Sedley & Brown-Plato _Theaetetus_ 145-147 (1)

Plato, "Theaetetus" 145-147Author(s): David Sedley and Lesley BrownSource: Proceedings of the Aristotelian Society, Supplementary Volumes, Vol. 67 (1993), pp.125-149+151Published by: Wiley on behalf of The Aristotelian SocietyStable URL: http://www.jstor.org/stable/4106986 .

Accessed: 22/01/2015 08:03

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

Wiley and The Aristotelian Society are collaborating with JSTOR to digitize, preserve and extend access toProceedings of the Aristotelian Society, Supplementary Volumes.

http://www.jstor.org

This content downloaded from 168.176.5.118 on Thu, 22 Jan 2015 08:03:03 AMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=black

http://www.jstor.org/action/showPublisher?publisherCode=aristotelian

http://www.jstor.org/stable/4106986?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp


PLATO, THEAETETUS 145-147

David Sedley and Lesley Brown

I-David Sedley

A PLATONIST READING OF THEAETETUS 145-147

Only one ancient commentary on Plato's Theaetetus has come down to us, and it happens also to be the earliest of all

philosophical commentaries to survive in any significant bulk. Before the papyrus scroll containing it gives out it has taken us, in its 75 columns, from the dialogue's opening at 142 down to the first discussion of flux at 152-3. Although it throws much light on the author's brand of Platonism, and on how he applies it to the Theaetetus in particular, the portion of the dialogue covered affords us relatively few chances to see him at work on that peculiarly twentieth-century preoccupation, the analysis of a Platonic argument. It is with that in mind that I have opted here to examine his treatment of one short passage rich in argument, 145d-147c.

Berlin Papyrus 9782 has received surprisingly little discussion since its original publication by Diels and Schubart in 1905.1 Its author-Anon., as I shall be calling him-has so far eluded identification.2 (For my part I feel confident only in asserting who he isn't-but even that need not concern us here.) For present purposes it is enough to say that he is a Middle Platonist, working some time

1 H. Diels, W. Schubart, 'Anonymer Kommentar zu Platons Theaetet', Berliner Klassikertexte II (Berlin, 1905). In the present article I follow the Diels/Schubart text except where otherwise indicated. My new edition of it, jointly with Guido Bastianini, is appearing in Corpus dei papirifilosofici greci e latini, vol. III (Florence, 1993). I am grateful for the generous collaboration of Guido Bastianini and of the general editor Fernanda Decleva Caizzi. I also thank audiences at the Institute of Classical Studies, London, and at Cornell and Duke Universities, for helpful reactions to earlier versions, and Myles Burnyeat, Pierluigi Donini and Joseph DeFilippo for written comments.

2 Diels linked him philosophically with Albinus, whom he believed in turn to be identical with 'Alcinous', the author of the Didaskalikos. H. Tarrant, 'The date of Anon., In Theaetetum', CQ 33 (1983), 161-87 argues for Eudorus; but contra now see J. Mansfeld, 'Two attributions', CQ 41 (1991), 541-4. In the forthcoming edition (see n. 1) I argue against his identification with anyone of whom we know much more than the bare name.



126 I-DAVID SEDLEY

between 50 BC and 150 AD, when the papyrus itself was written, and that he probably belongs nearer to the beginning of that period than to the end of it. It will do no harm to think of him as not unlike Plutarch in his philosophical outlook (albeit less sympathetic than Plutarch to the sceptical tradition within the Academy). He believes, not surprisingly, in the unity of Plato's thought, as well as, a little more unusually, in the unity of almost the entire Platonist tradition within the Academy.3 He also, like many of his contemp- oraries and successors, welcomes Aristotle as the Platonists' ally, and is ready to use Aristotelian dialectical tools in the analysis of Plato's text.4 These are among the features which we will see helping to colour his approach to the Theaetetus.5 The results may be occasionally perverse, but he is also enabled to exploit philosophical interest in the text where others have since overlooked it. I submit, in fact, that he has extracted more philosophical interest from Theaetetus 145-147 than all the modern commentators combined. Besides that, he affords us the rare opportunity to find out how the arguments in the Theaetetus could be read by a Platonist-someone who presupposed them all to be valid, and their conclusions true.

I

Wisdom and knowledge. SOCR. Isn't learning becoming wiser about what one learns? THT. Of course. SOCR. And it is wisdom, I take it, that makes the wise wise? (aopiq... 00opoi o01 00opo)

3 54.38-55.13. Cf. Plutarch's lost work, nEpi TOy paJlV ETvaI TOV &TTrro r•,6TwVOr 'AKCsaqpiav (no. 63 in the Lamprias catalogue).

4 This is possibly a respect in which he differs from Plutarch. Anon. is arguably more familiar with Aristotle's Topics than Plutarch. At least, F H. Sandbach, 'Plutarch and Aristotle', Illinois Classical Studies 7.2 (1982), 208-32, pp. 212-13, 217-19, doubts whether Plutarch had much direct acquaintance with the Topics, although contra see P. A. Vander Waerdt, 'Peripatetic soul-division, Posidonius, and Middle Platonic moral psychology', GRBS 26 (1985), 373-94, pp. 379-80 n.23. Moraux (see next note), p. 483 suggests that Anon.'s own knowledge of Aristotle came from a dialectical handbook; but that would hardly account for his verbatim quotation from the Topics at 24.30-25.29 (see below).

5 The Aristotelian input to this part of the commentary has been usefully examined by P. Moraux, Der Aristotelismus bei den Griechen von Andronikos bis Alexander von Aphrodisias ii (Berlin, 1984), 482-90, but I feel that he misses the strengths of Anon.'s treatment by failing to examine his strategy as a whole.



PLATO, THEAETETUS 145-147 127

THT. Yes. SOCR. And this is surely no different from knowledge? THT. What is? SOCR. Wisdom. Or isn't it true that in those respects in which people are knowledgeable they are also wise? THT. What of it? SOCR. Then knowledge and wisdom are the same thing? (145d7-e6)

Is this an argument at all? Is it anything more than a trivial

sequence of inferences, to warm up the interlocutor and perhaps to steer the topic of conversation in the desired direction? If it attracts

any comment at all from modem scholars, it is normally read as a simple gambit on Socrates' part, to change the topic from mathematical studies to knowledge itself. None, I am quite sure, has suspected that it conceals an argument for the Platonic thesis that learning is recollection. Yet that is how it presents itself to Anon.:

The argument is posed as follows (if one may pass over the setting up of some of the premises):

He who learns becomes wiser. He who becomes wiser recovers wisdom. But wisdom is knowledge. Therefore he who learns recovers knowledge.

But instead of this he uses the conclusion 'Therefore knowledge and wisdom are the same thing?', taking them as equivalent: for he said to him 'In those respects in which people are knowledgeable, they are also wise.'

