Comparative Hokan-Coahuiltecan Studies: A Survey and Appraisal 9783110887839, 9789027927170

Citation preview

JANUA LINGUARUM STUDIA M E M O R I A E NICOLAI VAN WIJK DEDICATA Series Critica, 4 edenda curat WERNER WINTER

COMPARATIVE HOKAN-COAHUILTECAN STUDIES A survey and appraisal

by MARGARET L A N G D O N

1974 MOUTON THE HAGUE'PARIS

Copyright 1974, Mouton & Co., Publishers, The Hague, The Netherlands. No part of this book may be translated or reproduced in any form by print, photoprint, microfilm, or any other means, without written permission from the publishers

LIBRARY OF CONGRESS CATALOG CARD NUMBER: 72-94480

Printed in Hungary

TABLE OF C O N T E N T S

1 Introduction

9

2 Setting the stage 2.1 John Wesley Powell 2.2 Future Hokan-Coahuiltecan families 2.2.1 Powell units 2.2.2 Other families

13 13 17 18 20

3 From hypothesis to (near) proof 3.1 Alfred L. Kroeber and Roland B. Dixon 3.2 Edward Sapir

22 22 36

4 Further extensions of the stock

48

5 Toward proof and establishment 53 5.1 Through the forties 53 5.2 The Symposium on American Indian Languages and "The Survey" 56 5.3 Recent developments in each family 61 5.4 Cross-family comparisons 69 5.5 Proto-culture and homeland 73 6 Hie future 6.1 The First Conference on Hokan Languages 6.1.1 Pomo 6.1.2 Yuman 6.1.3 Other Hokan languages

77 77 78 79 82

6

TABLE OF CONTENTS

6.1.4 Wider perspectives 6.2 Taking stock 7 Bibliography 7.1 Alphabetical listing by author 7.2 Index to the bibliography 7.2.1 Bibliographic information 7.2.2 Languages 7.2.3 Methodology, theory, general reviews 7.2.4 Prehistory and glottochronology 7.2.5 Ethnography 7.2.6 General 7.2.7 Acculturation 7.2.8 Anthologies 7.2.9 Reviews

85 86 89 90 110 110 110 113 114 114 114 114 114 114

1. 2. 3. 4.

Karok Shasta Chimariko Achumawi Atsugewi 5. Yana 6. Pomo 7. W a s ho 8. Esselen 9. Salinan 10. Chumash 11. Yuman 12. Seri 13. Coahuiltecan 14. Karankara 15. Tonkawa 16. Chontal 17. S u b t i a b a - T l a p a n e c 18. Jicaque

1 INTRODUCTION

The purpose of this monograph is to survey the literature pertaining to the group of American Indian languages included under the label Hokan-Coahuiltecan (also sometimes referred to as Hokaltecan)1, with specific emphasis on the evidence adduced for their common genetic origin. These lagguages are most highly concentrated in the present state of California, but have representatives spread over a wide geographical area - as far south as Nicaragua and as far east as Texas and northeastern Mexico. Some are separated from the nearest member of the group by great distances; even those in geographic proximity tend to be remarkably divergent in structure. The term STOCK is used here to characterize the group as a whole in that it is meant to include several distinct language families and LANGUAGE ISOLATES. The term FAMILY is reserved for groups of languages so closely related that their affinity is undisputed and can readily be ascertained by inspection of vocabulary items. The term language isolates (following the usage of HAAS e.g. 1966) refers to a language not so obviously related to others, i.e., a family with a single member. In addition to listing and summarizing the relevant contributions, this survey is critical in that it attempts to state not only the claims of each investigator, but also the quality and quantity of data on which the conclusions are based; the validity of results is also 1

The question of nomenclature is always a vexing one in an area where conflicting practices prevail. In the case of Hokan-Coahuiltecan versus the shorter and more euphonious Hokaltecan, I have chosen the longer and more awkward term as more descriptive of the degree of doubt that still remains about the actual relationship between Hokan and Coahuiltecan.

