378 Dr. Dorothy Wrinch and Dr. H. Jeffreys on Certain 



Another problem requiring explanation, related to the 

 last, is that of accepting the opinion of a class of 

 persons on matters, not of observation but of inference. 

 There is no longer any person whose individual know- 

 ledge covers all science as at present known, and most 

 scientific workers at some stage of their careers need 

 to seek the opinion of others on matters on which they 

 themselves are not expert. It is one of the most difficult 

 applications of the individual judgment to decide on the 

 probability to be attached in such a case. In a later paper 

 we hope to give an account of some principles that may 

 be adopted. 



The data we have to help us in the tasks of evaluating 

 probabilities in special cases^ and of choosing our primitive 

 postulates, are merely the ordinary current notions of 

 scientific validity. In other words, it is necessary to 

 interpret common-sense. Any result we offer must agree 

 with common-sense and with any results that can be 

 logically or mathematically deduced from common-sense. 

 It is necessary to make this statement, for otherwise it 

 might be maintained that our assumptions are unsound, 

 or that our results are untrue, or possibly that they 

 are not proved. We have to decide for ourselves 

 which results common-sense requires, then to consider 

 out of which assumptions these can best be obtained, 

 and then to see how far the assumptions themselves 

 fit in with common-sense. Thus there is no question of 

 "proving" a result, beyond showing that it is deducible 

 from some set of propositions that w^e consider plausible. 

 There is no question of making unjustified assumptions ; 

 for we make any assumption that we believe correct, and 

 then work out its consequences. If these results cover 

 anything in experience, the position is that a certain state- 

 ment A with vdiich common-sense agrees can be logically 

 obtained from an assumption B. We look into the latter, 

 which may or may not be in accordance with common- 

 sense. Then we say that A and B are equivalent. If 

 common-sense believes both, then we shall have shown 

 their logical connexion, If it believes A and the contra- 

 dictory of B, then we shall have shown that there is 

 something wrong in the position of the man in the street 

 in this domain, and that he must make his choice between 

 believing both A and B or neither. This decision may be 

 facilitated by working out other consequences of the two 

 alternatives. In either case, w T e are assisting common-sense 

 to build up a consistent and comprehensive attitude, and 

 that is the aim of science. 



