michelson: terms of relationships 183 



might safely be inferred from terms of relationship ; whereas such 

 marriages are fundamentally repugnant to the Piegans, and their 

 terms of relationship are new, not old. This is, of course, a con- 

 crete example of one of Sapir's points. Recently Swanton" quite 

 similarly brings forward data from Creek and Chickasaw which 

 prove the unsoundness of such inferences. 



The above has been cited to show that American ethnologists 

 generally have taken a united stand against Rivers' one-sided 

 attitude. I reopen the case because none of us has given abso- 

 lute proof that kinship terms are borrowed. In another place' 2 

 I have tried to prove that Cree has borrowed certain terms 

 from Ojibwa; and somewhat similarly that Peoria has been influ- 

 enced by Sauk, etc. I think the reasoning given there is sound, 

 and as near absolute proof as we can expect to have in the case 

 of prehistoric linguistic borrowing ; yet it is not absolute in so far 

 as we have no Cree nor Ojibwa records transmitted to us his- 

 torically, extending over several centuries, showing absolutely 

 that such borrowing took place. The same is true regarding 

 Peoria. English is a good language to draw on for illustrative 

 material to prove such a point, for it has been transmitted his- 

 torically for several centuries. Every Indo-European philolo- 

 gist knows that sister is Scandinavian in origin, and that cousin, 

 niece, nephew, aunt, and uncle are Romance. Furthermore all 

 our terms of -in-law are directly or indirectly due to the latter's 

 influence; grand-father and grand-mother are Romance in the first 

 member of their compounds. Similarly Albanian frat " brother" 

 is Romance in origin and is not a native word, as is shown by 

 the phonetics. In the same way Hungarian bar at " brother" is 

 borrowed from Slavic, a case of borrowing across linguistic stocks. 

 [Ojibwa nimpdpa, nimdrnd (Fort William) are other illustrations 

 of kinship terms borrowed across linguistic stocks.] These 

 facts, long known, are brought forward simply because they seem 

 to have escaped the attention of ethnologists. Delbriick 13 was 

 well aware of the fact that Indo-European terms of relationship 



11 Amer. Anthrop. n. ser., 18: 463. 1916 [1917]. 



12 Proc. Nat. Acad. Sci., 2: 297. 1916. 



13 Die Verwandtschaftsnamen. Leipzig. 1889. Also in Abh. phil.-hist. 

 Klasse sachs. Ges. Wiss., 11: 379. 1889. 



