1 ἐξ ἀχοῆ ζ, b8-c1, cannot mean ‘by hearsay,’ as Cornford and McDowell for example translate, for hearsay evidence is not admissible under 4th century Athenian law: cf. Harrison, A.R.W. The Law of Athens vol. II (Oxford: Oxford U. P. 1971) 145-6.Google Scholar The phrase must mean to contrast the eyewitness, who has the evidence of his own eyes, with the Juror, who can only listen to the eyewitness testimony of others.

2 Cf. Gorgias 454d1 ff., Meno 97c11 ff., Symposium 202a2 ff., Timaeus 51d5 ff.

3 Plato's general policy in the Theaetetus, as it is now most often perceived, is to argue for the philosophical conclusions at hand with a minimum of baggage which is extraneous for his immediate purposes: so for example Robinson, Richard ‘Forms and Error in Plato's Theaetetus,’ reprinted in his Essays in Greek Philosophy (Oxford: Oxford U.P. 1969) 39–73;Google Scholar Runciman, W.G. Plato's Later Epistemology (Cambridge: Cambridge U.P. 1962)Google Scholar chapter 2; Cooper, John M. ‘Plato on Sense-Perception and Knowledge (Theaetetus 184-186),’ Phronesis 15 (1970) 123–146,CrossRefGoogle Scholar and McDowell, John Plato Theaetetus (Oxford: Oxford U.P. 1973)CrossRefGoogle Scholar passim. We must be sure then in interpreting the present argument that we do not attribute to Socrates a more strongly restrictive notion of knowledge where a weaker would suit his purposes equally well.

4 Hintikka, p. 79 (Hintikka, Jaakko ‘Time, Truth, and Knowledge in Aristotle and Other Greek Philosophers,’ in his Time and Necessity, (Oxford: Oxford U.P. 1973) 62–92).Google Scholar Readings of the passage along these lines are at least as old as Schleiermacher: ‘[der] Unterschied zwischen der auch mittelbar zu erlangenden richtigen Vorstellung, und der allemal und in allen Dingen nur unmittelbaren Erkenntnis” (p. 177 of his introduction to the Theaetetus, Part 2, Volume 1, of his Platons Werke, 2nd. ed. (Berlin: 1817)). A similar view appears in Bluck, p. 32: the case of the Juryman suggests ‘that the distinguishing mark of knowledge is personal acquaintance with the truth’ (his italics)… ‘you may be said to have knowledge of the truth instead of mere true belief about it only if you have had some sort of personal experience of it’ (Bluck, R.S. Plato's Meno (Cambridge: Cambridge U.P. 1961)Google Scholar. Bluck believes that in this example, ‘Forms are not mentioned but we are probably meant to think of them’ [p. 214]; note that in his view, the only kind of knowledge we can have of forms is knowledge by acquaintance, which ‘could not be expressed in any kind of proposition’ [pp. 212f.]. See also Rorty, p. 228; ‘juries can have true opinion but not knowledge; direct witnessing of some sort is a necessary, though not a sufficient condition for knowledge’ (Amelie Oksenberg Rorty, “A Speculative Note on Some Dramatic Elements in the Theaetetus,” Phronesis 17 (1972) 227-238). Rorty suggests that Plato might here have adopted a less austere view of knowledge; in fact, however, ‘it is of course’ the position quoted that Plato actually pursues.

5 This second account of Plato's counter-example takes a much less restrictive view of what can count as knowledge. Even on this weaker view, however, we do not deny that in general the eyewitness holds a privileged position. He alone saw what happened, and no Juror however diligent who was not on the scene can later see what he saw. Nor can the event be repeated for the Juror's benefit. (for a discussion of this and related issues, see Thompson, Manley ‘Who Knows?’, Journal of Philosophy 67 (1970) 856–869.CrossRefGoogle Scholar) But the privileged position of the eyewitness hardly shows that he always knows, or than non-eyewitnesses never know. Suppose then that in some case other than that envisioned in the Theaetetus, the Juror knows. To be sure, there will still be a difference between the Juror's knowledge, which is inferential, and that of an eyewitness, which is direct and non-inferential. But we are not forced to suppose that only direct and non-inferential knowledge can be called knowledge at all.

