UNIVERSITY OF CALIFORNIA PUBLICATIONS - . IN AMERICAN ARCHAEOl-,LOGY AND' ]ETH"N'OL6GY- Vol. 16, No. 3,* pp. 47-118, 1 figure in: text -and snap September 27, 19 19 - LINGUISTIC GFAMILIES OF CALIFORNIA ROLAND B. DIXON ANDL A-.. : L. KROEBER, . i , - -;- UNIVERSITY 0CALIFORNIA PRESS BERIEBLEY- I I ? I . . I -, F o . , i I I '4 I - 0 - I I . -? - --- UNWE or CALIFORNIA P IAj-T-ONS_ -EPVART NT -0 ANBT IRPOL-OGY The following pullcain eallng with a egical and eto subjects issued wider -the diretion f the Dieprtment of thropology are seniin exchange for he publi- catibns: of: atopolical dpftm n mseums,an f u nadevotedto geeral - anthropolog- or to archae.olt and ology. Tey are for saloe at th piDl3d stted.. Ex-Achges shold be di d toThe Exchange Dparment, Tniversity Library, Eerkeley, California, UH A.- All -odA a remittances a houeds e add:essed to t Ut verity of &MItIIOAW AX OHE O AND' EENOLGGY.-A. L. Ere er: d Volume-1, 2$425; Vo- 2tl, sie $3.50e; Voue 12 andfollowing, --; pM5.0 each.;00 i - Mited as uim Qa. Pubi.,Am. AhEtb;. Pirice VolL 1-Life and Clu of thej pa, bg Pliny Earle Goddard. Pp. 188- plate s -1- -- 4.- -ip:. Sepbtember, 1603 ..-.. ...-v**^...'> oroz *w --w_o . . . . ........ 125 2--. :p2. TExts by Pliny Earie DGOcdaI?4. Pp. 89-368. March, L904 .......... 00 0 . . ,, , -, xt pp-1 3 & Vol 2. 1. The 1lrton of- the ltr Orek/ ays, by WVilliam 3. Sinedii. Pp. 1-47; p;lates-4. - _April,1904-. - .--w --g--,---tv---.t- ~ . _ 6 40 '' .he Ln'uae, o the Cosat of* ~alcrni fout of Sea YrEanlsco, by< A. L. --tebox. Pp. 2 wt - a m.p.i.u, 10- .............._ . 60 - - ~ ypes of n Cul ?tu re ern bY. L. Kureber. Pp.K 81-103. Jne,- :1904 ..... ...~. - .. ..... ..... W .*;--$..*-o*---w-t*----vM-------f-*S--. .---.r r-*---^-- 25 - -4.askeV DeBie A. rthe des of o tern atafornia --by A. L. -roebelr Pp 411 ate ..-. J y 105 ... .............. 75 5. Th okute - of oouth r l alona, by #., L. roXeber" . Pp. ; . . h-:: lKi377-. .a ara?yi. 1907 .-.-........ ......... .25 : 0 > f - In-dex 1))> 37-379-; 2. -0 - -v? -7--- - ~ Vol.3. Th0,;Mo.hologr-of the lupa L:nage, by y Earle Goddard. 344 pp. 0.. . - .- e .9 -3O50 V.4. 1.Te Ea t I torica1RelatLops betwuen.>a e nd Japan, from- o--- > ~- -: - docum-ent s prsrvedin Spainnd .Tapan, -by Zela- Nutetall Pp 1-47.---- wi ~'',* 2.'tf ContiuetiOn tn te Phsiw AnilXopolg of Calionibsd- on Oollc- * f rf -~ -- t1oz1 in tbQe Deia$miwo 0o of the Uivrsit e C:alifornia, - -~ ~~~n In t- U.-;- S.J Nat*iional ,b Zum hbIAes Hdicka. Pp 49-04 wit;h 3. The 1 and m--5-tablr p.at- Wn, 1 .- . .....an , y75 oitiuShonion-a Di tof ,ia, by A. Kroeber.Pp:o5-66 D~~~ "February 1.90t7 . ... ....9wW-.......-..........-.. .....~.. 1.50*w** t> k w@ * -a---\t# - ~~~ 4 .Inbe. Pp. -3iL8. -Sep13teme . 1907- .... .... .~.. . ...........~-- -s--->--.... . !W- ...75 6.-T6 eligloiif the Indias of Calfnia -b A. L. Eroeber. Pp 349-356. aepteme, 1e07 - ... ...._. P..... .. -........ . tIndes 1pp7 357-374 Vol.5. 1. The; Pbo o Part I, Th 4vidua Sound,b - PlT Earle l.oddard p 2, 'f. plates 1-.. Mach, 1907 ... . .i . *35 2 --. t.Navaho Mythsrayers and Sogs ith Text and 'JEaslations, by- Wash ingt. -.n ion Mattews, edited WbyV & WEaeG.ddard. Pp. 21-6. Stem -. - 3. K-to -1.xtso by ?1 Piy Earle Goddard. Pp.t 65-238, plate ?. Zbecembe 190 2.50 ;0' t'4:. Te MateIa Culitueof sthe 'laiuath Lae and Modoc In.diansof .Nor+-th N * - ---. .eastern .California an& Soutek Orgon, by S.. A. larrett. - Pp. 235-292, -- plte 1-2 J P n- e 1910o**w *-. .-. ... . . . -. .*-/S ----w r--w 75 5. - & The Cl3marlo . nliaisi -andiLaSn;Bage :by Rolan -.: DIo. Pp.- 2534.80k* Augst 190... . ..................... 1.07 Index, pp. 381-3484 VL 6. 1.The : E 01_tboG r4hyof t_ Tm ndNihborin Idian, y am .Alfre Ba.rrett. Pp.1-33 P. P. 0. . -5February 1908 ........ ...... 3.25 .:'--S' ..- \2.' The -Geogaph :and.Dlect o.f 'the~iIwok n.>ians b Samue Alfred - Bahret. Pp,3-68, .map 3by. - -Was,, 3. . On the Eince Goft$he Ocduption oi aJad Zlns bthe wk Indians by- A. X.f ErQe.er. ' Pp. 889-80 Nos 2 and. - in one ove ard e Tpy 1908- . e 90 .5 IndeX, pp. 881400. - * ; . f~~~~~~a - i 0 -by->, ,.-~ .av 1 0 . -23 ~ ,- ,AAWI ao- CACS / AhA EJ dAtAo'an *41#wAh lb% - l 4f,0tscaN O >~il^ttaN IN& 10- - V4100 fte lllt? UNIVERSITY OF CALIFORNIA PUBLICATIONS IN AMERICAN ARCHAEOLOGY AND ETHNOLOGY Vol. 16, No. 3, pp. 47-118, map 1, 1 figure in text September 27, 1919 LING-UISTIC FAMILIES OF CALIFORNIA BY ROLAND B. DIXON AND A. L. KROEBER CONTENTS PAGE HISTORICAL INTRODUCTION ...............................................48 THE PENUTIAN FAMILY .................................................55 Cognate stems.............55 Basic list .........................................................56 Other cognates ...................................................67 Changes other than sound shifts .......................................69 Expansions........................................................69 Contractions and reductions ........................................71 Metathesis .......................................................73 Assimilation ......................................................73 Analogizing and reduplication ......................................74 Summary ........................................................74 Sound equivalences ...................................................74 Stops ............................................................74 Palatal T series ...................................................77 P andK .........................................................77 S andK .........................................................78 L andK .........................................................79 Surd L ...........................................................79 R ...............................................................80 The palatal nasal .................................................81 Mixed vowels .....................................................82 Original Penutian sounds............................................84 Accent .............................................................84 Suffixes .............................................................89 Morphology .........................................................90 Vowel mutations ..................................................90 The verb and its affixes ............................................91 Pronominal elements ..............................................92 Number and case in nouns .........................................95 Form and use of stems .............................................97 Geographical organization ..............................................98 Historical interrelations of the Penutian languages .........................O OTHER FAMILIES ........................................................103 Hokan and "Iskoman" .................................................103 Washo ............................................................. 104 "Ritwan"............................................................112 Athabascan ..........................................................113 Shoshonean..........................................................114 Yukian.115 48 University of California Publications in Am. Arch. and Ethn. [Vol. 16 HISTORICAL INTRODUCTION Powell's classification and map of the linguistic families of America' allotted twenty-two families or parts of families to California. For a number of years this reckoning stood unaltered; for many years it was unchallenged even to its details. In 1905 one of the present authors showed the relationship of the Shastan dialects to Palaih- nihan (Achomawi and Atsugewi),2 and his combination, first under the name of Shasta-Achomawi and then of Shastan, came into general acceptance through its recognition by the Bureau of American Ethn- ology.3 This fixed the number of families in California at twenty-one. A few years later the same writer suggested, on the basis of some preliminary but fairly considerable body of material, that Chimariko and Shasta were sprung from a single stock;4 and his colleague sup- ported the view that Miwok and Costanoan must probably be con- sidered akin5-as, in fact, had been the custom before their separation by Powell. Neither of these consolidations, which would have reduced the families represented in the state to nineteen, was, however, advanced without reserve, and they failed of general acceptance. Several other possible cases of kinship presented themselves to the writers at various times, as between Wintun and Maidu, Maidu and Yokuts, Chumash and Salinan, and even Yuman and Esselen, and several brief lists of similarities were compiled by them. Once or twice allusions to these possibilities found their way into print.6 But, on the whole, the resemblances seemed scant and best interpretable as due either to coincidence or borrowing. By 1903 at least a slight body of information had been secured on the grammar of representative dialects of each of the twenty-two families then recognized in California, and the present writers pub- lished a brief monograph7 classifying these families into several types and subtypes on the basis of structure. They expressly emphasized that this classification was not genetic, and assumed that while the structural resemblances were undoubtedly actual, they must be inter- 1 Ann. Rep. Bur. Am. Ethn., VII, 1891. 2 Am. Anthr., n. s., vii, 213-217, 1905. 3Bur. Am. Ethn., Bull. xxx, 1907, 1910. 4 Present series, v, 337, 1910. 5 Ibid., ix, 259, 1910. 6 Ibid., p. 261. 7 Am. Anthr., n. B., V, 1-26, 1903. Dixon-Kroeber: Linguistic Families of Calif ornia preted as due to secondary influences exerted on each other by lan- guages of distinct origin. In the main, or at least at several impor- tant points, the non-genetic type classification then advanced coincides with the genetic classification presented in the present paper. That it is different in several cases is duie to the scant knowledge then extant as to most of the languages, a paucity that necessitated the basing of comparisons on a small number of comparatively outward features of structure, such as the presence and absence of pronominal affixation and a plural in nouns. That the suggestions as'to genetic relation- ship which this classification obviously bore were not adopted or even followed out by the authors 'was due to an essentially conservative attitude, a reflection of the views generally current for some twenty years after the publication of Powell's basic classification, views not yet wholly deprived of vitality, and certainly in the main justified by the sense' of order and fixity which Powell's work introduced as against the hesitations and random irresponsibilities of previous less systematic attempts. This conservative attitude was adhered to by the writers for a number of years-maintained, in fact, to a point that now seems to them to have been unnecessarily belated. 'As evidences of similarities between this and that language accumulated, they were indeed noted, but were consistently interpreted as instances of one unrelated lan- guage borrowing either material or machinery from another.8 Finally, it seemed desirable to bring together all the readily avail- able data and determine the exact degree and nature of the similarities. About two hundred and twenty-five English words were selected on which material was most likely to be accessible in reasonably accurate and comparable form, and the known native equivalents in sixty- seven dialects of the twenty-one stocks were entered in columns. Com- parisons were then instituted to determine all inter-stock similarities that seemed too close or too numerous to be ascribed to coincidence. The purpose of the study was threefold: first, to ascertain the nature and degree of borrowing between unrelated languages; second, to trace through these borrowings any former contacts or movements of language groups not now in contact; third, in the event of any rela- tionship existing between languages then considered unrelated, to determine this fact. Some similarities were soon apparent between nearly all the twenty- one stocks, "and with the completion of the comparative vocabularies 8 Compare, for instance, A. L. Kroeber, Anthropos, viii, 394ff., 1913. 1919] 49 50 University of. California Publioations in Am. Arch. and Ethn. [Vol. 16 the number of resemblances had become considerable. They seemed to show little, however. Families some distance apart on the map often had more stems in common than those in juxtaposition; if the remote group was regarded as once in contact with the one with which it shared most words, it must have been in contact also with others with which it shared but few words. Stems of the most diverse and scattering types of meaning showed similarities in distinct stocks: beyond onomatopoetic names of birds there was no one class of words that seemed more or less given than others to transference by loan. The results, in short, appeared completely meaningless. Finally, in a mood rather of baffled impotence, an interpretation of the cases of most abundant resemblance as due to genetic relation- ship was applied. At once difficulties yielded, and arrangement emerged from the chaos. The appended table 1 shows the first results. This table gives the number of stem resemblances common to every two of the twenty-one stocks, or to any dialects of every two of the twenty-one. It is evi- dent that the highest numbers occurred in two groups, both of which are marked in the table by enclosing frames or boxes. One group contained the interrelations of Karok, Shasta, Chimariko, Yana, and Pomo; the other of Wintun, Maidu, Miwok, Costanoan, and Yokuts. Every instance of two stocks sharing twenty or more words was found to be included in these two groups, as were the majority of instances of two stocks sharing ten or more stems. In other instances the number of words common to two stocks was absolutely small, but distinctly higher than the number shared by either with any other stock. Thus, apparently common to Chumash and Salinan, 12; to Chumash and any other language, 8; to Salinan and any other language, 6. Again, Yurok and Wiyot, as per the table, 4; Yurok and any other, 0 to 1; Wiyot and any other, 0 to 3. In both these instances the hint was reinforced by the fact of u'Uques- tioned structural similarity; and in the latter of the two, at least, a renewed search prompted by the growing suspicion quickly revealed an additional group of shared stems that had either not entered into the comparative vocabularies or whose resemblance had been over- looked. An arrangement of the highest numbers of table 1 in order, as in table 2, also proved significant. The first sixteen pairs comprised eight belonging wholly to one of the groups (A) marked off in table 1; five belonging wholly to the other group (B); one (X) consisting -of Dixon-Kroeber: Linguistic Families of California . . . . . . . . . . . . . . . . . . 