Page:IJAL vol 1.djvu/316

308 p-verbs embrace such ideas as and seems to suggest that the proper basis of classification is not so much transitive and intransitive as active and sta'tic, as in Haida-Tlingit, Siouan, and Chimariko. A more intensive study of the Salinan material, supplemented eventually by comparison with Chumash, Yuman, Seri, and possibly Coahuil- tecan-Tonkawa (cf. Comecrudo pa- verbs and Seri, like Salinan, adjectives in -), will doubt- less clear up this fundamental problem of Sali- nan morphology. The t- verbs (pp. 39, 40) seem most intelligibly explained as subordinates (conjunctives), morphologically nothing but nominalized forms, the /- being identical, as Mason suggests with reserve, with the common nominal /- prefix. This explanation gains force from the fact that the /-forms regularly replace p- and k- forms after " proclitic " and other prefixed elements. Thus, such a form as ram-t'-xwen THEN (HE) ARRIVED is really THEN- THE-ARRIV(ING), THEN (JT is) THAT (HE) ARRIV- ED; similarly me-t-amp' WHEN (IT) CAME OUT must be understood as TIME-THE-COMING our. Such constructions, rt need hardly be added, are common in America.

The negative verbal prefix ko-, k (pp. 41, 42) otters many points of similarity with the Chimariko negative xu,- .\-. The pronominal element follows in Salinan, regularly precedes in Chimariko. Dixon, however, remarks that the first person singular negative of verbs with y-, i- as first person singular pronominal prefix is generally xe-, the -e- replacing frequently the initial vowel of the stem. This feature is so isolated as to appear archaic ; it strongly, and perhaps significantly, parallels Salinan k-e NOT i.

The locative adverbsand prepositions (pp. 55- 57) are frequently characterized by certain prefixed elements (ina-\ tuma- ; urn-; /inn-; umpa-, tumpa- ; /-, //'-) which seem to me not quite fully understood by Mason. The most likely analysis, it seems to me, assumes a

��petrified noun *mna- PLACE, THERE, which may appear abbreviated to ma- or urn-, according to phonetic, perhaps accentual, conditions. To this element may be prefixed the article-like /-, while the demonstrative pa THAT may fol- low. The correctnessof this view is corroborat- ed by such an independent adverb as tumpa THERE, evidently t-um-pa THE-PLACE-THAT ; similarly, rnin-t'ca' IN THE WATER is to be un- derstood as r-iim-t'-ca' THE (r-<?-)-PLACE-THE- WATER. The element inn-, inn-, -nnia-\s cognate to ma- forms in Yana, Chimariko, and Pomo.

A detailed linguistic analysis of the first text (pp. 64-67) makes concrete in the mind of the reader what has been given in analytic form in the grammatical survey. This analysis is con- vincing in the main. The chief misunderstand- ings, if I may be allowed the term, are due to a failure to recognize in all cases the nominal /- prefix and to a tendency to cut loose the initial vowel of the stem or the pronominal " proclitic " vowel afld attach it to the preced- ing consonant. Thus, the form tiyaten', trans- lated as (THEN WHY) TO GO ALSO ? (freely, WHY SHOULD i COME ?) is analyzed as consist- ing of a general preposition //-, the stem ya, and the iterative suffix -Inn. Far more plausible is the analysis t-iya-lcn (\VHY) THE-GOING-ALSO? (stem iya, ia ; cf. Washo iye TO GO), possibly t-i-ya-ten (WHY) THE-I-GO-ALSO ? The " prepo- sition " //'- is probably a phantom.

In view of the rapidly increasing importance of lexical comparisons in American linguistics, the full Snlinan vocabulary included by Mason is in the highest degree welcome and will eventually constitute not the least valuable part of the book. It is precisely because of the grow- ing importance of comparative work that I have in this review emphasized- points of rela- tionship between Salinan and other languages of its group, for that it belongs to the group provisionally known as " Hokan " is now abundantly clear. Much more might have been

�� �