In lieu of an abstract, here is a brief excerpt of the content:BOOK REVIEWS 113 phers); nevertheless, I feel that the book would have been more effective pedagogically had they devoted more attention to them than they have. As a specific recommendation, I would suggest several short introductions at strategic places in the text devoted to a brief resum~ of the historical setting in which the philosophers to be discussed found themselves. Once again I should emphasize that, despite the criticisms (...) I have made, I think the book to be an excellent piece of work, which should prove quite successful as an undergraduate text in the history of philosophy. OLIVER JOHNSON University of California, Riverside Plato's Theory of Knowledge. By Norman Gulley. (London: Methuen & Co. Ltd.; New York: Barnes & Noble Inc. Pp. viii ~ 203.) Mr. Gulley gives an account of the development of Plato's theory of knowledge. The impetus to this development he finds in an inconsistency that underlies the attempt to combine an a priori method of analysis with a theory of knowledge based on the doctrines of recollection and the archetype-copy relation of Forms and sensibles. Only in the last chapter, "Mathematical Knowledge," does he suggest a theory of the relation of a priori to empirical knowledge which he thinks might have satisfied Plato. The major part of the work, the first three chapters, concerns the conflicts and readjustments occasioned by the underlying inconsistency. I shall discuss these chapters in terms of a few central points. In chapter I, GuUey attempts to get at the basis of the theoretical difficulties. He points out that Plato first considers philosophic method as independent of the theory of Forms and the theory of recollection (pp. 40 f). In the Meno it is a method of analysis, which is a forerunner of the Phaedo's method of hypothesis, of which the dialectical method is a further development. It is an a priori method. But the theories, which are first hypotheses examined by the method (Phaedo), become later "presuppositions for the effective practice of the method," as, for example, in the Republic. In "recollection" there is an intimate connection between Forms and sensiblcs, but in knowledge there is complete separation. This is what Gulley considers to be the source of the major difficulties. Since I do not find these difficulties, I shall concentrate on their exposition. Gulley notes in the Phaedo two meanings of "giving an account": (1) in terms of a method of propositional analysis; (2) in terms of participation of sensibles in Forms. His explanation of the first offers no difficulties, but of the second there are serious problems. His explication of the second is as follows: To "give an account" is "to explain one's possession of a Form by showing that it is necessarily implied by that judgment of the 'deficiency' of sensible characters which arises as soon as they are recognized" (p. 40). His argument for this interpretation is as follows: "... if to know a Form brings with it the ability 'to give an account', and if to know a Form is to be reminded of it by the sensibles which resemble it (74 b-c), the 'giving of an account' must be explicable within the conception of 'knowledge' implied in the use of the term in 74 b-c" (p. 46). He assumes that Plato is at least consistent. Now in 74 b-c Simmias agrees with Socrates that we have knowledge of equality "in and by itself" and that such things as equal logs and stones "lead us to conceive" of it. Only "recollection," not the preexistent knowledge of Forms is in question here. Since it is assumed that knowledge implies the ability to give an account of what is known, then this instance of knowledge, which occurs in conjunction with perception, implies a similar ability. But one cannot infer from this that the discussion of equality as Plato has discussed it, which is a discussion of it as an example of recollection, is the same thing as the giving an account of equality itself, much less that it is "the only type of account possible" within the theory of Forms. That it is the only possible type is in fact contradicted by the assumption of pre... (shrink)