0 0) 0- 0- cn toD 01 O O O O O coL o to 0 to to to. 0) 0: to) to toI 0 --4 to- 0) )-A 0 0 CO - 0 01 01 . = Cn Cn 00 to to 0 1 C t S 01 to C 00 to UV CO CO O= -1 -4 CO - CO 0n Ci t O- 00 CA -4 00 t O C C toD 0 o P CO CO 4 - t ts v r CO CO CO O 01 O c0 to c0 c0 4 00 t o cO 01 01 to co cO CO 00 co t t w C OD t o O C 04 04 0J C I to - to 0 c 00 I4 CO -to -tcoI 00 -1 oo 00 -o ts I o 0I -- O5 0 I c: 00 v -- v 00 cn CO 01 o to CO to v v to 01 CO Os 01 v O~ 01 s cn ~ ~ - C 0 t 0 to 01 to - 0 0 0 0 P- 0 I.. 0 ~- C to to o ~ to o - to - - c co 01 c co CZ P I 01 IO -' CO S to CS 01 Co t C t o t CO 0 1-- I 0 4CO 0 0 0 0 0 C) O Athabascan Yurok Wiyot Lutuami Karok - 0 0 Shasta Pt < - 0 Chimariko to t o 0 0 Yana O= C I - Pomo gv 0 , o Yuki o o 0 0 Washo - v-- 0 Wintun t00 C o 00 t Eo -4 - 0 0o Maidu co toi -' s o 01 to to ?t 0 0Miwok o c -1 -4 0cn c0n to to o 0 0 Costanoan CO to PP., CO 01 00 00 0 co C) CA cW PP. 01 00 to c-- cn -- to CO cn to -4 --4 to 01 0 co co CO 01 cn cn 0 --4 co cn 01 0~1 CO 0 toD 0 0 01 toD 0 to 00 0 CO 00 P- P- W OC 0 0 toD Yokuts Esselen Salinan Chumash Shoshonean Yuman Aggregate Stems represented 1919] 51 to Co to 00 CA ~-- 01 Cn o 01 01 cO 0o 00 oo -4 ,_ P-- CA oo oo ~-- C0 P-- oo- cnL 52 University of California Publications in Am. Arch. and Et'hn. [Vol. 16 an A language and a B language in territorial contact; and two (Y) consisting of a language-Yuki, included neither in group A nor in group B-with respectively an A and a B language, between which it lies wedged geographically. Again, Shasta and Chimariko had previously seemed related; but Shasta and Pomo presented more apparently common stems (21) than Shasta and Chimariko. So with Miwok and Costanoan: their similari- ties (27 in number), were equalled by those of Wintun and Miwok, and nearly equalled by those of Maidu and Yokuts (24) and Maidu and Wintun (23). TABLE 2 Order Families Resemblances Group 1 Miwok and Costanoan --------------------- 27 A 2 Miwok and Wintun --------------------- 27 A 3 Maidu and Yokuts -.----------------------24 A 4 Maidu and Wintun -.----------------------23 A 5 Shasta and Pomo --------------------- 21 B 6 Shasta and Chimariko -------------- 19 B 7 Yokuts and Costanoan- -------------------- 19 A 8 Yokuts and Miwok --------------------- 18 A 9 Maidu and Miwok --------------------- 18 A 10 Yokuts and Wintun --------------------- 17 A 11 Wintun and Pomro -------------------- 17 X 12 Yana and Pomo --------------------- 15 B 13 Wintun and Yuki --------------------- 14 Y 14 Pomo and Yuki -.----------------------14 Y 15 Chimariko and Pomo --------------------- 13 B 16 Shasta and Yana --------------------- 13 B In short, only two attitudes seerned consistent. One was the old one of regarding each of the twenty-one stocks as totally unrelated; the other, to unite some of them as per the groups outlined. In one respect the figures in tables 1 and 2 were quickly found unfavorably misleading. They might or might not be considered as including radical words due to a common origin: they certainly in- cluded words not due to such common origin but derived by loan. This element must obscure the incisiveness of the conclusions derivable from table 1. Further, the comparisons used being avowedly super- ficial, that is, not based on analysis, a certain number of false coin- cidences were bound to have crept in. For instance, the writers never supposed that the nine stems apparently common to Yokuts and Sho- shonean indicated any degree of genetic unity. Several of the nine were obvious cases of borrowing; the remainder were so few that they Dixon-Kroeber: Linguistic Families of California could scarcely be regarded as due to anything but accidental outward resemblances such as are bound to arise on the probability of chance. Now Yokuts, in the table, is credited with twenty-four stems shared with Maidu, with which it is thrown into a new family, and nine with Shoshonean, to which it is unrelated; a proportion which certainly seems unconvincing. But on elimination of the nine Shoshonean cases as not indicative of kinship and subtraction of an equal number from the Maidu-Yokuts total of twenty-four as probably also due to loan or coincidence, the proportion becomes 15 to 0; which is dis- tinctly more positive, particularly in view of the comparative paucity of the material examined and the necessarily somewhat mechanical nature of the comparisons instituted on a first survey. Actually, in this case, the proportion would be stronger than 15 to 0, because Shoshonean is in geographical contact with Yokuts and Maidu is not, so that the probabilities of borrowing would be less in the latter case and the figure properly subtractable from the gross twenty-four would presumably be less than nine. Another factor tended to obscure the first results. Table 1 deals only with stocks as units. Actually many of them were represented by data from several dialects or languages. To have included all these dialects in the table would have been exceedingly laborious and ren- dered the table rather unwieldy as a basis for'reference. As drawn up, a stem common to only two dialects of different stocks was, there- fore, entered exactly like a stem common to all the dialects of the two .stocks: each counted as a case. For example, when the Tiibatulabal dialect of Shoshonean, which is spoken in contact with Yokuts, has obviously borrowed the Yokuts word for ''elk," the same weight at- tached to this instaihe, in the table, as to the fact that the identical stem for "two" runs through all dialects of Yokuts and of Maidu. Now the comparatively high number of Wintun-Pomo (A-B), Yuki- Wintun (Y-A), and Yuki-Pomo (Y-B) resemblances, which alone break the uniformity of table 2, is clearly due in considerable measure to this source of error of the tabulations. The Wintun-Pomo similarities were largely between dialects of these stocks, not between the stocks as wholes; and about half of the Yuki-Wintun and Yuki-Pomo cases held only for a small detached offshoot of Yuki, the Wappo, which is flanked on the map by Wintun on one side and Pomo on the other. These considerations, accordingly, strengthened the authors in the conviction that the soundness of their inferences was greater than a merely cursory inspection of their results, as summarized in table 1, 1919] 53 54 University of California Publications in Am. Arch. and Ethn, [Vol. 16 would indicate; and thev proceeded to announce their findings, at first in mere abstract,9 and soon after with some slight indication of the nature of their evidence.10 It was necessary to find names for the new groups or families. To extend the designation of one member of each group to the entire group would have been as misleading, in the end, as it would have been to name the tongues of Europe " Sanskritan. " Binary designa- tions of the type of Indo-Germanic, Ural-Altaic, and Uto-Aztekan were likely to prove unrepresentative, and certain to be clumsy. There seemed no recourse, accordingly, but to new and therefore arbitrary designations of the types of "Semitic"; and genealogical as well as national appellations being wanting in the native Californian field, and none of a geographical character applicable, the family names proposed were based on forms of the numeral "two" in the families of speech involved. The result, then, of the investigations up to this point was the setting up of four larger families: Penuttian, consisting of the Wintun, Maidu, Yokuts, Miwok, and Costanoan families as previously recognized. Hokan, including Karok, Chimariko, Shastan, Pomo, Yana, and, by subse- quent addition, Esselen and Yuman. Iskoman, namely, Chumash and Salinan. Bitwan, or Yurok and Wiyot. One thing and another delayed the marshaling in print of the evidence on which this new classification rested. Meanwhile, two of the new "families" were themselves merged in others; for the validity of a third, sufficient proof has been forthcoming from another source; until, now, the authors' task has shrunk substantially to an examina- tion of their material from the five stocks named Penutian. The major part of this paper deals with this examination. The intro- duction, which herewith concludes, has been preposed to make clear the bearings of the Penutian findings to the work done in the past. Inci- dentally, the historical nature of this review may serve to establish that the authors proceeded not impetuously, but rather reluctantly, and step by step. In fact, as they look back now upon their partici- pation in the problem of the reduction of the multifarious Californian languages to a smaller number of original units, they are impressed by their almost undue conservatism. They seem to themselves to have been timid in approaching the possibility of relationship at all, and slow in uncovering the most patent cases even after they held the clue. 9 Science, n. s., xxxvii, 225, 1913. 10 Am. Anthr., n. s., xv, 647-655, 1913. Dixon-Kroeber: Linguistic Families of California THE PENUTIAN FAMILY COGNATE STEMS The list that follows gives one hundred and seventy-five Penutian stem resemblances that were found among the native equivalents of the two hundred and twenty-five English words with which this study was commenced. The material is arranged in the order which has become usual in briefer American vocabularies. A number of sets of cognates have been taken from their numbered place in this list and put at the end to allow of their being printed in columnar form, with a view to bringing out parallelisms that otherwise might fail to impress without detailed analysis and discussion. Thus, number 6 does not come between 5 and 7 in the main list, but heads the following series. Orthography. This probably speaks for itself to any Americanist. It may be said, however, that c is or resembles English sh; tc is English ch; R is the palatal nasal; x is a palatal fricative; t, d are palatalized; I is the glottal stop; k', t', t', p', tc' are glottalized stops and affricatives; L iS voiceless 1; y, a sonant fricative palatal; o, ii are unrounded mixed vowels. Sources: Maidu: manuscript records by R. B. Dixon. Yokuts: manuscript records by A. L. Kroeber. Wintun: chiefly manuscript records by R. B. Dixon; also by H. B. Wilson and A. L. Kroeber; also, published vocabularies of S. A. Barrett"l and J. W. Powell.12 Miwok: primarily, the vocabularies published by S. A. Barrett,13 supple- mented by manuscript material by R. B. Dixon, A. M. Tozzer, and A. L. Kroeber. Costanoan: various data, as assembled in a publication by A. L. Kroeber,14 subsequently increased by the San Juan Bautista material of Arroyo de la Cuesta as classified by J. Alden Mason.15 Abbreviations: W Wintun N Northern c Central SE Southeastern SW Southwestern s Southeastern and Southwestern 11 Present series, vi, 81-87, 1908. 12 In Stephen Powers, "Tribes of California," Contrib. to N. A. Ethn. II, 1877. 13 Present series, VI, 69-80, 362-367, 1908. 14 Ibid., ix, 243-248, 1910. 15 Ibid., xi, 399-472, 1916. 1919] 55 56 University of California Publications in Am. Arch. and Ethn. [Vol. 16 Abbreviations: Md Maidu NW Northwestern NE Northeastern s Southern (Nishinam) N Northwestern and Northeastern Y Yokuts N Northern v Valley K Kings River T Tule-Kaweah p Poso Creek B Buena Vista Mw Miwok s Southern Sierra c Central Sierra N Northern Sierra p Plains co Coast L Lake C Costanoan F San Francisco j San Jose CL Santa Clara CR Santa Cruz B San Juan Bautista (Mutsun) M Monterey (Rumsen) Basic List 1. Person. W c win (man), s wi-ta (man); Md N maidu, s maidii-k; Y K mayi, T mai; Mw s, C miwU, N miwu-k, P miu-ko, co ula-mitca. 2. Man (1). Y T taati (person), P taat (person), B toxy; Mw P cawe, co taiyi-c, L tai; C J, B tare-s, CL tare-c, CR tcare-s. 3. Man (2). Y N, V nono; Mw s, C, N nania. Cf. adjacent Mono (Shoshonean) nana. 4. Boy, Youth (1). W c k 'aina, s seri-ta, sw teurai; Mw P sali-nai, co hena-s, L hena-putu; C F ceni-s-muk, CR ala-cu, M cini-emk. 5. Boy, Youth (2). Y N, V, K, T notco, N nonkoi, P kohutem; C j kotco, B kotcinsix. 7. Woman, Girl (2). Mw (woman) s oha, C, N osa, P fucuufi; C (girl) i atsya-kic, B atsia-gnis, M atsy-amk. 8. Girl. Md NE konoi-be, s konai; Y N gaina, v gaita, T guyodum, P getep-al, gadap-cul, B guyoyanm; MW L koya, co kola-putu; C F kata. 9. Baby. W N ila, c elet, SE ilak, sw ilai; Md NW tsilak 'a, NE konoko; Mw s esellu-tki; C F oklucku-c. 10. Old man. W N kiemi-la, c kias, sw tciak (metathesis); Md NW kano, s oskon; Mw s humele-tki; C F hunta-x (old), J, CL hunta-te; perhaps also, by metathesis, Y v nioxelo, K moxolo, T moxodo. 11. Head (1). W sw dil; Md s tcul; Y K dool, T tot; Mw P tolo. 1]a. Head (2). W SE ana; Md N, S ono (s: hair); Y N otco, V oto, K, T, B oto (hair); Mw C, N hana; C J, CL, B uri (hair), CR uri, M ut. Dixon-Kroeber: Linguistic Families of California 12. Head (3). Mw co molu; C j motil, B moxel; Y (forehead) N pitciu, V pitiu, K tiliu, T tiidu, P peleu. 13. Hair, Forehead. W (hair) N tomoi, c tumoi; C (forehead) J tima-x, CL tima-tc, CR, B tima; and possibly Md sun. 15. Ear (2). W N, c, S mat; Md N, S bono. 16. Eye (1). W N tum, C LUi; Md N hini, s hin; Mw s huntu, C suntu, N SUtU, CO, L suit; C J hiny, CL, B, M xin, CR hin. 17. Eye (2). W SE sas, sw sa; Y N, V, T, K caca, P ceca, B sasa-'l. 18. Nose (1). W N, C sono, S Linik; Md NW sumo; Y N sinik, N, V, K tinik, T tiiuiiik, P tefiik, B tinak. 19. Nose (2). Md NE hiku-m; Mw N huku, P, CO, L huk; C F, CL, CR, B US, J hus, M Wus. 20. Mouth. W N, C, S kol, SE koL; Md NW kombo, NE simi, S sim; Y N, V, K, T cama, P cema, B sa m. 21. Teeth. W N, C, S Si ci; Md NW tsawa, NE, S tsiki; Y N, K te]i, v te'yi, T tafi; Mw s, C, N kutu, P, CO, L kfit; C F, J, CR, B, M sit, CL siit-em. 22. Tongue (1). W N, C, S tahal; Md N eni, s alili (assimilated); Y N talxats, v talxat, K taiXit, T tadxat, P talapi-s, B a'lada-s, K madat. 