228 michelson: American Indian languages 



little material at hand definitely to prove or disprove it. Such 

 as we have tends strongly to establish it. For example it is 

 patent that the post-positions of Wishram (Chinookan stock) 

 are due to the influence of Sahaptian, a distinct though contigu- 

 ous stock. Similarly, classifiers are common to the Salishan, 

 Wakashan, Chimmesyan, Koluschan, and Skittagetan stocks 

 which are at the same time contiguous. One may suspect, 

 indeed, that the first pair, and similarly the last pair, have differ- 

 entiated from a common ancestor. Yet at the very best we 

 would have a single striking morphological trait spread through- 

 out three stocks which otherwise have nothing in common in 

 either morphology or vocabulary, but only have resemblances 

 in sounds. It is too great a strain on the imagination to believe 

 that this is wholly the result of chance. If the accepted defini- 

 tion of stock is to remain, namely, a stock consist of one or more 

 languages all of whose sounds, morphology, syntax, and vocab- 

 ulary genetically have descended from a single ancestor, we 

 must admit at once that this trait which is held in common, is 

 due to borrowing. For the differences in vocabulary and 

 morphology are so enormous that it is inconceivable that the 

 present differences are solely due to later differentiation. The 

 vocabulary and morphology which cannot be explained at pres- 

 ent are just as important as the extremely small percentage 

 that can. There are a few other cases in which morphological 

 borrowings between stocks is plausible, but they are entirely 

 too few in number to warrant us at present in applying the prin- 

 ciple broadcast. So we have t© content ourselves in the mean- 

 while in pointing out structural resemblances between stocks, 

 such as between Esquimoan and Algonquian, which resemble 

 each other strikingly in their pronominal systems, and to a much 

 less extent in the building up of verbal stems, in the hope that 

 a careful and minute study of all the dialects of such stocks may 

 enable us definitely to affirm whether such traits are due to early 

 differentiation from a common remote ancestor or to compara- 

 tivoly recent borrowings. 



It must be admitted that recently there has been a decided 

 tendency among Americanists to consolidate such stocks as 



