Abbildungen der Seite
PDF
EPUB

apparently show only survivals, but it is quite possible that if we had reasonably full data for Washo and Seri, we should have to number these among the archaic languages as well. The following table of the distribution of the adjectival and verbal class-prefixes is necessarily even more fragmentary than the corresponding table for the absolutive noun affixes. It is given rather as the statement of a fundamental and far-reaching problem needing investigation than as a satisfactory formulation of the facts.

The lack of entries for Karok, Shasta-Achomawi, and Chontal and the paucity of entries for Yuman and Chimariko prove little or nothing because for none of these languages except the last have I anything available but the most fragmentary material. However, it is clear that the characteristic development of intransitive and transitive class-prefixes which we have in Subtiaba and Salinan, and perhaps in Washo, is absent in Chimariko and Yana. It looks, then, as though it was the southern, rather than the northern, Hokan languages that best preserved the archaic features of the Hokan system of verbal class-prefixes. This accords with the greater conservatism of the southern languages in the use of absolutive nominal prefixes.

If we ruthlessly eliminate at this preliminary stage of linguistic inquiry all evidence that is weak or ambiguous, we can still rest a reasonable case for an old Hokan system of "classifying" or generic prefixes in the noun, the adjective, and the verb on the following condensed table:

[blocks in formation]

CONCLUSION: FURTHER VISTAS

On the basis of the evidence reviewed in the preceding pages it seems difficult to avoid the conclusion that Subtiaba is a Hokan (or "Hokan-Coahuiltecan") language and that, as such, it is genetically related to widely different languages spoken far to the north (e. g. Shasta, in northern California) and to the northeast (e. g. Tonkawa and Karankawa, on the Texas coast). This evidence is lexical, phonological, and, above all, morphological. The typical form of stem in Subtiaba is of a recognized Hokan pattern; the alternation of forms with and without initial vowel is characteristically Hokan. The old system of Hokan nominal, adjectival, and verbal consonantal prefixes, defining the most fundamental classificatory notions, had begun to emerge before the present study of Subtiaba was undertaken, but the new evidence brought by this remote language of Nicaragua and southern Mexico has clarified the picture appreciably. Much, of course, remains obscure and in many directions there is everything to be done.

But can we stop with Subtiaba? In a previous communication20 I ventured to suggest that one of the major groups of American Indian languages is a large group extending east and west from the Atlantic to the Pacific. The name "Hokan-Siouan" was suggested for these languages. A tentative genetic scheme might be thus represented:

A. Hokan-Coahuiltecan

I. Hokan proper

1. Northern Hokan

a. Karok; Chimariko; Shasta-Achomawi

b. Yana

c. Pomo

2. Washo

3. Esselen; Yuman.

4. Salinan; Chumash; Seri

5. Chontal

II. Subtiaba (and Tlappanec)

III. Coahuiltecan: Tonkawa; Coahuilteco-Cotoname-Comecrudo;
Karankawa

20 A Bird's-eye View of American Languages north of Mexico, Science, n. s., Vol. LIV, 1921. p. 408.

[blocks in formation]

Such a scheme must not be taken too literally. It is offered merely as a first step towards defining the issue, and it goes without saying that the status of several of these languages may have to be entirely restated. Thus, Yuki is listed for the time being as coordinate with Hokan-Coahuiltecan, but it is not unlikely that it is really a specialized Hokan language which has undergone a great deal of phonetic decay and has been rather seriously influenced by contact with neighboring Penutian languages (Miwok and Wintun).

The evidence for this "Hokan-Siouan" construction is naturally morphological rather than lexical, though the lexical bonds that unite Natchez-Muskogian and Hokan, for instance, are by no means negligible. This evidence will be given in due time. It is of a general rather than specific nature, though specific elements constantly enter into the argument, and can hardly receive its due weight unless one contrasts the underlying "Hokan-Siouan" features with the markedly different structures that we encounter in Eskimo-Aleut, in Nadene, in Algonkin-Wakashan, and in Penutian.21 There is now reason to believe that some of the more archaic elements and classes of elements that are found in HokanCoahuiltecan also exist as survivals in languages spoken far to the east. Some very suggestive evidence has recently come to hand, for instance, which seems to indicate that Natchez and the Muskogian languages originally possessed a system of consonantal prefixes analogous to the old Hokan system that we have discussed. Thus, Choctaw la"sa "scar": mi"sa "to be scarred" is curiously reminiscent of such alternations as Subtiaba d-aca- “grass": m-a ca "to be green" and suggests an old nominal prefix - (cf.

21 The grammatical peculiarities of Uto-Aztekan are not so clearly differentiated from those of "Hokan-Siouan."

Chontal?) and an adjectival or intransitive m- (as in Yana, Pomo, and Subtiaba); for the latter element the evidence is already respectable. A surprising number of other consonantal alternations have been found in Muskogian and they present an important historical problem.

The example of Subtiaba is sufficient warning of the impossibility of drawing a preconceived boundary to the south. There is no reason whatever to believe that the "Hokan-Siouan" group as already defined will remain without further adjuncts in Mexico and Central America or perhaps even beyond. The addition of the Mayan languages seems rather more than less likely. To mention such possibilities is to make it clear that the real problems of American Indian linguistics have hardly been stated, let alone studied.

VICTORIA MEMORIAL MUSEUM,

OTTAWA, CANADA.

THE VILLAGE SITES IN TOLOWA AND NEIGHBORING AREAS IN NORTHWESTERN CALIFORNIA

IN

By T. T. WATERMAN

N THE year 1909 Dr. Kroeber dispatched me to Northern California to look into the native life existing among the Yurok. The fruits of my brief labors there have appeared in part in print, though one paper, "Yurok Culture," is still in storage in the Museum of the American Indian, Heye Foundation, New York. It contains the only detailed account of the northwestern California house. In the intervals of my business with the Yurok I used to look across the bay to distant Point St. George, dim, romantic and far away, and my soul took fire to wander thither and work with the Tolowa. Mr. Heye, to whose institution I committed my fortunes in 1921, actually sent me to the Tolowa territory to collect specimens for him, and accordingly during a month of that summer I lived in Crescent City, carrying out, at my own expense, some investigations on local ethnology. So little has been said about the Tolowa, that an essay on their habitat may interest the readers of the Anthropologist. They are an unusually interesting group, to me, very different in some respects from the Yurok.

Everyone, I think, is familiar with the fact that a somewhat peculiar way of living characterizes the northwest California tribes. Native life changes quite rapidly as one goes northward into Oregon. The Yurok, the Hupa, and the Karok have a somewhat highly specialized "house complex." For example, among the Yurok every house has a name. The Tolowa to the north of them have apparently no names for houses. In fact, when we leave the Yurok we have to pass almost to Alaska before we again find the custom of naming dwellings. There is, in fact, quite a sharp line both in this and in other matters between the Yurok and the Tolowa, although they live together on one stretch of coast. The Tolowa maintained close contact by trail over the mountains with the Karok, more so than they did with the Yurok who lived within view of them on the beach. "Money" made of dentalium

« ZurückWeiter »