142 INTRODUCTION TO PHILOSOPHY OF SCIENCE 



are merely occasional or frequent associations, and the only 

 laws possible are statistical. Hence the concepts interrelated 

 by such laws are not as determinately related to one an- 

 other as they would be in the case of universal laws. To 

 know that an event is occasionally or frequently associated 

 with another does not tell one as much about the event 

 as to know with what it is universally or inevitably associ- 

 ated. As a consequence, in descriptive science one can gain 

 no important information about a symbol by exploring its 

 intensional features, i.e., its relations to other symbols. 

 One determines what a symbol means not by definition 

 (intension) but by illustration (extension). In other words, 

 one determines the properties of an event by pointing to 

 the event, rather than through the indirect route of finding 

 some other event of known properties with which the event 

 in question is universally associated. 



This is equivalent to asserting that in descriptive science 

 extension is taken as regulative over intension. Every 

 symbol has its meaning determined by that to which it 

 refers empirically, not by other symbols to which it has 

 relations. In the language of the traditional logic, all prop- 

 ositions of descriptive science are synthetic propositions, not 

 analytic. A synthetic proposition is one whose predicate 

 contains something not contained in the subject; an analytic 

 proposition is one whose predicate can be obtained from 

 the subject by mere analysis of its meaning. The proposi- 

 tion, "All bodies have weight," would presumably be syn- 

 thetic, since its truth cannot be determined by analysis 

 of the subject term, i.e., by "body' one does not mean 

 "something possessing weight." On the other hand the 

 proposition, "All triangles are three-sided figures," would be 

 analytic, since its truth can be determined by a mere anal- 

 ysis of its subject term, i.e., by "triangle" one does mean 

 "a figure which has three sides." For a descriptive science 

 this distinction does not exist; all propositions are synthetic. 

 The truth of every proposition must be determined by an 

 examination of the cases to which it refers. No predicate 



