EXPLANATORY SCIENCE 199 



unless it refers directly to some other event, then one may 

 employ the word "description" in an equally generous sense 

 and say that Y describes y even though y is a hidden event. 

 By virtue of this use it becomes possible to say that x is 

 explained by y, i.e., explanation is a relation between events. 

 It seems best to employ the words in the strictest sense be- 

 cause this seems more precise and because it is a use which 

 satisfies both the positivist and the realist, i.e., description 

 is at least a relation of a symbol to a clearly given event, 

 whatever else it may be, and explanation is at least a relation 

 between symbols, whatever else it may be. According to this 

 interpretation the essence of the explanatory relation is to be 

 found in the implicative relation between the explanatory 

 symbol and the descriptive symbol. Description is a matter 

 of extension, while explanation is essentially a matter of 

 intension. Descriptive or empirical science consequently 

 emphasizes the relations of symbols to events, while ex- 

 planatory or rational science emphasizes the relations of 

 symbols to one another. Descriptive science is primarily 

 concerned with reference outside of itself, explanatory 

 science is primarily concerned with internal consistency. 

 This is the foundation for the distinction, to which reference 

 will be made later in the chapter, that empirical sciences are 

 real or existential while rational sciences are ideal or non- 

 existential. 



The character of these relations may be made clear by an 

 illustration. Suppose that X is the descriptive law, "Bodies 

 expand when heated." A strictly descriptive law would 

 never have this universal form, since all bodies could never 

 be observed, but this fact may be neglected for the present. 

 Then x symbolizes the actual fact of a large number of 

 associations between cases of the application of heat and 

 cases of ensuing expansion; the law may be said to describe 

 this fact. Suppose, now, that Y is a group of propositions 

 about molecules, viz., the group which would ordinarily be 

 said to reveal the defining characteristic of molecules. Since 

 this group of propositions implies the descriptive law, the 