For the fully articulated argument goes as follows. He who learns becomes wiser; he who becomes wiser recovers wisdom; therefore he who learns recovers wisdom. But since it was agreed that people are knowledgeable in those respects in which they are also wise, it will be conceded that knowledge and wisdom are the same thing. (16.14-17.3)6

Anon.'s analysis is not easy to follow, and has been severely criticised by those scholars who have discussed it, including in

6 At 16.29 I read [ os,

for [671; at 16.33-4

8i[q]p[pw].p[il]Ivoc (s•ilE]p[Oappl ]Ivoc

Praechter, 8l[an]o[vo6pE]Ivoc Moraux).



128 I-DAVID SEDLEY

particular Moraux in the second volume of his Aristotelismus.7 But none of these critics has noticed that Anon.'s version is about recollection, and they have implausibly taken the crucial expression iTrrTi~Jlqv vatappavElv (16.24-5, cf. 21-2, 40-1) simply to mean 'acquire knowledge'. The use of the identical expression at Meno 85d virtually guarantees the more explicit sense 'recover knowledge'. If we miss this, we make Anon.'s mistake sound like one of sheer incompetence. In reality, if he is wrong, he is at least creatively and interestingly wrong.

Anon. offers two versions, as can be seen in the following table.

Plato Anon. version A Anon. version B

(Tht. 145d-e) (16.14-25) (16.32-17.3)

1 Learn x = get wiser Learn = get wiser Learn = get wiser about x

2 Being wise is due to Get wiser = recover Get wiser = recover wisdom wisdom wisdom

3 Therefore learn = recover wisdom

4 Wise with respect to Wise with respect to x = knowledgeable x = knowledgeable with respect to x with respect to x

5 Therefore wisdom = Wisdom = knowledge Therefore wisdom = knowledge knowledge

6 Therefore learn = <Therefore learn = recover knowledge recover knowledge>

In Version A (16.14-25) he says he will omit some intermediate steps.8 He limits himself to giving the bare bones of the argument, including the real conclusion, 'He who learns recovers knowledge', which he has to admit (16.26-9) Plato failed to make explicit. Version B (16.32-17.3) is presented as a more articulated9 analysis

7 P. Moraux, op. cit. pp. 488-90; G. Invernizzi, 'Un commentario medioplatonico at Teeteto e il suo significato filosofico', Rivista di Filosofia Neoscolastica 68 (1976), 215-33, pp. 225-6; K. Praechter, review of Diels/Schubart in GGA 171 (1909), 530-47 (repr. in his Kleine Schriften (1973), 264-81), p. 541.

8 16.15-18. Version A omits (3), (4), and especially the move from (4) to (5) (which version B supplies at 17.3-24, see below).

9 Assuming my reading s1[q]p.[pw]p[lIyvor

at 16.33-4 (see n.6 above). The sense will be precisely that used at Aristotle, Top. 156a19-20, on the dialectical tactic of not stating interim conclusions, so as to make the syllogism less 'articulated' (81apOpoo66al).




of the intermediate steps, supplying one interim conclusion (3) and the premise (4) from which another of the premises (5) is itself derived, but this time neglecting to repeat the implicit overall conclusion (6).

A primafacie obstacle to understanding Anon.'s rewriting of the passage is the difficulty in seeing how recollection has been smuggled in. But a glance at line 2 of the table will reveal that he has detected an allusion to the theory in the expression oopi~a... oo0pol ol 0oopoi (145dl 1), which I translated 'It is wisdom... that makes the wise wise'-more literally 'It is by [or 'because of] wisdom... that the wise are wise'. Most readers of Plato will immediately associate this causal dative with Plato's regular mode of formal explanation, 'F things are F because of F-ness' (Phaedo 100d, Hippias Major 287c-d, 289d, Sophist 247a), which he presents either as some kind of truism, or as the modest metaphysical thesis that it is (for example) the presence of F-ness in F things that makes them F (locc. citt.; Hippias Major 292c-d, Lysis 217). Plato expects it to be valid for a wide range of values of 'F'. But Anon. has chosen to understand it as a causal thesis specific to knowledge or wisdom: wisdom is what makes one wise, in the sense that one must already have (latent) wisdom in order to become (actually) wise.

Should we laugh? One reason for not doing so is that the construal, far from being gratuitous wishful thinking, is part of a coordinated unitarian strategy for reading the whole dialogue. Anon. regards the Meno as the key text for understanding Plato's theory of knowledge, including that presented in the Theaetetus, and he has little difficulty in, for example, interpreting Socratic midwifery in the latter dialogue as the dialectical encouragement of others' recollection of innate knowledge. Indeed, as we shall see, his interpretation of the very next passage, on Theaetetus' first definition, will be drawing on this theory of knowledge.

Moreover, the absence of any explicit allusion to recollection need not disconcert him at all, since, as he points out elsewhere, Socrates himself says in the Meno 'Let it make no difference whether we call it teachable or recollectable' (Meno 87b-c, cited by Anon. at 56.11-31). Thus he is constantly on the look out for buried allusions to the doctrine. The very fact that readers are often left, as here, to work it out for themselves is in his view no more



130 I--DAVID SEDLEY

than a proper application of Socratic midwifery.10 Besides, any reader of Plato is virtually obliged to subscribe to some version of Anon.'s principle, enunciated in column 59: 'In inquiries he asks questions and does not make assertions, so that he posits neither a falsehood nor a truth; but to those well versed in his method he covertly (XEXqO6-rT[&])

indicates his own doctrine.' A further ground for respecting Anon.'s approach is that it leads

him, unlike all modem commentators I have consulted, to take the argument seriously. In particular, he is led to ask how Plato can justify his move from (4) to (5): 'Isn't it true that in those respects in which people are knowledgeable they are also wise?...Then knowledge and wisdom are the same thing?' Anon. comments:

But since it was agreed that people are knowledgeable in those respects in which they are also wise, it will be conceded that knowledge and wisdom are the same thing. For whereas from 'The same people are knowledgeable and wise' it cannot be deduced that 'Knowledge and wisdom are the same thing'-just as from 'The same people are literate and cultured' it cannot be deduced that literacy and culture are the same thing-from 'People are not knowledgeable in some respects and wise in others, but both in the same respects' it will follow that knowledge and wisdom are the same thing. (For so too, if people were literate in the same respects as those in which they are cultured, literacy and culture would be the same thing.) (16.41-17.24)

This is an acute observation of the care with which Plato has constructed his inference, and the point may well be a sound one. Clearly from the mere identity of wise people with knowledgeable people we would not be able to infer the identity of wisdom with knowledge, any more than from the identity of creatures that have livers with creatures that have hearts we can infer that hearts are livers. But Anon. holds that if two properties are coextensive not merely in belonging to precisely the same individuals but also in belonging to them in the same respects, then an identity relation does obtain between them.

