Plato the Pythagorean. A critical study of Kenneth Sayre - (2009)

International The Journal of the Platonic Tradition he International Journal of the Platonic Tradition 3 (2009) 58-77 brill.nl/jpt Plato the Pythagorean A Critical Study of Kenneth Sayre, Plato’s Late Ontology, A Riddle Resolved 1 Denis O’Brien Centre National de la Recherche Scientifique, Paris Kenneth Sayre’s A riddle resolved was rightly acclaimed, when it first appeared (1983), as ‘unique and illuminating’ (Daryl Pullman), conveying ‘exciting ideas in a sober and calm way’ (Daniel H. Frank), and not least as ‘well worth the price’ (I paid only £10 for my paperback copy at the time). It is, even so, a dense and difficult book. he density—and not for a moment would one wish it otherwise—stems from the large number of ancient texts, quoted in Greek and in translation, and commented upon in 1) Sayre, Kenneth M. Plato’s late ontology, A riddle resolved, with a new introduction and the essay “Excess and deficiency at Statesman 283C-285C”. Parmenides Publishing. 2005. xxviii + 362 p. Two ISBN numbers are given, with no means of distinguishing between them (both are described as paperbacks, printed on alkaline paper), ISBN-10: 1-930972-09-1 and ISBN-13: 978-1-930972-09-4. Price not marked. On the verso of the title page, where one would normally expect to find such information, no address is given for the publisher other than ‘the United States of America’. On p. 337 n. 23 of the volume itself, the location is given as Las Vegas. Sayre’s book was originally published as Plato’s late ontology, A riddle resolved, by Princeton University Press, Princeton, New Jersey, in 1983, x + 328 p., ISBN: 0-691-07277-9. he new publication has the same pagination for the main body of the text (pp. 1-318), but contains nonetheless a number of corrections, or would-be corrections, as will be noted in the course of this study (see esp. n. 11 below). he new publication should therefore have been described as a revised, or corrected, reprint of the main body of Sayre’s work. I regard it as a serious fault on the part of the publisher that this has not been done, and that the innocent reader is left to suppose, wrongly, that the main body of the text is identical to the text originally published in 1983. (he description ‘New ed.’, included on the verso of the title page under the heading ‘Library of Congress Cataloging in Publication Data’, will be taken, unless otherwise specified, as referring only to the inclusion of a new introduction and an additional essay.) © Koninklijke Brill NV, Leiden, 2009 DOI: 10.1163/187254708X397397 D. O’Brien / he International Journal of the Platonic Tradition 3 (2009) 58-77 59 detail, both for the meaning they bear in the immediate context and for their consistency, or lack of consistency, with each other. he difficulty is inherent in the long-standing riddle that Sayre sets himself to solve. How is it that the various summaries which Aristotle provides of Plato’s philosophy, notably in the Metaphysics, appear to bear so little resemblance to what we learn of Plato’s philosophy from the dialogues? One no doubt all too simple answer to the question would be to take seriously what we are told by the author of the Seventh Letter, very likely Plato himself, when he protests that what Plato thought never has been, and he hopes never will be, put into writing (cf. 341 b 5-c 5). If Aristotle is summarising Plato’s ‘unwritten doctrines’, can we do otherwise than take on trust whatever he tells us of them? So speaks the timorous soul. Sayre’s thesis is more robust and more complex. However unfamiliar it may seem to the modern reader, what Aristotle tells us of Plato’s philosophy is there, waiting to be found in the dialogues, if only we delve deeply enough. Sayre’s efforts, as one would expect, are centred on the second half of the Parmenides and the Philebus. From his study of those two dialogues, Sayre draws out a history of Plato’s abandonment of his earlier theory of forms in favour of a philosophy that is indeed, as Aristotle says it is, a Pythagoreanstyle derivation of sensible objects and numbers from the One and an Indefinite Dyad. he second part of the Parmenides becomes therefore, by a pretty paradox, a repudiation by Plato’s Parmenides of the monism of his historical namesake, in favour of Pythagorean ideas that the historical Parmenides had sought to controvert. At the same time, the second part of the Parmenides and the Philebus tell the story, for those who have ears to hear, of Plato’s own repudiation of the forms of the Phaedo and the Republic, in favour of a pre-Parmenidean Pythagoreanism that reduces, perhaps does away with altogether, any separation between forms and their sensible counterparts, and in so doing frees Plato from the insoluble problem that had plagued him, throughout the earlier dialogues, of explaining how particulars ‘participate’ in the forms from which they derive their many attributes. hat double repudiation, by Plato himself and by Plato’s Parmenides, is the key to Sayre’s solution of the riddle of Plato’s late ontology. How well does Sayre’s bold programme work out in practice? An answer to that question can come only from a detailed appraisal of the texts that Sayre has quoted, an exercise that is not as simple as it may seem, since the translations Sayre relies upon for the pursuance of his thesis are not always 60 D. O’Brien / he International Journal of the Platonic Tradition 3 (2009) 58-77 the most reliable.2 So it proves to be with Sayre’s examination of the first hypothesis of the second part of the Parmenides (137 c 4-142 a 8), supposedly founded on the claim that Unity ‘is not many but one itself ’, ἀ ’ αὐ ἶ α Sayre’s translation (p. 42) of ῖ (137 d 1-2). hose words can have the meaning that Sayre has given them only if αὐ is taken as an intensifying adjective, bearing on the complement of the infinitive ( αὐ , ‘one itself ’). It is surely far simpler, and far more natural, to take the sentence as Cornford has taken it, with αὐ as an independent pronoun, the subject of the infinitive, not the complement. Having driven the young Aristotle to admit that, if the one has parts, then it is many and not one (137 c 5-d 1), Parmenides confronts him with the contradiction (d 1-2): ‘And yet it (αὐ ) has to be, not many, but one (ἕ ).’3 he first hypothesis will indeed end up severing the links of the one, not only with the ‘many’, but with any property other than that of sheer unity, so as to leave it, ultimately, deprived even of existence. But that is not, as Sayre thinks it is (p. 42), because, in the sentence quoted, ‘ αὐ ἶ α is understood to mean that this Unity exists in and by itself ’. he words αὐ are not here a variant (minus the article) of the expression ῦ αὐ ῦ (137 b 3) or αὐ ἕ (143 a 6 et alibi). In the second hypothesis (142 b 1-155 e 3), the one does exist, so much so that, since ‘one’ and ‘existence’ are not identical, the ‘one that is’ turns out to be no longer one. However far the one is divided, so Parmenides argues, ‘one’ always clings to ‘being’ and ‘being’ always clings to ‘one’, but 2) If the following pages of this study give a large place to questions of translation, it is in response to the express instruction of the Associate Editor (not specifically in relation to Sayre’s work): ‘Please add a comment or two on the translations. I am mounting a onewoman campaign against poor translations. Too many people think that these things don’t matter.’ Cor ad cor loquitur. In Sayre’s defence, I may add that many of the translations that I shall venture to query in this course of this study are not his alone. Were he to be so minded, Sayre could, in several places, cite in his support any one of a number of illustrious names, not least that of Cornford (in his translations of the Parmenides, the heaetetus and the Sophist). I would, even so, maintain my comments. But no doubt I am over critical. he author of a dithyrambic review (Anthony C. Daly), quoted by the publisher on the back cover of the volume, singles out for special praise the ‘accuracy’ of Sayre’s translations. 3) See F. M. Cornford, Plato and Parmenides, Parmenides’ Way of Truth and Plato’s Parmenides translated with an introduction and a running commentary, in the series International library of psychology, philosophy and scientific method (Routledge & Kegan Paul, London, 1939), ad loc. (p. 116): ‘But it is to be one and not many.’ D. O’Brien / he International Journal of the Platonic Tradition 3 (2009) 58-77 61 without ‘being’ ever becoming ‘one’ and without ‘one’ ever becoming ‘being’. he result, so we are told, is that what was supposedly one ‘always necessarily is becoming two, and never is one at any time’, Sayre’s transla’ ἀ ἶ α (142 tion (p. 55) of ἀ e 7-143 a 1). I have to struggle to give meaning to Sayre’s English sen) tence. Plato’s Greek is pellucid. he coupling of a participle ( and an infinitive ( ἶ α ) does not indicate a simple conjunction (‘and’ in Sayre’s translation). It expresses a relationship of subordination, in the context most simply taken as a causal dependence (‘since’). he object in ques), never ‘is’ ( ἶ α ). ‘Necessity’ tion, since it is always ‘becoming’ ( bears on that same causal relationship, and not on the participle clause alone (as in Sayre’s translation, ‘always necessarily is becoming’). he mean) by an adverb, as Sayre has done, is ing, if we translate the noun (ἀ that ‘Necessarily’, the object in question, ‘is never at any time one, since it is always in the process of becoming two’.4 he emphatic opposition of the two temporal adverbs, one positive, for , ‘never’), ‘becoming’ (ἀ ί, ‘always’), the other, negative, for ‘being’ ( justifies the expanded translation of the second, negative, adverb as ‘never at any time’. he same opposition gives point to the use of a present, as dis, not ). If what is ‘never tinct from an aorist, participle ( at any time’ one ( ) does not succeed in ever actually ‘becoming’ two (the meaning that would attach to , aorist), it is because it is only ever ‘in a process of becoming’ ( , present), a process which is never brought to a term, and so continues ‘endlessly’ (cf. ἀ ί). he paradoxical result, stated in the sentence following, is that the ‘one that is’, far from being ‘one’, turns out to be ‘indefinite in multiplicity’ [. . .] ῆ ). Cornford sees in that expression the (143 a 2: ἄ ‘unlimited series’ of actual numbers.5 Sayre (p. 55) is no doubt right to insist that, at this moment in the argument, we have to do, not with number, but with a ‘numberless multitude’. Does it follow, even so, that Sayre’s 4) he sentence begins with ὥ (142 e 7), not included in Sayre’s quotation. We have to ‘understand’ a verb in the indicative, since the noun that follows (ἀ ) is in the nominative: ὥ ἀ [sc. ί]. Literally: ‘As a result (ὥ ), there is necessity.’ If we translate with an adverb: ‘As a result, necessarily.’ he infinitive that follows ( ἶ α ) is governed by ἀ , not by ὥ , and has for subject ‘the one that is being’ (or ‘the being that is one’), taken over from the preceding sentence (142 e 6-7). 5) Cornford, Plato and Parmenides, p. 138. 62 D. O’Brien / he International Journal of the Platonic Tradition 3 (2009) 58-77 ‘numberless multitude’ is therefore a precursor of the Indefinite Dyad that Aristotle and others attribute to Plato and the Platonists? Possibly. But before we yield to the siren voices calling to us from the Philebus and from the Metaphysics, we need to keep in mind the strategy of α ), Parmenides’ own the ‘laborious game’ (137 b 2: α α description of the whole of the second part of the dialogue. he one of the first hypothesis, in excluding multiplicity, ends up not existing at all (141 e 9-142 a 1). Conversely, the one of the second hypothesis exists, but ends up not being one (cf. 144 e 3-7). You pays your money, and you takes your choice. If the one is one, it does not exist. If it exists, it is not one. he joke may not be to everyone’s taste, but I am pretty sure that we are meant to be amused. However ‘laborious’, the game is after all a game. he two hypotheses are deliberately (and amusingly) complementary, and contradictory. But in that case how are we to tell how much of the game—which of the two conflicting hypotheses, if either—the author of the dialogue would have us take seriously? Sometimes, with an individual argument, we can tell, but only when the context gives us an obvious clue, as it does in a passage which undoubtedly looks back to the Phaedo. Socrates, in the Phaedo (102 b 8-c 5), is shorter than Simmias, not because Socrates is Socrates, but because he participates in shortness. Conversely, Simmias is taller than Socrates, not because Simmias is Simmias, but because he participates in tallness. An identical reasoning is adduced for ‘more’ and ‘less’ in the Parmenides (149 d 9150 a 1). If the one is either ‘more’ or ‘less’ than things that are other than the one, it is not in virtue of the fact that the one is one. Conversely, if things other than the one are either ‘more’ or ‘less’ than the one, it is not in virtue of the fact that they are not the one. Either relationship, so Parmenides concludes, as Socrates had done in the Phaedo, arises