23. Tongue (2). Mw s, co lemtep, c, P nepit (metathesis), N nepitu, L letip; C F lasek, J. B, lase, CL lasex-em, CR lasa. 24. Beard. W N -teeki, c -tseket, s sep-sep; Md N simpani, s masau (by meta- thesis?); Y N dabuts, N, V damut, T damot, K djamoe, P djamec, B damuc. 25. Chin. W s kaba; Md Nw yappa, NE tsawa-m-b6mim; Y N uguC-ul, awate-il, V, P awac, K awadji, awaic-il, T awaci, B yuku-l; C M awic. 27. Shoulder. W N salu-s, C, SE sala; Md N, S dadaka; Y N gapeal, v gapsai, tapal, K tatal, T tapad, P gepeil, B gapsai. 28. Hand (1). MW s, C, N tissu, s ukuc, N ukusu, P eku, CO, L uku; C F. Jy CL, CR, B (arm) isu, M (arm) is; and possibly W s, C, N sem, c, sw (arm) sala. 29. Hand (2). Y N bonoc, v bunduk, K poton, T putonf, B potoni; Mw (arm) s patcan, c woniotu; C M puts. 30. Arm. Mw s, P tawa, N tumalu, P tumal, co tauli, tali-k, L tauli-k; C CL ta]me-s (metathesis). 32. Breasts. W C, SW imit, SE imik; Md NW nuni (assimilated); Y N, P minite, V, T, B menit, K mintci-x; Mw s, C, N musu, P, CO, L mu. 33. Belly. W sw bus; Y N balik, K olok', lak'i-n; Mw L puluk; C M pitin. 34. Knee (1). W N puiyak, sw pomok; Md N pokosi (metathesis), s podok; Y N, K opuk, K UpUC, N bucon, v kuyut, T kuyoc-ud, P kuyoc-il, B kuyos-u '1. 35. Knee (2). W c, sw anak; Mw s hofioi, C, N hoiioyu, P honoi. 36, Knee (3). Mw L tokollo; C M dol-s. 38. Foot (2). W c kole; Mw N. P kolo, co koyo, ko, L kollo; C F kolo, CL, CR (leg) koro, B M koro. 39. Heart (1). Md NE honi, s hon; Y N, v hon-hon, K hono-hon, T honi-hon, p honiheni. 40. Heart (2). W N pudu-s, puris, s puru; C F, CR mini, J minyi-x, CL minitc. 41. Bone. Y N, K tc'ei, V, T, P tC'iI B ts'i-x; C J tayi, CL taxi-m, CR tcaye, B tati, M tcatc; and perhaps Mw s kuteute, C, N kutcutcu, co mutci. 43. Blood (2). W N dedeki (assimilated), c, s sak; Md N, s sede. 45. Fat. W N hom, SE hum; Md NE hiti, s h6t; Y N, V. K, T hexa, P hiya. 46. Skin. Y v daplan, K tapia, P p 'alai, B gapac, K totop, V, K tcului, T tcudui; Md (by metathesis) NW potani, NE posala, s potomn. 47. Faeces (1). W N tceni-s, c teeni, sw teni; Mw s kunat, c kunatu-s, N kunatu, P kuna. 1919] 57 58 University of California Publications in Am. Arch. and EtJhn. [Vol. 16 48. Faeces (2). Md NE piti; Y N, K pidik, v, T bidik. 49. Urine. W N tcunu-s; Md NE tsumni; Y N, V, T tcuyon, V teulon, K tuyun. 50. Sinew. Md NE paka; Y N, V, K pikil, T piked. 51. Horn. W c k'ili, sw teili; Y v icel, T iicad, P e'Cil, B ueul. 52. Tail (1). W N cUtU]n; Y N kut, v gut-'oc, K kot, T gut, P got. 53. Tail (2). Md N buku, s buk; Y B tapak. 54. Navel. W N nak; Md NW naka. 55. Chief. W c, s sektu; Md s huk; Mw P teeka. 56. Shaman, Ceremonial Initiate. W c hiyom, s yom-ta; Md N yomi; Mw C, N alini, P umise, L yomta; and possibly Y V, K antu, T aiitu, P efiteu. 57. House (1). W N, SW k'ewel, c k'el, s k'ewe; Md N hobo, s h6p6u; C F riwa, J ruwai, CL "gruwa-m," CR ruwa, B ruka, M ruk. 58. House (2). Y v, K, P ti, V tsi, T, B te; Mw s, c uteu, C, N, P, CO kotca. 60. Arrow (1). W N not, s noko; Md N noko. 61. Arrow (2). Y N, V, K, T t',uyo-c; C CL tawi-s-em, CR teemo, B temo-x, tiyo-s, M tep-s. 62. Knife. Md NW tsumi, NE tsami; Mw s cope, L tsitsa (assimilated); C F tipa, CR teipa, M tip. 63. Boat (1). Y N, V, K, T, P owon, B u'win; Mw s wote; C F wali, CL walin. 63a. Boat (2). W s nu; Md N n6; Mw L nu (borrowed from W s). 64. Boat (3). Md NE yaka, s wakai; Mw N saku, P saga, co caka. 65. Pipe (1). W SE dopo; SE topor; Mw P topo-kela; C J torepa, CL "trepa-m," cR torexo. 66. Pipe (2). W sw bomit; Md N pani (tobacco), s pan; Y v baum, N baum (tobacco), P bamuni, B bomots; Mw s, C, N paumma. 67. Pestle (1). W N satak, c cotok, SE taki, sw t 'usa; Md N.W suni. 68. Pestle (2). Md NE bayi, s bai; Y V polwoi, K palui, T padui; Mw P hopa, co pa, paya. 69. Sun (1). W N sas, SE con, sw sun; Y P hicta; Mw C, N hiema, P, CO, L hi; C F, CL, CR, B iemen, J hiemen. 70. Sun (2). Md N poko, s oko; Y N, V, Op, K Upu-C, T opodo, T upi-c (moon), P opo-c (moon). 71. Moon. Mw s, C, N, P kome, L kume-nawa; C F kolma, J, CL korme. 73. Night (1). Mw s kawulu-to, C, N kawulu, P, co, L kawfl. 74. Night (2). W N ken-wina, c leni, sw sinol; Y N tuyon, V, T, P, B toyon, K toyoni. 75. Cloud. W N, C k'a, s k'ir, k'ira; Y v k 'ilei, K k 'elei, T k 'iidai, P k 'eli, B p'iya-p'iya; Mw co ilau, ya-kal. 76. Lihtning. W N waluka; Md NW -winaka, NE wipili (metathesis), s wispil; Y N walma, v walam-wiya, K walam-wila'a, T wadam-wida; C F wilka-warap, CL wilka, Ca wilep. 78. Snow. W N yolo, c yola, s yol; Md N, S ko; Mw s, c, N, P kela; L tana; C J, CL wakan, CR wakani. 79. Fire. Md N, S sa; Y N, V ocit, K UCUt, T, B ucit, P ocot; Mw s, C, N wuke, P wiike, co wfuki, L wiki; C CL, M coto, B coton. 81. Ash. W N bukul, c puk, s put; Md NW pOpO (assimilated), NE pUpU; Y, T, P hapac. 82. Coal. W N, c kali, SE wali; Y N, V calu, K cal 'an, T capan, B soolin. 83. Water (1). Y N, V, K ilik, T idik, P elik, B ilak; Mw s, c, N kiku, P, CO, L kik (probably by assimilation); C, all dialects, si. 84. Water (2). W N C, S mem; Md N momi, s mom. 86. Mountain (1). W C tCoL, S tOL; Y N, K dulul. Dixon-Kroeber: Linguistic Families of California 87. Mountain (2). Md N yamani; Y V lomi't, T domit, P lomat, B lumit; Mw s, c leme. 88. Stone. W N, C SOn, SE s05, todui, tului, sw kodoi; Y N, v cilel, P xelul, B xalu'l. 89. Salt. W N, C, S weL, Md N, ? ba. 90. Wood. W N tCUs, c took', SE tok', sw tok; Md N tsa; Y N, v hit'ec, K, hit'ic, T hilt 'ac, P etis; Mw s huhU, C, N susu (assimilated). 91. Grass, Brush. Y v, K yawil, T yawud; C F yama, CR wai. 92. Sugar Pine. W N, C sumu; Md N, S SUmU; Y N, V tconoxi-C, T tcoonoxi-c. 93. Yellow Pine. Md N, S imi-m-tsa; Y v inil, T idfiit; and possibly W N kulem. kulem. 94. Digger Pine. W N tcati, c to 'oko, SE to 'usak, sw t 'uwa; Md NW, S toni-m-tsa; Y N, K ton, T tofia-c. 95. Buckeye. W c pasa; Md N, S polo-m-tsa; Y N, V dopin, K dupun, T dopofi. 96. kanzanita (1). W N pai, c paka; Md NW iipiU, NE epo; Y N apteu, V, T aptu, K uptu, P opto. 97. MafAzanita (2). W SE eya, sw eye; Mw s, C, N, P eye, CO, L eyi. 98. Tule (1) .a W SE sa-kulul, guye; Y N k 'oyi-c, V got, K k 'oyu-c, T got, P gatswei, kololi-s; Mw L kol. '99. Tule (2).a W C Lap, SW LOP, Laka; Mw co loko, cappa. 100. Tule (3).a W N pat-pat, SE p'ocak; Md NW pok-poko; Y N, v bumuk, v poton, K patak. 101. Dog. W N cuku, c oukut; Md N SiU, s suku; Y N tcexa, K tcecec, T tcejej, P toicuo, B tseses; Mw S, C, N tcuku, P teuteu; C CL tcutcu, Ca hi-teas. 102. Coyote. W sw ul (wolf), Md Nw oleli, NE helieni (wolf), s ole, hosai, Y B hulapinsa-s; Mw s aheli, c aseli, N oletcu, P oleti, co oye, L ole. 103. Fox. W N, C hau; Md N hawi, s hau; Y N, V, K, T au'dja, B gaudat. 105. Grizzly Bear. W N wima, C wemaL; Md N pano. 107. Deer. W N, C, s nop; Md s dope. 108. Elk. W sw lokoya, Y N, V, K, T coxgoi. 109. Jackrabbit. W N patkili-s, C patkili; Md NW, s palali, palal (rabbit- assimilated), NW, s boye; Y v capal, dopol, T dopod, tukuyun, P tokoo-c, B te'abakiu; Mw s, C, N eplali, P epali, co aule. 110. Bird. W N toil-toil, c toiltoit; Y N to'enpai, K denip, T detoip, P wiwitsi (assimilated), B we'tuts; Mw S, c teiteka (assimilated), N miteema-ti, P teiteipu-k (assimilated); C B teiri-t-min. 111. Buzzard. W N, C, s hus; Md NW hosi, NE mini-husi, s hos; Y N hots, hutawi, v hot, koteya; Mw s huhu, c husu, N, P teuhu (metathesis), L hus. 112. Crane (1). W N kat; Md N waksi, s k6u; Y N, V, K waxat, T waxit. 113. Crane (2). W N toriha-9; C B ture-s. 114. Turtle. W c an, SE anu-c, sw ano; Y N on; Mw s awan-ta, C, N awanna-ta, P awanna-i; C F auni-e-man, CL, CR, B auni-e-min, M xuno-s-xim. 115. Rattlesnake (1). W s tiwil; Md NW cola, s sola; Y V, K teel, T todd, P tiel, B tiwi-tiw; Mw L holo-mai. 116. Rattlesnake (2). W N Lak, C Letceu; Mw s, c lawati. 117. Frog. W N watak-mit, c, s watak; Md NW awani, NE wakat (metathesis); Y N watak-saai; Mw s wataku-na, C, N watak-sayi; C B wakarat-min, M wakatse-m (metathesis). 118. Fish (1). Y N lopitc, v lopit, K luput, T dopit, P lopat, B lupit; Mw s lapisai, C, N lapisayu, P pu, co elewi, lota; C CR helai. a There are three common species of Scirpus and Typha generally called "tule" in California. 59 1919] 60 University of California Publications in Am. Arch. and Ethn. [Vol. 16 119. Fish (2). W N diket, c teit, s tir, til; C M kinir. 121. Trout. W N sulot, c colat, sw ciya-tir, mol; Md N paliki, s pala; Y N, V, K dalim, T dadim; Mw C lewim, Co tolo, L hul. 122. Fly. W N xilit, lulit (assimilated), SE homota; Md NW amelulu, NE emelulu, 9 emululu (all assimilated); Y v monai, K monoyi, T p 'anue-kai, P p 'onits-kai, B p'anis-gai; C F momua, CL mumuri-x, CR mumura, B mumuri (all assimilated, evidently also onomatopoetic in form). 123. Spider (1). W N kerek; C B tiras-min. 124. Spider (2). Md NE makati, s boko; Y N, V, K medja, T mutea. 125. Louse. W N donus, c dono, SE peris, sw peri; Md NW di, NE pedesi, s pedes; Y N tihite, V, K tihit, T tehet, P tihet, B tahat; Mw s, C, N ketu, P ken, CO, L ket; C B kaxai, M kax. 126. Flea. W N kokus (assimilated), c kok'as, SE tunkuc, sw teoteo (assimi- lated); Md N, S t6kiisi; Y N, V, K bak'il, T bak 'id, P bak'el; Mw s, P, CO, L kuku, C, N kukusu, co kukus; C B, M por. 127. Grasshopper. W N, C nep, sw taram; Md N, S eni; Y N tanuwa-te, v tsanaiwu-c, K danaiwu-tc, T tcaneka-c; Mw s anut. 12& Worm. W N kye-s; Md NW kape, NE kayi, s pep; Y N, K k'eu, K k 'ewa-s-k'ewa-s, K, T gatu. 129. Name. Md N yawi; Y N, V, K, T hoyowo-c. 131. Black (1). W N tcululi (assimilated), c kut 'a, SE tiloki, sw silia; Md NE kol6-m, s kole-m; Mw C, N kululi, P kulu-lu (assimilated); C F colko-te, i eirke-wis. 132. Black (2) (cf. 73, Night). W c molok, SE multi; Y N motsko-n, v moteko, K motki-wik, P motka; Mw Co muluta, L mulu-mulu; C CL murtu-c-mini, Ca murtu-s-min, B hu-mulu, muretu, murtcu. 134. Large. Md s muku; Y N, V matek, v mayek, T met; Mw co omotak; C B mati-li-; and possibly W c komaso, N bohema. 135. Small. W N kuta, SE kucu-n, sw kutci, kuci; Md NW nukti; Y N goli-s, gote-tc, v godi; Mw s teini-miteu, c tuni-tci, P ititi (assimilated) L kucci (bor- rowed from W sw); C . kuteu-wic, CL kucue-min. 136. I. W N, C ni; Md N, S ni; Y, all, na; Y P gen ("my"); Mw, all, kanni; C F, j kana, CR, B kan, B, M ka; W s tcu. There are probably two elements, k- and n-, sometimes combined, sometimes used singly. 137. Thou. W, all, mi; Md, all, mi; Y, all, ma; Mw, all, mi'; C F, J, CL me-ne, B me-n, M me. 138. We. Mw s mahi, C, N, P maci, co mako; C J makin, CL maken, cR maksent, B makse, M mak. 139. This (1). W N, S eu; Md NW, S hede; Y N hi, v hihi, K ke, T xe, P, B Xi. 140. This (2). Md NE uni; Mw s, c, P ne, N nei, co ni; C CL, CR, M ne-pe, B ne-. 141. That (1). Md NW, S hodo; Y N, V ki, v, T ka, K kai, T xi, P XO, ko, B xuntu. 142. That (2). Md NE ani; Mw c, P, co no, N noi; C B nu-, M nu-pi. 143. That (3). W N, S pi; Md NW mo, NE, S mo; C B, M pina, nu-pi, ne-pe (this). 144. Who. W c popi, sw popa (assimilated); Md NW mene, NE homoni, s homoko; Y N, K, T, P, B wat, v watuk; Mw s, C, N mana, P, co manti; C F, CL mato, J mato, CR atu, B ate, M amp. 145. What. W N heke (where); Md N, s hesi; Y N hata, v han-uk, K hanta, T, P han, B han-wil, Mw P hiti, co hinti; C P hinto. 146. Up (1). W N, C olel; Mw s, C, N, CO, L lile; C J rini-mu, B rini. 147. Up (2). W s panti; Md NW hipi-nak, NE hipi-na; Y N tcipin, V, K, T tipin, B tipan; C M tapre. 148. Down. W N, C ken, s tcenti; Md NW kana-nak, NE kanai-na; Y B xoni-u; Mw P wanit; C B winimui, M winum. Dixon-Kroeber: Linguistic Families of California 149. Tomorrow (1) (Cf. 69, Sun). W N hima; Md NW himaduk; Y K, T hiemxac, B huimdja. 150. Tomorrow -(2). W s bina; Md NE bene-ki; Y v bayi-u. 151. One. W N, c kete-m, c kete-t, s ete-ta; Md NW wokte, NE sdti, s wite; Y N, T yet, v yet, K yetc, P, B yit; Mw s, c kefie, P kena-tu, CO, L kenne. 152. Two (1). W N, c pale-t, s pampa-ta (assimilated); Md N pene, s pen; Y N, K punoi, v ponoi, T, B pOnOi, P pufii. 153. Two (2). Mw s, c, N oti-ko, P oyo-ke, co ossa, L otta; C J utsi-n, CL uti-n, CR uthi-n, B utxi-n, M uti-s. 154. Three. W N, C panOL, SE ponoL-ta, sw punuL-ta; Md N, S sapwi; Y N, V, K, T copin, P Cop, B Copi; C J, CR, B kapha-n, CL kapa-n, M kape-s. 155. Five (1) (cf. 28, Hand). W N, C tean-sem, s ete-sem-ta; Md NW ma-tsani; Y N, V yit-sinil, K yit-icnil, T yut-ciniud, P yet-sili, B yit-sufiul, C J, R mi-cur, CL mu-cur. 156. Five (2). Md NE ma-wika, s ma-wik; Mw s ma-hoka, c ma-soka, M mo-soka, P ka-soko. 157. Six. Md N saitsoko, Y N, V, K, B tc'olipi, T tc'udipi, P tc'ilipi; Mw s, C, N temoka, P temepu; C J, CL, CR, M sake-n. 158. Eat (1). Y N, V xat; C M kat. 159. Eat (2). W N, C, S ba; Md N pe, s payel; Mw s, C, N UWU; C F amu-, J, CL, cR amai, C ama. 160. Drink. W SE elu, sw tete (assimilated?); Y N, V, K, T ukun, P okon, B ukon; Mw s, N uhu, c usu, P use, co, I. ussu; C F owato, J uweto, CL weto, CR uit, B uki-si, M uki-s. 162. Dance (1). W N tcuna, c ken-tcon, SE tono, sw ton, tolo; Md NW kamini; Y N, K, T ka 'im; Mw s kalane, c ka]anu, N kalte, co kawul; C CL to-ken, CR, B tcite, M tcit. 163. Dance (2). Md NE wete-mdo; Y N Woti, T woti-yot, P wodoy, B wutei. 164. Sing. W SE muhi, sw muhu, s muhe; Mw s, N mUli; C miili. 165. Sleep (1). W N xina, c keno, s k'ana; Y P k'eneu ("lie" in other Yokuts dialects). 166. Sleep (2). Y K entim, T efitam, B ifiati; Mw co etc, L ets; C J ete, CL etini, CR eteen, B eten, M etn. 167. See. W SE tc's, sw teowi; Md N tse; Y v cil, P dileu. 168. Kill. W N Loma, s limo; C J, CL, CR, B, M nimi. 169. Give. Md NW me, NE, S mei; Y N, V, K wan, V, T waki, P wa 'en, B wa'an; Mw co wa. 170. Laugh. Y N, K, T hai, v, B haya, P hayau; Mw co hawai. 171. Cry, Mourn. W N watca, SE wati, sw watu; Md NW won, NE wakdo; Y N, V, K wahil. 