LAW, CAUSE 363 



the law is not assured, while the latter forbid certainty be- 

 cause the specific features of the individual are not known. 

 Universal laws are distributively applicable to individual 

 cases, while statistical frequencies are collectively applicable. 1 

 When one states that all Irishmen are probably quick- 

 tempered he means that there is a certain probability that 

 each Irishman is quick-tempered; but when one states that 

 the frequency of quick-tempered Irishmen is 86 in 100 he 

 is attributing a certain number, a ratio, to a group as a 

 whole, and this states nothing whatsoever with reference to 

 each Irishman, except that he is a member of a group possess- 

 ing a certain collective property. Statistical frequencies 

 state distributions, and distributions are applicable, basi- 

 cally, only to collections. 



The summary of these various operational approaches to 

 the problem of scientific law will be made in connection 

 with the analysis of the scientific content of law. The 

 purpose of this reference to the work of Keynes and Nicod, 

 and to the problem of statistical correlations, has been to 

 show that the problem of induction is no longer concerned 

 with the establishment of necessary laws. The operational 

 techniques are directed to the ascertainment of probability, 

 either in the law itself or in the behavior of an individual 

 case. Probability depends on evidence, and the ascertain- 

 ment of evidence depends on the establishment of laws 

 connecting evidence of certain kinds either with laws of 

 certain degrees of probability or with predictions of certain 

 degrees of probability concerning individual cases. The 

 problem is one of great complexity, and is not likely to be 

 solved in the immediate future. The present discussion has 

 attempted merely to locate the problem in its proper context. 



Reference must be made, however, to an alternative 

 operational technique by which universality in a scientific 

 law may be achieved. Though this method is employed in 

 science with great frequency, there is some question as to 

 whether it should be called an operational technique. It 



1 See Chapter XIII, p. 262. 