only because largeness and smallness are ‘forms’ (149 e 9: ἴ ), existing independently of whatever it may be that is characterised by the form. So close a parallel can hardly be the result of chance. But in the following lines of the Parmenides (150 a 1-b 2) the theory is made nonsense of. ‘Smallness’, so Parmenides argues, if it comes to be present in the one, has either to be ‘stretched out’ through the whole of the one (150 a 4: α ), and is in that case equal to the one, or to ‘surround’ the one (ibid.: α), and is in that case larger than the one. In either case, ‘smallness’ is no longer ‘small’. he contradiction obviously arises because D. O’Brien / he International Journal of the Platonic Tradition 3 (2009) 58-77 63 the two verbs ( α and α, ‘stretched out’ and ‘surrounding’) present the form as spatially extended in the object that it characterises. But why should there be such an assumption? Why does the young Aristotle allow himself to be so easily persuaded by what would seem to us to be a caricature of the theory? he answer is not only that he has been deliberately chosen as someone who will answer Parmenides’ questions without making a fuss (cf. 137 b 6-8). he young Socrates, a far less compliant interlocutor, is just as much at a loss when Parmenides presents him with virtually the same conundrum, in the first part of the dialogue (131 b 3-c 1). Is a form present in its particulars in the way that a sail or awning is ‘spread out’ (cf. α ) over a crowd? Faced with that image, the young 131 b 8: α α Socrates is as helpless as the young Aristotle will be when faced with an almost identical image in the second part of the dialogue. Why is that? Why does the young Socrates allow himself to be sidetracked by what would seem to us to be a parody of the theory? Why are the two young friends both so easily taken in by the spatial metaphor? Sayre’s comment is as brisk as it is unilluminating (p. 24): ‘he sail is a poor analogy for the Forms, to which spatial dimensionality is totally foreign.’ Would that have been as obvious to a contemporary audience as it is to Sayre? In the introductory remarks to his analysis of place in the Physics, Aristotle observes that ‘everyone takes for granted that things that are, are somewhere’.6 It is true that, elsewhere (Phys. iii 4, 203 a 8-9), Aristotle excludes Plato from the generalisation, since he writes specifically that Plato’s forms are not ‘outside the heavens’, ‘for the simple reason that they are not “somewhere” at all’. And it is true that Plato himself, in the Timaeus, is emphatically of the opinion that it is only as ‘in a dream’ that ‘we claim that everything that there is must needs be somewhere, in a particular place and taking up a definite space, and that what is neither on earth nor anywhere in the heavens is nothing’ (52 b 3-5). But the two young friends of the Parmenides pretty obviously do share what Aristotle describes as the common belief that ‘things that are, are somewhere’. 6) Phys. iv 1, 208 a 29. In its context, the implication of the remark is that what is ‘somewhere’ occupies a certain ‘place’. he implication is made explicit in book three of the Physics. ‘What is somewhere is in place, and what is in place is somewhere’ (iii 5, 206 a 2-3). 64 D. O’Brien / he International Journal of the Platonic Tradition 3 (2009) 58-77 Does Parmenides’ comparison of the forms to an awning, in the first part of the Parmenides, do the contradictions attaching to ‘smallness’ in the second part of the dialogue, have a serious purpose in bringing home to a contemporary audience the feature of the theory that Sayre, as most modern readers of Plato, simply takes for granted—that the forms are not spatially extended? Were that, even if only in part, Plato’s purpose in the Parmenides, it would be difficult to tie in with Sayre’s reading of the dialogue. Why should Plato have tried to correct a misapprehension attaching to the theory of forms as put forward in the Phaedo, if his aim is to prepare the way for a new and very different ontology, where forms no longer exist apart from their particulars? Still more awkwardly, why should Plato have tried to win away the reader from the idea that ‘things that are, are somewhere’, when that very presupposition is firmly rooted in the Pythagorean ontology, supposedly about to replace the forms of the Phaedo? For Parmenides and his supposedly Pythagorean adversaries are pretty clearly prime examples of Aristotle’s generalisation. Parmenides’ ‘being’ is undoubtedly spatially extended (‘from the centre reaching out equally in every direction’, fr. 8.44: ἰ α ῃ), and in so far as his Pythagorean contemporaries identified ‘things’ with ‘numbers’, it is more than likely that they did so because they too thought of their numbers as subjected to what Sayre calls ‘spatial dimensionality’. Can the Parmenides still plausibly be taken as Plato’s preparation for his return to a Pythagorean ontology, if part at least of his purpose in the dialogue is, as it would seem to be, to cut away one of the crucial presuppositions that had underlain the earlier theory—the very theory to which he is supposedly about to return? Tempting though it is to pursue these perspectives, and difficult though it always is to tear oneself away from the fascinating complexities of the Parmenides, other dialogues clamour for attention, notably the two satellite dialogues, the heaetetus and the Sophist. Both texts are claimed by Sayre as indirect support for his thesis (pp. 206-238), but once again difficulties of translation obtrude. he first definition of the form of non-being in the Sophist, if we adopt Sayre’s translation (p. 233), turns on the ‘contrast’ or the ‘contradiction’ between ‘a part of the nature of Difference’ and ‘a part of that of Being’, ί ω α ῆ ῦὄ ἄ α 258 a 11-b 1: ἡ ῆ α ἀ ω ἀ ί . Sayre’s translation requires a repetition of both D. O’Brien / he International Journal of the Platonic Tradition 3 (2009) 58-77 65 ‘part’ and ‘nature’ (both ί and ω ) in the second term of the definition. But if we follow, as we are surely intended to, the definition given in the preceding lines of the negative form ‘non-beautiful’ (257 d 7-11), we see at once that the contrast lies with the whole of being, not with a part of being. ‘Non-beautiful’ is constituted by a part of otherness ω ). opposed to ‘the nature of the beautiful’ (257 d 11: ῆ ῦ α ῦ ‘Non-being’, likewise, is constituted by an opposition between ‘the nature ί ω ) and ‘the of a part of otherness’ (258 a 11: ῆ α nature of being’ (258 b 1: ῆ ῦ ὄ [sc. ω ]).7 hat simple opposition is obscured by Sayre’s