656 EXPERIMENTAL PSYCHOLOGY 



question is really one as to the basis of selection among the vari- 

 ous governing concepts which may be adopted. Such sources of 

 preference of course exist. Not all logical schemes for the subdi- 

 vision of a given set of phenomena are equally valuable; and the 

 classifications which as a matter of fact we make are those which 

 serve our practical or scientific purposes, not those whose appeal 

 rests solely upon logical simplicity or completeness. The purely 

 logical analysis of any subject-matter declines in importance as 

 the range of phenomena with which it deals is amplified; and in 

 every general classification of the sciences it will be found that 

 while within certain of its component groups the arrangement 

 appeals to the observer by the practical fitness of its affiliations, 

 in regard to other parts of the field the correlation of elements 

 seems to have been made with a view to logical completeness 

 only, instead of reflecting those associations which are most inter- 

 esting or most important for the scientist. 



The reason for this inadequacy does not lie in any failure to 

 carry out the divisions involved in the logical principle assumed, 

 nor, indeed, in any lack of value in the governing concept itself. 

 On the contrary, it arises from a deficiency which reappears as per- 

 sistently in the relations in which we view the objects of our scien- 

 tific activity as in those of our practical life. It is not our habit 

 except, indeed, when the problem of classification itself is in 

 question to carry a single logical principle continuously through 

 the whole series of related activities in which our thought is ex- 

 pressed. Within a narrowly limited field a given principle may 

 be of such dominating importance that the single system which 

 results from its application is of both logical and practical value. 

 But with every fresh extension a point is more nearly approached 

 where it will be disjjjaced in significance by some other concept; 

 in which case the usefulness of the logical classification will be 

 reduced in proportion as the set of relations prescribed by this 

 iresh point of view acquires greater importance in the mind of the 

 thinker. 



This dualism must always be faced by the systematizer if his 

 interest in the subject-matter be other than formal, and whose 

 is not? All logic worth the name is instrumental, not final. The 

 mental attitude which erects consistency into a principle of abso- 

 lute value is branded as scholastic pedantry; for all rational in- 

 terest in the logical relation of concepts roots somewhere in the 

 desire to understand the historical connections of things. Classi- 

 fication is good only in so far as it helps us to group things accord- 

 ing to their most important aspects. When it ceases to have this 

 value, we turn to some other principle which in its turn has gained 

 the ascendency. 