10 I develop this point in 'Three Platonist interpretations of the Theaetetus ', forthcoming in C. Gill, M. M. McCabe (edd.), Form and Argument in Later Plato (Oxford, 1994).




It would be advisable to take 'respects' here to cover all relevant respects. In the case of hearts and livers, their non-identity is rescued by the fact that one does not have a heart and a liver with respect to the same part of the body, or the same bodily functions. But in the example of literacy and culture, chosen by Anon., their non-identity might be argued rather on the strength of their different external objects-e.g., that you can be cultured, but not literate, with respect to music and painting. Conversely, to take a case of putative identity, Theaetetus' definition 'Knowledge is perception' could be successful only if knowing and perceiving were done with the same part of oneself-the sense-organs-and with respect to the same objects, the sensible qualities, as well as at the same times, by the same person, etc. Interpreted thus, with an appropriate degree of breadth,11 Anon.'s law of identity could come out as a distant forerunner of Leibniz's Law, at least as far as concerns the identity of properties.

This is the first of two passages in which we will see Anon. showing a special interest in the logic of identity. And although he is familiar with Aristotle's Topics, his law of identity does not closely echo anything there, not even in Top. vii.l, Aristotle's fullest account of identity rules.12 Rather, I am inclined to suspect, his sensitivity to the nuances of Plato's text at this point may reflect the Hellenistic debate about the Unity of the Virtues.

Plato's account in the Protagoras, and likewise that given by the founding Stoic Zeno, notoriously left their interpreters, ancient and modern,13 in disarray as to whether the proper Socratic thesis is one of the mere inseparability of the virtues (Chrysippus; Vlastos), or

11 Anon.'s analysis has the value of showing that at 145e3-4 it is too restrictive to translate 'If you know something, aren't you also wise about it?' (Waterfield; similarly Jowett, Apelt, Carhill, Kennedy, Levett, Cornford, Fowler, Benardete), thus limiting the 'respects' (conveyed in Greek by the simple accusative of respect) to the objects of cognitive states. McDowell's '... people are wise in precisely those respects in which they are knowledgeable' emerges as superior. Note too that Anon. has scrupulously recast as a proper biconditional Plato's 'in those respects in which people are knowledgeable they are also wise', which thus phrased would have allowed people to be wise without being knowledgeable.

12 Top. 152b25-9 is perhaps the closest Aristotle comes. As for a possible Platonic background, neither Ion 537 nor Rep. 477 seems enough to have generated the law, although both with hindsight might have been thought to rely on it.

13 Stoic evidence in Long/Sedley, The Hellenistic Philosophers (Cambridge, 1987), section 61. For the parallelism of the ancient with the modern debate, see M. Schofield, 'Ariston of Chios and the unity of virtue', Ancient Philosophy 4 (1984), 83-95.



132 I-DAVID SEDLEY

of their numerical identity (Cleanthes, Aristo; Penner, Taylor, Irwin). Anon. himself does not go along with the strict identity thesis, as propounded by the Stoic Aristo, but endorses the weaker doctrine of the inseparability (6VTOcKOAouioca) of the virtues.14 He is also keenly interested in the debate on the issue between Aristo and the mainstream Stoics (11.22-40). Now Aristo had held that there is a single virtuous state of mind, which counts as 'moderation' in relation to desire, 'justice' in relation to contracts, and so on.15 Anon.'s law of identity would fit perfectly into this debate, enabling the retort to be made to Aristo that the same state of mind may be both moderate and just, but that if it is moderate in respect of one thing, but just in respect of another, then moderation and justice will not be the same thing. Thus it is worth considering the possibility that Anon.'s very proper sensitivity to the precise terms of Plato's argument here, and the law of identity which he invokes, are the fruit of his evident involvement in the Stoic debate about the Unity of the Virtues.

II

Definitional complexity. THT. Well it seems to me that the things one can learn from Theodorus are kinds of knowledge-geometry, and the others you just listed-and again that shoemaking and the other skills of the craftsmen, both jointly and severally, are nothing other than knowledge. SOCR. With nobility and generosity, my friend, when asked for one thing you are giving many, and complex instead of simple.16 (146c7-d4)

This sounds like the interlocutor's standard first shot at definition by cataloguing examples, followed by Socrates' characteristically ironic response: how generous of you to give me many when all I asked for was one (cf. Meno 71e-72a). But what is the force of

14 At 9.44-10.1, where Diels/Schubart read rr6v[-T'6](Kohoue'

, the correct reading is 1 6vTr[ac]KoouOia.

15 Plutarch, VM 440E-441A.

16 Richard Janko has suggested to me the intriguing alternative construal 'when asked for one thing, you are giving many varied things (noA•X& bibx Kai TTOIKiXh), instead of something simple'. This possibility is not, however, contemplated by Anon.




Socrates' additional complaint,'complex instead of simple (rr I K iX•a 6rVTrI &rrnho)'? He might mean (1) a whole variety of items, instead of one; or (2) items each of which is complex, instead of a non- complex item.

Certainly (1), favoured by almost all modem translators, is a good Platonic usage.17 Understood this way, Socrates will be in effect repeating the point made by the words '...when asked for one thing you are giving many', and thus putting all the stress on the unity of the definition. But it is arguable that this theme, familiar enough from other dialogues,18 is not the dominant one in the criticisms which follow when Socrates is asked by Theaetetus to explain his complaint. Rather, the principal objection is that a definition like 'Knowledge is shoemaking', once the definiens 'shoemaking' is itself properly analysed as 'knowledge of making shoes', turns out to give only non-essential information about knowledge, and to threaten circularity. Thus19 the main thrust of the objection could turn out to rest on the concealed complexity of such items as 'shoemaking'. And that could favour (2).

Anon., at least, decidedly opts for the latter reading:20 Socrates has asked 'What is knowledge?', i.e. what is it for knowledge to be knowledge? But Theaetetus has enumerated many (e.g. geometry, music), and has gone wrong both in not representing the essence of knowledge, and in enumerating many. But he would have been wrong even if he had listed one, by representing the species instead of the genus, defining what is prior through what is posterior. It is as if someone were to ask 'What is an animal?' and he were to reply 'Man, horse'. For the things which are prior are predicated of the things which fall under them, but not vice versa. And genera are participated, and are more simple, while species participate, and are more complex. For man is both an animal and in addition rational and mortal. Hence 'You are giving many' refers to the enumeration of kinds of knowledge. But 'complex instead of simple' is because he has listed species and these are more complex than the genus; he would have said the same even if Theaetetus had

17 Campbell cites Philebus 12c as a partial parallel. 18 E.g. Meno 72, Euthyphro 6d-e. 19 As Campbell, ad loc., concedes, despite favouring (1). 20 Also favoured by S. Benardete, The Being of the Beautiful (Chicago/London, 1984), 1.8,

94, though by no other modem translator I have consulted.



