1 See Casteñeda, H-N., (a) ‘Plato's Relations, Not Essences or Accidents, at Phaedo 102b2-d2,’ Canadian Journal of Philosophy, 8 (1978) 39–53CrossRefGoogle Scholar; (b) ‘Plato's Phaedo Theory of Relations,’ Journal of Philosophical Logic, 1 (1972) 467-80; (c) ‘E1 análisis de Platón de las relaciones y de los hechos relacionales,’ Critica, 14 (1971) 3-18.

2 The above principles and example are drawn from Castañeda (1972), 471.

3 Castañeda sees this, (1978), 43, where he correctly notes that, ‘this irreducible relational aspect [the derivative connection between Simmias and Socrates in (4) above] does not relate particulars, but connects Form instantiation factors [immanent characters].’ This aspect of Plato's Phaedo theory will be brought out in the following section.

4 Lewis, Frank, ‘Parmenides on Separation and the Knowability of the Forms: Plato Parmenides 133a ff.,’ Philosophical Studies, 35 (1979)CrossRefGoogle Scholar. has done much to clarify the structure of the ‘worst difficulty’ argument by making these general principles explicit. I differ with his interpretation of these principles, as will be seen.

5 One might object here that there is no reason to suppose that an analysis of relations utilizing the notion of immanent characters has been offered by Plato; i.e., that since ‘in us’ is used in contrast to terms which characterize Forms (e.g. 'itself,’ ‘that which is’) all that Plato is doing is contrasting the relations of phenomenal objects with respect to those of Forms. In answer to this let it be granted that 1) if x is in us then x is not a Form. I will then conceed that 2) there is no analysis of ‘in us’ in the text (nor is there an analysis of the types of predicates Forms are capable of receiving). But the text makes clear that the things ‘in us’ resemble Forms, that we possess them but not Forms, and that we are named after them. If the above objection is correct we should be able to translate all such talk in terms which refer only to sensible particulars. I cannot see that we are able to do this. While ἐν ἡμῖν at 133e5 seems Just to mean things like sensible particulars, a consideration of the entire argument makes clear that ἐν ἡμῖν is used ambiguously. I am concerned in this paper with its use where what appears to be picked out by the term are the manifestations of the Forms which particulars have ‘in’ them, and which are not identical to particulars; i.e., immanent characters.

6 Per Parmenides’ suggestion at 133dl-2.

7 Castañeda, (1978), 43

8 As noted by Plato: 1f one of us is master or slave of another … [he] is master or slave of another man’ (Parm. 133d7-el).

9 From here on I am only concerned with dyadic relations.

10 This reading is also strongly suggested by Rep. 438e11-439a3, where thirst is placed in the class of things which are ‘of something’ (relational). Plato points out that we may say that ‘it [thirst] is what it is in relation to something [drink],’ and that the ‘it is’ refers to thirst. Thus ‘thirst is thirst for drink’ is a correct reformulation and asserts that thirst is the same thing as thirst for drink. It does so by means of reiterating the subject term ‘thirst.' This interpretation of principle B defeats Lewis’ contention that Parmenides' argument fails on disanalogies between two statements concerning Formchains (statements of ‘proper separation’): (1) Geometrical knowledge is knowledge of figure, and (2) Mastery is mastery of slavery (117-18). To begin, (1) is not correctly formulated. I have shown that dyadic Form-chain statements are such that their predicative phrase must contain the name of the Form which is named in the subject position. Statement (1) does not observe this stricture and by mentioning three Forms (Geometrical Knowledge, Knowledge, and Figure) is not a statement concerning a dyadic Form-chain. To specify the relation Lewis has in mind in a statement analogous to (2), (1) must be recast as: ‘Geometrical knowledge is geometrical knowledge of known figure.’ Such a formulation is not open to Lewis’ criticisms.

11 This would seem to be true by definition, since any immanent character must be ἐν ἡμῖν - Plato's usual formula for immanent characters - and this rules out their being in (possessed by) a Form. Furthermore, immanent characters are only capable of being individuated in respect of the particular possessing them, e.g. ‘the largeness in Simmias’ picks out a particular immanent character. Now someone might try to claim that ‘Largeness itself is large in virtue of possessing largeness (an immanent character) in itself.’ But even if this made sense it would be wrong, for it is implausible to maintain that Largeness is large for any other reason except that it is so in virtue of its own nature. It can also be argued that to possess an immanent character is to resemble the Form responsible for that immanent character. But to be a resemblance of some Form is to imperfectly resemble that Form, and Forms are not imperfect. Hence, Forms do not possess immanent characters (cf. Phaedo 74c-75d, Parm. 134c). Finally, if Plato understands immanent characters to be nothing but the sensible exemplifications of Forms - and this seems reasonable - then since Forms are not apprehensible by means of the senses they cannot possess them. It should be noted here that if the possession of some immanent characters by Forms could be made possible for Plato, then there would be a way in which the ‘worst difficulty’ could be defeated. As will be seen, an immanent character of the requisite sort might be ‘the known ἐν ἡμῖν,’ which if possessed by a Form would establish the relation of knowledge with some knower possessing 'knowledge ἐν ἡμῖν.’ But within the context of this argument it cannot simply be assumed that Forms are ‘by their very nature’ knowable, in the sense that they simply are capable of possessing ‘the known ἐν ἡμῖν.’ To do so would beg the question.

