184 michelson: terms of relationships 



are linguistic phenomena. Several terms that correspond pho- 

 netically occur in so many Indo-European languages that it is 

 evident they belonged to the Indo-European parent language. 

 The social organization of the peoples speaking the historical 

 languages had nothing to do with it, as is shown by their diverse 

 social organization. Nor can it be said that such German com- 

 pounds as Schwieger-mutter, Schwieger-vater, Schwieger-tochter 

 have replaced the Old High German words by reason of a change 

 in social organization. Similarly, the fact that the Slavic word 

 for " father" (Old Bulgarian oibcb) has a different termination 

 than Greek arra, Gothic atta, is of linguistic significance, not 

 sociological. These data support my contenion referred to pre- 

 viously. 4 



Let us return once more to Rivers' position. On looking 

 over the tables given at the end of volume 1 of Rivers' History of 

 Melanesian Society, I am convinced that after all, he may have 

 just as much a linguistic and disseminative problem as a sociologi- 

 cal one. For example, the social organization of Mota, Banks 

 Islands, and Eddystone, Solomon Islands, is entirely different; 

 nevertheless the terms for father, elder brother, and younger 

 brother are evidently the same. Furthermore, the distribution 

 of slightly varying forms of the words tama, Una (father and 

 mother respectively) even across linguistic stocks, points in the 

 same direction. As I am not a specialist in Melanesian and 

 Polynesian linguistics, I regret that I can not thresh this out to 

 the end, and can only indicate a problem for others to solve. 



