THE DISCOVERY OF LAWS 61 



intended when it is asked how we have such knowledge 

 from experience of the past. 



And because such a question is meaningless the answers 

 given to it are meaningless also. They always consist 

 in some attempt to prove from very abstract and obscure 

 premisses a doctrine called by the high-sounding title of 

 Uniformity of Nature ; it is argued that, for some reason 

 to be found in transcendental philosophy, nature must 

 be such that what is true of her in one part, in one region 

 of space or at one period of time, must be true of her in 

 any other part. But the value of such a doctrine depends 

 entirely on the meaning attributed to " nature." If the 

 world means merely the non-human, external world of 

 common sense (as in Chapter II), then the doctrine is 

 simply untrue. Nature, in this sense, is not uniform ; 

 there are events which happen once and never happen 

 again ; and it is precisely because there are such events 

 that we distinguish between past and future. If it were 

 v true that " history repeats itself," there would be 

 no history ; history is the record of events which have 

 not repeated themselves and the proverb like almost 

 all proverbs merely represents an attempt to obtain, 

 by an epigrammatic form, credence for an assertion which 

 nobody would otherwise believe. It is true that many 

 lu'ch do not so repeat themselves, and perhaps 

 >ortant of these events, are characteristically 

 human and do not, therefore, form part of common-sense 

 ;e are enough non-recurrent events, 

 which have nothing to do with man, to distinguish 

 1 future and thus to controvert the 

 all nature is uniform in all its p, 



If. n by " nature " in this 



connexion the carefully scnitini/.rd nature of science, 

 then the doctrine merely states thai nature is nai 



>rld of science is charac- 

 d by an by the 