134 I-DAVID SEDLEY

only named one kind of knowledge, say geometry. For knowledge is something simple, as if participated and uncompounded in relation to the things that fall under it, whereas geometry is complex, as participating and compounded. (18.11-19.20)

The idea of the species as definitionally composite, as com- pared with the incomposite genus, has a firm foothold in Middle Platonism and Neoplatonism,21 and an Aristotelian background.22 It probably just struck Anon. as the obvious way to read Socrates' remarks. Even so, it is not easy to see what general doctrine he could take Plato to be invoking when he castigates the complexity of Theaetetus' definition. He can hardly suppose the implication to be that the definiens must always itself be strictly 'incomposite' in the relevant sense. Clearly few definientia are.23 The force of the criticism must rather, on this reading, be that 'Knowledge is shoemaking' fails as a definition because shoemaking, as a specimen branch of knowledge, is more definitionally complex than is appropriate to the simple genus 'knowledge': its definition must include the genus knowledge, but also add a further differentia. This asymmetry between definiendum and definiens might best be expressed in terms of priority and posteriority (as Anon. himself sees, 18.25ff.): it is incorrect, as Aristotle would put it,24 to define a term by means of terms posterior to it. But here, as Anon. reads it, the same point is being expounded instead in the language of generic simplicity versus specific complexity.25 Knowledge pure and simple is being mis- leadingly placed on a level with something as complicated in conception as a specific branch of knowledge.

21 Proclus, In Parm. 650-1, on Platonic ouvaywy6 as the collection of a definition out of its components, is a particularly striking example.

22 Cf. Aristotle, Met. Z 12. 23 In 'Knowledge is perception', welcomed by Socrates as well-formed, the definiens is

simple, but only verbally; conceptually it is complex (esp. 156). And all the subsequent definitions in the dialogue have unquestionably complex definientia.

24 E.g. Aristotle, Top. vi.4. 25 This distinction should, despite a strong surface resemblance, probably be kept quite

separate from Anon.'s view that the Theaetetus is concerned with 'simple incomposite knowledge', as opposed to a 'composite' or 'systematic' knowledge (i.e. irrT-r1T6q

in the sense 'a science') like geometry (especially 14.45-15.18, 17.25-32). There 'composite' means not definitionally but structurally complex. (For a third usage of 'simple knowledge', see Alcinous, Didaskalikos 155.33 Whittaker.)




III

Definition and tautology.

SOCR. When you say shoemaking, are you speaking of nothing other than knowledge of the manufacture of shoes? THT. Nothing other. SOCR. What when you say carpentry? Are you speaking of nothing other than knowledge of the manufacture of wooden objects? THT. Again, nothing other. SOCR. Then aren't you in both cases defining what each is the knowledge of? THT. Yes. SOCR. But the question, Theaetetus, was not what things knowledge is of, or how many kinds of knowledge there are. We didn't ask out of a desire to count them, but to know what knowledge itself is. (146d6--el0)

Theaetetus has defined knowledge as shoemaking and carpentry. Socrates objects that shoemaking and carpentry are in turn equivalent to knowledge of certain things; hence Theaetetus has only told us that knowledge is knowledge of certain things, not what knowledge is. The inference has rightly been regarded with suspicion.26 But Anon.'s highly discerning analysis arguably comes closer to rescuing it than any of his modem successors has done. He starts with a paraphrase, followed by a brief gloss:27

'Asked what knowledge is, by replying "shoemaking" you mean nothing other than knowledge of the manufacture of shoes. For shoemaking is knowledge of the manufacture of shoes. Likewise also if you say carpentry you mean nothing other than knowledge of manufacturing wooden objects.' Because that is the proprium (ilov) of shoemaking. (19.30-46)

26 Cf. McDowell ad loc., p. 114, and, for a more constructive set of criticisms, M. F. Burnyeat, 'Examples in epistemology: Socrates, Theaetetus and G. E. Moore', Philosophy 52 (1977), 381-98.

27 It is essential to separate it into these two components, as I have done with my use of quotation marks for the paraphrase, because otherwise the concluding gloss, 44-6, would make an impossible sequel to the sentence immediately preceding it. (Diels/Schubart were driven by this consideration to emend OKUTIKrTg to TEKTOVIKIg in 45-6. But that is unnecessary, once it is noticed that Anon. regularly juxtaposes paraphrase with comment without signalling the transition.)



136 I-DAVID SEDLEY

To make sense of what follows, it is crucial to see that this paraphrase is more loaded and less innocent than might at first appear. Where Plato uses the neutral term (ppa'6ic, which I have translated 'speak of', Anon. has substituted XYEIS, which I have translated 'mean'. Thus interpreted, Socrates is claiming that when Theaetetus says 'shoemaking' his word does not merely pick out something which happens to be identical to knowledge of making shoes, but actually means 'knowledge of making shoes'. And that is the point of Anon.'s comment, 'that is the proprium of shoemaking'. In Aristotle's usage a proprium (Ziov) had tended to be a non-definitional peculiar characteristic.28 But the technical sense which the term had gone on to acquire in Hellenistic philosophy restricted it to essential or definitional peculiar characteristics.29 This latter sense must be the one intended here, if the remark is to be of any relevance to the preceding paraphrase. And, if so, Anon. has supplied the remark as a ground for taking Socrates to be eliciting the actual meaning of Theaetetus' definiens, and not merely some further identity that obtains for it. 'Shoemaking' means 'knowledge of making shoes', because that is its proprium or defining characteristic.

Without seeing this, it would be hard to appreciate what is at issue in Anon.'s next paragraph:

This is a misunderstanding, they say: for the object and the definition are convertible, but the definition does not mean exactly the same as the name. For if one person asked 'What is a man?', and the other replied 'A rational mortal animal', just because a rational mortal animal is a man we won't say that when asked 'What is a man?' he replied 'A man'. For in that case if you ask someone 'Whose son was Achilles?' and he replies 'Achilles was the son of Peleus', we will have to say that when asked 'Whose son was Achilles?' he replied 'Achilles'. This was what he replied, but accidentally. (19.46-20.24)30

This fascinating anticipation of G. E. Moore's Paradox of Analysis must be approached carefully in order to grasp what, in context, is the point being made.

