turn to the last and shortest chapter of the Aristotle section, chapter 12. Here we have a single opening sentence from Iamblichus, but it appears that he has written more than a single sentence to close the chapter, which also closes his whole Aristotle section. Here is the entire chapter: Chapter 12. If we need, not only to reach this conclusion on the basis of the parts of success, but also to establish it on a higher basis, of success as a whole, let us state explicitly that philosophizing stands in the same relation to success as it does to our being good or bad. For all the things that are for this or because of this are to be valued by all, and of the things that make us successful some are as necessities, others are enjoyable. Thus we take the position that success is either wisdom and a certain expertise, or virtue, or great enjoyment, or all these things. Thus if it is wisdom, clearly only philosophers will have a successful life; and if it is virtue of the soul or enjoyment, even then it will belong to them alone or most of all, for the most dominant thing in us is a virtue, and wisdom is the most pleasant of all when one is compared to another. And similarly, even if one says that all these things together are the same as success, it is to be defined as wisdom. Hence everyone who is able to should do philosophy [φιλοσοφητον]; indeed, either this is living perfectly well, or it is, at any rate, the greatest cause of it, so to speak, in our souls. But here, perhaps due to our race being unnatural, it is di¶cult to learn about things and investigate them, and even perceive them, due to lack of natural talent and unnatural living; but if we were ever able to find salvation again there where we have been set loose from, clearly we would all do so easily and with pleasure. For as it is now, we neglect the good things and carry on doing the necessities, most of all those most regarded as happy by most people; but if we were to take the heavenly road and settle our lives

Created on 4 July 2005 at 14.34 hours page 279

Authenticating Aristotle’s Protrepticus

279

on our companion star, then we would be doing philosophy, truly living, and observing spectacles indescribable in beauty, gazing with the soul fixedly at the truth and observing the authority of the gods, rejoicing and continuously deriving enjoyment from the observing, taking pleasure apart from all pain. So in this way, approaching in turn every kind of success, we discover that doing philosophy makes a contribution for us; for this reason it is worthwhile taking it up as being the most excellent and most appropriate thing for us. (Iambl. Protr. ch. 12, 89. 7–90. 15)

The opening sentence needs to be stripped away as the obvious work of Iamblichus; it is meta-textual, turgid, and full of the technical words (‘on the basis of’) which he tends to use in this work. The last sentence, however, is not the extent of Iamblichus’ closing remarks; he ends with an allusion to a Platonic idea (‘perhaps due to our race being unnatural’), followed by his own Neopythagorean elaboration. We need to make the separation of voices some distance up135 from the end, after the word ‘souls’ at 89. 25, if only because of the evident, indeed palpable, drop in the rhetorical level. This has been well expressed by Werner Jaeger: To suggest that Aristotle is the author of the conclusion actually found in Iamblichus136 . . . is to let desire stifle critical reflection. Enthusiastic the sentences may be, and even inspired; but it is not the controlled enthusiasm of Aristotle, who never forgoes the strict rhythm of his apodictic advance, and values form higher than the highest inspiration, often as his arguments perceptibly overflow with the latter. Most of the details of Iamblichus’ passage could indeed well have been taken from the Protrepticus . . . but the loose and merely associative conjunction of these notions into an edifying summons to the other world, the confusion of ideas that can be detected in them, the sacerdotal unction with which the writer introduces some of Plato’s ceremonial words, the presence of certain distinctly Neo-Platonic phrases like ‘the heavenly path’ and ‘the realm of the gods’, and lastly the excessive loquacity of the conclusion, with its inability to come to an end—all these things betray retouching by Iamblichus. (Entwicklung, 79, Eng. trans.)

This opinion of the authorial persona of Iamblichus is well ex135 A good parallel is the extended opening remarks to chapter 18 in the Plato section (113. 22–114. 7). 136 Where this conclusion starts was somewhat in dispute: in his edition of the fragments of Aristotle’s dialogues, Walzer divided the text at 89. 22 (60. 7 Pistelli) and marked the concluding part with the reference in parentheses; Ross translated the sentences from 89. 22 to 90. 2 as also coming from Aristotle; but During found • the right division, marking the conclusion as starting at 89. 26.

Created on 4 July 2005 at 14.34 hours page 280

280

D. S. Hutchinson @ Monte Ransome Johnson

pressed and well judged, in our view, especially the point made by Jaeger that Aristotle’s text is always making a steady advance in argument, in contrast to which the contributions of Iamblichus seem to be entirely motionless road signs. The same was very clear in the Plato chapters. When we strip away this extended concluding section, the conclusion we are left with is that ‘everyone who is able to should do philosophy [φιλοσοφητον]; indeed, either this is living perfectly well, or it is, at any rate, the greatest cause of it, so to speak, in our souls’ (89. 22–5). This conclusion is reached, at its last stage, by means of an argument about the three kinds of lives: whether one’s life is centred on pleasure, excellence, or wisdom, one should do philosophy. Aristotle develops the idea of the three main candidate ways of life in no fewer than three places in the extant works: Eudemian Ethics 1. 1–5, Nicomachean Ethics 1. 5, and Politics 7. 1–3. This is evidently the ultimate conclusion of an extended overarching line of reasoning that began back in the ‘three lives’ argument introduced near the beginning of Iamblichus’ chapter 7 (71. 18–22). In our discussion of that chapter (subsection 3.2 above) we saw Aristotle refer these to his ‘exoteric arguments’. It should also be noted that Aristotle, in his treatise on rhetoric, spells out that the discussion of the components of happiness is key to protreptic speech: ‘all exhortations [προτροπα] and counter-exhortations are about happiness and the things which conduce to it or are detrimental to it’ (Rhet. 1. 5, 1360b9–11). So we need to conclude that the book from which the chapter 12 material was excerpted is the same book as the ‘exoteric arguments’ from which the chapter 7 material was drawn. Which published work was the source of the chapter 12 material? The clue is again in the single-word conclusion that appears to be highly distinctive of this book of Aristotle’s: φιλοσοφητον. We saw above (p. 272) that this appears in chapter 6 (and elsewhere), a chapter that surely must be attributed not just to Aristotle but specifically to his Protrepticus; so chapter 12 must be attributed to the Protrepticus. And if chapter 6 and chapter 12, then all the intervening chapters too must come from Protrepticus, as we discovered in the Plato chapters that Iamblichus does not jump around between works; between any two blocks of quotation from the same work, every intervening block of quotation must also be from that work, as we said above (pp. 240–1). That the blocks of quotation in every

Created on 4 July 2005 at 14.34 hours page 281

Authenticating Aristotle’s Protrepticus

281

chapter (not just 6 and 12) need to be attributed to Protrepticus is a conclusion that will be buttressed by supporting arguments in our next and last section.

4. Summary, conclusions, and new status of the question Chapters 9, 10, 11, and 12 consist, we saw, of long passages quoted from Aristotle, surrounded by thin comments by Iamblichus (generally only one sentence); this is similar to his way of quoting from Plato in chapters 14–18. We also saw long passages quoted from Aristotle in the other chapters, but with a complication: in chapters 6, 7, and 8 we detect ‘bridge’ passages, similar in kind to those found in chapters 13 and 19, in which Iamblichus returns to his own voice in order to provide a transition between two separate Plato passages. So over the seven chapters of Iamblichus we detect fourteen openings and closings, and seven bridge passages written by Iamblichus himself—twenty-one contributions. The rest, we contend, is faithfully quoted material from Aristotle—over 500 lines of material quoted, sometimes perhaps slightly modified, from Aristotle’s Protrepticus. Here is a synopsis of the section in which Iamblichus uses Aristotle as a source: Ch. 6. Opening—the correct use of goods as tools—bridge—the priority of causes and elements—bridge—why the wise person has authority— bridge—philosophy is worth the trouble—closing. Ch. 7. Opening—the ‘ergon’ argument—bridge—comparison between observation and vision—closing. Ch. 8. Opening—commonly held beliefs—bridge—wisdom the opposite of lunacy, puerility, drunkenness, sleep, death—bridge—value of property—bridge—worthlessness of commonly valued goods illustrated through parables about the eyes of Lynceus, the Orphic rites, the Tyrrhenian pirates, and the Isles of the Blest—closing. Ch. 9. Opening—teleological argument from ends in accordance with nature; infinite regression of the utilitarian conception of education; ‘intelligence is the god in us’—closing. Ch. 10. Opening—usefulness of philosophy to politics—closing. Ch. 11. Opening—the equivocal predicates ‘living’, ‘perceiving’, ‘seeing’, and ‘knowing’; being wise makes us most fully alive and provides the truest pleasure—closing. Ch. 12. Opening—summary protreptic through the ‘three lives’ argument—Iamblichean elaboration and closing.

Created on 4 July 2005 at 14.34 hours page 282

282

D. S. Hutchinson @ Monte Ransome Johnson

We can infer from this synopsis that much may have dropped out of Iamblichus’ version, because any chapter division or bridge indicates a gap in the source. And if Aristotle’s work was a dialogue, then we can infer that all (or most) traces of that have been erased. That being said, it seems to us that the rest of these arguments, what amounts to about a dozen blocks of unequal length, have all been extracted from Aristotle’s Protrepticus. But why do we think so? After all, we observed from the study of Plato parallels that Iamblichus quotes from multiple sources, and that was our reasonable prediction. Most of the fourteen blocks in these seven chapters have as the best candidate for an original Aristotle home the Protrepticus, with the notable exception of the last block in chapter 8, the one containing the eyes of Lynceus, the Tyrrhenian pirates, and the claim attributed to Anaxagoras that ‘intelligence is the god in us’. Passages in this famous block of text have been claimed to come not from the Protrepticus but the Eudemus; and other passages have been claimed to come not from the middle but the end of the Protrepticus. Neither claim can be true, as the following arguments will show; but if one’s method is to look at each sentence or passage individually, not as blocks carefully assembled in a literary construction, one will not be able to reach more solid conclusions. During our study, we noted five separate themes that link various chapters to each other as being from the same work. When we assemble these links, we find that they define a gridwork of literary unity that cannot reasonably be taken apart. Chapters 7 and 12 are linked by the ‘three lives’ argument mentioned in chapter 7 and concluded in chapter 12; chapters 6 and 10 are linked by their discussion of the practical and political benefits of wisdom; chapters 6, 7, 8, and 12 are all linked by the signature concluding slogan φιλοσοφητον (n. 126); chapters 7 and 11 are linked by elaborate comparisons between perceptual vision and intellectual vision; chapters 7 and 9 contain linked discussions of teleology and the ergon argument.

Created on 4 July 2005 at 14.34 hours page 283

Authenticating Aristotle’s Protrepticus 6 7 8 9 10 11 12 * * * * * * * *

* * * *

283

‘three lives’ argument practical benefits of philosophy φιλοσοφητον perception and intellectual vision teleology and the ergon argument

When we remember that Iamblichus uses a natural sequence, in which he never returns to a work having used it, we can show that each of these chapters is attached to each other by at least two separate links in the gridwork. Of all the chapters, the most solidly attached to the rest is our chapter 8.We have no choice but to place chapter 8 in the same work as the chapters before and after it; and this is in virtue of five separate arguments. All five of these apparent links would have to be illusory for the content of chapter 8 to be detachable from the work. But now that we know that it is all one work, we hardly need to slow down and ask which one. The one work that comes into consideration as possibly containing all this content is Aristotle’s Protrepticus. This conclusion is bolstered by considering evidence that seems, oddly, never to have been brought into play for this topic: what Aristotle recommends in Rhetoric 1. 5–6 for authors of protreptic discourses reads like a recipe for the kind of exhortation that we find in various of these Aristotle chapters of Iamblichus’ Protrepticus; for details, see above, nn. 99, 125, and 134. Does it make sense to think that Iamblichus uses di·erent techniques with Aristotle (only one work) and with Plato (several works)? Of course it does: Iamblichus did not have to search across a variety of works of Aristotle to find a rich treasure trove of protreptic arguments, whereas Plato, who did not really believe in the genre (and may have been positively hostile to it),137 never wrote a Protrepticus true to genre, and scattered his protreptic arguments and reflections across his corpus. Here again, it is the di·erence in the works and authors whom he is quoting that accounts for the di·erence of technique. Iamblichus seems to have made a sevenchapter section out of various works by Plato and combined them 137 S. R. Slings (ed.), Plato: Clitophon, Edited with Introduction, Translation, and Commentary [Clitophon] (Cambridge, 1999), 127–64.

Created on 4 July 2005 at 14.34 hours page 284

284

D. S. Hutchinson @ Monte Ransome Johnson

into a unified protreptic section, in order to mirror the unity he found ready-made in Aristotle’s Protrepticus. So we have excellent reason to believe that chapter 8 is a central part of the Protrepticus; but others have argued that the Tyrrhenian pirates passage should be ascribed to Aristotle’s lost dialogue Eudemus, so we need to consider their arguments. The idea goes back a long time (to 1840),138 but never gathered much real evidential support, and by the time it was championed most recently, by Brunschwig and Flashar, independently in two papers in the 1960s,139 the argument rested on a thin set of considerations. The macabre pessimism of the passage seemed to earlier scholars to fit poorly with an enthusiastic exhortation to philosophy such as Protrepticus, and the mythological treatment of the theme of life as a punishment fits well with some quoted passages from Aristotle’s Eudemus. To these considerations Brunschwig (‘Pirates’, 186–9) added a very indirect one drawn from a mention of the idea that life on earth is for us a punishment, at the beginning of Cicero’s lost work Consolatio, which probably drew on Crantor’s lost work On Grief, which is mentioned in pseudo-Plutarch, Consolatio, just before that source makes a long quotation from Aristotle’s Eudemus. These seemed to Brunschwig su¶cient to reopen the question of the provenance of the passage, but the earlier considerations are fallible, he admits, and for them to be decisive we would have to be confident that their pessimism is irreconcilable with the optimism shown by other fragments of Protrepticus; but ‘nothing is easier to reconcile with optimism than pessimism’ (‘Pirates’, 189). Flashar repeats the old ideas that the horrible Tyrrhenian pirates story does 138 A. B. Krische, Die theologischen Lehren der griechischen Denker: Eine Pr•ufung der Darstellung Cicero’s (Forschungen auf dem Gebiete der alten Philosophie, 1; G•ottingen, 1840), 17; J. Bernays, Die Dialoge des Aristoteles in ihrem Verh•altnis an seinen u• brigen Werken (Berlin, 1863), 144; E. Heitz (ed.), Aristotelis fragmenta (Paris, 1869), 49–50; Rose, in Opera omnia Aristotelis, v. 1480b20–39 (but Rose changed his mind in 1886 and attributed the passage to Aristotle’s Protrepticus); O. Gigon, ‘Prolegomena to an Edition of the Eudemus’, in During and Owen, Mid• Fourth Century, 19–33 at 27–8. In his later collection of fragments for the revised complete works of Arisotle, Gigon evidently abandoned hope of attributing any of the Iamblichus material to any particular lost work of Aristotle, calling them collectively ‘protreptic commonplaces excerpted from several dialogues’ (O. Gigon (ed.), ‘Librorum deperditorum fragmenta’, in Aristotelis opera, iii (Berlin and New York, 1987), 302b13–318b8). This opinion is expressed most recently by Most, ‘New Fragments?’, 208 n. 35. 139 Brunschwig, ‘Pirates’, 185–9; Flashar, ‘Platon und Aristoteles’, 71–3 (independent from each other, as Flashar had not noticed Brunschwig’s article).

Created on 4 July 2005 at 14.34 hours page 285

Authenticating Aristotle’s Protrepticus

285

not make for good upbeat advertising, such as is needed for a protreptic to philosophy, and that the content is appropriate to what we know of Eudemus.140 It seems that these more recent claims have not made any advances over the old subjective impression shared among scholars; and this unsystematic way of reasoning cannot stand up to the example of Plato’s Phaedo, which does manage to combine those themes in a most remarkable and e·ective way. We saw that chapter 8 is not at the end of the work, but somewhere in the middle; since it is clearly a rhetorical climax, it seems to mark the end of an intermediate portion of the book. The order of the work must be the one as given here; but does this make sense? Of course it does. The rhetorical structure of the work is quite lost to us, and many imaginable and plausible scenarios could contain this material in this order. Phillip De Lacy ventured the following appreciation of the material when laid out in this order: The passages seem to be from a rhetorical piece designed to persuade an audience of privileged young men who are preparing for a political life, whose families own slaves . . . that wealth and power are not enough. Philosophical wisdom, which is easily acquired, uses the best part of the soul to open up a vision of the universe that is enjoyable for itself, with no further end in view. But they must also be assured that attaining this beatific vision does not entail abandoning the political life. Philosophical wisdom will help them succeed in all purposeful activities and will indeed put them at the top of the hierarchy of useful pursuits. Thus in both contemplation and action philosophical wisdom will make their lives successful, upright, and pleasant. This summary follows in general the order of excerpts in Iamblichus.141

The cogency of this summary indicates that there is no prima facie need to reorder the line of thought preserved in Iamblichus. The most famous agent of reordering, During, himself remarked: • ‘It is clear that any such attempt [at reconstruction] must be purely speculative . . . My aim has been to arrange the fragments in a reasonable order. Since it is unlikely that we shall ever be able either to prove or disprove our case, I do not attach much importance to this aspect of our problem.’142 It would be only fair to consider 140 He adds as a new consideration a separate echo of the Eudemus in chapter 12 (89. 26): ‘But here, perhaps due to our race being unnatural, it is di¶cult to learn about things and investigate them.’ Yet this is in the voice of Iamblichus, who was in the habit, as we saw in the Plato chapters, of decorating his own openings and closings with colourful and familiar ideas from the author in question. 141 Letter to D.S.H., 16 May 2001. 142 During, Attempt, 37. •

Created on 4 July 2005 at 14.34 hours page 286

286

D. S. Hutchinson @ Monte Ransome Johnson

During’s reasons in favour of his scrambled reordering, but he • provides none.143 There is no need to follow During in this idiosyncratic reorder• ing, but perhaps we should consider the widely held view that the ending of chapter 8 closed Aristotle’s work. ‘Only one thing is pretty certain, namely that the last sentence of the Protrepticus is preserved in B 110 [ = 78. 19–79. 2 Des Places]’.144 Again, there is no argument for this point, probably because previous generations of scholars, beginning with Hirzel, had become convinced of it, on the basis of a (misunderstood) report from Augustine about how Cicero’s dialogue ended,145 together with a (frail) inference that the note on which Cicero’s dialogue ended would have been the same as that on which Aristotle’s dialogue ended (because Cicero modelled his work as a protreptic).146 During may have simply felt that the point • 143 If we make the assumption that he worked in the simplest way, then the result could have been achieved with a relatively short series of steps: move chapter 8 to the end, move three large chunks of text from elsewhere into chapter 6, and import three single sentences. The rationale for moving chapter 8 will be discussed below; the other transpositions seem to have been to move the first part of chapter 9 (to 82. 20) and the last part of chapter 5 (from 65. 1) to after the first block in chapter 6 (67. 14), and to move the remaining bit of chapter 9, together with chapter 10, to near the end of chapter 6 (70. 9). The final changes seem to have been even weirder: prefacing the middle chunk of chapter 6 with a single sentence taken from near the beginning of Iamblichus, DCM (B 52), prefacing the chunk 9b+10 with a single sentence removed from the beginning of chapter 7, and replacing that removed sentence with one of his own, an entirely invented sentence (B 58). The result is his new arrangement (‘6a’ means ‘first chunk of chapter 6’): 6a, 9a, 5b, 6b, B 41 prefacing 9b+10, B 52 prefacing 6c, B 58 prefacing 7, 11+12, 8. 144 During, Attempt, 37. • 145 ‘Commending this contemplative wisdom . . . Cicero says at the end of the dialogue Hortensius: “To us . . . who spend our lives in philosophy this is a great hope—that if that by which we feel and think is mortal and perishable, we shall have a happy setting . . . and a rest from life; if, on the other hand, as the ancient, the greatest, and far the most famous, philosophers thought, we have minds eternal and divine, then we should reflect that the more these minds have been constant in their courses—in the use of reason and in the desire of discovery—and the less they have mixed and implicated themselves in the vices and errors of mankind, the easier will be their ascent and return to heaven.” Then, adding this very clause and summing up his argument, he says: “Wherefore—to bring my speech at last to an end—if we wish either to be quietly extinguished when we have lived our life with these skills, or to move without delay from this to a far better home, all our interest and concern must be bestowed on these studies” ’ (Trin. 14. 19. 26, trans. after Ross fr. 10c). 146 On this inference, the extreme scepticism of Rabinowitz, Sources, 23–7, is warranted; yet this ‘communis opinio’ (Flashar, ‘Platon und Aristoteles’, 71 n. 56) is still current: ‘If, then, the Hortensius was “modelled on a protreptic pattern” (ad exemplum protreptici) and if Aristotle’s Protrepticus was one of its main examples, a passage from the Protrepticus which we know was imitated at the end of the

Created on 4 July 2005 at 14.34 hours page 287

Authenticating Aristotle’s Protrepticus

287

had somehow been established, so that the last sentence of Cicero’s dialogue struck him, oddly, as ‘modelled on the last sentence of the Protrepticus, but entirely di·erent in purport’.147 Indeed, they do make di·erent points: Aristotle’s conclusion is that if we had to exist without cognitive awareness, we might as well not bother to live; Cicero’s conclusion is that, whether death is extinction or departure, we must seriously apply ourselves to the study of philosophy, a conclusion that could just as easily have come from Plato as from Aristotle. Slings tries to make the argument explicit as to why the ‘large majority of scholars’ hold it to be the last paragraph of the work, but adds no new considerations and carries no fresh conviction.148 One of the main reasons why the ending of chapter 8 was taken to be the ending of the work needs to be addressed head on. Slings repeats what was said at the outset by Hirzel,149 and never challenged, that since the rhetorical climax at the end of the chapter can hardly be more pointed, it must have been a peroration, and there is nothing for the dialogue to do at this point but finish. Yes, this is possible, but it is by no means necessary that a rhetorical climax should come at the end of such a work. To think that this is a probative inference is to fly in the face of evident facts about Plato’s rhetorical climaxes, which often come in the middle of his works. For example, the central speech of his Theaetetus150 contains a tremendous rhetorical climax (Iamblichus quotes the digression, including its climax, in chapter 14, as we saw above), but it would be a mistake to conclude from Iamblichus that this speech closed Plato’s work. Likewise, the great speech in Phaedrus is crucial to the meaning of the dialogue, climactic, and central.151 Hortensius may be plausibly located near the end of the Protrepticus itself’ (Slings, Clitophon, 338). 147 During, Attempt, 267. • • 148 Slings, Clitophon, 336–8, the ‘large majority’ comprising R. Hirzel, ‘Uber den • Protreptikos des Aristoteles’ [‘Uber den Protreptikos’], Hermes, 10 (1876), 61–100 at 94–5; Rose, Fragmenta, fr. 61; Jaeger, Entwicklung, 101 n. 1 (Eng. trans.); Bignone, Perduto, i. 97–8; During, Attempt, 37; id., ‘Problems’, 169; Schneeweiss, ‘Der Pro• treptikos’, 228; S. Mansion, ‘Contemplation and Action in Aristotle’s Protrepticus’, in During and Owen, Mid-Fourth Century, 56–75 at 67 n. 1. • • 149 Hirzel, ‘Uber den Protreptikos’, 94–5. 150 David Sedley (Midwife, 80) argues that the digression is central to the Theaetetus both for its interpretation and in its geometrical, or ‘logometrical’, position. 151 The point that the midpoints of Platonic works are of particularly high significance can be extended quite generally, to include all of his long works (Meno

Created on 4 July 2005 at 14.34 hours page 288

288

D. S. Hutchinson @ Monte Ransome Johnson

We must surely put a stop now to all this talk of reordering the work and shifting its parts around: the technique of Iamblichus in Plato was a common-sense technique, to work in a natural sequence; it makes sense to assume that he used this common sense with Aristotle too; and no argument for reordering is even remotely probative. The ordering is provided for us by Iamblichus, and it provides a solid skeleton for us to work with in conducting a future reconstruction.152 Much has been achieved, but we would jeopardize our achievements by exaggerating them, so we need to remind readers of what is still unknown or unproven. We do not know how much of the Protrepticus is still missing. We do not know its overall rhetorical or literary structure. We do not know whether it had characters and speakers and dialogue, or not. But we do have reasonable confidence in a number of conclusions: that Iamblichus was always faithful to the thought of Aristotle, and usually to his language, that he was faithful to the order of the text, and that his quoting pure blocks in natural sequence defines for us in an objective way the limits of the passage that he is quoting. It is at this point that our reasoning is most open to error: it is quite possible that some of our decisions about where to separate the voices are wrong, and it is quite improbable that we have made every separation in the right place. Future scholarship will probably clarify these decisions and solidify most of them; but we cannot now claim to have made every right choice. and longer, with the exception of Laws and perhaps Philebus): each of these contains crucial dramatic episodes in their exact centres. To the cases of Theaetetus and Phaedrus, where the crucial dramatic episode at the centre is a revelation of Platonic doctrine in an extended and impressive speech, we might add the ‘central books’ of the Republic, which comprise the middle third of Republic 2–10, and the speech of Timaeus, which comprises the middle two-thirds of Timaeus/Critias (though the symmetry of this work is masked by its having been broken into Timaeus and Critias). 152 We cannot do that work of reconstruction here, but pause only to o·er two important consequences. The passage in chapter 8 that calls on us to honour the ‘god in us’ looks back to Plato’s Timaeus and forward to Aristotle’s Nicomachean Ethics, the two texts that David Sedley has convincingly argued (‘Becoming like God’, 335–9) express the same key idea in di·erent words; now we can also see that this key statement of Aristotle’s faith comes at a central point of his Protrepticus, a good place to express a key idea in an attractive way. Secondly, the skeleton that is revealed to us in Iamblichus’ Aristotle chapters dictates the exact places to situate the two columns of papyrus fragments recently convincingly but interrogatively ascribed to the Protrepticus by Glenn Most in ‘New Fragments?’

Created on 4 July 2005 at 14.34 hours page 289

Authenticating Aristotle’s Protrepticus

289

The one serious worry which we needed to keep in mind for a while can now be dispelled or lessened: we noticed that for a page in chapter 13 Iamblichus used a pastiche construction instead of a natural sequence of block quotations, drawing his material from various sources in order to make a transition from the Phaedo material to the Theaetetus material. But if he worked from only one text of Aristotle, as we can now see, he would not have needed to construct any artificial pastiches to make a transition between works. Yet it remains possible that he did so, in which case there may be a few zones in the seven Aristotle chapters which are more like paraphrase than quotation. So that is the limit of the uncertainty: either strict quotation or strictly modified quotation, but just possibly, perhaps, a paraphrase or two. But even if so, there is no reason at all to suspect that any of Aristotle’s thoughts has been distorted, not even in the paraphrase. If our conclusions are true, then the history of the recovery of the Protrepticus seems to have progressed like this. Ingram Bywater noticed that large parts of Iamblichus’ Protrepticus contained quotations from Aristotle’s work of the same name. Many of Bywater’s critics were rightly sceptical of his use of Cicero and other material to authenticate the passages he claimed to be recovering. But the critics of his theory that Aristotle’s work could be recovered from Iamblichus became excessively focused on the problems of using sources other than Iamblichus’ Protrepticus. An artefact of this excessive focus was the methodologically flawed and inconclusive criticism of Rabinowitz,153 who ventured to predict what a close study of the Aristotle chapters would reveal, having started with a handful of disputable fragments rather than an analysis of the Plato chapters. It is sobering to realize how wrong he was: A prediction might be ventured that a painstaking study of those portions of the ‘excerpt’ [viz. chapters 6–12] which have not been subjected to analysis in the preceding chapter [viz. all of it except one argument at 69.11–19] will 153 It seems that Rabinowitz’s ultimate motive was to undermine the developmental hypothesis of Jaeger, and his interest in the philosophy of Aristotle’s Protrepticus, if he had any, does not make itself known in his work. The real point of Rabinowitz’s extraordinary screed seems to be revealed in the last note on the last page of his book: ‘It is noteworthy that Jaeger’s famous theory of Aristotle’s “Entwicklung” rests squarely on the assumption that the fragments of the Protrepticus are authentic. If they are not, however, this theory would seem to be shaken -νω κ τω.’

Created on 4 July 2005 at 14.34 hours page 290

290

D. S. Hutchinson @ Monte Ransome Johnson

reveal therein a complex array of doctrines derived from many sources, and, further, that when such a study has been fully carried out, the last ground for the erroneous notion of an ‘excerpt’—our ignorance of Iamblichus’ actual sources—will have been removed. Only then will the unjustifiable ascription of passages from near and far to the lost Protrepticus be checked; and only then will the true poverty of our information concerning the work be realized. (Rabinowitz, Sources, 95)

We venture to claim that we have ‘fully carried out’ the right kind of ‘painstaking study’ of the correct object of enquiry, namely the parallel groups of chapters quoting both Plato and Aristotle. It was to save other scholars the uncertainty of knowing what a painstaking study would reveal that made us be so very explicit and painstaking. The study needed to be done, and seen to be done. During was successful (in the minute analysis of his commentary) • in showing that the diction and philosophy contained in the sections of Iamblichus’ Protrepticus that had been identified by Bywater were thoroughly Aristotelian, and he rightly realized that the Iamblichus material was the key to any successful recovery. But he failed to extricate Iamblichus’ interpolations, and he worsened things by dividing the work into fragments of arbitrary length, by reordering them according to what he thought was reasonable, and by insisting that he could know the literary form of the work (asserting that it was a letter and not a dialogue). Both of these deficiencies of his work—his reordering scheme, and his claim to know the one thing that cannot be known from the source in question—stem directly from his failure to study Iamblichus’ methods of chapter construction and excerption of Plato’s dialogues. The first person to have seen this clearly, either Flashar or his graduate student, failed to deliver the study asserted by him and acknowledged by all later scholars to be necessary to answer the important exegetical questions, and so even he was led into error. That brings us to the current situation. These blocks of Aristotelian philosophy languish in the back of large volumes and old tomes of fragments, virtually unused, even though no one has mounted any convincing arguments against considering them to be words of Aristotle. A study of the Protrepticus in which these blocks are contained reveals that it is probable that Iamblichus maintained a high degree of fidelity to his own sources. It is pretty clear, from both the form and the philosophical content, that we are dealing with a published work, the Protrepticus mentioned in many

Created on 4 July 2005 at 14.34 hours page 291

Authenticating Aristotle’s Protrepticus

291

consistent reports, and duly listed in all the ancient catalogues of Aristotle’s works. All in all, the judgement of the young Bywater (‘Lost Dialogue’, 66) has been vindicated. His first words on the topic can also serve as our last words: The Fragment incorporated by Iamblichus would seem to be substantially homogeneous and consecutive, Aristotelian in its contents and with the Aristotelian manner everywhere visible in the style; at times, too, there is a vigor, a refinement, in other words, an originality about the expression which precludes the idea that we are reading a compilation by some inferior hand. So far, then, we are justified in considering it a part of an independent work of Aristotle’s—a work which on external grounds I have endeavored to identify with the Dialogue called the Protrepticus.

University of Toronto @ St Louis University B I B L I O GR A P HY Allan, D. J., review of Rabinowitz, Sources, in Classical Review, ns 9 (1959), 124–7. ‘A Passage from Iamblichus in Praise of the Contemplative Life’, Archiv f•ur Geschichte der Philosophie, 57 (1975), 246–68. Alline, H., Histoire du texte de Platon (Paris, 1915). Augustine, Confessions, trans. H. Chadwick (Oxford, 1991). Barnes, J. (ed.), The Complete Works of Aristotle [Revised Aristotle] (2 vols.; Princeton, 1984). (trans.), Porphyry: Introduction, with a Commentary (Oxford, 2003). Bernays, J., Die Dialoge des Aristoteles in ihrem Verh•altnis an seinen ubrigen • Werken (Berlin, 1863). Berti, E., La filosofia del ‘primo’ Aristotele (Padua, 1962), 2nd edn. (Milan, 1997). (trans.), Aristotele: Protreptico. Esortazione alla filosofia (Turin, 2000). Bignone, E., L’Aristotele perduto e la formazione filosofica di Epicuro [Perduto] (2 vols.; Florence, 1936). ‘Nuove testimonianze e frammenti del “Protrettico” di Aristotele’, Rivista di filologia e d’instruzione classica, 64 (1936), 225–37. Bos, A. P., ‘Aristotle’s Eudemus and Protrepticus: Are they Really Two Di·erent Works?’, Dionysius, 8 (1984), 19–51. Brunschwig, J., ‘Aristote et les pirates tyrrh‹eniens (›a propos des fragments 60 Rose du Protreptique’ [‘Pirates’], Revue philosophique de la France et de l’‹etranger, 88 (1963), 171–90. (ed.), Aristote: Topiques, i. Livres I–IV (Paris, 1967).

Created on 4 July 2005 at 14.34 hours page 292

292

D. S. Hutchinson @ Monte Ransome Johnson

Burnet, J. (ed.), Platonis opera, vol. i (Oxford 1900). (ed.), Platonis opera, vol. iv (Oxford 1902). Bywater, I., ‘On a Lost Dialogue of Aristotle’ [‘Lost Dialogue’], Journal of Philology, 2 (1869), 55–69. Carlini, A., Studi sulla tradizione antica e medievale del Fedone [Tradizione del Fedone] (Rome, 1972). and Strachan, C., Studi dell’Academia Toscana di Scienze e Lettere, 129 (1992), 147–67. Casaglia, M. (trans.), Aristotele: Protreptico. Esortazione alla filosofia (Florence, 2001). Chroust, A. H. (trans.), Aristotle: Protrepticus. A Reconstruction (Notre Dame, 1964). Cobet, C. G., ‘Platonica’, Mnemosyne, ns 2 (1874), 241–82. Cooper, J., and Hutchinson, D. S. (ed.), Plato: Complete Works [Complete Plato] (Indianapolis and Cambridge, 1997). Des Places, E. (ed.), Jamblique: Protreptique (Paris, 1989). ‹ ‘Pr‹edicats univoques et pr‹edicats analogiques dans le “Prode Strycker, E., treptique” d’Aristote’, Revue philosophique de Louvain, 66 (1968), 597– 618. • Diels, H., ‘Uber die exoterischen Reden des Aristoteles’ [‘Exoterische Reden’], Sitzungsberichte der Berliner Akademie der Wissenschaften, 19 (1883), 477–94. Dillon, J., ‘Iamblichus of Chalcis (c. 240–325 a.d.)’ [‘Iamblichus’], Aufstieg und Niedergang der r•omischen Welt, 36.2 (1987), 862–909. Dodds, E. R. (ed.), Plato: Gorgias. A Revised Text, with Introduction and Commentary [Gorgias] (Oxford, 1959). Duke, E. A., et al. (eds.), Platonis opera, vol. i [1995 OCT] (Oxford, 1995). D•uring, I., ‘Problems in Aristotle’s Protrepticus’ [‘Problems’], Eranos, 52 (1954), 139–71. ‘Aristotle in the Protrepticus’, in Autour d’Aristote: recueil d’‹etudes de philosophie ancienne et m‹edi‹evale o·ert a› A. Mansion (Louvain, 1955), 81–97. Aristotle in the Ancient Biographical Tradition (G•oteborg, 1957). (ed.), Aristotle’s Protrepticus: An Attempt at Reconstruction [Attempt] (Studia Graeca et Latina Gothoburgensia, 12; G•oteborg, 1961). (trans.), Der Protreptikos des Aristoteles (Frankfurt a.M., 1969). and Owen, G. E. L. (eds.), Aristotle and Plato in the Mid-Fourth Century: Papers of the Symposium Aristotelicum Held at Oxford in August, 1957 [Mid-Fourth Century] (G•oteborg, 1960). Festa, N. (ed.), Iamblichus: De communi mathematica scientia [DCM] (Leipzig, 1891; suppl. U. Klein, repr. Stuttgart, 1975). Flashar, H., ‘Platon und Aristoteles im Protreptikos des Jamblichos’ [‘Platon und Aristoteles’], Archiv f•ur Geschichte der Philosophie, 47 (1965),

Created on 4 July 2005 at 14.34 hours page 293

Authenticating Aristotle’s Protrepticus

293

53–79; repr. in P. Moraux (ed.), Fr•uhschriften des Aristoteles (Darmstadt, 1975), 247–69, and in Flashar, Eidola: Ausgew•ahlte Kleine Schriften (Amsterdam, 1989), 297–323. Follon, J. (trans.), Aristote: Invitation a› la philosophie (Paris, 2000). Furley, D., review of Rabinowitz, Sources, in Journal of Hellenic Studies, 89 (1959), 178–80. Gigon, O., ‘Prolegomena to an Edition of the Eudemus’, in D•uring and Owen, Mid-Fourth Century, 19–33. (ed.), ‘Librorum deperditorum fragmenta’, in Aristotelis opera, iii (Berlin and New York, 1987), 283a41–286b39, 302b13–318b8. Heitz, E. (ed.), Aristotelis fragmenta (Paris, 1869). • • Hirzel, R., ‘Uber den Protreptikos des Aristoteles’ [‘Uber den Protreptikos’], Hermes, 10 (1876), 61–100. Hutchinson, D. S., The Virtues of Aristotle (London, 1986). Jachmann, G., ‘Der Platontext’, Nachrichten von der Akademie der Wissenschaften in G•ottingen, phil.-hist. Kl. 11 (1941), 225–389; repr. in id., Textgeschichtliche Studien (K•onigstein/Ts., 1982), 581–745. Jaeger, W., Aristoteles: Grundlegung einer Geschichte seiner Entwicklung [Entwicklung] (Berlin, 1923; English trans. R. Robinson, Oxford, 1948). Johnson, M., Aristotle on Teleology (Oxford, 2005). Krische, A. B., Die theologischen Lehren der griechischen Denker: Eine Pr•ufung der Darstellung Cicero’s (Forschungen auf dem Gebiete der alten Philosophie, 1; G•ottingen, 1840). Lloyd, G. E. R., Aristotle: The Growth and Structure of his Thought (Cambridge, 1968). Mari, M. (trans.), Anonimo di Giamblico: La pace e il benessere. Idee sull’economia, la societ›a, la morale (Milan, 2003), with introduction and commentary by M. Mari and preface by D. Musti. Mansion, S., review of Rabinowitz, Sources, in Revue philosophique de Louvain, 56 (1958), 316–20. ‘Contemplation and Action in Aristotle’s Protrepticus’, in D•uring and Owen, Mid-Fourth Century, 56–75. Menn, S., ‘The Origins of Aristotle’s Concept of νργεια: νργεια and δναµις’ [‘Origins’], Ancient Philosophy, 14 (1994), 73–114. ‘Aristotle’s Definition of Soul and the Programme of the De anima’, Oxford Studies in Ancient Philosophy, 22 (2002), 83–139. Minca, B., and Partenie, C. (trans.), Aristotel: Protrepticul (Bucharest, 2005). Moraux, P., review of Rabinowitz, Sources, in L’Antiquit‹e classique, 27 (1958), 460–4. Most, G., ‘Some New Fragments of Aristotle’s Protrepticus?’ [‘New Fragments?’], in Studi su codici e papiri filosofici: Platone, Aristotele, Ierocle

Created on 4 July 2005 at 14.34 hours page 294

294

D. S. Hutchinson @ Monte Ransome Johnson

(Studi e testi per il Corpus dei papiri filosofici greci e latini, 6; Florence, 1992), 189–216. Mutschmann, H., ‘Inhaltsangabe und Kapiteluberschrift im antiken • Buch’, Hermes, 46 (1911), 93–107. Nightingale, A. W., Spectacles of Truth in Classical Greek Philosophy: Theoria in its Natural Context (Cambridge, 2004). O’Meara, D., Pythagoras Revived: Mathematics and Philosophy in Late Antiquity (Oxford, 1989). Platonopolis (Oxford, 2003). Pistelli, H. (ed.), Iamblichi Protrepticus (Leipzig, 1888; repr. Stuttgart, 1967). Rabinowitz, W. G., Aristotle’s Protrepticus and the Sources of its Reconstruction, vol. i [Sources] (Berkeley, 1957). Reeve, C. D. C. (trans.), Plato: Republic (Indianapolis and Cambridge, 2004). Rose, V. (ed.), Aristoteles Pseudepigraphus (Leipzig, 1863). (ed.), ‘Aristotelis qui ferebantur librorum fragmenta’, in Opera omnia Aristotelis, v (Berlin, 1870), 1463–589. (ed.), Aristotelis qui ferebantur librorum fragmenta [Fragmenta] (Leipzig, 1886; repr. 1966). Ross, W. D. (trans.), Select Fragments = The Works of Aristotle, vol. xii (Oxford, 1952). (ed.), Aristotelis fragmenta selecta (Oxford, 1955). Sa·rey, H. D., and Westerink, L. G. (eds.), Proclus: Th‹eologie platonicienne, vol. i (Paris, 1968). Schneeweiss, G., ‘Der Protreptikos des Aristoteles’ [‘Der Protreptikos’] (diss. Munich, 1966). Sedley, D., ‘Is Aristotle’s Teleology Anthropocentric?’ [‘Teleology Anthropocentric?’], Phronesis, 36 (1991), 179–96. ‘ “Becoming like God” in the Timaeus and Aristotle’ [‘Becoming like God’], in T. Calvo Mart‹§nez and L. Brisson (eds.), Interpreting the Timaeus-Critias (Proceedings of the IV Symposium Platonicum; Sankt Augustin, 1997), 327–39. The Midwife of Platonism: Text and Subtext in Plato’s Theaetetus [Midwife] (Oxford, 2004). Slings, S. R. (ed.), Plato: Clitophon, Edited with Introduction, Translation, and Commentary [Clitophon] (Cambridge, 1999). (ed.), Platonis Respublica (Oxford, 2003). Spoerri, W., review of Rabinowitz, Sources, in Gnomon, 32 (1960), 18–25. Walzer, R. (ed.), Aristotelis dialogorum fragmenta (Florence, 1934). Wilpert, P., review of Rabinowitz, Sources, in Archiv f•ur Geschichte der Philosophie, 42 (1960), 101–6.