inversion of the order of words in the Greek. Where the Stranger speaks of ῆ α ί ω (258 a 11), Sayre writes, not of ‘the nature of a part of Difference’, but of ‘a part of the nature of Difference’. It is easy enough to see why. How can any one ‘part ’of Difference have a ‘nature’ that is not the ‘nature’ of any other ‘part’? Negative forms do of course all differ in so far as the second term in the opposition differs (‘being’, ‘beauty’, ‘justice’. . .). But, in so far as they are all ‘parts’ of the same form (the form of Difference), how can the various ‘parts’ not have the same ‘nature’? Despite the order of ί ω ), the meaning, so it is all too words in the Greek ( ῆ [. . .] easy to suppose, must surely be ‘a part of the nature’, not ‘the nature of a part’. he opposition, as in Sayre’s translation, will lie between ‘a part of the nature of Difference’ and ‘a part of that of Being’. But no. he deviation from the text, trivial though it may seem, is at the root of Sayre’s misunderstanding. he reason for the slight oddity in the ί order of words that Plato has chosen (258 a 11: ῆ α ω , ‘the nature of a part of otherness’) is that we cannot but ‘hear’ the word which ends the first expression ( ω ) repeated as the complement of the feminine article in the second term of the definition (258 b 1: ῆ 7) Although the definition of ‘non-being’ (258 a 11-b 1: a ‘part of otherness’ opposed to ‘the nature of being’) is obviously intended to match the definition of ‘non-beautiful’ (257 d 7-11: a ‘part of otherness’ opposed to ‘the nature of the beautiful’), the former is not simply a generalisation of the latter, as Sayre, in these pages (pp. 232-234), would seem to suppose. ‘Non-being’ is clearly identified as a separate form in the lines that follow the Stranger’s definition (258 b 6-c 5). Moreover, the anatomy of the two forms is not the same. he form of non-being, like all other forms, including all other negative forms, participates in being, whereas the form of non-beautiful does not participate in the beautiful. But for these and other refinements of the Stranger’s theory of negation and non-being, the reader must look elsewhere. 66 D. O’Brien / he International Journal of the Platonic Tradition 3 (2009) 58-77 ῦὄ [sc. ω ], ‘the nature of being’). he same is not true of the penultimate word in the first term of the definition ( ί ). With the order of words in the Greek, there is no encouragement to ‘hear’ a repetition of ί following the reference to ‘being’ ( ῦ ὄ ), and therefore no encouragement, and certainly no compulsion, to understand the definition in the way that Sayre has done, with the second term in the definition construed as including a reference both to ‘a part of being’ and to ‘the nature of being’. If the order of words in the Greek had been as it is in Sayre’s translation (‘a ω ί ), the syntax of part of the nature of otherness’, ῆ α the sentence would have been radically altered. In the second term of the definition (258 b 1: . . . α ῆ ῦὄ ), we would have had to supply from the inner ear, as Sayre has done, both ω and ί , ω because it is required by the feminine article ( ῆ ), and ί because the word could not but have lingered in the inner ear as the last word in the expression preceding. he second term in the definition would therefore [sc. ω ]), have been not, as in fact it is, ‘the nature of being’ ( ῆ ῦ ὄ but ‘a part of the nature of being’ ( ῆ ῦ ὄ [sc. ω ί ]). he obvious parallel between the form of ‘non-beautiful’ and the form of ‘nonbeing’ (both of them opposed to a form, and not to ‘the part’ of a form) would have been lost, and with it Plato’s whole conception of non-being as a form which, like the form of being itself, is universally participated.8 Sayre’s understanding of the Stranger’s second definition is problematical both for the text he has adopted and for the translation he has given of it.9 Non-being, if we adopt Sayre’s text and translation (p. 234), is defined as ὂ ἕ α αὐ ῆ [sc. ῆ α ω ] ἀ (258 e 2-3), ‘each part of the Different opposed to “what is”’. ‘Each part’, in Sayre’s rendering, cannot be other than a translation of 8) If the part of otherness that is the form of non-being were opposed to only part of the form of being, then only whatever participated in that part of being would be at once ‘being’ and ‘non-being’. But it is essential to the Stranger’s theory that everything (other than the form of being itself ) is both ‘being’ and ‘non-being’, ‘being’ because it participates in being, ‘non-being’ because it is not identical to being. But again, for these refinements of the Stranger’s theory, the reader must look elsewhere. 9) In this and in the two paragraphs following, my remarks apply, exceptionally, to the text given in the original printing of Sayre’s volume (1983). An attempt seems to have been made to correct this text in the reprint (2005), but the result is a meaningless and impossible jumble. See n. 11 below. D. O’Brien / he International Journal of the Platonic Tradition 3 (2009) 58-77 67 [. . .] ἕ α . But that sequence of words is impossible. When is used with the article, it has to be placed in the so-called predicative position, either before the article (ἕ α ) or after the noun ( ἕ α ). With the text that Sayre has printed, the opposition would have to lie between ‘the part of the Different’ ( αὐ ῆ [sc. ῆ α ω ]) and ‘each being’ ( ὂ ἕ α ). But the text itself is wrong. he unanimous reading of the manuscripts is not, as in the text that Sayre has printed, ἕ α (accusative), but (genitive). Sayre’s accusative is taken from Simplicius’ commentary on the Physics, where Plato’s definition has been deliberately altered to match the definition of non-being given by Plotinus in the Enneads.10 With the genitive of the manuscripts restored to the text (258 e 2-3), the opposition lies between ‘the part of the Different’ ( αὐ ῆ [sc. ῆ α ω ]) and ‘the being of each thing’ ( ὂ ). All that is no mere pedantry. With the unanimous reading of the manuscripts restored to the text of the second definition (258 e 2: , not ἕ α ), and with both definitions translated correctly, ‘the nature of being’ in the first definition (258 a 11-b 1: ῆ ῦ ὄ [sc. ω ]) matches ‘the being of each thing’ in the second definition (258 e 2-3: ὂ ). he two expressions are not synonymous, and are not meant to be. he difference is the difference between form and particular. he ‘part of otherness’ that constitutes the form of non-being is opposed both to being as form (258 b 1: ‘the nature of being’) and to being as particular (258 e 2: ‘the being of each thing’). Far from the theory of forms in the Sophist pointing away from the earlier dialogues, as Sayre would have us believe, the distinction between form and the particularisation of form is still an essential feature of the theory.11 ἕ α 10) Enn. ii 4 [12] 16.1-3, echoed by Simplicius, Phys. 238.26-27. he reading of the manuscripts ( not ἕ α ) is repeated by Simplicius in an earlier passage of his commentary (Phys. 135.26-27). he error (ἕ α for ), perpetuated even in the latest Oxford text (E. A. Duke et alii, see n. 23 below), has been well ventilated in the recent literature. 