No. 558 J 



CAUSES AND DETEIOIIXEL'S 



355 



spects, and in a greater number of ways, the causal analy- 

 sis may be carried to any extent desired. In this way is 

 reached, so far as it can be reached, that complete state- 

 ment of all the things on which a given process or result 

 depends, with its accompanying "mental model " of the 

 process, — that is commonly set forth as the aim of scien- 

 tific investigation. At the same time, by classifying all 

 the various sorts of preceding differences ("causes") 

 and the corresponding succeeding differences ("effects"), 

 we obtain general rules or laws. 



7. But the statements or mental models of given proc- 

 esses referred to above can never be really complete in 

 the sense of specifying everything that must have oc- 

 curred in order that the given result should appear. For 

 all differences between cases we find preceding differ- 

 ences, and so backward indefinitely. If this infinite re- 

 gress appears unsatisfactory, it is the constitution of 

 nature that is at fault. But any given investigation 

 seeks, for definite purposes, to trace the determining 

 differences back only to a certain stage. The investiga- 

 tor commonly finds that after a time the preceding dif- 

 ference of conditions passes into a field through which 

 he is not interested in tracing it; as when a biologist 

 finds a result to be due to a preceding difference in tem- 

 perature. 



8. Expressed accurately, the principle underlying all 

 this is: Every succeeding difference in perceptual condi- 

 tions is experimentally bound up with a preceding dif- 

 ference in perceptual conditions. This may be called 

 the postulate of experimental analysis. Cause or de- 

 terminer, and effect or thing determined, are both dif- 

 ferences between specifiable cases. In common usage 

 the term cause or determiner is loosely employed to ex- 

 press that which is added, or that which is subtracted, to 

 produce one case from the other; it may, therefore, as 

 well be the absence of something as the presence of 

 something. Thus, the determiner for blueness of eyes, 

 as compared with brownness of eyes, is, loosely, but con- 