Created on 4 July 2005 at 14.34 hours page 295

ARI S TOTLE ON WHY P LANTS CANNOT P ERCEI VE DAMIAN MURPHY

 paper pertains to the debate between literalists and spiritualists concerning Aristotle’s theory of perception. I investigate why Aristotle thought that plants cannot perceive. Aristotle says that it is because they do not have a mean and cannot receive forms without matter (DA 424A32–B3). A crucial question is whether ‘to receive forms without matter’ means ‘to receive the forms of the sensible object without receiving matter from the sensible object’ (literalism) or ‘to receive the forms of the sensible object without being a·ected materially’ (spiritualism).1 As an explanation of why ã Damian Murphy 2005 I wish to thank Nick Denyer and David Sedley for many extremely helpful comments and suggestions on earlier drafts of this paper. I also wish to thank Myles Burnyeat for discussing with me many of the issues arising in the paper. 1 The terms ‘literalist’ and ‘spiritualist’ are coined by M. Burnyeat, ‘Is an Aristotelian Philosophy of Mind Still Credible? A Draft’ [‘Draft’], in M. Nussbaum and A. Rorty (eds.), Essays on Aristotle’s De Anima [Essays] (Oxford, 1995), 15–26. I take Burnyeat himself to be the most prominent spiritualist. His position is elaborated in M. Burnyeat, ‘How Much Happens When Aristotle Sees Red and Hears Middle C?’, in Nussbaum and Rorty, Essays, 422–34, M. Burnyeat, ‘Aquinas on “Spiritual Change” in Perception’, in D. Perler (ed.), Ancient and Medieval Theories of Intentionality (Leiden, 2001), 129–154, and Burnyeat, ‘De Anima II 5’. The spiritualist position is also represented by T. Johansen, Aristotle on the Sense-Organs (Cambridge, 1998), and, provided we accept Burnyeat’s interpretation of his work, Aquinas. The two advocates of the literalist position on whom I shall focus in this paper are Sorabji and Everson. See R. Sorabji, ‘Body and Soul in Aristotle’, Philosophy, 29 (1974), 63–89; id., ‘Intentionality and Physiological Processes: Aristotle’s Theory of Sense-Perception’ [‘Intentionality’], in Nussbaum and Rorty, Essays, 195–226; id., ‘Aristotle on Sensory Perception and Intentionality: A Reply to Myles Burnyeat’, in Perler, Ancient and Medieval Theories of Intentionality, 49–61; S. Everson, Aristotle on Perception [Perception] (Oxford, 1997). Other advocates include T. Slakey, ‘Aristotle on Sense Perception’, in M. Durrant (ed.), Aristotle’s De Anima in Focus (London and New York, 1993), 75–89; M. Nussbaum and H. Putnam, ‘Changing Aristotle’s Mind’, in Nussbaum and Rorty, Essays, 27–56; S. Cohen, ‘Hylomorphism and Functionalism’, in Nussbaum and Rorty, Essays, 57–74. What I call the

Created on 4 July 2005 at 14.34 hours page 296

296

Damian Murphy

plants cannot perceive, reception of forms without matter is most readily understood in the spiritualist way. I argue that attempts by literalists to defend their position by appeal to DA 3. 13 are incorrect. They take DA 3. 13 as evidence that plants cannot receive sensible forms without receiving matter, on the grounds that this passage says that plants consist of Earth,2 which is essentially cold and dry, and thus cannot receive the forms of the cold and dry because they possess these already, and cannot receive the forms of the hot and moist without ceasing to consist of Earth. I argue that this argument can prove too much and is mistaken on the grounds that the chemical elements are not characterized simply as, for example, cold-to-the-touch and dry-to-the-touch, but by more specialized notions of cold and dry. Thus, while Earth is essentially chemically cold and chemically dry, it can change with respect to qualities as perceived by touch. For similar reasons I also argue that their consisting of Earth does not help us understand why plants do not have the sort of mean which is necessary for perception. I go on to o·er an account of a spiritualist persuasion of why Aristotle thought that plants cannot perceive. I argue that plants lack a mean inasmuch as they lack a fixed degree of hotness intermediate between the extremes of hot and cold. Since perception crucially involves the registering of the di·erence between the state of the sense organ and the state of the sense object, variation in the material state of the body of a plant prevents it from acting as a sense organ. 1. Summary of the mainline debate It is generally agreed that, for Aristotle, when one perceives red, there is some sort of alteration (
λλοωσς τις, DA 416B34). The question is whether this is (or necessarily involves) an ordinary literalist position allows for many variants, aimed at demonstrating Aristotle’s allegiance to modern theories in the philosophy of mind. For example, Slakey claims that seeing red is type-identical to the eye becoming red, both Cohen and Nussbaum and Putnam claim that seeing red is to be identified in functional terms and is contingently identical with the eye becoming red, and Everson argues that seeing red supervenes upon the organ of sight becoming red. 2 I use capitalized ‘Earth’, ‘Water’, ‘Fire’, and ‘Air’ when I wish to highlight that I am talking about the pure elements. Lower-case usage is neutral between the element of that name and the more common or garden stu· of that name.

Created on 4 July 2005 at 14.34 hours page 297

Aristotle on Why Plants Cannot Perceive

297

alteration (such as those described in Physics 3. 1–3 and GC 1) wherein the sense organ gains one quality and loses another (for example, it changes from being transparent to being red), or whether it is only an extraordinary alteration, akin simply to awareness of the quality in question. I should note that I am phrasing the question in the language adopted by Burnyeat3 in order to avoid any confusion between ‘physical’ in the modern sense and the Aristotelian φυσικ ς. I use the term ‘spiritualist’ for interpreters of Aristotle who attribute to him the following position: perceptual awareness is basic and in need of no further explanation;4
λλοωσς τις refers to the transition from the sense faculty not perceiving to the sense faculty actually perceiving; there is no change on the material side. I use the term ‘literalist’ for interpreters of Aristotle who attribute to him the following position: a necessary condition of perception is that there is an ordinary alteration on the material side. Their motivation for claiming this is their interpretation of what Aristotle says at DA 403A3–B19—that, in addition to the formal explanation, there is a material explanation to be given of perception. In practice the literalist treats his claim that perception requires some material alteration as implying the claim that perception requires the sense organ literally to take on the sensible quality of the object that it perceives. For example, a necessary condition of seeing red is that the organ of sight change from not being red to being red, a necessary condition of perceiving an object as hot is that the organ of touch change from not being hot to being hot. Thus, underlying perception there is an ordinary material alteration in which one quality is lost and one quality is gained. It should be noted that I am interested simply in the debate as it relates to the interpretation of Aristotle. It must be accepted that on either interpretation, in the absence of some modification, Aristotle’s theory of perception will be deemed inadequate by modern, post-Cartesian standards. 3 M. Burnyeat, ‘De Anima II 5’, Phronesis, 47 (2002), 28–90. 4 e.g. Burnyeat, ‘Draft’, 26: ‘the flesh, bones, organs, etc. of which we are composed are essentially alive, essentially capable of awareness’ (emphasis original). Or Burnyeat, ‘How Much Happens When Aristotle Sees Red and Hears Middle C?’, 423: ‘If you ask why an animal with eyes open in clear daylight sees the coloured object in front of it, Aristotle’s reply, I maintain, is just this: such is the nature of sight and colour.’

Created on 4 July 2005 at 14.34 hours page 298

298

Damian Murphy

One major point of debate between literalists and spiritualists is the following: What is it to receive sensible forms without matter (τ δεκτικν τν ασθητν εδν νευ τς λης, DA 424A18–19)?5 The spiritualist claims that to receive form without matter is for awareness of red to take place without any corresponding alteration in the matter of the sense organ. The literalist argues rather that the reception of form without matter means that the eye becomes red (i.e. undergoes an ordinary alteration) without receiving any matter (e.g. particles of red paint) from the object of perception. On the literalist reading, Aristotle insists that perception be ‘without matter’ to distinguish his theory from those of the likes of Empedocles and Democritus, in which perception involves the reception in the sense organ of emanations from the object. It is important to be clear about the spiritualist’s position. He does not deny that there are necessary standing conditions of a material kind for perception. For example, he thinks that the transparency of the organ of sight is necessary for sight.6 Similarly, he acknowledges that there must be an unobstructed view of the object. These are material conditions, but they are not alterations at all and hence not ordinary alterations. The spiritualist can also allow that certain material changes are necessary for perception. For example, if the eyelids are closed, it is necessary that they be opened before seeing can take place. The spiritualist simply denies that it is necessary for the perceptual organ itself to undergo an ordinary alteration that determines precisely what the perception is a perception of. Opening the eyes is a necessary condition of any seeing whatsoever. For the literalist, on the other hand, the organ of sight literally becoming red is a necessary condition of seeing red. I do not intend to argue for the spiritualist’s position generally. Rather my method will be first to grant the literalist his assumptions and show that on these assumptions he cannot o·er an adequate account of why Aristotle thought that plants cannot perceive. Secondly, I show that I can o·er a good explanation of why Aristotle 5 Cf. 424B1–2, 425B23–4, 434A29–30. 6 e.g. Burnyeat, ‘How Much Happens When Aristotle Sees Red and Hears Middle C?’, 422–3, his conditions (a) and (b).

Created on 4 July 2005 at 14.34 hours page 299

Aristotle on Why Plants Cannot Perceive

299

thought that plants cannot perceive on the spiritualist’s assumptions. 2. The problem of plants κα δι τ ποτε τ φυτ οκ ασθνεται, χοντ τι µ ριον ψυχικν κα πσχοντ τι #π τν $πτν· κα γρ ψ'χεται κα θερµανεται· α(τιον γρ τ µ) χειν µεσ τητα, µηδ* τοια'την
ρχ)ν ο+αν τ ε(δη δ-χεσθαι τν ασθητν,
λλ πσχειν µετ τς λης. (DA 424A32–B3) [It is also clear] why plants do not perceive, although they have a certain psychic part and are a·ected to some extent by objects of touch, for they are both cooled and warmed. The reason is that they have no mean, and no principle such as to receive the forms of sensible objects, but are a·ected together with the matter.

The big debate here between literalists and spiritualists is as to what πσχειν µετ τς λης refers to. The literalist takes it to mean that plants are a·ected together with the matter of the sensible object. The spiritualist understands it to mean that plants are a·ected together with their own matter. We may note that the claim that plants cannot perceive is common in Aristotle (e.g. De somno 454B28); however, it is only in De anima that he attempts to explain why they cannot. Strictly speaking, I assume that here Aristotle is explaining only why plants cannot perceive tangible qualities. Since plants clearly do not possess the organs of the other senses, it is obvious that they do not see, hear, smell, or taste. However, since animals perceive tangibles through most parts on the outside of their bodies, Aristotle must rule out the possibility that (as Plato thought, Tim. 77 a–c) the bodies of plants are also able to perceive tangibles. DA 424a32–b3 arguably o·ers two reasons why plants cannot perceive. First they do not have a mean, secondly they cannot receive forms without matter. It is not immediately clear whether Aristotle is o·ering two distinct explanations, o·ering two separate facts that collectively amount to a single explanation, or essentially o·ering the same explanation in di·erent words. I focus for most of this paper on the claim that a plant cannot perceive because it cannot receive forms without matter. In Section 8 I briefly consider the claim that a plant cannot perceive because it does not have a mean. When I come to o·er my own solution (Sections 9 and 10), I treat the two claims as part of a single explanation.

Created on 4 July 2005 at 14.34 hours page 300

300

Damian Murphy

What it is for a plant to have a mean is just as much of a problem for the spiritualist as for the literalist, especially if we recall that Aristotle has just linked the perceptual mean with the phenomenon of ‘blind spots’ for the sense of touch (424A2–7). This might be thought to favour the literalist because of the following sort of consideration. Since for the literalist, actually perceiving something as, for example, hot requires an ordinary alteration in the sense organ, and the sense organ cannot undergo ordinary alteration to the state it is already in, we cannot perceive qualities in objects which are the same as those the sense organ already has. Thus on the literalist reading there is an explanation of the endoxon7 that we do not perceive the hotness8 of things equal in hotness to the sense organ, and clear content can be given to the notion of a mean. The spiritualist, however, struggles to explain why there should be ‘blind spots’. For if the organ of touch is basically capable of awareness, why should it not be capable of perceiving objects as hot as itself? As for the reception of forms without matter, Sorabji accuses Burnyeat of o·ering a circular explanation of why Aristotle thought plants cannot perceive.9 To say that plants cannot perceive because they have no principle such as to receive forms of tangible objects without being a·ected in their own matter does not really explain why plants cannot perceive, it merely restates that they cannot perceive. However, it is not clear that Sorabji does any better in explaining why plants cannot perceive. He claims that perception requires the reception of forms without the reception of matter and that plants are unable to receive forms without receiving foreign matter. But on the literalist account it is not clear why the fact that plants must receive foreign matter together with sensible forms should prevent perception. Admittedly, Aristotle thinks that perception does not involve reception of matter, but his comment may be merely a matter-of-fact one aimed at those who believed that perception required emanations from the object. That perception, as a matter of fact, does not involve reception of matter does not in itself explain why perception could not involve reception of matter. 7 Burnyeat, ‘Draft’, 21 n. 3, cites Theophr. De sens. 2 as evidence for this. 8 I use ‘hotness’ to name the qualitative range between and including the extremes of hot and cold. I use ‘moistness’ to name the qualitative range between and including the extremes of moist and dry. I deliberately wish to avoid the seemingly more convenient term ‘temperature’ because Aristotle did not have access to such a precise notion. 9 Sorabji, ‘Intentionality’, 217.

Created on 4 July 2005 at 14.34 hours page 301

Aristotle on Why Plants Cannot Perceive

301

I shall go on to o·er (Section 9) a spiritualist interpretation of the reception of forms without matter on which it is clear why a plant’s being a·ected in its own matter prevents perception. There is also a further, more fundamental, problem for the literalist. It is undeniable that we sometimes perceive a plant as cold and sometimes as hot. How else could Aristotle claim to know that they are ‘cooled’ and ‘warmed’? From this we can infer that plants are sometimes cold and sometimes hot. For perception of special sensibles is unerring (DA 428A11, 430B28). So let us posit a plant which is first cold and later hot. From the standpoint of everyday thinking, that plant has undergone ordinary alteration and has received the form of the hot. Since, again from the standpoint of everyday thinking, there is no reason to suspect that in all such alterations a plant must have received external hot matter, we can conclude that on (at least some) such occasions the plant receives forms without receiving matter. But Aristotle is here saying that the reason why plants do not perceive is that they do not receive forms without receiving matter. On the face of it this suggests that the literalist’s interpretation of ‘receiving forms without matter’ is wrong. On the other hand, the spiritualist can admit the everyday view that plants receive forms (in the literalist sense of ‘receiving forms’) without receiving foreign matter. For, on the spiritualist reading, this is not what Aristotle means when he says that plants do not receive forms without matter. Rather, what plants do not do is receive forms without undergoing alteration of their own matter. That is, they cannot undergo extraordinary alteration. At first sight the literalist is on the back foot with respect to Aristotle’s claim that plants cannot perceive because they cannot receive forms without matter. Sorabji admits that we perceive plants sometimes as hot and sometimes as cold, but is forced to claim that Aristotle thought that plants are a·ected only by taking in foreign matter, just as they do in fact take in matter as nutriment from the soil.10 Burnyeat claims that there is no evidence that Aristotle thought this,11 against which Everson and Sorabji cite DA 435A19–B1. Everson’s argument is the most sophisticated that has come to my attention to be o·ered by a literalist on why plants cannot perceive. He denies that plants receive forms at all and hence denies that they 10 Sorabji, ‘Body and Soul in Aristotle’, 74 n. 28; ‘Intentionality’, 217; cf. Everson, Perception, 89. 11 Burnyeat, ‘Draft’, 23–4.

Created on 4 July 2005 at 14.34 hours page 302

302

Damian Murphy

receive forms without receiving matter. However, I wish to argue that appeal to DA 435A19 ·. does not help the literalist. It is to this passage, and Everson’s argument, that I now turn.12 3. The appeal to 435A19–B1 /στε τν µ*ν τοιο'των στοιχεων οθ*ν 0ν ε(η σµα το1 ζ34ου. οδ* δ) γ5ϊνον. πντων γρ 7 $φ) τν $πτν 8στν /σπερ µεσ της, κα δεκτικν τ ασθητ5ριον ο µ νον 9σαι διαφορα γς εσν,
λλ κα θερµο1 κα ψυχρο1 κα τν λλων $πτν $πντων. κα δι το1το το:ς ;στο:ς κα τα:ς θριξ κα το:ς τοιο'τοις µοροις οκ ασθαν µεθα, 9τι γς 8στιν, κα τ φυτ δι το1το οδεµαν χει α(σθησιν, 9τι γς 8στιν. (DA 435A19–B1) And so none of these elements could be the body of an animal. Nor could it be of Earth. For touch is a sort of mean of all tangible qualities, and the sense organ is able to receive not only the distinguishing characteristics of Earth, but hot and cold and all the other tangible qualities. And on account of this we do not perceive with bones13 and hair and such parts, because these are of Earth, and for this reason plants have no perception, because they are of Earth.

The kernel of Everson’s argument14 is that Earth, according to Aristotle’s theory of chemistry as expounded in De generatione et corruptione, is essentially cold and dry, and so cannot vary in hotness or moistness and still be Earth.15 Since plants are of Earth (DA 435A19–B1), they cannot undergo ordinary alteration of hotness or 12 Everson, Perception, 86–9. Sorabji, ‘Intentionality’, 216, also makes appeal to DA 435A20–3 in trying to explain why plants cannot perceive. I should note that the appeal of the literalist, and in particular of Everson, to 435A19–B1 has received severe criticism at the hands of J. Magee, ‘Sense Organs and the Activity of Sensation in Aristotle’, Phronesis, 45 (2000), 306–30. His criticisms of the reconciliation of the fact that we perceive plants sometimes as hot and sometimes as cold with the claim that they do not undergo ordinary alteration tell decisively against Everson’s position. However, he does not make the criticisms which I consider fundamental, namely those of (1) in argument A, subsection 3.1 below. 13 Burnyeat claims that bones are capable of awareness: see n. 4 above. I presume that, in the light of DA 435A19–B1, Burnyeat can abandon this claim without detriment to his overall position. 14 Everson, Perception, 86–9. 15 Everson, Perception, 88, says: ‘Something consisting simply of earth cannot be, or have, an organ of touch since it cannot change in the relevant ways. In his discussion of the characteristic qualities of the four elements in de Generatione et Corruptione, Aristotle has told us that . . . earth is cold and dry (II. 3, 330B3– 5). These are necessary qualities of [earth]: an element cannot lose its distinctive qualities without ceasing to be that element—it is by qualitative change that the elements are converted into each other. Something which consists of earth cannot be anything other than cold and dry.’ I take it that by ‘earth’ Everson means what I express by the capitalized ‘Earth’.

Created on 4 July 2005 at 14.34 hours page 303

Aristotle on Why Plants Cannot Perceive

303

moistness. Everson takes ‘having a mean’ as more or less the same thing as being able to receive forms without matter, inasmuch as an organ of touch is equally disposed to receive either of, for example, hot or cold. That Aristotle accepts the same theory of chemistry in both De anima and De generatione et corruptione is clear from the deliberate allusion back to the latter at DA 423B27–31. Upon this point the literalist and I are in agreement. However, we di·er in our interpretation of Aristotle’s chemistry. I shall discuss a di¶culty which DA 423B27–31 provides for my interpretation in subsection 7.3. Now I outline the literalist’s first argument. 3.1. Argument A This is an argument that plants cannot perceive tangible qualities because they cannot undergo ordinary alteration with respect to tangible qualities,16 which is a necessary condition of such perception. While the substance of the argument is Everson’s, the particular formulation below is my own. (1) Anything which undergoes ordinary alteration of hotness or moistness cannot, at all times, consist purely of Earth. (2) A plant, at all times, consists purely of Earth. Therefore, (3) A plant cannot undergo ordinary alteration of hotness or moistness. (4) Reception of the forms of tangibles requires ordinary alteration of hotness or moistness. Therefore, (5) A plant cannot receive the forms of tangibles. 16 For convenience I use ‘tangible’, when strictly speaking the argument concerns only tangible qualities in the dimensions of hotness and moistness. So, prima facie, even if successful, Everson’s argument leaves open the possibility that plants can perceive e.g. hardness or roughness. This could be seen as an objection to Everson’s argument since Aristotle wishes to eliminate the possibility of plants perceiving by touch generally. However I do not wish to pursue this line of criticism. First, although Aristotle’s claim is stated quite generally, when elaborating it, his examples always concern hotness and moistness. Secondly, Aristotle thinks that other tangible qualities are derived from hot, cold, moist, and dry (GC 2. 2). Perhaps the inability to perceive hotness and moistness would prevent perception of hardness and roughness.

Created on 4 July 2005 at 14.34 hours page 304

304

Damian Murphy

(6) Perceiving tangibles requires receiving the forms of tangibles (without receiving matter). Therefore, (7) A plant cannot perceive tangibles. 3.2. The reconciliation of argument A with everyday thinking Everson does not argue that plants fail to receive forms without matter because they receive only forms with foreign matter (as we can see, his argument is that they fail to receive forms without matter because they fail to receive forms), but he does still claim that when we perceive plants becoming hot it is because they take in foreign matter. To explain the obvious fact that a plant does get hotter in the sun, Everson argues that this happens because what we commonly describe as a plant becoming hot is really an influx of hot foreign matter (e.g. hot air), while the plant itself remains as it always was. What a·ects the organ of touch of the person who touches the plant is not the plant itself but the foreign matter that the plant has acquired. We may compare the fact that we should not claim that I have altered just because some matter temporarily in my stomach, such as garlic, causes a certain smell to be present wherever I am. Although (4) and (6) will not be accepted by the spiritualist, I shall grant the literalist his characteristic assumptions for the sake of argument, and instead attack (1) and (2). 4. Against (2) The meaning of 435A19–B1 cannot be that plants, hair, and bones consist entirely of Earth, but rather that they are predominantly of Earth. At Meteor. 389A11–16 Aristotle17 lists various bodies which consist of Water and Earth but predominantly of Earth. This list includes hair, bones, and certain types of plant and plant parts. At

17 I assume that Meteorologica 4, even if misplaced with the first three books of Meteorologica, is by Aristotle. Alexander of Aphrodisias makes this claim in his commentary on Meteorologica 4 (179. 1–6 Hayduck). E. Lewis, Alexander of Aphrodisias: On Aristotle Meteorologica 4 (London, 1996), 1–15, summarizes the arguments for taking Meteorologica 4 as genuine.

Created on 4 July 2005 at 14.34 hours page 305

Aristotle on Why Plants Cannot Perceive

305

Meteor. 384B30–118 he says that the homoeomerous19 bodies consist of both Water and Earth, both in plants and in animals. It is important to see the significance to the likes of Everson of the claim that plants consist purely of Earth at all times. For otherwise we might have said that even though plants consist purely of Earth for most of the time, when a hot object is in contact with the plant there is a temporary change away from the normal state of the plant. This, we might have argued, is consistent with maintaining that the plant is of Earth tout court. However, in view of the evidence of Meteorologica 4 this option is unattractive. 4.1. Argument A* Given this objection to (2), we might try to revamp the argument, to reach the intermediate conclusion (3), as follows: (1*) Anything which undergoes ordinary alteration of hotness or moistness cannot, at all times, consist of 90% Earth and unvarying and precise proportions of the other elements. (2*) A plant, at all times, consists of 90% Earth and unvarying and precise proportions of the other elements. Therefore, (3) A plant cannot undergo ordinary alteration of hotness or moistness. Argument A* now follows argument A, using (4) and (6) as premisses to reach (7). The selection of 90% is arbitrary. The point is that, if the literalist is to argue in this fashion, he must select some precise proportion of Earth and some precise proportions of the other elements. For if the proportions are merely to remain within some range, there is nothing to prevent the plant from undergoing an ordinary alteration of hotness or moistness. I must make it clear that I do not wish to ascribe argument A* to Everson himself. However, given that Aristotle does not hold premiss (2), I construct this revised argument to investigate the idea that a plant’s chemical constitution is such as to prevent it from receiving tangible forms. 18 We may also note that at De longitudine et brevitate vitae 467A6–9 Aristotle says that plants live longer because they are Earthy and have less Water. Taken in the most natural way, this implies that plants have some Water. 19 By ‘homoeomerous’ Aristotle means something of which any part, however small, consists of the same stu· as the whole. This conception is most distinctly opposed to the atomistic conception of matter.

Created on 4 July 2005 at 14.34 hours page 306

306

Damian Murphy 5. Against (1*)

The most fundamental criticisms of argument A* I shall bring against (1*). The first is that the thinking behind (1*) would seem to rule out the possibility of ordinary alteration with respect to hotness and moistness even for the organ of touch of an animal. Secondly, I shall argue, there is an equivocation in the senses of ‘hot’, ‘cold’, ‘moist’, and ‘dry’ employed in (1*) and (4) and that (1*) is false for the senses required to make (4) true on the literalist reading. This second criticism applies equally to argument A with respect to premisses (1) and (4). In order to make these criticisms it is necessary to enter into some fairly lengthy discussions of Aristotelian chemistry. It is to this that I now turn. 5.1. Aristotelian chemistry: contraries and elements At GC 329A27–35 Aristotle claims that the contraries hot and cold and moist and dry are more basic to explanation than the chemical elements.20 Earth is cold and dry, Water is cold and moist, Fire is hot and dry, Air is hot and moist (GC 330B1–5). For the moment I leave open what exactly Aristotle means by ‘hot’, ‘cold’, ‘moist’, and ‘dry’. I take up my disagreement with Everson on this matter in Section 7. For Aristotle there are two types of chemical reaction. First there is elemental transformation: one element becomes another element by changing with respect to one or more of its contraries (GC 2. 4). Secondly there is what I shall term ‘Aristotelian-mixing’ (GC 2. 7). The process of Aristotelian-mixing is such that when two or more homoeomerous stu·s are thoroughly mixed the resultant stu· is itself homoeomerous (rather than consisting of juxtaposed particles of the initial ingredients). Furthermore, the process of Aristotelianmixing is such that the capacity or power of the Aristotelian-mixt (I shall coin this term for what results from Aristotelian-mixing) is proportional to the capacities or powers of the ingredients.21 For 20 Despite Aristotle’s terminological qualms, I shall use ‘element’ to denote a simple and basic chemical stu·, i.e. Earth, Water, Fire, and Air. 21 We should note that Aristotelian-mixing includes examples both of chemical compounding and of what modern chemists would call ‘mere blending’. This will count as Aristotelian-mixing provided that the resultant stu· is such that any spatial part, however small, is qualitatively identical to any other spatial part. For example, for Aristotle, salt fully dissolved in water would count as homoeomerous.

Created on 4 July 2005 at 14.34 hours page 307

Aristotle on Why Plants Cannot Perceive

307

example, if a hot element and a cold element are mixed in a ratio of 2 : 1, then (if the conditions are suitable) the result is a stu· which is neither completely hot nor completely cold but more hot than cold in capacity in proportion to the ratio in which the components were mixed (GC 334B8–16). For clarity, let us introduce two notations for classifying chemical stu·s in Aristotle. The first is in terms of proportions of each of the four chemical elements, the second in terms of the underlying degrees of hotness and moistness of the stu·.22 Elemental formulae While GC 2. 1 gives the impression that contraries are prior in explanation to the chemical elements, elsewhere (e.g. Meteorologica 4) Aristotle classifies homoeomerous stu·s in terms of the proportions of the chemical elements.23 With this practice in view, I now introduce elemental formulae. 22 The idea behind, and many of the specifics of, this mathematical approach to Aristotle’s chemistry relies on the work of P. Needham, ‘Aristotelian Chemistry: A Prelude to Duhemian Metaphysics’ [‘Aristotelian Chemistry’], Studies in History and Philosophy of Science, Part A, vol. 27.2 (1996), 251–69, and, in particular, K. Fine, ‘The Problem of Mixture’, in F. Lewis and R. Bolton (eds.), Form, Matter, and Mixture in Aristotle (Oxford, 1996), 82–182, in which many of these formulae, and their interrelations, are explicated more fully. 23 There is in general a metaphysical di¶culty in attributing to Aristotle the view that a homoeomerous stu· consists of Earth, Air, Fire, and Water. For he insists that the elements are present in the homoeomerous stu· only δυνµει. Because of the very nature of a homoeomerous stu·, Aristotle cannot o·er the sort of explanation a modern chemist would o·er of the fact that water consists of hydrogen and oxygen. While a resolution of this problem is beyond the scope of this paper, I note that Aristotle (especially in Meteorologica 4) does talk about homoeomerous stu·s consisting of di·ering proportions of Earth and Water. I take this to provide justification for the use of elemental formulae. (What I have translated as ‘consists of’ is typically just 8κ followed by the genitive, or often just the genitive. When the 8κ is present I take this to be the first sense of 8κ which Aristotle has in mind when he discusses the various senses of 8κ at Metaph. 1023A26–B11. For one of his examples there is that all things which can be melted are 8ξ δατος.) Roughly speaking I propose a deflationary reading of, for example, the claim that gold consists of 45% Earth, 53% Water, 1% Fire, and 1% Air. I take this to be true i·, in principle, gold could be made through a series of mixings in which the total amounts of the elements used stand in these proportions. Alternatively, the claim is true i·, in principle, in a series of extractions from gold, the total amounts of elements extracted were of such proportions. It is also the case that the net capacity of the Aristotelian-mixt must stand in relation to the proportion of the elements. So, for example, if gold had the above formula, it should be very cold and moderately moist. Deflationary readings are advocated by J. Bogen, ‘Fire in the Belly: Aristotelian Elements, Organisms and Chemical Compounds’, in Lewis and Bolton, Form, Matter, and Mixture in Aristotle, 183–216, and Needham, ‘Aristotelian Chemistry’. However, Fine, ‘The Problem of Mixture’, o·ers a more realist account, on which the elements can be said to exist in the Aristotelian-mixt.

Created on 4 July 2005 at 14.34 hours page 308

308

Damian Murphy

Every chemical stu· S is mapped onto an ordered quadruple <e, w, f, a>, where e is the fraction of Earth in S, w the fraction of Water in S, f the fraction of Fire in S, and a the fraction of Air in S. (0 ≤ e ≤ 1, 0 ≤ w ≤ 1, 0 ≤ f ≤ 1, 0 ≤ a ≤ 1, e+ w+ f+ a = 1.) The method for determining the elemental formulae of chemical stu·s is as follows. base clauses The elemental formulae of the pure elements are: Earth Water Fire Air

<1, 0, 0, 0> <0, 1, 0, 0> <0, 0, 1, 0> <0, 0, 0, 1>.

general clause When a series24 of Aristotelian-mixings takes place such that: (i) at the first stage of mixing, S , whose elemental formula is 1 <e , w , f , a >, is mixed with S , whose elemental formula is 1 1 1 1 2 <e , w , f , a >, in some proportion, and, 2 2 2 2 (ii) at each subsequent stage of mixing, the stu· that results from the previous stage of mixing is mixed with either S or 1 S in some proportion, and 2 (iii) the total amount used of S is p and the total amount used 1 1 of S is p , 2 2 then the stu· S that results from the final stage of mixing will have 3 the elemental formula <e , w , f , a >, where: 3 3 3 3 e = [p ÿ e + p ÿ e ]/(p + p ) 1 1 2 2 1 2 3 w = [p ÿ w + p ÿ w ]/(p + p ) 3 1 1 2 2 1 2 f = [p ÿ f + p ÿ f ]/(p + p ) 3 1 1 2 2 1 2 a = [p ÿ a + p ÿ a ]/(p + p ) 3 2 1 2 2 1 2 Qualitative formulae Given that Aristotle has said that the contraries are more basic 24 The complication of a series of mixings is necessary to make sense of something consisting of a very small, but non-zero, amount of some element. For in certain extreme proportions Aristotelian-mixing does not take place. Instead the lesser ingredient is overwhelmed by the dominant ingredient and transformed into it. This is most amply demonstrated by Aristotle’s startling claim that when one drop of wine is added to 10,000 gallons of Water, Aristotelian-mixing does not take place but rather the wine is transformed into Water (GC 328A24–31).

Created on 4 July 2005 at 14.34 hours page 309

Aristotle on Why Plants Cannot Perceive

309

than the elements, we also need to develop a notation for classifying chemical stu·s in terms of the contraries. Each chemical stu· S is mapped onto an ordered pair , 0 ≤ r(h) ≤ 1, 0 ≤ r(m) ≤ 1, where h is the degree of hotness of S and m is the degree of moistness of S, and r(h) is the real number assigned to degree of hotness h and r(m) is the real number assigned to degree of moistness m. The method for determining the qualitative formulae of chemical stu·s is as follows. base clauses r(extreme cold) = 0 r(extreme hot) = 1 r(extreme dry) = 0 r(extreme moist) = 1 general clause When a series of Aristotelian-mixings takes place such that: (i) at the first stage of mixing a stu· whose degree of hotness is h is mixed in some proportion with a stu· whose degree of 1 hotness is h , and 2 (ii) at each subsequent stage of mixing, the stu· that results from the previous stage of mixing is mixed in some proportion with either a stu· whose degree of hotness is h or a stu· 1 whose degree of hotness is h , and 2 (iii) the total amount used of stu· whose degree of hotness is h is p , and the total amount used of stu· whose degree of 1 1 hotness is h is p , 2 2 then the stu· that results from the final stage of mixing will have degree of hotness h , where: 3 r(h ) = [p ÿ r(h )+ p ÿ r(h )]/(p + p ). 1 1 2 2 1 2 3 r(m ) is calculated in a similar fashion mutatis mutandis. The general 3 clause is derived from Aristotle’s claim about Aristotelian-mixing at 334B8–16. The elements are characterized by extreme qualities We should be clear that the elements possess their qualities in the extreme. Water is not merely cold-ish or moist to an extent, it is

Created on 4 July 2005 at 14.34 hours page 310

310

Damian Murphy

cold in the extreme and moist in the extreme. This is also the case mutatis mutandis for the other elements. Such is the most plausible reading of Aristotle’s overall discussion of elements and contraries in GC 2. If this were not the case, there would be a causal difficulty. If, let us say, Fire were only moderately hot, what would there be to explain thermal qualities hotter than those of Fire? If we had to posit an additional stu·, which was extremely hot, then we might legitimately wonder why Fire was classed as an element at all. Aristotle implies this in the argument at Metaph. 993B24–6: =καστον δ* µλιστα ατ τν λλων καθ> ? κα το:ς λλοις #πρχει τ συν4νυµον (ο@ον τ π1ρ θερµ τατον· κα γρ το:ς λλοις τ α(τιον το1το τς θερµ τητος). And a thing has a quality in higher degree than other things if in virtue of it the similar quality belongs to other things (for example, fire is the hottest of things: for it is the cause of the heat of all other things).

The relationship between elemental formulae and qualitative formulae Given these two notations, we must enquire how they are related. In view of what we have said about Aristotle’s chemistry, we need the following two theses to be true. While I take them both to hold, I shall argue that they hold only for the appropriate senses of ‘hot’, ‘cold’, ‘moist’, and ‘dry’. weak thesis of determination of qualitative formulae by elemental formulae The weak thesis (as I shall abbreviate it) is as follows. If S is a sample of a pure element, then: If the elemental formula of S (Earth) is <1, 0, 0, 0>, r(h) = 0 and r(m) = 0. If the elemental formula of S (Water) is <0, 1, 0, 0>, r(h) = 0 and r(m) = 1. If the elemental formula of S (Fire) is <0, 0, 1, 0>, r(h) = 1 and r(m) = 0. If the elemental formula of S (Air) is <0, 0, 0, 1>, r(h) = 1 and r(m) = 1.

Created on 4 July 2005 at 14.34 hours page 311

Aristotle on Why Plants Cannot Perceive

311

strong thesis of determination of qualitative formulae by elemental formulae25 The strong thesis (as I shall abbreviate it) claims that the qualitative formula of any homoeomerous stu· is determined by its elemental formula. If S is a sample of a homoeomerous stu· whose elemental formula is <e, w, f, a> and h is the degree of hotness of S and m is the degree of moistness of S, then the qualitative formula of S will be , where: r(h) = f+ a. r(m) = w+ a.

6. First criticism of (1*) My first criticism of (1*) is as follows. (1), as used in argument A, requires only the weak thesis to hold. Unfortunately for the literalist, argument A must be abandoned because (2) is false for Aristotle. When we consider argument A*, (1) is replaced by (1*), and (1*) does need the strong thesis. But if the strong thesis is to be the basis of (1*), we can use the argument to prove too much. For we can prove that in all cases there is no ordinary alteration with respect to tangible qualities without change in the proportion of elemental constituents. For the sake of argument, let us temporarily assume that ‘hot’, ‘cold’, ‘moist’, and ‘dry’ are used univocally in premisses (1*) and (4) to refer to the ordinary, to-the-touch qualities that we perceive, and that the strong thesis holds for these qualities. The argument for (1*) will be as follows: (8) At any time, anything which consists of 90% Earth and precise proportions of the other elements will have some precise degree of hotness and some precise degree of moistness. 25 We may note that there is, in general, no unique way to determine the elemental formula from the qualitative formula. For example, a stu· whose qualitative formula is <0.5, 0.5> could have the elemental formula <0, 0.5, 0.5, 0> or the elemental formula <0.5, 0, 0, 0.5>. In the case of pure elements, and only in the case of pure elements, we can uniquely determine the elemental formula from the qualitative formula. One way out of this di¶culty, proposed by Needham, ‘Aristotelian Chemistry’, is to appeal to the characterization of elements in De caelo as light and heavy. For some suggestions as to how we might generally make sense of an Aristotelian-mixt consisting of elements see n. 23.

Created on 4 July 2005 at 14.34 hours page 312

312

Damian Murphy

It is important to appreciate that the precise degree of hotness and precise degree of moistness are determined from the precise proportions of Earth and the other elements in accordance with the strong thesis. (9) Anything which undergoes ordinary alteration of hotness or moistness cannot, at all times, have an unvarying degree of hotness and an unvarying degree of moistness. By the notion of ordinary alteration, if there is ordinary alteration of hotness there must be a change in the degree of hotness, and, if there is ordinary alteration of moistness, there must be a change in the degree of moistness, so the degrees of hotness or moistness cannot be the same at all times. From (8) and (9) we can deduce (1*): (1*) Anything which undergoes ordinary alteration of hotness or moistness cannot, at all times, consist of 90% Earth and unvarying and precise proportions of the other elements. The crucial premiss is (8), which follows from the strong thesis. Given that our choice of 90% Earth was to some extent arbitrary, and we did not even specify the proportions of the other elements, but merely required that there be precise proportions, the same argument should work for chemical stu·s generally. In particular, it should be applicable to the chemical stu· of which the organ of touch in an animal consists. Let us say, for the sake of argument, that its elemental formula is <0.25, 0.25, 0.25, 0.25> at all times. We can re-create the above argument to show that such an organ cannot undergo ordinary alteration. Thus the literalist who embraces argument A* is forced to claim, not only that plants do not undergo ordinary alteration of hotness or moistness, but that nothing does. Hence on the literalist reading there would be no perception of tangibles whatsoever. It is only the organ of touch that is directly vulnerable to this line of attack. When the ordinary alteration of an organ is with respect to hotness or moistness, which are defining characteristics of elements, we must say also that there is a change in the elemental constitution of the organ. On the other hand, if the ordinary alteration is with respect to colour or smell or anything which is not a defining characteristic of elements, then it is prima facie possible

Created on 4 July 2005 at 14.34 hours page 313

Aristotle on Why Plants Cannot Perceive

313

for the elemental constitution to remain the same, but for the form (e.g. the colour) to change. One might retort, no doubt correctly, that the organ of touch need not, for Aristotle, have a precise proportion of the elements, but that a range is permitted and variation can take place within that range. The same form (i.e. the functional form of the sense organ) can be instantiated with small variations in the matter. But in this case, why can we not allow small variations in the hotness and moistness of plant tissue, while maintaining that it is still the same stu·? And if we can allow this, then we have failed to explain why plants cannot perceive. It is open to the advocate of argument A, who relies only on the weak thesis, to deny the strong thesis and thus block the claim that a parallel argument rules out all ordinary alteration without change in elemental formulae. However, the advocate of argument A* cannot do this, for he needs the strong thesis himself.

