268
��INTERNATIONAL JOURNAL OF AMERICAN LINGUISTICS
��VOL. I
��in the nouns is the word for CRANE, dni in Tlingit, and dd in Kato, and found fairly frequently in other Athapascan dialects. A rela- tion between Tlingit n and Athapascan s or i is all that is lacking. Since the word is almost an isolated case of phonetic correspondence of nouns of identical meaning, as will appear below, it seems more logical to consider the word one that Tlingit has borrowed from a neighboring Athapascan dialect.
Almost equally exact is the agreement of Tlingit (aw FEATHER, and fa which, in Atha- pascan frequently translated FEATHER, is eve- rywhere restricted to the larger stiff feathers of the wings and tail. The more exact rendering, then, would be PLUME. I am told there is a corresponding restriction in meaning in Tlingit.
In only one instance is there an indication of several nouns with the same phonetic cor- respondences. Tlingit Saq' and Athapascan ts'in mean BONE; Tlingit sax means HAT and Chi- pewyan tc'a DANCE-HAT ; and Tlingit six' and Jicarilla ts'ai mean DISH. In these cases there is agreement only between the initial conso- nants, the other sounds varying. A dance-hat is probably something quite different from simply a hat. The case is too weak to be con- vincing, and, unsupported as it is, carries very little weight. Two other fair agreements appear in the list, - - Tlingit can OLD PERSON, and Beaver con OLD AGE ; Tlingit ci SONG, and Chi- pewyan an SONG.
Out of over three hundred monosyllabic nouns gathered by Professor Boas, most of which have clear-cut meanings, one hundred and fifteen have been matched with Athapascan words of identical or closely-related meanings. Some of the unmatched Tlingit nouns have meanings too general or too specialized to be matched satisfactorily with Athapascan forms. In many instances the Athapascan nouns of corresponding meanings are dissyllabic and have descriptive meanings, and are therefore
��not comparable with the monosyllabic, non- descriptive nouns of Tlingit. This tendency to replace the simple nouns with longer descrip- tive terms is very pronounced in Athapascan. The unmatched Tlingit nouns, then, do not weaken the case for genetic relation. However, five fairly satisfactory agreements out of one hundred and fifteen which have been matched in meaning do not present an impressive pro- portion.
In addition to these, Sapir lists the follow- ing :
��ATHAPASCAN
��TLINGIT
��~t e > Xe grease ex grease
-;t> tooth HX tooth
-k.'a arrow g!a point
-onaye elder brother bitnx man's older brother
t'e^ night t'a't night
/o' fish tUn'k! cohoes
mis cheek wye cheek
no place of retreat, nn fort island
Of these xe GRESSE is unfamiliar to me in Athapascan ; and Tlingit c'.y (Boas 'ex) is given the meaning FISH-OIL. The Athapascan word for TOOTH is -70, or -wo ; Boas gives for Tlingit 'iix. Sapir gives q!a as meaning POINT, and compares it with Athapascan k!a ARROW. Swan- ton's texts and Boas', wordlist give the meaning POINT OF LAND. The connection in meaning would appear far-fetched. The Athapascan word k!a does not refer to the point ot the ar- row, but to the shaft, since the separable pointed end, either of stone or wood, has a different name. Boas has i'uk"' COHOE-SALMON, which leaves only the vowel n to carry the phonetic similarity; for glottalixed /' and k' are very distinct from / and k without glottalixa- tion. In Athapascan, no does not mean a PLACE OF RETREAT, it means an ISLAND.
Dr. Sapir has assumed that the primitive Na-
�� �
