ANALYSIS OF SCIENTIFIC CONCEPTS 251 



or as the measure of motion; these suggest alternative routes 

 of derivation. The verbal similarity is a source of error, for 

 if the routes of derivation are different the concepts them- 

 selves would presumably be differently defined; at least 

 the concepts should be assumed to be different until the 

 routes of derivation have been demonstrated to be equiv- 

 alent. 



In the third place, a proper understanding of the opera- 

 tional derivation of any symbol demands that one ascertain 

 whether the method has been that of construction or that 

 of hypothesis. The distinction between these two methods 

 has already been clarified. 1 But the decision in any given 

 case is difficult to make. A symbol may be considered a 

 construction if it contains only such content as may be 

 derived from its empirical foundation through the use of 

 the required operation; but a symbol may be considered 

 an hypothesis if it contains such additional content as must 

 have been obtained through the use of creative insight. 

 The distinction is, of course, a fine one, and requires a pre- 

 cision in techniques which is not easy to attain. Further- 

 more, the difficulty is augmented by the fact that, although 

 one may know in a general way what the operational deriva- 

 tion has been, scarcely anyone knows specifically what the 

 rules for the employment of the operation are; hence he is 

 not in a position to estimate the operation as having been 

 correctly or incorrectly performed. The point is, simply, 

 that one does not know in any very precise way what ab- 

 straction, association, serial extension, and the rest are; 

 consequently he is not able to formulate any rules according 

 to which symbols should be given meanings through the 

 use of such techniques. As a further consequence, therefore, 

 one is not able to say when a symbol has been increased 

 through imaginative activities. This makes the discussion 

 of the operational derivation of scientific symbols essen- 

 tially unsatisfactory. The best that can be done is to throw 

 out certain suggestions, indicate obvious inadequacies, and 



1 Chapter IX, pp. 185-186. 



