The Method ^ 



I have used this principle to test the validity 

 of distinctions as they are made by commonly 

 used words in each science. It is too often 

 assumed that distinctions correspond to words, 

 and that when an array of words has been 

 defined, they will be justified by facts in the 

 real world. But language has had so long 

 and so loose a development that the mere 

 presence of words is no evidence of the real- 

 ity of the distinctions they represent. Usage 

 also gives so broad a meaning to many words 

 that they lap over each other in every con- 

 ceivable way. They must be pruned to the 

 real differences as they exist in the parallelism 

 of expression in two related fields. The word 

 " emotion," for example, has a great variety of 

 meanings. It can be restricted to one only 

 when we define it by its physical parallel; for 

 since thought and matter are parallel, emotion 

 has its exact physical expression. We may 

 be sure the word stands for something when 

 it is defined in terms of this physical parallel. 

 We can be sure of nothing if usage and cus- 

 tom alone are relied on. 



This method of defining may err in unduly 