6. Woman, tGirl (1) 1Tr N pO k -ta s po ki -ta sw 0 ki -ta Md N, S ku le Y N, v mu k'e la T mu k'a -c B mu k'e -s Mw co ku leyi L pO tsi C CR tmu xa -c B mu ku r -ma 61 1919] 62 University of California Publications in Am. Arch. and Ethn. [Vol. 16 14.' Ear (1) tu k to k su k -a'l to 1 ko to ko -su a lo k co lo to tu k -sus e tco o tco tu x -s 26. Neck do ki do lu tu ku -tuku koyi kuyi kii -sok mi ki -c mi k'i -c mu k'a -c lo ba to pa he le ke he le ki la nai ra nai ha r ko ka tc k 31. .Nails kaha i k'a ! tc'a i tca i tsi' bi bi tsi go tco yi ti pI tci tu r tu r tu r (Metathesis) -c (Borrowed?) -em -is 37. Foot, tLeg (1) wo to n wu to fl ta na -hi Y N, V, T K, P B Mw s C, N P, CO, L p C J, CL CR B M W N C Md NW NE S Y N V, K T Mw c N, P CO L C F J, CL CR M W N C BE SW Md NW NE, S Y B Mw P. L CO' C J, CR CL B . Y K T B Dixon-Kroeber: Linguistic Families of California ttu fnu ttu na ha te ha ta ha ta 42. Blood (1) Y N, V K, B T p C F, i, CL,CR J M Mw s, P, CO, L C, N W N C SE SW Md N S Y N, T, P v B K Mw Co, L C J, M W N C S Md NW NE S *Y N, V T p B Mw c p CO, L C J CL pa ya x he pa ho pa hi ba pa ya n pa tia n pa tea n ki tca u ki tca wu 44. Liver ki la te L. tci la tce la ku la ku la di p di ' p di 'x p da la pi ku Ila Si ne -C (Borrowed?) 59. Bow ku lu 1 (Assimilation) ku I sa k nu n (Assimilation) pa n da pa n da ka pa n da k da la p da li p da. ya p ge ts 4i -I ga tsi fii -1 so llo ku ta nu ka ko no co no k ta nl ka -m Mw s, c, N p B, C, N C CL CR -m -c 1919] 63 64 University of California Publications in Am. Arch. and Ethn. [Vol. 16 72. Star LU yu k La ra k tsa tu -s ti la -men ta ti -men to ti -mel lu Ilu 10 16 ts'a ya ta -s tca i ta -c tca i to -c tc'o i to -c tso yo ti -s tso yo to -s tca la tu to le 77. Rain lu he lu ha yu ru nu ka u mu tca ho ma u pa aa nau 'a mne a mani i nam i nlam 80. Smoke W N nu k c no k SE mo ko SE ma ka sw no lo Md N Su ku s su k Y V, K, T, B mo da k p mu t ka Mw s ha ki C, N ha ki And possibly: p ka li CO, L ka I (Assimilation) (Assimilation) W N, C SE N SE SW SW Md NW NE Y N V K T p B Mw s L W N C SW Mw s, c, N S p CO, L C F J, CL CR M M -sa -su 1919] Dixon-Kroeber: Linguistic Familie8 of California 65 85. Earth W SE wi la k Md N ka wi Y T ha' pi 1 p ha ma n Mw s, C, N,P, L wa Ili C F, J, CL wa re CR, M pi re .B mu n 104. Bear W N, C tsi L SE tcu ku i SE si ka SW ci la i Md N mo de Y N, V U lu i N U yu m v mo lo i K do noho p p t'e ne i Mw co, I. ku le C J, CL, CR, B, M 0 re s 106. Puma W N pa ti t SE wo k'e n sw pa te Md N pe ku ni Y N, V, K we he ci t T WO ho ci t p wi hi ci t B WO ho sa t Mw co upu k so L upu k se C M he k s 120. Salmon W N, c nu t N nu r s hu r SE hu I C F tci ri k CL tci pa 1 (Metathesis) CR hu ra k B u ra ka M u ra k 66 University of California Publications in Am. Arch. and Ethn. [Vol. 16 130. White W N ka ya N xa ye C LU yu ke SE tc'a m SE SR m Sw tca ma sw tea lo ki Md NVw eda. hi NE ko ko -pem (Assimilation) 8 ko Y N, V djo lo 1 (A.ssimilation) K djo ye yi (Assimilation) K do lo li (Assimilation) T tco do d (Assimilation) P teu fna n B tea wa fi MW C, N ke le li (Assimilation) C B tea 1 ka B pa 1 ka M pa xe la -St (Metathesis) 1,33. Red W N te di c te de ki BE wa ra k sw tu lu ka Md NW e la la k -no (Assimilation) NE lak la k -pe (Assimilation) Y N tco yi ki -n v go li k -lai N ha bi 1 kai v xa bi 1 wai v ba tsi gi -n K ha ba ka T- ba dji ki -n p ho lo ts gai (Metathesis) B xa' wa 1 Mw C, N we te ti (Assimilation) p w-fw te te (Assimilation) co u lu ta C J pu 1 . te -wis M yu r tsi -st 161. Run W s wi ti li Md NW ye wo NE wi le no Y K la wi t T da wi t Mw S. N hu wa te c ku wa tu co hic wa te Dixon-Kroeber: Linguistic Families of California Other Cognates As an example of how readily comparative material accumulates as soon as the formal comparative tables are abandoned on which the preceding list of resemblances is based, some kinship terms may be cited.'6 These will be found on the following page. To those familiar with the consistency with which the designations of the primary kinships run through the Indo-Germanic languages, this list may seem feeble evidence. But among the Indians of the Californian region no such conservatism exists. Within different parts of the same stock as formerly recognized the plan of kinship reckoning sometimes differs profoundly; and terms are often utterly different in form and meaning. Such variations, for instance, occur in Shoshonean,17 a stock which philologically is compact and has never been considered anything but a unit. By comparison the foregoing Wintun, Yokuts, Miwok, and Costanoan terms of relationship show all the connection that could be expected. The following are also significant: Maidu (s) North South East, up-stream West, down-stream tocim komo noto tai Yokuts (T) xucim (N, toxil) xomot not datu There are a number of parallels in verb stems, outside of the mate- rial that has already been reviewed. YokUts (T) 18 wejifi, not find yom, cuckold mok, swallow danl (d ex 1), hear dadik, hang bax, fear t 'of, drown dMtc, be cold to, throw at kem, embrace tcuyo, urinate Costanoan (B) 19 waian, miss, err yume, deceive, cheat muiku, swallow without chewing nam, nanm, hear, listen to laki, hang poiko-ni, frighten titci, drown tur-si, be cold tu, strike in eye kapal, embrace teala, urinate 16 Present series, xi, 437, 1916; xii, 172, 1916; xii, 352, 368, 1917. 17 Ibid., xii, 367, 1912. 18 Ibid., II, 246, 1907. 19 Ibid., xi, 441, 1916. 1919] 67 68 University of California Publications in Am. Arch. and Ethn. [Vol. 16 pq ~ ~ ~ ~ ~ _ c a c a S E-4 ~ ~- -0 0 z -4 0 n - 0 -e 4- a fi .m S a ci aa3 a~~~~~~~~a E 4z ce ce~~~~~~~ co 0~~~~~ p-4 4-4C ~a a * a a *i a : rc = a a a 2 ev .ca Q a a a ?=toS Dixon-Kroeber: Linguistic Families of California Maidu (NE) 20 kot, kut, divide, ket, graze hi, smell kai, fly pedi, steal a, say to, burn not, bend Yokuts (T) 18 day, step, kick dadik, hang dan, hear ko, hit with hand tumi, throw taw, pay teaw, tcanl, shout xapit, pull out Maidu (NE) 20 huk, whistle tap, squeeze tus, break tse, see heno, w6no, die Yokuts (T) 18 k6t, tear hon, smell (Mw, hukaye, smell) hay, fly mud, cheat (Mw, motca, hide) ah, cry tud, burn nokum, bend Miwok (N) 21 tekme, kick yotki, hang toloye, hear kute, knock with hand dobe, throw tiwa, buty kaunie, shout k6pa, pull Miwok (N) 21 kuyage, whistle tupi, press taswa, break ete, see yina, yunu, kill CHANGES OTHER THAN SOUND SHIFTS Before the sound equivalences of Penutian can be examined advan- tageously the scope of several disturbing factors must be considered. These are: expansions, contractions, assimilation, metathesis, analog- izing, and reduplication. Expansions Expansions of one syllable to two occur in all the Penutian lan- guages, but are uncommon in the two western ones, Wintun and Costanoan. Yokuts is decidedly the most given to the practice, more than a tenth of its words, if the list here dealt with is representative, manifesting an initial increment. Since no Penutian language pos- sesses prefixes, except for a few semi-etymological elements preposed to the Maidu verb, these increments are of phonetic origin. Most of them consist of a mere vowel, or a vowel preceded by h-. which occa- sionally is replaced by x- or k-. 20Bur. Am. Ethn., Bull. XL, 723, 1910. 21 Present series, Ix, 288, 1911. 1,919] 69 70 University of California Publications in Am. Arch. and Ethn. [Vol. 16 Thus, wood, Y hi-t'ec, W tcus, tok' (for t'ok?), Md tsa, Mw susu (by assimilation). Again, ash, Y ha-pac, W puk, put, Md pupu (assimilated); water, Y i-lik, Mw kik (assimilated), C si; boat, Y o-won, Mw wote, C wali, walin; blood, Y K, B he-pa, Y N, V payax, C payan, Mw kitcau; horn, Y i-cel, W teili, k'ili. In most of these instances the vowel or syllable added initially by Yokuts is com- pensated for by a vowel, more rarely by a syllable, lost at the end of the word. The operative process is therefore similar to that which, within a single Yokuts dialect, produces entim-ji and inietm-ac; amd-in and amad-ad; amat-ji and aut-ud; tc6nic and tcanj-in; wuiktcamni and wiikatemina; naxamic, naxmica, and naxamca; onmid, unimda, onemadi, and onimidi. These changes of quality and position of vowel within Yokuts are in many cases accompanied by a shift of accent, which indeed is probably the primary cause of most of the vocalic mutations. Now it is significant that in the cases cited, and apparently generally in words to which Yokuts prefixes a syllable as compared with the other Penutian languages, the added element is not a slurred approach but the stressed part of the word. Compare i'-cel with Wintun tci'li; and, between Yokuts dialects, he'-pa and pa'yan. There can be little doubt, accordingly, that the numerous increments of Yokuts are really such and are connected with a set of inclinations otherwise manifest within the language. More difficult to understand are a few cases in which Yokuts possesses an initial syllable with a dental stop lacking in other lan- guages; as, buckeye, Y do-pin, du-pun, Md polo-m-tsa, W pasa; and jackrabbit, Y do-pol, ca-pal, tc'a-bakiu, W patkili, Md (rabbit) palal, Mw e-pali, e-plali, a-ule. But here again Yokuts loses at the end the equivalent of the syllable it adds at the beginning. Of somewhat different type is an occasional Yokuts expansion of the vowel of a first syllable into two. Thus, girl, Y guyoyam, kaina, Md konai, Mw koya, kola, C kata; boy, Y P kohutem, N, V, K, T notco, C kotco; louse, Y tihit, Md di, W donu, Mw ken, ketu, kaxai. Other cases of Yokuts increments will be found in 53, 85, 110, 129, 133. It should be added that Yokuts increments are usually confined strictly to that language, whereas when other languages show an addi- tion this more frequently reappears in a second tongue also. To be specific: thirteen of sixteen Yokuts increments are peculiar, but only eleven out of an aggregate of nineteen in the four other languages.' Miwok increments appear in stems 26, 68, 106, 118, 134, 135, 102, Diwxo-Kroeber: Linguistic Families of California 109, and perhaps 110. They consist of a vowel, usually without h-. All but the last three examples are confined to the Plains, Coast, and Lake dialects. The three Sierra dialects, although geographically between Yokuts and Maidu, appear to entertain an aversion to initial expansions. Maidu expansions are discernible in 85, 117, 122, 130, 133, 144. All but the last appear in the Northwestern dialect; the Northeastern and Southern participate less regularly. Wintun increments are probable in 56, 122, 146, and Costanoan in 26, 101, and perhaps 118. Contractions and Reductions Loss of the initial syllable by contra-ction or outright apocope is sometimes difficult to distinguish from an initial increment, when a stem is traceable in only two or three languages. There are however undoubted cases, like fish, Mw P pu, s lapisai, co lota, Y lopit; grass, brush, C cR wai, F yawa, Y yawil; sun, W sas, con, Y P hicta, Mw hiema, hi, C iemen; and bow, W s nun, N kulul, Mw c solloku, etc., as shown in the columnar presentation of stem 59. Other certain or probable cases are 6, 26, 28, 31, 33, 37, 46, 58, 67, 73( ?), 77, 78, 79, 80, 96, 112, 121, 127, 147, 154, 159, 164. The uncertainty whether any given form represents an apocope in one language or an increment in another is compensated for by the certainty that the tendency of each language in the two opposing directions is in inverse ratio. Yokuts, with most increments, has fewest losses. Wintun adds rarely, but lops off most frequently. Costanoan alone departs rather uncommonly from the Penutian average by either expanding or reducing the head of the word, and may, therefore, be regarded as preserving intact the initial syllable of the original Penutian stems more frequently than any other lan- guage. Thus: Increment Apocope Total Net changes Wintun .... 12 3 15 -9 Maidu -9 6 15 -3 Yokuts -3 16 19 +13 Miwok 9 8 17 -1 Costanoan -4 3 7 -1 In summary, so far as the first syllable of stems is concerned, Wintun tends most to shorten and Yokuts to lengthen or preserve its words. 1919] 71 72 University of California Publications in Am. Arch. and Ethn. [Vol. 16 The same inclinations appear in medial and final syllables. Wintun again cuts them away; Yokuts maintains them almost without ex- ception. Maidu lops many words at the end, but almost invariably preserves the middle one of three syllables, however much altered. Examples of the loss of such syllables, or their consonant, so far as is possible to determine in the incomplete knowledge of Penutian suffixes, are: Medial Syllable Smoke, W nuk, moko, Y modak, Md suk, Mw haki-su Breasts, W imit, Mw musu, Y minitc, menit, Md nuni Night, W ken-, leni, sinol, Y toyon Flea, C por, Y bak 'il, W tunkuc, kok 'ac, Mw kuku Puma, W patit, Y wehecit, Mw upukse, Md pekuni See also 14, 33, 34, 36, 44, 49, 57, 65, 71, 100, 114, 118. Final Syllable Teeth, W si, Md tsiki, Y teli, C sit, Mw kutu Hand, C puts, Y putoni, Mw (arm) patean Blood, Md sede, W dedeki, sak (ex s-d-k) Faeces, Md piti, Y pidik Sinew, Md paka, Y pikil Pestle, Md suni, W satak, taki (ex s-t-k) Cloud, W N, c k 'a, s k 'ira, Y k'ilei, Mw ya-kal Mountain, W tCoL, toL, Y dulul Mountain, Mw leme, Y lomit, Md yamani Grass, C yawa, Y yawil Sugar Pine, W, Md sumu, Y tconoxi-c Manzanita, Md iipii, Y aptu, W paka, pai Fox, W, Md hau, Y au 'dja Salmon, W nur, hur, hul, C hurak, uraka See also 22, 23, 24, 30, 32, 37, 38, 48, 50, 68, 69, 72, 79, 85, 87, 100, 101, 117, 128, 130, 131, 133, 167, and perhaps 44 and 57. It is possible that some of these forms are better explained bv the addition of a suffix in some languages-as perhaps -I in Yokuts-than by the loss of a stem syllable in others. But the variety of final syllables that appear and disappear is too great to allow this explana- tion to be extended to more than a fraction of the cited cases. The tendency 'toward apparent loss of the characteristic element of any syllable of the stem is summarized in these figures: Initial Medial Final syllable syllable syllable Total Wintun -. 12 9 13 34 Maidu - .--9 2 14 25 Yokuts -3 3 1 7 Miwok -9 5 8 22 Costanoan -4 3 4 11 Total -- - - 37 22 40 99 Dixon-Kroeber: Linguistic Families of California Metathesis Outright interchange of the consonants of two successive syllables is not infrequent, especially in Maidu. Yokuts is again most con- servative. Old man, W sw tciak, c kias, N kiemi-la, Md NW kano, Mw s humele-tki, possibly Y moxelo. Tongue, Mw c, P nepit, N nepitu, L letip, s, Co lemtep, c lase, etc. Beard, Md s masau, N simpani, W s sep-sep, Y damut, djamoc. Nails, Md NW tsi 'bi, NE, S bitsi, Y B gotcoyi-c, Mw co pitci. Belly, C M pitin, Mw L puluk, Y N balik, W sw bus. Knee, Md N pokosi, s pokok, W puiyak, pomok, Y opuk, kuyut.. Skin, Md potom, potani, posala, (ex t-p-l or t-p-n), Y daplan, tapia, p'alai, totop, etc. Pipe, W topor, dopo, C torepa, etc. Lightning, Md wipili, wispil (ex w-l-p), W waluka, Y walma, C wilka, wilep. Buzzard, Mw N, P tcuhu, c husu, L hus, Y hot, hots, Md h6si, W hus. Frog, Md NE, s wakat, W watak, Y watak-saai, Mw wataku-na, C wakatse-m. Salmon, C CL tcipal, F teirik, CR hurak, W hul, nur, etc. Red, Y P holotsgai, N habilkai, v xabilwai, N tcoyikin, W sw tuluka, c tedeki. As between sleep, Y entim, iniati, C eten, etn, metathesis is probable, but in which direction is difficult to decide. Assimilation Assimilation of consonants is about as frequent as metathesis. It is restricted almost wholly to sounds of successive syllables. Juxtaposed consonants are confined to the interior of words in Penutian, are rather uncommon even there, and apparently are less inclined to modify each other when in direct contact than when separated by a vowel.22 22 All Penutian language show a few cases of consonant collocations of the type nt or mb which seem to be secondary expansions of an original nasal or original stop. Thus, Mw s huntu, eye, Md, C, hin; Md N simpani, beard, Y damut, djamoc, W s sep-sep. But this process is of sporadic occurrence. In general, juxtapositions are due to two sets of causes. The first is the trans- position of a vowel, or its loss; as breasts, Y, v, T, B minit, K minteix; smoke, Y, V, K, T, B modak, P mutka; sleep, Y B iilati, K entim, C B eten, M etn; black, C F colko-te, W SE tiloki. Such organic juxtapositions, or modifications of the pure stem, are abundant iii the two southern Penutian lanuages, but rare in Wintun. The second kind of consonant collocation is due to word building, by suffix, composition, or reduplication. Thus, knee, C M dol-s, Mw L tokollo; what, Y B han-wil, T, P han; beard, W s sep-sep. Juxtaposition of consonants due to these causes appears to be about as frequent in the Penutian languages as that of the preceding class, except in Yokuts, which is in general most given to vowel shiftings and mutations but has few suffixes and almost no compounds. Initial consonant combinations are not tolerated by any Penutian language. A few contrary examples in Costanoan, like CK "gruwam," house, and "trepam," pipe, are almost certainly faulty orthographies, as shown by the cognate forms ruwa and torepa in near-by dialects. Costanoan, and this is true of several of its dialects, is the only Penutian language that permits final combinations: Monterey evinces the trait most markedly. The "doubled" consonants of Miwok, as in kanni, 1, awannata, turtle, cappa, tule, walli, earth, 1919] 7:3 74 University of California Publications in Am. Arch. and Ethn. [Vol. 16 Tongue, Md s alili, Y talxat, a 'lada-s. Shoulder, Md dadaka, Y gapeal, gapasai. Breasts, Md NW nuni, Y minitc, etc. Knife, Mw L tsitsa, C tcipa, tip, Md tsami. Ash, Md! NE pUpU, W put, puk, Y hapac. Fly, W N lulit, Md NW amelulu, C CR mumura, W SE homota, Y monoyi, p 'anuck 'ai. White, Md NE koko-pem, Y djolol, djoyeyi, dololi, Mw C, N keleli, W c Luyuke, sw tcaloki, Md NW edali, C B tcalka. See also 43, 59, 72, 83, 109, 110, 126, 128, 131, 133, 135, 144, 152, 160, and perhaps 90. The number of instances in the vocabularies used is Maidu 11, Miwok 8, Wintun 8, Yokuts 2, Costanoan 1. Analogizing and Reduplication Analogizing has obviously affected the numerals, pronouns, and demonstratives in several Penutian languages. Whether its influence may be traced to any notable extent in nouns and verbs can only be ascertained through more intensive study. Reduplication of any sort is rather rare in Penutian. There are a few cases of its occurrence in stems: Yokuts hon-hon, heart, p'ia- piya, cloud, tiwi-tiw, rattlesnake, k'ewas- k'ewas, torm, Wintun sep- sep, beard, tcil-tcil, bird, pat-pat, tule, Maidu pok-poko, tule, Miwok mulu-mulu, black, possibly Yokuts nono, man, caca, eye. Maidu may have more cases, but they do not occur in the foregoing list. Summary All in all, it appears that in stem modifications due to other causes than sound shifts Costanoan and Yokuts are most conservative of original Penutian form, though the latter inclines somewhat to elabor- ation; that Wintun and Maidu are most given to reductions of conse- quence; and that Maidu shows the largest number of other distortions. SOUND EQUIVALENCES Stops All known dialects of the three "Pen" languages possess three classes of stops: surds; sonants; and glottalized surds. The surds are less aspirated than in American English, perhaps, or the sonants are proved by the non-Miwok cognates to be mere lengthenings of originally simple consonants, apparently under the influence of the stress accent 'falling on a short vowel immediately preceding. There is no clear case of one of these long Miwok consonants springing from assimilation of two aistinct sounds brought into contact. Dixon-Kroeber: Linguistic Families of California 7 more briefly voiced: at any rate the two classes are not infrequently confounded in orthography. The glottalized stops are produced rather feebly, on the whole; enough, at any rate, for the application of the term "fortes" to them to be misleading. These sounds have also been recorded with considerable inconsistency. In both "Uti" languages, the three types of stops have come to coincide. At least, the data are so uniform on this point that if differences in voicing, breath, or glottalization survive in the stops of these two languages, they must be near the threshold of perceptibility.23 So far as can be judged, under the orthographical difficulties indi- cated, the three classes of stops have shifted rather than maintained themselves consistently in the Pen languages. This is particularly clear for the sonants. The only apparent case of a voiced stop per- sisting in dialects of all three languages is 150, tomorrow, W s bina, Md NE bene-ki, Y v bayi-u. Maidu sonant stops seem to be generally represented by nasals in Wintun, and by either nasals or surd stops in Yokuts; 1 and y count as equivalents of n in this connection, so far as origin of Maidu d is concerned. Compare the regular conversion of generic Yokuts 1 to d in the Tule-Kaweah group. W Md Y 15 m b 107 n d 1 n d y 59 1. n, (s) nd 1, y, (fi) 125 n, r d t 104 1, (L, k) d 1, y, n 68 b p 53 b p Yokuts sonant stops remain stops but have more frequently been written as surds in Wintun and regularly so in Maidu. The Miwok and Costanoan equivalents are usually stops. W Md Y 124 k dj, tc 8 k g 66 b p b 11 d tc d, t 25 b 'p. w g, w C: w 33 b b,- 135 k k g 23 One of the authors has recorded sonant stops in Northern Miwok (present series, Ix, 279, 1911) and in Costanoan (ibid., Ix, 250, 1910), but the incon- sistency of his orthography leads him to doubt whether there is more than one class of stops in either language. 1919] .75 76 University of California Publications in Am. Arch. and Ethn. [Vol. 16 126 59 121 133 24 110 86 108 k, t, tc k (s, m) k tc, s tc t, tc k t p p k s b g, d d g, k, w dj, d d, tc', t d xg Mw: t, h Glottalized stops have probably been recorded even less depend- ably than voiced ones. Half or more of those written in Yokuts ap- pear also in the corresponding Wintun words. Where there is dis- crepancy, the orthography may be at fault-the Wintun data being from a variety of observers. Maidu presents fewer parallels, and these show either unglottalized surd stops or sounds other than stops. In fact, the Maidu vocabularies utilized happen to reveal almost no instances of glottalized sounds, although the class is well established for the language. Miwok and Costanoan normally have surd stops where Yokuts or Wintun have glottalized ones. Md Y ts V k' g, k' k' k k', g k k' k k' tc', ts ts tc26 k h k', k s tc' m p', m t' tc' Mw S k k k k, p tc, t tc, t k k C k t k t s m t, t (tc) t, t, tc The known relations of Penutian stops can be provisionally and incompletely summarized as follows, labials being used as examples: w m b p P' Md b b p p y m p b b p' Mw m p p p p C m p p p 24 tc'ok, t 'ok, conjecturally reconstructed from recorded tcok', tok'. 25 ky for k' I 26 Buena Vista dialect only. w tc', t'24 k' g, k kyk' ky 25 k, k' k ts, t, L k', tc', k, tc k' s m 90 75 98 165 128 126 26 72 31 106 33 157 122 61 41 Dixon-Kroeber: Linguistit Famities of California Palatal T Series Original Penutian seems to have possessed two series of t sounds: one produced against the teeth, the other against the gums or anterior palate, probably with an arched tongue. These two sets of sounds persist in Yokuts and Costanoan, where they are readily distinguished, the posterior t, rendered t in this paper, having been variously written tr, thr, ths, tc, and the like. In Miwok the two series appear to be still distinguishable, especially in the Coast and Lake dialects; but in the Sierra idioms they are so nearly alike that their orthographical sepa- ration has either not been attempted or has been accomplished with obvious inconsistency. For Maidu and Wintun there are no reports of t sounds in more than one position. That the distinction of t and t is original in Penutian, in spite of its absence from the two northern languages, is indicated by several circumstances: First, both sounds coexist in the majority of Penutian tongues; second, these languages belong to both the Pen and the Uti branches; third, t several times appears in the same stems in Yokuts and Costanoan: for instance, 11a, 12 144, 166, and perhaps 132; finally, the equivalents of t in the three northern languages are other sounds more frequently than t, especially s. Y t=Md s, 18, 24, 80, 111, 112, 115; h, 147; s, t, 46; ts, 21; k, 34; 1, n, d, P, etc., 11, lia, 27, 59. Y t =Mw s, 32, 59, 111, 118; h, 80, 115; tc, 29, 58, 166; k, 21, 59; n, lla. Y t C s, 21; s, t, 59; ts, 29; t (), 147. Y t =W s, 21, 111; ts, 24; k, 29; L 18; 1, n, 11, lla; -, 80, 147; t, 24, 32, 112, 115. It thus appears that t is represented by t in Yokuts, t or some- times s in Costanoan, s in Maidu, s, h, or tc in Miwok, and t, s, or other sounds in Wintun. P and K Equivalences between p and k, or between labials and palatals, are common, and rather striking in view of the complete absence of labialized palatals of the kw type from all existing Penutian forms of speech. That p and k do positively correspond is clear from the number of instances in which they substitute interdialectically within a language. Each of the Penutian languages presents such cases. 1919] 77 78 University of California Publications in Am. Arch. and Ethn. [Vol. 16 w k, p k, p k k' p k p, d k, tc, t kpm p, m Md Y mp, m m k,p m k', p' p, b p, k k k' P,- k p p p,d t t b k,- p,- w, k Mw k, p p, w k, p k, p k, p k k k C k, p k,- k k p k p p k, x, k,- k There is no great regularity; but on the basis of prevailing type of sound, the labial and palatal would seem to have been original one about as often as the other; while Yokuts, and next to it Miwok, to have been most given to alteration. S and K A very apparent sound equivalence is between s and k. This holds good between Miwok and Costanoan in both directions, and again between Yokuts and Wintun, so that it may well be regarded as a fundamental trait of Penutian phonology. Md Y Costanoan k, Miwok s k, ts k Costanoan s, Miwok k k k k s Yokuts s, Wintun k s - s s s s, x s Various s S Mw C s s s k k k k k k k, t s s s s, tc s k s s s, 24 99 76 75 109 26 110 157 42 125 126 23 59 127 w . 9 138 160 19 28 83 101 20 51 82 81 88 90 154 123 4 1 k k k', tc k k, t k, t, s k, s k k', tc, s Dixon-Kroeber: Linguistic Families of California The s is often, in Yokuts, the modified sound written c; and, espe- cially in Miwok, alters dialectically to h; but the shift between k and s seems to have occurred through t or tc rather than h or x. The first two groups in the list give the impression of k being the original sound, the third of s; but the ind.ications remain to be substantiated. If they are correct, Maidu perhaps preserves the original sounds throughout. L and K A curious correspondence, both within Wintun and between it and other languages, is that of k and 1. Explanatory transitions are lack- ing. The 1 occurs on both sides. Smoke, W N nuk, c nok, SE moko, sw nolo, Y modak. K is original. Bear, W SE tcukui, sika, sw cilai, N, C tSiL, Y ului, moloi, t 'enei, Mw kule, C ores. L is original. Drink, W N elu, Y ukun, Mw uhu, usu. Sing, W s muhi, Mw muli, may be cited in this connection. Fish, W N diket, c teit, s tir, til, C M kinir, is not wholly certain. Neck, W N doki, c dolu, s tuku-tuku, is also open to some doubt, on account of forms like Mw co heleke. Water, Y ilik, Mw kik; white, Y djolol, W sw tcaloki; black, W N tcululi, Mw kululi, W SE tiloki, C F colko-te, are more probably instances of assimila- tion; although assimilation, in most languages, generally occurs between sounds having one or more qualities in common; so that even these cases might indi- cate that Penutian felt I and k to be less far apart than we feel them to be. Surd L Surd L occurs in all Wintun dialects, but in no other Penutian language. Records sometimes indicate it as affricative; but it is known to be often a continuant, and may be always so. According to the available data the occurrence of the sound is rather irregular, each dialect in turn showing it in words in which the others have sonant 1 or other correspondences. Thus, koL, kol, mouth; teLi, teela, kila, liver; tsiL, silai, sika, bear; Loma, limo, kill; in which the surd 1 is confined respectively to SE, C, N and c, N. The predominating equivalent in Penutian is 1: 44, 86, 104, 116. Sometimes this 1 is perhaps a secondary development from n or from m through n: 20, 168. A second equivalence that seems fairly well established is with s (=h), tc, t. Compare 16, 18, 72, 130, and perhaps 99. A supernumerary final -L in 89, 105, 154 suggests a Wintun suffix. Thus, 105, W N wima, c wemaL, Md pano. 1919] 79 80 University of California Publications in Am. Arch.. and Ethn. [Vol. 16 R All five Penutian languages have 1, but r is found only in the two western ones, Wintun and Costanoan. As one of these is of the Pen and the other of the Uti branch, the question is fair whether r may not be an original Penutian sound distinct from 1, though now merged with it in the three eastern tongues. Among the stems considered, r is common to Wintun and Costanoan in scarcely half a dozen. Two are correspondences confined, so far as known, to Northern Wintun and San Juan Bautista Costanoan- extreme dialects several hundred miles apart: 113, crane, toriha-s, ture-s; 123, spider, kerek, tiras-min. Of these the latter must be accepted with reserve. In two other stems Wintun varies between t, r, and 1, Costanoan has r: 119, fish, W diket, tcit, tir, til, C kinir; 120, salmon, W, nut, nur, hur, hul, C, tcinik, hurak, uraka, urak, also tcipol, possibly by metathesis for tcilap ex tcilak ex tcirak. A last case shows both languages varying between r and 1; 133, red, W, tedi, tedeki (cf. d =r below), warak, tuluka, C, pulte-wis, yurts-ist. Stems in which Wintun alone shows r are also not very numerous: 4, boy, W, r, n; Mw, 1, n; C, 1, Di. 40, heart, W, r, d; C, n, ny. 72, star, W, r, 1, y, -; Md, 1; Y, y, i; Mw, 1. 