354 INTRODUCTION TO PHILOSOPHY OF SCIENCE 



require, at least according to the prevailing view, the notion 

 of necessity, and the problem arises as to what can be the 

 empirical foundation for this notion. Is necessity a property 

 of scientific law merely because, as Hume maintained, the 

 correlation occurs repeatedly? If so, it is hard to see how 

 necessity can be found in a collection of cases if there is 

 not the slightest hint of necessity in the individual case. 

 Or is necessity a property of scientific law because, as 

 Whitehead maintains, each individual case exhibits neces- 

 sity? If so, it is hard to see how one can ever be mistaken 

 as to a causal connection. The view which will be taken on 

 this issue can be more clearly indicated by turning to the 

 third aspect of law, viz., that which is involved in the notion 

 of repetition. 



(3) It seems unquestionable that a single correlation or 

 association of events does constitute a fact about each of 

 the events. An event is not merely what it is "in itself 

 but what it is in its associations with all other events. 

 Hence to know with what events a given event is once 

 associated, occasionally associated, frequently associated, uni- 

 versally associated, and even never associated, contributes to 

 the understanding of the nature of the given event. Further- 

 more, it seems hard to deny that the information contributed 

 in each of these cases is different in degree. One may call 

 this feature of an association its empirical necessity. An as- 

 sociation which occurs only once has a low degree of empirical 

 necessity; associations which occur occasionally, frequently, 

 and universally have continuously increasing degrees of 

 empirical necessity. 



At the empirical level little or no attention is paid to the 

 degree of empirical necessity involved in any association. 

 Unless a specific problem is involved cases are not counted, 

 and exceptions are not noted. Presumably no attention is 

 paid to any association unless it occurs at least occasion- 

 ally. This accounts for the fact that a law is customarily 

 described as a repeated association of events. The opera- 

 tional transformation of the notion of empirical law involves, 



