44 ANALYSIS AND CRITICISM 



this idea, not as something estabhshed, but as a beacon.'2i The 



discussion here is, of course, pre-Darwinian (1856). 



M. Paul's methodological point is that, presumably without 

 ever leaving the region of fact and without bringing in any meta- 

 physical 'entities', he had arrived at a 'unique formula, generative 

 definition, whence will be derived, through a system of progressive 

 deductions, the orderly multitude of other facts'. 22 This had been 

 accomplished by the process of abstraction with which we are 

 already famihar (Chapter II, 'The Problem of Method'): 'A 

 group having been formed, we extricate some general fact from 

 it by means of abstraction. We concede hypothetically that it is the 

 cause of the others. Knowing the properties of the causes, we 

 verify whether it has them; if it doesn't have them, we will try 

 the hypothesis and the verification on its neighbours, until we 

 find the cause. Reuniting a group of causes or generative facts, we 

 seek to find by the same procedure which one engenders the 

 others. . . . Abstraction, hypothesis, verification, these are the 

 three steps of the method. No more are needed, and they are all 

 needed. '23 Furthermore, M. Paul claims— and this is the im- 

 portant consequence for Taine's philosophy of criticism— when- 

 ever you use this method, 'you discover a hierarchy of necessities; 

 there is such in the moral world as in the physical world. A civihza- 

 tion, a people, an age, has a definition, and all their characteristics 

 or their details are nothing but the sequel and developments. '24 

 The example of Rome is used to illustrate this point, more or less 

 repeating material from the Essay on Livy (final chapter of Part 



One). 



At this juncture, M. Paul has passed beyond analysis of method 

 and the idea of cause to speculations which may give us pause. 

 Some elements in his construction— which may result perhaps 

 from his method, perhaps from a tendency to apply that method 

 too ghbly— seem of questionable vahdity to the present-day 

 reader. For example, his summary of the Roman type uses the 

 notion of phrenology— then a popular theory, of course— to find 

 an example of a 'cause': 'A dry and plain spirit, which is probably 

 the effect of the original structure of the brain, a persevering and powerful 

 circumstance, which created the necessity in this people for think- 

 ing of its interest and of acting together, produced and 

 strengthened beyond measure the egoistical and pohtical 

 faculty.' 25 Two elements of the famous formula ('race' and 

 'milieu') are here invoked, but our ideas of race and heredity 



