588 P. K. RAY'S DEDUCTIVE LOGIC. desirable in an elementary treatise : since, on the one hand, con- troversy on fundamental points is to the beginner distracting and confusing ; while, on the other hand, the dogmatic neglect of all opposed views may lead to narrowness and incompleteness of treatment. At the same time, the method of reconciliation is in danger of leading to inconsistencies, and it may involve a certain straining of terminology ; from the latter of which difficulties at least Dr. Kay has not entirely escaped. I do not, for example, think that any defence can be found for the statement that " regarded objectively, that is, as something existing in things or objects, a concept is an attribute or a collection of attributes in which a number of individual things or objects agree" (p. 5). This passage involves a confusion of phraseology, if nothing more; and it is calculated to suggest to the student a meta- physical doctrine which it is hardly probable that the author himself holds. Is it not again misleading to say that "in Material Logic ... a concept must be a collection of actually existing attributes, a judgment an actual relation between two real concepts " (p. 15) ? On the whole Dr. Ray's own exposition of Logic is on nominalist rather than on conceptualist lines. He says little concerning the psychological character of concepts, but he de- votes two chapters to a discussion of terms, following in the main Mill's method of treatment. The following statement indeed sounds ultra-nominalistic, " With a system of perfect definitions, a science would attain perfection" (p. 61). The con- text, however, shows that the author supposes the denotation of the terms fixed before their connotation, and he implies that we must know completely the characteristics of the objects with which the science is concerned before our definitions can be re- garded as perfect. Now up to a certain point it is doubtless true that as a science makes progress, its definitions make progress too ; but it should, I think, be added that at least in the more abstract sciences the 'definitions may reach their culminating point of perfection before the same can be said of the sciences themselves. Dr. Ray charges Mill with treating verbal proposi- tions as of no importance ; but whatever may be the justification for this charge, he himself appears to go too far in the other direc- tion when he says that " a verbal proposition does not merely explain the meaning of a name, but expresses, like a real proposi- tion, a relation between phenomena or attributes " (p. 117). The writer must be again thinking of the case where we seek to give a connotation to a name of which the denotation is already definitely fixed. But this is hardly what Mill has in view when he is talking of verbal propositions ; and we ought not to overlook his treatment of definitions in his Fourth Book. He shows clearly how in practice definition is often a question of things, and not merely of words. In his treatment of abstract names Dr. Ray departs from the
