22 MENTAL EVOLUTION IN MAN. 



Here, then, we have stated, with all the common-sense 

 lucidity of this great writer, what we may term the initial or 

 basal distinction of which we are in search : it is that " proper 

 difference" which, narrow at first as the space included be- 

 tween two lines of rails at their point of divergence, " at last 

 widens to so vast a distance" as to end almost at the opposite 

 poles oi mind. For, by a continuous advance along the same 

 line of development, the human mind is enabled to think 

 about abstractions of its own making, which are more and 

 more remote from the sensuous perception of concrete objects; 

 it can unite these abstractions into an endless variety of ideal 

 combinations ; these, in turn, may become elaborated into 

 ideal constructions of a more and more complex character ; 

 and so on until we arrive at the full powers of introspective 

 thought with which we are each one of us directly cognisant. 



We now approach what is at once a matter of refined 

 analysis, and a set of questions which are of fundamental 

 importance to the whole superstructure of the present work. 

 I mean the nature of abstraction, and the classification of 

 ideas. No small amount of ambiguity still hangs about these 

 important subjects, and in treating of them it is impossible to 

 employ terms the meanings of which are agreed upon by all 

 psychologists. But I will carefully define the meanings which 

 I attach to these terms myself, and which I think are the 

 meanings that they ought to bear. Moreover, I will end by 

 adopting a classification which is to some extent novel, and 

 by fully giving my reasons for so doing. 



Psychologists are agreed that what they call particular 



kind is performed mainly through the agency of verbal or other conventional signs 

 (as we shall see later on), and it is owing to a clearer understanding of this 

 process that Realism was gradually vanquished by Nominalism. The only 

 difference, then, between Locke and Berkeley here is, that the nominalism of the 

 former was not so complete or thorough as that of the latter. I may remark that 

 if in the following discussion I appear to fail in distinctly setting forth the doctrine 

 of nominalism, I do so only in order that my investigation may avoid needless 

 collision with conceptuallsm. For myself I am a nominalist, and agree with Mill 

 that to say we thmk in concepts is only another way of saying that we think in 

 class names. 