28 Aristotle, Top. 101b19 ff. 29 Long/Sedley section 32.

30 At 20.23-4 I1 read 6Ih' IX ourPip3nKEv, where Diels/Schubart have &X &hw ou•ppipqKIEV.




Evidently an objection to what precedes is being quoted: 'It has been misunderstood, they say (TrrcpEKSi6E KTOI I, (patOiv)'. Either (a) it is an objection to the paraphrase and explanation just offered, as being a misunderstanding of Socrates' words; or (b) it is an objection to Socrates' understanding of Theaetetus' words. If (a), then we may take the objectors to be themselves Platonists. For, since their complaint is that the interpretation saddles Socrates with a false inference, they are presumably themselves committed to making the inference come out as valid. It is hard to see who but Platonists would be thus committed.31 On the other hand, if (b) were correct, the objectors would be directly criticising Socrates' inference, and could hardly be Platonists. They might perhaps be Peripatetics. I have found no way of choosing between these alternative readings, beyond a hunch that the failure to identify the critics as opponents (merely 'they say'), and Anon.'s subsequent accommodating attitude to them, makes it more likely that they are Platonists, and that we are here witnessing an internal school debate about how best to do justice to the master's words.

The first objection (19.46-20.14) is that although a definition expresses a reciprocal relation of some kind,32 it is a mistake to treat that relation as one of synonymy. Synonymy is taken to imply intersubstitutability salva veritate. But if definiens and definiendum are intersubstitutable salva veritate, then they can be substituted for each other even within the definition itself. In which case anyone who asserts a definition is, absurdly, asserting a mere tautology.33

Applied to the argument at Tht. 146d-e, this is supposed to show that when Theaetetus says 'Knowledge is shoemaking' Socrates cannot legitimately object that, since shoemaking is itself definable as knowledge of making shoes, Theaetetus' definition is

31 In theory, they could be Stoics, using the Theaetetus as an authentic document about Socrates, to whom (unlike to Plato) they did impute authority. But I cannot see much plausibility in this. Admittedly the use of

6VT•oTrpi(pE at 19.47-20.1 could recall the

Stoic Antipater's views on definition (Long/Sedley 32B 2), but it is also Aristotelian (Top. 154b2-3).

32 For 6vro-rp&pElv, see Long/Sedley 1.94, on Antipater. The point must be that although the relation of being-definable-as is not symmetrical, the definiens and definiendum can properly be said of each other.

33 The paradox chosen seems to be this, viz. the reduction of 'x said that a is b 'to 'x said that a is a ', rather than the reduction of 'a is definable as b' to 'a is definable as a '.



138 I-DAVID SEDLEY

semantically equivalent to 'Knowledge is knowledge of making shoes'.

The critics then seem to continue with a supporting example (20.15-24). If this intersubstitution were legitimate, then even where the relation expressed by the copula is not definitional but still one of identity, intersubstitutability salva veritate would have to be assumed. If 'Achilles' and 'the son of Peleus' were universally intersubstitutable, then saying that Achilles was the son of Peleus would be the same as saying that Achilles was Achilles. This time a reason is given why the substitution is illegitimate: it is only 'accidentally' (cx, ou I3AU3PIKE) that, in speaking of 'the son of Peleus', one speaks of Achilles.

One might suppose the point to be that although Achilles and the son of Peleus are identical, it is only an 'accidental' identity. Such a diagnosis would owe much to Aristotle, Soph. El. 179a36 ff., where accidental identity explains why you can know Coriscus and yet not know the approaching man, who happens to be Coriscus. But its relevance here is not entirely clear. For one thing, even granted that Achilles is only accidentally the son of Peleus, it would be unfortunate if the solution worked only for accidental identities. Ideally, even statements of essential identities, such as '2 + 2 = 4', should not be reducible to assertions of tautologies. For another thing, the lesson of the Achilles example is clearly meant to be applicable to the first example, the definition of man; but whatever kind of identity might be expressed by 'Man is a rational mortal animal', it can hardly be an accidental identity.

Perhaps then it is better to understand 'accidentally' as modifying, not the identity relation between Achilles and the son of Peleus, but the identity relation between the assertion 'Achilles was the son of Peleus' and the assertion 'Achilles was Achilles'. It is only accidentally that in making the former assertion one is also making the latter.

If we reconstruct the argument this way, we are saved from having to ask what kind of identity underlies the substitution. Suppose someone says that 4 is the number of the evangelists. Even though the identity 2 + 2 = 4 is no accident at all, it is perfectly legitimate to deny that this person has said, other than accidentally, that 2 + 2 is the number of the evangelists.




Although this notion of accidental saying makes ready sense, it does not reflect the standard Aristotelian use of oup3PE P K6O, which is normally reserved for connexions which fail to obtain always or for the most part. We must think instead of a weaker sense, better rendered 'incidentally', such as is found at De interpretatione 21a25-8, where we are told that to say of Homer that he 'is a poet' is to say of him that he 'is', but only 'incidentally' (KaTd

oup p3ErPKOC,); or at Physics VI 10, 240b8-12, where the movement of a boat involves the 'incidental' movement of its passenger.

The general lesson is a very valuable one. Let the '=' sign stand for any kind of identity at all that falls short of bare synonymy or self-identity-whether definability-as, or accidental identity, or essential identity. The critics' claim is that if you say 'x is y', and if as a matter of fact y = z, it does not follow that (other than incidentally) you either said or meant 'x is z'.34 And this must be right. When Socrates later says 'Mud is earth mixed with water', even granted that as a matter of fact water is identical with, or definable as, H20, it does not follow that Socrates either said or meant that mud is earth mixed with H20. A modem analysis of this phenomenon would no doubt point to the context, that of 'saying' or 'meaning', as being referentially opaque, and disallow the substitution for that reason. But these ancient critics have, at least, achieved a major step forward in identifying the fallacy itself.35

It seems fruitless to speculate what, if any, alternative reading of the passage was proposed by the critics (assuming they were Platonists). What we do get, though, is Anon.'s reply to their criticism:

What I say, then, is that this is not the aim of his argument. Rather, it indicates that Theaetetus' answer was not to the point. For when asked about a thing itself-and this is understood in the what-it-is category-his own answer was in the category of relation. For being of such-and-such is relative. In addition he is outlining a

34 In this they are making an advance on Aristotle, who treats definiens and definiendum as without qualification intersubstitutable (Top. 142a34-b6, A Pr 49b5).

35 I only know one other ancient occurrence of this tautological fallacy, in a passage where Alexander (Mantissa 159.22-6), apparently unaware that it is fallacious, uses it to oppose the identification of virtue with happiness: 'If someone who asks whether virtue is sufficient for happiness is not asking a silly question, while someone who asks whether virtue is sufficient for virtue is asking a silly question, the two questions are not the same. And if so, virtue and happiness are not the same.'



