ANALYSIS AND SYNTHESIS 43 



polar, subordinate processes of destruction and repair. The point 

 is that, by seeing nutrition as the cause of an entire group of facts, 

 we simpHfy our reasoning, without thereby departing from the 

 facts, though we temporarily ignore them. This group of facts 'is 

 composed only of consequences. I no longer have any need to 

 notice it, I will find it again by reasoning when necessary. It is 

 five hundred facts reduced to a single one. In subsequent re- 

 searches, I shall have to pay attention only to the summary and 

 generative fact.' is This 'summary and generative fact' is not a 

 Platonic Idea, however, but a natural type or universal, derived 

 by abstraction from scientific observation and analysis. 



Taine's tendency was to retain deductive habits derived from 

 his rationalistic heritage and attempt to fuse them with induction 

 in his own special way. Thus, M. Paul's analysis of nutrition 

 includes a number of extremely general propositions (such as 

 'nutrition is a cause') and then proceeds to 'verify' them by con- 

 sidering various sets of facts (such as 'the relations and the nature 

 of the operations and of the organs'). i9 His claim is that the 

 former and the latter are so related that from the former 'one can 

 deduce' the latter. But this is exactly the point at issue. Does not 

 our very understanding of nutrition consist of these sets of facts? 

 Deduction is used here in a sense quite diflferent from that in- 

 volved in the syllogistic reasoning of formal logic. 



M. Paul then, somewhat tentatively, develops a theory which, 

 by analog)' from biology to psychology, is of central importance 

 to Taine's theory and practice of criticism, namely, that the 

 type or species, 'a fixed and Hmited form, which endures from 

 generation to generation', functions as a cause: 'How to know 

 whether it is an eflfect or a cause? By admitting, hypothetically, 

 that it is an efifect, then by verifying or refuting that hypothesis 

 through experience. If the function determines the type, one 

 ought to deduce from the function the existence, variations, and 

 persistence of the type. If the former is lacking, the latter ought to 

 be lacking. If the former varies, the latter ought to vary. If the 

 former persists, the latter ought to persist. If not, the latter is 

 independent of the former.'20 a preliminary examination of some 

 of the evidence— such as the persistence of vestigial organs which 

 no longer perform a function— reveals that probably the species 

 'is^ not a derived and dependent thing, but independent and 

 primitive'. However, at this point Taine backtracks somewhat: 

 'I am sketching a method, I am not advancing a theory. Consider 