12 Plato gives us one such statement at 134a9-bl: ‘knowledge in us is [knowledge] of only the things that are in us.'

13 This is the element implicit in Castañeda's explanation of the ‘irreducible relational aspect’ of particulars (cf. n.3). That is, on this principle one element involved in the fact expressed by ‘Simmias is taller than Socrates’ is that Simmias' tallness ἐν ἡμῖν is towards (πρόζ) Socrates’ shortness ἐν ἡμῖν (cf. Phaedo 102c10-d2).

14 I borrow this term from Lewis, which designates his corresponding version of principle C. I have tried to be careful in formulating this law to make it clear that Parmenides has said nothing which would involve him in denying the relation of participation between particulars and Forms.

15 I understand the statement descriptive of the relevant Form-chain here to be 'Knowledge is Knowledge of the Known’ (134a3-4). While the term in this statement which designates the Form the Known is of course a commentator's coinage, principle PA6 demands it. Furthermore, such coinage is Justified by analogy with the Mastery-Slavery case and on the principle that ‘there exists a Form for every collection of things [e.g., known objects] we call by the same name’ (Rep. 596a5-8. cf. Parm. 133all-b2). The term ‘the Known’ Just makes explicit the implicit relational character of what should be one Form, described as being ‘that very thing which is truth (real)’ (134a3-4). Truth (the real) is, after all, that which knowledge must be in relation to in order to be knowledge.

16 It would be interesting to see whether there are considerations which would show that a tripartite theory of relations employing immanent characters is logically preferable on Plato's view to a bipartite one utilizing only Forms and sensible particulars. Since the present version of the ‘worst difficulty’ argument crucially involves a commitment to immanent characters, I have suggested at the close of this paper that an analysis of whether Plato has such a commitment must be made before it can be said that Parmenides’ argument fails. I am currently at work on such an analysis. It should also be noted, however, that the 'worst difficulty’ would seem to be capable of being reformulated so as to cause the same sort of problems for a bipartite theory.

17 This includes a popular objection to the ‘worst difficulty’ argument (cf., Cornford, F., Plato and Parmenides, [London 1939). 99Google Scholar; Gulley, N., Plato's Theory of Knowledge, [London 1961], 118Google Scholar; Vlastos, G., ‘The Third Man Argument in the Parmenides,’ in Allen, R.E., ed., Studies in Plato's Metaphysics [London 1965). 258)Google Scholar. It is roughly this. A particular knower for Plato is not a sensible particular. Rather, that which knows is the soul per the doctrine of ἀνάμνησιζ. The soul, when it knows, is something χάθ´ αὑτό (in respect of sensible things such as the body) and akin to the Forms. Hence, the relation of knowledge is one between Forms and Form-like souls not requiring immanent characters and not, thus, subject to the law of factual separation. But of course, the more Form-like a soul is, the less possible it becomes to speak of any person's particular soul knowing something and being a knower. On the other hand, the less Form-like and more individualized a soul becomes, the more likely it is to possess immanent characters (e.g. ‘the Justice in the soul’) which account for its particular characteristic of being a knower, thus subjecting it to PA9. This sort of objection fails to come to grips with PA9 itself and does not take note that Plato's doctrine of ἀνάμνησιζ provides an explanation of our already having knowledge of the Forms - which is precisely what is being argued against here. The Appendix to this paper discusses an interesting objection made to my view made by William Prior.

18 I believe Lewis’ tentative endorsement, 112, of Forrester's criticism of Parmenides’ argument ('Arguments and [sic] Able Man Colud [sic] Refute: Parmenides133b-l34e,’ Phronesis, 19 (1974) 233-7) shows that he does not fully appreciate the independent status of the law of factual separation. On Forrester's reading, the statement of ‘proper separation,’ ‘Knowledge is knowledge of the known’ establishes by itself - without reference to an independent law of factual separation - that the Form the Known only has a correlative relation with Knowledge itself, and hence, with no knower in this world. That, on his reading, does not then forbid that some knower might know some Form in the domain of the Known. This cannot be right however. For if ‘proper separation' (my principle PA6) were the only principle supporting Parmenides’ argument as Forrester seems to suppose, then there would also be no principle with which to oppose the suggestion that a master might master some Form other than Slavery, e.g., Lionhood or Largeness itself. Since this suggestion is absurd, the law of factual separation, which does dispose of the suggestion, must be understood to be a principle in its own right. That is, a principle possessing grounds independent of a ‘principle of proper separation’ for its Justification; very like those I have suggested.

19 As Parmenides points out ‘we do not [i.e., cannot] rule over them [the gods and presumably Slavery itself] by virtue of rule ἐν ἡμῖν (134el-2). This explanation further supports my position that it is the nature of immanent characters (i.e., what sorts of relations they may serve as constituents in) which provides a Justification for the law of factual separation.