140 I--DAVID SEDLEY

dialectical theorem, that what is prior should not be defined through what is posterior, since if one defines it through its class-members one will be defining it through itself. For if someone were asked 'What is a man?' and replied 'Socrates', he has defined it through itself, because Socrates is a man. Likewise too if someone, asked 'What is knowledge?', replied 'Shoemaking', he has defined it through what is posterior. For knowledge of the manufacture of shoes is a (kind of) knowledge. (20.24-21.13)36

As I understand him, Anon.'s reply is that his analysis does not saddle Socrates with the fallacious reductive strategy envisaged by the criticism. That is, Socrates is not trying to reduce what Theaetetus said, or meant, to something that he did not say or mean. Rather, Socrates wants to analyse Theaetetus' definition into a form which will lay bare its concealed logical deficiencies.37 And these deficiencies are twofold: (a) the definiens is in the wrong category; (b) the definiens is posterior to the definiendum. Both are strongly Aristotelian. Let us take them in turn.

First the categorial point. Socrates, having recast Theaetetus' definition as 'Knowledge is knowledge of making shoes (etc.)', objects that his question had not been what knowledge is knowledge of, but what knowledge itself is (146e7-10). According to Anon.'s diagnosis, Socrates' recasting had the quite innocent aim of exposing a category mistake. The question about knowledge was in the what-it-is category, the answer was in the category of relativity.

Now metaphysically this has little to do with Aristotelianism. Anon. may appear to associate knowledge itself with the what-it-is category, i.e. the category of substance, and its species (shoemaking, etc.) with the category of relativity. Aristotle, by contrast, places knowledge itself in the category of relativity,38 since it is

36 At 20.25 Diels and Schubart reported the sequence Touvuv and adopted Wilamowitz's emendation of it to Trovuv; but this latter is in fact tne correct reading of the papyrus.

37 The device of exposing deficiencies by substituting an equivalent phrase for a term in the proposed thesis is itself one recommended in Topics i.5 and called pET76 Any r (Top. 112a21-2, 130a38-b3, 147bl3-15). Anon. seems to be echoing that recommendation at 22.7-13, using the same technical term; but unfortunately, if so, his comment must have become misplaced in the text: it probably belonged a column earlier, after 21.13, and referred to the present passage of Tht.

38 Categories ch. 7. This difference between Anon. and Aristotle is observed by Moraux, op. cit. p. 486. Note also that whereas Anon. tries to drive a wedge between defining knowledge and stating what its objects are (also at 2.32-52, where his entire reading of




inseparably correlated with its object, the knowable, but its species in a different category, that of quality.39 He places neither the genus nor the individual species of knowledge in the category of substance.

Fortunately, however, Anon.'s point has little to do with metaphysics. His reliance on Aristotelianism here is, not unreasonably for a Platonist,40 a methodological one. Hence his approach to category theory owes more to the Topics than to the Categories.41 In the Categories the primary category is limited to bona fide Aristotelian substances, and the other categories are viewed as classifying the items which can be predicated of those substances. But in Topics i.9 Aristotle says 'Someone who indicates what-it-is (T6 Ti 7 oTi) sometimes indicates substance, sometimes quality, sometimes another of the categories',42 going on to explain that if I point at a colour and say 'That is white' or 'That is a colour' I am stating what-it-is by indicating a quality, not a substance. As Ackrill puts it,43 'In this passage [i.e. as opposed to the Categories ], where the question "what is it?" is thought of as addressed to items in any category, Aristotle can no longer use 'what is it' as a label for the first category but employs the noun for "substance".' (For a similar application of category theory, cf. Metaphysics Z 4, 1030a17 ff.)

This precisely sets the background to Anon.'s non-metaphysical use of the category theory. Significantly it is a use which, according to Michael Frede's recent work,44 would accurately reflect the theory's origins in Platonic dialectic. Anon. is not interested in the place of knowledge as such within the category scheme, but in the way that the categories can be used to grade different kinds of questions and answers about any given object of inquiry, regardless of its own categorial status. You can ask what a thing is, how big

the dialogue depends on the distinction), Aristotle (Top. 146b5) considers a definition of knowledge deficient if it doesn't specify its object(s).

39 Cat. 11a20-36. 40 Thus Alcinous, Didaskalikos 159.43-4 Whittaker, places the category theory (which he

traces back to Plato, cf. Plutarch, An. procr 1023e-f) in his section on dialectic, between the treatments of sophisms and etymology.

41 For a more metaphysical use of category theory, cf. 68.1-15. 42 Top. 103b27-9. 43 J. L. Ackrill, Aristotle's Categories and De Interpretatione (Oxford, 1963), p. 80. 44 M. Frede, 'Categories in Aristotle', in D. J. O'Meara (ed.), Studies in Aristotle

(Washington D.C. 1981), repr. in Frede's Essays in Ancient Philosophy (Minneapolis 1987).



142 I-DAVID SEDLEY

it is, where it is, and so on. But question and answer must be located within the same category. It was by contravening this assumed rule that Theaetetus' definition failed, and Socrates' paraphrase did no more than expose the category mistake.

The second deficiency that Socrates' paraphrase is supposed to have brought to light is that Theaetetus has defined what is prior by means of what is posterior. Anon.'s textual warrant for this must be in the ensuing argument at 147a7-b8, which uses the example of clay but is meant to be applicable, mutatis mutandis, to Theaetetus' definition: if the definiendum appears in the definiens, as it does in 'Knowledge is knowledge of making shoes (etc.)', or in 'Clay is oven-makers' clay (etc.)', then the definition fails because it presupposes understanding of precisely the term yet to be explained.

There seems to be a slide here from an epistemological to an ontological principle. Plato's argument sounds primarily epistemological: if you don't understand the term 'knowledge', you don't understand the expression 'knowledge of making shoes' either. Anon.'s analysis sounds ontological: you cannot properly define a thing by means of things which are logically posterior to it, since to do so is circular (you will have defined it 'through itself').

It might be said, in favour of Anon.'s reinterpretation, that it exonerates Plato of the most seriously fallacious move in his argument. 'Understand' introduces an opaque context, in which substitution of equivalent terms is not normally acceptable. If you understand the word 'water', and water = H20, this does not entail that you understand the formula 'H20'.45 Likewise, even if shoemaking = knowledge of making shoes, you can understand the word 'shoemaking' without necessarily understanding the formula 'knowledge of making shoes'. And although Plato does often enough fall victim to such fallacious substitutions,46 Anon. can be expected to be far more alive to their dangers,47 if only thanks to

45 See Burnyeat, art. cit. 389-90.

46 See Burnyeat, loc. cit., who compares Tht. 191d, 203c-d. Other possible occurrences include Gorg. 468d1-e5, Meno 77e5-78b2, Phd. 74b7-c6.

47 This virtually guarantees that at 21.16-24 Praechter (who otherwise greatly improved on the Diels/Schubart reading there) was wrong to retain 06

[T•] (22), where I propose o[u].