10

INTRODUCTION

discussed, when appropriate, in the light of methodological and theoretical considerations. While the main emphasis is on the verification of Hokan-Coahuiltecan as a genetic unit, no attempt is made to exclude typological and areal considerations as they have direct, though often negative, bearing on the question of genetic affinity. The comprehensive Bibliography lists, in addition to items dealing with matters of classification, the pertinent descriptive work on individual languages, with particular emphasis on recent work. A map identifying the location of the various groups discussed is appended. Although all known suggestions for extension of the stock by the addition of individual languages or by consolidation with other stocks are surveyed in a separate chapter (Chapter 4), they are not submitted to detailed scrutiny. The present study will make clear that a cogent evaluation of such far-flung relationships is best postponed until such time as the more central relationships are better understood. Problems of classification are of central importance in the field of Amerindian linguistics as a bewilderingly large number of languages dot the linguistic map of the New World. Attempts to bring some order into this situation (most frequently by postulation of genetic relationship) have therefore been a recurrent preoccupation of those scholars concerned with the native languages of America, not because this is eternal or immutable, or based on any ideal principle, but because it is an invaluable tool, a fixed point in what would otherwise be a chaos, and therefore a practical necessity (KROEBER 1913: 400).

Several schemes have been proposed ranging from conservative groupings not exceeding the level of "family" (e.g., POWELL 1891) to the postulation of single origin for all American languages (e.g., BRINTON 1891, R A D I N 1919, SWADESH 1954a). Most influential was that of SAPIR (1929a) whose investigations led him to propose a classification of the languages of North America into a small number of stocks and superstocks. In his scheme, Hokan-Coahuiltecan is a stock grouped with other stocks into the Hokan-Siouan

INTRODUCTION

11

superstock. The importance of SAPIR'S work in the Hokan-Coahuiltecan story is such that two distinct periods must be recognized, the first leading to and culminating in SAPIR'S classification (Chapters 2 and 3), the second covering subsequent developments (Chapters 4, 5, and 6) and clearly indebted to him whether the aim is to further expand the stock or, more importantly, to consolidate and/or amend his outline by more detailed and exact work. No absolute criteria exist as to what constitutes incontrovertible evidence of genetic relationship and therefore no absolute answer can be given to the question: is there indeed such a genetic unit as Hokan-Coahuiltecan ? What we can deal with are various levels of probability that a certain hypothesis is correct. The terminology and criteria proposed by HYMES ( 1 9 5 9 ) provide a framework within which such questions can be discussed and will therefore be freely used. The three stages in genetic classification are hypothesis, proof, and establishment [emphasis mine]. A hypothesis of relationship may be presented as just that. Often enough, a claim of actual connection is made, but so long as no evidence is published, or the published evidence does not meet criteria of proof we have only a hypothesis. When no inference other than that of genetic connection is possible, a relationship is proven. (The amount and kind of evidence necessary to eliminate alternative inferences may vary greatly from one case to the next.) To establish a relationship is to detail its specific content by extensive correspondences a n d reconstructions. (HYMES 1959: 52)

This survey on the whole is organized chronologically, but occasionally the presentation departs from strict chronology to insure better topical cohesiveness. The history of the classification presented makes no claim of originality. Some of the same ground has been covered at various times and with varying emphases by others (e.g., HAAS 1 9 6 3 , 1 9 6 4 ; JACOBSEN 1 9 6 4 , 1 9 6 6 ; BRIGHT 1 9 5 4 ; OLMSTED 1 9 6 4 ) , and I have benefited greatly from these sources in the preparation of this monograph. People too numerous to mention have added to and commented on a first version of the bibliography. Leanne H I N T O N drew the map. Leanne H I N T O N and

12

INTRODUCTION

Pamela MUNRO deserve special thanks as cheerful and independent research assistants at various stages in the progress of this work. A Junior Faculty Summer Research Fellowship granted by the University of California made it possible for me to devote full time to this project during the summer of 1970.

