3OO
��INTERNATIONAL JOURNAL OF AMERICAN LINGUISTICS
��VOL. I
��TWO PHONETIC SHIFTS OCCURRING IN MANY ALGONQJJIAN LANGUAGES'
By TRUMAN MICHELSON
��I STATED in the " Journal of the Washington Academy of Sciences ", 4 : 404, that the inter- change of 6 before consonants, and aw before vowels, was universal in Fox. This it not quite accurate, for aw shifts to a, not 6, before cer- tain consonants. An examination has revealed that the same (or closely allied) shifts occur in many Algonquian languages. Specifically the languages in which I have thus far been able to establish that the shifts take place are Fox, Kickapoo, Cree (see below), Montagnais (see below), Shawnee, Ojibwa, Algonkin, Potawatomi, Peoria, and Delaware. Since these shifts are shared by so many Algonquian lan- guages, and since these languages are in subs- tantial agreement in the shifts, it is clear that these changes must be very ancient, and presu- mably in their beginnings go back to the Algonquian parent-language. I have derived my examples, for Fox, from Jones's Texts (references by page and line)and my unpublish- ed texts, and notes in a few cases (for the principle differences between Jones's and my phonetics see p. 54 of this Journal) ; for Kic- kapoo, Jones's Tales (references by page and line) ; for Cree, Lacombe's grammar and dic- tionary ; for Montagnais, Lemoine's grammar and dictionary ; for Shawnee, Gatschet's manu- scripts in the Bureau of American Ethnology and my early Shawnee notes ; for Ojibwa, Baraga's grammar and dictionary and Jones's Texts, Volume I (references by page and line); for Algonkin, Cuoq's grammar and dictionary;
��i . Printed by permission of the Secretary of the Smith- sonian Institution.
��for Potawatomi, photostat copies of Gailland's dictionary ; for Peoria, Gatschet's manuscripts in the Bureau of American Ethnology ; for Delaware, Zeisberger's grammar. These sources are of greatly varying quality, not to speak ot quantity. Moreover, it has not been possible for me to control their phonetics in all cases : hence it is that I cannot formulate definite laws covering all the languages concerned. Nor do I claim to give exhaustive rules for even those languages with which I am tolerably familiar. It would be an easy matter to obtain full data in the field ; in the office, it means the reading of hundreds of pages of texts, without being sure of completeness. What I wish to do is to establish the shifts and give such rules as I can, in the hope that others will assist in gathering materials which will enable complete laws to be formulated, and especially to find out whether these same shifts occur in other Algonquian languages. The fol- lowing table shows the provisional results :
��Fox
��Kickapoo ... aw Shawnee. . . au'
��> o before);, '/,./, | <'/,],. y,v>
> <7 before g, k [= g], t
> o before n, 'k, tc
> a before g
> u, o [= (ij before /
> a before g, k [= g}
��Cree aw, a'w[aw] } > d [ti] before /, k [=^
( tch
Montagnais. J > I [= ,/| before *
> 6 before ;;, * (Fox 'A Ojibwa a', rfw [dty] ^=,7,1^
> d before g, s, /, rf
��2. Terminally ; Gull Lake dialect -dt' according to Michelson.
�� �
