investigation condemned to remain in the realm of less or more plausible ‘judgement’ or ‘opinion’ (doxa).
(208d1–4), that it is ‘the brightest of the things in the heaven that orbit the earth’, while that of Theaetetuswould (209b2–c10) pick out, not shared features like his having a snub nose and bulging eyes, but the unique features of his snub nose and his bulging eyes. To the extent that these examples focus on particulars rather than universals as objects of knowledge, that they offer no system for arriving at the final differentiation, and that the unique featuresthat they select could well be regarded as inessential, they do not convey deep Platonic insights, but rather confirm Socrates' implied disclaimer when he calls this a popular criterion. The Platonizing implications of the definition are, in a way that by now we have come to expect, to be found in the Platonic subtext, not in Socrates' own words.
As in previous cases, we may expect that Socrates' grounds for rejecting the definition will convey the kind of critical insight that made him the midwife of Platonism. His opening contention is as follows (208e7–209d3). Being able to distinguish something from all other things is, in itself, a necessary condition even for having true judgement about it (for reasons which we have amply considered in connection with the falsity puzzles, pp. 123–5 above): until you have marked something off from all other things—at the very least by judging truly how it differs from them—it cannot even feature in your thinking asthe individual thing that it is.
Thisnow (209d4–210b3), finally, leads to a fatal dilemma about the cognitive relation in which the knower must stand to the added logos. If, on the one hand, it isenough to have true judgement about how the thing differsfrom other things, the added logos isredundant, because merely to have that thing featuring in your judgement in the first place already requiresat a minimum that you judge truly how it differsfrom other things. If, on the other hand, the knower's cognitive relation to the logos has to be stronger than true judgement, it seems that it will be necessary to know how the thing differsfrom all other things. And that yieldsthe circular definition that knowledge istrue judgement plus knowledge of the object'sdifference from other things.
In this way, Socrates' parting shot is a thoroughly Socratic one: the rejection of a proffered definition on groundsof its concealed circularity. Thiswasthe motif that underlay hisvery first refutation in the Theaetetus, hisrejection of the definition of knowledge as ‘shoemaking (etc.)’, on the ground that it istantamount to the circular ‘Knowledge is knowledge of making shoes, etc.’ (pp. 23–7
above). We saw there how it reflects critical techniques which Plato strongly associates with Socrates. By a kind of ring composition, that same Socratic insight has put in a reappearance right at the end of the dialogue, laying low the final, and most promising, definition of knowledge that Socratesand young Theaetetuscould conjure up.
It isnot hard to work out that thissame objection isequally threatening to all definitionsof knowledge astrue judgement plussomething.263Whatever the something may be—whether justification, analysis, warrant, differentiation, or, asin the Meno (98a1–8), ‘calculation of the cause’—the same problem will threaten to arise. It is not enough to have mere true judgement about the extra something, or simply to assert it, or for it merely to exist. The knowing subject can stand to it in no cognitive relation weaker than that of knowing it. And as soon as this is recognized, circularity sets in. Knowledge will be defined astrue judgement plusknowledge of something.
Although it isnormal to credit the discovery of thisviciouscircle to Plato in the Theaetetus, it in fact had an earlier history in his Socratic writings. In the Meno, a dialogue largely devoted to scrutinizing Socratic ideas, Socrates famously proposes (98a1–6) that correct judgement becomesknowledge when one ‘binds’ it by ‘calculation (logismos) of the cause (aitia)’. He next rephrases this relation itself in causal terms: it is ‘because of binding’ (desmōi, 98a8) that knowledge differsfrom correct judgement, where ‘because of’ is expressed by a Greek dative, a standard Platonic way of indicating a cause.264 And hisimmediately ensuing remark addsan eloquent twist in the tail. Here isthe relevant sequence (98a6–b5):
SOCR. … And thisiswhy knowledge issomething more valuable than correct judgement. And it isbecause of binding that knowledge differsfrom correct judgement.
MENO. It certainly doesseem that way, Socrates.
SOCR. And yet I myself am speaking not asone who knows, but asone who isguessing. What I don't think ispure guesswork is that correct judgement
263 Cf. Bostock (1988: 238–40). Since Plato wrote, a strategy has emerged for escaping this problem by appeal to epistemological externalism—the denial that only factorsof which we are or can be aware have a bearing on justification. But as far as I can see it would be anachronistic to treat Plato as allowing for any such possibility.
264 For Plato's causal locutions see Sedley (1998a : 115). Compare in particular Phd. 101b4–6, where ‘Ten ismore numerousthan eight by [dative] two’ isexplicitly interpreted ascausal.
and knowledge are different. If there'sanything else that I would claim to know—and there are preciousfew things of which I would claim that—this is one thing that I would add to the list of those that I know.
This subtle variation on a familiar theme, Socrates' disavowal of knowledge, is used to generate a paradox which looks far too carefully crafted to be dismissed as accidental.265Socrates here clearly distinguishes a Difference Thesis from a Causal Thesis. That is, he distinguishes the thesis that correct judgement and knowledge are different, from the thesis that a certain kind of binding isthe cause of that difference. Yet he also claims to know that the Difference Thesisistrue.
How does he know? According to the Causal Thesis, in order to know the truth of the Difference Thesis he would have to have bound it by calculating the cause—that is, by calculating the cause of the difference between correct judgement and knowledge. But that is exactly what he denies that he has done: his proposal as to what causes the difference between correct judgement and knowledge is, he hasnow confessed, a mere guesson hispart, and very far from being one of the tiny handful of thingsthat he would claim to know. Whatever he may mean by ‘guess’ in this context, it inevitably fallsfar short of anything that could be called ‘calculation’, or indeed of anything that could be thought adequate to underwrite a knowledge claim.
To put the paradox in its most succinct form, Socrates has asserted (a) that he knows that these two cognitive states differ; (b) that the cause of their differing is something at which he can do no more than guess; and (c) that to know something you have to have calculated its cause. In view of the scarcely deniable incompatibility between guessing and calculation, it seems impossible that all three of (a)–(c) should be true.
In thisway, Socrateslaysa double snare. He warnsus of the suspect status of his implicit definition of knowledge, by calling itskey element a mere guess. But he also, even more subtly, drawsour attention to where itsfault may lie, by spotlighting the same question that will later be brought to the surface in the Theaetetus: if knowledge istrue (or
‘correct’) judgement plus something, what is the minimum required cognitive relation between the knowing subject and the something? Socrates' confession that his own cognitive
265 The planting of such booby traps is a recurrent feature of the Meno. See p. 26 n. 41 above for another example.
relation to that extra something is, in this instance, one of mere guesswork both undermines the definition itself and, more specifically, invites the killer question: if not mere guesswork, what is the required cognitive relation between the knowing subject and the extra something?
In thisway, the Theaetetus' final failure to define knowledge as a species of true judgement builds on a suspicion which Plato had long ago depicted hisSocratesasharbouring. I should add here that even asearly asthe quintessentially Socratic Charmides Plato had indicated, if only in passing, Socrates' and his interlocutor's shared assumption that knowledge (epistēmē) and opinion or judgement (doxa) have different objects266—the primary basis of the radical separation between these two cognitive states which Plato would impose once his two-world metaphysics was in place.
Even now, then, asthe Theaetetus' arguments draw to a close, what we are witnessing is the working out of a recognizably Socratic agenda, with results that we are encouraged to see as pointing inexorably towards the Platonist epistemology of the Republic.