EXPLANATORY SCIENCE 211 



late set constitutes a ground for severing its connection with 

 empirical data. So far as descriptive science is concerned, 

 the answer here is definite; intension is never regulative over 

 extension, hence a proposition is true or false according as 

 it is descriptively true or false, regardless of its deducibility 

 from a postulate system. But the tendency in rational science 

 is to make intension regulative over extension, and hence to 

 say that a proposition which is capable of being deduced 

 from a postulate set is true no matter what its descriptive 

 reference may be. This is essentially equivalent to con- 

 sidering the relation between the Y and the X proposition 

 as analytic, and as depending therefore upon linguistic 

 factors rather than upon data. If the postulates do imply 

 the propositions in question, it is part of their meaning that 

 they do so, hence nothing empirical could ever constitute a 

 refutation of this fact; the postulates imply the propositions 

 in exactly the same sense that the meaning of the word 

 "triangle' 1 contains the meaning of the phrase "three- 

 sided figure." The only alternative is to consider the relation 

 between the Y and the X propositions as synthetic, and as 

 depending therefore upon empirical data rather than upon 

 linguistic factors. If the postulates imply the propositions 

 in question it is only an empirical fact that they do so, hence 

 nothing in the character of language or logic could legislate 

 this away; the postulates imply the propositions in exactly 

 the same sense that the meaning of the word "body' is 

 connected with the meaning of the phrase "possessing 

 weight." According to this latter conception a descriptive 

 proposition is always true or false in an empirical sense, and 

 nothing in the character of postulate sets could make a 

 descriptive proposition which is empirically true false, or one 

 which is empirically false true. 



This results in the development of a peculiar sensitivity in 

 explanatory sciences. Being highly integrated, they suffer 

 extreme shocks when new information necessitates a change. 

 For a change in any one part of the system requires modifica- 

 tions in all other parts. Mathematics experienced such a 