On his reconstruction, Anon. would take Socrates at 146e4-5 to be accusing Theaetetus' definition of amounting to the tautology that knowledge is knowledge, and thus undo the




the tip-off of the anonymous critics he has just quoted. Hence his reinterpretation obeys the widely-invoked principle of charity, whereby we don't attribute a fallacious argument to Plato where his text can plausibly bear an alternative, innocent reading.

But is the reinterpretation plausible? It would, I think, be highly plausible to someone, like Anon., whose rule-book is Aristotle's Topics. In Topics vi.4, when Aristotle expounds his principle that a truly scientific definition will define the posterior by means of the prior, this is presented in terms of the requirement that the definiens be 'more intelligible (yvwpipWCTEpov)'. But by this he turns out, not surprisingly, to mean 'more intelligible absolutely (yvwpip-

r•Epov 6ArrA•r)', since usually it is that which is posterior, e.g. the

species of the definiendum, that will be 'more intelligible to us'. Hence all Anon. need do in order to secure his reinterpretation is take Socrates' talk of 'understanding' in this passage to refer to that idealised scientific understanding which Aristotle describes as 'absolute', and any division between intelligibility and logical priority is erased. If that guess is right, Anon. has once again been able to place the Aristotelian Organon at the service of Plato's text.

It may seem disappointing to find Anon. legislating out of existence the epistemological paradox which the passage is often thought to propound, namely the version of the so-called Socratic Fallacy48 according to which if you don't know what the genus is- say, knowledge-then you don't know what its species- shoemaking, geometry, etc.-are either. But as a matter of fact, he does turn out to have views which bear directly on that supposed fallacy. For he endeavours to explain to us just how far the understanding of a term or concept does, and how far it does not, rely on the ability to define it.

To see this, we must start with the passage of the Theaetetus which appears to defend the Socratic Fallacy:

SOCR. Suppose someone asked us about some banal, familiar thing, such as what clay is. If we replied to him 'Potters' clay, oven-makers' clay, brick-makers' clay', wouldn't we be ridiculous?

good work that precedes. It would, in any case, be a quite implausible understanding of Socrates' words.

48 Cf., most recently, G. Vlastos, 'Is the "Socratic Fallacy" Socratic?', Ancient Philosophy 10 (1990), 1-16.



144 I-DAVID SEDLEY

THT. Perhaps. SOCR. First, I suppose, in thinking that the questioner understands our answer, when we say 'clay', whether we add 'dollmakers" or any other craftsmen. Or do you think someone has any understanding of the name of anything, when he doesn't know what the thing is? THT. None. SOCR. Then nor does someone ignorant of knowledge understand knowledge of shoes. THT. No. SOCR. Then whoever is ignorant of knowledge does not understand shoemaking, or any other craft. THT. That is so. (147al-b9)

On Socrates' question 'Do you think someone has any understanding of the name of anything, when he doesn't know what the thing is?', Anon.'s comment is as follows:

For if he is ignorant of the thing he will not know the thing's name either, because the name is a thing's sign. Hence whoever is ignorant of the thing will not know its sign either. Epicurus says that names are clearer than definitions, and indeed that it would be absurd if instead of saying 'Hello, Socrates' someone were to say 'Hello, rational mortal animal'. But definitions are used neither for greeting nor as being more concise than names, but for unravelling our common conceptions. And this is not possible without grasping each genus and the differentiae. (22.31-23.12)

Anon.'s view on the function of definitions is characteristically Middle Platonist. The Hellenistic idea of natural 'common conceptions' (KOIV(ai VVOaI I, npoA••E q) of things was married to the Platonic theory of recollection, the result being an account whereby innate common conceptions--common, that is, to all rational beings--can be articulated into full-blown definitions, especially with the help of a dialectician's skill in intellectual midwifery.49 Thus definitional understanding is not a prerequisite of our ability to understand a word, as regards either ordinary linguistic under-

49 For Anon.'s use of this theory, see also 46.34-48.35, 52.44-54.13, 55.14-33, 56.11-37, 57.11-42. And cf. Alcinous, Didaskalikos 155.17-29.




standing of the word itself or a general grasp of the thing it names. It is required only for fully articulated dialectical understanding.

This is then applied to Socrates' actual argument. On Socrates' inference at 147b7, 'Then whoever is ignorant of knowledge does not understand shoemaking', Anon. comments:

He does not mean that without knowing knowledge it is not possible to know shoemaking,50 but that if someone posits that shoemaking is knowledge, but is ignorant of what knowledge is, he will also not possess the idea of shoemaking to the extent that it is called knowledge. (23.15-25)51

It is easy here to see how a unitarian, reading the supposed Socratic Fallacy in the light of the theory of recollection, can present Socrates with a sanitised and perfectly innocent formulation of it. When you do not know the definition of a thing, it need not follow that you are totally ignorant of it and of its species or instances, let alone of the very meaning of its name. You simply lack full dialectical understanding of it. You cannot be altogether ignorant of it, because the conception of it is innate in you, even if not yet properly articulated, and provides a sound basis for your everyday unreflective understanding of words.52 In the present passage it is not linguistic understanding but only dialectical understanding-an understanding of shoemaking qua species of knowledge-that Socrates means to deny to anyone ignorant of the definition of knowledge.53

50 I.e. 'know what shoemaking is'. I doubt if there is any reference here to another possible misunderstanding, viz. 'know how to make shoes'. To read Anon. that way would be to miss the real philosophical interest of his distinction.

51 At 23.17(-18) I correct Diels and Schubart's br[i1 OTriv / <o0K E"0oTV> to 06[K] EOTIV.

52 Some other Middle Platonists would deny this: especially Alcinous, Didaskalikos 154.29-156.13, who associates everyday understanding with a separate, empirically derived faculty, 'doxastic (as opposed to epistemonic) reason'; also probably Plutarch. In support of this reading of recollection (including the Plutarch evidence), see D. Scott, 'Platonic anamnesis revisited', CQ 37 (1987), 346-66. But Anon., like most modern interpreters of the Phaedo, holds that pre-dialectical innate ideas already afford us access, albeit fuzzy, to their objects. See especially 46.43-9: 'Natural conceptions need articulation. Before that, although they focus (in4rr•Xouoi) on things as a result of possessing traces of them, they do not do so clearly.' However, he differs from his modern counterparts in apparently not counting this pre-dialectical use of concepts as 'recollection', to judge from 57.11-42, where, equating 'pregnancy of soul' with recollection, he allows that not every soul is pregnant in every incarnation.

53 For a largely similar reading of the present passage (though without recollection), see Burnyeat, art. cit.



146 I-DAVID SEDLEY

So on Anon.'s interpretation there is no argument here against our ability to operate successfully with a word even when we have no grasp on its definition. Socrates is merely pleading for the priority of the genus in definitional understanding (presumably to be equated with that idealised understanding which Aristotle intended with his expression 'more intelligible absolutely'). The interpretation is arguably consonant not only with the text of Tht. 146-7, but also with the sequel, where a version of the 'Socratic' fallacy is rejected as sophistical (196d ff.), but where the vicious- ness of definitions which fail to observe the priority rules is reasserted in the final argument (210a).

