THE DIALECTIC DIVISIONS OF THE MOQUELUMNAN FAMILY IN RELATION TO THE INTERNAL DIFFERENTIATION OF THE OTHER LIN

GUISTIC FAMILIES OF CALIFORNIA

By A. L. KROEBER

As a linguistic area California is noted for the number of its distinct stocks or families. With a few doubtful exceptions these families stand practically as they were originally determined, and the most recent investigations as yet give but little indication that their number, which is above twenty, will be materially reduced through the discovery of identities resulting from deeper study.

As to the number and nature of the subdivisions of these families there is however still much confusion. Comparatively abundant material in the form of vocabularies has been extant for many years, but the character of this material is such as to give rise to as many doubts as positive determinations. The vocabularies were collected at different places and at different times by various observers, many of them untrained, using the crudest orthographies, and at times very imperfectly acquainted with the modes of life and the ideas that shape and condition the vocabulary of any people. In many cases it is therefore not impossible that dissimilar vocabularies represented similar or even identical dialects. Instances of this sort are not lacking, as is witnessed by the several Wishosk or Wiyot vocabularies, which are only interpretations of one undifferentiated language. On the other hand, until direct evidence has been brought to the contrary, it is always possible that the dissimilar word-lists represent dialects that are actually different. Between these two possibilities there is little room for any certainty in each case until more satisfactory information has become available.

On the one hand there has been a tendency among scholars not personally acquainted with the native languages of California, to regard the families of the state, which in their territorial extent and

the number of individuals comprised by them are often more nearly the equivalent of the language or dialect elsewhere, as the representatives of such less fundamental divisions, which through peculiarities of culture or environment have in California become differentiated into distinct stocks. Following out this view it is tacitly assumed that the California families show no noteworthy internal differentiations, and that the total number of dialects occurring in the state is virtually summed up by the number of linguistic families. On the other hand the diversity in point of families, the apparent differences between published vocabularies of the same family, as well as general statements by observers who have made no linguistic records, have given rise to the idea that the diversity of families is only indicative of a more general and much greater diversity of speech. According to this view the families are subdivided into languages, and these into dialects, making the total of varying forms of speech in the state an enormous number. Bearing on this question, the present paper summarizes the results of recent study of several California families, with especial reference to conditions existing in one of their number — the Moquelumnan — that is among the least known, and with a view to determining as far as is possible at present the nature and degree of the internal or dialectic differentiation of families throughout the state.

The Maidu, who adjoin the Moquelumnan people on the north, according to Professor R. B. Dixon speak three well-marked dialects or languages, which he has named northeastern, northwestern, and southern. Within each of these three dialects local differentiations are very slight. For exact knowledge of the degree of difference between the three dialects, and of such minor variations as there may be within each, it will be necessary to await the publication of the results of Professor Dixon's linguistic investigations.

The Shasta and Achomawi of northeastern California have recently been united by Professor Dixon into a single family.' Within this family he recognizes six subdivisions, several of them on the verge of extinction and surviving in only a partial state. Two of these divisions fall within the limits of the former Achomawi

1 See American Anthropologist, 1905, VII, 213.

and four within the former Shasta family. The brief vocabularies which he has published in support of his unification show all six subdivisions to be so markedly different from one another as to constitute true languages. According to Professor Dixon's personal statement, however, none of these six languages appears to have been appreciably differentiated into dialects.

In northwestern California are situated several families of unusually small territorial extent: the Yurok, Karok, Wishosk, and Chimariko. The principal body of the Yurok, occupying all the territory along the Klamath and part of that on the coast, speak a uniform dialect. In the strip of coast extending southward from the mouth of the Klamath there are three dialects, at Gold Bluff, Orick, and Trinidad, successively more divergent from the principal form of Yurok speech. The differences are, however, not very great, and even the Trinidad dialect was intelligible to the river Yurok. The Karok possessed a uniform dialect over the greater part of their territory. Those in the north about Happy Camp are said to have spoken a quite distinct dialect, which does not appear even to have been recorded. This dialect is said to have been at least partly unintelligible to the main body of the Karok, and must thus be regarded as constituting a separate language within the family. Nothing is said by the Indians of any intermediate forms of speech connecting the two languages. The Chimariko, who were a small body of people, spoke only one dialect. The same is true of the Wishosk, who, though probably more numerous, also occupied a restricted territory.