11) he reader who has had the patience to read the three paragraphs preceding will no doubt be as bewildered as I am when I tell him that, in the reprinting of Sayre’s volume (p. 234), the final letter of the word ἕ α (cf. Soph. 258 e 2) has been changed to an upsilon, but with the accent of the word still given, though now impossibly, as proparoxytone, and with the translation ‘each part’ also unchanged. I can only venture the (admittedly uncharitable) explanation that someone (whether the author or the publisher or some 68 D. O’Brien / he International Journal of the Platonic Tradition 3 (2009) 58-77 Sayre fares no better with his analysis of the false logos, in the following pages of the dialogue (260 a 4-263 d 5). Aristotle repeats Plato’s definition of truth and falsity as saying, of what is, that it is, and of what is not, that it is not (truth), as opposed to saying, of what is, that it is not, and of what is not, that it is (falsity).12 ‘he fact that Plato and Aristotle can agree on the nature of truth and falsehood dramatizes the absence of any dependence in Plato’s account upon the concept of participation’ (Sayre, p. 238). Not a bit of it. he same definition of truth is to be found in a non-philosophical context in Xenophon’s Anabasis, where ‘to tell the truth’ ῦ α ) is again to speak ‘of things that are, as things that are, and of (ἀ things that are not, as things that are not’.13 he passage from Xenophon shows clearly enough that these are conventional definitions. In the Metaphysics as in the Sophist, they are to be taken as the explicandum, not the explicans. he repetition that Sayre has remarked upon shows only that Plato and Aristotle are concerned with the same problem, not that their solution to the problem is the same or even similar. Adoption of what was, at the time, a conventional formula is no indication at all that, in the Sophist, Plato already has in his sights a new ontology that would somehow bring him closer to Aristotle, or closer to the Pythagoreans. For a writer usually so clear and forthright, Sayre is strangely ambivalent in the account he gives of the heaetetus. On one level, all is sweetness and and ἄ which loom large in Sayre’s light. he two words ῆ account of the Parmenides (143 a 2: ἄ [. . .] ῆ ) appear twice in the ‘secret doctrine’ of the heaetetus (156 a 1: [. . .] ἄ , 156 a 8-b 1: [. . .] ἄ α) and are glossed by Sayre in essentially well-meaning but misguided reader) has attempted to change the reading of the text from an accusative to a genitive case, but without realising that, in doing so, he has to change the accent of the word (the genitive of ἕ α , with a final long syllable, has to be paroxytone, not proparoxytone) and seemingly without appreciating that, if he changes the Greek text, then he has perforce to change the translation that goes with it. 12) Aristotle, Met. Γ 7, 1011 b 26-27. Cf. Soph. 240 d 6-241 a 2, 263 b 2-d 5. I leave aside here such details as the difference between singular (‘what is’, ‘what is not’, Aristotle) and plural (‘things that are’, ‘things that are not’, Plato). Earlier in the Sophist, Plato uses a singular (237 a 3-4) to make the same point. 13) Xenophon, Anabasis iv 4.15 (of a messenger who can be counted on to bring back reliable news). D. O’Brien / he International Journal of the Platonic Tradition 3 (2009) 58-77 69 the same way. he ‘numberless multitude’ of the second hypothesis of the Parmenides (Sayre, p. 55) reappears in the heaetetus as a ‘total lack of determinate character’, ‘a mass or multitude with no limitation whatever, whether with respect to number or to characteristic’ (Sayre, p. 218). he reader not unnaturally assumes that, if the second hypothesis of the Parmenides can be taken as looking forward to Plato’s late Pythagorean ontology, then the same will be true of the ‘secret doctrine’ of the heaetetus, and is therefore not surprised to learn, in the concluding paragraph of Sayre’s account of the heaetetus (p. 218), that Plato ‘accepted’ the secret doctrine ‘as an account of the realm of becoming that was sufficiently detailed and plausible to provide the basis for further deliberations regarding the relationship between Forms and sensible objects’. But there is a dark cloud to that conclusion, for we are also told that the world as described in the secret doctrine is a world ‘without the Forms’ (Sayre, ibid.). How can Plato have adopted a world without forms as the basis for a new understanding of the relation between forms and particulars? An obvious answer may seem to be to restrict the world without forms to the sensible world, leaving open the possibility of there being another world, where there are forms. But that is easier said than done. Socrates ᾶ attributes to the adherents of the secret doctrine the belief that ἦ , followed by α ἄ α ῦ ὐ (156 a 5). Sayre’s translation (p. 207) ‘All is motion’, followed by ‘and nothing other than ᾶ ) and misplaces the refermotion’ minimises the role of the article ( ence of the demonstrative pronoun ( ῦ ). It is not ‘all’ but ‘the all’, ‘the universe’ (156 a 5: ᾶ ), that is said to be movement, while the neuter pronoun (156 a 5: ῦ ) cannot have for antecedent the feminine noun that precedes ( ), as in Sayre’s translation (‘nothing other than motion’). he antecedent of the demonstrative pronoun must be, or at ᾶ . Socrates’ point is, first, least must include, the neuter expression that ‘he universe is movement’ (156 a 5: ᾶ ἦ ), followed by ‘and apart from that, there isn’t anything else at all’ (ibid.: α ἄ α ῦ ὐ ). he verb (ἦ , translated by a present tense, on the supposition that this is a ‘philosophical’ use of the imperfect) has a predicative meaning in the ᾶ as subject and as complefirst part of the sentence, with ment, but is ‘understood’ with an existential meaning in the second part α ῦ ), ‘there is of the sentence, literally: ‘Other than that’ (ἄ 70 D. O’Brien / he International Journal of the Platonic Tradition 3 (2009) 58-77 nothing’ ( ὐ [sc. ἦ ]).14 he conjunction of ᾶ and α ῦ , with ᾶ as the antecedent of the demonstrative pronoun (156 a 5: α ῦ , ‘apart from this’, i.e. apart from the universe), may seem pleonastic (without special qualification, to say that there is ‘nothing other than the universe’ is simply to repeat that the universe is the universe). But ᾶ to mean ‘the all’ or ‘the universe’ was still, at the time the use of when Plato was writing, something of a neologism, and the repetition, far from being otiose, would have served to reinforce the meaning of a relatively new addition to the language.15 An exact translation (especially of ᾶ ) is needed for an appraisal of Sayre’s paraphrase, when he seeks to restrict the application of Socrates’ words to the sensible universe (p. 216): ‘All that exists in the sensible universe is motion.’ hat cannot be the implication of the words in the text. ‘he all’ that is the universe is all there is. For the adherents of the secret doctrine, it is not only ‘the sensible universe’ that is movement, it is ‘the universe’ tout court, the one and only universe there is, that is movement. Apart from the universe that is movement, there is no other. ‘he all’ that is movement is all there is. hat conclusion has only to be stated clearly for Sayre’s interpretation, no less than that of Cornford, to be seen to be impossible. A world, a universe, deprived of form, a world made up exclusively of movement, is not a world that Plato could have ‘accepted’ (Sayre, p. 218, quoted above). A he expression ἄ ὐ ἤ or ὐ ἄ ἤ (‘nothing other than’) is a standard form of words, with ἄ α (‘other apart from’) a common variant for ἄ ἤ (‘other than’). With ᾶ the antecedent of ῦ , the meaning has to be existential, not copulative, as it is in Sayre’s version. It would be meaningless to say, of ‘the all’, that ‘apart from that’, and therefore ‘apart from itself ’, ‘it is nothing’ (ἦ copulative). he meaning has to be: ‘Apart from that’, i.e. ‘apart from the all’, ‘there is nothing’ (ἦ , existential). In the translation adopted above (‘apart from that, there isn’t anything else at all’), the negation has been made to bear on the verb only because this gives a slightly more natural form of words in ὐ α ῦ English. ‘At all’ is added to mark the difference, in Greek, between ἄ and ἄ α ῦ ὐ , where ὐ has been separated from ἄ and placed last in order to bring out the point of the remark. he sentence that begins with ᾶ (‘the all’) ends, appropriately, with ‘nothing’ ( ὐ ). (he answer has to be ‘nothing at all’, if you start with ‘everything’, and ask what else there is.) 15) For the origins of Plato’s use of the expression ( ᾶ , ‘the universe’), see my contribution to the fourth Symposium Platonicum Pragense, ‘Le Parménide historique et le Parménide de Platon’, Plato’s ‘Parmenides’, Proceedings of the Fourth Symposium Platonicum Pragense, ed. A. Havliček and F. Karfík (Prague: Ο ΟΥ Ε Η, 2005), pp. 234-256. 14) D. O’Brien / he International Journal of the Platonic Tradition 3 (2009) 58-77 71 world, a universe, which excludes the forms is the antithesis of Plato’s world, and in the heaetetus it is introduced as such, by way of implicit demonstration that, in a world where there are no forms, not only can there be no knowledge (the central theme of the heaetetus); there cannot even be any adequate perception of the many things that we do so obviously see and feel around us. If, to take only one example, there is no largeness ‘in nature’ (the terminology of the Phaedo), then there can be no largeness ‘in us’, with the result that we cannot even perceive one person as taller than another. . . Far from looking forward to Plato’s new ontology, the secret doctrine of the heaetetus is a description of the nightmare world that would close in on us if there were no forms. Sayre claims (p. 216) that the proliferation of detail in Socrates’ elaboration of the secret doctrine is a sign that Plato developed the theory by way of support for his own beliefs. But no: the detail is not an indication of Plato’s own acceptance of the theory (Cornford), nor even of his own provisional acceptance of the theory (Sayre). Plato adds the detail so that we may see, in detail, the absurdity of the world from which he would release us.16 With the Philebus and the Politicus (the latter dialogue given extensive coverage in the new essay added to the reprint), Sayre is at last on firmer ground, in so far as both dialogues include a nexus of ideas and expressions that do undoubtedly look forward to what we are told of Plato’s philosophy by Aristotle and others.17 Even so, difficulties remain. What precisely is the new ontology that supposedly underlies both the Philebus and the 16) I do no more than skim the surface of my contribution to the sixth Symposium Platonicum Pragense (17-20 October 2007), provisionally entitled ‘How tall is Socrates? Relative size in the Phaedo and the heaetetus’. he Proceedings of the Symposium, given over to a study of the heaetetus, will be published in due course by Ο ΟΥ Ε Η Press (Prague). 17) he new essay on the Politicus was first published in electronic form in volume 5 of the Journal of the international Plato society. he references to ‘the great and the small’ and ‘the indefinite dyad’ in Alexander, hemistius, Philoponus and Simplicius, given as Appendix A of the electronic version, have not been included in the reprint. Were they ever to be published in printed form, they would need to be extensively revised. As they stand, they include two references with a non-existent line number (Alexander), one reference to a passage which has nothing to do with the question in hand (again Alexander), two references which overlap with the lemma (Philoponus), one reference which is exclusively to the lemma (Simplicius), and one reference to a passage which is no more than a verbatim quotation from the text of Aristotle (again Simplicius). 