IV

Mud. Finally, we come to Socrates' own specimen definition, that of mud (which in the interests of clarity I shall substitute for 'clay' from here on):

SOCR. Moreover, although able to give a banal and brief reply, he is going round by an interminable54 route. Just as also in the question about mud, a banal and simple answer is presumably to say that earth mixed with moisture would be mud, and to forget about the question what it is of. (147c3-6)

Anon.'s comment exemplifies an occupational hazard of all harmonisers of Plato and Aristotle, the problem of what to do when Aristotle directly criticises Plato:

Aristotle criticises definitions of this kind, e.g. that of snow as frozen water, that of mud as earth mixed with moisture, and that of wine as putrefied water (in Empedocles' words, 'water putrefied in wood'). For neither is snow water, he says, nor is mud earth, nor indeed is wine any longer water. 'For one must not' he says 'accept55 this for all things of which the genus is not truly predicated, but for those of which the genus assigned is truly predicated.'

54 Anon. comments (24.7-10): 'because particulars are in a way infinite (arrn[E ]pa)'. This allusion to the doctrine of Philebus 16c ff. is less promising than it at first looks, because species, such as Theaetetus has listed, are not particulars, and would therefore on most interpretations of the Philebus count there as a determinate many, not antirIpa.

55 The papyrus has 6Irro8 IKTrOV for Aristotle's xrroSEKTEov; this makes little sense, and is best treated as a simple scribal error, not a varia lectio.




Let us concede that in the case of wine the genus is not truly given by saying that it is water. For someone will say, 'Even if it is granted that it was previously water, now at least it no longer is, and it certainly does not change back into water.' How in the case of mud can we avoid saying that it is earth affected in such-and-such a way by moisture? For earth remains, and that is also why when the moisture dries up it will be earth again. Just as in the case of wind, having doubted whether one should say that it is 'air in movement', he added 'But if after all one should in this case concede that it is air in movement...', so too in the case of mud he should have said the same. For the definitions were assigned on the same basis. (24.30-25.29)56

Aristotle (Top. 127a3-19) had included the Theaetetus 'mud' example in a list of ill-formed definitions which allegedly treat the material origin of the definiendum as if it were its genus. The genus must be truly predicable of the definiendum, but water is not truly predicable of snow or wine, therefore snow cannot be defined as 'frozen water' nor wine as 'putrefied water'. By the same token, mud cannot be 'earth mixed with moisture'. And wind should not be defined as 'air in movement': it is not the air itself, since that same air can remain after the wind has gone; rather, it is the movement of the air.

What are Anon.'s options? He might dismiss the predicability principle itself as invalid. But that would run quite contrary to the authoritative status he assigns to the Topics. Or he might deny that Socrates intends the mud definition as a serious definition at all. But that would be at best a Pyrrhic victory.

Or again he might have anticipated a favourite Neoplatonist tactic and argued that Aristotle's criticism is directed only at those who misunderstand Plato's mud definition as one through genus and differentia. After all, Aristotle is himself also keenly interested in definitions of the form 'a = b + c',57 and not only does the mud definition lend itself to such analysis but it would also, thus understood, anticipate the form of the later definition in the Theaetetus, 'Knowledge is true judgement plus an account'. Anon.'s failure to

56 At 25.9 I read o6, with Praechter, in place of bi. At 25.19 the pap. reads [nr]vov-ros, and there is no need to follow Diels and Schubart in emending this to

nv6pacfoq. 57 Top. vi.13.



148 I-DAVID SEDLEY

pursue this promising path probably reflects his own preoccupation with the relevance of the whole passage to the proper roles of genus and species in definitions.

Thus Anon. is left with the option of accepting Aristotle's general criticism but proposing that this particular definition ought to have been exempted from its scope. His strategy is to take the wine definition as the paradigm case: wine cannot be said to be water, since even if it originally came from water it is not water now and can never turn back into water. In that case, mud is different, he argues, since not only does it come from earth, but the earth must remain present in it because as soon as it dries out it becomes earth again.58

I would expect Aristotle to reply that the earth, like all material ingredients in a mixture, is present in the compound only as a potentiality.59 But here the Platonic theory of the elements, whereby the mixture is a mere juxtaposition of particles of the two elements earth and water, would probably be enough to save Anon. from having to concede the point. Thus, whether or not Anon. realises it, his residual disagreement with Aristotle may come down to one of physics, not dialectical method at all.

However, the most intriguing part of Anon.'s strategy is his exploitation, at the end, of Aristotle's hesitation as to whether the definition of wind as 'air in movement' really is guilty as charged. All we have to do, he suggests, is take Aristotle's hesitation about wind to extend to the very similar definition of mud, and we have him back on our side once more. A very impressive exercise in damage limitation!

V

Conclusion. Without meaning to endorse all that he says, I have tried hard, perhaps too hard, to show up Anon.'s analysis in its most favourable light. In applying to the text a preconceived view of Plato's thought, and in probing the arguments with the best currently

58 He would, of course, have to say something equivalent about snow. And if he had known Metaphysics VIII 2, he could have pointed out in his support that at 1043a9-10 Aristotle himself accepts the definition of ice as 'frozen water'.

59 GC 327b22 ff. But if earth is the dominant element, Aristotle might allow that it remains in actuality too: see e.g. GC 321a34-b2.




available analytical tools, despite their anachronistic precision and technicality, he is anticipating what has now become a standard approach to Platonic arguments. And in one respect his approach even has advantages over twentieth-century analytic treatments of Plato. The Topics, Anon.'s trusty handbook, must represent work on which Aristotle was engaged within Plato's own school, not perhaps at the time the Theaetetus itself was written, but almost certainly in its immediate aftermath. Its evident appropriateness to the task Anon. has set himself is, we might say, no accident.

But the principal reason why we should take this commentator seriously is that, unlike most of us, he is a card-carrying Platonist, reading the text of Plato with a respect and commitment which few of us can hope to match. It would be unwarranted complacency to assume that we have nothing to learn from him.



PLATO, THEAETETUS 145-147

David Sedley and Lesley Brown

II-Lesley Brown

As the author was invited at short notice to give the reply to Dr. Sedley's paper, it was not possible to include her paper in this volume. Copies of the paper will be available at the Joint Session, and it will be published in Part 1 of the 1994 Proceedings.



Sedley & Brown-Plato _Theaetetus_ 145-147 (1)

Documents

jstor terms

jstor archive

david sedley

platos theaetetus

terms conditions of

amall use subject

aristotelian societystable

supplementary volumes