2 S E T T I N G T H E STAGE

Three main trends, each centering around a dominant figure, and each characterized by a distinct position with respect to linguistic classification, emerge from a bird's eye view of the period leading to the postulation of the Hokan-Coahuiltecan stock. (1) The POWELL classification of languages north of Mexico into 58 families, based on LEXICAL comparisons only. ( 2 ) The TYPOLOGICAL classification of RROEBER - DIXON, ultimately leading to the postulation of genetically related groups. (3) The SAPIR classification into superstocks, with evidence for genetic affinity adduced, not only from the lexicon, but from ARCHAIC MORPHOLOGICAL features. The contribution of POWELL, though it does not in any way make hypotheses above the family level, is so important as the basic foundation for all further work that it will occupy most of this chapter.

2.1 J O H N W E S L E Y

POWELL

Although the history of the classification of Amerindian languages begins much earlier, a firm starting point for the present review is POWELL'S famous synthesis Indian linguistic families of America north of Mexico (POWELL 1891), now conveniently available in paperback (BOAS - POWELL 1966). As has often been pointed out (for a recent reiteration see ELMENDORF 1965), in spite of the generally poor quality and scarcity of the data available, the great merits of this classification are, on the one hand, its exhaustiveness, and, on the other, the very conservative approach used in determin-

14

SETTING THE STAGE

ing genetic relationship, i.e., only lexical similarities are accepted as proof of cognation. With minor exceptions, POWELL'S "families" are still considered valid units by all concerned. It is worth noting also that, in addition to a list of all families, with geographical location and member "tribes", the work includes some other important features, among which are an annotated list of bibliographical sources and also, consonant with POWELL'S systematic scheme of nomenclature, for each family all names previously used to designate the family and the bibliographical reference in which the specific name appears. In addition, there is a map summarizing the findings (unfortunately not included in the paperback edition), and some explicit statements of POWELL'S theoretical views, a brief discussion of some aspects of which is in order at this point. For our purposes, it seems appropriate to consider only such assumptions as have a direct bearing on the job of classification; an analysis of POWELL'S philosophy of language as a whole is well beyond the topic of this monograph. The leading theme guiding the proposed classification is clear: only similarities in the vocabulary are to be used as proof of cognation. Thus: A single language is called a stock or family when it is not found to be cognate with any other language. Languages are said to be cognate when such relations between them are found that they are supposed to have descended from a common ancestral speech. The evidence of cognation is derived entirely from the vocabulary [emphasis mine]. Grammatic similarities are not supposed to furnish evidence of cognation, but to be phenomena, in part relating to stage of culture and in part adventitious [ . . . ] . It therefore becomes necessary in the classification of Indian languages into families to neglect grammatic structure, and to consider lexical elements only. But this statement must be clearly understood. It is postulated that in the growth of languages new words are formed by combination, and that these new words change by attrition to secure economy of utterance, and also by assimilation (analogy) for economy of thought. In the comparison of languages for the purposes of systematic philology it often becomes necessary to dismember compounded words for the purpose of comparing the more primitive forms thus obtained. The paradigmatic words considered in grammatic treatises

SETTING THE STAGE

15

may often be the very words which should be dissected to discover in their elements primary affinities. But the comparison is still lexic, not grammatic. A lexic comparison is between vocal elements; a grammatic comparison is between grammatic methods, such, for example, as gender systems. (BOAS - POWELL 1966 : 87)