7. The second criticism of (1*) Regardless of whether we employ (1) as derived from the weak thesis or (1*) as derived from the strong thesis, A-type arguments are unsound. For the senses of ‘hot’, ‘cold’, ‘moist’, and ‘dry’ (hotto-the-touch etc.) required to make (4) true do not support either thesis.26 First I shall say a little about what I mean by, for example, ‘hot-tothe-touch’. Given that, for Aristotle, sense perception is unerring and there are perceptible qualities in objects which are themselves causally e¶cacious,27 when we perceive an object as hot, it is because that object is hot, just as when we perceive an object as red, it is because that object is red. This, however, does not enable us to say straightforwardly that the object is hot in every sense. For Aristotle outlines a variety of senses of ‘hotter’ at PA 648B11–649B7. He also notes that just because x is hotter to the touch than y, it does not follow that y cannot be hotter than x in some other sense. 26 Everson, Perception, 88, and Sorabji, ‘Intentionality’, 214–16, assume that the weak thesis, at least, does hold for the to-the-touch senses. 27 While these claims have aroused controversy in themselves, it is open equally to the spiritualist and literalist to accept them, e.g. Everson, Perception, ch. 3.3, and Burnyeat, ‘De Anima II 5’, 45, 58.

Created on 4 July 2005 at 14.34 hours page 314

314

Damian Murphy

τι θερµ τερον µ*ν κατ τ)ν $φ)ν τ ζ-ον δωρ, ψ'χεται δ* θAττον κα π5γνυται το1 8λαου. (PA 648B30–2) Again boiling water is hotter to the touch than olive oil, but becomes cold and solidifies more quickly than olive oil.

The sense in which an object is hot because it is perceived as hot, I designate ‘hot-to-the-touch’. I shall use ‘hot-to-the-touch’, ‘coldto-the-touch’, etc. to refer to objective states of the object. This might initially sound a little odd to us since we think of temperature as an objective state of the object and hot-to-the-touch as a subjective quality which only very roughly corresponds to temperature. (For example, a stone floor at 10 C is perceived as colder-to-thetouch than a carpeted floor at 10 C.) However, this is not Aristotle’s way of looking at things. For he does not have a refined notion such as temperature and he takes perception to be unerring. Thus what I label ‘hotness-to-the-touch’ is, for Aristotle, an objective state of the object. It is almost too obvious to mention that the sense of ‘hot’, ‘cold’ ‘moist’, and ‘dry’ required for (4) is the to-the-touch sense. For on the literalist reading, the form of the object is the very same form as that which the sense organ literally acquires. However, when we turn to premisses (1) and (1*), it is not the case that Earth is necessarily extremely cold-to-the-touch or extremely dry-to-thetouch. For, I shall argue, there are some qualities called ‘cold’ and ‘dry’, having the extremes of which is a necessary and su¶cient condition for being Earth. But these qualities are not the ordinary to-the-touch qualities mentioned in (4). I shall henceforth designate these qualities ‘chemical qualities’ and speak of, for example, ‘chemically cold’. There is nothing to prevent Earth becoming hot-to-the-touch while still being chemically-cold and still being Earth. The putative sense organ which is made of Earth can undergo ordinary alteration on coming into contact with an object which is hot-to-the-touch, and itself become hot-to-the-touch. Hence their being of Earth is not su¶cient for the literalist to explain why plants cannot perceive. I intend to show by examination of Meteorologica 4, that both theses of the determination of qualitative formulae by elemental formulae are false for the to-the-touch senses of ‘hot’, ‘cold’, ‘moist’, and dry’.

Created on 4 July 2005 at 14.34 hours page 315

Aristotle on Why Plants Cannot Perceive

315

7.1. The strong and weak theses are false for the to-the-touch senses of ‘hot’, ‘cold’, ‘moist’, and ‘dry’ General considerations If Aristotle did accept that the strong thesis held for the to-thetouch senses, his whole mode of expression in Meteorologica 4 would be misleading. Aristotle is perfectly happy to talk about gold melting or water becoming hot. He does not say, for example, ‘What people commonly describe as water becoming hot is really a transformation of some of the Water into Fire28 or a mixing in of bits of Fire from an external source.’ Nor does he say, for example, ‘What people commonly describe as gold melting is really a transformation of some of its Earth into Water.’ Indeed, if this were the case, it would be very strange to think that gold was meltable because it consisted predominantly of Water (Meteor. 389A7–9). On the model I wish to reject, he should have said that gold is meltable because it is such that its Earth can be readily transformed into Water. It is clear from Meteorologica 4 that Aristotle does indeed think of testing to see whether stu·s are predominantly Earthy or predominantly Watery, that is, whether they are chemically moist or chemically dry, by testing what dispositions or capacities they have. He does not say that we should simply investigate how moist (i.e. fluid) or dry (i.e. solid) a stu· is in its current state. At 389A4–7 he claims to have given a complete account for determining whether a stu· is Earthy or Watery.29 The method involves determining whether the stu· has certain passive capacities or incapacities. He says: κα λλοις οκειοτ-ροις πθεσιν, 9σα τ3 πσχειν λ-γονται, λ-γω δ> ο@ον τ τηκτν κα πηκτν κα καµπτν κα 9σα λλα τοια1τα· πντα γρ τ τοια1τα παθητικ, /σπερ τ #γρν κα τ ξηρ ν. το'τοις δ> Bδη διαφ-ρει ;στο1ν κα σρξ κα νε1ρον κα ξ'λον κα φλοις κα λθος κα τν λλων =καστον τν Cµοιοµερν µ*ν φυσικν δ* σωµτων. ε(πωµεν δ* πρτον τν
ριθµν ατν, 9σα κατ δ'ναµιν κα
δυναµαν λ-γεται. (Meteor. 385A4–12) And [mixed bodies are distinguished] by their more characteristic affections, which express their aptitude to be a·ected. I mean, for example, 28 Aristotle is willing to o·er this sort of explanation when he deems it appropriate. For example, at Meteor. 380A23–5 he says that everything that ripens changes from being Airy to being Watery and from being Watery to being Earthy. 29 J. Lennox, ‘Commentary on Sorabji’s “The Greek Origins of the Idea of Chemical Combination” ’, Proceedings of the Boston Area Colloquium in Ancient Philosophy, 4 (1988), 64–75, draws attention to the importance of Meteor. 4. 7–11 in diagnosing the chemical composition of homoeomerous stu·s.

Created on 4 July 2005 at 14.34 hours page 316

316

Damian Murphy

being meltable and solidifiable, and bendable and other such things. For all such qualities are passive, just like the moist and the dry. And by these qualities bone, flesh, sinew, wood, bark, stone, and each of the other natural homoeomerous bodies di·er. Let us first enumerate these qualities expressing the capacity or incapacity of a thing.

It is clear that to determine whether something is chemically moist or chemically dry, we do not simply touch it, we must investigate how it will behave in certain circumstances. For example, will it melt if heat is applied? Will it bend if a suitable force is applied? If Aristotle had thought that the strong thesis held for the to-thetouch senses, he need only have said that in order to determine whether something is Earthy or Watery one should feel it with the hand and determine whether it is dry-to-the-touch or moist-tothe-touch. To the general considerations from Aristotle’s text we might add a theoretical consideration. Any scheme which does not allow something to become hotter-to-the-touch or colder-to-the-touch without either undergoing a change of its own chemical constitution or the entering of some foreign hot or cold matter is very strange. It makes for great di¶culty in talking of the persistence of homoeomerous stu·s, for which identity criteria might reasonably be stated in terms of chemical composition. The classes of stu· in Meteorologica 4 Meteorologica 4 contains a discussion of the properties of various stu·s which consist primarily of Earth and Water in varying proportions (i.e. e+ w is almost but not exactly equal to 1). We should note that at GC 334B30–1 Aristotle says that every Aristotelianmixt (at least in the sublunary sphere) contains all four elements, so we should presume that the stu·s of Meteorologica 4 also contain traces of Air and Fire.30 However, the classes of stu· considered by Aristotle in Meteorologica 4 are based upon the relative amounts of Earth and Water. It is not surprising that Aristotle’s characterization of homoeomerous stu·s should be rather rough and ready. He does not venture on such rash claims as, for example, that gold contains three parts Water to one of Earth, but contents himself with dividing homoeomerous stu·s into a few broad classes. However, he has 30 Although this does not force us to claim that there cannot be pure samples of a single element. For a single element is not an Aristotelian-mixt.

Created on 4 July 2005 at 14.34 hours page 317

Aristotle on Why Plants Cannot Perceive

317

not been very careful with how he has named his classes of homoeomerous stu·s in Meteorologica 4. He mentions the following classes: (C1) (C2) (C3) (C4)

Consists of Water. Consists of Water and Earth but predominantly of Water. Consists of Earth and Water but predominantly of Earth. Consists of Earth.

I contend that, despite the terminological variation, there are only two classes: å (C2) and (C3) are mutually exclusive. å (C1) = (C2), å (C3) = (C4). A general argument for this conclusion is that Aristotle uses his names of these classes interchangeably. For example, at 384A11–14 he tells us that wine, urine, vinegar, lye, and whey consist of Water and Earth but predominantly of Water. However, at 389A9–11 he says that urine, vinegar, lye, whey, and some wines consist of Water.31 At 383B18–20 he says that clay, salt, and nitron consist of Earth and Water but predominantly of Earth, but at 388B12–13 he says that these same stu·s consist of Earth. I shall henceforth name (C1) and (C2) ‘the Watery class’ and (C3) and (C4) ‘the Earthy class’. Two counter-examples to the thesis of determination of qualitative formulae by elemental formulae for the to-the-touch senses the gold and brine counter-example Let us assume that the sense of ‘moistness’ in question is moistnessto-the-touch. Starting with Aristotle’s claims in Meteorologica 4 and the thesis that elemental formulae are su¶cient to determine qualitative formulae, anything which is Watery must be more moist than anything Earthy (assuming the proportion of Air is the same in each). Hence gold should be more moist-to-the-touch than brine. Yet the evidence of our sense organs tells us that brine is more moist-to-the-touch than gold. Given the innocuous assumption that the ordering of ‘more-moist-than’ is asymmetric, this shows that the most charitable interpretation of Aristotle is that relevant 31 He also says at 382B13 that wine, urine, and whey consist of Water.

Created on 4 July 2005 at 14.34 hours page 318

318

Damian Murphy

senses of ‘moist’ and ‘dry’ are not moist-to-the-touch and dry-tothe-touch. It may be possible to defend the claim that the chemical sense is the to-the-touch sense along the following lines. Admittedly, in Meteorologica 4 Aristotle generally mentions only the two predominant elements (Earth and Water) in a homoeomerous stu·. However, one might still claim that variations in the proportions of Air and Fire make a crucial di·erence. So, for example, the elemental formulae of gold and brine might be as follows: Gold <0.45, 0.53, 0.01, 0.01> Brine <0.4, 0.35, 0.01, 0.24> Hence: r(moistness of gold) = 0.54 r(moistness of brine) = 0.59 On this suggestion brine has more Earth than Water, but also has a substantial amount of Air, the other moist element. There is no reason to attribute this belief to Aristotle; I introduce it simply in my role as devil’s advocate. I am reluctant to embrace this line of argument. For presumably Aristotle did not mention the proportions of Air in the first place because they were minimal and incidental to explaining the characteristics of the homoeomerous stu·s in question, but this attempt to escape the counter-example does not treat the proportion of Air as merely incidental. the ice counter-example First let us assume that ice is the same chemical stu· as Water. For Aristotle did seem to think that ice was Water (e.g. Post. An. 95A16–21). Given they are the same stu·, they ought to have the same essential chemical characteristics. In particular, assuming that chemical stu·s are characterized by to-the-touch characteristics, ice and Water ought to be equally cold-to-the-touch. This is manifestly not the case. Ice is always colder-to-the-touch than Water. I conclude from this that being extremely cold-to-the-touch is not an essential characteristic of Water. So Water must be extremely cold in some more refined fashion. In this case, we can admit that ice and Water are the same stu·, inasmuch as they have the same

Created on 4 July 2005 at 14.34 hours page 319

Aristotle on Why Plants Cannot Perceive

319

essential chemical properties (they are both extremely chemically cold) and claim that they di·er only in incidental properties (one is colder-to-the-touch than the other). Modern chemistry makes a similar sort of claim. For it claims that ice and water are the same stu·, inasmuch as they have the chemical formula H O, and that 2 they di·er only in incidental properties, one being solid, the other liquid. On the assumption that ice is not chemically identical to Water (for example, because when Aristotle says that ice is Water, he does not mean to imply that they have exactly the same elemental formula, or he means that ice consists of common or garden water rather than pure elemental Water) we can still consider the following argument. (I1) (I2) (I3) (I4)

Water is extremely cold, i.e. r(hotness of Water) = 0. Hence nothing can be colder than Water. Yet ice is colder-to-the-touch than Water. Therefore the sense in which Water is extremely cold is not the to-the-touch sense.

The most obvious way to challenge this argument is to deny (I3). denying (I3 ) The idea is that although ice is undoubtedly colder-to-the-touch than common or garden water, it is not in fact colder-to-the-touch than elemental Water. For example, let us imagine that common or garden water has the elemental formula <0.01, 0.9, 0.01, 0.08> and ice has the elemental formula <0.08, 0.9, 0.01, 0.01>. This would explain why ice is colder and more solid than common or garden water. Each contains the same proportion of Water, but common or garden water has more Air and ice has more Earth. The problem is that this suggests that the real reason why the ice is solid is because of the Earth, and yet Aristotle does not mention this in his account of ice. If, on the other hand, the elemental formula of ice is something like <0.01, 0.97, 0.01, 0.01>, where ice is in fact closer to pure elemental Water than is common or garden water, then the most plausible extrapolation will be that pure elemental Water is in fact dry-to-the-touch like ice,32 contrary to the claim, which the advocate of the to-the-touch interpretation must make, that Water 32 Aristotle says that ice is actually but accidentally dry at PA 649B11–13.

Created on 4 July 2005 at 14.34 hours page 320

320

Damian Murphy

is extremely moist-to-the-touch. While it is legitimate to posit that elemental Water may have properties slightly di·erent from those of common or garden water, it is implausible to suggest that it has very di·erent properties. If elemental Water is more like ice than it is like common or garden water, it is rather misleading to call it ‘Water’. As a last desperate measure the defender of the claim that elements are characterized by to-the-touch tangible qualities might argue that Aristotle simply stipulates that Water is whatever is extremely cold-to-the-touch and extremely moist-to-the-touch. A priori stipulation would be one way to precisely define the elements, but I do not believe that Aristotle is doing this or that he should be doing this. His manner of introducing the elements at GC 330A30– b5 suggests that he is not stipulating in this fashion, especially the fact that he feels the need to justify his assignment of hot and moist to Air, stating that Air is a kind of vapour. If positing elements in his system is to have any value, their crucial characteristics must be determined a posteriori. Therefore I shall try not to convict Aristotle of redefining a priori the elements in terms of hotness-to-the-touch and moistness-to-the-touch. If he were guilty of this, the system of elemental notation would be redundant. The discussion in Meteorologica 4, where stu·s are classified in terms of proportions of elements, suggests this is far from being the case. Modified notions of heat Nowadays we are reluctant to accept hot-to-the-touch as a quality of the object at all, and are able to disambiguate various precisely defined modern physical notions which bear some relationship to being hot-to-the-touch. Given this, we might be inclined, while accepting that cold-to-the-touch is not a defining characteristic of Aristotle’s Earth, to claim that possessing a certain temperature or some combination of temperature and conductivity etc. is in fact the appropriate characteristic. To this thought we might add claims such as ‘Surely Aristotle could have meant to introduce the notion of standardized conditions for making perceptual judgements?’ For example, he could have insisted that, in judgements by touch of the hotness of an object, the area of contact or the volume of the object be equal in all cases. But these refined notions of hot-to-thetouch are no more suitable for defending the thesis that a stu·’s qualitative formula is determined by its elemental formula than

Created on 4 July 2005 at 14.34 hours page 321

Aristotle on Why Plants Cannot Perceive

321

the original rough and ready notion. Indeed, if by some miracle of time travel we could give to Aristotle precise modern notions, and indeed precise modern instruments, such as thermometers, the thesis that a stu·’s qualitative formula is determined by its elemental formula would not even hold for a precise modern notion such as temperature. Provided we take hotness and moistness to be states, neither thesis will hold. I take a state to be a property such that whether an object possesses that state at a particular time can in principle be verified without reference to, or knowledge of, the properties of the object at other times or in other possible worlds. Clearly in our usual employment of ‘hot’ we do think of it as a state. If I say that some water is hot, this is consistent with it remaining hot or becoming cold in the future, and with it being hot or cold in counterfactual scenarios. In order to avoid the ice counter-example, we require a notion of heat on which water, in its typical state, is roughly as hot as ice. But, to take a precise modern notion, this is certainly not the case for temperature. Equally, to take a more Aristotelian notion, nor is it the case that ice heats up or cools down another object roughly as much as water does.33 However we improve upon the notion of hot-to-the-touch, so long as we take ‘hot’ and ‘cold’ to refer to states as opposed to capacities we will not remove the counter-examples. Equally it is not the case that, on the one hand, gold is dry-tothe-touch, but on a more precise notion of moistness, is moist. For example, at GC 329B30–1 Aristotle does o·er a more precise sense, saying that the moist is that which, being readily determinable in shape, is not determinable by any limit of its own. 7.2. Hot, cold, moist, and dry as natural capacities By way of defending my own position I should say a few more words about what senses of ‘hot’, ‘cold’, ‘moist’, and ‘dry’ do characterize the elements. The programme I am proposing for revising the sense of the tangible qualities which we attribute to the elements is not akin to, for example, replacing the notion of hot-to-the-touch with the precise concept of temperature. I am not proposing reduction, elimination, or any other refinement of hot-to-the-touch per se. 33 This is o·ered as a means of determining which of two objects is hotter at PA 648B12–14.

Created on 4 July 2005 at 14.34 hours page 322

322

Damian Murphy

Rather, I take hot-to-the-touch to be first and foremost a state of the object, and believe the chemical elements to be characterized by natural capacities, not states. A fully worked distinction between states and natural capacities in the field of tangible qualities requires a more detailed discussion than space permits. Here I shall merely indicate the direction such a discussion would take. That something is hot by natural capacity means more than to say that it is potentially hot. Water is potentially hot, for it can be made actually hot by an external source such as a stove. Something is hot by natural capacity when it has its own internal source of becoming hot. So charcoal is hot in this sense, even if at the time it is cold-to-the-touch. Boiling water, while hotto-the-touch when it is boiling, is cold by natural capacity, because, left to its own devices, it quickly becomes cold. While the notion of natural capacity is di¶cult to articulate precisely, it is fundamental to Aristotle’s general mode of scientific explanation. Evidence that ‘hot’, ‘cold’, ‘moist’, and ‘dry’ can name natural capacities is provided by parts of Aristotle’s discussion of the various senses of these terms at PA 648B11–649B20. In particular, we might note: λλον δ* τρ πον θερµν πε'κη κα τ πονα, τ3 ταχD µεταβλλειν ες 8ν-ργειαν πυρ ς. (PA 649A27–9) In another way pinewood and fatty substances are hot, by quickly changing into the actuality of fire.

Although pinewood and fatty substances, in their most common state, are no hotter to the touch than other things, there is a sense in which they are hot because they have a greater natural capacity to go on fire and hence to become actually hot-to-the-touch. At this stage it is necessary to consider a passage which might seem to run counter to my argument. 7.3. DA 423b27–31 $πτα µ*ν οFν εσν αG διαφορα το1 σ4µατος HI σµα· λ-γω δ* διαφορς αJ τ στοιχε:α διορζουσι, θερµν ψυχρ ν, ξηρν #γρ ν, περ Kν ερ5καµεν πρ τερον 8ν το:ς περ τν στοιχεων. τ δ* ασθητ5ριον ατν τ $πτικ ν, κα 8ν 3K 7 καλουµ-νη $φ) #πρχει α(σθησις πρ4τ3ω, τ δυνµει τοιο1τ ν 8στι µ ριον. The distinguishing characteristics of body qua body are tangible. By distinguishing characteristics I mean those which di·erentiate the elements: hot and cold, dry and moist, about which we have spoken before in the treatise about the elements. And the sense organ for them, the organ of

Created on 4 July 2005 at 14.34 hours page 323

Aristotle on Why Plants Cannot Perceive

323

touch, that in which the sense of touch as it is called primarily resides, is the part which potentially has these qualities.

Sorabji interprets 423B27–31 as claiming that the same to-the-touch qualities (e.g. hot-to-the-touch) as characterize the chemical elements must be literally (according to him) received by the sense organ during perception.34 I have argued on the basis of De generatione et corruptione and Meteorologica 4 that Aristotle’s chemical elements are not characterized by to-the-touch qualities. However, it might be thought that this passage is an independent piece of evidence against my interpretation of Aristotle’s chemistry. Sorabji35 takes this passage to be about the material composition of the organ of touch. This is made absolutely clear if one reads down to 424A10, and I accept the interpretation. The problem for me is that the ‘them’ (ατν) in b30 clearly refers to to-the-touch tangible qualities. For these are the qualities that we perceive. In that case it might appear that the literalist is justified in taking b28–9 as endorsing the traditional view of Aristotle’s chemistry, that the essential characteristics of each chemical element are to-the-touch tangible qualities. For the purposes of this paper, my sole interest in the passage is to show that it is consistent with my account of Aristotle’s chemistry. This passage provides a second di¶culty for my interpretation, but one upon which I do not wish to dwell. Line b31 says that the organ of touch is potentially hot-to-the-touch, cold-to-the-touch, etc. This, and what follows down to 424A10, might be most naturally read in a literalist fashion.36 However, Burnyeat has shown that there is an alternative, spiritualist reading of the passage.37 It is not my intention to go over these arguments again here. I am taking for granted that Burnyeat’s general interpretation is at least possible and arguing that it has much greater success than the literalist interpretation in explaining why Aristotle thought that plants cannot perceive. Su¶ce it to say that I understand b31 in a spiritualist way. Returning to Aristotle’s chemistry, I accept that the ‘them’ of 34 Sorabji, ‘Intentionality’, 214, says: ‘There are three reasons why I think this first part of the passage cannot be handled by those who deny that Aristotle is referring to a literal taking on of temperatures and other qualities. First, a relevance must be supplied for the sudden reference in the middle of DA to GC and its doctrine that hot, cold, fluid, and dry are the defining characteristics of the four elements.’ 35 Ibid. 36 As Sorabji does (‘Intentionality’, 214–15). 37 Burnyeat, ‘Draft’, 20–1.

Created on 4 July 2005 at 14.34 hours page 324

324

Damian Murphy

b30 refers to to-the-touch tangible qualities. Yet I still maintain that it is natural capacities to be hot, cold, moist, and dry that differentiate the elements. While Aristotle could have been clearer on the point, what he does say at DA 423B27–31 is reasonable enough if my interpretation of his chemistry is correct. When he says that the very same tangible qualities as we perceive di·erentiate the elements, he is not saying anything false, for the natural capacities for these qualities do indeed di·erentiate the elements.38 But none of this entails that the current to-the-touch state of the sense organ can be determined simply from its elemental formula, in accordance with the strong thesis. In a way the claim about the material composition of the organ of touch is a weak one, that it must itself have tangible qualities, but it is not an insignificant claim for Aristotle to make. For the other sense organs can be devoid of all the qualities they are to perceive (for example, the organ of sight is colourless); only the organ of touch must have qualities in the ranges it is to perceive. Since the organ of touch consists of the elements, it must actually have to some extent a natural capacity to be hot-to-the-touch. But in order to have this natural capacity it must actually already have some degree of hotness-to-the-touch. For things which do not already have some degree of hotness-to-the-touch are not the sort of things to have to any extent a natural capacity to be hot-to-the-touch. Since we perceive states of objects and not their natural capacities, the literal acquisition of the tangible state of the object by the sense organ, which the literalist claims is necessary for perception, is compatible with the sense organ having the same natural capacity throughout the ordinary alteration, and thus is compatible with the organ being entirely of Earth or of 90% Earth. Therefore DA 423B27–31 still does not provide a licence for type A arguments. At this stage I take an opportunity to defend my overall argument against the literalist’s account of why Aristotle thought that plants cannot perceive. Even if one disagrees with my characterization of Aristotle’s chemical elements in terms of natural capacities, the literalist cannot run a type A argument. Even granting that the weak thesis holds for to-the-touch qualities, argument A fails because 38 I also understand DA 435A22–3 in a similar fashion. When Aristotle talks of the distinguishing characteristics of Earth, and implies that these are the qualities with which the sense of touch is concerned, this is true inasmuch as Earth is essentially characterized by having a natural capacity to be cold-to-the-touch and a natural capacity to be dry-to-the-touch.

Created on 4 July 2005 at 14.34 hours page 325

Aristotle on Why Plants Cannot Perceive

325

plants do not consist purely of Earth. Even granting that the strong thesis holds for to-the-touch qualities, argument A* fails because it allows us to prove too much. Again, while I believe that Aristotle’s chemical elements are characterized by precise extremes of hot, cold, moist, and dry, and this was crucial to the ice counter-example, to deny that elements are characterized by extremes is of no help to Everson, although the precise nature of my criticisms would need to be altered. First let us imagine that, although Earth is not cold in the extreme, it is moderately cold to some precise and unvarying degree. This still does not alter the fact that argument A* allows us to prove that ordinary alteration of hotness is generally impossible without change in chemical constitution. If, on the other hand, Earth’s being cold is compatible with a range of degrees of hotness, then something entirely of Earth, or of 90% Earth, can undergo ordinary alterations in respect of hotness within that range, and type A arguments do not even get o· the ground.

8. Plants cannot perceive because they do not have a mean Many explanations of what it is for a plant not to have a mean appeal to the theses of determination of qualitative formulae by elemental formulae and take the qualitative formulae to concern to-the-touch tangible qualities. Inasmuch as they do this, they are also incorrect. Such arguments claim that plants cannot perceive because being of Earth precludes their being in a material state which is necessary for perception. This sort of approach as to how we might understand 435A19–B1 is suggested by Sorabji.39 8.1. Argument B (2*) A plant, at all times, consists of 90% Earth and unvarying and precise proportions of the other elements. (10) Anything which, at all times, consists of 90% Earth cannot, at any time, be in a mean state with respect to hotness and moistness. (11) Anything which is to perceive hotness and moistness must 39 ‘Intentionality’, 216.

Created on 4 July 2005 at 14.34 hours page 326

Damian Murphy

326

be in a mean state of hotness and moistness (at least) at some time. Therefore, (7) A plant cannot perceive. Since I have already argued against (2), I state the argument using (2*) as the starting point. We may note that this argument does not make use of assumptions specific to the literalist.40 The assumption behind (10) is that, whereas being of 90% Earth entails that a plant must have fairly extreme qualities, if something consisted of, let us say, 25% of each element it would be in a mean state of hotness and moistness. Let us apply our second and fundamental criticism of (1*) to argument B. Is there equivocation in the senses of ‘hotness’ and ‘moistness’ employed in premisses (10) and (11)? (10) is most naturally interpreted as concerning the chemical senses and as such is true. The idea is simply that anything consisting of 90% of a single element (as opposed to roughly 25% of each of the four elements) cannot be in a mean state. (11) can be made true on any reasonable interpretation of what Aristotle says about the blind spot at DA 424A2–7 (and the literalist’s interpretation of this has considerable prima facie plausibility). For the blind spot phenomenon shows that the organ of touch has intermediate to-the-touch tangible qualities. Hence there is an equivocation. However, unlike the case of argument A*, it might be possible to remedy this. On the version of the argument that is neutral between spiritualism and literalism, which claims merely that it is a standing condition for perception that the organ be in a mean state, it would be possible to reinterpret (11) as claiming that the organ, in order to perceive, must be in a mean state with respect to the chemical qualities of hot, cold, moist, and dry. While the blind spot phenomenon 40 Premiss (11) is deliberately worded so as to be consistent with two more precise versions of the argument. One version, which even a spiritualist might be tempted to endorse, makes the claim that anything which is to perceive must be in a mean standing condition (regardless of whether or not it undergoes ordinary alteration of hotness and moistness). Cohen, ‘Hylomorphism and Functionalism’, 67, o·ers a version of the argument in which premiss (11) is replaced by the more specific claim that perception of tangible qualities requires an ordinary alteration which starts out from a mean of hotness or moistness. Thus, for Cohen, the reason that plants do not perceive is not the implausible claim that they fail to undergo ordinary alteration at all, but that they fail to undergo ordinary alteration of the right sort.

Created on 4 July 2005 at 14.34 hours page 327

Aristotle on Why Plants Cannot Perceive

327

shows us that the organ of touch happens to have intermediate to-the-touch qualities, what is essential is that it have intermediate chemical qualities. In this case my fundamental criticism is avoided. My own solution along spiritualist lines is a version of this claim. However, the literalist will wish to retain the to-the-touch reading of (11) while replacing (10) with: (10*) Anything which, at all times, consists of 90% Earth cannot, at any time, be in a mean state with respect to hotness-tothe-touch and moistness-to-the-touch. The thinking behind this would be as follows. While we might acknowledge that the evidence of Meteorologica 4 shows that the senses of ‘hotness’ and ‘moistness’ employed in chemistry are not the to-the-touch ones, or indeed states at all, it might be that the senses of ‘hotness’ and ‘moistness’ relevant to chemistry are incompatible with certain to-the-touch states. So, for some reason yet to be explained, being of 90% Earth, while consistent with a wider range of to-the-touch qualitative states than the strong thesis would allow, is still incompatible with being in a mean state with respect to the to-the-touch qualities at any time. This line strikes me as rather implausible. For water (which is presumably almost entirely Water) can even be boiling, as well as being of intermediate temperature. Again, some things which are Watery, such as gold, are very dry-to-the-touch. I shall not o·er a detailed criticism of (10*) here, but I think it is fair to say that the onus is on the literalist to defend (10*) if he elects to o·er it in his argument. 8.2. Recapitulation I have argued that a plant does not consist purely of Earth and hence certain arguments about why plants cannot perceive tangibles must be modified. Argument A does not survive its modification to A* because once we introduce the strong thesis we can prove too much. Most importantly, I have shown that the sense of ‘hot’, ‘cold’, ‘moist’, and ‘dry’ employed in characterizing the elements is not the to-the-touch sense. Hence arguments A and A* are unsound. For if we are to avoid equivocation and thus make the arguments valid, premisses (1) and (4) cannot be true together. A parallel criticism may be brought to bear against argument A* that (1*) and (4) cannot be true together. Similarly, in argument B premisses

Created on 4 July 2005 at 14.34 hours page 328

328

Damian Murphy

(10) and (11) are each most readily made true on a di·erent sense of ‘hotness’ and ‘moistness’. Attempts to rectify this either import an understanding of the mean which does not particularly help the literalist, or rest on a large promissory note. We must be careful to see where all this leaves the overall debate. We have not directly advanced the case of the spiritualist. The literalist is still left with his initial claim that, for Aristotle, a plant alters with respect to hotness and moistness only when the plant receives foreign matter. That this is, as a matter of fact, the only way in which plants undergo ordinary alteration is as ridiculous as it always was. I now turn to my account of why Aristotle thought that plants cannot perceive, which shows that the spiritualist understanding of receiving forms without matter can o·er a good and non-circular explanation of this fact.

9. What is it for plants to fail to have a mean? I suggest that to say that a plant does not have a mean is to say that it does not have any fixed degree of hotness-to-the-touch which is to be the benchmark against which the degrees of hotness-to-thetouch of objects are to be judged. The sense organ is described as a mean inasmuch as it is in a fixed state of hotness-to-the touch in between the more extreme degrees of hot and cold that it is to perceive. So the problem is not that plants are supposed to have extreme, as opposed to intermediate, degrees of hotness and moistness, but that their own degree of hotness and moistness is subject to uncontrollable variation.41 I shall first discuss an example case where the state of the sense organ may have some bearing on what sensible quality we perceive an object as having. 9.1. The three-buckets experiment To illustrate the role that a fixed mean plays in perception, let 41 That the fixed degrees of hotness and moistness of an organ of touch should be intermediate rather than extreme can be explained by what I term the ‘next best thing’ hypothesis. Ideally, for Aristotle, the sense organ is neutral between the sensible qualities it is to perceive by being devoid of them all (for example, the organ of sight is colourless). However, since the organ of touch cannot lack every degree of hotness or every degree of moistness (DA 423B27–31), the next best thing is for it to have intermediate degrees. This teleological explanation is neutral between literalism and spiritualism.

Created on 4 July 2005 at 14.34 hours page 329

Aristotle on Why Plants Cannot Perceive

329

us consider the simple experiment of putting one’s hand in three buckets of water in turn, one hot, one cold, and one tepid. If one puts the hand in any, removes it, and waits a while, and then puts it in another, one judges it for what it is. If, however, one goes directly from hot to tepid, one erroneously judges the tepid as cold, and if one goes immediately from cold to tepid, one erroneously judges the tepid as hot. That the di·erence between the state of the sense organ and the state of the sense object is crucial to what quality we perceive an object as having is acknowledged by both literalist and spiritualist in their explanations of Aristotle’s account of the blind spot phenomenon. Whatever the reason for it, that which is hotter-to-the-touch than the organ of touch we perceive as hot, that which is colderto-the-touch than the organ of touch we perceive as cold, and that which has the same degree of hotness as the organ of touch we do not perceive as hot or cold at all.42 The very fact that we can call some judgements erroneous and others correct is because for us the sense organ occupies a fixed position (from which it is temporarily dislodged after feeling an object of a di·erent temperature under certain circumstances). A plant does not have the ability to bring it about that its body returns quickly to its normal temperature. For a plant has no control over its own body temperature. 9.2. Plants Let us grant for the sake of argument that, all other things being equal, plants have a sort of ‘perception’, perception*. That is, let us grant them, per impossibile, a central sense organ and all the other things needed for making perceptual judgements. They do not, however, have a fixed mean. Therefore there would be no criteria by which a plant could distinguish the truth. For the plant is always in the position we are in while rapidly moving our hands from one bucket to another in the three-buckets experiment. Hot to a plant is simply hotter than the plant’s current temperature, cold is colder than the plant’s current temperature. That the plant’s current temperature varies results in the plant’s inability to use perception* as a basis of transtemporal comparisons or for forming any unqualified judgement. Whether a plant perceives* something 42 Corroboration is provided concerning the perception of hard and soft at Meteor. 382A11–21.

Created on 4 July 2005 at 14.34 hours page 330

330

Damian Murphy

as hot or cold depends on what state the plant happens to be in at the time. Plant perception* is not a method by which the plant can gain any reliable information about the object of perception. I shall briefly forestall a possible misunderstanding. ‘But all that human perception tells us is whether something is hotter or colder than our sense organ. How do we gain reliable information about the object of perception?’ We can accept that in one sense whether something is hot or cold is relative to us, but at least it is relative to us in a systematic and fairly constant fashion. What is hot (hot relative to us) on Tuesday is hot (hot relative to us) also on Saturday. However, it is not necessarily the case that what is hot relative to a plant on Tuesday is still hot relative to a plant on Saturday. This is the di·erence. For humans the three-buckets experiment is artificially contrived and represents abnormal conditions; for plants such conditions are normal. For Aristotle a perceptive faculty must satisfy something like the following condition: å A perceptive faculty is able to make, in normal conditions, true perceptual judgements of an unqualified nature. So, for instance, the sense of sight can judge that something is red, the sense of touch can judge that something is cold. There is no qualification to these judgements and, in normal conditions (e.g. when the subject is not mad or in pain, or has not immediately been feeling the hotness of something else) such judgements are always correct (e.g. DA 418A11–12). Since for a plant there are no normal conditions, because its own body temperature is not fixed but rather it acquires the temperature of its surroundings, any perception* is just as much false as true, and as such does not qualify as perception. This line of thinking is also applicable to other tangible dimensions. A plant varies in its degree of moistness, being moist in spring but dry in late summer. There will also be seasonal variations in its degree of hardness. 9.3. Evidence in Aristotle My suggestion as to what the mean, which plants fail to have, actually is relies on the idea that animals can maintain their body temperature to a greater extent than plants, and thus have an internal principle which enables them to be una·ected to a su¶cient

Created on 4 July 2005 at 14.34 hours page 331

Aristotle on Why Plants Cannot Perceive

331

extent in their matter by the forms of sensible objects. King gives an account of how Aristotle thought that animals maintain body temperature.43 Concoction of food in the stomach produces heat, and respiration produces cooling. A plant has neither of these functions. (Food is concocted outside the plant.) Plants are able to survive greater fluctuations in temperature than animals because they do not need to maintain their body temperature, but this lack of maintenance of body temperature prevents perception of hotness. A crucial assumption here is that Aristotle thought that all animals at least had better control over their body temperatures than plants. For he wishes to maintain that all animals possess the sense of touch (e.g. DA 434B13–14). Furthermore, it is important that Aristotle realized that plants su·ered marked variations in body temperature. King44 argues on the basis of De iuv. 470A27–32 that Aristotle thought that plants are more vulnerable to their environment than animals and as such are subject to seasonal withering and dying out. When Aristotle talks about some animals being hotter than others, he means that generally they have greater natural heat, and he does not mean that they are hotter in the to-the-touch sense. According to King,45 a feature of having a greater degree of natural hotness is that ‘they are not so much at the mercy of the vagaries of their surroundings—as plants are to an extreme degree’. On this account we could allow that those animals with a lower degree of natural heat have less control over their body temperatures, but more control than plants. From this we should expect that those animals with the greatest degree of natural heat, e.g. humans, had the most accurate sense of touch, because the fluctuations of the hotness of the sense organ were least. A piece of evidence in favour is that at PA 660A11–13 Aristotle says that mankind has the sense of touch most able to make perceptual discriminations. In this section I have outlined what it is for an organ of touch 43 R. King, Aristotle on Life and Death [Life] (London, 2001), ch. 4 (particularly sects. 4.4 and 4.5). For example, on p. 104 he says: ‘to see how touch is preserved, we have to ask why living things stay hot. “Hot” means some quality in the range between hot and cold. That is to say, “hot” is to be seen as a mean. Since touch requires a mean, and a mean is to be understood as a balanced state between extremes on a quality range such as hot–cold, the process of nutrition, in preserving the balance of the living thing also preserves the capacity for perception.’ 44 Life, 107. 45 Ibid. 84.

Created on 4 July 2005 at 14.34 hours page 332

332

Damian Murphy

to be in a mean state, why such a fixed mean state is necessary for perception, and why, in broad terms, animals but not plants can preserve a fixed mean state. Now I turn to explaining the specifics of the key De anima passages.

10. How my approach deals with DA 424A32–B3 and 435A19–B1 10.1. DA 424a32–b3 I noted (p. 300) that Sorabji accused Burnyeat of circularity in his interpretation of why Aristotle thought that plants cannot perceive. The advantage of my interpretation is that it links together the claim that plants do not have a mean with the claim that they have no principle such as to receive tangible forms without being a·ected in their own matter, and in doing so provides a genuine explanation along spiritualist lines of why they cannot perceive. A plant cannot perceive because it does not have a fixed mean to use as a benchmark against which to make perceptual judgements. It does not have a fixed mean because its own matter is a·ected by sensible objects. That is, a plant becomes literally hot-to-the-touch and cold-to-the-touch. In some cases the e·ect of a tangible object upon a plant will be so severe that there is no di·erence at all to register. For example, on a hot day the degree of hotness of the plant’s body is the same as that of the surrounding atmosphere, hence there is no di·erence to register and the plant cannot perceive that the air is hot.46 The same applies mutatis mutandis on a cold day. That the sense organ should become literally like the object of perception not only is not a necessary condition of perceiving but also is a su¶cient condition of not perceiving. For in such cases there is no di·erence to register. Not all cases will result in absolutely no di·erence between the degree of hotness of the plant and the degree of hotness of the sense object. Let us imagine that I hold an ice cube against the trunk of a tree. It is implausible to say that the whole tree must acquire the degree of hotness of the ice cube. However, even if the tree were to perceive* the cold of the ice cube, because of the lack of fixed mean, this would not count as perception. The reason that 46 That this would happen is implied by Aristotle’s discussion of the blind spot phenomenon at DA 424A2–7.

