PROBLEMS OF ANALYSIS AND CRITICISM 137 



"Gallic tradition'", but we recall, with him, too, that 'At one 

 time the best representative of Gallic wit was Heinrich Heine, a 

 German Jew; today, it might well be Aldous Huxley.' ^i Or are 

 we to take care of such examples by falling back on the old dodge 

 that 'the exception proves the rule'? 



At this point, too, we may well recall the difficulties presented 

 by Taine's ambiguous references to abstraction (Chapter IV). 

 William James, in a review of On Intelligence, stated the problem 

 as follows: 



'Every real abstraction is an extract (to use his happy termino- 

 logy) from a multitude of particular things or events which may 

 differ as to their other details. The British school says the things 

 are "similar" as to this character, but Taine affirms the common 

 character to be literally the "same" in all, thus giving it a sort of 

 ontologic status, a real existence differing from that of individuals 

 and events only in possessing superior stability and permanence. 

 The beauty and value of these abstract characters, or generalized 

 extracts, is that they are fertile, for they contain wrapped up in 

 them — sometimes obvious, sometimes latent, and to be discovered 

 only by a keen analytic eye — further properties, other abstrac- 

 tions. . . . The class of abstractions to which he is not thus indul- 

 gent differs from the former chiefly in its infecundity. It is that 

 o{ substances, such as matter, the ego, the faculties of the mind, and 

 what may be called the dynamic entities, as power, necessity, 

 cause, force, etc. Here his nominalism stands firm. . . . His best 

 and deepest reason for rejecting [this] class of abstractions is that 

 they really explain nothing. . . . We find M. Taine constantly 

 forgetting this point of view. . . .'^2 



Is the Master Faculty a false 'substance' or a 'real abstraction'? 



Finally, if we have defined what we mean by a 'universal and 

 permanent cause', and shown it to be a 'true' abstraction, how can 

 we be sure that the hierarchy we establish on its basis is the only 

 valid one? The possibility always remains — stated for criticism 

 by the 'perspectivism' of Wellek and Warren — that for other 

 purposes, and from other points of view, other hierarchies might 

 be set up and other causes found to be universal and permanent. ^^ 

 In other words, how are we to decide the metaphysical issue 

 which, as Stephen C. Pepper has recently shown in persuasive 

 detail, 54 underlies the conflicts of critical systems? 



These questions are important for our study of Taine because 