I have quoted this paragraph almost in full as it is clear that POWELL'S notion of lexical comparison is a rather sophisticated one, that it assumes a good deal of what by other criteria are considered grammatical processes, that it allows what for all intents and purposes is the use of internal reconstruction, and that it accepts as proof of cognation phonological equivalences, but not syntactic. It can be seen that this view of cognation comes close to that held by the most revered of historical linguists lacking only the requirement of regular recurrent sound correspondences which, it must be recalled, had only been asserted in full rigor in Europe as recently as 1876, a mere 15 years before the publication of the classification (LESKIEN 1 8 7 6 : 2 8 ) . When we also recall that the work in question is the result of "labors extending through more than twenty years of time" (BOAS - POWELL 1 9 6 6 : 2 1 5 ) , it is clear that it could not have incorporated the insights of the Neogrammarian position. Now, from theory to practice, it is of interest to ask how this underlying principle was applied in the specifics of the classification, since the work itself does not exhibit the comparative lexical information on which it is based. While acknowledging the great efforts of a number of collaborators, POWELL takes personal responsibility for the classification itself. We must therefore assume that he perused the comparative material for all languages involved, and made decisions of cognation in each case. It is doubtful that this is the absolute truth in all cases, as may be illustrated by the fate of Seri, a language which was to be eventually considered Hokan. POWELL lists it as a Yuman language, but later consensus removed it from that family, giving it the status of a language isolate. POWELL'S bibliographic reference at this juncture is G A T SCHET ( 1 8 8 3 , 1 8 8 6 ) , two sections of a detailed survey of "Der

16

SETTING THE STAGE

Yuma-Sprachstamm" (GATSCHET 1877b, 1883, 1886, 1892), which included also the publication of comparative word lists collected by a variety of investigators. GATSCHET (1883: 133) states Die mir vor einiger Zeit vom Prf. Wilh. Herzog in Oppau (Rheinpfalz) mitgetheilte Vermuthung, dass das Seri ein Yuma-Dialekt sei, habe ich vollkommen bestätigt gefunden und habe daher diese Sprache dem Yumastamme beigezählt.

He specifically discusses the items he considers cognate (a total of 12) with the Gila and Colorado dialects and mentions others which point to similarities with other Yuman languages. He assumes that Seri is most closely related to the more westerly of the Yuman languages. Clearly this is the source of POWELL'S decision, and not his detailed perusal of the vocabularies, which, while they can be interpreted as showing the resemblances alluded to by GATSCHET, also show how skewed the percentage of similarities from Seri to any Yuman language is as opposed to any two Yuman languages to each other. GATSCHET also writes [ . . . ] das Seri ist ein von den Gila- und Colorado-Dialekten sehr weit abstehender Dialekt und Indianern jener Gegenden nur in wenigen Ausdrücken verständlich. (GATSCHET 1883: 133)

Nowhere of course does POWELL specify how many and how exact the similarities should be before he will accept them as proof of cognation. One must conclude that POWELL used the available suggestions of his predecessors when they appeared to be based on at least some tangible evidence, and that elsewhere he, like all comparativists, had an intuitive feeling for what is and what is not cognate. Seri was not allowed to remain a Yuman language for very long, since GATSCHET soon reversed his position: At present the chances stand entirely against genealogical affinity of Seri with Yuma (GATSCHET 1900: 558)

although he cites six forms which make connection with Yuma

17

SETTING THE STAGE

plausible, though obviously remote. In the same article, GATSCHET also rejects the inclusion of Pericü and Waikuri (of Southern Baja California) in the Yuman family where they had been listed by POWELL, on the very sensible grounds that Pericü - extinct and unrecorded - cannot be linguistically classified, and that the scant data available on Waikuri (BAEGERT 1773) show no observable resemblances to Yuman languages. Here again, POWELL had apparently taken a suggestion from the literature at face value.

2.2 F U T U R E H O K A N - C O A H U I L T E C A N

FAMILIES

Using the POWELL groups as a starting point, those families which were to be eventually included by SAPIR into his Hokan-Coahuiltecan stock (with the addition of Jicaque, which most recent classifications include in the stock as well) will now be listed to give a clear definition of the number of languages and families involved as well as their respective geographical distribution. The list may be used in conjunction with the attached map, as the ordering is from North to South and the number and main family name constitute cross references to the items on the map. The list consists mainly of POWELL units, 14 of them, one of which - Yuman - has been further divided into Yuman and Seri, for reasons discussed above. To these must be added three more families not covered by POWELL because their location wholly outside the geographical area covered by him excluded them by definition. We are therefore dealing with a core of 18 families. Of these, 11 are represented wholly or in part in California, and account for one half of the total number of California families {i.e., 22); 10 of the 18 are true language isolates; seven are completely extinct, two probably so, and one has less than five speakers. These figures in their starkness contain the clues to some of the particularly difficult problems encountered in attempting confirmation of any proposed classification. Language isolates are notoriously unsatisfactory units for deep comparisons, and extinct languages, even if well recorded - as is the case for only

