Page:Mind (New Series) Volume 6.djvu/31

 ON THE INTERPRETATION OF PLATO'S PARMENfDES. 15 plurality but of a single all-embracing whole. The argu- ment which yields this result, though perfectly intelligible, is rather obscurely expressed, and has been most unjustly branded by Stallbaum with the name of fallacia, so that it may be as well to reproduce it in extenso. Each part, we say (157 D), must be part not of a mere multiplicity but of a single whole. For it is easy to see that if the multiplicity of which the unit is taken to be a part does not in some way form one coherent whole, the only way in which the unit can be a part of it is by being a part of each separate unit which is contained in it, and so, among others, of itself. Thus you have to meet a dilemma. If the part be primarily a part of each of the constituents of the multiplicity taken by itself, you have of course conceded the point at issue ; but if it be not a part of each and every separate constituent, then neither is it a part of them all taken together on the assumption, that is, that all the constituents taken together form a mere plurality and not a single whole. This is of course at once manifest in any concrete case. Where one element appears to enter as a component into several inde- pendent systems, as, for instance, when the same man is a member of several unconnected organisations, it can only be part of them all because it is part of them each. Thus a man may be at one and the same time an English subject, a Roman Catholic and a Teetotaler, but he is only all three at once because he is also each separately. And so we may see, in any case we choose to take, that what is a "part" is always primarily and directly a part of some one definitely organised whole. Or, as Plato phrases it : " Every part is a part not of a diverse and heterogeneous multiplicity, but of some one single reality (juta? rti/o? t'Sea? teal ei>o<? TIVOS) which we call a whole, and which is a perfect unity consti- tuted by all the parts ". Thus we may say of the manifold of existence, as of every lesser multiplicity, that it is " one complete whole of parts," or in a word that it, like the one reality of hypothesis 2, exists in the form of a system. And once more (157 E-158 A), not only is this true of the whole of existence but of every subordinate part in it. For when we speak of " each " part our very language recognises " each " as being itself a unity distinct from its companions. What is not " one" cannot be called " each ". But again none of these parts is mere undi versified unity, for they all were by our hypothesis " other than unity" (raa TOV ei/o<?). They are therefore not identical with it, and each " part " of the manifold, since it is neither a bare unit nor yet nothing at