75, cloud, W, r, -; Y, 1, y, d; Mw, 1. 77, rain, W, r, 1; Mw, n, m, p; C, n, mn, nm, nl. 125, louse, W, r, n; Md, d, -; Y, t; Mw, t, n; C, x. It appears that Wintun r corresponds to n more frequently than to 1, both within and without that language; and that d, y and t are also equivalents. Costanoan has r more frequently. W Md Y Mw C 2 man t,t,xy w, y, i r 6 woman - 1 1, 1r, hla head n n t n r, t 26 neck d, t m 1, t r, 1, tc 31 nail i y r 38 foot 1 l,y r,1 44 liver 1, L 1 1,',- 1 r 57 house k' h r 65 pipe d, t t tr, tor 71 moon r, 1 85 earth 1 1,n 1 r, n 104 bear 1, L, k d 1,n, y 1 r 110 bird 1 n, (tc) m, (tc) r 122 fly t y, c, ts r,- Dixon-Kroeber: Linguistic Families of California 126 flea s,- s 1, d s,- r 131 black 1, t' 1 1 r,l 132 black, night 1 t, tc 1 r 146 up 1 1 r 147 up n n n r 155 five m n n,fn,1 r The equivalent is 1 most frequently in Miwok, rather less often than other sounds in Wintun and Maidu, and uncommonly in Yokuts. The next most usual correspondences are n, d, y, and t. There is nothing in this evidence to militate specifically against the assumption of r as an original Penutian sound, but very little to support such a view. If secondary, r has clearly developed from both l aind n, and has probably realtered to both these sounds as well as to d. San Francisco Costanoan usually, but not always, replaces r of other Costanoan dialects by 1. The Palatal Nasal The palatal nasal i occurs in the three Sierra dialects of Miwok (S, C, N), and the three southerly hill dialects of Yokuts (T, P, B).27 The regular equivalent in other Miwok and Yokuts dialects is n. The in dialects of Miwok and Yokuts are not in contact, but separated by a tract of Northern and Kings River Yokuts dialects. All the Penu- tian n dialects border on Shoshonean speech, which generally pos- sesses in. As in the case of mixed vowels, an alien origin for the sound is therefore indicated. It is of course conceivable that ni is an ancient Penutian sound which happens to have become altered to n in all except a few dialects; but this is unlikely because the Miwok and the Yokuts n- rarely occur in the same stems. Miwok tuiiu, leg, Yokuts wutofi, foot, is the only instance in the comparative vocabulary used. On the other hand, we have 29, Mw c wonotu, s patcan, Y T putoln. Moreover, ni corresponds to other sounds than n in Wintun, Maidu, and Costanoan: 21 Y taiii, teyi, teli, Md tsiki, Mw kutu, C sit 151 Mw kene, kenne, W kete, Md wite, Y yet. 155 Y -cinud, -sinil, W -sem, Md -tsani, C -cur Finally, -i is unknown in Miwok and distinctly rare in Yokuts in initial position; which fact further argues against its being an original sound. 27 The sound in occurs in Maidu, but usually through modification of m by following k, and apparently never organically. 1919] 81 82 University of California Publioations in Am. Arch. and Ethn. [Vol. 16 It appears, therefore, that ni developed in Miwok and Yokuts through an analogous stimulus, probably acquaintance with Sho- shonean; that this influence operated independently in the two lan- guages; and in the main affected different words. The factors that determined under which conditions original n or other sounds became altered to ni or remained unchanged are wholly unknown. Mixed Vowels With the exact consonant equivalences of the Penutian tongues so little uncovered, it is hardly profitable to compare the vowels in general. There is, however, one class whose distribution makes an inquiry into their origin desirable. These are the unrounded mixed vowels usually written ii and o, although in the Miwok sources chiefly drawn upon in this paper they appear as u' or u. These vowels have not been.heard in Wintun and Costanoan, in Lake Miwok, nor in Northern, Valley, and Kings River Yokuts. They are restricted to Maidu, to the four interior dialects of Miwok, and to two of the.four hill dialect groups of Yokuts. Now these Penutian dialects are all in the Sierra Navada, across whose crest they adjoin, for the most part, on several Shoshonean dialects-Northern Paiute, Mono, and Kern River-which contain a mixed vowel also produced with unrounded lips. This Shoshonean vowel, usually represented by the character ii or i, has recently been shown by Dr. E. Sapir to be the normal Shoshonean equivalent of Uto-Aztekan e, which is wanting from the dialects in which the ii or i appears. No other language in the Cali- fornian region is known to possess sounds of the same class, except Hokan Chumash, which is also in contact with Shoshonean, although with more southerly dialects than those enumerated. There is thus a high prima facie probability that the ii vowel or vowels of Shoshonean and of the easternmost Penutian dialects and of the one Hokan language have had a single, common origin, their spread being an instance of diffusion from one to another or two other unrelated linguistic families. The sound occurring more widely in Uto-Aztekan than in Penutian or Hokan, its source in that family is suggested. The query now arises, how far intra-Penutian comparisons bear out these indications. To begin with, Penutian u and o correspond to so many other vowels, and Uto-Aztekan u or i so regularly to e, that the probability of their origin in the latter family is strongly confirmed. Dixon-Kroeber: Linguistic Families of California The Penutian ui-dialects, however, lie in two groups: a larger, com- prising the overwhelming bulk of Maidu and Miwok speech; and a smaller, consisting of a minority of the Yokuts dialects. As these two groups are geographically discrete, they are best considered separately. In Yokuts ii and 6 occur most frequently in the Tule-Kaweah dialects, in which they are almost uniformly the equivalents of Northern, Valley, and Kings i and e respectively, although i and e are by no means lost. This establishes the distinctness of ii and o from each other, as opposed to the unity of the Shoshonean sound variously rendered by it, i, o, u. The factors that caused their rise in Tule- Kaweah are wholly obscure: the Shoshonean influence was almost certainly their ultimate stimulus, but not their determining cause. In Buena Vista these vowels crop out less regularly, but can also hardly have been anything but secondary. This dialect group bordered on Chumash instead of Shoshonean speech areas. For Maidu and Miwok the situation is more intricate. Instead of the simple rule u ex i, 6 ex e, there is an irregular variety of corre- spondences. The nature of these even suggests that it is by no means certain that ii and 6, u and iu, are really distinct sounds. It seems possible that their difference is only in the apperception of recorders. There is, it is true, a somewhat greater number of cases of Maidu ii equating with other Penutian u and i, of 6 with o and e; but the ten- dency is not sufficiently preponderant to dispel all doubts. Moreover, ii equates not only with the back vowel u and the front vowel i, but not infrequently with a. Similarly, Miwok u is sometimes clearly u and sometimes clearly i in Costanoan, whereas it appears to be almost anything in the Pen languages. Finally, Maidu ii, 6 and Miwok u, u scarcely ever equate with each other. Shoshonean example may there- fore be indirectly responsible for their presence also; but if so, it has certainly affected the two languages diversely and probably at differ- ent times. The principal available cases in the basic list of this essay have been collected in the attached tabulation, arranged so as to show what little regularity is visible. It may be added that the Yokuts sounds present no correspondences with either the Maidu or Miwok ones. It may therefore be concluded that the Penutian mixed vowels are of alien origin; that they are restricted to border groups of dia- lects; that three times they sprang up under the same stimulus, but probably independently, in as many Penutian languages; and that their history in these three tongues, although still almost wholly obscure, was quite different. 1919] 83 84 University of California Publications in Am. Arch. and Etha. [Vol. 16 Maidu u .i 6, u 0 6 6, i,6 ii, i ol, u tii, 6 6, (u) a, a e a *Coast dialect: u.' Yokuts e, (a) i, (a) U, 0 a i 0, 11 e e e a, (i) a, (u) a (ya) a Miwok *i * u u u u e a a u, li e, a u, u u, (u) a u i, e, a u U, u o, u u,l - , i u o, u u, 4, u, i tProbably assimilated. Original Penultian Sounds The consonant system of Penutian may be hypothetically recon- structed as follows, with doubtful sounds in brackets: k tc t t p k' [tc'l tp P' g [dj] [d] d b w, y, h [xl Ic] s [r] n I m The original vowels probably did not go beyond u, o, a, e, i. These may or may not have been differentiated as to openness. ACCENT Accentuation of words is by stress in Penutian, but has been too irregularly recorded to make its representation throughout this paper worth while. Accent is commonest on the penult and antepenult, but extends from the last to the fourth from the last syllable; it varies Penutian , e i, (u) 0, u u, o, a U, i u 0 u 0, u e e a a a, u, 1 a, u, a u u u e o, (u) u a, (e) 6 44 104 126 126 131 111 107 101 45 57 151 112 96 72 72 81 16 21 47 161 90 164 159 72 79 Costanoan U, a i .0 0 0 u, a u, (i) i [o] u Dixon-Kroeber: Linguistic Families of California in the same stem between different languages; and even between dia- lects within one language. Suffixed and compounded words alter their accent and must be distinguished from stems. The rules governing accent are most fully apparent in Yokuts. Its normal place is the penult. From this it recedes to the antepenult, but never farther; nor does it seem ever to rest on the final syllable. It would appear that the accent is not inherent in any one syllable of the stem; not only because it shifts toward the end as certain syllables are added, but because as stems are expanded or reduced as between one dialect and another the accent moves to maintain the same relative place from the end. Thus: mu'k 'ac, muk 'e'la ta'lxatl, tala'pis, a 'la'das ga'peal, ge'pcil, ta'pad, gapa'sai tea'itac, tsaya'tas, tsoyo'tis There are exceptions, but mostly open to some doubt. ge'tsail, ga'tsifiil to'yon, to'yoni to'kco, do'pod, tu'kuyun In case a syllable is internally amplified it may retain its accent; for example, ko'hutem, compare no'tco, Costanoan ko'tco; do'nohop, in which -noho- seems to represent -lo- or -lu-; fio'ho'o, possibly from the same root, or perhaps felt as a disyllable: fioho'. Somewhat anal- ogous is ho'nohon, dialectic for usual ho'nhon. Diphthongs, even such as we should regard as mere juxtapositions of vowels, seem to be felt as single sounds in Yokuts: pi'tiu, ti'liu, tcu'lui, pa'dui, po'lwoi, k'i'lei, u'lui, mo'loi, t'e'nei, co'xgoi, mo'nai, te'ufian, po'noi, and so on. Accordingly, several words that might seem accented on the fourth syllable from the end, really have the stress on the antepenult: p'a'nuck'ai, tc'a'bakiu, pale'uyami. Case suffixes as a rule do not alter the accent. The plural suffix usually causes a shift. Tense and mode endings of verbs differ. One class, composed chiefly of short i-suffixes, leave the accent; the other, mostly long-vowelled28 a-suffixes, operate like the plural in a.29 Thus: Noutn Plutrals napa'tum, na'ptim-a naxa'mic, na'xmic-a (objective naxa'mc-a) wiiktca'mni, wiika'temin-a 28 Or perhaps causing lengthening of the vowel immediately preceding the suffix. 29 Present series, II, 190, 200, 212, 1907. 85 1919] 86 University of California Publi,eations in Am. Arch. and Ethn. [Vol. 16 xoco'mo, xo'cim-a komo'yic, ko'myic-a mo'koi, moko'y-o guyo'dum, go'yum-a bafikala'tci, bafie'klatc-i hitwa yu, hite'way-i tce'jej, tcija'j-a mu'k 'ac, muk 'e'c-i afitu, aflta'w-i a'xid, axe'd-i tcolinok, tcuye'nak-i a'gac, agesc-i o'nmid, one'mad-i (but oni'mid-i) Verb Suffixes almid, a'md-in, almid-ji, ama'd-ac, ama'd-ad elintim, ilintim-in, elfitim-ji, inieltm-ac, ifieltm-ad ha'wid, ha'ud-inin, ha'wid-ji, hawa'd-ac, hawa'd-itc t 'u'i, t 'u'y-on, t 'u'i-ju, t 'uit 'u'y-ud, t 'oyat 'ya'-nitc, t 'uyu't 'yu-wuc yi'win, ye'un-in, ye'win-ji, yuwe'n-itc, yuw-a na yom, yo'm-un, yo'm-jo, yuma'-ac, yuma'-ad, yuma'-niite It is clear that change of accent is closely associated with a loss and appearance of vowels, with an alteration of vowel quality, and perhaps with length of vowels. Which of these several factors is primary, it is difficult to say. But it is certain that the majority of the more notable cases of the vocalic mutation or ablaut that is so characteristic of Yokuts, are, or once were, accompanied by shiftings of the accent. In their pure stem form, Yokuts words seem to be normally accented on the penult. Maidu accent is more variable than Yokuts, and falls from the final to the fourth or fifth syllable from the end. Nouns and verbs differ, it would seem; but in both, vowel quantity is an important if not the deciding factor. Nouns appear to go by the rule that the accent falls on the long vowel nearest the end; if there is no long vowel the accent comes on the first syllable of the word. Ono'mus', ono, tsbl' buu',mini', kowo', tolil, pitl', suimi', haw!', yoso', mak6', kayl', kono', hemi', sukil', bos6', pok6', etc. tsi'tsi, na'na, tsalwa, kfl'yi, ma'wa, we'pa, fl'yi, etc. kono'ko, posa'la, ole'li, pekuini, pala'li, awA'ni, hom6'ni, homo'ko, etc. p6'tani, ya'mani, hell'eni, hl'pinak, hi'maduk, ame'lulu, etc. hi'ni, si'mi, e'ni, yi'mi, h6iti, m6'de, wi'sla, t6li, etc. tsi'laka, si'mpani, pa'ndaka, la'klakpe, ka'nanak (cf. kanal'na in another dialect), polo'loko, pa'kpaka. e'lalakno, tsi'statatk6 Dixon-Kroeber: Linguistic Families of California There are, however, some exceptions: sede', paka', kosi', kula', heme', lulu', bobo', popo' pa'no, kl'wi, hb'16 wili'll yui'tdihli, olo'lok6, witti'mtiimi Compound nouns tend to accent the first member, irrespective of long syllables in the second: mo1m-pAn6, sutlm-b'ukui, balt-sawi, toll'-wakA, hilnim-butuj, malidiim-bom6 Personal and deinonstrative pronouns follow the same rule as nouns. nisalm, nise&m, nisa', nise'ki, amaWm, ainatkan, uinllnan mi'nki, mi'ns6m, m6'yem, mi'ntseki, m6'seki, m6'im Verbs are more difficult to judge, because they appear to contain few long syllables. At least, few long vowels are marked in the avail- able data. The accent therefore recedes; but its most frequent place varies between the first and second syllables according to factors that are obscure. ha'ndakdaudom, i'nnotimoto, te'lapsito, ba'poldon, ba'yaldau, sii'dakdau, wa'daston, la'ptikinuweas, tW'sbokiteam. hil'tulsipdom, 6'puldon, s6'wodom, bo'dakdaudom, bfi'dutmin, huiimitdom, si'dakdau, wui'suwalakano, m&'datotsoia, b&'kasdau, he'appinpintsoia, we'yedonkfkan. hala'pnodom, s6hu'nb6ktsoia, bole'kwodoidom, bopo'ldon, b6y611kin, batsa'psito, huitu'lto, huide'ktodom, siwa'waitoweten, sika!lamak6kan, wa 'a'pdau, kapwi'- limdoinodom, oiko'iyahaka'ankano, wiwo'doinatseno. hayo'sip, sike'yodom, siita'nodom, yea'ssito, w6w6'hanodom, woda'kas, 6n6'lutmakas, 6y&madom. yui'lipsito, y6'n6n, wi'aspin, wita'tamotopi, s6'w6kitsoia, w6'tanfi. 