As to the Athabascan dialects of northwestern California available information is lacking. Professor P. E. Goddard, who has made a special study of these languages, groups the Hupa and a few of the neighboring Indians into a dialectic division more or less divergent from the remaining dialects.

Southward of this region, in and about Mendocino county, are the Yuki, whose speech has four principal forms. In a detached territory in the south are the Wappo, who apparently have been separated from the other Yuki for a long period, as their language is very distinct. They are aware of many similarities existing between their speech and the Yuki proper, but cannot understand

the latter. The Wappo themselves distinguish four dialects of their language, but Mr S. A. Barrett, who has carefully examined these dialects, fails to find any appreciable difference between them.

The main body of the Yuki speak three dialects: that of the Yuki proper, the adjacent Huchnom, and the coast Yuki. The coast Yuki are separated from the Yuki proper by Athabascans, and their dialect has the appearance of being a comparatively recent but rapidly diverging offshoot from that of the Yuki. Within the Yuki proper there are several minor dialectic variations the degree of which has not been determined, though it does not appear to be considerable.

Adjoining the Yuki on the south are the Pomo, whose dialects have been carefully determined by Mr S. A. Barrett. He has found seven of these dialects, which differ to such a degree in the forms of many of their words, and show distinct radicals for so many other words, that they should perhaps be considered languages rather than dialects. With the exception of one slight subdialect he finds no diversification of any of these seven languages. This is the more noteworthy as a number of distinct villages or political units were comprised under each language, and as several of the languages extended over territories of quite diverse environment, such as the coast and the interior. Mr Barrett's study was made with the special purpose of determining the total amount of dialectic variation existing within the family, and has been exhaustive. While his results cannot be expected to apply to all other families of the state, they are no doubt indicative of conditions existing in a number of them.

Among the Wintun no such systematic studies have been made as among most of the families heretofore enumerated. Partial investigations however reveal three well-marked groups. The central one of these occupies very nearly the territory covered by Glenn and Tehama counties; the other two occupy all the Wintun area respectively to the north and south. It does not appear that there were any considerable differentiations within each of these three dialects or languages.

The Washo, who, though primarily a Nevada people, occupied a limited territory in east central California, are said by themselves to have spoken one language without dialectic modifications.

On reviewing the information thus available as to the northern part of the state, it will be seen that the internal differentiation of the families is much less than has often been assumed. Several families are without dialects and none show more than six or eight. It is especially noteworthy that in most cases the dialects are on the one hand quite sharply distinct from one another, and on the other show but little or no internal variation. The statements frequently made as to the change of native languages encountered every few miles as one travels in California, with a constant gradual increase of differences, are thus on the whole unfounded so far as northern California is concerned.

In the central and southern parts of the state a smaller number of families occupy a larger territory. In several of these a greater number of dialects occur than in any family of northern California. Several stocks are so much reduced in numbers that it is doubtful whether the total number of their dialects and the degree of similarity and dissimilarity of these can be ascertained.

This is the case among the Costanoan stock, which inhabited the coast region between San Francisco and Monterey. Vocabularies have been published or procured from the Indians at the seven Spanish missions in this territory, which show that distinct dialects were spoken at each of these places. There is every indication that the total number of dialects in the family was larger. The information that there is or that can still be obtained however represents only points or isolated limited portions of an area, so that the exact determination of the dialectic groups of the stock and of their subdivisions cannot be attempted.

The Esselen family, which consisted of a small number of people inhabiting a restricted territory on the coast south of Monterey, is now extinct. There is nothing to show that this language was dialectically modified.

The Salinan family, south of the Costanoan and Esselen, is also very little known. The native speech at the two missions in Salinan territory, San Miguel and San Antonio, was different. The two dialects could not have been altogether mutually intelligible. There is nothing to indicate that there were any other Salinan dialects, but on the other hand it cannot be stated positively that such other dialects were lacking.