Created on 4 July 2005 at 14.34 hours page 333

Aristotle on Why Plants Cannot Perceive

333

the plant lacks a fixed mean is because it is too readily a·ected in its own matter. 10.2. DA 435a19–b1 We must still explain the claim at DA 435A19–B1 (p. 302 above) that plants cannot perceive because they are Earthy. Why should Earthy things be particularly prone to externally induced fluctuations in temperature? The passage is one of which no entirely adequate interpretation has come to my attention in the literature. Problems with Sorabji’s interpretations of the passage Even if, as I believe, I have done enough to show that Everson’s use of the passage to argue that plants cannot receive forms is mistaken (even if one were not to accept my analysis of Aristotle’s chemistry), Sorabji also o·ers two interpretations of the passage. His second interpretation is as follows (this is essentially what I label ‘argument B’): Alternatively, Aristotle may be using the idea of qualities received in a less usual way to refer to the organ’s standard qualities, not to those which it temporarily assumes during perception. He will be saying that the organ is standardly characterized by an intermediate blend of hot and cold, of fluid and dry, etc., and cannot just be cold and dry. That too would confirm part of what was said above but would throw no light upon what happens to the organ at the moment of perception. (‘Intentionality’, 216)

The idea is that the organ of touch, in its resting state, has a mean degree of hotness-to-the-touch, and for some reason, being of Earth precludes this. I agree with this much. However, as Sorabji himself admits, on such an interpretation Aristotle at 435A19–B1 is not talking about the act of perception. Thus this interpretation does not entail literalism. While I disagree with Sorabji as to why being of Earth precludes having a mean degree of hotness-to-the-touch, and thus prefer my own interpretation, his second interpretation o·ers a good enough account of Aristotle’s specific claim that their lack of a mean prevents perception in plants. I shall concentrate my fire on his first interpretation. If ‘receive’ here refers as usual to the perceptual process, there will have been some carelessness, because cold and dryness [i.e. the distinguishing characteristics of Earth] are precisely what plants, being already cold and

Created on 4 July 2005 at 14.34 hours page 334

334

Damian Murphy

dry, could not receive. If such carelessness is accepted, there will be further confirmation of my claim that the reception of perceptible form is the literal taking on of heat, cold, fluidity, and dryness, etc. (‘Intentionality’, 216)

As it stands, the text gives the impression that if the sense organ consisted of Earth, then it could receive the cold and dry (the distinguishing characteristics of Earth), but not all the other tangibles. On Sorabji’s interpretation of 423B27–424A1047 the exact opposite obtains. Something cold-to-the-touch and dry-to-the-touch should be able to receive the hot and the moist, but unable to receive the cold and the dry. (For the moment I am willing to grant, despite my previous arguments, that Earth is usually cold-to-the-touch and dry-to-the-touch. However, so as to avoid my first criticism of (1*) in Section 6, I shall assume that something which is predominantly of Earth can temporarily become hot-to-the-touch while still being of Earth.) I shall grant to Sorabji that Aristotle has been careless. However, whereas Everson’s argument, if it had been successful, would have explained why plants cannot perceive any degree of hotness or moistness, Sorabji’s explanation fails to do this. Let us recall from Section 2 (p. 300) what the literalist has to say about the blind spot (424A2–7). By this reasoning the literalist can explain why a plant, being of Earth, cannot perceive cold and dry. However there is no reason why something of Earth should not perceive anything that is hotter-to-the-touch than Earth. Again if, as I think he does, by ‘of Earth’ Aristotle means merely ‘almost entirely of Earth’, then, on the literalist’s account of the blind spot, an Earthy sense organ should have a blind spot for things that are very, but not extremely, cold-to-the-touch. On this account the literalist can explain only why plants cannot receive certain specific degrees of hotness and moistness. He cannot explain why they cannot receive any degrees of hotness and moistness. However, in this passage Aristotle wants to say why Earth (or a preponderance of Earth) could not form a perceiving organ of touch at all. In explaining this (as against explaining why there is a blind spot), there is nothing to make literalism seem plausible. My interpretation My account of this passage is along the following lines. Earthy things do not have the ability to maintain their temperature be47 Sorabji, ‘Intentionality’, 214.

Created on 4 July 2005 at 14.34 hours page 335

Aristotle on Why Plants Cannot Perceive

335

cause they have very little natural capacity to be hot. The reason that plants do not perceive is not that they are in a fixed cold-to-thetouch state, but that they do not have the sort of natural capacity (which animals have) for producing heat to counteract environmental factors. Something Earthy very easily becomes as cold as an object in contact with it, and cannot of its own accord restore an intermediate degree of hotness. While something Earthy might not acquire the precise degree of moistness of something in contact with it, without su¶cient natural heat, for example, to concoct the moisture, it is a·ected to a significant extent. So in each case something Earthy cannot adequately maintain a benchmark against which to perceive the tangible qualities of objects.48 We may note on this score that Aristotle suggests that all the sense organs must contain Fire because perception is impossible without heat (DA 425A6). While DA 435A19–B1 is di¶cult to interpret in isolation, we can make good sense of it in terms of our interpretation of DA 424A32–B3. This is preferable to trying to interpret DA 424A32–B3 in terms of 435A19–21, a passage which the literalist is forced to admit betrays carelessness. I now gloss my translation of 435A21–B1 to illustrate the basics of my interpretation. Afterwards I shall deal with some complications. a21–2: ‘For touch is a sort of mean of all tangible qualities.’ The organ of touch must have fixed and intermediate degrees of hotness, moistness, hardness, etc. a22–4: ‘and the sense organ is able to receive not only the distinguishing characteristics of Earth, but hot and cold and all the other tangible qualities’. The sense organ is able to receive not only the forms of cold and dry but also the forms of hot and cold and all the other tangible qualities.49 It is able to receive these forms in a spiritual way without being a·ected in its own matter. I note that ‘receiving

48 Although Aristotle does not o·er this reason himself, I claim that an organ of touch could not consist entirely (or almost entirely) of any single element for similar reasons. For example, if the organ of touch were Fire, then in the presence of a hot object it would very easily become hot-to-the-touch and could not return to an intermediate state of its own accord. Aristotle does not o·er such reasons in DA 3. 13 because he has already o·ered other reasons why the organ of touch could not consist of Fire, or Air, or (implicitly) Water (435A10–19). 49 See n. 38.

Created on 4 July 2005 at 14.34 hours page 336

336

Damian Murphy

forms without matter’ has not been forgotten at the end of DA 3. For Aristotle mentions this at 434A30.50 a24–B1: ‘And on account of this we do not perceive with bones and hair and such parts, because these are of Earth, and for this reason plants have no perception, because they are of Earth.’ It is on account of the fact that bones, hair, and plants do not have fixed and intermediate degrees of the tangible qualities and are incapable of receiving tangible forms in a spiritual way that they lack perception. A di¶culty with my interpretation I believe what I have said is a good solution for the problem of why plants or any organisms which consist entirely of Earth (or predominantly of Earth)51 do not perceive. However, it is not clear how it explains the lack of perception by bones and hair. On my reading Aristotle is saying that bones and hair do not perceive on account of the fact that they do not maintain a fixed mean state and cannot receive in a spiritual way all the tangible forms. This is very similar to his explanation of why plants cannot perceive at 424A30–B2. My interpretation is clear on how these features prevent perception. Where my interpretation struggles is in explaining why the fact that bones and hair are Earthy should result in their not being able to maintain fixed means and lacking the ability to receive in a spiritual way all the tangible forms. The specific problem for my interpretation is that, while bones and hair in themselves are Earthy, they are parts of the body of an animal and so it ought to be possible for their temperature to be 50 My interpretation must still face up to the strange suggestion that an Earthy sense organ could perceive cold and dry, but not hot and moist. I believe that Aristotle here could have in mind his criticism of the theory of Empedocles that like is perceived by like (DA 410A27–B1). Empedocles would have expected an Earthy sense organ to be especially receptive of the distinguishing characteristics of Earth. At 435A22–3 Aristotle corrects this position, insisting that an organ of touch must be able to receive all tangible forms, not only those which characterize what it is made of. The problem for Sorabji was that the combination of what Aristotle says about the blind spot with the traditional interpretation of his chemistry entails that an Earthy sense organ should be specifically unable to perceive the cold and dry because it would itself already be cold-to-the-touch and dry-to-the-touch. My account fares better. For on this account the distinguishing characteristics of Earth are that it is chemically cold and chemically dry, and this does not entail that it is cold-to-thetouch and dry-to-the-touch. Thus an Earthy sense organ, if there were such a thing, would not be specifically unable to perceive cold and dry. 51 Aristotle’s overall mission in this passage is to explain why an animal could not consist entirely of Earth (or almost entirely of Earth).

Created on 4 July 2005 at 14.34 hours page 337

Aristotle on Why Plants Cannot Perceive

337

maintained by other organs of the animal. It is not clear why being of Earth as such should prevent a particular body part (as opposed to the whole organism) from perceiving through touch. If it were a necessary condition of something’s being an organ of touch that it could itself maintain a mean degree of hotness, then being predominantly of Earth would explain the lack of perception of bones and hair. Some support to this line of thought is given by King, who suggests that while the heart is ultimately responsible for the maintenance of natural heat, individual body parts concoct their own nourishment. King says: Heat is produced in the heart insofar as food fuels the heat in the heart. But this does not mean that this heat is transmitted to all parts of the body. The heat in the heart produces nourishment, which in its final stages is blood, which is transmitted to the rest of the body. ‘Every part works on and concocts food with its natural heat.’ (JSVM [ = De iuv.] 469B11–13.) This nourishment keeps the heat going everywhere like logs on a fire. So, although all other parts depend on the work of the central organ—in sanguineous animals, on the heart for blood—each part ultimately concocts its own food with its own heat (De Generatione Animalium 784A34–B3).52

Earthy things have less natural heat and are less able to maintain a mean degree of hotness-to-the-touch by using nourishment in the blood. It might well be the case that in the extreme, bones and hair are entirely dependent upon surrounding organs and systems for the maintenance of their degree of hotness-to-the-touch. Thus they are not suited to being organs of touch. King also suggests specifically that an organ of perception preserves its own heat, rather than being preserved by the heart.53 While I admit this line is somewhat speculative, I note that the literalist does not comment on the specific case of why being Earthy prevents bones and hair from perceiving. To mention but a few difficulties with the literalist’s position, are we to think that bones and hair must always be very cold-to-the-touch? Or are we to think that the hair becomes hot only when foreign matter enters it, as the literalist thinks happens in the case of plants? I hope to have shown 52 King, Life, 97. 53 Ibid. 102: ‘“For nothing is capable of perception without heat.” (425A6.) This view of perception provides an additional motivation for the desire to see the seat of perception and nutrition as one: nutrition preserves a balanced heat, and this heat plays a role in perception. The perceiving body preserves itself, in so far as it nourishes itself.’

Created on 4 July 2005 at 14.34 hours page 338

338

Damian Murphy

that overall my interpretation can make better sense of the passage than literalism, and continues to do so even if, despite Section 7 of this paper, we allow the literalist the claim that Earth is essentially cold-to-the-touch and dry-to-the-touch.

11. Conclusion In this paper I have addressed a small aspect of the debate between the literalist and the spiritualist. I have argued, on the basis of an examination of Aristotle’s chemistry, that the literalist does not have an adequate account of why Aristotle thought that a plant cannot perceive. The spiritualist can make good sense of what Aristotle says, linking the claim that a plant lacks a mean with the claim, as the spiritualist understands it, that perception is the reception of forms without being a·ected in the matter. On this interpretation we can make sense of both DA 424A32–B3 and 435A19–B1. Trinity College, Cambridge

B I B L I O GR A P HY Bogen, J., ‘Fire in the Belly: Aristotelian Elements, Organisms and Chemical Compounds’, in Lewis and Bolton, Form, Matter, and Mixture in Aristotle, 183–216. Burnyeat, M., ‘Is an Aristotelian Philosophy of Mind Still Credible? A Draft’ [‘Draft’], in Nussbaum and Rorty, Essays, 15–26. ‘How Much Happens when Aristotle Sees Red and Hears Middle C?’, in Nussbaum and Rorty, Essays, 422–34. ‘Aquinas on “Spiritual Change” in Perception’, in Perler, Ancient and Medieval Theories of Intentionality, 129–154. ‘De Anima II 5’, Phronesis, 47 (2002), 28–90. Cohen, S., ‘Hylomorphism and Functionalism’, in Nussbaum and Rorty, Essays, 57–74. Everson, S., Aristotle on Perception [Perception] (Oxford, 1997). Fine, K., ‘The Problem of Mixture’, in Lewis and Bolton, Form, Matter, and Mixture in Aristotle, 82–182. Johansen, T., Aristotle on the Sense-Organs (Cambridge, 1998). King, R., Aristotle on Life and Death [Life] (London, 2001). Lennox, J., ‘Commentary on Sorabji’s “The Greek Origins of the Idea of

Created on 4 July 2005 at 14.34 hours page 339

Aristotle on Why Plants Cannot Perceive

339

Chemical Combination” ’, Proceedings of the Boston Area Colloquium in Ancient Philosophy, 4 (1988), 64–75. Lewis, E., Alexander of Aphrodisias: On Aristotle Meteorologica 4 (London, 1996). Lewis, F., and Bolton, R. (eds.), Form, Matter, and Mixture in Aristotle (Oxford, 1996). Magee, J., ‘Sense Organs and the Activity of Sensation in Aristotle’, Phronesis, 45 (2000), 306–30. Needham, P., ‘Aristotelian Chemistry: A Prelude to Duhemian Metaphysics’ [‘Aristotelian Chemistry’], Studies in History and Philosophy of Science, Part A, vol. 27.2 (1996), 251–69. Nussbaum, M., and Putnam, H., ‘Changing Aristotle’s Mind’, in Nussbaum and Rorty (eds.), Essays, 27–56. and Rorty, A. (eds.), Essays on Aristotle’s De Anima [Essays] (Oxford, 1995). Perler, D. (ed.), Ancient and Medieval Theories of Intentionality (Leiden, 2001). Slakey, T., ‘Aristotle on Sense Perception’, in M. Durrant (ed.), Aristotle’s De Anima in Focus (London and New York, 1993), 75–89. Sorabji, R., ‘Body and Soul in Aristotle’, Philosophy, 29 (1974), 63–89. ‘Intentionality and Physiological Processes: Aristotle’s Theory of Sense-Perception’ [‘Intentionality’], in Nussbaum and Rorty (eds.), Essays, 195–226. ‘Aristotle on Sensory Perception and Intentionality: A Reply to Myles Burnyeat’, in Perler (ed.), Ancient and Medieval Theories of Intentionality, 49–61.

Created on 4 July 2005 at 14.34 hours page 340

page 340 is blank

Created on 4 July 2005 at 14.35 hours page 341

ARI S TOTELI AN TELEOLOGY LINDSAY JUDSON

 is November, and the long, dry summer is past. The rain is falling, and the farmer knows that this means that the corn he has just planted will grow.1 Is the rain, in Aristotle’s view, a process which has the growth of crops—or of plants more generally—as its end, or is it a process which has a quite di·erent end (or no end at all), and which the farmer merely exploits for his own agricultural purposes? Neither answer is, on the face of it, incoherent. The general can be pleased at the upsetting e·ect which an eclipse has on the enemy troops, and with su¶cient astronomical knowledge he might even have planned his campaign so as to exploit this effect: but no one believes that Aristotle would take the eclipse to be even partly for the sake of upsetting the troops.2 On the other hand, the farmer can be pleased at the ripening of his olives, which is for Aristotle a paradigm case of a natural teleological process. In the absence of a belief either in divine providence or in the existence of a global teleological Nature,3 however, what we might call the biocentric view, that the rain falls for the sake of plant life in general, and hence for the sake of all animal life, seems quite bizarre. Equally hard to find plausible is a special form of biocentrism—the anthropocentric view that the rainfall is for the sake of the growth of the farmer’s crops because the natural world is naturally—rather than providentially—organized around the life of human beings. ã Lindsay Judson 2005 An earlier version of this paper was given at the University of Catania: I am grateful to Franco Repellini and Mario Vegetti for comments during the discussion. I am also grateful to David Charles and David Sedley for their extensive comments on subsequent versions. 1 For details of the ancient farming year, see R. Osborne, Classical Landscape with Figures: The Ancient Greek City and its Countryside (London, 1987), 13–15; I am grateful to Simon Price for this reference. 2 Perhaps an eclipse is not for the sake of anything: Metaph. Η 4, 1044B3–14. 3 David Sedley comes close to ascribing this latter idea to Aristotle: see pp. 360–1 below.

Created on 4 July 2005 at 14.35 hours page 342

342

Lindsay Judson

None the less, the attribution to Aristotle of one or other of these views of natural teleology has been defended by John Cooper, David Furley, and David Sedley.4 Cooper argues for the ‘plain’ biocentric view: he thinks that seasonal rainfall and other natural processes are for the sake of life in general, and not merely for the sake of human life; for Sedley, Aristotle’s natural teleology is thoroughly anthropocentric: he endorses the biocentric view, but also sees all or most of animal life as being in turn for the sake of human life.5 Furley’s main concern is to defend the thesis common to both of these views, that the rainfall is (at least partly) for the sake of the growth of crops; the main aims of the present paper are to criticize this thesis and to o·er a fresh interpretation of the key passage, the opening section of Phys. 2. 8 (198B10–199A8).

I Aristotle’s argument at 198B34–199A86 seems to run as follows: (1) If something occurs which works to the benefit of some4 J. Cooper, ‘Aristotle on Natural Teleology’ [‘Natural Teleology’], in M. C. Nussbaum and M. Schofield (eds.), Language and Logos: Studies in Ancient Philosophy Presented to G. E. L. Owen (Cambridge, 1982), 197–222; D. Furley, ‘The Rainfall Example in Physics ii. 8’ [‘The Rainfall Example’], in A. Gotthelf (ed.), Aristotle on Nature and Living Things (Pittsburgh and Bristol, 1985), 177–82; D. Sedley, ‘Is Aristotle’s Teleology Anthropocentric?’ [‘Aristotle’s Teleology’], Phronesis, 36 (1991) 179–96. 5 I have not encountered a defence of what we might call ‘limited anthropocentrism’—the idea that the rainfall is only for the sake of agricultural crops. 6 δνατον δ το τον χειν τν τρπον. τα τα µ ν γρ κα πντα τ φσει  αε οτω γγνεται  ς !π τ πολ, τ#ν δ$ π τχης κα το α&τοµτου ο&δ(ν. ο& γρ π τχης ο&δ$ π συµπτ)µατος δοκε* ειν πολλκις το χειµ#νος, λλ$ !ν +π κνα· ο&δ καµατα +π κνα, λλ$ /ν χειµ#νος. ε ο0ν  π συµπτ)µατος δοκε*  1νεκ του ε2ναι, ε µ3 ο4ν τε τα τ$ ε2ναι µ5τε π συµπτ)µατος µ5τ$ π τα&τοµτου, 1νεκ του /ν ε6η. λλ µ3ν φσει γ$ !στ τ τοια τα πντα, ς κ/ν α&το φα*εν ο7 τα τα λ(γοντες. στιν 8ρα τ 1νεκ του !ν το*ς φσει γιγνοµ(νοις κα ο0σιν. (‘Yet it is impossible that this should be the true view. For these and all natural things come about in a given way either always or for the most part; but of not one of the results of luck or chance is this true. We do not ascribe to luck or mere coincidence the frequency of rain in winter, but frequent rain in summer we do; nor heat in summer, but only if we have it in winter. If, then, it is agreed that things are either the result of coincidence or for the sake of something, and these cannot be the result of coincidence or chance, it follows that they must be for the sake of something; and that such things are all due to nature even the champions of the theory which is before us would agree. Therefore action for an end is present in things which come to be and are by nature.’)

Created on 4 July 2005 at 14.35 hours page 343

Aristotelian Teleology

343

thing, either it does so by coincidence or it happens because of the good it does. (2) Coincidences are necessarily exceptional occurrences: they happen neither always nor for the most part. (3) The beneficial arrangement of teeth in animals (and of course the occurrence and beneficial functioning of many other animal parts) occurs either always or for the most part. So (4) The development of teeth in this arrangement (and the coming to be of many other animal parts) is for the sake of the animal’s life and well-being.7 Cooper takes premiss (3) as evidence that Aristotle’s argument rests on the ‘alleged fact’ that the species of organisms which we find in the world are eternal; this alleged fact is itself a ‘reasonable inference’ from observed facts concerning the harmonious and ecologically stable interaction of animal and plant phenomena.8 He argues that ‘always’ in premisses (1)–(3) must be construed in a strong sense which involves the notion of regular recurrence throughout all time; for otherwise premiss (3) is (according to Cooper’s view of Aristotle) compatible with our being in a Democritean world, in which present regularities are merely temporary and a matter of chance—and so Aristotle’s argument will not work. Cooper argues that the basis for Aristotle’s belief in the existence of teleology in nature is the alleged fact of species permanence together with the claim that: if it is a fundamental fact about the world, not derivable from other natural principles, that it maintains forever these good life-forms, then the processes by which it does so, being processes by which some good is achieved, are for the sake of the outcomes.9 7 Note that on this reading of the argument Aristotle does not claim that all regularities are for the sake of something, but only that the regular occurrence of a beneficial outcome is (in this way it is parallel to the idea that chance events are events which happen neither always nor for the most part and which are also beneficial: see Phys. 2. 5–6). Even this claim he need not hold without exception: the argument requires only that in general such regularities are for the sake of something. 8 ‘Natural Teleology’, 202–4 and 209: I discuss this inference below. 9 Ibid. 213. This form of teleological explanation of the eternity of species must be distinguished from (although it is compatible with) the explanation which Aristotle gives at GA 2. 1, 731B20–732A3, and GC 2. 10, 336B25–32. If Cooper’s account were right, his ‘goals actually existing in nature’ would constitute the mechanisms, as it were, by which the eternity of organic species is brought about, while the argument

Created on 4 July 2005 at 14.35 hours page 344

344

Lindsay Judson

Taken together, these claims yield the conclusion that regularly occurring rainfall is for the sake of plant growth (and consequently for the sake of all life-forms which depend ultimately on plants). Cooper’s way of understanding Physics 2. 8 presents Aristotle with an unenviable dilemma, however. Aristotle is supposed to be willing to concede that processes of ‘regular’ generation could be no more than a long run of good luck even if they have been occurring throughout recorded history, provided only that they have not been occurring eternally. But the concession that temporary ‘regularity only throughout history’ can be a matter of chance undercuts the basis of the ‘reasonable inference’ which Cooper ascribes to Aristotle—the inference from the ecological stability of the present order to its permanence—precisely because that present stability is compatible with our being in a Democritean world. If Aristotle insists that the inference is none the less reasonable, his materialist opponents can presumably advance an analogue of it to explain why the ‘run of good luck’ is in fact permanent: it is permanent because the ecosystem is self-sustaining.10 To say this would of course be to o·er an overarching explanation of the run of ‘luck’; but this explanation would not appeal to the existence of goals in nature, and hence would not involve any concession to Aristotelian teleology as Cooper construes it.11 Perhaps Cooper can reply that the materialist cannot advance such an explanation, because the inference to permanence is sound only if one takes the present harmonious

in the GA and GC would explain why these mechanisms exist at all (by showing that the eternity of species is better than non-eternity). 10 Sarah Waterlow o·ers the materialists a similar argument in her discussion of Physics 2. 8 in Nature, Change, and Agency in Aristotle’s Physics: A Philosophical Study [Nature, Change, and Agency] (Oxford, 1982), 77–9: see p. 351 below. 11 Cf. C. Witt, Substance and Essence in Aristotle: An Interpretation of Metaphysics VII–IX (Ithaca, NY, and London, 1989), 92–3, and L. Judson, ‘Chance and “Always or for the Most Part” in Aristotle’ [‘Chance’], in Judson (ed.), Aristotle’s Physics: A Collection of Essays (Oxford, 1991), 73–99 at 87–8. It is not impossible, of course— though it seems unlikely—that in 2. 8–9 Aristotle does assume that the main outlines of his world-picture, including the permanence of (most) species, are correct, and is merely arguing that teleological explanation of organic phenomena is required given that basic world-picture. If this were so, Cooper’s account of Aristotle’s lines of argument in these chapters would not be open to the objection I have raised (though there would then be no need to see the appeal to the instantiation of species throughout all time as conveyed by the sense of ‘always’, since the permanence of species would be being assumed in advance of the argument). But Cooper would still face the di¶culty outlined in the next paragraph.

Created on 4 July 2005 at 14.35 hours page 345

Aristotelian Teleology

345

order not to be due to chance; but then Aristotle’s argument too blatantly begs the question against his opponent.12 For these reasons Cooper’s interpretation of ‘happening always’ does not yield a good account of Physics 2. 8 (nor of 2. 9, for the same reasons). Cooper’s other claim, that Aristotle believes that any regularly occurring process which helps to maintain life-forms occurs for the sake of those life-forms, is also questionable. Of course Aristotle thinks that some of the processes which help to maintain life-forms are for their sake; but it does not follow that he thinks that everything needed to explain species permanence happens for the sake of those species, since he might think that there are restrictions on which processes admit of the appropriate teleological explanation. I shall return to this question later. A less ambitious argument than Cooper’s might seem to be available, however—one which is deployed by Furley and endorsed by Sedley. It is undeniable that the winter rainfall benefits the crops in a regular way by enabling them to grow to maturity and thus to fulfil their nature; and however we understand Aristotle’s notion of happening ‘always or for the most part’,13 it is also undeniable that he thinks that rainfall in the winter happens always or for the most part.14 Thus it appears to follow that winter rainfall must be for the sake of the growth of plants: for according to premiss (1) the rainfall must either be a coincidence or for the sake of the crops, and according to premiss (2) it cannot be a coincidence if it occurs always or for the most part.15 With this argument to hand, it might seem surprising that anyone has ever thought that Aristotle did not believe that rain was for the sake of the growth of plants. There are a number of reasons why we should be reluctant to accept Furley’s conclusion, however. First, the view it ascribes to Aristotle is, as I have said, 12 The line of argument for the permanence of species which Cooper actually sketches on Aristotle’s behalf seems to fall into both traps. Cooper writes, ‘[i] One observes in the world itself, then, no internal disharmony or imbalance that could lead to its eventual destruction; and since [ii] there is nothing outside it that could attack it and cause its disintegration, it seems only reasonable to believe that in these respects no change is to be anticipated’ (‘Natural Teleology’, 204, numbering added). Point [i] can be happily appropriated by the materialist in the way that I have suggested, while to assert point [ii] is already to deny the truth of Empedoclean or Atomistic physics. 13 For discussion see Judson, ‘Chance’. 14 See 198B34–199A3; the example of hot weather (though not that of rain) recurs at Metaph. Ε 2, 1026B30–6. 15 See Furley, ‘The Rainfall Example’, 179; Sedley, ‘Aristotle’s Teleology’, 182–4.

Created on 4 July 2005 at 14.35 hours page 346

346

Lindsay Judson

quite implausible, even against the background of his teleology of animal parts and functions, unless one supplements it with some view generally thought un-Aristotelian—the view that the world is governed by divine providence, or the view that it is the work of a cosmic Nature. Second, it appears to allow in far too much within the scope of natural teleology. Suppose that avalanches regularly cut o· access to a certain Alpine valley, and that one of two constantly warring tribes regularly takes advantage of this to regroup in the safety of the valley: by the present argument the avalanches happen for the sake of providing safety for the tribe. Likewise, if examination candidates regularly use their summer hayfever as an excuse for poor performance, the hayfever occurs for the sake of providing an excuse; and so on.16 Third, Aristotle appears to o·er the rainfall example precisely as a case in which the benefit is not a matter of teleology: χει δ$ ποραν τ κωλει τ3ν φσιν µ3 1νεκ του ποιε*ν µηδ$ :τι β(λτιον, λλ$ <σπερ ει = Ζε?ς ο&χ :πως τν σ*τον α&ξ5σAη, λλ$ !ξ νγκης . . . (198B16–19) A di¶culty presents itself: why should not nature work, not for the sake of something, nor because it is better so, but just as Zeus rains, not in order to make the corn grow, but of necessity?

Furley replies to this last point that it is possible to take the rainfall example as the objector’s, not as Aristotle’s: Aristotle is presenting an aporia, and so we can read him as saying: suppose an objector says: ‘why should not nature work, not for the sake of something, nor because it is better so, but just as Zeus rains, etc.’ . . .

This is unconvincing: there is no hint that we are to understand the illustration of the aporia in this ‘distanced’ way. Moreover, the second illustration—the rain rotting the corn at 198B21–3—is clearly meant as a straightforward case of the lack of a teleological connection, and it would be very strange if Aristotle presented his 16 Of course (as David Sedley has pointed out to me) there are ways to block these undesirable consequences by restricting the sorts of regularities and/or beneficial consequences to which the argument is meant to apply; and I shall argue below precisely that premiss (1) must be understood in a restricted way. But the decisiveness that Furley’s argument appears to o·er for the claim that Aristotle takes winter rainfall to be for the sake of the crops stems from its unrestricted reading of the premisses: once it is envisaged that Aristotle is applying restrictions, then it becomes an open question once again whether he is including or excluding the rain’s e·ect on the crops.

Created on 4 July 2005 at 14.35 hours page 347

Aristotelian Teleology

347

opponent’s case in this way—using without comment one example which constitutes uncontentious common ground, and one which embodies the very disagreement at issue. Furley claims that the case of the rotting corn is meant as an argument by the objector for the non-teleological interpretation of the first rainfall example, and that the presence of this argument shows that the first example is indeed meant to be contentious. But first, Aristotle introduces the second example with =µοως δ κα (‘similarly’), not with γρ (‘for’): the two examples are presented in parallel. Second, on Furley’s account, Aristotle ought not to find it so obvious that the rain is not for the sake of rotting the corn: if disgruntled farm-workers regularly leave some of their masters’ corn out to rot in the rain, then on Furley’s account the rain is for the sake of spoiling the corn: this is simply another example, like the avalanche and hayfever cases, of the implausible consequences of the view Furley ascribes to Aristotle.17 Fortunately, Furley’s reading of our passage is by no means obligatory: the dichotomy of ‘by chance or for the sake of something’ (199A3 ·.) need not be meant to hold for everything which has a beneficial outcome and which happens always or for the most part. Premiss (1), in other words, may be intended to apply only to a restricted range of cases; in particular, it need not apply to the rain and hot weather mentioned in the previous sentence (198B36–199A3)— they may be examples introduced simply to illustrate the point that happening rarely is a necessary condition of happening by chance.18 The question, then, is: what is the restriction and how can it be motivated? On what basis can Aristotle think that the dichotomy does apply to the case of teeth if it does not apply to that of the regular and beneficial occurrence of seasonal rain? If the former is for the sake of easier eating, why is the latter not for the sake of the growth of corn? 17 Sedley argues, apropos of this passage, that Aristotle’s use of the phrase ‘Zeus rains’ at 198B18 shows he ‘has in mind someone’s attempt to reinterpret as mechanical a phenomenon popularly viewed as an act of god’, and thus that ‘right from the start the issue is one of specifically anthropocentric teleology’ (‘Aristotle’s Teleology’, 185). But even if Sedley is right as to why the phrase ‘Zeus rains’ appears, what would have been at issue would, clearly, have been anthropocentric providence: so even on this account Aristotle has as much reason to distance himself from the implications of ‘Zeus rains’ as his ‘mechanist’ opponent (cf. R. Wardy, ‘Aristotelian Rainfall or the Lore of Averages’ [‘Aristotelian Rainfall’], Phronesis, 38 (1993), 18–30 at 20). 18 Cf. T. H. Irwin, Aristotle’s First Principles (Oxford, 1988), 522 n. 18.

Created on 4 July 2005 at 14.35 hours page 348

348

Lindsay Judson

It might be suggested that the di·erence between the two types of case is that the growth of the corn already has a teleological explanation in terms of human purposes. This response will not do for a number of reasons. (1) There is no obvious reason why there cannot be teleological overdetermination, nor why di·erent ends cannot form a conjunctive final cause. (2) There are parallel cases in which it is hard to imagine Aristotle denying the applicability of natural teleology—cases such as breeding livestock, where human beings clearly make use of processes which are naturally teleological in their own right. (3) The seasonal rain promotes the growth of many plants not cultivated by humans: the view that Aristotle takes the rain to be for the sake of these is equally unattractive. Nor can the di·erence be that, in the case of the rain but not in the case of the teeth, we can give an adequate explanation without having to appeal to any final causes; for the issue in 2. 8 is precisely whether this is true of the case of the teeth. The answer I wish to propose is that Aristotle intends premiss (1) to apply only to the generation, functioning, and parts of natural substances. In the case of the arrangement of teeth, the processes which bring about the good outcome, and which Aristotle will say are for the sake of that outcome, are part of the natural functioning of a single substance which is also the primary beneficiary of the outcome.19 This is clearly not so in the rain case: neither the rain which falls nor the meteorological processes which cause it are primarily responsible for the presence of the seeds or young plants (though of course the fact that rain regularly falls in this particular place may be partly responsible), and the good outcome is not a good for—does not constitute or bestow a benefit on—the rain, clouds, etc. If the rain case did resemble the case of the teeth in these respects, then we would indeed have reason both to see the rain/corn processes as being processes in a single whole and to see the successful growth of the corn as a per se final cause of the rain.20 19 Of course other processes external to the substance are required too: the point of the present suggestion is that for Aristotle these processes form part of the backdrop of normal conditions against which the teleological drama is played out by individual natural substances (cf. P. Pellegrin, ‘Les ruses de la nature et l’‹eternit‹e du mouvement’ [‘Les ruses’], in M. Canto-Sperber and P. Pellegrin (eds.), Le Style de la pens‹ee: recueil de textes en hommage a› Jacques Brunschwig (Paris, 2002), 296–323). 20 I would suggest that we should see a similar line of thought at work in the case of those organisms whose normal form of coming to be is spontaneous generation— i.e. the testacea and certain species of fish, insects, and plants (see GA 3. 11; HA 5. 1, 539A15–25; cf. HA 5. 15–16, 19, 31–2; 6. 15–16). I hope to discuss this elsewhere.

Created on 4 July 2005 at 14.35 hours page 349

Aristotelian Teleology

349

If Aristotle does intend this restriction to apply, how should we think of it as figuring in the argument? The key sentence is at 198B34–6, where Aristotle introduces premisses (1)–(3), and says ‘For these and all natural things [τα τα µ ν γρ κα πντα τ φσει] come about in a given way either always or for the most part . . .’. On my reading τα τα refers precisely to the restricted range of things (natural substances, their parts, generation, etc.) and not to everything that happens by nature. That this is the natural way to understand τα τα is conceded even by Furley: ‘the immediate antecedent to which τα τα most naturally attaches itself in the preceding paragraph is the parts of animal bodies’.21 Furley none the less thinks that τα τα must be understood as referring to natural things without restriction, for two reasons.22 First, the cases which Aristotle uses to support premiss (2) fall outside the restricted range—rainfall and hot weather. I agree that, on my reading, Aristotle illustrates premiss (2) with cases which (though of course natural) fall outside the scope of the argument: τα τα µ ν γρ κα πντα τ φσει should be understood as ‘these things (animals, their parts, etc.), and indeed everything that is or happens by nature . . .’.23 But this is hardly a problem. The good e·ects of the winter rain were introduced at the outset as a foil to the case of the teeth, and so it is quite natural for Aristotle to turn to it, and to its ‘partner’ the seasonal summer heat,24 as illustrations here. Indeed, it might have seemed positively desirable to use examples from outside the range of cases for which Aristotle is claiming that teleological explanation is required. Second, Furley argues, if Aristotle had intended premiss (1) in the restricted way, he would have been obliged to explain the nature of the restriction, and to say why the premiss was true for the cases to which it did apply—‘and there is of course no trace of this explanation’.25 I suggest that Aristotle does not say anything about these things because he takes his restriction of premiss (1) to 21 ‘The Rainfall Example’, 180. 22 Furley’s view of the reference of τα τα is endorsed by Sedley (‘Aristotle’s Teleology’, 182–3) and by Wardy (‘Aristotelian Rainfall’, 19–21), who thinks that Aristotle does take (all) rainfall to be for the sake of something—but that what it is for the sake of is the rain’s reaching its natural place. 23 I am grateful to David Charles for suggesting this reading of πντα τ φσει. Note that while the subject of 1νεκ του /ν ε6η at 199A5 must, on my reading, be the restricted range of natural things, τ τοια τα πντα at a6 could be understood either in the restricted or in the unrestricted way. 24 For a good account of this, see Sedley, ‘Aristotle’s Teleology’, 185–6. 25 ‘The Rainfall Example’, 180–1.

Created on 4 July 2005 at 14.35 hours page 350

350

Lindsay Judson

natural substances and their parts as obvious.26 From the outset, in other words, the issue for Aristotle is whether teleology is required for substances and their parts: it is this view to which he thinks Empedocles’ ox-progeny—and Democritus’ atoms—pose a threat. We should read premiss (1), then, as restricted to the workings of natural substances: if this is correct, Aristotle is not committed to the extension of teleology to the seasonal rain, hayfever, or the convenient avalanche. It has to be admitted that this reading of 2. 8 requires us to see Aristotle as setting up the debate with his opponent(s) to some extent on his own terms: a careful Empedoclean—or a careful atomist—should not perhaps concede that animals are substances.27 But this sort of tipping of the scales is entirely characteristic of Aristotle’s discussions with his predecessors. Note that I am not suggesting that Aristotle is trading on the idea of an essential connection between being a natural substance and being teleologically explicable. All that Aristotle need invoke here is a rough-and-ready conception of natural substance to which all or most parties might agree.28 Whatever Aristotle’s considered position on the metaphysical connection between natural substance and teleology, the argument in 2. 8 should be seen as working towards it rather than from it. Does the premiss that animals and plants are substances none the less beg the question against Aristotle’s opponents, since they might say that it is indeed part of their position that only the elements (or atoms) are, even in these general terms, natural substances?29 Aristotle can at least reply (1) that he has independent grounds for (and the weight of common opinion in favour of) this premiss, and (2) that the present argument shows, in e·ect, part of the high price of giving it up, as we shall see in the next section.30 26 Perhaps not surprisingly, Aristotle’s general analysis of chance and luck in Physics 2. 4–6 also depends crucially on the distinction between the workings of substances and the e·ects of conjunctions of causes: see Judson, ‘Chance’. 27 For a closely related idea see S. Sauv‹e Meyer, ‘Aristotle, Teleology and Reduction’ [‘Reduction’], Philosophical Review, 101 (1992), 791–825 at 812–16; see also n. 38 below. 28 Cf. the ‘rough-and-ready’ usage of substance (in general) which Aristotle deploys in Metaph. Λ 1. 29 This is not to say that this was in fact the position of Aristotle’s opponents, not least because of possible underdetermination of their position(s) vis-›a-vis the notion of natural substance; but it is not an implausible response to o·er on their behalf (indeed thinkers who deny the reality of coming into existence may be under some pressure to assign ontological posteriority to temporary compounds). 30 For point (1), see Metaph. Ζ 2 and Η 1; I discuss point (2) below, pp. 352–5. In

Created on 4 July 2005 at 14.35 hours page 351

Aristotelian Teleology

351

II How then should we construe Aristotle’s argument? What lies behind premiss (1), construed as restricted to natural substances and their parts, is that Aristotle thinks that there can be no explanation of regular benefit within the processes of an individual substance without teleological explanation. Debate has focused on the idea that Aristotle thinks that his opponents cannot explain why features such as the normal arrangement of teeth regularly come about. The problem then is to see why he thinks that they cannot explain this. Aristotle is thus seen as conceding that Empedocles’ idea that ill-formed animals simply die o· might explain why at some putative early stage there were not many ill-formed animals alive,31 but as insisting that it cannot explain why well-formed animals are the norm in all subsequent generations: that would require in addition the idea that animals reproduce to type. It is then hard to see how Aristotle can have even a remotely satisfactory argument. A good statement of the problem is given by Waterlow: [Aristotle’s materialist opponents] could have grounded the combination [of the material factors in stable and self-reproducing compounds] in the natures of the factors themselves: these stay combined because they are such as to interact to form a stable system . . . It results because the factors, once together, a·ect each other in such a way that they physically cohere.32

One view is that Aristotle simply envisages opponents who will cave in and agree that regular reproduction according to type is incompatible with the idea that animals come to be by the chance assemblage of parts: thus Aristotle takes Empedocles as not even trying to appeal to Waterlow’s argument.33 The di¶culty with this the case of Cooper’s interpretation, by contrast, the charge of question-begging constitutes a serious di¶culty. This is because according to Cooper, Aristotle’s argument rests on his insistence that species and other natural regularities are eternal: this is a premiss for which there is no obvious independent support, and one which Aristotle’s opponents have no reason to endorse, and every reason to reject. 31 Cf. Irwin, Aristotle’s First Principles, 107. 32 Nature, Change, and Agency, 79. 33 So David Charles (‘Teleological Causation in the Physics’, in Judson, Aristotle’s Physics, 101–28 at 111–13), who thinks that the argument is radically ad hominem—i.e. against a strangely ‘flat-footed’ opponent (112 n. 15). Charles thinks that Aristotle intends the argument to have only this very modest target, and that his ‘basic argument’ in defence of teleology does not come until 199B14–18 (113 n. 16). I would prefer to see both arguments as serious defences of teleology.

