Oxford Studies in Ancient Philosophy: Volume XXIX: Winter 2005

  • 40 57 3
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Created on 4 July 2005 at 14.33 hours page iii


VOLUME XXI X winter 2005

Created on 4 July 2005 at 14.33 hours page i


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



Xenophon at his Most Socratic (Memorabilia 4. 2)



Socrates’ Deliberative Authoritarianism



The Unity of Virtue in Plato’s Protagoras



Shame, Pleasure, and the Divided Soul



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



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


Aristotle on Why Plants Cannot Perceive



Aristotelian Teleology



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



Index Locorum


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


    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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


‘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


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


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, 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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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]


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

  2 ∆


[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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


     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


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


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


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


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


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


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


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 πα ων (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


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


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


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


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


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


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


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


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


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


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


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


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


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


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 (ασχρο-, 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


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


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


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


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


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


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


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 (ασχ)νεσθαι, 47 d 2) one’s views. So Socrates seems to ascribe to Crito a particular Principle of Moral Deliberation: (PMD*


) 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*


) 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 ασχ)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


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


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


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


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


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


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 (ατας 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


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


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


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


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


  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


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


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


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


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 (σοφ α) and bravery (νδρε α) 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'( παραδε γµατ 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


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


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 +σ η . . . δ κη; 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


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


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


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


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 σοφ α and νδρε α, 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


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, 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


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 σοφ α and νδρε α 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


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


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


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 (σοφ α) 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@ νδρε ας, 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


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 σοφ α (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 σοφ α, but it does not explicitly exclude it either.

Created on 4 July 2005 at 14.33 hours page 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 (σοφ α) 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 (σοφ α) 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 σοφ α 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@ σωτηρ αν διB παντς δξης Fρθς τε κα@ νοµ µου 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


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. Jρ στων, SVF i. 351], -νθρωποι τοτοις, πλεονα 8ν ε#ρνην κα κακ2 λ ττω κατ9 α?το@ς Aσεσθαι. τ2 µBν ο=ν κακ2 ο "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


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


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


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


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


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


‘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


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


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


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


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


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


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


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


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 (81. 5–6).

Created on 4 July 2005 at 14.34 hours page 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 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


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


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


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


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


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


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


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


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


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


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


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


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


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 , 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


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 , 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


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


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


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


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


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


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


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


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 * * * * * * * *

* * * *


‘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


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


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


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


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


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


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


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


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


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


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


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


 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


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


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


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


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


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


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


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


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


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


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


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


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


Damian Murphy

Every chemical stu· S is mapped onto an ordered quadruple , 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


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 , is mixed with S , whose elemental formula is 1 1 1 1 2 , 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 , 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


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


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 , r(h) = 0 and r(m) = 0. If the elemental formula of S (Water) is , r(h) = 0 and r(m) = 1. If the elemental formula of S (Fire) is , r(h) = 1 and r(m) = 0. If the elemental formula of S (Air) is , r(h) = 1 and r(m) = 1.

Created on 4 July 2005 at 14.34 hours page 311

Aristotle on Why Plants Cannot Perceive


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 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 could have the elemental formula or the elemental formula . 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


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 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


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


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


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


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


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


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 Brine 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


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 and ice has the elemental formula . 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 , 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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


 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


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


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


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


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


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$ 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


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’


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


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’


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 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


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’


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


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’


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


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’


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


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’


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


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’


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


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’


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


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’


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


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


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


Created on 4 July 2005 at 14.35 hours page 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


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


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


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


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


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


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


Created on 4 July 2005 at 14.35 hours page 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


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