72 D. O’Brien / he International Journal of the Platonic Tradition 3 (2009) 58-77 Politicus, and quite how does it relate to the more familiar ontology of the Phaedo and the Republic? By way of answer to those two questions, Sayre assures us (p. 184) that, in both theories, ‘Forms are standards or paradigms by which sensible objects are given names and identified’, with the difference that, in the Philebus, ‘the doctrine of the Forms as paradigms is given a literal and unproblematic sense’. To illustrate the supposedly ‘unproblematic’ paradigms of the Philebus, Sayre quotes (p. 184) ‘the numerically exact interval of the meter, yard, or cubit’, in so far as such an interval ‘provides a standard against which to measure the length of individual sensible objects’. It would be cavilling, and foolish, to object that Plato never refers to a meter or a yard. But is it over-scrupulous to object that nowhere, in the course of the later dialogues (excluding therefore Alcibiades I ), does Plato ever refer to a cubit or a dactylos, taken in isolation as a unit of measurement? In the pages of the Politicus, Plato does indeed look back to the problem of relative size which had loomed large both in the Phaedo and in the preliminaries to the ‘secret doctrine’ of the heaetetus, and does now introduce a notion that had been noticeably lacking in the two earlier dialogues. ‘Larger’ and ‘smaller’, ‘more’ and ‘less’, so the Stranger argues, apart from their reciprocal relationship, require ‘the being necessary for generation’ ῆ ω ἀ α α α ὐ α ), a ‘being’ which is none (283 d 8-9: other, so he tells us, than ‘the condition of due measure’ (283 e 3: ῦ ύ ). (For both expressions, I repeat Sayre’s translation, p. 320.) he Stranger’s ‘due measure’, so Sayre tells us (p. 186), is comparable to the standard Cubit (or standard Yard or standard Meter): As the standard Yard can be known in numerical proportion to other linear measures, so the Forms Health, Beauty, and so forth, can be known by exact specification of the due measure by which they are constituted. hose words conclude Sayre’s resolution of the riddle of Plato’s late ontology. he doubting homas cannot but suppress a sigh, a sigh not only of disbelief, but of puzzlement. How can any standard, and therefore arbitrary, unit of measurement, whether yard, meter or cubit, possibly be taken as emblematic of the new Platonic form more pythagorico? How can the theory of forms as ‘paradigms’, in a ‘literal and unproblematic sense’, have included, even as an illustrative analogue, a ‘standard’ Cubit or a ‘standard’ Dactylos? Does not the specification of ‘measure’ as ‘due measure’ require D. O’Brien / he International Journal of the Platonic Tradition 3 (2009) 58-77 73 some element of form, necessarily absent from the arbitrary divisions of the measuring rod (or its equivalent, some part of the human body, whether hand, arm, foot or finger)? Even if we restrict our attention to ‘linear measures’, or more generally to the dimension of ‘size’, it is difficult to see how so arbitrary a measure as a ‘yard’ or a ‘cubit’ could be invested with any intrinsic significance. It is true that the explanation of comparative size in the Phaedo (Socrates is ‘small’ because he is ‘smaller’, Simmias is ‘tall’ because he is ‘taller’) does leave what we may see as an obvious conceptual gap, and does therefore call out for completion. We ‘perceive’ Socrates as ‘small’ only when we perceive him as ‘smaller’ than Simmias; but we may well wish to conclude, as well or instead, that Socrates is ‘small’, not only when compared to Simmias or to Phaedo, but in relation to some standard size—‘small’, for a man of his age and condition. Plotinus has an explanation of such ‘standard’ size in terms of form. For plants and animals, specifically for ‘man’ and ‘horse’, the appropriate ‘size’ ) is a property conveyed by the imposition of form, with all the ( imprecision consequent upon the conjunction of form and matter, accompanied, as it inevitably is, by a degree of ‘deformation’ when something ends up taller or shorter than it would have been if the form had been more successfully instantiated.18 he theory of ‘due measure’, as put forward by the Stranger of Plato’s Politicus, we may well suppose to have been intended as an answer to the same problem. But in that case would the theory of ‘due measure’ have excluded the presence of form? Plotinus is a firm believer in almost the whole panoply of Platonic forms as making up the hypostasis of Intellect. His conception of size as a function of form is intended as an extension, not a replacement, of the theory of forms. Was the same possibly true of Plato? he author of the Philebus and the Politicus does undoubtedly introduce new ideas. But does he therefore mean to abandon the old? One of the more puzzling features of Aristotle’s Pythagoreanising Plato is that he is made to speak (Met. A 6), almost indiscriminately so it might seem, of both ‘forms’ and ‘numbers’. Oddly enough, the same conjunction, of word and idea, appears explicitly in the Timaeus. Faced with the 18) Plotinus, Enn. ii 4 [12] 11 and Enn. iii 6 [26] 16. he remarks on ‘deformation’ are my own, added simply to fill out the theory in the light of the obvious consequence of Plotinus’ general understanding of the relation of matter to form. 74 D. O’Brien / he International Journal of the Platonic Tradition 3 (2009) 58-77 irregular movements of ‘the receptacle of becoming’, the Demiurge brings to birth the world of our perceptions by the imposition of ‘forms and numbers’ (53 b 5: ἴ α ἀ ῖ ). But the Timaeus also contains one of the most explicit, and certainly one of the most assertive, declarations of Plato’s belief in forms (51 b 6-52 a 7). Since Sayre believes that Plato has abandoned the theory of forms, voiced so vigorously in the Timaeus, in favour of the new ontology foreshadowed in the Parmenides, and all pervasive in the Philebus and Politicus, it is in his interest for the Timaeus to be placed chronologically before the Parmenides, preferably before the heaetetus and the Sophist, and certainly before the Philebus and Politicus. Along with G. E. L. Owen, Sayre therefore goes to great lengths on the question of dating, in order to free the later dialogues from ‘the shadow of the Timaeus’ (Sayre, p. 256, quoting Owen). But what if the shadow were a shining light? Is it possible that Plato’s late ontology, embracing both forms and numbers as in the Timaeus, was intended, not as a replacement, but as an extension, of his earlier ideas? I return, one last time, to Sayre’s translation of the text. When the Stranger of the Politicus has won young Socrates’ assent to the notion of comparative size (283 d 11-e 2: what is larger is so in relation to what is smaller, and vice versa), he introduces the notion of ‘due measure’ with words that come out in Sayre’s translation (p. 320) as: ‘Wouldn’t we say, rather, that there are things exceeding the condition of due measure, or exceeded by it. . . ?’ ‘Rather’, given emphasis by being placed between commas, undoubtedly implies, in English, that the new idea is marked as a replacement for the old. here is no such implication in the Greek. he Stranger’s question, in Greek, takes the form, 283 e 4-5: ἆ ’ ὐ αὖ . . . he adverb (αὖ), literally ‘again’, may indeed be used to mark an opposition, but it may also be used simply to indicate accumulation, or addition, as in Robin’s translation of the same text: ‘Ne dirons-nous pas, cette fois. . . ?’19 hat such is in fact the Stranger’s meaning, in this passage of the Politicus, is made abundantly clear by his next question (283 e 8-12), where 19) L. Robin, Platon, Œuvres complètes, Traduction nouvelle et notes, in the Bibliothèque de la Pléiade (Paris, 1950), vol. ii, ad loc. (p. 382). I quote the pagination of the volume open before me. Robin’s second volume was first published in 1943, and has been re-issued a number of times since, with various changes in format and pagination. D. O’Brien / he International Journal of the Platonic Tradition 3 (2009) 58-77 75 there is a fully-fledged ‘rather’ (283 e 10: ᾶ ), but where the word serves to indicate the difference between size taken exclusively as a reciprocal relation, as opposed to ( ᾶ , ‘rather’, the comparative adverb introducing the second and preferred option) the two explanations taken together, ‘large’ and ‘small’ as a reciprocal relation, and ‘large’ and ‘small’ in relation to ‘due measure’. In these two sentences (283 e 3-6 and 283 e 8-12), the new theory of ‘due measure’ is not presented as an alternative to the notion of comparative size, as put forward in the Phaedo and the heaetetus. When the young Socrates is asked to admit the validity of the new idea (‘due measure’), he is not asked to forego the conception of size as a reciprocal relation, a conception to which, only a moment before, he [sc. ῖ]). had given his unqualified assent (283 e 2: As so often, insisting on the exact translation of a single word, especially a word seemingly so trivial as αὖ (‘again’, ‘cette fois’), may seem impossibly pernickety. But here, as so often, the stakes are high. Does the new theory (‘due measure’) replace the old (‘large’ and ‘small’ as a reciprocal relation), or are the two theories introduced, not only as compatible, but as complementary? Sayre’s ‘rather’, in the first sentence quoted (283 e 3-6), would come down firmly on the side of replacement. But that is not the meaning of the Greek. Once we take the two sentences together (283 e 3-6 and 283 e 8-12), it is clear that the Stranger asks the young Socrates to admit the new notion, not to exclude the old. Is that why Aristotle writes, seemingly so glibly, of Plato the Pythagorean? Is the answer to the riddle of Plato’s late ontology that it was only ever intended to complement, not to supplant, the ontology of the earlier dialogues? *** he reprinting of Sayre’s work contains a few corrections. he lacuna in the table of contents has been made good, and at least in one place (p. 213) an obvious error in the translation of the Greek has been corrected.20 But all the many misprints in the Greek, some of which make the text unintelligible, still stare at one from the page, as does the foolish system of accentuation adopted in the original edition by the printers of the Princeton 20) I forbear to mention again the bungled attempt made at a correction of Sayre’s text on p. 234. See n. 11 above. 76 D. O’Brien / he International Journal of the Platonic Tradition 3 (2009) 58-77 University Press, with enclitic accents left in place even when the enclitic word is not included, and oxytone words printed barytone when they fall at the end of a quotation. It is no defence to say that the accents have been quoted ‘as they stand in the text’. A word that is printed with two accents in a running text, because it is followed by an enclitic, cannot but lose the second accent if it is printed without the enclitic. An oxytone becomes barytone only because it has another word, not an enclitic, following it, with no pause in the meaning. If the following word is not included in the quotation, then the final word of the quotation cannot but revert to being oxytone. hose are elementary rules of Greek accentuation, and it is folly for printers to think they can ignore them.21 A second, less obvious, but no less regrettable, feature carried over from the original edition is the author’s choice of lineation in his references to Plato. For the first text quoted in this study, from the second part of the Parmenides, I give the reference as 137 d 1-2, whereas Sayre gives (p. 42), for the same words, the reference 137 d 2-3. he discrepancy presumably arises because Sayre is following the lineation of the Loeb edition. his is as irritating for the French reader as it is for the English reader when French authors quote, as they sometimes do, the lineation of the Budé series. May I strongly urge both parties to stick to the lineation of Burnet’s Oxford text?22 21) When an oxytone word (other than the interrogative ) is included in the body of a running text (i.e. as anything other than the final word in its clause or sentence, and except when the word following is an enclitic), it is ‘put to sleep’ ( α ), as the Greek grammarians rather prettily put the point (the acute accent is replaced by a grave accent). When the word is not embedded in a running text (when it falls at the end of a sentence, or before a comma or colon), it remains oxytone. If an oxytone word is extracted from a context where it had been barytone, and is placed at the end of a quotation, it is therefore only logical that it should revert to being oxytone. 22) J. Burnet, Platonis opera recognovit brevique adnotatione critica instruxit, in the series Scriptorum classicorum bibliotheca oxoniensis, 5 tomi (Oxonii, 1st edn 1900-1907, 2nd edn, 1905-1914). he lineation in the two editions is the same. It is merely quixotic to attempt to return to the lineation of any earlier edition, and impossible to adopt, consistently or easily, the lineation of Henri Estienne’s edition, from which our page numbers are taken (Henricus Stephanus, Platonis opera quae extant omnia, 3 vols, Geneva, 1578), since the letters dividing Estienne’s pages into horizontal divisions are only approximately calibrated to the line divisions. D. O’Brien / he International Journal of the Platonic Tradition 3 (2009) 58-77 77 his may of course raise a problem in the future, since the lineation of the first volume of the new Oxford edition of Plato differs considerably from Burnet’s two editions.23 But whether the line numbers in references to Plato should therefore be adapted to the new edition is a question that will have to be addressed only when, if ever, the new edition is complete. By that time, I for one will almost certainly be safely in the grave. 23) E. A. Duke et alii, Platonis opera recognoverunt brevique adnotatione critica instruxerunt, in the series Scriptorum classicorum bibliotheca oxoniensis, tomus i (Oxonii, 1995).