6 Myles Burnyeat, ‘A Paradox in Plato's Distinction Between Knowledge and True Belief (Theaetetus 200d-201c),’ read at the Triennial Meeting of the Joint Committee of Greek and Roman Societies, Cambridge, August 1, 1978. I do not know if Burnyeat still holds the views I attribute to him here, and in any case, he is not responsible for the use I make of them. A revised version of his paper (which I have not seen) is to appear in Proceedings of the Aristotelian Society, Supplementary Volume 44 (1980).

7 Thompson, op. cit., note 5.

8 The best we can do is to suppose that reference to Justification is buried in ἱ χανῶζ b3, or in the idea that the Juror is ἄχροζ, c6, or that he has performed his Job well, c2, but this falls well short of what is needed.

9 The concept of well-justified true belief, as distinct from knowledge, is the natural complement of our assumption that in principle a Juror, who is not an eyewitness, can never know. No change in the circumstances of deliberation, then, could bring a Juror to know the truth in the technical sense reserved for the eyewitness. At the same time, however, the system is not capricious. To convict, a Jury must have a high degree of certainty. They must have better than mere true belief: that is, they must reach well-justified true belief. This concept is institutionalized in our practice by various constraints both on how the Jury is chosen, and on how it operates. There is little room, however, for the notion of well-justified true belief in the Athenian system of Plato's day, for the mechanics of the system do not allow it. There is at least prima facie the same difference between the eyewitness and the Juror whose business it is to evaluate testimony (cf. ἐζ ἀχοῆζ, b8-cl). But Juries are large (from 200 to 500 members), and no particular attempt is made to exclude partial Jurors beyond the heliastic oath to vote ‘without fear or favour’ sworn by all dikasts at the beginning of each year. Jurors are not disqualified by any prior knowledge or opinion they may have of the matter at hand, and there are not the same restraints on what evidence is admissible. Finally, Athens is usually agreed to be one of the cities in which no deliberation is allowed before the Jurors vote (Aristotle Politics 1268b9 ff., cf. Harrison, op. cit., p. 164). The absence of a notion of well-justified true belief does not by itself prove that the Athenians also lacked our assumption that a Juror by definition cannot know. But the two conceptions are so closely tied together that the conclusion seems probable, in the absence of clear evidence to the contrary. For Athens, therefore, in an idealization of the courtroom process, a Juror might well attain to actual knowledge. As we have seen, no Athenian would find grounds in his own legal system for supposing that only well-justified true belief is ever available.

10 One last defence of the view that, for Plato, eyewitnessing is at least necessary for our knowledge of empirical fact comes from Plato's description of the counter-example at b7-8: ὅταη ὁιχαίωζ πϵισθῶσιν ὁιχαοταὶ πϵρὶ ῶν ἰὁόντι μόνον ἔστιν ϵἰὁἐναι, ἅƛƛωζ ὁὲ μή, ‘whenever Jurors are persuaded in accordance with Justice on matters which only an eyewitness can know, and which cannot be known in any other way.’ But this evidence is entirely equivocal. On one reading, Plato is speaking wholly generally, asserting of the entire class of distant events that only eyewitnesses can have knowledge of them. But it is equally plausible that he means to introduce a restriction on the class of historical events to those which only an eyewitness can know, while excluding those cases in which a Jury of non-eyewitnesses can still get knowledge of what took place. So the passage does not force the conclusion that, for Plato, eyewitnessing has a necessary role in knowledge.

11 Cf. Timaeus 51e1-3.

12 Cf. Gorgias 454d1-455a7. It is a hallmark of persuasion as opposed to teaching in the Gorgias, not Just that it imparts belief rather than knowledge in an audience, but also that the man who uses persuasion himself lacks knowledge on the subject (458e-459c).

13 I am indebted to Myles Burnyeat, Robert C. Cummins, and Keith Quillen for helpful discussion of earlier drafts of this paper.