1. I leave open here the question how far Plato's depiction of Socrates in the earlier dialogues can be relied upon as evidence for the historical Socrates, and how far Aristotle may have been misled by taking it as such. See "Did Plato write Socratic Dialogues", CO 31 (198 1), ), pp. 305- 320. 2. There seems to be no recognition of this, the primary metaphysical influence on Plato, anywhere in Aristotle's work. Guthrie's attempt to explain the omission(A History of Greek Philosophy IV, p.35) misses the point, which is that Aristotle's perspective on his predecessors is not in our sense historical. The early Academy's enthusiasm for Pythagoras has led him to overestimate Pythagorean influence on Plato's own thought. Of course it may be through contact with Pythagoreans in Italy that Plato first encountered the Eleatic theory of Being. (See my article cited in the next note.) And Plato's view of mathematics may reflect Pythagorean influence. But the doctrine of Forms can be stated without reference to mathematics; it cannot be stated without reference to the Eleatic view of Being. 3. That Pythagorean doctrine is itself scarcely attested outside of Aristotle's report. For my own guess as to its historical basis in the fragments of Philolaus, see "Pythagorean Philosophy before Plato" in The Pre-Socratics, ed. A.P.D. Mourelatos, (New York, 1974), p. 174.
4. For quite a different evaluation of Aristotle's report by reference to the dialogues, see T.I. Irwin, "Plato's Heracliteanism," The Philosophical Quarterly 27 (1977), pp. 1-13. I am not convinced that what Irwin calls "a-change" (namely, the co-presence of opposites in a single subject without reference to time) was ever regarded by Plato as a case of change or flux. In Greek as in English, the words for motion and change imply a temporal dimension.
5. I assume that the Cratylus is roughly contemporary with the Phaedo or with the tenth book of the Republic, and that the doctrine of Forms in the Cratylus is essentially the same as in these two dialogues. This is a controversial view which I cannot defend here. 6. See my review of Paul Woodruff's book on the Hippias Major, forthcoming in Oxford Studies in Ancient Philosophy.
7. Sym. 187A. The quotation of D. 51 is contaminated here by an echo of the antithesis "convergent divergent" from D. 10. It is really surprising how many editors of Heraclitus have been willing to follow Zeller in using this loose quotation in Plato to correct the literal text of fr. 51 transmitted by Hippolytus, particularly since the "correction" involves a semantic loss, namely the term for "agreement," /)on!o
8. See The Art and Thought of Heraclitus (Cambridge 1979) pp. 168, 223.
9. See the image of the kyklos in Heraclitus D. 103; the cyclical thought is to be found in very many fragments. Compare kylkos in Phaedo 73B 1.
10. I do not doubt the Heraclitean authenticity of D. 82-83, although I doubt the Platonic authorship of the dialogue which contains the quotation. See also D. 61; D. 9, 13, 37, 110-111. Similar "relativistic" remarks are put in Protagoras' mouth at Prot. 334a-C.
11. J. Burnet, Plato's Phaedo on 70E 1.
12. There is a memorable echo of this Heraclitean conception of the phenomenal world, qualified by Plato's characteristic dualism, in the famous passage of the Theaetetus where Socrates is praising the philosophic life in terms that echo both the Phaedo and the Gorgias. Theodorus says: "If you could persuade all men, Socrates, as you persuade me, there would be more peace and fewer evils among human beings." Socrates replies: "But Theodorus, it is not possible for evils to be destroyed, for it is necessary that there always be something opposed to the good. But evils have no seat among the gods; they necessarily inhabit this our region of mortal nature. Therefore one should try to flee from here to there as quickly as possible." (Theaet. 176A; I am grateful to Martin Andic for reminding me of the Heraclitean character of this passage.)