PROBLEMS IN COMPARATIVE PHILOLOGY 59 



due to simultaneous changes which rise independently in the members 

 inhabiting a certain area. Its collective character is, if I understand 

 Wundt aright, due to the fact that the causes for its existence are 

 uniformly present in many members, who, therefore, at about the 

 same time independently hit upon the same change. According to 

 this view, every change starts and spreads as a primary change. The 

 other theory makes a sharp distinction between primary and second- 

 ary changes and explains the spread of a change as due to the adop- 

 tion of the innovation by the rest of the speech-community. Having 

 been originated elsewhere, the change is afterwards accepted and 

 mostly unconsciously imitated. In two ways this latter view 

 (which is shared, among the latest writers, by Delbriick, cf. his Ein- 

 leitung in das Studium der indogermanischen Sprachen, 4th edition, 

 p. 149) seems to me to have an advantage over the other. In the 

 first place, it bridges the gulf which separates phonetic changes from 

 those of a semantic and syntactical kind. For it seems hard to 

 maintain that the change in a syntactical construction or in the 

 meaning of a word owes its universality to a simultaneous and inde- 

 pendent primary change in all the members of a speech-community. 

 By adopting the theory of imitative spread, all linguistic changes 

 (formal as well as semantic) may be viewed as one homogeneous 

 whole. In the second place, the latter view seems to bring linguistic 

 changes into line with the other social changes, such as modifications 

 in institutions, beliefs, and customs. For is it not an essential charac- 

 teristic of a social group that its members are not cooperative in the 

 sense that each member actively participates in the production of 

 every single element which goes to make up either language, or 

 belief, or customs? Distinguishing thus between primary and 

 secondary changes and between the origin of a change and its spread, 

 it behoves us to examine carefully into the causes which make the 

 members of a social unit, either consciously or unconsciously, willing 

 to accept the innovation. What is it that determines acceptance 

 or rejection of a particular change? What limits one change to a 

 small area, while it extends the area of another? Before a final 

 decision can be reached in favor of the second theory of imitative 

 spread it will be necessary to follow out in minute detail the mechan- 

 ism of this process in a number of concrete instances; in other words 

 to fill out the picture of which Tarde ( Les Lois de I' Imitation) sketched 

 the bare outlines. If his assumptions prove true, then we should have 

 here a uniformity resting upon other causes than the physical uni- 

 formity that appears in the objects with which the natural sciences 

 deal. It would enable us to establish a second group of uniform 

 phenomena which is psycho-physical in its character and rests upon 

 the basis of social suggestion. The uniformities in speech, belief, and 

 institutions would belong to this second group. 



