HOKAN. This term unites as genetically related a number of North American Indian linguistic stocks, scattered over a large area and previously considered distinct. Dixon and Kroeber in 1913 (Amer. Anthropologist, vol. xv.) coined the name from the word for "two" in some north Californian dialects, including in the family Shasta (q.v.), Achomawi, Karok, Chimariko, Pomo (q.v.), Yana, Esselen, Yuman (q.v.), and subsequently Salinan, Chumash (q.v.), Washo, all in or near California, and Seri (q.v.) in Sonora and Tequistlatecan in southern Mexico. Sapir (Intern. Journ. Am. Linguistics, vol. i. 192o) added Coahuiltecan Karan kawa, Tonkawa, and in 1925, (Am. Anth., vol. xxvii.), Subtiaba in Nicaragua, suggesting further that Siouan, Muskogi, Caddo and Iroquois would ultimately prove to be related to Hokan-Coahuilte can. The historic significance of these linguistic determinations, if confirmed, is considerable, but some authorities (Boas, God dard, Michelson, Yhlenback) remain unconvinced of the validity even of the narrower Hokan family in California, though others (Rivet, W. Schmidt) accept the full findings. (A. L. K.)