Created on 4 July 2005 at 14.35 hours page 352

352

Lindsay Judson

reading is that it makes Aristotle set out to defeat far too easy an opponent. Another view is that Aristotle thinks that any good explanation of how it is that animals reproduce according to type will require teleologically laden notions of form and of its role.34 If that is right, Aristotle’s argument would not be merely ad hominem,35 but it is still unsatisfactory: the crucial challenge presented by Waterlow has not been met, but only pushed back a step, and Aristotle mentions nothing here which is relevant to meeting it.36 I wish to suggest that Aristotle has a di·erent line of attack in 2. 8. The issue for Aristotle’s opponents is not whether they can explain the regular coming to be of teeth arranged in the way that they are, and so forth; rather it is that they are committed to excluding too much from their scientific account of animals and plants. The facts that the arrangement of teeth, the presence of organs, the occurrence of processes such as respiration, and so on, serve the life of the animal in question will simply not appear in their account of the animal’s nature at all, as they take these things to be a matter of chance, and chance things do not figure in a scientific account. This reading is to be distinguished from those which see Aristotle as arguing that appealing to the relevant benefits is required in order to explain why the arrangement of teeth regularly comes to be (though of course Aristotle thinks that it is required).37 On my account Aristotle is claiming that even were the opponents able to explain without reference to ends and benefits the regular coming to be of teeth arranged in a way which is beneficial, the fact that the arrangement is beneficial simply will not figure in the opponents’ scientific account of the natures of animals—the beneficial outcome 34 See e.g. J. Lear, Aristotle: The Desire to Understand (Cambridge, 1988), 38–40. 35 Though Aristotle does have strong objections to direct specifically at the accounts of Empedocles and Democritus: see GA 1. 18, 722B8 ·., and GA 5. 8. 36 Waterlow’s own solution in e·ect makes the issue ad hominem against Empedocles once more: she sees Aristotle as meeting the challenge by claiming that Empedocles’ four-element theory (as understood by Aristotle) cannot yield per se unities (and hence a fortiori cannot explain reproduction according to type); cf. DA 2. 4, 415B28–416A8. 37 This idea can be found in Irwin, who thinks that Aristotle’s argument is ad hominem against Empedocles, and that the point is that even Empedocles’ story about the teeth really appeals to the benefit of suitably arranged teeth in explaining why most living sheep (at any moment) have them (Aristotle’s First Principles, 107). It can also be found in Sauv‹e Meyer, who sketches the argument that Empedocles is committed to invoking the benefit that suitably arranged teeth bring when he explains why we can expect future generations of sheep to have suitably arranged teeth (‘Reduction’, 811).

Created on 4 July 2005 at 14.35 hours page 353

Aristotelian Teleology

353

is, for the opponents, just a matter of chance.38 Aristotle’s appeal to the regularity of these excluded matters is meant to show that their exclusion is unacceptable. Without mention of these things our account of animals and plants will be hugely impoverished; moreover, what seems to be at the root of this impoverishment is a decision to privilege some of the regularities which naturally strike us over others in principle—because they do not involve the bestowing of benefits. This argument is not beyond challenge, of course—one of the central goals of modern biology has been to develop accounts which do not exclude these regularities while not embracing Aristotelian teleology—but it seems a plausible basis for Aristotle’s defence of teleology here: without such an account his opponents’ position seems inadequate.39 Three objections might be made at this point. The first is that the focus of the chapter is on the question whether things such as the growth or rotting of the corn and the coming to be of the arrangement of the teeth happen by chance: does this not show that Aristotle takes his opponent to deny that these things have any explanation?40 I think that this objection is misplaced. For Aristotle,41 38 To this extent I agree with Sauv‹e Meyer (‘Reduction’, 812–25) that Aristotle is confronting what he takes to be, in her terminology, an ‘eliminativist’ position (though in other respects our accounts are quite di·erent: see n. 40); but I am not sure that this use of ‘eliminativism’ is helpful. As a supposed alternative to reductionism one might naturally take eliminativism to have, or at least to threaten to have, ontological implications; but as she defines it (‘Reduction’, 822), eliminativism is just the claim that certain admitted phenomena have no (scientific) explanation, and hence is irrelevant to ontology unless one also adopts the scientistic premiss that only what figures in a scientific account of the world is a candidate for existence. 39 David Sedley has suggested that the opponents can include in their zoology the fact that the arrangement of teeth is beneficial—as part of their explanation of why the species continues to exist. That may be so, but these facts will still (to Aristotle’s way of thinking) throw no light on the natures of the organisms concerned; and of course for Aristotle it is these natures which form the primary object of zoology. 40 Sauv‹e Meyer (‘Reduction’, 798–812) adopts this line, and argues that while Aristotle’s opponents may claim that there are explanations of these things, they are in fact—from Aristotle’s point of view—committed to denying this, since they are committed to denying that things of this sort have a per se or intrinsic cause (i.e. a cause of the appropriate sort located in a single substance). One di¶culty for this interpretation is that Aristotle again gives himself a too easy opponent. The opponent seems to be a serious one because he thinks that the arrangement of teeth can be explained; but a serious opponent would take issue with the restrictive view of causation on which Aristotle is supposedly relying here precisely because on the opponent’s account the combination of ‘material’ causes does explain the outcome. Sauv‹e Meyer’s Aristotle has thus loaded the dice too heavily against his opponents here. 41 The details of the analysis which follows have an Aristotelian cast, but the

Created on 4 July 2005 at 14.35 hours page 354

354

Lindsay Judson

ascribing events or other things to τχη or τ α&τµατον (luck or chance) is not a matter of denying that the event in question has an explanation or a per se cause, but of a¶rming that between the event and the benefit which is the key constitutive element in the piece of good luck one sort of connection (incidental causation) holds while another sort of connection (per se causation) does not.42 It is this which explains a striking feature of talk about luck and chance—in Greek as in English—that events and processes which contribute to a chance result, but which themselves have unimpeachable per se causes, can also be properly ascribed to chance or luck.43 We can see this at work in Aristotle’s discussion of chance in chapters 4–6. His principal example is that of a man who goes to the market place and happens to meet someone who owes him money. Aristotle is quite happy to characterize other thinkers as prepared to say that the man went to the market by chance (το !λθε*ν π τχης ες τ3ν γορν: 196A3–5), and so too is Aristotle in his own right (‘in these circumstances he is said to go by chance’ (λ(γετα γε ττε π τχης !λθε*ν: 197A2–3)), even though both parties think that there will be a perfectly standard explanation—one which Aristotle will count as a per se cause—of the man’s going to the market.44 So when Aristotle has his opponents here say that the arrangement of teeth comes to be as a matter of chance, what they are represented as denying is that the arrangement comes to be for the sake of the relevant benefit to the animal: this leaves it open whether this is because there is no explanation of why the arrangement comes to be in the first place or, as I am suggesting, because the connection with the benefit to the animal is a non-explanatory connection. The second objection is that Aristotle’s ‘materialist opponents’ general form of the analysis reflects ordinary Greek usage—and ordinary English usage, for that matter. (There are further di·erences between τ α&τµατον as Aristotle understands it (as involving benefit or harm) and our ‘chance’, which has a wider application (for discussion see Judson, ‘Chance’, 76–82), but these do not a·ect the present point.) 42 In the case of the benefit itself, the ascription of it to luck or chance correspondingly involves, for Aristotle, the a¶rmation of a relation of incidental causation to at least one member of a particular set of other items, and a denial that the benefit is the per se ‘e·ect’ of any members of the set: see Judson, ‘Chance’, 89–98. 43 This feature is pointed out by W. D. Ross, Aristotle’s Physics: A Revised Text with Introduction and Commentary (Oxford, 1936), 516–17. 44 α6τιον τ βολεσθαι γορσαι !λθντα (196A5); κα το !λθντα κοµσασθαι τ ργριον, µ3 τοτου 1νεκα !λθντα, 8πειρα τ πλEθος· κα γρ δε*ν τιν βουλµενος κα δι)κων κα φεγων κα θεασµενος (197A15–18).

Created on 4 July 2005 at 14.35 hours page 355

Aristotelian Teleology

355

might respond that their account seems impoverished only from Aristotle’s teleological viewpoint: if we do not beg the question by assuming that viewpoint, we can see that their account works, and does so more economically than Aristotle’s own. If this objection is sound, then on my interpretation too the challenge has simply been pushed back a step.45 But on my account the opponents’ more economical account is successful (if at all) only because it rules out of court a vast range of important facts about beneficial outcomes: these facts are salient even prior to the adoption of a commitment to teleology, and it is not simply prejudging the issue to think that they should figure in a properly comprehensive and explanatory biology. The third objection is that it is a fact, just as well established and regular as those about the arrangement of teeth, etc., that rainfall (in due season) helps plants to grow: surely these ‘rain-facts’ ought to appear in our scientific account too? If this is right, does my argument not require Aristotle to say that the rainfall is for the sake of the plants? The answer is that these ‘rain-facts’ will figure in biology at another point, namely when biology talks about the teleological adaptation of the plants to their environment.46 So the facts about the rain’s beneficial e·ect do appear in biology without his having to see the rain as for the sake of the crops. Obviously no such move is available in the case of the teeth: apart from the animals whose teeth they are, there simply are no other teleologically ordered substances whose ends can be appealed to in the way that plants can be appealed to in explaining why it is that rain is regularly beneficial.

45 David Charles suggested this objection. 46 For examples of adaptation in this sense see e.g. PA 3. 14, 674A22–B5 (the camel has a hard palate and several stomachs because its food is thorny and woody), and 4. 12, 694B12–17 (some birds have long legs and long toes because they live in marshes and have to walk across boggy ground, and ‘nature makes the organs to fit the function, not the function to fit the organs’); for the more complex cases of seasonal migration, hibernation, and aestivation see HA 8. 12–17 (596B20 ·.). There is an important di·erence between the present suggestion about the rain case and that of Aristotle’s opponent’s citing ‘teeth facts’ to explaining why the species continues to exist (see n. 39 above). For the opponent, as Aristotle sees it, the citing of the teeth facts floats entirely free of the animal’s nature; Aristotle’s own citing of ‘rain facts’ as part of the account of how an organism is well-adapted to its environment, by contrast, is firmly anchored in the nature of the organism concerned.

Created on 4 July 2005 at 14.35 hours page 356

356

Lindsay Judson III

I would like now to look beyond Physics 2. 8 to four other passages. Three of these have been used to support biocentrism and/or anthropocentrism, while one (which I shall deal with third) has been used to attack them. The first passage is a famous one from the Politics: <στε =µοως δEλον :τι κα γενοµ(νοις οητ(ον τ τε φυτ τ#ν ζG)ων 1νεκεν ε2ναι κα τ 8λλα ζG#α τ#ν νθρ)πων χριν, τ µ ν Hµερα κα δι τ3ν χρEσιν κα δι τ3ν τροφ5ν, τ#ν δ$ γρων, ε µ3 πντα, λλ τ γε πλε*στα τEς τροφEς κα 8λλης βοηθεας 1νεκεν, Iνα κα !σθ3ς κα 8λλα Jργανα γνηται !ξ α&τ#ν. ε ο0ν K φσις µηθ ν µ5τε τελ ς ποιε* µ5τε µτην, ναγκα*ον τ#ν νθρ)πων 1νεκεν α&τ πντα πεποιηκ(ναι τ3ν φσιν. (Pol. 1. 8, 1256B15–22) Hence it is clear that in the same way we should think that, in things which have come to be too, plants exist for the sake of animals, and that the other animals exist for the sake of human beings—domesticated ones both for their usefulness and for food, and of wild animals, if not all, most, at any rate, for food and other help—so that there might be clothing and other tools from them. If, then, nature makes nothing either incomplete or pointless, it is necessary that nature has made all of these things for the sake of human beings.

This is, on the face of it, a firm commitment, not only to the view that plants and other animals are for the sake of something beyond themselves, but also to the view that they are for the sake of human beings—the anthropocentric view for which Sedley argues.47 But I do not think that the passage can be relied on as evidence for this view: Aristotle is pretty clearly making these claims from the point of view of household managers and statesmen (το*ς οκονµοις κα το*ς πολιτικο*ς, 1256B38): this opens up the possibility that he takes these ‘for the sake of’ claims to be true from that viewpoint. Given the naturalness of the household and the polis, Aristotle may think that it is entirely in order, when looking at the world from the standpoint of a household manager or statesman, whose proper business it is to make certain provisions for human life, to take certain sources of food or of building materials as genuinely (even naturally) for the sake of human diet or human shelter, and to argue 47 Cf. Sedley, ‘Aristotle’s Teleology’, 180–1. Note that for this reason this passage presents a di¶culty for plain biocentrists such as Cooper as well as for opponents of the idea that rainfall is for the sake of the crops.

Created on 4 July 2005 at 14.35 hours page 357

Aristotelian Teleology

357

to this from a parallel about nature’s provision of nourishment for animals’ o·spring. Nothing follows, however, about the status of these things from the standpoint of biology—or of Aristotelian first philosophy. In a similar way, I would argue, when Aristotle says in the De interpretatione that the past is (absolutely) necessary, he is speaking from the agent’s point of view. What is past cannot from the agent’s point of view, deliberating about what to do now, be otherwise—she cannot change it—however contingent it is from the point of view of natural science, or even of metaphysics.48 Again, one could imagine someone with a broadly Democritean view of colour, as not any sort of part of what there is in reality, none the less thinking it legitimate—not as a fac«on de parler, and not as a matter for a dispositional analysis—to say, when speaking from our perspective, that grass is (really) green. From certain other points of view (e.g. that of dogs), this would not be true; nor would it figure at all, according to the Democritean, in correct scientific or metaphysical accounts of the world. These cases are meant to illustrate the general idea of a judgement’s being legitimate from a certain, non-arbitrary point of view: it is not part of my claim that they are exactly parallel to each other or to the type of claim I detect in the Politics passage. All that we need to disarm the Politics passage for our present purposes is that this sort of reading is a plausible option; which reading we actually adopt would then be determined by our view of Aristotle’s natural teleology, and not the other way round. However, there is some reason internal to the passage to prefer my reading of it. If Aristotle were starting from the assumption of an anthropocentric teleology, we should expect him to think that the whole of nature is for the sake of human beings. Seen in that light, his shuffling around the question of how many of the wild animals are for our sake—‘and of wild animals, if not all, most, at any rate, for food and other help’—would be at best unfortunate, and at worst inconsistent. If, as Sedley puts it, ‘All sublunary nature is organized in a cooperative hierarchy, in which the lower benefits the higher’,49 why are there animals which are not for the sake of humans—which, indeed, seek to destroy us? Of course, explanations could be devised; 48 See Judson, ‘La bataille navale d’aujourd’hui: De Interpretatione 9’, Revue de philosophie ancienne, 6 (1988), 5–37 at 27–8. Of course, one does not need to accept my claim about the sea battle passage in order to accept my claim about the Politics passage. 49 ‘Aristotle’s Teleology’, 196.

Created on 4 July 2005 at 14.35 hours page 358

358

Lindsay Judson

but it is much easier to see Aristotle’s words here as instead reflecting the viewpoint of the household manager or statesman: he is simply considering how many animals can be used for human ends, and reacting to the thought that some animals are, for example, too fierce to be so used.50 I think that this interpretation can also be used to explain the second passage, Phys. 2. 2, 194A33–6:51 !πε κα ποιο σιν α7 τ(χναι τ3ν λην α7 µ ν Lπλ#ς α7 δ ε&εργν, κα χρ)µεθα ς Kµ#ν 1νεκα πντων +παρχντων· !σµ ν γρ πως κα Kµε*ς τ(λος (διχ#ς γρ τ οM 1νεκα· ε6ρηται δ$ !ν το*ς περ φιλοσοφας). For the crafts too make their matter—some make it without qualification, others make it suitable to work with—and we use everything as being for our sake; for we too are in a way an end (for for the sake of something is said in two ways: this has been explained in On Philosophy).52

One could avoid any suggestion of anthropocentrism by taking the ς at a34 and the πως in the following line to indicate subversion: we use things as if they are for our sake, we are an end in a fashion.53 But if we take ς and πως in a non-subversive way, there still need be no 50 Of course the normative conditions of a natural household or πλις may rule out the use of certain other animals which could otherwise be put to use: these conditions may, for instance, rule out cannibalism. So in a di·erent way these animals will not be for the sake of human life either. But this is clearly not what is in Aristotle’s mind here. Other commentators are more dismissive of this passage than I think we need be. Nussbaum invokes the idea that ‘it assumes an anthropocentric vantage point and asks what use various parts of the natural world are to man in his e·orts to establish himself in the world’, but still dismisses the result as ‘a preliminary phainomenon, from the human-practical viewpoint, not a serious theoretical statement’ (M. C. Nussbaum, Aristotle’s De Motu Animalium: Text with Translation, Commentary, and Interpretive Essays [De Motu Animalium] (Princeton, 1978), 96; cf. Pellegrin, ‘Les ruses’, 321). Balme thinks that ‘Aristotle means only that A is for the sake of B in the sense that B cannot happen without A’ (D. M. Balme, Aristotle’s De Partibus Animalium I and De Generatione Animalium I (with Passages from II. 1–3): Translated with notes [De Partibus Animalium] (Oxford, 1972), 96). Wardy simply embraces as unsurprising the idea that ‘Aristotle’s political doctrine must be badly out of kilter with his natural philosophy’ (‘Aristotelian Rainfall’, 22–3). 51 I am grateful to David Sedley for drawing my attention to this passage. (My punctuation is di·erent from Ross’s: I start the parenthesis after τ(λος rather than after !σµ(ν. But nothing hinges on this di·erence—if anything, my punctuation is more favourable to Sedley’s reading of the passage than Ross’s is.) 52 Presumably these two ways are those distinguished in DA 1. 4, 415B20–1, and Metaph. Λ 7, 1071B2–3: if something is said to be for the sake of X, this can be because X is the goal or because it is the beneficiary of a teleological process. 53 It might be argued that the next clause militates against this, since it introduces two straightforward and un-Pickwickian ways of being an end (so Sedley, ‘Aristotle’s Teleology’, 189). But it might be that the connection between the two clauses is

Created on 4 July 2005 at 14.35 hours page 359

Aristotelian Teleology

359

commitment to anthropocentrism, since Aristotle may once again be speaking from a certain naturally demarcated point of view—in this case that of the craftsman.54 The third passage occurs at the end of Phys. 2. 7: κα πντως ποδοτ(ον τ δι τ, ο4ον :τι !κ το δε νγκη τδε (τ δ !κ το δε  Lπλ#ς  ς !π τ πολ), κα ε µ(λλει τοδ σεσθαι (<σπερ !κ τ#ν προτσεων τ συµπ(ρασµα), κα :τι το τ$ Nν τ τ Nν ε2ναι, κα διτι β(λτιον οτως, ο&χ Lπλ#ς, λλ τ πρς τ3ν Oκστου ο&σαν. (198B5–9)55 We must explain the why in every way, namely, that from this necessarily that (‘from this’ either without qualification or for the most part); that this must be so if that is to be so (as the conclusion from the premisses); that this was the essence; and because it is better thus—not without qualification, but with reference to the essence of each thing.

This is generally thought of as Aristotle’s ‘basic teleological axiom’56—and with good reason, since Aristotle observes it almost universally throughout his biological works (I return to this point below). Some commentators take the ‘axiom’ as decisive evidence against teleological biocentrism, for it appears to rule out what we might call extra-species teleology—i.e. teleological explanation of a feature of or behaviour by X which appeals to the good of members of another species. David Balme is a typical example: ‘This [i.e. the axiom] cannot be reconciled with a doctrine that the function of grass is to be eaten by sheep, and of sheep to be eaten by men.’57 Furley tries to avoid this conclusion by reading the axiom as requiring merely that we do more than claim that the final cause is good—we must say what thing it is good for.58 In a similar vein Sedley holds that Aristotle means only to rule out ‘the idea of an intrinsic cosmic good which is not anybody’s good at all’.59 This interpretation involves taking ‘with reference to the essence of each thing’ to mean something like ‘with reference to the essence of each thing whose benefit is part of the teleological explanation’. This weaker, and the idea of the end as the beneficiary is being introduced because the loose way in which we are an end is that we are beneficiaries who none the less are not that for the sake of which the matter exists, precisely because there is no teleological connection of the appropriate sort at work in this case. 54 For a di·erent non-anthropocentrist reading of the passage, see Wardy, ‘Rainfall’, 27–9. 55 Cf. GA 5. 8, 789B2–5; IA 2, 704B12–18. 56 J. G. Lennox, Aristotle: On the Parts of Animals [On the Parts of Animals] (Oxford, 2001), 341. 57 De Partibus Animalium, 96. 58 ‘The Rainfall Example’, 181–2. 59 ‘Aristotle’s Teleology’, 190.

Created on 4 July 2005 at 14.35 hours page 360

360

Lindsay Judson

reading is not impossible, but it is much more natural to understand the axiom as meaning that when giving the final cause of X’s being F we must specify how being F is good for X. Had Aristotle intended Furley’s and Sedley’s weaker reading, we would have expected him to say ‘not without qualification, but with reference to the essence of something’, rather than ‘with reference to the essence of each thing’. Moreover, Aristotle’s practice in the biological works is (with one exception which I shall discuss below) in conformity with the stronger principle, that teleological explanation of X’s being F must appeal to the good of X. Now this is not quite the end of the matter, since the axiom still leaves open the possibility that X’s being F might be for the sake of a di·erent species, Y, providing that benefiting Y in the relevant way is itself good for X.60 In other words, we need to distinguish between extra-species teleology which is nonconformist—in which the appeal to the good of the other species does not lead back to the good of X, and so the axiom is violated—and the conformist sort in which the appeal to the good of something other than species X is an intermediate step in the ultimate appeal to X’s good, and so conforms to the axiom. The possibility of such conformist extraspecies teleology o·ers little comfort to the plain biocentrist, however. I can think of no passage in which Aristotle suggests that—in general—X’s benefiting other species also benefits X, and it seems very implausible that this should be true.61 And even if the rainfall does have a final cause—to reach its natural place—it is fantastical to suppose that its beneficial e·ect on plants is good for the rain too. Nor, I think, is it any comfort to the anthropocentrist. Sedley appeals to it in this way: Each being serves both its own ends and—to put it very crudely—those of the next link in the food chain. Rain per se may fall in order to return to its natural place; but rain falls where and when it regularly does in order to make plants grow. Grass per se grows in order to flourish as grass; but grass 60 It also leaves open the possibility that X’s being F might be for the sake of the order and beauty of the cosmos as a whole, providing that contributing to the order and beauty of the cosmos in the relevant way is itself good for X . This, indeed, is how I think we ought to construe Aristotle’s various claims about what we might call ‘cosmic teleology’—about animals’ imitation of ‘the everlasting and divine’ in DA 2. 4, 415A26–B7 (cf. GC 2. 10, 336B25–32; GA 2. 1, 731B18–732A12), the motions of the heavenly bodies in Metaph. Λ 8, and the good of ‘the nature of the whole’ (K το :λου φσις) in Λ 10. But this is a story for another day. 61 Cf. Sedley, ‘Aristotle’s Teleology’, 192.

Created on 4 July 2005 at 14.35 hours page 361

Aristotelian Teleology

361

grows where and when it regularly does in order to feed animals. A deer per se grows in order to become a mature deer; but deer grow where and when they regularly do in order to feed man. Thus all the good ends served are those of individual species, but in each case the ends of the higher species play an orchestrating role.62

There seem to be two di·erent lines of thought here. One is that the anthropocentric scheme satisfies the teleological axiom because all the goods served are those of some species or other; but this relies on the weak reading of the teleological axiom as requiring only that the good cited be good for something. The other line of thought relies on a very sharp distinction between what rain/plants/deer do (to be explained in terms of their own good) and the ‘orchestrating’ of when and where they regularly do it, to be thought of as not something those species themselves do. What Sedley goes on to say about ‘global nature—the nature of the entire ecosystem, so to speak’63 suggests that this is what he principally has in mind. It is not, he says, the nature of the rain which is exhibited in the ‘providential [sic] winter rainfall’,64 but the nature of the world as a whole, construed as analogous to a household or polis.65 But now two problems emerge. First, we might well doubt that there is or can be such a thing, for Aristotle, as a global nature which is the subject of these orchestrating activities. Second, once again only the weak reading of the teleological axiom is satisfied. Perhaps we should suppose that the axiom applies only to species, and thus that global nature is beyond its scope: but we would still expect the good at which global nature aims to consist in the beauty and order of the totality of the natural world, and not in the activities of one of its species—even the highest one.66 It now looks as if, even if we granted that the rain fell for the sake of the growth of plants, the only evidence for anthropocentrism rather than some form of cosmic teleology or (mere) biocentrism would be the Politics passage. 62 Ibid. 190–1. 63 Ibid. 192 ·. 64 Ibid. 192. 65 Sedley develops this idea in ‘Metaphysics Λ 10’, in D. Charles and M. Frede (eds.), Aristotle’s Metaphysics Λ: The Proceedings of the Fourteenth Symposium Aristotelicum (Oxford, 2000), 327–50; the wider issues he raises cannot be dealt with here (cf. n. 60 above). 66 Of course, on such a picture, it could be part of the proper functioning of global nature to provide properly for the life of human beings; but (1) the benefit to the latter would still be subordinate to the goodness of global nature; (2) the idea that the analogous provision which global nature would provide for other species would primarily be for the promotion of human life is again unmotivated.

Created on 4 July 2005 at 14.35 hours page 362

362

Lindsay Judson

To sum up. The ‘axiom’ of 2. 7 does not altogether rule out interspecies teleology, as Balme and others have thought; but (unless we read it in the weak way advocated by Furley and Sedley) it is none the less a serious problem both for plain biocentrism and for Sedley’s anthropocentrism. Like Physics. 2. 8, the fourth passage also concerns a placement of teeth: χει δ κα περ τ στµα διαφορς. τ µ ν γρ κατ$ ντικρ? χει τ στµα κα ες τ πρσθεν, τ δ$ !ν το*ς +πτοις, ο4ον οI τε δελφ*νες κα τ σελαχ)δη· κα πτια στρεφµενα λαµβνει τ3ν τροφ5ν. φανεται δ$ K φσις ο& µνον σωτηρας 1νεκεν ποιEσαι το το τ#ν 8λλων ζG)ων (!ν γρ τAE στρ(ψει σG)ζεται τQλλα βραδυνντων· πντα γρ τ τοια τα ζGωοφγα !στν), λλ κα πρς τ µ3 κολουθε*ν τAE λαιµαργRα τAE περ τ3ν τροφ5ν· SRTον γρ λαµβνοντα διεφθερετ$ /ν δι τ3ν πλ5ρωσιν ταχ(ως. πρς δ τοτοις περιφερE κα λεπτ3ν χοντα τ3ν το Sγχους φσιν ο&χ ο4ν τ$ ε&διαρετον χειν. (PA 4. 13, 696B23–34)67 Fish also di·er with respect to the mouth. In some the mouth is placed straight across and towards the front; but in others it is on the underside— e.g. in the dolphins and selachians [i.e. sharks]—and they turn belly up to seize their nourishment. And nature appears to do this not only for the sake of the preservation of the other animals (for during the turn the other animals escape, because of the delay—for all such creatures are carnivorous), but also in order that they do not follow their gluttonous ways in relation to nourishment; for if they could grasp it easily, they would be destroyed owing to being quickly filled up. And in addition to these reasons, the nature of the snout, being curved and narrow, is unable to have much of an opening in it.

This passage violates the teleological axiom of Physics 2. 7: it invokes a benefit to a di·erent species (in fact to a wide range of other species) as a final cause of the location of the mouth in sharks and dolphins—apparently a case of what I have called nonconformist extra-species teleology. The passage is of no direct help to the an67 With this we should compare HA 8. 2, 591B23–30: το*ς µ ν ο0ν 8λλοις χθσιν K θ5ρα τ#ν Kττνων καταντικρ? γνεται το*ς στµασιν, :νπερ πεφκασι τρπον νε*ν· ο7 δ σελαχ)δεις κα ο7 δελφ*νες κα πντες ο7 κητ)δεις πτιοι ναππτοντες λαµβνουσιν· κτω γρ τ στµα χουσιν. δι σ)ζονται µTλλον ο7 !λττους· ε δ µ5, πµπαν /ν Uλγοι δοκο σιν ε2ναι· κα γρ K το δελφ*νος Uξτης κα δναµις το φαγε*ν δοκε* ε2ναι θαυµαστ5 (‘The other fish catch the smaller ones in their mouths straight ahead, in their natural swimming position. But the selachians and dolphins and all cetaceans [i.e. whales, etc.] turn over on their backs to catch them; for their mouth is below. For this reason the smaller fish tend to survive more. Otherwise it is thought that there would be altogether few; for the dolphin’s quickness and capacity for eating seem to be astonishing’).

Created on 4 July 2005 at 14.35 hours page 363

Aristotelian Teleology

363

thropomorphic position, however, for two reasons. First, there is no suggestion here of any subordination of the τ(λος, or end, of the sharks or dolphins to that of the smaller fish—still less of any connection with the τ(λος of human beings. Second, this case does not exemplify the general pattern of ‘orchestration’ suggested by Sedley, since it is one and the same feature which is good for the sharks themselves and for the smaller fish.68 Should we none the less take the passage as showing that Aristotle does not really hold to the teleological axiom? It is hard to think that it does show this, given that it is unique: nowhere else in Aristotle’s biological works does he appeal to the good of a di·erent species when giving a final cause. It would be di¶cult to accept that Aristotle thought that nonconformist extra-species teleology was indeed legitimate in biology, and yet throughout his vast account of animal species he found no case in which to apply it—nor even any case in which it was worth raising the question whether it was applicable—except this one. As Lennox says, Aristotle’s practice in zoology ‘is presumptive evidence against Sedley’s thesis’.69 It is much easier, therefore, to ‘defuse’ the sharks passage in some way or other and to leave the teleological axiom intact. Commentators have made a number of suggestions as to how the passage might be played down;70 I 68 See pp. 360–1 above. Of course it is consistent with Sedley’s account that there should be teleological overdetermination of a single feature; my point is that the PA passage does not give any support to the further idea that extra-species teleology is (usually) a matter of cosmic nature arranging distinct properties or patterns of behaviour not fixed by the nature of the species. 69 ‘Material and Formal Natures in Aristotle’s De Partibus Animalium’, in Lennox, Aristotle’s Philosophy of Biology: Studies in the Origin of Life Science (Cambridge, 2001), 182–204 at 201 n. 5. Sedley replies that the zoological works are concerned only with the animal’s nature, which is centred on the realization of its own form (‘Aristotle’s Teleology’, 195). But (1) this commits Sedley to finding the shark passage as out of place as his opponents do; (2) it still seems more than extraordinary that, though Aristotle is supposed to think of cosmic nature as orchestrating where and when animals behave in many ways for the sake of human life—and often, subordinately, for the sake of other species which in turn serve human needs—this is never relevant to zoology. 70 In his Clarendon commentary, Balme suggests that Aristotle means only the weak claim that the location of the mouth is necessary for the survival of the small fish (De Partibus Animalium, 96; cf. Nussbaum, De Motu Animalium, 95: Aristotle is careless, and means only that the survival of the smaller fish is a result, not an end; this also seems to be Pellegrin’s view (‘Les ruses’, 311)). Elsewhere Balme says that the passage is sarcastic, and aimed at ‘popular providentialists like Xenophon’ (‘Teleology and Necessity’, in A. Gotthelf and J. G. Lennox (eds.), Philosophical Issues in Aristotle’s Biology (Cambridge, 1987), 275–85 at 278–9; cf. ‘Aristotle’s Biology Was Not Essentialist’, in Gotthelf and Lennox, 291–312 at 299). Lennox

Created on 4 July 2005 at 14.35 hours page 364

364

Lindsay Judson

have two new ones. (1) Aristotle has been given the extra-species claim by his informants (or has inferred that it must be what they think), and takes it seriously as an endoxon or common opinion; but he does not really accept it, and the ‘but also’ clause represents his own analysis.71 (2) Aristotle thinks that the preservation of the smaller fish actually benefits the sharks and dolphins too: if they ate all the small fish today, there would be none left for them to eat tomorrow. This would bring the passage back into conformity with the teleological axiom. Aristotle does not, of course, say this; but the thought behind it may be implicit in the remark about the sharks’ and dolphins’ ‘gluttonous ways’. The sharks passage, in any case, as Hume says about his missing shade of blue, ‘is so singular, that it is scarcely worth observing, and does not merit that for it alone we should alter our general maxim’.72 I conclude that there is little reason to accept, and a great deal of reason to reject, the view that the rain falls to make the crops grow: when the farmer sees the November rain, what he can expect in the spring is simply the growth of plants well adapted to their environment, not an instance of biocentrist teleology. Christ Church, Oxford

B I B L I O GR A P HY Balme, D. M., Aristotle’s De Partibus Animalium I and De Generatione Animalium I (with Passages from II. 1–3): Translated with Notes [De Partibus Animalium] (Oxford, 1972). ‘Teleology and Necessity’, in Gotthelf and Lennox, Philosophical Issues in Aristotle’s Biology, 275–85. ‘Aristotle’s Biology was Not Essentialist’, in Gotthelf and Lennox, Philosophical Issues in Aristotle’s Biology, 291–312. (On the Parts of Animals, 341) says that the problem would have been defused if Aristotle had deployed the distinction we met in Physics 2. 2 between the beneficiary of a feature and the τ(λος (see above, p. 358 and n. 52): but using this distinction would not by itself solve the problem, as the axiom requires the beneficiary in this sense to be the possessor of the feature. 71 Note the use of φανεται (‘nature appears to do this . . .’) at b27, and compare the use of δοκο σιν in the HA passage (see n. 67 above)—though of course the φανεται introduces both final causes, and not just the extra-species one. 72 David Hume, Enquiry concerning Human Understanding, ed. L. A. Selby-Bigge and P. H. Nidditch (Oxford, 1975), 21.

Created on 4 July 2005 at 14.35 hours page 365

Aristotelian Teleology

365

Charles, D., ‘Teleological Causation in the Physics’, in Judson, Aristotle’s Physics, 101–28. Cooper, J., ‘Aristotle on Natural Teleology’ [‘Natural Teleology’], in M. C. Nussbaum and M. Schofield (eds.), Language and Logos: Studies in Ancient Philosophy Presented to G. E. L. Owen (Cambridge, 1982), 197–222. Furley, D., ‘The Rainfall Example in Physics ii. 8’ [‘The Rainfall Example’], in A. Gotthelf (ed.), Aristotle on Nature and Living Things (Pittsburgh and Bristol, 1985), 177–82. Gotthelf, A., and Lennox, J. G. (eds.), Philosophical Issues in Aristotle’s Biology (Cambridge, 1987). Hume, David, Enquiry concerning Human Understanding, ed. L. A. SelbyBigge and P. H. Nidditch (Oxford, 1975). Irwin, T. H., Aristotle’s First Principles (Oxford, 1988). Judson, L., ‘La bataille navale d’aujourd’hui: De Interpretatione 9’, Revue de philosophie ancienne, 6 (1988), 5–37. Aristotle’s Physics: A Collection of Essays (Oxford, 1991). ‘Chance and “Always or for the Most Part” in Aristotle’ [‘Chance’], in Judson, Aristotle’s Physics, 73–99. Lear, J., Aristotle: The Desire to Understand (Cambridge, 1988). Lennox, J. G., Aristotle: On the Parts of Animals [On the Parts of Animals] (Oxford, 2001). ‘Material and Formal Natures in Aristotle’s De Partibus Animalium’, in id., Aristotle’s Philosophy of Biology: Studies in the Origin of Life Science (Cambridge, 2001), 182–204. Nussbaum, M. C., Aristotle’s De Motu Animalium: Text with Translation, Commentary, and Interpretive Essays [De Motu Animalium] (Princeton, 1978). Osborne, R., Classical Landscape with Figures: The Ancient Greek City and its Countryside (London, 1987). Pellegrin, P., ‘Les ruses de la nature et l’‹eternit‹e du mouvement’ [‘Les ruses’], in M. Canto-Sperber and P. Pellegrin (eds.), Le Style de la pens‹ee: recueil de textes en hommage a› Jacques Brunschwig (Paris, 2002), 296–323. Ross, W. D., Aristotle’s Physics: A Revised Text with Introduction and Commentary (Oxford, 1936). Sauv‹e Meyer, S., ‘Aristotle, Teleology and Reduction’ [‘Reduction’], Philosophical Review, 101 (1992), 791–825. Sedley, D., ‘Is Aristotle’s Teleology Anthropocentric?’ [‘Aristotle’s Teleology’], Phronesis, 36 (1991), 179–96. ‘Metaphysics Λ 10’, in D. Charles and M. Frede (eds.), Aristotle’s Metaphysics Λ: The Proceedings of the Fourteenth Symposium Aristotelicum (Oxford, 2000), 327–50.

Created on 4 July 2005 at 14.35 hours page 366

366

Lindsay Judson

Wardy, R., ‘Aristotelian Rainfall or the Lore of Averages’ [‘Aristotelian Rainfall’], Phronesis, 38 (1993), 18–30. Waterlow, S., Nature, Change, and Agency in Aristotle’s Physics: A Philosophical Study [Nature, Change, and Agency] (Oxford, 1982). Witt, C., Substance and Essence in Aristotle: An Interpretation of Metaphysics VII–IX (Ithaca, NY, and London, 1989).

Created on 4 July 2005 at 14.35 hours page 367

ON THE US E OF STOI CHEI ON IN THE S ENS E OF ‘ ELEMENT’ TIMOTHY J. CROWLEY

     says that Empedocles is the first to name fire, air, water, and earth as the four kinds of stoicheia, or ‘elements’ (Metaph. Α 4, 985A32; cf. 984A8). But it is well known that Empedocles does not call fire, air, water, and earth stoicheia; rather, he calls them the ‘roots of all things’ (πντων ιζ µατα) (31 B 6 DK).1 The use of the term stoicheion in the sense of ‘element’ or ‘principle of nature’ is usually believed to be a later innovation. How much later is a matter of some dispute. In a fragment preserved by Simplicius, Aristotle’s pupil Eudemus of Rhodes seems to identify Plato as the first to call the ‘elementary principles of natural and generated things’ stoicheia.2 Diels, in his study of the development of the term ‘element’ in Graeco-Roman philosophy, reviews the evidence for Eudemus’ claim and concludes that before Plato nobody had used the term stoicheion with reference to ‘the physical principles’.3 Some commentators, however, have challenged this conclusion. A common view is that the atomists were the first to use the term for the principles of things; Diels himself acknowledges this possibility, and it is repeated by Burnet.4 But Burnet also thinks it ã Timothy J. Crowley 2005 The research of which this paper is a partial product was made possible by a Jacobsen Fellowship from the Royal Institute of Philosophy, and a Scatcherd European Scholarship from the University of Oxford. For these awards I am very grateful. I would also like to thank Michael Frede and David Sedley for their helpful comments on earlier drafts of this paper. 1 Empedocles gives mythological names to the roots in fragment 6, but names them as fire, earth, air, and water at 31 B 17. 18 DK. 2 Simpl. In Phys., proem. 7. 10–17 Diels; Eudemus fr. 31, in F. Wehrli, Eudemos von Rhodos, 2nd edn. (Die Schule des Aristoteles, 8; Basel, 1969). Cf. D.L. 3. 24. 3 H. Diels, Elementum (Leipzig, 1899), 17. See also W. Burkert, ‘Stoicheion’, Philologus, 103 (1959), 167–97 at 174–6; cf. W. Schwabe, ‘“Mischung” und “Element” im Griechischen bis Platon’, Archiv f•ur Begri·sgeschichte, suppl. 3 (1980), 147–9. 4 Diels, Elementum, 13 n. 1; J. Burnet, Early Greek Philosophy [EGP], 4th edn.

