868 



SCIENCE 



[N. S. Vol. XXXVIII. No. 990 



tifie knowledge were practically inconsider- 

 able. His great merit lay in his making 

 men see that science was in fundamental 

 need of a new method. The method he sug- 

 gested was not adopted ; but his analysis of 

 the need was the signal for the search 

 which has ended in modern science. 



I need not take you further through the 

 long history. It is sufficient to my purpose 

 to point oixt that primitive man first devel- 

 oped by experience a way of his own for 

 observing and fixing in mind external phe- 

 nomena, that the Greeks seized upon their 

 own and their predecessors' observations 

 and sublimed experience into theory, that 

 Telesio and Bacon and others taught man- 

 kind the insuiificiency of Greek methods and 

 the need of new ones, and that modern 

 science came into being and fulness of 

 stature through generations of workers who 

 sought to put, and succeeded in putting, 

 the new ideas into the form of effective tools 

 of advancement. 



From this brief historical account it is 

 seen that the method of experimental sci- 

 ence has itself grown through experiment. 

 The style of argument employed by Plato, 

 for instance, has been entirely superseded 

 by another and better. Man had to learn 

 by the experience of failure how to ascer- 

 tain the true relations of phenomena. In 

 other words, there was no " preestablished 

 harmony" between the mind and the phe- 

 nomena it had to interpret of such char- 

 acter as to lead the former to a ready ex- 

 planation of the latter. 



Our progres in this respect has been over 

 a hard and long and rough road. We go 

 a very short distance, relatively, into our 

 past to find the time when methods were 

 uniformly employed in science which are 

 now known to be quite untrustworthy. 

 What is the bearing of this fact on our con- 

 fidence in the conclusions of science? In 

 order to answer this question properly we 



shall have to analyze briefly the general 

 nature of scientific investigation as at 

 present practised. 



In the first place, scientific demonstra- 

 tion starts from data which involve the 

 ever-present inexactness which is due to 

 experimental error. In the nature of things 

 it is impossible that the argumentation 

 should ever have an exact basis to rest 

 upon; and consequently all conclusions 

 must again be tested by a direct appeal to 

 phenomena. In another important respect 

 also the method is essentially different from 

 that employed in mathematics. Here intui- 

 tion is a fundamental guide in argument 

 as well as in discovery; and a "proof" 

 whose leading elements are grounded in 

 intuition is accepted with a confidence at 

 least equal to that which is accorded to one 

 characterized by mathematical precision 

 and rigor. 



One result. of this inexact basis and espe- 

 cially of this loose method of argumenta- 

 tion is that the conclusions reached often 

 are primarily of the nature of inference 

 from examples. They have little or none 

 of the compelling property which attaches 

 to mathematical conclusions. 



In other words, scientific (as opposed to 

 mathematical) truth is not necessary truth. 

 It is in the nature of things that the experi- 

 mental scientist can not give us absolute 

 truth. This is no criticism of his work; it 

 is not his province to give us absolute truth 

 — even if such a thing were supposed to 

 exist. 



What then is the purpose of the experi- 

 mental scientist! His province is to enable 

 us to get around among the phenomena of 

 the external world, to predict what will 

 happen under a given set of circumstances. 

 He will accomplish this end by studying 

 the relations among phenomena. He does 

 not need to know their ultimate explana- 

 tion ; it is sufficient if he can find the essen- 