16kdikn6mA'kas, tusbota'dom. There is no trace of a difference between classes of suffixes, as in Yokuts, some drawing the accent toward themselves, others allowing it to recede. It is however clear that diphthongs count as single and normally short sounds; which is as in Yokuts. Vocalic mutations differ from those of Yokuts. They do not seem to occur according to any system of true ablaut; and they are confined mostly to verbs, which are largely built up from monosyllabic stems. So far as can be judged at present, accent is not connected with vocalic changes in Maidu. The one factor that is rather certainly associated with stress is vowel quantity. If Wintun has ablaut, it has remained concealed. The process is, therefore, certainly less important than in Yokuts and Maidu. The recessive accent is also no longer traceable. Final accents appear in '87 1919] 88 University of California Publications in Am. Arch. and Ethn. [Vol. 16 both Northern and Southern dialects. As such accents are rare in Penutian they would seem to be of secondary origin in Wintun. This is confirmed by the fact that within Northern Wintun the accent of some words differs according to local subdialect. Hay Fork of Trinity River Upper Sacramento River so1no Lini'k ta'hal daha'l ku'ril (rib) tsuri'l olpu'ris (breath) o'lporu The speech of Hay Fork thus seems more inclined to preserve the ancient Penutian accent, that of the Upper Sacramento readier to shift it toward the ends of words. In line with this difference is the fact that antepenultimate accent is at best rare, perhaps wholly lack- ing, in the Upper Sacramento dialect, very common in Hay Fork. Southeastern Wintun shows words accented on the last, second, and third from the last syllable. As in the North, the determining prin- ciple remains obscure. A few Wintun words bear the accent on the first of four or five syllables, as Hay Fork su'getopi, fish net, but these are probably derivatives or compounds.' Miwok data are scant. It is certain that antepenultimate accent is common. But as vowel quantities have not been recorded system- atically, and the slight indications of vocalic mutation are too vague to be beyond doubt, the factors determinative of accent remain more obscure than in Yokuts and Maidu. Suffixes of case, and per- haps others, do not affect the accent of stems. This rule may account for some instances of accent on the first of four syllables: ka'wulu-to, hu'mele-tki, la'pisay-u. Penultimate stress is not unknown in words of three syllables: ali'ni, ahe'li, -epla'li, oti'ko, na'a'tca. In all such words the accented syllable seems to contain a long vowel. On the other hand, whenever the accent falls on the fourth syllable from the last all the following vowels seem to be short: ke'nnekakii, tolo"teakii. The probability is therefore rather strong that the accent of Miwok, like that of Maidu, is determined by vowel quantities. The lengthened consonants characteristic of most Miwok dialects are certainly associated with accent. They do not seem to appear invariably after each stressed vowel, but when they occur it is only in that position: wa'lli, ka'nni, awa'nnata, ese'llutki. In Costanoan, stems of two syllables prevail. But antepenultimate accent occurs, most formative, locative, and modo-temporal suffixes Dixon-Kroeber: Linguistic Families of California8 being added to the root without causing a shift of accent. In fact, there is no certain case of any Costanoan word being accented other than on the first syllable of its stem. Much more attention must be bestowed than has been usual here- tofore before the history of accent and of the factors related to it in the Penutian languages becomes clear. In summary, the following may however be suggested. Accent seems basically to be recessive, falling on the penult if this or the final syllable is long, on the antepenult if the last three syllables are short. Miwok and Maidu have carried this principle to the point of letting the accent recede even farther, if no long vowels follow. On the other hand, Maidu sometimes draws the stress to the final syllable itself, if this is long. Wintun also has come to know final accent, but on short vowels, so that the development in this language is likely to be distinct from the Maidu development. Yokuts has associated its accent deeply with its ablaut system. Miwok has lengthened many consonants under the influence of stress. SUFFIXES Various word-forming suffixes are clearly indicated by the com- parative lists, in which they have been designated by preposed hyphens. Many of these endings have been determined previously in studies of individual dialects, but are confirmed here, and several appear to be common to two or more of the five languages. Maidu, which favors short stems, shows by- far the fewest of these endings; but Costanoan, whose words are also compact, and which with the possible exception of Yokuts is grammatically the least synthetic lan- guage of the family, is well provided with word-forming suffixes. -s, Yokuts -c. This is the most frequent of Penutian suffixes, common to all languages except Maidu. A few instances follow; there are numerous other stems in the list on which the ending appears in one language only. Mw and C: 2 Y and C: 6, 26, 31, 61 W and C: 113 W and Mw: 47 W and Y: 72, 109, 128 Miwok -su, -hu, 14, 28, 80, may be the same suffix. Yokuts -a'l, -il, -1, -c-ul, 8, 14, 25, 34, 59, 145. Wintun -1, 74, 81, may be the same. Costanoan -x (F, J) and -tc (J, CL, CR), usually appear dialectically on the same stems: 10, 13, 40, 73. Individual Yokuts dialects occasionally show an -x or -hi which may be either a suffix or a phonetic increment: 32, 37, 41, 42. 1919] 89 t90 University of California Publioations in Am. Arch. and Ethn. [Vol. 16 -n is frequent in Santa Clara Costanoan: 21, 31, 37, 41, 57, 59, 61, 65, and occurs in Monterey: 114, 117. It is the nominative or absolute suffix of nouns in Maidu. Whether the -m in Maidu, 19, 131, is this or a true etymological suffix is not clear. -n, Yokuts: 33, 34, 46, 132, 133, perhaps 82. -min, -mini, is a Costanoan noun-forming suffix, common on adjective and verbal stems: 132, 135. It occurs also on several animal names: 114, 117, 123, which other languages indicate to be probably substantival. All of these are San Juan Bautista, but in 114 F, CL, and ca show the same ending. It is tempt- ing but seems hazardous to connect Wintun -men, -mel, of 72. A connection between Yokuts -t of 103, Miwok -ti of 110, Costanoan -t (B) of 110, -te (F) of 131, and -tei (B) of 73 is possible, but the randomness of occurrence compels serious reserve in acceptance. Southern Wintun has -ta on names of persons in 1, 4, 6 (N also), 56, and -tu, probably the same, in 55. A connection with Southern Wintun -ta on numerals in the count is proble- matical. Costanoan -wis and -ist, 131, 132, 133, 135, are previously known. Costanoan has -mak, -emk, -ma on several words denoting persons: 4, 6, 7. This seems essentially a plural suffix. The same may be said of Miwok -k, -ko in 1; Southern Maidu -k on the same stem may be connected. But Wintun -k (SE) in 85 and 94 is probably distinct, since it occurs on inanimate nouns. The same is true of Miwok (Co, L) -k in 30. Southern Miwok -tki, -tcki, in 9, 10, seems to be a diminutive. Besides Wintun -ta, -t, there are two endings of numerals that may be either suffixes or endings standardized under the influence of the analogizing that must tend to be powerful in any series of words habitually repeated in a fixed sequence: Miwok (and sometimes Maidu) -ka, -ko, -ke, and Costanoan, Yokuts and Wintun -n, -1, -L. MORPHOLOGY Vowel Mutations Yokuts possesses as one of its most distinctive features an intricate and subtle but regular system of ablaut mutations of vowels,30 pre- viously mentioned as connected with accentuation. Maidu is also characterized by rather abundant vocalic stem changes, though the probably complex rules under which these take place have not yet become clear.31 Miwok shows probable traces of the same process.32 That it has not yet appeared in Costanoan and Wintun may be due to its absence or to our imperfect knowledge of them. As vowel muta- tions thus occur in at least three branches of the Penutian family, they may be accepted as representative of its original state. Judging by their character in Yokuts, they are as "inflectional" in type as those of Indo-Germanic or Semitic. 30 Ibid., ii, 173-179, 190-191, 208-214, 230-231, 235-236, 289-292, etc., 1907. 31 Bur. Am. Ethn., Bull. XL, 686-687, 695, 699, etc., 1911. 32 Present series, ix, 288, 1911. Dixon--Kroeber: Linguistic Families of California The Verb and Its Affixes Suffixes are the only affixes found in Yokuts, Miwok, Costanoan, and Wintun. Maidu superficially presents a marked contrast with a liberal array of prefixes. It proves, however, that these are all of a single class, namely, instrumental modifiers of the verb. A number of them, moreover, are apparently recent in origin as prefixes, since they can 'themselves be used as verb stems.33 It is therefore likely that these elements will prove to be or originally to have been the first members of binary compound verbs. Of those that do not occur as separate stems, two show a clear relationship in form to the inde- pendent nouns denoting the corresponding part of the body, a fact that stamps them as evidently until recently noun stems used'in composi- tion with verbs, and indicates that the remaining elements of the same type may have originated in a similar manner.34 It is also a fact that Maidu is in contact with Washo and Achomawi-Atsugewi, both pre- fixing languages and both possessing elements of this very kind, so that their development in this tongue may have been stimulated by the influence of contact. At any rate, it is clear that Penutian'as a whole may be set down as having been orginally, and still being distinctively, suffixing only. Maidu is also exceptional, in the light of our p'resent knowledge, in the possession of another class of affixes to verbs, though these follow the general rule of the family in being suffixes: elements denoting spatial location or direction of motion.35 The same tendency toward greater concreteness of expression that caused the growth of the instru- mental prefixes, may be responsible for the development of this class of suffixes, which were again stimulated, perhaps, by the influence of foreign but adjacent idioms. The Penutian verb is, therefore, -apart from its stem mutatioins and personal endings, simple: its other modifications consist only of the addition of a series of suffixes which may be described as temporal, modal, and formative. By the last term are meant causatives, inchoa- tives, desideratives, and similar elements that in the so-called inflec- tional languages would probably be described as indicative of con- jugational classes. All the verbal suffixes of the group are accordingly 33Bur. Am. Ethn., Bull. XL, 693-698, 1911. 34 Ibid., p. 695. 35 Bur. Am. Ethn., Bull. XI,, 700, 1911. Not a trace of such elements occurs in Yokuts, nor according to the material of Arroyo de la Cuesta, who knew the language well, in Costanoan. The authors looked for them in Wintun and Miwok, so that in spite of the fragmentary knowledge of these languages, their absence from them can not be charged to mere oversight. 1919] 91 92 University of California Publications in Am. Arch. and Ethn. [Vol. 16 comparatively abstract and formal. The greatest departure from the average is displayed by Maidu, which not only seems to possess the largest number of suffixes, but to embrace those of most concrete and adverbial force, such as several denoting "nearly," completely," and "only." The tense suffixes even of similar dialects of any one of the five Penutian languages seem to be rather variable.36 Close similarities between the languages themselves can therefore not be expected, either for the temporal or the modal and formative elements. Nevertheless, enough suffixes can be found to connect all the languages (table 3). Some of the forms given as parallel in this table may resemble each other only superficially and not genetically; but enough remain to establish a fundamental identity of apparatus. The continuative has been included not because the endings are similar but because in Maidu and Yokuts this tense is independently formed from distinct stems but in the same manner: by the verb meaning to "be" or "live." The passive forms are perhaps also not related; but a true passive, expressed by a simple suffix, is such a rarity in the wide field of American languages that its occurrence in four of the five tongues examined should be enough to carry a heavy presumption of their unity. It is significant that it is again Maidu which is exceptional. The preterite tense elements apparently distinguish, in most of the languages, not only completed from progressing action, but also dif- ferent degrees of remoteness of time. The well defined and limited range of meaning of these verb-suffixes as a class is notable, and varies remarkably little in any of the five languages. In addition to the elements included in the table, there have been found, in any or all languages of the group, only incep- tive, benefactive, excellentive, interrogative, imperative, subjunctive ("when, while after"), present, future, and participial elements. Pronominal Elements In view of the fact that four of the five languages under considera- tion have previously been described as pronominally non-incorporative, it may surprise when the authors now state their belief that the Penutian verb was originally provided with personal endings. Miwok has not only a full but a variegated array of pronominal verb-suffixes ;37 36 Compare, for Yokuts, present series of publications, ii, 282, 337-338, 1907; for Miwok, ix, 309-311, 1911; for Costanoan, II, 72-74, 1904, and ix, 252-253, 1910. 