Created on 4 July 2005 at 14.35 hours page 368

368

Timothy J. Crowley

plausible that Plato may have taken this use of the term from the Pythagoreans;5 indeed, according to Ryle, Sextus Empiricus identifies stoicheion, in the sense of an ultimate material element, as a term of Pythagorean origin.6 A more speculative suggestion is o·ered by Lagercrantz, who imagines that some anonymous Athenian master teacher introduced the term for the specific purpose of explaining Empedocles’ doctrine of the four ‘roots’; thus Empedocles’ four ιζ µατα become the four stoicheia.7 But if there is disagreement about who is to be credited with the first use of stoicheion in the sense of ‘element’, there tends to be a general consensus as to the reason why this particular term came to be used in this sense. For whether one thinks that Plato, or the atomists, or anyone else, is responsible, it is widely assumed that this use of stoicheion is metaphorically derived from some other, more familiar, use of the term. In particular, it is usually believed that stoicheion primarily means ‘letter of the alphabet’, and that, by comparing the principles of nature and natural things to the letters that constitute a word, the former also come to be called stoicheia.8 In what follows I examine what we might call the ‘internal’ evidence for the view that Plato is the first to use stoicheion in the sense of ‘element’. By the internal evidence I mean the evidence that is available in Plato’s dialogues. The first time Plato uses stoicheion in this sense is generally agreed to be at Theaet. 201 e;9 but he also uses the term in the Timaeus and in the Sophist. I argue that the (London, 1930), 336 n. 4. See also C. H. Kahn, Anaximander and the Origins of Greek Cosmology [Anaximander] (New York, 1960), 120; I. During, Aristotle’s • Protrepticus: An Attempt at Reconstruction [Protrepticus] (Studia Graeca et Latina Gothoborgensia, 12; G•oteborg, 1961), 202; G. Vlastos, Plato’s Universe (Oxford, 1975), 67 n. 4. Cf. J. B. Skemp, Plato’s Statesman: A Translation of the Politicus of Plato with Introductory Essays and Footnotes (London, 1952) 161 n. 1. 5 Burnet, EGP, 228 n. 1. 6 G. Ryle, ‘Letters and Syllables in Plato’, Philosophical Review, 69 (1960), 431– 51 at 431. The reference, not given by Ryle, must be to M. 10. 249–50 ( = Against the Physicists 2. 249–50). 7 O. Lagercrantz, Elementum (Uppsala, 1911), 17–18. 8 Kahn writes: ‘In Greek, as afterwards in Latin, this expression [στοιχεον] is based on a comparison of the physical principles to the letters of the alphabet (the primary meaning of στοιχεα)’ (Anaximander, 120). Cf. Diels, Elementum, 13 n. 1, and 19; Burkert, ‘Stoicheion’, 175; Burnet, EGP, 336 n. 4; W. Charlton, Aristotle’s Physics Books I and II, Translated with an Introduction and Notes (Oxford, 1970), 46. 9 See F. M. Cornford, Plato’s Theory of Knowledge [PTK] (London, 1935), on Theaet. 201 e: ‘This is said to be the first occurrence [of στοιχεα] as applied to the elements of physical things’ (143). See also A. E. Taylor, A Commentary on Plato’s Timaeus [Timaeus] (Oxford, 1928), 306.

Created on 4 July 2005 at 14.35 hours page 369

On the Use of Stoicheion in the Sense of ‘Element’

369

relevant passages in these dialogues most naturally indicate that Plato is appealing to an already current usage of stoicheion, rather than introducing a novel sense of the term. Hence I am broadly in agreement with those commentators who have urged that the use of stoicheion in the sense of ‘element’ or ‘constituent of natural things’ pre-dates Plato. But what distinguishes my argument from similar views regarding the use of stoicheion is that I have no interest in trying to identify any of Plato’s predecessors as sources of this sense of the term. Partly this is because I do not believe there is su¶cient evidence to substantiate any such identification.10 More importantly, however, it seems to me that the very attempt to specify an ‘introduction’ of the use of stoicheion in the sense of ‘element’ is misguided. For what I also want to argue is that this use of stoicheion is a standard, or ‘ordinary’, usage of the term. In other words, I reject the common assumption that the use of the term stoicheion in physical, metaphysical, or more generally cosmological contexts, in the general sense of principle of body, is a metaphorical derivation from some other use of stoicheion. What of the Eudemus fragment? I do not think its evidence can be ignored,11 or dismissed as mistaken.12 After all, Eudemus 10 Cf. M. Burnyeat, The Theaetetus of Plato [Theaetetus] (Indianapolis, 1990), 152: ‘for a number of pre-Socratic theorists . . . the word “element” was coming to be a vogue term’. But the only evidence that Burnyeat can muster to support this claim is Plato’s use of stoicheion in the Sophist and Timaeus. The evidence for an atomistic provenance of this use of stoicheion is based upon some remarks of Aristotle’s, at Metaph. 985B15 and GC 315B6–15 (cf. De caelo 1. 7, 275B31–276A1). But, remarkably, in neither passage does Aristotle actually use the term stoicheion. For a critique of the view that the atomists employed the term stoicheia in the sense of the elements of all things, see Taylor, Timaeus, 307. The only evidence that stoicheion is of Pythagorean origin is that provided by Sextus Empiricus (see n. 6). According to Sextus, the Pythagoreans say that those who are doing genuine philosophy are like those who examine language: as the latter begin by investigating the stoicheia, because words are composed of syllables, and syllables of letters, so the true physicist ought firstly to enquire into the stoicheia of the universe. This, it must be said, appears more concerned with an issue of methodology than with anything else. 11 Some commentators appear content to ignore Eudemus’ claim as evidence for the first philosophical use of the term stoicheion. For instance, Vollgra·, Koller, and Lohmann, in their detailed studies of the term, pay very little attention to the fragment; see W. Vollgra·, ‘Elementum’, Mnemosyne2, 4 (1949), 89–115; H. Koller, ‘Stoicheion’, Glotta, 34 (1955), 161–74; J. Lohmann, Musik^e und Logos: Aufs•atze zur griechischen Philosophie und Musiktheorie (Stuttgart, 1970). Cf. also A. Lumpe, ‘Der Begri· “Element” im Altertum’, in Archiv f•ur Begri·sgeschichte, 7 (1962), 285–93. Similarly, recent English-language studies of the Theaetetus make little or no reference to Eudemus; see e.g. J. McDowell, Plato’s Theaetetus [Theaetetus] (Oxford, 1973); D. Bostock, Plato’s Theaetetus [Theaetetus] (Oxford, 1988); Burnyeat, [See p. 370 for n. 11 cont. and n. 12

Created on 4 July 2005 at 14.35 hours page 370

370

Timothy J. Crowley

is usually considered to be a reliable source. But it seems to me that the significance of the fragment ought to be judged in the light of the internal evidence, rather than vice versa. Hence I have little to say about Eudemus in this paper; my focus is, in the main, restricted to a consideration of the use of stoicheion, in the sense in question, in Plato’s dialogues. I begin, however, not with Plato but with Aristotle, and his account in the Metaphysics of the meaning of the term stoicheion. This seems to be the best way to start, because here Aristotle lays down quite clearly the various meanings of the term, as well as noting the further possibilities of its use.

1. Aristotle on the meaning of stoicheion Aristotle o·ers the following account of the meaning of the term stoicheion in chapter 3 of Metaphysics Delta, a book whose contents he elsewhere refers to as ‘the discussion of the number of ways in which things are called what they are’ (Metaph. Θ 1, 1052A15; Ε 4, 1028A4). It is useful at this point to consider the entire chapter: An element is said to be the first constituent from which something is composed, indivisible in form into another form, for instance the elements of utterances [φωνς στοιχεα] are those from which the utterance [ φων]13 is composed and into which it is ultimately divisible, and which are not further divisible into utterances di·erent in form; but if they do divide, the parts are of the same form, for example, as a part of water is water—this is not the case for a syllable. In the same way also those who speak of the elements of bodies mean the things into which bodies ultimately divide, and which are not further divisible into things other in form; and whether such things are one or many, they call these elements. Closely resembling this also are what are said to be the elements of geometrical propositions Theaetetus. But cf. also D. Sedley, The Midwife of Platonism: Text and Subtext in Plato’s Theaetetus (Oxford, 2004), 155–6. 12 Cf. D. Furley, The Greek Cosmologists, i. The Formation of the Atomic Theory and its Earliest Critics (Cambridge, 1987), 151. Furley is prepared to state that, if this use of the term pre-dates Plato, then Eudemus must be wrong to say that Plato first gave the name stoicheia to the elements. 13 Aristotle defines  φων as a kind of sound made by animate beings, at DA 420B5, cf. 29; also HA 535A27, PA 664B1. But it is only the sounds of the human voice that have stoicheia. A stoicheion is indivisible, but what is most distinctive about a stoicheion is that it is a constituent of a compound. The indivisible sounds of animals do not combine to make a composite sound, hence, although these are indivisible sounds, they do not constitute anything and are not stoicheia (see Poet. 20, 1456B20–2).

Created on 4 July 2005 at 14.35 hours page 371

On the Use of Stoicheion in the Sense of ‘Element’

371

[τ τν διαγραµµτων στοιχεα],14 and generally those of demonstrations; for the first demonstrations that are present in many demonstrations are called elements of demonstrations; such as the primary syllogisms, out of three terms through one middle. And, metaphorically, anything that is one and small, and which has many uses, is called an element, thus also the small and simple and indivisible is called ‘element’. From this comes the view that the most universal things are elements, because each of these, being one and simple, exists in many things [ν πολλος πρχει], or everything or nearly everything; also the one and the point are believed to be principles. Since, then, the things that are called ‘genera’ are universal and indivisible—for these do not have a definition—some call the genera ‘elements’, and more so than the di·erentiae, because the genus is more of a universal. For to whatever the di·erentia belongs, the genus follows, while to whatever the genus [belongs], the di·erentia does not always [follow]. So it is common to all cases that the element of each thing is the first constituent in each thing. (Metaph. ∆ 3, 1014A26–B15)15

In this passage Aristotle explains that stoicheia, in the most general sense, are the first things out of which composite items are constituted; they are the first things because they are indivisible into further, prior things. This is the general meaning of stoicheion (cf. De caelo 3. 3, 302A10–21). Aristotle then considers three examples of 14 W. D. Ross, in Aristotle’s Metaphysics: A Revised Text with Introduction and Commentary [Metaphysics] (2 vols.; Oxford, 1924), i. 234, refers to Asclepius (174. 9) for the point that διαγρµµατα are geometrical propositions, and not figures. See also T. L. Heath, Mathematics in Aristotle (Oxford, 1949), 205–6. J. L. Ackrill, Aristotle’s Categories and De Interpretatione: Translation with Notes (Oxford, 1963), 111, writes: ‘Many geometrical “propositions” are in fact solutions to construction problems (e.g., Euclid I, 1, 2, 3); and the construction of appropriate diagrams plays an important role in the proof of theorems (e.g. the theorem of Pythagoras, Euclid I, 47).’ 15 στοιχεον λ!γεται ξ ο# σ$γκειται πρ του νυπρχοντος 'διαιρ!του τ( ε)δει ε*ς +τερον ε,δος, ο.ον φωνς στοιχεα ξ /ν σ$γκειται  φων0 κα1 ε*ς 2 διαιρεται 3σχατα, κενα δ4 µηκ!τ6 ε*ς 7λλας φωνς 8τ!ρας τ( ε)δει α9τν, 'λλ κ:ν διαιρται, τ µ;ρια <µοειδ, ο.ον =δατος τ> µ;ριον =δωρ, 'λλ6 ο9 τς συλλαβς. <µοAως δ4 κα1 τ τν σωµτων στοιχεα λ!γουσιν οB λ!γοντες ε*ς 2 διαιρεται τ σ µατα 3σχατα, κενα δ4 µηκ!τ6 ε*ς 7λλα ε)δει διαφ!ροντα· κα1 ε)τε Dν ε)τε πλεAω τ τοιαEτα, ταEτα στοιχεα λ!γουσιν. παραπλησAως δ4 κα1 τ τν διαγραµµτων στοιχεα λ!γεται, κα1 Gλως τ τν 'ποδεAξεων· αB γρ πρται 'ποδεAξεις κα1 ν πλεAοσιν 'ποδεAξεσιν νυπρχουσαι, α#ται στοιχεα τν 'ποδεAξεων λ!γονται· ε*σ1 δ4 τοιοEτοι συλλογισµο1 οB πρτοι κ τν τριν δι6 +ν>ς µ!σου. κα1 µεταφ!ροντες δ4 στοιχεον καλοEσιν ντεEθεν I :ν Dν Jν κα1 µικρ>ν π1 πολλ KL χρσιµον, δι> κα1 τ> µικρ>ν κα1 MπλοEν κα1 'διαAρετον στοιχεον λ!γεται. Gθεν λλυθε τ µλιστα καθ;λου στοιχεα ε,ναι, Gτι +καστον α9τν Dν Jν κα1 MπλοEν ν πολλος πρχει N πOσιν N Gτι πλεAστοις, κα1 τ> Dν κα1 τ0ν στιγµ0ν 'ρχς τισι δοκεν ε,ναι. πε1 οPν τ καλο$µενα γ!νη καθ;λου κα1 'διαAρετα (ο9 γρ 3στι λ;γος α9τν), στοιχεα τ γ!νη λ!γουσA τινες, κα1 µOλλον N τ0ν διαφορν Gτι καθ;λου µOλλον τ> γ!νος· (/ µ4ν γρ  διαφορ πρχει, κα1 τ> γ!νος 'κολουθε, (/ δ4 τ> γ!νος, ο9 παντ1  διαφορ. Mπντων δ4 κοιν>ν τ> ε,ναι στοιχεον 8κστου τ> πρτον νυπρχον 8κστ(ω.

Created on 4 July 2005 at 14.35 hours page 372

372

Timothy J. Crowley

the usage of stoicheion, and he distinguishes these usages from metaphorical applications. Aristotle elsewhere describes a metaphorical application of a term as a ‘strange’ or unusual application of that term.16 He often contrasts metaphorical applications of terms with those applications where the term is used in the ‘ordinary’ way, i.e. strictly (κυρAως), or properly (ο*κεAως). By the ‘ordinary’ use Aristotle means the real, or actual, sense, the sense in which everybody uses the term.17 So, for instance, in the Poetics he explains ‘strange’ words (ξενικ;ν), or strange applications of words, such as metaphors, as ‘everything apart from the ordinary’ (πOν τ> παρ τ> κ$ριον, 1458A23–5).18 The first three examples of the use of stoicheion in Metaphysics Delta, then, ought to be understood as examples of ordinary, or non-metaphorical, usages of the term. The things that are ordinarily called stoicheia are (1) the things into which syllables are divisible; (2) the things into which bodies are divisible; and (3) the things into which geometrical propositions are divisible, or the principles of proofs or demonstrations— that is, the propositions whose proof is involved in the proof of other propositions.19 Now clearly these things are called stoicheia homonymously. For they are all called stoicheia, but they are not defined in the same way (see Cat. 1A1–4). The stoicheia of syllables, which are phonemes or letters, belong to the science of grammar (see Cat. 14B1–2; cf. Poet. 1456B20–2); the stoicheia of bodies belong to the science of nature or physics; and the stoicheia of geometrical propositions belong to the science of geometry. They are, however, associated homonyms,20 because, while the definition pertaining to stoicheia is di·erent in each case, there is a shared core meaning, which is the general meaning of stoicheion: ‘that first, indivisible constituent out of which something is composed’.21 So, for in16 At Poet. 1457B6 ·. he defines µεταφορ (which is not exactly equivalent to metaphor) as ‘a word used in a strange way’. 17 See LSJ s.v. κ$ριος; also e.g. Poet. 1457B3–4; cf. Rhet. 1404B5–6. For the use of ο*κεος and κ$ριος see Rhet. 1404B31–2 (cf. 1410B12–13). 18 For further contrasts between µεταφορ and ordinary usage, see Poet. 1457B1–4, also 1458A33, 1458B17; Top. 123A33–6, 139B34, 158B12; Rhet. 1410B12–13; MM 1. 26, 1192B15–16. 19 Aristotle o·ers a similar distinction of the three senses of stoicheia at Metaph. Β 3, 998A23–8. 20 Associated, as opposed to discrete, homonyms. On this distinction see ch. 1 of C. Shields, Order in Multiplicity: Homonymy in the Philosophy of Aristotle [Homonymy] (Oxford, 1999). 21 C. Kirwan, in Aristotle: Metaphysics Books Γ, ∆, and Ε. Translated with an

Created on 4 July 2005 at 14.35 hours page 373

On the Use of Stoicheion in the Sense of ‘Element’

373

stance, the constituent of a syllable and the constituent of a body have the name stoicheion in common, but they do not have the same definition; nevertheless, there is something in their definitions that they do share, and that is that they are constituents of compounds and indivisible into further constituents that are di·erent in form.22 The ‘transfer’ (µεταφορ) or metaphorical application of stoicheion is said to be to anything that is small and simple and indivisible, and that has many uses (1014B3–4).23 This transfer opens the way towards applying the name stoicheion to things that are most universal. Thus, for example, the point and the One, or unit (τ> +ν), might be called stoicheia. Here Aristotle presumably has in mind the Platonists’ use of stoicheion.24 Aristotle occasionally indicates that the Platonists use the term stoicheia with reference to the elements of number, i.e., the One and the Great and the Small (1087B13, 1091A10), or the One and the Unequal (1087B9). For instance, in Metaphysics Nu he says that they call the principles of numbers stoicheia (τς 'ρχς 2ς στοιχεα καλοEσιν, 1087B12–13). It would seem, then, that Aristotle regards the Platonist use of stoicheion as metaphorical. This point is not without some pejorative implications; Aristotle is not averse to criticizing the Platonists for appealing to metaphors (Metaph. 991A20 ·., 1079B24 ·.). The chapter concludes in the same way that it began, with Aristotle repeating the core meaning of stoicheion (1014A26, b15). But one might doubt that metaphorical applications of stoicheion also share the core meaning. For the notion of being the first constituent of a compound seems to be, if not absent (for the point might be Introduction and Notes, 2nd edn. (Oxford, 1993), 128, says that the first sense of stoicheion at 1014A26 refers to material substances, e.g. fire, air, water, and earth. But this is mistaken. The account of the term stoicheion at 1014A26, i.e. indivisible constituent of compounds, is the general meaning of stoicheion, a meaning that is shared by each one of the ordinary usages of the term. The use of stoicheion with reference to the material constituents of bodies is the second example of the usage of the term. 22 Shields o·ers the following explanation of core-dependent homonymy: ‘x and y are homonymously f in a core-dependent way i·: (i) they have their name in common, and (ii) their definitions do not completely overlap, but (iii) they have something definitional in common’ (Homonymy, 106). 23 A metaphorical use of stoicheion, in the sense of being one and indivisible and with many uses, may be its use in the sense of τ;πος, i.e. an argument widely applicable; cf. Top. 120B13, 121B11, 151B18 (this is Bonitz’s suggestion, reported by Ross, Metaphysics, i. 295). 24 Ross, Metaphysics, i. 295: ‘Aristotle is referring here to Pythagorean and Platonic views’; cf. Metaph. 986A1, 998A20 ·., 1028B25–8, 1069A26–8.

Created on 4 July 2005 at 14.35 hours page 374

374

Timothy J. Crowley

considered, by some, to be a constituent of lines and planes; cf. Phys. 215B18–19), then certainly less crucial in these applications of stoicheion. As Aquinas points out in his commentary on this passage, universals are not constituents of things; rather they predicate the substance of a thing.25 Having said that, it may often be di¶cult to distinguish ordinary from metaphorical usages of the term. There are glimpses of the use of stoicheion to capture a notion of ‘principle’ on the remarkably few occasions that the term is found in the extant writings of Plato’s contemporaries. Both Isocrates and Xenophon use the term in the sense of the ground rules, or first parts, of some subject or discipline. Isocrates refers to the most important stoicheia of good government, and of the stoicheia of rhetoric; Xenophon uses the term in the sense of the ‘first things’.26 Stoicheion here means the first parts, or the ground rules, of some thing or discipline. We find this use also in Aristotle, for instance when he, like Isocrates, refers to the parts of rhetoric as the stoicheia (Rhet. 1. 2, 1358A35). Again, in the Organon Aristotle often employs stoicheia in the sense of ‘elementary rules’ (Top. 4. 9, 147A22; see also 4. 3, 123A28; 6. 5, 143A13; SE 172B21, b31; 174A21). It might be thought that these examples provide evidence that there is an incipient or rudimentary sense of principle or arch»e connoted by the term stoicheion.27 But are these examples instances of metaphorical use?28 Fortunately, thanks to the Delta chapter, we can pick out at least three usages of stoicheion that Aristotle explicitly identifies as ordinary or non-metaphorical. For the sake of convenience, let us call the first of these the ‘alphabetic’ sense,29 the second the ‘elemental’ sense, and the third, referring as it does to the principles, axioms, and postulates of geometry, the ‘geometric’ sense. By the ‘elemental’ sense, at this point, I intend nothing more technical than the use 25 Cf. Aquinas, Commentary on the Metaphysics of Aristotle , trans. by J. P. Rowan (Chicago, 1964), vol. 1, bk. 5, Lsn 4, Sct 804, p. 317. 26 Isocr. Ad Nicolem 16. 7, Ad filium Iasonis 8. 8; Xen. Mem. 2. 1. 1. 9. A somewhat di·erent, and rather singular, use of the term is Aristophanes’ use of stoicheion in the sense of the individual measures or units of a sundial (Eccl. 652). But this use may o·er the best clue to the original meaning of the term—a question which I do not pursue. 27 See Diels, Elementum, 17, 22. 28 For Diels, Isocrates and Xenophon are using stoicheia metaphorically in the passages cited above (Elementum, 17). 29 ‘Alphabetic’ is perhaps not completely satisfactory, but it seems more appropriate than ‘grammatical’ or ‘linguistic’, and more familiar than e.g. ‘phonic’.

Created on 4 July 2005 at 14.35 hours page 375

On the Use of Stoicheion in the Sense of ‘Element’

375

of stoicheion to refer to a constituent, or a principle, of a body, rather than to the constituent of a syllable or a proposition of geometry. It is important to emphasize that these three senses of the term stoicheion are independent of each other, and hence that each can be understood without reference to the other. For instance, stoicheion in the elemental sense is not presented as being dependent upon, or metaphorically derived from, stoicheion in the alphabetic sense; both senses are ordinary usages of the term.30 This reflects Aristotle’s practice elsewhere. In De caelo 3. 3, for instance, Aristotle’s definition of stoicheion in the elemental sense is presented without reference to stoicheion in either its alphabetic or its geometric sense (302A10–21). The context of the definition in the De caelo is cosmology, or in general, the study of nature, and Aristotle uses the term stoicheion as an appropriate technical term of the study of nature. Similarly, in the Poetics he identifies stoicheion as a part of speech (or language, λ!ξις) and defines it accordingly, without reference to either of the other senses of the term (1456B20–2). Of the three ‘ordinary’ usages of stoicheion that Aristotle identifies, I think it is quite clear that the alphabetic and the geometric senses were familiar to Plato and his contemporaries. Many examples of the alphabetic sense can be recognized in Plato’s works. In the Cratylus, for instance, Socrates refers to ‘the alpha and beta and the other stoicheia’ (431 a, cf. 393 d, 426 d, 433 a–b; cf. Phileb. 17 a–18 d). The use of the term stoicheia in geometry, in particular as the title for treatises on geometry, was made famous by Euclid, but the ‘geometric’ sense of stoicheion may well have been familiar since the fifth century; in his commentary on Euclid, Proclus reports that Socrates’ contemporary Hippocrates of Chios was the first to write an ‘Elements’ of geometry (Procl. In Eucl. 66. 4–8 = 42 A 1 DK). Theudius’ ‘Elements’, the immediate precursor to Eu30 For a very di·erent reading of Metaphysics Delta 3, see M. Crubellier, ‘Metaphysics Λ 4’, in M. Frede and D. Charles (eds.) Aristotle’s Metaphysics Lambda (Oxford, 2000), 137–60. Crubellier suggests that what I have called the elemental, geometric, and metaphorical uses of stoicheion are all derived metaphorically from what he calls the ‘original’ reference, i.e. the alphabetic sense (142). But Crubellier ignores the core meaning of the term with which Aristotle bookends the chapter; and he attempts to force on certain terms, such as <µοAως and παραπλησAως, by which the elemental and geometric usages are introduced, an unnatural tone of qualification. Furthermore, he translates καA at b3 as ‘hence’, thereby removing the contrast that, as I have argued, Aristotle intends between ordinary and metaphorical usages of stoicheia. I find Crubellier’s general claim, that Delta 3 betrays Aristotle’s ‘distrust’ of the term stoicheion, to be without foundation.

Created on 4 July 2005 at 14.35 hours page 376

376

Timothy J. Crowley

clid’s Elements, was the geometry textbook of the Academy.31 It is not certain that these pre-Euclidean ‘Elements’ were actually called Stoicheia.32 Proclus does not say that they were. But it is likely that by Euclid’s time, Stoicheia was the traditional title for treatises on geometry. Whether or not the ‘elemental’ sense is a recognized usage of stoicheion by Plato’s contemporaries, not to mention his predecessors, is, of course, less clear. On the standard interpretation of Eudemus’ evidence, Plato himself is responsible for coining the ‘elemental’ sense of stoicheion. Moreover, it is usually thought that he arrives at this novel sense by transferring the term from its familiar ‘alphabetic’ sense, rather than from its (presumably also familiar) ‘geometric’ sense, to the context of the study of nature. Ross, for instance, writes: ‘in Plato [stoicheion] often means an element of spoken language, answering to γρµµα, an element of written language, and in Theaet. 201 e it is metaphorically used of the elements of any complex whole’.33 So, on this view, the elemental sense of stoicheion is metaphorically derived from the alphabetic sense. But if stoicheion had not been used in this way before, then, presumably, it would have been quite clear to Plato’s audience that the term is being used in a new, and unusual, way. It wears its metaphorical derivation on its sleeve, as it were.34 Now, as we have just seen, in Metaphysics Delta Aristotle is careful to distinguish the ordinary usages of stoicheion from metaphorical transfers. But he presents the elemental sense of stoicheion as one of the ordinary, and therefore non-metaphorical, usages of the term. It seems quite remarkable that Aristotle would identify as a non-metaphorical usage of a term a usage that was introduced by its progenitor in such a way that its metaphorical origins are not merely apparent, but explicit (cf. Theaet. 202 e). Could the metaphorical origins of the ‘elemental’ sense of stoicheion have been forgotten by Aristotle’s time?35 Such a suggestion seems incredible; for Plato’s innovation, if we admit that it is such, would have 31 See Ross, Metaphysics, i. 234. 32 Note that Diels, Elementum, 27, argues that the titles cannot be considered as evidence for the dating of the mathematical use of stoicheion. 33 Metaphysics, i. 137–8; cf. n. 8 above. 34 Cf. Kahn: ‘[with Plato’s use of stoicheion] the comparison to the letters was still clearly borne in mind’ (Anaximander, 120). 35 ‘Only with Aristotle does στοιχεα appear as an abstract expression whose metaphorical value has been largely forgotten’ (Kahn Anaximander, 120).

Created on 4 July 2005 at 14.35 hours page 377

On the Use of Stoicheion in the Sense of ‘Element’

377

had to be relatively recent. For Metaphysics Delta, and also the De caelo—where the term stoicheion is used in the context of physics without qualification or apology—are both usually considered to be early works, while the dialogues where Plato uses the term in the elemental sense are of his middle and late periods.36 In fact, Aristotle already seems to be using stoicheia in the sense of ‘first principles of nature’ in the Protrepticus, which is traditionally reckoned to be one of his earliest works (B 35).37 If Plato did indeed introduce the elemental sense of the term stoicheion, then I think one ought to expect some indication of Aristotle’s awareness of this innovation. But there is no such indication. Aristotle thinks that the alphabetic and the elemental senses of stoicheion share a core meaning, but nowhere in the corpus does it seem even to have occurred to him that one was derived from the other. There are places where he might have indicated that this is the case, if indeed it were; but he consistently fails to do so.38 There is the suggestion, made by Ross, that the problematic phrase ‘so-called elements’ (τ καλο$µενα στοιχεα, τ λεγ;µενα στοιχεα) is evidence of Aristotle’s acknowledgement that this use of the term stoicheion is not yet firmly established in the context of natural philosophy.39 This suggestion is presumably made under the influence of the view that Plato had recently introduced the elemental sense of the term. If Ross were right about the phrase ‘so-called 36 The chronology of Aristotle’s works is something of a minefield, but most commentators would agree that the De caelo is an early work; Metaphysics Delta is also often regarded as an early work: it is thought to be earlier than many of the other books of the Metaphysics, and perhaps even earlier than the physical writings; see Ross, Metaphysics, vol. i, p. xxv. For a succinct historical survey of the issue of the chronology of Plato’s dialogues, see L. Brandwood, ‘Stylometry and Chronology’ [‘Stylometry’], in R. Kraut (ed.), The Cambridge Companion to Plato (Cambridge, 1992), 90–120. 37 During, Protrepticus, 210. During argues that Aristotle is using the popular • • language of the Academy. On the dating of the Protrepticus, see ibid. 287–8. 38 For instance, in Metaphysics Μ 10 he sets up an analogy between the stoicheia of syllables and the stoicheia of substances, but the analogy works because the stoicheia of syllables and of substances share a core meaning, and not because the latter are derived metaphorically from the former (1086B22–3). See also Metaph. Ζ 17, 1041B13–31, esp. b16–17. Cf. 1013B17 (same as Phys. 195A16), and GA 722A32. 39 Ross, Metaphysics, i. 294. Note that Ross changes his mind on the interpretation of τ καλο$µενα στοιχεα, in Aristotle’s Physics: A Revised Text with Introduction and Commentary (Oxford, 1936), 484; there he moves towards an interpretation of the phrase that is close to that in H. H. Joachim, Aristotle: On Coming-to-Be and Passing-Away: A Revised Text with Introduction and Commentary (Oxford, 1926); see e.g. Joachim, 137.

Created on 4 July 2005 at 14.35 hours page 378

378

Timothy J. Crowley

elements’, then we ought to expect to find ‘elements’ prefaced by ‘so-called’ most particularly in the early works, such as the De caelo and the Protrepticus. But we do not find this at all. Aristotle appears to be confident that everyone will understand what ‘element’ means when used in the context of a discussion of nature or cosmology.40 Now it is well to point out that many terms, scientific concepts in particular, are originally introduced into scientific discourse as metaphors; and over time the original, literal meaning tends to be forgotten and the terms may lose their metaphorical associations, and become ‘dead metaphors’. So it need not be very remarkable that Aristotle would identify as an ordinary, non-metaphorical usage of a term a usage that was originally metaphorical. But for the elemental sense of stoicheion to be accepted as an ordinary usage so quickly is certainly remarkable. Could it be a simple linguistic fact that Plato’s innovation has been very successful, to the extent that Aristotle regards the ‘elemental’ sense as one of the ordinary usages of the term stoicheion, and is uninterested in, perhaps unaware of, its metaphorical origins? Such rapid terminological establishment could be explained, for instance, as being due to a combination of Plato’s great personal influence and, perhaps, a renewed interest in physics and cosmology after the Timaeus. But this suggestion must be rejected. For if we turn now to examine Plato’s own use of stoicheion in the elemental sense, it seems that he is taking up a usage that is already established among his own contemporaries—a usage that is, as Aristotle would say, ‘ordinary’.

2. Stoicheia in the Timaeus The dialogues where Plato uses stoicheion in the elemental sense are the Theaetetus, the Sophist, and the Timaeus. The Theaetetus is probably the earliest of these,41 and the term stoicheion occurs for the first time at 201 e. It is introduced in a somewhat qualified way; 40 Moreover, Ross’s suggestion would not explain why later Greek philosophers continue, albeit infrequently, to use the expression τ καλο$µενα στοιχεα; see e.g. S.E. PH 3. 62; Plot. Enn. 3. 1. 3. 2. 41 Presuming the Theaetetus to be earlier than the Sophist, Politicus, and Timaeus, which is to agree with the conventional ordering of the dialogues and with recent stylometric studies; see e.g. Brandwood, ‘Stylometry’. But see G. E. L. Owen, ‘The Place of the Timaeus in Plato’s Dialogues’, Classical Quarterly, ns 3 (1953), 79–95, for an argument that the Timaeus pre-dates the Theaetetus.

Created on 4 July 2005 at 14.35 hours page 379

On the Use of Stoicheion in the Sense of ‘Element’

379

Socrates seems rather di¶dent about using stoicheion in this sense. But this is the closest Plato comes to o·ering an introduction of the term in the sense of element or primary physical ingredient.42 For in the Sophist, which according to the most widely accepted chronologies would be the next dialogue to feature the elemental sense of stoicheion, the term is used just once, and without fanfare. The Eleatic Stranger refers to ‘those who make all things come together at one time, and divide into parts at another . . . divided into a limited number of elements [στοιχεα] and out of these put together again’ (252 b).43 This ready use of stoicheion in the elemental sense would suggest that it is a familiar usage. One gets a similar impression from the Timaeus (48 b–c). But the apparent familiarity with the elemental sense of stoicheion in each of these dialogues does not discount in principle the possibility that Plato himself introduced this sense at Theaet. 201 e. Before we take a closer look at the Theaetetus, however, let us examine the evidence of the Timaeus. It seems better to start with the Timaeus because the relevant passage is less problematic than the Theaetetus passage, and indeed it may be of help in understanding the latter. In the crucial passage, Plato, through his spokesman Timaeus, is referring to the popular, but unreflective, belief that fire, air, water, and earth are the principles and elements of everything. This belief is unreflective because no one knows, or has attempted to explain, the nature of these bodies. Plato claims that anyone who thinks seriously about the matter will see that fire, air, water, and earth cannot be regarded as the most basic constituents of things. He writes: So one must see the nature of fire and water, air and earth before the generation of the heaven and the properties of this; for as the case stands, no one up to this time has revealed the generation of these, but as if people know what is fire and each of the others, we say they are principles, positing them as elements of the whole [στοιχεα τοE παντ;ς]; though it is not even right for someone with even a little intelligence reasonably to compare them even to the class of syllables. (48 b–c)44 42 Precisely what sort of things the stoicheia at Theaet. 201 e are meant to be is a matter of some dispute: they are often thought to be logical atoms or conceptual constituents, rather than physical ingredients. See e.g. G. Ryle, ‘Logical Atomism in Plato’s Theaetetus’, Phronesis, 35 (1990), 21–46. 43 In this passage Plato is probably referring to Empedocles; see Diels, Elementum, 20; Lagercrantz, Elementum, 17. 44 τ0ν δ0 πρ> τς ο9ρανοE γεν!σεως πυρ>ς =δατ;ς τε κα1 '!ρος κα1 γς φ$σιν θεατ!ον

Created on 4 July 2005 at 14.35 hours page 380

380

Timothy J. Crowley

The view that fire, air, water, and earth are the constituents of things appears to be something of a commonplace by Plato’s time.45 But the significant feature of the above passage is that it seems these four ‘Empedoclean’ elements are commonly or popularly called stoicheia by Plato’s contemporaries. In other words, the elemental sense of the term stoicheion is already available. For what irks Plato here is precisely that people tend to call fire, air, water, and earth the archai and stoicheia of everything. This, for Plato, is a mistake. The meaning of archai is ‘the original sources or principles of things’; and of stoicheia, ‘the ultimate constituents of things’. Plato would agree that fire, air, water, and earth are constituents, but he wants to deny that they deserve to be called the ultimate constituents, the archai or stoicheia, of everything. Hence Plato is criticizing the contemporary usage of these terms. Clearly, if stoicheion is already regularly used in the elemental sense by Plato’s contemporaries, then Plato himself cannot be responsible for introducing this sense of the term.46 This might seem a natural reading of the passage. But perhaps the matter cannot be so easily settled. A problem for this reading is that one might think that stoicheion at Tim. 48 b–c is metaphorical, and clearly so. Fire, air, water, and earth are so far from being stoicheia, Plato says, that they are not even like syllabai, or syllables. The term syllab»e has obvious grammatical, or what I have been calling alphabetic, connotations; the most common meaning of syllab»e is ‘a compound of stoicheia, in the alphabetic sense’, i.e. ‘a compound of phonemes (or letters)’. But Plato is not making the point that fire, air, water, and earth are not even compounds of phonemes; rather, at 48 b–c syllab»e is being used as a metaphor for a (minimally) complex body. But if syllab»e is being used metaphorically, then one might think that the use of stoicheion at Tim. 48 b–c is likewise metaphorical. In particular, it may be argued that the term stoiα9τ0ν κα1 τ πρ> το$του πθη· νEν γρ ο9δεAς πω γ!νεσιν α9τν µεµνυκεν, 'λλ6 Vς ε*δ;σιν πEρ Gτι ποτ! στιν κα1 +καστον α9τν λ!γοµεν 'ρχς α9τ τιθ!µενοι στοιχεα τοE παντ;ς, προσκον α9τος ο9δ6 :ν Vς ν συλλαβς ε)δεσιν µ;νον ε*κ;τως π> τοE κα1 βραχW φρονοEντος 'πεικασθναι. 45 Cf. Tim. 49 b–c. For further evidence in Plato’s works that fire, air, water, and earth are popularly regarded as the material constituents of things, see Phileb. 29 a 10–11, Crat. 408 d, and cf. Prot. 320 d. 46 Cf. Burkert, who does not think that Tim. 48 b–e is of relevance for the question of the usage of stoicheion; hence he believes that this passage does not a·ect the validity of Eudemus’ report (‘Stoicheion’, 176).

Created on 4 July 2005 at 14.35 hours page 381

On the Use of Stoicheion in the Sense of ‘Element’

381

cheion ordinarily has the alphabetic sense, but it does not ordinarily have the elemental sense. Thus, Taylor, in his commentary on the Timaeus, glosses στοιχεα τοE παντ;ς as ‘literally the “ABC of everything”’.47 It is but a short step to conclude that the elemental sense is a sense that has been introduced recently, perhaps by Plato himself, by a metaphorical derivation from the alphabetic sense. Before examining this suggestion, it is important to clarify just what Plato is getting at by the comparison of fire, air, water, and earth with syllables. He believes that fire, air, water, and earth are not the ultimate constituents of things, and to make this point clearly and sharply, he denies that they are even as basic as syllabai. In other words, fire, air, water, and earth have constituents, and as such they cannot be regarded as genuine stoicheia; but even these immediate constituents are not the genuine stoicheia, because they can be analysed into further, even more basic, constituents. The constituents of a syllable, on the other hand, are the stoicheia, i.e. stoicheia in the alphabetic sense, because a syllable is the first thing that stoicheia constitute. The point is that fire, air, water, and earth are not comparable to syllables, because, unlike syllables, fire, air, water, and earth are complex phenomena that admit of more than one division before the genuine stoicheia are reached. Plato explains why this is so a little later in the dialogue. At 53 d he says that it is clear to everyone that fire, air, water, and earth are bodies, and that all bodies are solids. Furthermore, all solids are bound by surfaces, and all surfaces are divisible into scalene and isosceles triangles. These triangles are held to be elementary (53 c– d); for from these all other triangles come to be, e.g. the equilateral triangles that make up the surfaces of fire, water, and air. There is a question as to whether the triangles are the ultimate principles: at 53 d Plato says that ‘archai more ultimate than this only the god knows and such a man who is loved by god’. But it is nevertheless clear why the common or popular notions of fire, air, water, and earth are not even comparable to syllables; they are not even the first compounds of the ultimate stoicheia. Fire, air, water, and earth are already at several removes from the basic triangles, and these latter may not even be the ultimate archai and stoicheia of things. 47 Taylor, Timaeus, 306. Similarly H. D. P. Lee, in his popular translation of the Timaeus (Harmondsworth, 1965), renders στοιχεα τοE παντ;ς ‘the alphabet of the universe’. See also R. G. Bury, Plato, vii. Timaeus (London and Cambridge, Mass., 1929), 110 n. 1.

Created on 4 July 2005 at 14.35 hours page 382

382

Timothy J. Crowley

Notably, however, Plato does proceed to treat the basic triangles as the stoicheia of things (54 d; 55 a, b; 56 b; 57 c; 61 a). Now I do not wish to deny that Plato appeals to metaphor in his attempt to undermine the popular identification of fire, air, water, and earth as the στοιχεα τοE παντ;ς. The reference to syllabai certainly involves an allusion to the alphabetic sense of stoicheion. Perhaps it should be noted that the reference to syllab»e is not necessarily a reference to alphabetic syllables: just as stoicheion has a core meaning of ‘primary constituent’, so it is sometimes suggested that syllab»e has a basic or primitive sense of ‘that which is held together’—that is, of several things held together; hence a composite or complex, as opposed to a simple, object.48 But this use seems to be rare, and in any case the force of the critique at 48 b–c is lost if syllabai simply means an object composed of other things. For what would it mean to say that fire, air, water, and earth ought not to be compared even to the class of complex objects?49 Clearly, then, the term syllab»e is being used metaphorically, to name that first thing that is composed of the physical or natural stoicheia. But I do not think that it follows that the use of stoicheia in the expression στοιχεα τοE παντ;ς is also metaphorical. It makes better sense, both of the passage itself and of the subsequent use of the term, to think that stoicheion is already ordinarily used in the elemental sense, a sense that is independent of the alphabetic sense of the term; and hence that Plato in the Timaeus is deliberately trying to overturn the popular belief that fire, air, water, and earth are the stoicheia of everything.50 And the way he sets about doing so is by playing on the ambiguity of the term stoicheion, in particular by alluding to one of the other ‘ordinary’ senses of the term, namely, 48 See LSJ s.v.; also McDowell, Theaetetus, 239, and Burnyeat, Theaetetus, 340 n. 1. 49 It would have to be specified that syllab»e means ‘complex object that is composed of simples’, i.e. of stoicheia; the alphabetic sense does this job perfectly. 50 Compare the brief cosmological aside in the Philebus (29 a–b), a dialogue which is generally regarded to be among Plato’s latest (in many aspects of style it is close to the Laws, which is almost certainly Plato’s final dialogue: see Brandwood, ‘Stylometry’, 114). Here Socrates refers to fire, air, water, and earth, but he does not refer to them as stoicheia. If indeed it were the case that Plato introduced the use of stoicheia as a way to refer to such things as fire, air, water, and earth, then it would certainly be a little surprising that he refrains from using his innovation in the Philebus. An admittedly speculative suggestion is that Plato deliberately refrains from using stoicheia to refer to fire, air, water, and earth in the Philebus precisely because he has, in the Timaeus, criticized the popular identification of fire, air, water, and earth as stoicheia.