18

SETTING THE STAGE

a few - cannot provide the additional data which new hypotheses would need for their confirmation. That none of these drawbacks were sufficient to hamper the search for deep relationships will be amply demonstrated below. For each family, the list indicates the name of the family (if two names are listed, the first is the one most commonly in use today, the second in parentheses is POWELL'S; if there is but one name, POWELL and modern practice coincide), the geographical location, the number of languages and their names as well as the number of dialects (if reported), and whether or not still spoken. In distinguishing between "languages" and "dialects" I have conformed to the predominant usage in each of the areas; in cases of conflicting usage, I have made the choice according to my understanding of the facts. The difference between a language and a dialect is essentially of minor interest from a linguistic point of view (in cases of clearly different entities, the question does not arise; in cases of a dialect continuum shading into different languages, clear-cut divisions are often not possible), but a consistent way of talking about entities is indispensable in matters of classification. The list of families, now follows. 2.2.1 Powell units (Quoratean): Northern California along the banks of the Klamath River, also more recently in Scott Valley; one language - Karok - with no significant dialect variations observable at the present time; less than 100 speakers (BRIGHT 1957). (2) SHASTA (Sastean): aboriginally " [ . . . ] a part of the Rogue River Valley in S. Oregon and the Scott Valley, the Shasta Valley, and a stretch along the Klamath River in California" (SILVER 1966b: Abstract); one language - Shasta - four main dialects; fewer than five speakers. (3) CHIMARIKO (Chimarikan): Northern California, western portion of Trinity County; one language - Chimariko - with two reported dialects; extinct. (1) KAROK

SETTING THE STAGE

19

(Palaihnihan): Northern California along the Pit River; two languages - Achumawi2 with nine dialects, Atsugewi with two; still spoken by a total of a few dozen speakers (4) ACHUMAWI-ATSUGEWI

(OLMSTED 1966).

(5) YANA (Yanan): Northern California, contiguous areas of Shasta and Tehama Counties; one language - Yana - with four dialects (Northern, Central, and Southern Yana, and Yahi); all extinct (SAPIR - SWADESH 1960). (6) POMO (Kulanapan): Northwestern California, drainage of Russian River and extending inland to include Clear Lake, plus a smaller noncontiguous area to the northeast on Stony Creek; seven languages all called Pomo preceded by the geographical qualifiers Central, Northern, Northeastern, Eastern, Southeastern, Southern, and Southwestern (also called Kashaya); all still spoken with the exception of Northeastern. (7) WASHO (Washoan): straddling the boundary between California and Nevada around Lake Tahoe; one language - Washo; still spoken. (8) ESSELEN (Esselenian): small area along the California coast south and east of Monterey Bay; one language - Esselen; extinct. (9) SALINAN: along the coast of California, directly south of Esselen territory; one language - Salinan - two reported dialects named after the local Missions (Antoniano and Migueleno); probably extinct at this writing. (10) CHUMASH (Chumashan): Santa Barbara islands and adjacent coastal territory of Southern California from San Luis Obispo to Ventura and inland; six languages, all called Chumash with qualifiers mostly based on the names of the Missions established in their territory, i.e., Ventureno, Barbareno, Ineseno, Purisimeno, 2

The name Achumawi accurately describes only one of the nine dialects. (1964) disambiguates the dilemma by using Ajumawi for the specific dialect and Achumawi for the language as a whole. The people who speak the Achumawi language are known as the Pit River Indians. The term Palaihnihan has recently become a close competitor for Achumawi-Atsugewi due to OLMSTED'S revival of the POWELL name. The two labels are synonymous and both will appear in later sections of this monograph. OLMSTED

