76 DOGMATISM AND EVOLUTION 



turn, it is equally true that "concepts without percepts are 

 empty." The conditions of the possibility of experience, except 

 as they receive some particular filling in the actual course of 

 experience, are without significance. Their objective validity 

 depends on their reference to given objects. But while the valid- 

 ity of any universal proposition depends on its reference to con- 

 tingent fact, the content of the universal proposition, as such, 

 is supposed by Kant to contain no element of contingency. That 

 is to say, while the universal must depend for its validity on 

 particular given experiences, the meaning of the abstract univer- 

 sal as such remains the same no matter what the particular 

 filling may be. But a proposition which remains unchanged in 

 meaning, no matter to what contingent particulars it may have 

 reference, is a necessary truth pure and simple. 



It is the purity of the categories of the understanding, that is, 

 their absolute separation from all the particularity of perception, 

 which makes necessary the device of the schematism. Each 

 category has in itself a certain sort of meaning, that is, it may be 

 formally defined in purely universal terms; but as thus defined 

 it can never be applied to any objects of experience. It is, how- 

 ever, capable of being given an interpretation in terms of per- 

 ception; and it is as so interpreted (schematized) that it enters 

 into experience. What connection there is between the formal 

 definition of the category and its schema, is left wholly unex- 

 plained. Thus substance is defined as "that which may be con- 

 ceived as subject, without itself being predicate of anything else." 

 As schematized, however, it becomes the permanent in time. 

 Certainly the ground of connection between these two meanings 

 is far to seek. Indeed, as is well known, Kant says of the schema- 

 tism that it is "an art hidden in the depths of the human soul, 

 the true secrets of which we shall hardly ever be able to guess 

 and reveal." 



We have only to state Kant's position clearly, in order to see 

 both its near affinity to, and its divergence from, rationalism. 

 In showing that no particular contingent proposition can be 