37 Present series, ix, 283-286, 297-306, 1911. Dixon-Kroeber: Linguistic Families of California a Qt _ 0 O O C 0a e 4 . a 0 o o 0 C+. C+ C4. ~ ~ 0~~~~~~~~~~C CZ 02. 16Q C:~~~~~~~~C M I I P O W--P -l C P * 0 2 1C C D 2 CD 02 I I -0- m4. - k. *. - ,I. p 5p o 0 02 COP 0, ' F-- o 02" 0 02. - Z 0 CR. CR -:< a o o 4 z rJ20 N9 D'j 02~- 1919] 93 02 I 11 94 University of California Publications in Am. Arch. and Ethn. [Vol. 16 and in Costanoan,38 Wintun,35 and Maidu40 some fragments survive. It is true that the complex Sierra Miwok pronominal endings, which are altered for tense, may easily be a special development of these dialects. But the variety of the few suffixes in Costanoan, Wintun, and Maidu, and their lack of obvious relation to the independent pro- nouns in those languages, give them the appearance of remnants of a once larger array of forms, rather than the character of rudimentary beginnings of a conjugation evolving from stems through enclitics to suffixes; and this makes it almost certain that at least part of the Miwok pronominal suffix formation is old. The course of development of Penutian, in respect to this feature, is therefore much more prob- ably toward the analytical stage than toward the synthetical. Yokuts, whose pronouns are all distinct, though accentless, is accordingly the exceptional member of the family in the present particular, though to all practical purposes Costanoan is equally isolating. The Costanoan and Miwok suffixes indicate that the original pro- nominal suffixes were objective as well as subjective.41 As regards the possessive pronominal relation, the former status of the group is not clear. In all of the languages except Miwok the possessive pronoun is now invariably a separate word,42 and its expres- sion by suffixes in Miwok may be nothing more than a special and recent formation by analogy with the abundant pronominal endings of verbs. In the independent pronoun the distinction between the inclusive and exclusive first 'personal plural has not yet been reported for Costanoan or Maidu, but is certainly found in the three other lan- guages, and therefore seems normal. A pronominal dual is well marked in Yokuts, Wintun, and Maidu, in fact gives some indication of being basic to the plural in Maidu43 and Yokuts. 44 For Miwok a dual is at least not yet disproved, and may in some forms also underlie the plural.45 The elements indicative of number in the pronoun differ 38 Ibid., II, 72-73, 1904; Ix, 252-253, 1911. 39 Putnam Anniversary Volume, p. 463, 1909. 40 Bur. Am. Ethn., Bull. XL, 719, 1909. 41 Compare the passages last cited. 42 A few Yokuts terms of relationship commence with a syllable in n- which appears to.be a crystallized pronominal element of the first person. Compare this series, xii, 352, 353, 1917. 43 Bur. Am. Ethn., Bull. XL, 709-711, 1911. The same seems to be true of Wintun: compare dual nelit, malit, pulit, plural nelet, malet, pulet, with Southern Maidu dual nas, mam, mosam, plural nes, mem, mosem. 44Present series, II, 223, 1907. 45 Ibid., ix, 303, 1911. Dixon-Kroeber: Linguistic Families of California greatly in the several languages, evidently because the number ele- ments have been combined with case endings and then distorted by strong analogizing influences. Number and Case in Nouns The noun is marked in three of the five languages by a plural suffix for at least animate substantives; in all five, except for partial deviations in Maidu and Costanoan, by an objective and a possessive case; and in all five except Wintun by about four or five other instru- mental and local cases, which both by their form, and through their not being accompanied by an indefinitely larger number of locative endings of concrete meaning shading into semi-independent postposi- tions, indicate that they are true cases and not adverbial postpositions. This last feature is particularly important, and is only one more instance of the remarkable similarity of inner character and tendency which this family bears to Indo-Germanic. The number and case suffixes are given in table 4, and will be seen to be rather similar. Wintun and Maidu have lost the plural suffix. Maidu is also without an objective, replacing it by a subjective, whose sign, -m, is the same as the objective in Wintun. As the Maidu sub- jective is a sign of the intransitive as well as of the transitive agent, and is usually given with nouns standing alone, it really corresponds closely to the peculiar nominative of Indo-Germanic. The possessive is originally -n. Maidu has replaced this by -ki, and Costanoan has lost the case in both noun and pronoun. The instrumental occurs in all the languages, and is not only similar in all, but apparently of importance. In Wintun it is the only non-syntactical case remaining. A general locative is marked by t followed by a vowel: Yokuts -u evidently represents former -tu. An ablative and terminative are formed in Yokuts and Maidu by an n-, in Miwok and Costanoan by an m- element. In Maidu the two are similar in form; in Miwok and Costanoan, either identical, or, when formally distinguished, not sharply separated in meaning. It is therefore likely that these two cases are derived from a common base or are originally only one, the distinction between this element and the locative being one of motion as distinct from position. A comitative in k also seems a common heirloom of the family, though not universal. One or two instances of additional reported cases in San Juan Bautista Costanoan,46 such 46 Ibid., ix, 252, 1910. 1919] 95 96 University of California Publications in Am. Arch. and Ethn. [Vol. 16 .> . q 'I .. . a e a CE ca aa1 = .O I I I I ca -r -. a I4v W rn 4z 9 I 1 -4-,;,1 1 ,.:? :? 0 M 4.-.) 1 1 1 0 I c a4 :0 :0 :0 l4 I I a - "$-3?.3 Dec* 19 .................. .............* 5. ~.The EUliks X.andFg helhnoun, by N1 0 C.,Non. Pp. 357426, lates a 650 - -- -; X Bancr:ft-Ibta,t-LL--. :*e..-$. M-lQ8 ........... ............................. ....... -... 7 -~5- Ine,P.427443 ,VoL 8.o1 A __Mls1oueq#ord fth aifr Ind nffo ~Mznoipt Inth Bancrtoft: LibOrry y& L. Er-br.P 47 My . ..... .25 -- 42. TierEthbzograpIiy of tbi abflla tndians, by A L. Z:Eroeber. Pp. 2946-08 0 ~lates 1.l15. July, 190 .2?...*.. .75? *X?- - 3. TEhe egon of thle Luis4fo an- Dieguefoidia of Souen California; ~f by - -nbta2ce goddar.Duboe. Pp 9-8, plfis 16.19.iSr- Jun, 90 i...*.. 1.25; 4. r- ThoOultue l >ofthe Lu1tofi[ Ind1gus by pSte d r -rcmaii. Pp. 1];877 0 - - - p h2 plteg 20- Aus 1998. ----- --- ------- .. ...... .. 50 5t. &-N,o.t;es on ShoSb6neaii Diiectsoi SothernCator by AIL.or. Pp; . 235-269. etebr-i<909 .._ . ..- . . v .- . .3t 0. heeo s Pactl,ced- of DiegueEo Idlans, b T JnT.oWaterman. Pp.O 27-35, pdateis 21-, 1910 ...-S.e.. - _ 80 Index, bpp 859m0 k. - V-oL 9?.1 Ya- ~Tetsa;- b EdSvard S4l,together wit;h T:ana- Mtha -olliected ~by ~ - --;-bland0B. Dixon Pp. 1-235. I..ebru.t.. I .....,t.... . 2.50. ---=2. The Obumash' an* ostanoa Laguages, b A. Z,. Kreer.. ?. 37-27 - - -- The Languages of the t ofr Cai aNori of SanhrAucscq, by A. -L. >rq6ber' Pp. ~733e-4 Ian4 m-ap.- rl 1911 ... ..-. .- .1.-- 5-- r -In de pp.p 437-439. - ~ Vol. L--Y.1. Phonetic VobXtituents of tA1e jN~ioe Laaguages of alifordia,, bt A. L - roa eber. Pp.-12 May, 1911 . . .......... -10 .Tie Phionetic Elewents of theNrte ate Langu,ge, by T. T. Water- --~~~ ~~~P~ P_'' .- - - m>f --b 134 oimtyt v> ..61--- ~ AQI"&p.144 la; 1D5. -eerberug41............--..-.. . 3. PhonetiC Elements of the 1libae Il,6guage(byA.. I, Koeb; Pep.45-96, plte ---< As6-20.--November, 1911...r.......!.-.-. . . --.r-. ?5 4. Tih etbnology of the Salin an by - A.ae M . Pp% 91,2A0- -; plate 37.;b ........Deeu 11 d..............1. 75 5.-- -'pg VebSte by JanS6 Do lore, ;. 241,-263;. A,ugustv, 144t. ...1........................:......... .25i - * Notes- on thie Chilula ndi-as of wt oia,'-b PliSny are- a. o drd -s P p 265-288,elates 3841. Aprg1W14 .. . .30 -- -; 7.: Obiua Texs -by -Plin Earile. (4ard.S Pp. 2s89-379. -:{ombr 19t4 ..... 1X.00 InVex, 1Ph u81-5 . - e of Vol.1. Elment` eof tPhe iato Langge, by Pif b Earle od rd -?p. 1i-7 pLs '145. Otobter, 1912 . ... ..........2.00 2; -Phoftl E 1mets of the Die Lagge, b A. L. ooeber and 3. P. -; f- -8~Eardgou Pp. 177-188. Apr--il1914 --........ .....~...,........... -_ -.. .- .1Q- --.o Selarn _e.theea,an-3 n,b4.L.reer p 79.9. 5r _At ,S6e~5 at>-(6--4i8.Pery, 1915. ............. .~.............................,................ 1.0 8~~ .- Dichotomous Social Orgnztionin Suha C3u*entral Caifrna by E3dward--- 'Winlows Gifford. 291-296. February, 19 .... ... ...... --05 -6. !e Delineation of the-Da8i te Azt Aldect by T. Wat - -- man. -r 297-93t m arch, 1916 o-.o .....24-21. LgI-st, ..1 1.00 Z 7. Thie Muiu D)ialect of Qsaoa i Dased on tihe Vocab y o Do la Cu" . -<~~b 3 ~ X. Alden Mason. P e 3,>*@ 'p., *'39-4 JnoG 2p4-479. ........ .. ......... < ''r-7- CUua T'xts -b-Piy 14 1 * < - g ~~~~~~~~~~~~al.06ftr p .w.y*~f%*0 UNIVERSITY OF cALrFORNIA PUBLICATIONS- ('oTlNUED)I.u VoL 12. 1 ompition of Califrnia Sheliond by Edward Winslow Gmord. P. June, - 1~8. .-...29 ........-. FEb?., t'........... .... .4. .-..... . . . .. . 40 .. - , C m,Wff -**;f*t eAirwt It***wo- - 5---ra e - 3:.- 3Arpaho aleteby A. L. Eroeber. i P -138. June, 1916: ........ .70 ---- 4. Miwok cioel by dard 6Winslow f. p.. 9-194. Jnni, 1916.55 -5. On P1otting th Iec -osf: the Voice, Ib Corneliu B:. Boadley. Pp. 195 218, platOs; Otob er, ...................... .... ....... .2. . Ttlabal and Aa*aiisu ZiipT,sbLdward Wiislow Gifford.. Pp. 21t'9-.2?4 b7eus ryl*-, . . .1917 ...... .. ...... . 30 .0 7. Bdee tion to-the Study f nAncient ieiian Social OTgaizas tion, by T.1A T.Watermg. Pp. 249 P2ebruaxry,g _917 .......'........ .35 -- -I '8 w6k Myt'lds, ?y Edwar4 IwGifford. Pp. 283338, plateo6. May, . ..... ................. ...... .55l 9. COaliforia ZinShpystems, A. L. Kbroebe Pp. 9 . May, --1917 ......?................. .0 :- 10. -Cteremoniesof tie pomo -Izdansby St. a A. garret. \ . 397441,y 8 te t- flIures. July, 1917.................. ......A..45 . 1. Pazze -; ar -)or~ .y S. A. Barrett. Pp. 44,-465, late 7. July, ' 19'Br'7. 2544 - ne,p.467-473.- VoLl. . 1 IThe Position of tana in tklo '3okan Stock, byE. Sai Pp.,1.34. July, -T1 .h'naId~ yT. !I`.Watra. Pp3-02 plates 120.. Feobruary 1918 .......----.-.-- . ...... .. ... .. ... .75 . 3. Yai Ar,ey, b Saxtoii T . Pope. .103-.15'2, plate - 21-37.- Marchk, 1918 .75 4. Ya"a Terii pf 4R4latotshlp, by- dwad Sadir. IPp 153-73. Mch,1918 .25 Vol.: 14. -1. The; Langutage of teSalni 1nia by iJ- Alden jEMaon. Pp 4. 14154. -* r 0 Jan$uary 1 918 . ..- *w-. . .. . 1.75 ;2. Clansa ado ltie in Southern fnia, by Edward Wisow Gifford --:t p. 2,. 1 X fl itexi. Ma.- ................. .- . .... .75 3.F Ethnogeo^g?aphy a8nd Achaeoleg~of the Wiyot Territoxy;, by Lltew3lyn L . L - ;,oud. Pp 2>1-46, plates 11, 15s tebxt4lgures.I) ecembeif, 1918 ................................... . 2,50 >.-.4. -The Wintu Hesi Ceremony, by 8g.. A. Barrett. -Pp. -437488, plates 22.23, . fgurs a t3ext of 0ar- 1919 -.. . .Gford. 7 -t 5. The- Geneic: EeltilsQnsbp: of *the lForth American...................... Ind4an Langwages,- byg- Paul -a- - PP, 48S-502i. Ma, 919.^ ..................s>-. ..,. .15 Voh 1:5. 1. Ifugao kaw, bt 3B. P.-:Earto Pp.- 1486, plai 1-33. F ebruary, 1919O.2.^.0 2 J-2. Nabaloi Son y. U. Mossad A.L. .toeber. Pp. 187-206. May, 919 .20. Vol. 1[6. 1,l Myths. of til Soter Ser iok, by S.T-A. Barrett.: Pp. 1-28 March-,, 0:- 2.- The Matrlflneal Complex, by Robert H.-Lowie. P. 29-45. -Marchl, 1919....aQ - .15 3.he Lingstic FamiiJes of aifori by Rpa -.: DixoA and A.Lb. * ;oeber- Pp.7-8, m 1l 1ire in We?. Sepember, 1919 ..-.; 75 -Volmes now conzpleted.. i , - -R t ; a P : lume 1. ^ 19084904. 378, pae d 39 plate ........................... - >. _ . $.5 Volume 2os. X-i9044 07. 393 _agn. .*. 21 pl.tes ...... ...... .......... S. . 5 . . . . ... -Vo;u 3.; 19b5, ^The I?rpolg oi: t-he 1>upa L age, 34 pages . S-.50w V olume 4. 19004997. 8 74t page, With b5 tables 10 plaes ziud map ; .......................... ............ 3.,..50 . olu e 0.- - 190 400pages,nithi 8maps b..... ... .-.. ..50 Volume 9< 1910-1911. 439 pagus XXAy . ......... ............. ...... Volume 0. 191-4 3 pages andr 41 plas -.-...%...A . ... . 3.5 Vole IL 1911-191 479t pages-ass-&l i* -d A 6 -205. plte . .s . . .250 -N -'S Volume; 12. :1910-191 47. 3 ;.pages and plates...... . -........ . . , , , .509 --2-~ Lst 5i47gX *----ww_*-*,* - a,,. =ote:-.--The Uive.olt ofalifornia Publication fare offered In ex0hs:a for the pblibi- cationssof-leaxed s;octies ad itlverss,a librarlos. tnlet lists of all the publ.eatioxm of -tl WtXniers1ty wile ie~tl upon requ%es. FOr sample copies, -' st. L f Npublicabtions - or- other. inforiation, address the M.tARA;GZ- QF5 TE U1aTIVESbTeY ~PRESt, -vELBY, C-IGEl:ilA,- U. @S. A. -47Al ma tte snt -in exchange shou 75 e '0ddreS?d to oin le -11 M B . OALWFOR134. 1y^. [A v - , U n . .-, A.- .......... -,- - -A ht' t, P , J,i%-.., ' 20~ ,- ,'??--!N . * - I , " , - -,' . :, , 1- . , . - . ., " , -1 " - -1 ? - . . . f - , " , , ., ., - ' ' I I- -?2 -' -.- 4 -- - ? I z -7 - - , .. 7 , - , - , " 'f? - , , . -, " I _ -\', -1 .?,, ; -? I - , . - - , f '' . ? V.? ,