Created on 4 July 2005 at 14.35 hours page 383

On the Use of Stoicheion in the Sense of ‘Element’

383

the alphabetic sense. To say that they are ‘not even syllables’ ought to be taken as an ironic jibe at the common view that fire, air, water, and earth are the stoicheia of all things. Plato here is drawing an analogy—albeit an analogy he wants to undermine—between fire, air, water, and earth as the constituents of bodies, and phonemes or letters as the constituents of syllabai, or syllables. For it to work, however, stoicheia must be a term that is commonly used to refer both to fire, air, water, and earth and to phonemes or letters. Hence, at 48 b–c syllabai must be understood in relation to stoicheia; but, on the other hand, stoicheia must be understood not only in relation to syllabai, but also according to its elemental sense. In fact, while the passage and its context indicate that stoicheion is used commonly in the elemental sense, it is quite clear that the compounds of physical elements are not commonly called syllabai. For whereas Plato continues later in the dialogue to use the term stoicheia to refer to what he considers to be the genuine elements, namely, the basic triangles, he does not use the term syllabai after 48 b–c. This would be surprising if he were borrowing both terms from alphabetic discourse. If both stoicheia and syllabai are being used metaphorically, then why not refer to the first things that the genuine stoicheia constitute, e.g. the equilateral triangles or the surfaces, as syllabai? I would suggest that Plato does not do so because stoicheion is a term that has a recognized elemental sense, whereas syllab»e is not; syllab»e, as noted above, may have a primitive sense of ‘that which is held together’, i.e. a complex object; but to speak of ‘physical or material syllabai’ would evidently be quite unusual, in a way that ‘physical or material stoicheia’ is not. Hence the term that is being used metaphorically in Tim. 48 b–c is not stoicheia, but syllabai. It is illuminating to compare at this point the brief report of the Platonic doctrine of generation that Aristotle o·ers in his Protrepticus. The doctrine of generation that Aristotle reports is similar to, but not quite the same as, that of the Timaeus. It is presumably drawn from Plato’s ‘unwritten doctrines’. According to Aristotle, the Platonists analyse the generation of substance in the following way: ‘lines [come to be] from numbers, planes from lines, solids from planes, and what they call syllables from elements’ (fr. 33. 9). In other words, at bottom there are numbers, from which everything else comes to be, because these come to be lines, and lines come to be planes, and planes come to be solids; then we have

Created on 4 July 2005 at 14.35 hours page 384

384

Timothy J. Crowley

stoicheia coming to be syllables. There are two points of significance in this passage to which I want to draw attention. Firstly, we have here further evidence that stoicheia in the elemental sense commonly refers to fire, air, water, and earth. For Aristotle readily identifies the ‘solids’ as the stoicheia, and the ‘solids’, as Plato argues in the Timaeus (53 d), are fire, air, water, and earth. Evidently, then, even in a very early work like the Protrepticus, Aristotle thinks it natural to call fire, air, water, and earth stoicheia. At the same time, however, he feels the need to qualify what presumably is a very singular and unusual use of the term syllabai. This, indeed, is the second point of significance. Aristotle uses the phrase ‘what they call syllables’ (αB Xνοµαζ;µεναι συλλαβαA). What is it that Plato and the Platonists call ‘syllables’ in the context of the explanation of the generation of substance? According to the Protrepticus, what they call syllables come to be from the elements; that is, what they call syllables are composites of fire, air, water, and earth. By qualifying the term in this way, however, Aristotle is clearly indicating that syllabai is being used in an unusual sense; and to use a term in an unusual, or strange, sense is, of course, to apply the term metaphorically. The Platonists, in other words, use ‘syllables’ as a metaphor for ‘complex bodies’; they adopt a term appropriate to one sphere of discourse, namely, grammar (or, more generally, language), and transfer it into another, namely, physics (and metaphysics). The obvious inference from this and the preceding point is that, whereas stoicheion does have an elemental sense, syllab»e does not ordinarily have a sense corresponding to the elemental sense of stoicheion. To speak of syllabai as compounds of elemental stoicheia, then, is to speak metaphorically. Our examination of the Timaeus o·ers little indication that Plato is the first to call the elements or principles of things stoicheia. On the contrary, the relevant passage suggests that the term is already in use, in this sense, among Plato’s contemporaries. This is a conclusion that is supported by Aristotle’s account of stoicheion in Metaphysics Delta, and also by his Protrepticus. But what is particularly interesting about the passage we considered from the Protrepticus is that it o·ers evidence that Plato introduced a di·erent term into discussions of nature and the composition of natural things, namely, syllab»e or ‘syllable’. Moreover, this use of syllab»e seems to be identified by Aristotle as a metaphorical derivation, presumably from the more familiar, alphabetic, sense of the term.

Created on 4 July 2005 at 14.35 hours page 385

On the Use of Stoicheion in the Sense of ‘Element’

385

3. Stoicheia in the Theaetetus Let us now consider the use of stoicheion in the Theaetetus. If, as is usually thought, the Theaetetus is earlier than the Sophist and the Timaeus, then the occurrence of stoicheion at 201 e is the first occurrence of the term in the elemental sense in the extant literature. But is it also the locus of the first ever use of stoicheion in this sense? The context is a discussion of knowledge, and the di·erence between knowledge and true belief. In order to clarify the distinction between what is knowable and unknowable, Socrates recounts a ‘dream’: ‘for I seem to have heard from some people that the first stoicheia, as it were [οBονπερεA], out of which we and other things are composed, have no account’ (201 d–e).51 Now, at first glance one might think that this passage o·ers good evidence that Plato did not introduce the term stoicheion in the elemental sense, precisely because the term is used by Socrates with reference to someone else’s theory. Unfortunately the matter is not quite so clear. For one thing, it is very di¶cult to identify the author, or authors, of this ‘dream’ theory.52 Some commentators have suggested that Socrates is referring to a Pythagorean theory, or the theory of atomism; others have inferred, from the remark that Socrates has dreamt the theory, that Plato is thinking of a theory held by a contemporary and promulgated after Socrates’ death.53 Still others, however, suggest that Plato is simply inventing the theory.54 Certainly, if Plato is making it up, then he may well be using stoicheion in the elemental sense for the very first time. If, on the other hand, Socrates is referring to a theory held by a historical figure, it does not necessarily follow that the author of the ‘dream’ theory actually used the term stoicheion. Hence, whether we think that Socrates is reporting a theory, or the 51 γY γρ αP δ;κουν 'κο$ειν τινν Gτι τ µ4ν πρτα οBονπερε1 στοιχεα, ξ /ν µες τε συγκεAµεθα κα1 τZλλα, λ;γον ο9κ 3χοι. 52 All references to the author of the ‘dream’ are in the singular, except the first, which is in the plural; cf. 202 e. 53 Antisthenes is often suggested as the author of the theory. Cf. Cornford, PTK, 143–4; W. K. C. Guthrie, A History of Greek Philosophy, v. The Later Plato and the Academy (Cambridge, 1978), 114–15; Bostock, Theaetetus, 202; Burnyeat, Theaetetus, 164–5. For Lewis Campbell, The Theaetetus of Plato: A Revised Text with English Notes [Theaetetus] (Oxford, 1883), and Taylor, Timaeus, 306–7, Socrates is dreaming of Pythagoreans. 54 See McDowell, Theaetetus, 234, 237. Cf. Diels, Elementum, 19; Burkert, ‘Stoicheion’, 174.

Created on 4 July 2005 at 14.35 hours page 386

386

Timothy J. Crowley

theory is a literary fiction, it remains a distinct possibility that this is the place where Plato introduces the elemental sense of stoicheion. If it were the case that Plato is here using the term for the first time in its elemental sense, then one question that we might reasonably ask is: why use this term? Plato would no doubt be aware of his contemporaries’ use of stoicheion in the primitive sense of ‘principle’, e.g. the stoicheia of rhetoric or of good government (see Section 1). But the usual view is that he is drawing on the alphabetic sense of stoicheion; and, indeed, there does seem to be some support for this view. At 202 e Socrates suggests to Theaetetus that, in order better to understand the ‘dream’ theory, they ought to take as ‘hostages’ the models (τ παραδεAγµατα) used by the author of the theory. He immediately identifies the ‘models’ as ‘the stoicheia and syllabai of grammar’ (or of writing: τ τν γραµµτων στοιχε τε κα1 συλλαβς). This certainly appears to be very strong evidence that the use of stoicheion in the sense of ‘element’ is metaphorically derived from the alphabetic sense of the term. But before trying to evaluate this evidence, let us return to 201 e. The interpretation of Theaet. 201 e depends to a large extent on how we understand the phrase τ οBονπερε1 στοιχεα. The rather mysterious expression οBονπερεA may indicate that Socrates is introducing, quite self-consciously, the term stoicheion, and in particular that he wants clearly to mark that he is introducing it metaphorically. In e·ect, he may be saying something like this: ‘the first “letters”, if you will, from which we and other things are composed’.55 Obviously this would be the favoured reading if we think that Plato is responsible for introducing the elemental sense of stoicheion. Now this reading need not imply that the ‘dream’ theory is his own invention. The ‘dream’ theory may be an actual historical theory that Socrates is reporting, but one that does not employ the term stoicheion; in order to explain the theory Socrates supplies the term himself. So, by using the expression οBονπερεA, Socrates would be flagging his transfer of stoicheion from its familiar use in the alphabetic sense to an unusual, or ‘strange’, use to describe the ultimate constituents of things.56 If Socrates is reporting an actual historical theory, however, and 55 Bostock renders οBονπερεA ‘if I may so call them’ (Theaetetus, 204). 56 For Diels, Elementum, 19, and Burkert, ‘Stoicheion’, 175, the alphabetic sense prepares and legitimates the new and strange transfer of the term stoicheion to the elemental sense.

Created on 4 July 2005 at 14.35 hours page 387

On the Use of Stoicheion in the Sense of ‘Element’

387

in particular, if he is repeating the term stoicheion as it is used by the progenitors of the dream theory, then the use of οBονπερεA may indicate that he is unhappy with the way the term is being used. He may be unhappy, for instance, because he believes that the ‘dream’ theorists identify the wrong sort of things as stoicheia. This suggestion would tie in very nicely with Plato’s criticism in the Timaeus of the use of archai and stoicheia with reference to fire, air, water, and earth. But, unfortunately, this seems unlikely: the things identified in the Theaetetus as stoicheia, however we may interpret them, seem comfortably to match the requirements for elemental status, unlike the Timaean conception of fire, air, water, and earth. There is perhaps another possibility, however, one which does not require that we make a decision upon the provenance of the ‘dream’ theory. Why Socrates uses the expression οBονπερεA may ultimately resist complete clarification, but nevertheless I think it can be established that the expression does not have to imply either reticence over the use of the term stoicheion or a warning that one is about to trade in metaphors. Plato, remarkably, never uses this expression anywhere else, but he does use the similar expression VσπερεA, and interestingly, he uses this latter expression in the Cratylus with reference to stoicheia. So let us turn now to the Cratylus, to see if it throws any light upon Theaet. 201 e. For much of the Cratylus Socrates is conducting an etymological enquiry into names, or words (Xν;µατα).57 He has been analysing names into smaller component names, and he now begins to wonder at what point one should call a halt to such a procedure. For instance, once one has analysed a name into its components, someone might ask for the names from which the component names are formed, and so on indefinitely. The question is: when would the answerer be justified in stopping? Socrates asks his interlocutor Hermogenes: Wouldn’t it be when he reaches the names that are, as it were, the elements [2 Vσπερε1 στοιχεα] of all the other statements and names? For if these are indeed elements [στοιχεα], it cannot be right to suppose that they are composed out of other names . . . if we ever get hold of a name that isn’t composed out of other names, we’ll be right to say that at last we’ve reached 57 The Greek [νοµα has no exact English translation: in the Cratylus, Xν;µατα include proper names, nouns, adjectives, and, at 427 c, adverbs.

Created on 4 July 2005 at 14.35 hours page 388

388

Timothy J. Crowley

an element, which cannot any longer be carried back to other names (422 a– b, trans. Reeve)58

In what sense is Socrates using stoicheion here? Both before and after this passage, Socrates frequently uses stoicheion in the familiar alphabetic sense, e.g. at 393 d–e, and then again at 424 c–e, 426 d, 431 e, and 433 a. In these passages stoicheion straightforwardly means ‘letter of the alphabet’ or ‘phoneme’. But at 422 a–b it is not used in the alphabetic sense. A stoicheion now means a certain sort of name or a word, in particular one of the primary names into which other names may be analysed, and which do not admit of analysis into names that are more primary. But if stoicheion at 422 a–b is not used in the alphabetic sense, can we at least say that it is derived metaphorically from the alphabetic sense? I think to say so would be an error. A metaphorical use is, strictly speaking, the transfer of a term, belonging to one sphere of discourse, to another sphere, where it is used ‘strangely’. But if ‘letter’ is the controlling sense of stoicheion, then to refer to certain kinds of words as stoicheia is not metaphorical. For it would be to use a term that belongs to a certain sphere of discourse, in this case grammar, in a di·erent sense within that very same sphere of discourse. And while this would, perhaps, be a ‘strange’ use, it would be so only because it is confusing. It seems rather that at 422 a Socrates is using the term stoicheia neither in the alphabetic sense, as he has earlier in the dialogue, nor in a sense derived from the alphabetic sense; but rather that he is now drawing on a more basic, and yet also familiar, sense of stoicheion. This, I would suggest, is the general sense of ‘primary constituent’. And this is a familiar sense: as Aristotle says in De caelo 3. 3, what everyone in every case would understand by stoicheion is ‘something into which compound things can be analysed’ (305B17–18).59 In the Cratylus compound names are analysed into their primary names, and so to call these latter the stoicheia of the former would be a typical, or ‘ordinary’, use of the term. Occasional uses of stoicheia, qualified by VσπερεA, 58 C. D. C. Reeve, Plato’s Cratylus: Translated, with an Introduction and Notes (Indianapolis and Cambridge, 1998). 59 Cf. Aquinas: ‘every one intends to mean by “element” something such as has been described, no matter what the field, for example, in corporal speech, and in demonstrations, in which the principles are called “elements” that are not resolved into other principles’ (‘Aquinas’ Exposition of Aristotle’s Treatise on the Heavens’, trans. by Father Pierre Conway and F. R. Larcher (unpublished but circulated in photocopied form; New York, 1963–4), bk. 3, Lec 8, Sct 600, p. 3-27).

Created on 4 July 2005 at 14.35 hours page 389

On the Use of Stoicheion in the Sense of ‘Element’

389

can be found in later Greek authors, and again on these occasions there is no need to think that the term must be understood as being metaphorically derived from the alphabetic, nor for that matter any other, sense of stoicheion. Plutarch, for instance, says that hope of reward and fear of punishment are ‘the elements, as it were’ (Vσπερε1 στοιχεα) of virtue (De liberis educandis 12 c 12). He means, surely, that these things are the primary constituents of virtue— there is certainly no need to interpret what he means according to the analogy of letters of the alphabet (cf. Iambl. Comm. math. 22. 13; Myst. 1. 4. 21). Does this mean that Plato in the Cratylus is drawing on the core meaning of stoicheion as identified by Aristotle in Metaphysics Delta? Yes and no. Consider again that the core meaning of stoicheion, stated fully, is ‘primary constituent, into which other things may be analysed, and which is not itself analysable into things different in form’. Plato calls the primary names stoicheia at 422 a–b because the primary names do not admit of analysis into names that are more primary. But even if analysis cannot retrieve further, more basic, names, it remains possible to analyse the primary names into other things that are di·erent in form. Indeed, Plato proceeds in the Cratylus to do just that; he analyses the names into the primary sounds, the syllables and their stoicheia, i.e. stoicheia in the alphabetic sense (424 b–e). So there appears to be a lack of fit between such stoicheia and the core meaning of the term. But arguably it would be out of place, in an enquiry into the primary names, to proceed all the way down to letters or phonemes. In order to analyse the correctness of these primary names, a di·erent sort of investigation is called for; and this investigation does indeed involve examining the stoicheia of the primary names. By identifying the primary names as stoicheia, on the grounds that further analysis will not produce further names, Socrates is perhaps operating with a general notion of stoicheion that lacks the stricter sense that Aristotle later gives it, namely, that it should be indivisible into things di·erent in form. This notion is general enough to be applied to a wide variety of things, just in case these things are simple and constitute some complex item.60 But in any case it is nevertheless clear that the term is being used with, and depends upon, reference to the core meaning of stoicheion; the way it is used in the Cratylus may lack precision, but this does not render it metaphorical. 60 Cf. n. 13.

Created on 4 July 2005 at 14.35 hours page 390

390

Timothy J. Crowley

Now how does the Cratylus aid our understanding of Theaetetus 201 e? What I would suggest is that the expression οBονπερε1 στοιχεα in the Theaetetus ought to be understood along the lines of Vσπερε1 στοιχεα; in other words, it is an indication that Socrates is drawing on the core sense of stoicheion. Hence the use of stoicheia in the ‘dream’ theory ought to be readily understandable without presuming an appeal to metaphor. This suggestion, however, faces an immediate problem, namely, Socrates’ claim at 202 e that the elemental sense of stoicheion is based on an alphabetic model. But let us consider Theaetetus’ reaction to Socrates’ claim. Theaetetus agrees, apparently after some hesitation, that the author of the dream theory had the alphabetic sense of stoicheion in mind as the model of the stoicheia of the theory. Now it may well be significant that Theaetetus seems to hesitate before agreeing with Socrates on the derivation of stoicheion. Prior to 202 e, Socrates says that the things composed of stoicheia are complexes (συγκεAµενα), and a few lines later he refers to these complexes as syllabai or ‘syllables’ (202 b). As we have seen, Aristotle in the Protrepticus claims that the Platonists commonly call compounds of elements syllabai. This use of syllab»e, I argued in Section 2, is a metaphorical derivation from the use of the term in alphabetic or generally linguistic contexts. But at 202 e it seems that Socrates has to clarify to Theaetetus his intention to consider the terms stoicheia and syllabai in their alphabetic senses. Perhaps this is simply because Socrates initially expresses his intention rather obscurely. But, on the other hand, it may imply that before 202 e they had been using these terms in some other sense, or at least that Theaetetus had understood stoicheion and syllab»e in some sense other than their alphabetic senses. If this were the case, then it would follow that the ‘dream’ theory as presented at 201 e to 202 c is readily understandable if stoicheion and syllab»e are taken simply to have the rather general senses of ‘constituent’ and ‘that which is constituted’; in other words, senses that are independent of the alphabetic sense. This may seem rather a lot to hang on an apparent hesitation. But in any case Theaetetus’ agreement regarding the derivation of stoicheion is certainly not enough to establish that the elemental sense is derived from the alphabetic sense. This is, after all, a Platonic dialogue, not a treatise; as in the Timaeus, Socrates may simply be playing on the fact that stoicheia has this ambiguity, that it has the meaning both of constituents, generally speaking, and of

Created on 4 July 2005 at 14.35 hours page 391

On the Use of Stoicheion in the Sense of ‘Element’

391

phonemes or letters. By 204 a Socrates seems consciously to depart from his alphabetic model, as he emphasizes that what he has said about stoicheia and syllabai holds both for γρµµατα and for ‘all other things’. The argument from this point on would appear to concern the relationship of parts to wholes considered generally, i.e. without special reference to the peculiar characteristics of alphabetic stoicheia.61 But why focus on an alphabetic model in the first place? A reason why he does so may be available at 206 b. Here Socrates states that it is by examining those stoicheia, and syllabai, of which we have experience that we can draw conclusions about the general class of stoicheia. In other words, Socrates fastens onto the familiar alphabetic sense of stoicheion in order more e·ectively to highlight certain, and to Socrates’ mind unsatisfactory, features of the stoicheia of the ‘dream’ theory (202 e–204 a). The latter evidently involves the use of stoicheion with reference to things with which we are less familiar. But that does not mean that the use of stoicheia in this sense is an unfamiliar use. At least two points emerge from this discussion of the Theaetetus. Firstly, it seems unlikely that, without the evidence of Eudemus’ report, anyone would have identified this passage at Theaet. 201 e as the first use of stoicheion in the elemental sense. On the other hand, it must be conceded that it is very di¶cult to establish conclusively from the evidence of the Theaetetus itself that stoicheion is already commonly used in the sense of ‘element’. This is because at Theaet. 201 e and onwards it is very di¶cult to say with certainty when, or even whether, Socrates is using stoicheion in the elemental sense.62 It seems best to think that the sense that Socrates has in mind throughout is that of the core meaning of the term.63 There is a core, or general, meaning of the term stoicheion, which is the basic sense of ‘primary constituent’, and there are certain uses of this term that are ‘ordinary’ or familiar, and one of these is the alphabetic sense. Nevertheless, on the basis of Aristotle’s evidence, and also that of the Timaeus and the Sophist, it seems clear that the elemental sense of stoicheion is also an ‘ordinary’ sense, and moreover it is a sense that is already available for use by Plato. In the 61 Cf. D. Gallop, ‘Plato and the Alphabet’, Philosophical Review, 72/3 (1963), 364–76 at 373. 62 Diels, Elementum 19, and Lagercrantz, Elementum, 17, suggest that at 201 e and thenceforward stoicheion has a double sense, of both ‘letter’ and ‘basic material’. 63 Cf. Campbell’s sober assessment of 201 e and the subsequent passages, Theaetetus, 213 n. 1.

Created on 4 July 2005 at 14.35 hours page 392

392

Timothy J. Crowley

Theaetetus and Timaeus Plato turns to the alphabetic sense of stoicheion in order to clarify certain points about the sort of things that his contemporaries identify as the stoicheia of nature and natural things. It may be that the elemental sense is not as familiar as the alphabetic sense and perhaps not even as familiar as the geometric sense; but it is an ‘ordinary’, as opposed to a metaphorical, sense none the less. It is not, however, that any use of stoicheion other than the alphabetic, geometric, or elemental senses must be deemed metaphorical. If what is called a stoicheion is the primary constituent of something, then that use of stoicheion is not metaphorical. So, for instance, the use of stoicheia at Crat. 422 a is not metaphorical. This point, I think, also answers a question raised earlier, regarding the use of stoicheion by Plato’s contemporaries Isocrates and Xenophon in the sense of ‘elementary rules’ or parts of a particular discourse. Such uses are not metaphorical, because they are based upon the core meaning of the term stoicheion. Admittedly, one might not think of a set of rules or the fundamental branches of a discipline as ‘constituents’ of that discipline, in the way, for instance, that phonemes constitute syllables. But what these uses do suggest is the organization of stoicheia into an order, or a comprehensible whole. The core sense of stoicheion, then, is that of a basic part of a whole.64 University College, Oxford

B I B L I O GR A P HY Ackrill, J. L., Aristotle’s Categories and De Interpretatione: Translation with Notes (Oxford, 1963). Aquinas, St Thomas, Commentary on the Metaphysics of Aristotle, trans. by J. P. Rowan (Chicago, 1964). ‘Aquinas’ Exposition of Aristotle’s Treatise on the Heavens’, trans. by Father Pierre Conway and F. R. Larcher (unpublished but circulated in photocopied form; New York, 1963–4). Bostock, D., Plato’s Theaetetus [Theaetetus] (Oxford, 1988). Brandwood, L., ‘Stylometry and Chronology’ [‘Stylometry’], in R. Kraut (ed.), The Cambridge Companion to Plato (Cambridge, 1992), 90–120. Burkert, W., ‘Stoicheion’, Philologus, 103 (1959), 167–97. 64 Indeed, it is sometimes thought that stoicheion literally means a member of a series or ‘one of a row’, from στοχος, a row; see Diels, Elementum, 58; cf. McDowell, Theaetetus, 239.

Created on 4 July 2005 at 14.35 hours page 393

On the Use of Stoicheion in the Sense of ‘Element’

393

Burnet, J., Early Greek Philosophy [EGP], 4th edn. (London, 1930). Burnyeat, M., The Theaetetus of Plato [Theaetetus] (Indianapolis, 1990). Bury, R. G., Plato, vii. Timaeus (London and Cambridge, Mass., 1929). Campbell, L., The Theaetetus of Plato: A Revised Text with English Notes [Theaetetus] (Oxford, 1883). Charlton, W., Aristotle’s Physics Books I and II, Translated with an Introduction and Notes (Oxford, 1970). Cornford, F., Plato’s Theory of Knowledge [PTK] (London, 1935). Crubellier, M., ‘Metaphysics Λ 4’, in M. Frede and D. Charles (eds.), Aristotle’s Metaphysics Lambda (Oxford, 2000), 137–60. Diels, H., Elementum (Leipzig, 1899). D•uring, I., Aristotle’s Protrepticus: An Attempt at Reconstruction [Protrepticus] (Studia Graeca et Latina Gothoborgensia, 12; G•oteborg, 1961). Furley, D., The Greek Cosmologists, i. The Formation of the Atomic Theory and its Earliest Critics (Cambridge, 1987). Gallop, D., ‘Plato and the Alphabet’, Philosophical Review, 72/3 (1963), 364–76. Guthrie, W. K. C., A History of Greek Philosophy, v. The Later Plato and the Academy (Cambridge, 1978). Heath, T. L., Mathematics in Aristotle (Oxford, 1949). Joachim, H. H., Aristotle: On Coming-to-Be and Passing-Away. A Revised Text with Introduction and Commentary (Oxford, 1926). Kahn, C. H., Anaximander and the Origins of Greek Cosmology [Anaximander] (New York, 1960). Kirwan, C., Aristotle: Metaphysics Books Γ, ∆, and Ε. Translated with an Introduction and Notes, 2nd edn. (Oxford, 1993). Koller, H., ‘Stoicheion’, Glotta, 34 (1955), 161–74. Lagercrantz, O., Elementum (Uppsala, 1911). Lee, H. D. P. (trans.), Plato: Timaeus (Harmondsworth, 1965). Lohmann, J., Musik^e und Logos: Aufs•atze zur griechischen Philosophie und Musiktheorie (Stuttgart, 1970). Lumpe, A., ‘Der Begri· “Element” im Altertum’, Archiv f•ur Begri·sgeschichte, 7 (1962), 285–93. McDowell, J., Plato’s Theaetetus [Theaetetus] (Oxford, 1973). Owen, G. E. L., ‘The Place of the Timaeus in Plato’s Dialogues’, Classical Quarterly, ns 3 (1953), 79–95. Reeve, C. D. C., Plato’s Cratylus: Translated, with an Introduction and Notes (Indianapolis and Cambridge, 1998). Ross, W. D., Aristotle’s Metaphysics: A Revised Text with Introduction and Commentary [Metaphysics] (2 vols.; Oxford, 1924). Aristotle’s Physics: A Revised Text with Introduction and Commentary (Oxford, 1936).

Created on 4 July 2005 at 14.35 hours page 394

394

Timothy J. Crowley

Ryle, G., ‘Letters and Syllables in Plato’, Philosophical Review, 69 (1960), 431–51. ‘Logical Atomism in Plato’s Theaetetus’, Phronesis, 35 (1990), 21–46. Schwabe, W., ‘ “Mischung” und “Element” im Griechischen bis Platon’, Archiv f•ur Begri·sgeschichte, suppl. 3 (1980). Sedley, D., The Midwife of Platonism: Text and Subtext in Plato’s Theaetetus (Oxford, 2004). Shields, C., Order in Multiplicity: Homonymy in the Philosophy of Aristotle [Homonymy] (Oxford, 1999). Skemp, J. B., Plato’s Statesman: A Translation of the Politicus of Plato with Introductory Essays and Footnotes (London, 1952). Taylor, A. E., A Commentary on Plato’s Timaeus [Timaeus] (Oxford, 1928). Vlastos, G., Plato’s Universe (Oxford, 1975). Vollgra·, W., ‘Elementum’, Mnemosyne2, 4 (1949), 89–115. Wehrli, F., Eudemos von Rhodos, 2nd edn. (Die Schule des Aristoteles, 8; Basel, 1969).

Created on 4 July 2005 at 14.35 hours page 395

I NDEX LOCORUM Aeschines Orator Against Timarchus 87: 164 n. 55 Aeschines Socraticus Alcibiades VIA.50 Giannantoni: 65 n. 46 Aetius • 1. 13. 1: 7 n. 8 1. 15. 3: 10 1. 15. 7: 11 1. 15. 8: 11 1. 17. 3: 7 n. 8 4. 9. 6: 23–4 n. 34 5. 22. 1: 7, 8 n. 12, 21 5. 26. 4: 8 n. 12 Alcmaeon, 24 DK A 5: 26 n. 38 A 10: 26 n. 38 Alexander of Aphrodisias In Aristotelis Meteorologicorum libros commentaria, ed. Hayduck 179. 1–6: 304 n. 17 In Aristotelis Topicorum libros octo commentaria, ed. Wallies 149. 9–15: 197 n. 4 Aquinas In Aristotelis De caelo, trans. Conway and Larcher bk. 3, Lec 8, Sct 600, p. 3-27: 388 n. 59 In Aristotelis Metaphysica, trans. Rowan vol. 1, bk. 5, Lsn 4, Sct 804, p. 317: 375 Aristo SVF i. 351: 133 n. 57

Aristophanes Ecclesiazousae 652: 374 n. 26 735: 5 n. 3 Aristotle Categories 1a1–4: 372 14a36–b2: 275 n. 129 14b1–2: 372 De anima 403a3–b19: 297 403b25–7: 250 404b8–15: 29 410a27–b1: 336 n. 50 412a27–412b10: 271 n. 122 415a26–b7: 360 n. 60 415b18–20: 271 n. 122 415b20–1: 358 n. 52 415b28–416a8: 352 n. 36 416b34: 296 418a11–12: 330 418b20–6: 20 420b5: 370 n. 13 423b27–424a10: 334 423b27–31: 303, 322 ·., 328 n. 41 424a2–7: 300, 326, 332 n. 46, 334 424a10: 323 424a18–19: 298 424a30–b2: 336 424a32–b3: 295, 299, 332 ·., 335, 338 424b1–2: 298 n. 5 424b28–9: 323 424b30: 323 424b31: 323 425a6: 335, 337 n. 53 425b23–4: 298 n. 5 428a11: 301 430b28: 301 434a29–30: 298 n. 5 434a30: 336 434b13–14: 331 435a10–19: 335 n. 48

Created on 4 July 2005 at 14.35 hours page 396

396

Index Locorum

435a19 ·.: 302 435a19–b1: 301, 302 ·., 304, 325, 332 ·., 335, 338 435a20–3: 302 n. 12 435a21–b1: 335 435a21–2: 335 435a22–4: 335 435a22–3: 324 n. 38, 336 n. 50 435a24–b1: 336 De caelo 268a13–14: 264 n. 111 271a33: 262 n. 109 275b31–276a1: 369 n. 10 302a10–21: 371, 375 305b17–18: 388 De generatione animalium 722a32: 377 n. 38 722b8 ·.: 352 n. 35 731b18–732a12: 360 n. 60 731b20–732a3: 343–4 n. 9 779b9: 5 n. 6 779b15–20: 21, 32 784a34–b3: 337 789b2–5: 359 n. 55 De generatione et corruptione 315b6–15: 369 n. 10 324b26–35: 23–4 n. 34 328a24–31: 308 n. 24 329a27–35: 306 329b30–1: 321 330a30–b5: 320 330b1–5: 306 330b3–5: 302 n. 15 334a26–31: 6 334b8–16: 307, 309 334b30–1: 316 336b25–32: 343–4 n. 9, 360 n. 60 De incessu animalium 704b12–18: 359 n. 55 De iuventute 469b11–13: 337 470a27–32: 331 De longitudine et brevitate vitae 467a6–9: 305 n. 18 [De mundo] 396b12: 264 396b13: 13 De partibus animalium 645b14: 271 n. 122 648b11–649b20: 322 648b11–649b7: 313 648b12–14: 321 n. 33 648b30–2: 314

649a27–9: 322 649b11–13: 319 660a11–13: 331 664b1: 370 n. 13 674a22–b5: 355 n. 46 694b12–17: 355 n. 46 696b23–34: 362 696b27: 364 n. 71 De sensu 437b23–438a5: 26 439b18 ·.: 18 440a10–12: 20 De somno 454b28: 299 Eudemian Ethics 1215b18–1216a3: 253 1216a2–10: 253 n. 98 1217b22: 245–6 n. 95 1218b32–6: 246 1218b34: 245–6 n. 95 1220a8–11: 272 n. 124 1244b23–1245a10: 251 1244b26 ·.: 250 Historia animalium 492a5: 5 n. 6 535a27: 370 n. 13 539a15–25: 348 n. 20 591b23–30: 362 n. 67 596b20 ·.: 355 n. 46 Magna moralia 1185a2: 253 n. 98 1192b15–16: 372 n. 18 Metaphysics 982b2–7: 274 984a8: 367 984b15–20: 257 985a32: 367 985b15: 369 n. 10 986a1: 373 n. 24 991a20 ·.: 373 992b13–18: 275 n. 130 993b24–6: 310 998a20 ·.: 373 n. 24 998a23–8: 372 n. 19 1001b26–1002b12: 275 n. 130 1010b11–14: 172 1013b17: 377 n. 38 1014a26–b15: 370–1 1014a26: 372–3 n. 21, 373 1014b3–4: 373 1014b15: 373 1023a26–b11: 307 n. 23 1026b30–6: 345 n. 14

Created on 4 July 2005 at 14.35 hours page 397

Index Locorum 1028a4: 370 1028b15–27: 275 n. 130 1028b25–8: 373 n. 24 1029b1–12: 275 n. 131 1041b13–31: 377 n. 38 1041b16–17: 377 n. 38 1044b3–14: 341 n. 2 1052a15: 370 1060b6–19: 275 n. 130 1069a26–8: 373 n. 24 1071b2–3: 358 n. 52 1076a28: 245–6 n. 95 1076b11–39: 275 n. 130 1077a24–36: 275 n. 130 1079b24 ·.: 373 1080a12–b36: 275 n. 130 1085a7–b34: 275 n. 130 1085a7–9: 275 1086b22–3: 377 n. 38 1087b1–13: 373 1087b9: 373 1087b13: 373 1090b5–32: 275 n. 130 1091a10: 373 1095b2–3: 275 n. 131 Meteorologica 374a7–8: 20 381b6: 264 382a11–21: 329 n. 42 382b13: 317 n. 31 383b18–20: 317 384a11–14: 317 384b30–1: 305 385a4–12: 315 388b12–13: 317 389a4–7: 315 389a7–9: 315 389a9–11: 317 389a11–16: 304 Nicomachean Ethics 1085a18: 69 1102a26–8: 246 1102a26: 245–6 n. 95 1102b2–11: 253 n. 98 1104b9–11: 154–5 n. 36 1104b12–14: 162 n. 51 1104b30–2: 139 1111b17: 154 n. 34 1135a3–5: 264 n. 112 1140a3: 245–6 n. 95 1143a8–9: 271 n. 123 1144a36–b7: 277 n. 132 1170a16 ·.: 250

1174a1 ·.: 253 n. 98 1178b18–20: 253 n. 98 1181a12–19: 264–5 Physics 194a21–2: 264 194a33–6: 358 194a34: 358 195a16: 377 n. 38 196a3–5: 354 197a2–3: 354 197a15–18: 354 n. 44 198b5–9: 359 198b10–199a8: 342 198b16–19: 346 198b18: 347 n. 17 198b21–3: 346 198b34–199a8: 342 198b34–199a3: 345 n. 14 198b34–6: 349 198b36–199a3: 347 199a5: 349 n. 23 199a6: 349 n. 23 199a16: 264 199b14–18: 351 n. 33 215b18–19: 374 217b31: 245–6 n. 95 Poetics 1456b20–2: 370 n. 13, 375 1457b1–4: 372 n. 18 1457b3–4: 372 n. 17 1457b6 ·.: 372 n. 16 1458a23–5: 372 1458a33: 372 n. 18 1458b17: 372 n. 18 Politics 1113a33: 277 n. 132 1138b21–5: 277 n. 132 1138b32–4: 277 n. 132 1254a33: 245–6 n. 95 1256b15–22: 356 1256b38: 356 1277a14–16: 277 n. 132 1278b31: 245–6 n. 95 1323a21–b21: 247 1323a22: 245–6 n. 95 Posterior Analytics 71b33–72a5: 275 n. 131 95a16–21: 318 Prior Analytics 68b35–7: 275 n. 131 scholium in MS Paris, BN Cod. Par. gr. 2064: 197 n. 4

397

Created on 4 July 2005 at 14.35 hours page 398

398

Index Locorum

[Problemata] 910a12–15: 21 Protrepticus: 193 ·. fr. B 33. 9 During: 383 • fr. B 35 During: 377 • fr. B 41 During: 286 n. 143 • fr. B 52 During: 286 n. 143 • fr. B 58 During: 286 n. 143 • fr. B 110 During: 286 • fr. 61 Rose: 287 n. 148 fr. 10a Ross: 255 fr. 10b Ross: 256–7 fr. 10c Ross: 286 n. 145 Rhetoric 1358a35: 374 1360b4–7: 69 1360b9–11: 280 1360b14–18: 248 1360b24–9: 246 1361a23–5: 268 n. 120 1362b2–12: 248 1362b29 ·.: 254 n. 99 1363a17: 277 n. 132 1363a21–4: 272 n. 125, 277 n. 134 1404b5–6: 372 n. 17 1404b31–2: 372 n. 17 1410b12–13: 372 nn. 17, 18 Sophistici elenchi 172b21: 374 172b31: 374 174a21: 374 Topics 105b20–1: 276 116a13–17: 277 n. 132 120b13: 373 n. 23 121b11: 373 n. 23 123a28: 374 123a33–6: 372 n. 18 129b33–4: 268 n. 118 130a19–21: 268 n. 119 139b34: 372 n. 18 143a13: 374 147a22: 374 151b18: 373 n. 23 158b12: 372 n. 18

Contra Iulianum 4. 15. 78: 256–7 De Trinitate 14. 19. 26: 286 n. 145 Boethius Consolatio philosophiae 3. 8: 197 n. 2, 255 Cicero Brutus 18. 70: 14 Codex Atheniensis 1249 fo. 110r: 20 David Prolegomena to Philosophy, ed. Busse 9. 2–12: 197 n. 4 Democritus, 68 DK A 135: 10, 11 Diocles of Carystus fr. 8 Wellmann: 12 Diodorus Siculus 2. 8. 2: 6 n. 7 Diogenes of Apollonia, 64 DK A 19: 11 A 29a: 11 Diogenes Laertius 3. 24: 367 n. 2 5. 22, title 12: 197 n. 3 7. 91: 201–2 n. 24 7. 129: 201–2 n. 24 7. 161: 133 n. 57 Dionysius of Halicarnassus De Isaeo 4: 5 n. 5

Asclepius 174. 9: 371 n. 14

Dissoi Logoi 3: 45, 59 n. 40 3. 4: 59 n. 40

Augustine Confessions 3. 4. 7–8: 197 n. 2

Elias Prolegomena to Philosophy, ed. Busse 3. 17–23: 197 n. 4

Created on 4 July 2005 at 14.35 hours page 399

Index Locorum Empedocles, 31 DK A 34: 6, 7 n. 8 A 37: 2 A 43: 6, 7 n. 8 A 57: 20 A 69a: 10 A 70: 8 n. 12 A 72: 2 A 78: 7, 8 n. 12, 21 A 86: 7, 8 n. 12, 10, 21, 23–4 n. 34, 24, 29, 30, 33 A 87: 23–4 n. 34 A 90: 23–4 n. 34 A 91: 21, 32 A 92: 10, 22–3 B 2: 3 B 3: 3 B 6: 2, 367 B 17: 2 B 17. 18: 367 n. 1 B 17. 21: 29 B 21: 2 B 21. 1–6: 9–10 B 21. 3: 19 B 23: 4, 7, 13, 14 B 33: 1 B 34: 6 B 35: 2 B 50: 21 B 61: 2 B 67: 19 B 71: 3–4 B 71. 4: 6 B 84: 24–5 B 93: 18 B 94: 20 B 96. 4: 6, 7, 8, 9, 19 B 98: 2, 7, 21 B 107: 29 n. 45 B 107. 1: 6 B 109: 29, 29–30 n. 46 B 112: 11 Euclid Elements 1. 1: 371 n. 14 1. 2: 371 n. 14 1. 3: 371 n. 14 1. 47: 371 n. 14 Eudemus, ed. Wehrli fr. 31: 367 n. 2

399

Euripides Helen 233: 6 n. 7 Troades 111: 6 n. 7 Galen [De humoribus] xix. 495 Kuhn: 11 • In Hippocratis De natura hominis commentaria xv. 49 Kuhn • = CMG 5. 9. 1. 27. 24: 7 n. 8 xv. 32 Kuhn • = CMG 5. 9. 1. 19. 7–12: 6 xv. 49 Kuhn • = CMG 5. 9. 1. 27. 24: 7 n. 8 De placitis Hippocratis et Platonis v. 627 Kuhn • = CMG 5. 4. 1 2. 462. 1–19: 29–30 n. 46 Herodotus 1. 98: 5 n. 3 2. 96: 6 n. 7 Hippocrates Medicus De natura hominis 4–7: 12 De o¶cio medici 25: 6 n. 7 De ossium natura 12: 6 n. 7 Hippocrates of Chios, 42 DK A 1: 375 Historia Augusta, ed. Hohl ii. 97. 20–2: 197 n. 2 Homer Odyssey 5. 248: 6 n. 7 5. 361: 6 n. 7 20. 17–18: 161 n. 46 Iamblichus De communi mathematica scientia 22. 13: 389 26–7: 202 n. 25 82. 19: 277 n. 133 De mysteriis 1. 4. 21: 389

Created on 4 July 2005 at 14.35 hours page 400

400 Protrepticus, ed. Des Places 41. 3–8: 204 41. 8–17: 204–5 48. 26–7: 204 61. 11: 272 n. 126 64. 28: 272 n. 126 65. 1: 286 n. 143 67. 14: 286 n. 143 67. 20–68. 14: 270 67. 23–68. 14: 271 67. 23: 245 n. 92, 272 68. 2–3: 270 n. 121 68. 2: 272 n. 126 68. 7: 271 68. 11–14: 274 68. 11–12: 272 68. 11: 271 68. 13: 271 68. 14–69. 19: 273 68. 14–21: 270 68. 14–18: 271 68. 14: 274 68. 18–21: 271, 276 68. 21–69. 2: 276 68. 21: 274 68. 28–69. 2: 274 69. 2–3: 273 n. 128 69. 3–10: 276 69. 10–19: 276 69. 20–70. 19: 276 69. 20–4: 270, 271 69. 27: 276 70. 9: 286 n. 143 70. 20–71. 12: 277 70. 20–2: 270, 271 70. 22: 245 n. 94 70. 23–4: 270 n. 121 70. 25: 277 n. 133 71. 9–12: 271 71. 13–22: 244 71. 18–22: 280 71. 18–21: 245 71. 20–1: 272 n. 126 71. 22–72. 3: 246 72. 3–72. 14: 246–7 72. 14–73. 17: 248 73. 17–74. 7: 249 74. 7–75. 13: 249–50 74. 20–1: 250 75. 14–24: 252 75. 14–16: 252 75. 15–16: 245 n. 92 75. 25–76. 26: 252–3

Index Locorum 76. 26–77. 11: 254 77. 12–27: 255 77. 12–13: 244 77. 12: 245 n. 94 77. 27–79. 6: 255–6 78. 12–79. 2: 257 78. 19–79. 2: 286 78. 21: 272 n. 126 79. 3–6: 256 79. 3–4: 245 n. 93 79. 7–84. 6: 258–61 79. 7–80. 20: 262 79. 7–8: 245 n. 92 80. 5 ·.: 262 n. 110 80. 7–9: 264 80. 18–20: 264 80. 20–81. 11: 262 81. 5–6: 259 n. 105 81. 11–82. 20: 262 81. 12: 260 n. 106 81. 18: 260 n. 107 82. 17: 260 n. 108 82. 20: 286 n. 143 84. 7–86. 9: 264 84. 9: 245 n. 92 84. 24–7: 278 85. 20–1: 262 n. 109 86. 4–5: 251 86. 10–89. 6: 265–7 86. 11–12: 245 n. 92 89. 5: 272 n. 126 89. 7–90. 15: 278–9 89. 8: 245 n. 92 89. 22–90. 2: 279 n. 136 89. 22–5: 280 89. 22–3: 272 n. 126 89. 22: 279 n. 136 89. 25: 279 89. 26: 279 n. 136, 285 n. 140 90. 16–110. 20: 225 90. 18–96. 10: 228 90. 18: 230 91. 19–21: 230 94. 17–19: 229 96. 10–13: 225 98. 25–30: 226 99. 2: 226 99. 20–100. 20: 234, 241 100. 19–20: 245 n. 91 100. 21–105. 27: 207 100. 23–4: 211 101. 14–15: 208 103. 22–4: 210

Created on 4 July 2005 at 14.35 hours page 401

Index Locorum 104. 105. 105. 105. 106. 107. 107. 107. 108. 108. 110. 111. 111. 113. 113. 113. 113. 113. 114. 114. 114. 115. 115. 119. 121. 124. 131. 142. 143.

