282 



SCIENCE 



[N. S. Vol. XXVII. No. 686 



anthropological consideration, is apparent 

 rather than real. 



The anthropologist needs no justification 

 for directing his attention first to language. 

 Not only is language recognized as the 

 necessary means to a really exact under- 

 standing of the life of any people ; it is also 

 the most generally useful instrument of 

 anthropological classification, and one of 

 the aids to historical knowledge which is at 

 times of the most fruitful service when all 

 other methods, even archeology, fail. Par- 

 ticularly in a region like California, where 

 the multiplicity of languages is so marked, 

 and where an absence of other means of 

 segregation and grouping is customary, 

 does an understanding of the linguistic re- 

 lationships become indispensable. 



Thirty years ago the number of distinct 

 linguistic stocks in California was pretty 

 accurately given as about twenty, and 

 Powell's great systematizing work fifteen 

 years later determined the number as only 

 slightly larger. Since then no entirely new 

 languages have been discovered. We may 

 therefore say with certainty that the num- 

 ber of native linguistic families will never 

 be regarded as greater than it is now. On 

 the other hand, recent studies show very 

 little tendency to reduce the total number 

 of stocks. The Shasta and Achomawi 

 have been found related, but this a ffini ty 

 had been at least suggested many years 

 before. That here and there languages, 

 such as Porno and Chimariko, have certain 

 important words in common with others, 

 sufih as Shasta, or with one another, is 

 not necessarily an indication of relation- 

 ship. It seems that almost every stock in 

 California has at least a few words in com- 

 mon with neighboring or more distant 

 languages, but that such words represent a 

 borrowing or diffusion. This is established 

 not only by the small number of such 

 words, but also by the fact that they are 

 frequently common to more than two 



languages. It would seem as if words of 

 certain significances had been particularly 

 liable in California to spread as loan words 

 to unrelated languages. Even if some 

 further unifications of languages now con- 

 sidered distinct should in future be made 

 by the students of California linguistics, 

 there is no reason to suppose that such a 

 reduction in the number of stocks will be 

 proportionately greater than elsewhere in 

 the North American continent. It is only 

 necessary to recall that a number of con- 

 servative scholars believe, or have pro- 

 claimed, the relationship of Natchez with 

 Muskogi, of Selish with Kwakiutl, of 

 Sahaptin with one or ruore of the neighbor- 

 ing languages, of Seri with Yuman, of 

 Shoshonean with Piman and Nahuatl, to 

 become convinced that any analogous con- 

 clusions which may be reached in Cali- 

 fornia will not be special, but will form 

 part of a general reduction in the number 

 of distinct linguistic branches, which is 

 almost certain to take place as knowledge 

 gradually increases. 



As regards the minor divisions of 

 language, system has pretty well replaced 

 chaos in California. It can not be pre- 

 tended that all the dialects are even toler- 

 ably well known, but at least the number 

 of dialects has been definitely determined 

 in all regions where the practical extinction 

 of the Indians has not made such a deter- 

 mination almost impossible. We know not 

 only the total number of divisions of each 

 linguistic family at the time of first contact 

 with the whites, but also something of the 

 relative degree of divergence of these divi- 

 sions. The loose statements formerly some- 

 times made that the number of unrelated 

 dialects of one stock was often very great, 

 and that these dialects showed a gradual 

 continuous change from one end of the ter- 

 ritory of a family to the other, have been 

 found to be entirely erroneous. In most 

 cases the divisions of each family are few 



