334 MICHELSON: ALGONQUIAN phonetic SHIIfTS 



keneno'tawi "dost thou understand me," awita ncnoHo'fiAgu's"" 

 "he might not understand us (incl.)," keneno'ton^^' "I under- 

 stand thee," keneno'tdg^'^'^' "he understands thee;" keneno'tdtlpen'"''' 

 "we (incl.) understand each other;" d'mdnd"kawuHc'' "then 

 many attacked him" [really an independent passive in 

 formation], d'mdnd'^kdguHc'^ "then he was attacked by many;" 

 keteminawdf" "if he takes pity on him," keteminawi "take pity 

 on him," ketemino' nagd^^^ "I take pity on you," kctemind'k'^ 

 "he takes pity on thee," nenl'cwiketemind'gdpen""' "we (excl.) 

 are both pitied." From the above it follows that -aw- appears 

 as -0- before -n-, -t-, [in final syllables], -He-, -'k-; as -d- before 

 -g-, -t- [normally], -'s-. I have evidence to show that -aw- also 

 has the same shifts before the same consonants (or their phonetic 

 correspondents) in Cree, Sauk, Kickapoo, Shawnee, Ojibwa, 

 Potawatomi, Ottawa, Algonkin, Peoria, Delaware, Penobscot, 

 and Passamoquoddy. Hence it is quite certain that these shifts 

 took place in the Algonquian parent language. I call attention to 

 the fact that this is the first time in the history of American lin- 

 guistics that phonetic shifts in the parent language of any Amer- 

 ican linguistic stock have been pointed out. 



I have previously shown that n (or its phonetic equivalent) 

 changes to c (or its phonetic equivalent) before i which is the ini- 

 tial sound of a new morphological unit in Sauk, Fox, Kickapoo, 

 Shawnee, Menomini, Ojibwa, Algonkin, and Peoria. The change 

 of 5 to c under the same conditions occurs in Sauk, Fox, Kickapoo, 

 and Ojibwa. The interchange of a and d is found in Sauk, Fox, 

 Kickapoo, Shawnee, and Peoria. Further researches may show 

 that these shifts occur in other Algonquian languages also, and so 

 perhaps may Ukewise be referred to the Algonquian parent lan- 

 guage. The contraction of -wa- to -o- occurs in a number of 

 Algonquian languages, but the evidence of Cree is unfavorable 

 to the theory that this particular shift is to be ascribed to the 

 Algonquian parent language. As the shift occurs in languages 

 which are geographically all contiguous it is possible the shift 

 has simply spread, so as to be practically pan- Algonquian rather 

 than proto- Algonquian.^ 



3 See this Journal 4: 402. 1914; Amer. Authropol. n. ser. 15: 470. 1913; Inter- 

 nat. Journ, Amer. Ling, i: 50. 1917. 