3–7 6: 209 17: 209 28–110. 2: 214 24: 217 n. 51 8: 217 n. 51 13: 217 n. 51 24–5: 214 n. 46 21: 217 n. 51 22–109. 4: 215 3–111. 17: 216 18–113. 21: 218 21: 245 n. 90 22–115. 21: 221 22–114. 7: 279 n. 135 22–7: 221 27: 222 28: 222 4: 222 6–7: 221, 245 n. 90 7–22: 222–3 11–17: 223 22–121. 6: 235 21: 239 1–6: 239 16: 272 n. 126 16–132. 15: 205 22: 272 n. 126 25: 272 n. 126

Isocrates Ad filium Iasonis 8. 8: 374 n. 26 Ad Nicolem 16. 7: 374 n. 26 Antidosis 80: 264 n. 114 Megarians test. 25 D•oring = II A 32 Giannantoni: 133 n. 57 Olympiodorus In Platonis Alcibiadem commentaria, ed. Creuzer 144: 197 n. 4 Papyri P. Lit. Lond. 145: 231 Parmenides, 28 DK B 8. 41: 2

401

Philoponus In Aristotelis De anima libros commentaria, ed. Hayduck 176. 30–2: 8 n. 12 178. 6–8: 9 217. 10–15: 32 344. 33–7: 20 In Aristotelis Physica commentaria, ed. Vitelli 705. 22: 245–6 n. 95 Plato Alcibiades I 103 b–104 c: 47 104 b–c: 47 n. 21 105 b–c: 47–8 n. 22 109 e–110 d: 69 116 e–118 b: 63 120 e–124 b: 65 121 a: 68 123 d: 47–8 n. 22, 65 124 b: 66 129 b–131 b: 67 131 c: 47–8 n. 22 132 a: 255 n. 100 Apology 19 e 4–20 a 2: 86 n. 21 20 b 5: 86 n. 21 20 d 8–e 2: 86 n. 21 21 b 2–c 2: 94 21 b 6–7: 91 21 b 9–c 2: 91 21 c 1–2: 91 21 d 6–8: 94 21 e 2–4: 94 22 a 4: 94 22 c 4–d 3: 85 22 c 9–d 4: 91 23 a 5–c 1: 84, 87 23 a 5–7: 94 23 b 2: 85 23 b 4–c 1: 83 23 b 6–c 1: 89 23 c 1: 94 24 e 1–25 c 3: 103 n. 48 28 a 5–b 2: 108 28 b 4–29 c 1: 96, 97 28 b 4–6: 88, 94 28 b 6–c 1: 100–1 n. 45 28 b 8–9: 88 28 c 6–d 5: 89 28 d 5–9: 89 28 d 5–6: 88 n. 25

Created on 4 July 2005 at 14.35 hours page 402

402 28 d 6–9: 96 28 d 9: 96 28 d 10–29 a 2: 89 28 e: 57 28 e 4–29 a 2: 89 28 e 5: 89, 92 29 a 2–4: 89 29 b 5–7: 88, 94, 95 29 b 6–7: 92 29 b 6: 89 29 b 7: 88 n. 24 29 c 7–d 5: 94 29 d 2–4: 89 30 a–b: 67 30 a 1–2: 89 30 a 5–7: 90 30 a 5: 89, 92 30 a 6–c 2: 94 30 a 6–7: 89 30 a 7–b 5: 89 30 a 7–b 4: 108 30 e 1–31 b 5: 84 30 e 1–31 a 2: 94 31 a 3–7: 85–6 n. 20 31 a 3: 85–6 n. 20 31 a 10–b 1: 85–6 n. 20 31 c 8: 90 31 d 2: 90 31 d 3–4: 90 32 a 1–3: 90 33 a 5–6: 83 33 b 3–5: 108 33 c 4–8: 89–90 n. 27 33 c 5: 89, 92 37 c 5–d 2: 108 37 e 3–38 a 3: 93 37 e 3–4: 93 37 e 5–38 a 8: 93 38 a 1–8: 93 38 a 5–6: 78, 85 n. 19 40 a 5–6: 90 40 b 2: 90 40 b 7–c 4: 90 40 c 2: 90 41 a–c: 68 41 c 9–d 3: 234 Charmides 153 a: 57 160 e 4–5: 155 n. 37 167 e 1–2: 154 n. 34 169 e–171 c: 63 171 d–172 a: 63 172 c–175 a: 63

Index Locorum 175 c 9–d 6: 84 Clitophon 407 a–e: 41 n. 7 410 a–b: 56 n. 36 Cratylus 393 d–e: 388 393 d: 375 408 d: 380 n. 45 422 a–b: 388, 389 422 a: 392 424 b–e: 389 424 c–e: 388 426 d: 375, 388 427 c: 387 n. 57 431 a: 375 431 e: 388 433 a–b: 375 433 a: 388 434 b 1: 5 n. 3 Crito 44 b 8–c 5: 102 45 d 8–46 a 1: 102 46 b 1–48 a 10: 103 46 b 4–6: 111 46 e 1: 103 47 a 2–48 a 10: 96 47 a 3: 103 47 a 4: 104 47 a 7: 104 47 a 8–d 7: 98 47 a 9–10: 104 47 a 13–b 12: 104 47 b 5: 103 47 b 9: 103 47 c–48 a: 67 47 c 1–7: 104 47 c 1–3: 103 n. 47 47 c 8–d 3: 100–1 n. 45 47 d 1–5: 104 47 d 1: 103 47 d 2: 82, 103 47 d 4–7: 81 47 d 6–48 a 4: 104 47 d 10: 103 47 e: 67 48 a 5–7: 105 48 a 5–6: 81 48 a 6: 82 48 a 8–10: 103 48 a 9: 103 49 a–e: 55 49 a 4–e 3: 99 49 c 10–d 5: 107

Created on 4 July 2005 at 14.35 hours page 403

Index Locorum 49 c 10–d 2: 108, 111 49 d 2: 103 n. 48 49 d 3–5: 108, 112 50 d 1–51 c 5: 98 50 d 1–e 1: 98 50 e 1–51 a 2: 98 51 b 2–4: 98 Epistle VII 341 b–344 c: 51 Euthydemus 277 e–278 b: 121 278 e–282 b: 67, 69 281 a–e: 67 Euthyphro 6 e 4–7: 82 6 e 10: 120–1 n. 17 11 b–d: 68 11 d 5–6: 85 11 d 6–7: 84 11 e 7 ·.: 120–1 n. 17 12 a 1: 120–1 n. 17 Gorgias 449 a: 165 n. 57 449 c–d: 165 n. 57 449 c: 165 n. 57 455 c–d: 165 n. 57 455 d ·.: 165 n. 57 458 b–c: 165 n. 57 458 d 7–e 1: 165 n. 57 460 a 3–4: 165 n. 57 463 d 4: 144 n. 18 464 b 1–465 d 6: 81, 98 465 a 2–6: 82, 83 467 b 3–4: 143 n. 16 467 b 10: 142 n. 15 468 a ·.: 133 n. 55 469 b: 140 n. 8 469 c 7: 142 n. 16 470 a: 143 n. 17 470 c 4–5: 142 n. 15 470 d: 140 n. 8 470 d 5: 142 471 e 1: 142 n. 15 472 d–e: 140 n. 8 473 b 12–c 5: 143 473 d 3: 145 n. 22 473 e 1: 142 n. 15 473 e 4–5: 142 n. 15 474 b 2–5: 140, 141, 147, 151–2 n. 29 474 d–475 b: 140 474 d 6–8: 140–1 n. 10 474 e 2–3: 140–1 n. 10

403

475 a 3: 140–1 n. 10, 145 n. 21 475 c 3: 144 475 e 3–5: 140, 141, 147, 151–2 n. 29 477 a–e: 67 477 a 1–4: 145 477 c–e: 145 479 b 6–78: 144 480 e–481 b: 57 n. 38 480 e 1–2: 141 n. 11 480 e 1: 149 481 b: 163 481 b 4–5: 57 n. 38 482 b 5–6: 151 n. 28 482 c ·.: 141 n. 12 482 e 2: 163–4 n. 54 483 a–b: 160 n. 45 483 a 6: 163–4 n. 54 483 a 7 ·.: 164 483 a 7–8: 166 n. 58 483 a 7–b 4: 160 n. 45 483 b 4–5: 165 483 c 2: 165 483 e 4–484 a 2: 166 483 e 6: 166 484 d 1–2: 160 n. 45 485 a 5: 160 n. 45 486 a 4 ·.: 160 n. 45 487 a 3: 150 487 e 4–5: 150 489 b 8: 160 n. 45 491 e–492 a: 149 491 e 6 ·.: 160 n. 45 492 e–494 a: 205 492 e 3–494 b 6: 218 492 e 3–7: 218 492 e 7–494 a 2: 218 492 e 7: 213 n. 43, 219 n. 53, 237 492 e 8: 237 493 a ·.: 162 n. 50 493 a–494 a: 149 493 a 1: 237 nn. 84, 86 493 a 6: 237 n. 84 493 a 7: 237 n. 85 493 b 1: 237 nn. 84, 86 493 b 4: 220 493 b 5: 237 493 c 4–5: 220 493 c 4: 237 n. 86 493 c 7–d 1: 220 493 d 2–4: 220 493 d 5–6: 220 493 e 7: 237 nn. 84, 86

Created on 4 July 2005 at 14.35 hours page 404

404 493 e 8: 237 nn. 84, 86 494 a 3–b 7: 219 494 c–e: 149 494 c 5: 150 494 d 2–4: 150 494 d 4: 151 494 e: 146 n. 23 494 e 7–8: 150 495 a 2: 150 495 a 5–6: 150 495 b 4–5: 150 n. 27 495 b 5: 151 495 e–497 d: 149 497 e–499 b: 149 499 b 5: 151 499 b 6–8: 151 500 a 2–6: 106 500 a 4–6: 107 500 a 8–b 5: 106 500 e 4–501 c 5: 83 501 a 1–b 1: 109 n. 53 503 a: 144 n. 18 503 a 7–9: 144 n. 18 504 a–505 b: 205 504 a 2–c 10: 106 504 a 8–505 b 12: 221 504 a 8–d 7: 222 504 d 1–e 5: 106 504 d 5–505 a 9: 224 504 d 5–e 4: 81 504 d 5–9: 224 504 d 6: 82 504 e 4–5: 224 504 e 6: 224 504 e 7: 237 nn. 84, 86 504 e 8: 237 n. 86 504 e 9–505 a 1: 224 504 e 9–10: 224 n. 56 505 a–b: 162 n. 50 505 a 3–5: 224 505 a 3: 237 n. 86 505 a 6: 237 nn. 84, 86 505 a 9–11: 224 505 b 1–12: 223 505 b 4: 237 505 b 7: 237 505 b 11: 237 n. 84 506 c–508 a: 205 506 c 6–508 a 8: 235 506 c 6–8: 236 506 c 7–8: 236 506 c 9: 236 n. 80 506 d 2: 236 n. 80

Index Locorum 506 d 4: 236 n. 80 506 d 6: 237 n. 85 506 d 8: 236 n. 81 506 e 1: 213 n. 43, 237 506 e 2: 236 n. 81 506 e 4: 236 506 e 5: 236 n. 80 507 a 1: 236 n. 80 507 a 2–4: 236 n. 81 507 a 3: 236 507 a 4: 236 507 a 5: 236 507 a 7: 236 n. 80 507 b 4: 236 n. 80 507 c 1: 236 507 c 5–7: 236 n. 82 507 c 8–9: 236 n. 82 507 c 9: 237 n. 85 507 e 2: 237 n. 85 507 e 5: 237 n. 84 507 e 6: 236 508 a 3: 236 508 a 4–6: 236 n. 82 508 a 7–8: 236 508 c 2: 82 508 d 4: 160 n. 45 508 e: 144 n. 18 511 c–512 b: 67 512 a–b: 67 515 c 4–517 a 6: 87 n. 23 517 b 5–c 2: 86 n. 21 517 c: 144 n. 18 517 c 5–518 c 1: 83 517 c 7–518 c 1: 81, 98 517 d 6–518 a 1: 96 518 a 1–5: 96 518 c 1–e 1: 105 522 a: 157–8 n. 41 522 c 4–6: 160 n. 45 527 e 3–5: 216 548 e ·.: 161 553 d 4–6: 161 Hippias Major 284 d 1–e 9: 100 n. 43, 103 n. 48 299 a: 144 n. 19 304 d 5–e 3: 107 Hippias Minor 366 c: 53 n. 30 373 c 6–375 c 3: 97 376 b 4–c 6: 61 Laches 181 a: 57 184 e 11–185 a 3: 81, 87

Created on 4 July 2005 at 14.35 hours page 405

Index Locorum 184 e 12–185 a 3: 83 185 a 1: 82 185 b 11: 82 185e 4: 82 186 c 1: 84 186 e 1: 84 186 e 2: 84 187 d 1–188 a 5: 93–4 n. 34 189 e 1–3: 83 189 e 3–190 c 6: 82 190 b 7–d 1: 83 191 d 6–e 1: 154–5 n. 36 192 c–d: 144 n. 18 194 a 1–5: 132 198 a 7–9: 131 n. 51 199 c 3–4: 125–6 n. 30 199 e 11: 84 200 a 1–201 b 5: 109 200 c 7–d 4: 84 200 e 3–6: 87 201 a 2–b 5: 83 201 a 2–6: 81 201 a 3–4: 87 201 a 5–6: 87 Laws 633 c 9–d 1: 154–5 n. 36 660 e–661 c: 205 660 e 2–661 d 3: 239–40 660 e 2–661 c 5: 235 660 e 2: 239 661 b 4–5: 240 661 c 5–8: 239 661 e 6–662 a 7: 239 671 c: 166 n. 59 732 e 3: 270 Lysis 223 b 5–7: 84 Menexenus 236 b–c: 237–8 246 d–247 b: 205 246 d 1–247 b 7: 235, 238 246 d 1–3: 238 246 d 4: 238 n. 87 246 d 5–6: 238 n. 88 246 d 8: 238 n. 87 247 a 3: 238 n. 87 247 a 4: 238 n. 87 247 a 7: 238 n. 87 247 b 1: 238 n. 87 247 c 1: 238 247 e: 238 247 e–248 b: 205 247 e 5–248 b 4: 235, 238

247 e 5: 238 247 e 7–8: 234 248 b 1–2: 238 n. 88 249 d–e: 238 Meno 76 c 4–d 5: 22–3 77 b ·.: 144 n. 18 84 b–c: 40 87 d–89 a: 67 96 c 1–100 c 2: 87 n. 23 97 e 2–98 a 8: 109 98 a 4–5: 109 99 b 11–100 b 4: 87 n. 23 [On Justice]: 45, 49 374 c–d: 58 Phaedo 60 c–61 b: 51 n. 27 64 a–69 d: 205 64 a 4–69 d 2: 225, 227 64 a 4–65 d 2: 228 64 a 6–c 3: 230 64 b 9: 231 n. 68 64 c 4–8: 230 n. 65 64 c 7: 232 n. 74 64 c 8–d 2: 230 n. 63 64 d 2–4: 230 n. 65 64 d 4: 232 nn. 74, 75 64 d 5–9: 230 n. 64 64 d 5: 231 n. 67 64 d 11–12: 230 n. 65 64 e 2–4: 230 n. 63 64 e 7–8: 230 n. 63 65 a 3–4: 230 n. 63 65 a 4: 229, 231 n. 67 65 a 5: 233 n. 76 65 a 8–10: 230 n. 63 65 b 1–3: 230 65 b 1: 230 n. 65 65 b 2: 230 n. 65 65 b 3: 232 n. 74 65 b 6–9: 230 n. 63 65 b 9: 230 n. 65, 231 n. 66 65 c 1–4: 230 n. 64 65 c 5–6: 231 n. 69 65 c 5: 231 n. 67 65 c 7: 232 n. 72 65 c 8: 231 n. 66 65 c 10: 229 n. 62 65 d 4–e 1: 228 65 e–67 b: 227 65 e 1–2: 230 n. 65 65 e 2–67 d 10: 228 65 e 6–7: 230 n. 63

405

Created on 4 July 2005 at 14.35 hours page 406

406 65 e 8: 231 n. 69 65 e 9: 232 n. 74 66 a 7–b 1: 230 n. 64 66 a 7: 229, 232 n. 74 66 b 2: 231 n. 66 66 b 3: 231 nn. 66, 68 66 b 4: 231 n. 69, 232 66 b 5: 231 n. 66 66 b 6: 232 n. 72 66 c 2: 231 n. 68 66 c 4: 231 n. 66 66 c 8–9: 231 n. 69 66 d 5: 231 n. 68 66 e 3: 231 n. 66 66 e 5: 231 n. 66 67 a 1: 231 n. 71 67 a 6: 231 n. 67 67 a 8: 231 n. 66 67 b 3–c 4: 230 n. 64 67 b 3: 229 67 b 4: 231 n. 68 67 c 5–d 2: 230 n. 65 67 d 1: 232 n. 74 67 d 3–6: 230 n. 64 67 d 5: 231 n. 67 67 d 9: 231 n. 67 67 d 11: 229 67 d 12–68 c 4: 228 67 e–68 b: 227 67 e 5–68 b 6: 228 67 e 5: 229 68 c 5–69 d 2: 228 68 c 5–6: 230 n. 65 68 c 5: 229 68 c 7–8: 230 n. 63 68 c 8: 231 n. 69 68 c 10–12: 230 n. 65 68 d 1–2: 230 n. 63 68 d 2: 231 nn. 67, 69 68 d 4–5: 230 n. 63 68 d 6: 232 n. 72 68 d 8: 231 n. 66 68 d 12: 232 n. 72 68 e–69 d: 227 68 e 3: 232 n. 72 68 e 4: 232 n. 72 68 e 7: 231 n. 69 69 a–c: 131 n. 50 69 a 3: 231 n. 66 69 a 5–6: 230 n. 63 69 a 6: 229, 231 n. 68 69 a 8: 213 n. 43, 233 n. 76 69 a 10: 231 n. 67

Index Locorum 69 b 2–3: 231 n. 69 69 b 6: 231 n. 67 69 b 7: 231 n. 66 69 b 8: 231 nn. 66, 67, 71 69 c 2: 231 n. 67 69 c 3: 233 n. 76 69 c 4: 232 n. 72 69 c 8: 232 n. 74 82 a 7: 227 n. 61 82 a 11–b 3: 131 n. 50 82 b 10–84 b 3: 225, 226 82 b 10: 226, 227 82 c 1: 232 82 c 2: 227 n. 61 82 c 3: 231 n. 67, 232 n. 73 82 c 4: 231 n. 66 82 c 9–d 2: 227 n. 60 82 c 9: 227 n. 61 82 d–84 a: 231 82 d 2: 231 n. 66 82 d 3: 231 n. 66 82 d 8–9: 227 n. 60 82 d 8: 227 n. 61 82 e 7: 227 n. 61 83 a 4: 231 nn. 66, 69 83 a 6: 231 n. 69 83 b 4: 231 n. 69 83 b 6: 231 n. 69 83 b 7–8: 213 n. 43, 233 n. 76 83 b 7: 232 n. 74 83 c 2: 231 n. 69, 232 n. 73 83 c 4–5: 227 n. 60 83 c 6: 231 nn. 66, 68, 233 n. 76 83 c 7: 231 n. 66 83 c 8–10: 227 n. 60 83 c 8: 213 n. 43 83 d 2: 231 n. 69 83 d 3–4: 227 n. 60 83 d 8: 231 n. 69 83 d 9: 231 n. 69 83 d 10: 232 83 e 1: 231 n. 66 83 e 4: 227 n. 60 83 e 5: 227 n. 61 83 e 6–84 a 2: 227 n. 60 83 e 6: 231 n. 69 84 a 5: 231 n. 69 84 b 1: 213 n. 43, 233 n. 76 84 b 3: 226, 231 n. 67 100 c 4–6: 123–4 n. 27 102 d: 123–4 n. 27 107 c–d: 205 107 c 1–d 5: 225

Created on 4 July 2005 at 14.35 hours page 407

Index Locorum 107 c 1–2: 226 107 c 1: 226 nn. 58, 59, 231 n. 69 107 c 2: 226 n. 59, 231 n. 69 107 c 3: 226, 231 nn. 66, 69 107 c 5: 226 n. 59, 231 n. 69 107 c 6: 226 nn. 58, 59, 231 n. 68 107 d 2–5: 226 n. 59 107 d 4: 226 n. 59, 231 n. 67 107 d 7: 226 n. 58 114 c 6–9: 234 114 d 8–115 a 3: 234 Phaedrus 246 b 2: 156 247 d 5–e 2: 126 n. 35 253 d 4: 156 253 d 6: 155 n. 37, 156 254 a 2–3: 156 n. 39 254 c 4–5: 156 n. 39 256 a 6: 156 n. 39 274 b–277 a: 51 274 c ·.: 129 Philebus 17 a–18 d: 375 29 a–b: 382 n. 50 29 a 10–11: 380 n. 45 36 c ·.: 161–2 n. 48 49 b: 56 n. 34 Protagoras 313 d 5–e 2: 81 313 d 5–e 1: 107 313 e 5–d 5: 106 319 a 7–320 c 1: 87 n. 23 320 d: 380 n. 45 322 c 2: 121 323 a 1–2: 121 324 d 7–325 a 4: 123–4 n. 27 325 a 2: 117 325 b 5–d 1: 115 329 b–c: 127 329 c ·.: 127 329 c–330 b: 117 ·. 329 c–d: 126 329 c 2–d 2: 118 329 c 5–d 1: 117 329 c 6–d 1: 116 n. 3 329 c 7–8: 127 329 c 7: 118 329 d 3–4: 118 329 d 4–8: 116 n. 3, 118 329 d 5–6: 119 n. 14 329 d 6–8: 126 n. 31 329 d 8–e 2: 118 329 e–330 a: 127

407

329 e 2–4: 116 n. 3, 118 329 e 2: 118 329 e 5–330 a 2: 119 330 a 3–4: 119 330 a 4–b 1: 119 330 b 1–2: 119 330 b 3–6: 119, 121 330 c–e: 123–4 n. 27 330 c 2–e 2: 123, 126 330 d 5–e 1: 123 330 d 8–e 1: 123–4 n. 27 330 e 7–331 a 5: 121 330 e 7–331 a 1: 117, 123 331 b 1–3: 120 331 b 4–6: 121, 122 331 b 5–6: 123 n. 26 332 a 2 ·.: 121 333 b–334 c: 122 333 b 4–5: 122 333 b 5–6: 122 333 b 6: 121 n. 20 334 c–338 e: 122 338 e–347 a: 122 347 a–348 c: 122 349 a–c: 117 349 b–c: 126, 127 349 b 1: 117 n. 8 349 b 3–4: 123 349 c 2–5: 119 n. 15 349 c 4–5: 116 n. 3 350 c 4–5: 122, 125–6 n. 30 350 c 6 ·.: 122 352 c 2–d 4: 129 n. 47, 131 352 d 4: 129 353 b 1–3: 130 n. 48 353 b 2: 125 356 c 4–357 c 1: 123–4 n. 27, 128 357 b 5–6: 128 359 a 2 ·.: 125, 130 360 a 4–5: 126 n. 32 360 b 2–3: 126 n. 32 360 b 7: 126 n. 32 360 d 4–5: 125 360 e 6–361 d 6: 84 360 e 8–361 a 3: 82 361 a 4: 126 361 b 1–2: 124 n. 28 361 b 5–6: 126 n. 36, 129 361 c 1–d 6: 86 Republic bk. 1: 45, 51 331 c: 59 n. 40 331 e–336 a: 55

Created on 4 July 2005 at 14.35 hours page 408

408

Index Locorum

338 c 4–d 2: 91 n. 30 347 b 5–e 7: 86 n. 21 345 b 7–e 4: 86 n. 21 347 d 4–6: 82 354 a 12–c 3: 84 375 a–c: 56 n. 34 377 b 2: 166 377 c 3–4: 166 378 c 2: 166 n. 60 388 d 6: 166 n. 60 395 c–d: 167 n. 62 401 b–c: 167 n. 62 401 e 4 ·.: 166 n. 60 401 e 4–402 a 4: 167 403 c 6: 166 n. 60 410 b–412 a: 163 n. 53 420 c: 5 n. 3 427 e: 120 n. 16 427 e 6–11: 130 428 c 11–d 3: 132 n. 53 429 b 9–c 1: 132 n. 52 429 c 2: 132 n. 53 429 c 7–8: 132 n. 53 430 a 3–b 2: 154 n. 33 430 b 2–3: 132 nn. 52, 53 430 c 3–5: 133 n. 54, 159 n. 43 438–9: 133 n. 55 438 a 3: 133 n. 55 439 d 8: 154 n. 34 440 a 3: 153 440 a 8–b 4: 153, 155 n. 37 440 b 1–2: 156 441 a 7–b 1: 166 441 a 9–b 1: 158 441 b: 161 n. 46 441 c 1–2: 154 n. 32 441 e 4–6: 132 441 e 4–5: 130 442 a 5–b 2: 153 n. 31 442 b 11 ·.: 130 442 b 11–c 8: 120 n. 16 442 b 11–c 4: 153 442 b 11–c 3: 132 442 c 1: 132 n. 52 442 c 5–8: 130, 132 442 c 10–d 1: 130 443 c 9–444 a 2: 130 443 e 6–7: 130 445 c 5–6: 130 449 a 1–2: 130 452 e 1–2: 144 n. 18 469 b–471 c: 56 n. 34 471 a: 57

479 a: 123–4 n. 27 484 c 7–8: 131 n. 51 484 d 7: 131 n. 51 491 a–495 c: 47 492 c 6–8: 168 n. 65 500 d 7–8: 131 n. 50 504 a–e: 133 n. 54 514 a–517 c: 205 514 a 1–517 c 4: 214 514 a 1: 214 n. 44 514 b 5: 214, 217 n. 51 514 b 7: 217 n. 52 514 b 8: 214 514 b 9: 217 n. 51 515 a 4: 214 nn. 44, 46 515 a 5: 214 n. 44 515 a 9–b 1: 214 n. 46 515 a 9: 214 n. 44 515 b 2–c 3: 215 515 b 4–5: 213 n. 43 515 b 10: 215 515 c 1: 214 n. 44 515 c 3: 214 n. 44 515 c 4: 214 n. 44, 217 n. 52 515 c 5: 217 n. 51 515 d 6: 217 n. 51 515 d 8: 214 nn. 44, 46 515 e 4: 214 515 e 5: 214 n. 44 515 e 7: 214 516 a 3: 214 n. 45 516 a 4: 214 nn. 44, 46 516 b 1: 217 n. 51 516 b 3: 214 516 b 6: 217 n. 51 516 b 7: 214 516 c 3: 214 n. 44 516 c 7–517 c 4: 215 516 c 8: 217 n. 51 516 d 4–e 8: 215 516 d 7: 217 n. 52 516 e 1: 214 n. 44 516 e 3: 214 n. 44 516 e 6: 214 n. 44, 217 n. 52 516 e 8: 217 n. 51 517 a 2: 217 n. 51 517 a 4: 217 n. 51 517 a 7: 214 n. 44 517 a 8: 214 517 b 1: 214 n. 45 517 b 5–7: 214 517 b 8: 217 n. 51 517 c 3: 217 n. 51

Created on 4 July 2005 at 14.35 hours page 409

Index Locorum 518 b–519 b: 205 518 b 7: 217 n. 51 518 b 8–519 b 5: 216 518 c 1: 217 n. 51 518 c 3: 217 n. 49 518 c 4: 217 n. 49 518 c 5: 217 nn. 51, 52 518 c 9: 217 n. 51 518 d 1–2: 217 n. 49 518 d 3: 217 n. 49 518 d 5: 217 n. 51 518 d 7: 217 n. 50 518 d 8: 217 n. 49 518 d 9 ·.: 131 n. 50 519 a 2: 217 n. 51 519 a 6: 217 n. 49 519 b 2: 217 n. 51 536 a 2–3: 131 537 d–539 e: 40 559 b 9–10: 163 n. 52 580 d ·.: 161–2 n. 48 586 b 7–8: 161–2 n. 48 586 d: 159 586 d 8–e 1: 161–2 n. 48 589 b 1–3: 163 n. 52 602 c ·.: 137–8 n. 2 611 b 10–c 5: 133 n. 54 Sophist 228 b 2–4: 159 n. 42 230 a–231 b: 40 252 b: 379 Statesman 277 c: 5 n. 3 283 c ·.: 129 Symposium 205 a: 69 210 b–c: 47–8 n. 22 219 e: 57 220 e: 57 222 a 8–b 4: 47 n. 20, 48 Theaetetus 152 b: 184 n. 15 171 a–b: 172, 173–4 173 c–177 b: 205 173 c: 207 173 c 8: 207 n. 32 173 c 9–177 b 7: 207 173 c 9: 211 173 d 6: 213 n. 42 173 d 7: 212 nn. 39, 40 173 e 4: 213 n. 42 173 e 5: 212 n. 40 174 a 3–5: 208

174 a 3: 207–8 n. 33 174 a 5: 212 n. 39 174 a 7: 212 n. 40 174 b 3: 212 n. 39 174 b 4: 212 n. 39 174 b 7–c 1: 209 174 b 8: 207–8 n. 33 174 b 9: 212 n. 39 174 c 1: 212 n. 39 174 c 7–8: 212 n. 39 174 d 5: 212 n. 39 174 e 4: 212 n. 39 175 a 7–b 1: 212 n. 39 175 b 1: 212 n. 39 175 b 7–8: 209 175 b 7: 207–8 n. 33 175 c 4: 212 n. 40 175 c 5: 212 n. 40 175 c 6: 212 n. 39 175 c 7: 213 n. 42 175 c 8: 212 n. 39 175 d 4: 212 n. 40 175 d 6: 212 n. 39 175 d 7–e 2: 210 175 e 1: 212 n. 39 175 e 3: 212 nn. 39, 40 175 e 4: 213 n. 42 175 e 5: 212 n. 39 175 e 6: 212 n. 39 175 e 7: 212 n. 40 176 a 1: 212 n. 40 176 a 2–6: 210 176 a 2–4: 207–8 n. 33 176 b 2: 210 176 b 3: 212 n. 40 176 b 5: 213 n. 42 176 b 7: 212 n. 40 176 c 2: 213 176 c 3: 212 nn. 39, 40 176 c 6: 212 n. 39 176 e 2–3: 209 176 e 2: 207–8 n. 33 176 e 4: 212 n. 39 177 a 2: 212 n. 39 177 a 9–b 1: 209 177 a 9: 207–8 n. 33 177 b 2: 213 n. 42 177 b 3: 212 nn. 39, 40 177 b 4: 210, 212 n. 40 177 b 7–8: 207 178 c–d: 172 201 d–e: 385

409

Created on 4 July 2005 at 14.35 hours page 410

410

Index Locorum

201 e: 368, 376, 379, 385, 386, 387, 390, 391 202 b: 390 202 c: 390 202 e–204 a: 391 202 e: 376, 386, 390 204 a: 391 206 b: 391 Timaeus 48 b–e: 380 n. 46 48 b–c: 379, 380, 382, 383 49 b–c: 380 n. 45 53 c–d: 381 53 d: 381, 384 54 d: 382 55 a: 382 55 b: 382 56 b: 382 57 c: 382 61 a: 382 77 a–c: 299

Porphyry De abstinentia 4. 20. 28: 5 n. 5 Vita Pythagorae 30: 7 n. 9

Pliny the Elder 35. 29: 8 35. 30–49: 13 n. 18 35. 50: 13, 14 35. 60–1: 16 35. 79 ·.: 16 35. 92: 13

Simplicius In Aristotelis Physica commentaria, ed. Diels proem. 7. 10–17: 367 n. 2 695. 34: 245–6 n. 95 In libros Aristotelis De anima commentaria, ed. Hayduck 68. 10–14: 8 n. 12 68. 10–12: 9 176. 32–177. 4: 7 n. 9

Plotinus 1. 2. 1–2: 159 n. 43 3. 1. 3. 2: 378 n. 40 Plutarch De communibus notitiis 1060 d: 256 n. 102 1061 d: 256 n. 102 De defectu oraculorum 433 b: 18 n. 31 436 b–c: 13 436 b: 5 n. 5 De E apud Delphos 393 c: 5 n. 5 De gloria Atheniensium 2: 5 n. 5, 16 De liberis educandis 12 c 12: 389 Quaestiones naturales 39: 20

Ptolemy-el-Garib title 1–2: 197 n. 3 Proclus In primum Euclidis Elementorum comentarii 66. 4–8: 375 Quintilian 12. 10. 4: 16 Sextus Empiricus Adversus mathematicos 10. 249–50: 368 n. 6 Pyrrhoneae hypotyposes 3. 62: 378 n. 40

Sophonias In libros Aristotelis De anima paraphrasis, ed. Hayduck 32. 18–32: 7 n. 9 32. 18–23: 8 n. 12 32. 19–20: 9 Stobaeus 1. 49. 58: 226 n. 58 4. 32. 25: 197 n. 2 Stoicorum Veterum Fragmenta, ed. von Arnim i. 200: 133 n. 57 i. 351: 133 n. 57 Suda s.v. φιλοσοφεν, Φ 414 Adler: 197 n. 4

Created on 4 July 2005 at 14.35 hours page 411

Index Locorum Syrianus In Metaphysica commentaria, ed. Kroll 101. 29–102. 25: 203 n. 30 Theognis 132: 120–1 n. 17 Theophrastus De sensibus 1–2: 30 7: 10, 24 8: 33 10: 7, 21, 29 n. 45 11: 7 12: 23–4 n. 34 17: 29, 30 23: 8 n. 12 27: 33 43: 11 59: 10 73–5: 11 79: 10 Vita Hesychii title 14: 197 n. 3 Vitruvius 7. 7–14: 13 n. 18 Xenophon Apology 6: 67 27: 68 Cyropaedia 1. 6. 27–41: 59 n. 40 Hipparchus 9. 1: 54 Memorabilia 1. 1–2: 47–8 n. 22 1. 2. 12–16: 47 1. 2. 24–5: 47 1. 2. 29–38: 44 1. 2. 29–30: 48 1. 2. 40–6: 44 1. 2. 42–5: 70 n. 49 1. 2. 56–61: 51 1. 3. 3: 51 1. 3: 47–8 n. 22 1. 4: 41, 42, 43 1. 5: 42 1. 6. 14: 51, 54 1. 6: 42, 43

2. 2. 2. 2. 2. 2. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 3. 4. 4. 4. 4. 4. 4. 4. 4. 4. 4. 4. 4. 4. 4. 4. 4. 4. 4. 4. 4. 4. 4. 4. 4. 4. 4. 4. 4. 4. 4. 4. 4.

1. 1–20: 45 n. 16 1. 1. 9: 374 n. 26 1. 20: 51 1. 21–33: 51 6: 56 6. 22–7: 56 1. 6: 56 n. 34, 59 n. 40 4: 70 5: 64 5. 4: 40 n. 5 6. 14: 70 6: 47–8 n. 22, 64 7: 47–8 n. 22 8: 60 n. 41 8. 1: 45 n. 16 9. 4–5: 68 9. 5: 60 9. 8: 56 n. 34 9. 10–11: 70 12: 67 12. 6–7: 67 1: 47 1. 2: 47 n. 20, 67, 70 2: 39 ·. 2. 1: 48 2. 5: 50 2. 10: 51 2. 11–18: 55 2. 11: 47 n. 20 2. 13: 52 2. 14: 57 2. 15: 52 n. 28, 57, 58 2. 16–18: 58 2. 18: 53, 58 2. 19–20: 59 2. 19: 58, 59 2. 20: 53 2. 21: 62 2. 22–3: 66 2. 22: 68 2. 25: 53 2. 26–9: 63 2. 28: 64 2. 30: 66 2. 33: 67 2. 34: 68 2. 35: 69, 70 2. 36–9: 69 2. 39: 66, 70 2. 40: 50 n. 26, 71 3: 54 3. 14: 67 4: 45

411

Created on 4 July 2005 at 14.35 hours page 412

412 4. 5: 54 4. 5. 6: 68 4. 6: 54 4. 6. 5–6: 60 n. 41 4. 6. 12: 70 n. 49 4. 6. 13–14: 54–5 n. 32 4. 6. 15: 54–5 n. 32 4. 7. 1: 54–5 n. 32 4. 7. 9: 67 4. 8. 1: 67 4. 8. 10–11: 56, 58 n. 39 Oeconomicus 1. 1–15: 70

Index Locorum 1. 5–15: 45 n. 16 1. 23: 57 16. 1: 54 21. 2: 70 Symposium 3. 4: 68 n. 48 3. 6: 52 5: 45 n. 16 8. 10–36: 67 Zeno SVF i. 200: 133 n. 57