THE NATURE OF SYMBOLS 71 



to insert into any symbolic scheme elements which are not 

 known to have any direct representative value; the symbolic 

 system adds to nature, and hence ceases to be strictly ade- 

 quate. Thus the realm of symbols is more extensive than the 

 realm of events. 



This control of symbolic schemes by human factors, 

 apparently working at odds to the theoretical ideal of 

 knowledge, has seemed to many to argue for the necessity of 

 recognizing an essentially pragmatic factor in knowledge. 

 Knowledge, it is claimed, must be adequate as a representa- 

 tive of events, but it must also be adequate to the knowing 

 demands of the human organism. Often, as in the cases 

 mentioned above, these two types of control are in conflict, 

 and one is obliged to make a choice between them. Conse- 

 quently it is one of the tasks of the philosophy of science to 

 make an analysis of all scientific schemes with a view to 

 showing which of their elements are traceable to that which 

 is to be known and which are traceable to the knowing 

 operations. The results of some of these analyses will be 

 discussed in Chapter VIII. 



If one grants, then, that an adequate symbolic scheme is 

 isomorphic with reference to the realm of events, and that 

 every scientific system is more or less adequate depending 

 on the extent to which the knowing operations transform 

 and create, the problem immediately arises as to whether 

 there may be degrees of adequacy within the scope of 

 isomorphism. Are there some types of symbol which are 

 more adequate than others? Apparently, one-to-one correla- 

 tion implies only a highly abstract identity, hence there 

 need be no specific similarity between the elements and 

 relations of the one system and those of the other. What 

 seems to be required is merely that if there are different 

 elements and different relations in the one system, there must 

 be the same differences in the other. This seems to be about 

 all that one can say with regard to the general referential 

 character of symbols. But what may be said as to the 

 adequacy of specific kinds of symbols? 



