56 HUMANISM m 



point out that the simple in psychology can only mean 

 what it is no use to analyse further. 1 In other words, the 

 distinction of simple and complex is always relative 

 to the purpose of the inquiry. The elements out of 

 which the complex states of mind are put together do 

 not exist as psychic facts. In the actual experiencing, 

 most states of consciousness form peculiar and recognisable 

 wholes of experience, which feel ( simple. Thus the taste 

 of lemonade is emphatically not the taste of sugar plus 

 the taste of lemon ; though of course it is by squeezing 

 the lemon and dissolving the sugar that we compose the 

 lemonade and procure ourselves the taste. The ex 

 periences which really are complex to feeling are 

 comparatively rare, as as e.g. when we feel the struggle 

 of incompatible desires. On the other hand, when we 

 reflect upon our experience, it is easy enough to represent 

 it all as complex, and to break it up into factors, which, 

 we say, were present unobserved in the experience. But 

 the justification of this procedure is that it enables us 

 to control the original experience, and the factors which the 

 4 analysis arrives at are whatever aids this purpose. It 

 is in no wise incumbent on us to go on making distinc 

 tions for their own sake and from inconsistent points of 

 view, without aim and without end. Indeed the practice, 

 though it seems to form the chief delight of some 

 philosophers, must be pronounced to be as such trivial, 

 irrelevant, and invalid. We have a right therefore to 

 declare simple and ultimate what it is useless to treat as 

 complex for the purpose in hand, and in this instance 

 we shall do well to avail ourselves of this right. For an 

 analysis of the valuation true and false, whether or 

 not it is possible for other purposes, would hardly be 

 germane to logic. 



1 I owe this definition to Prof. A. W. Moore s excellent account of the 

 functional theory of knowledge in Locke in the Chicago University Contributions 

 to Philosophy, vol. iii. p. 23. 



