472 THE TLINGIT INDIANS [eth. ann. 26 



RELATIONSHIP BETWEEN THE TLINGIT AND HAIDA 



LANGUAGES 



The grammatical structure of the Tlingit and Haida languages will 

 be given elsewhere." The possibility that thej' are related in some 

 waj' was suggested first by Professor Boas, whose reasons therefor are 

 given in the Memoirs of the International Congress of Anthropologv* 

 (p. 3-I-0-343). With the assistance of material recently collected, the 

 present writer is now able to undertake a more exhaustive comparison 

 along certain lines. 



1. Phonetics. In this respect the two languages agree closely with 

 each other and with other languages in the same region. In l)oth, 

 however, labials are used less generally, and those which occur seem 

 usually to be onomatopoetic. In Haida m is found with some fre- 

 quency, j'^ rarely except when it is evidently onomatopoetic, and li has 

 not been found more than live times altogether. In Tlingit /// appears 

 in a very few onomatopoetic expressions, but p and h do not occur in 

 words of native origin. Tlingit is furthermore distinguished from 

 Haida by a great expansion in the use of sibilants, in the absence of / 

 which is replaced by «, and in the absence of n. The Tlingit language 

 tends to shorten its vowels, resembling in this respect the Masset dia- 

 lect of Haida, and differing from that spoken at Skidegate and the 

 other southern towns. Obscure u and o sounds are used continually. 

 Following is a comparative list of the consonants in the two languages: 



Harmonic sound changes are few in Haida and still fmver in Tlingit. 

 The Ilaitia transposition of / and I is naturally wanting from Tlingit, 

 where / does not exist, and the interchanges between </ and x or ,/■ also 

 seem to be wanting. There is, however, an interchange of e and dj 

 or tc similar to the change in Haida from .y to dj and vice versa. On 

 the other hand, m and o sounds in Tlingit tend to change a following- 

 vowel into a similar sound, a phenomenon noticed elsewhei'e in North 

 America but not observable in Haida. 



a. Roots and stems. Tlingit words arc nearly always resolvable 

 into single syllables or single phonetics, each with a separate signifi- 

 cance. Haida constructions are longer and more ponderous, but so 

 many words may be resolved into grammatic syllables that all Haida 

 stems may be regarded as having been primitively monosyllabic. In 



"In the Handbook of American Indian Languages, not yet published. ("Chicago, 1894. " Very rare. 