2*

20

SETTING THE STAGE

Obispeno, and Island; all extinct - the last speaker of Barbareno died in 1965 (BEELER 1970a). ( 1 1 ) YUMAN : southernmost part of California including most of San Diego and Imperial Counties, extending into adjacent areas of Baja California, Mexico (with southernmost limits unclear and most languages extinct), northeast and east along the Colorado River and into Arizona, and southeast into small adjacent areas of Sonora, Mexico; eight languages - Walapai-Yavapai-Havasupai (considered three separate languages by some), Mohave, Yuma, Cocopa, Maricopa, Diegueno, Paipai (Akwa'ala), Kiliwa - dialect differentiation reported for most of them, particularly Diegueno; all still spoken. Among extinct languages of Baja California, Cochimi is usually considered to be clearly Yuman. (12) SERI: Tiburon Island in the Gulf of California and adjacent areas of mainland Sonora, Mexico; one language - Seri - several dialects; all still spoken. ( 1 3 ) COAHUILTECAN : contiguous areas of southwestern Texas and northeastern Mexico along the Gulf coast; at least six languages, the most commonly cited being Coahuilteco, Comecrudo, Cotoname; all extinct. ( 1 4 ) KARANKAWA (Karankawan): along the Texas coast northeast of Coahuiltecan, centered around Matagorda Bay; one languageKarankawa; extinct. (15) TONKAWA (Tonkawan) : Central Texas, north of Karankawa (the people were later moved to the vicinity of Tonkawa, Oklahoma); one language - Tonkawa - several dialects reported; there were six speakers in 1929 (HOLTER 1946b); now probably extinct. 2.2.2 Other families (Tequistlatecan): in the province of Oaxaca, Mexico; one language - Chontal (also called Tequistlatec) - with two main dialects (Highland and Lowland); still spoken. (17) SUPANEC (Subtiaba-Tlappanec), two geographically separated languages: Tlappanec of Guerrero, Mexico, with two main (16) CHONTAL

SETTING THE STAGE

21

dialects, still spoken, and Subtiaba, on the Pacific Coast of Nicaragua with a small offshoot in Lenca country in El Salvador, extinct. The two languages are very closely related and some investigators consider them dialects of a single language. (18) JICAQUE: along the northern coast of Honduras; one language - Jicaque -,two dialects; probably still spoken (HÄGEN (1943) reported one isolated community).

3

F R O M H Y P O T H E S I S TO (NEAR) P R O O F

3.1 A L F R E D L. K R O E B E R A N D R O L A N D B. D I X O N

From the newly founded (1901) Department of Anthropology at the University of California, Berkeley, poured forth in rapid succession a profusion of studies on the native languages of California, due to the efforts of several members of the Department, among whom Alfred L . KROEBER was without a doubt the leading spirit, untiring fieldworker, and broad synthesizer. His collaboration with DIXON produced a series of articles and monographs on the languages of California which for the first time put California linguistics on a respectable scholarly footing and clarified considerably the well-known complexity in the area where 22 of Powell's families were represented. This group's contributions fall naturally into three distinct (though not mutually exclusive) classes: (a) descriptive, consisting of word lists, grammatical sketches, and some short texts, (b) typological, and (c) genetic-comparative. Among the descriptive studies, and limiting ourselves to those of the contributions which are relevant for the future Hokan stock, we may note (a)

KROEBER ( 1 9 0 4 ) : grammatical sketches of Ineseno Chumash (based on material elicited by the author from one of the few remaining speakers), Salinan, and Esselen. His attempts to collect Esselen data convinced him that the language was by then extinct and the grammatical notes are based on the existing sources, all the data of which are published. Most extensive is

FROM HYPOTHESIS TO (NEAR) PROOF

23

the material collected by HENSHAW and alluded to by him in his 1890 article announcing the discovery of this new linguistic family (HENSHAW 1 8 9 0 ) ; (b) KROEBER ( 1 9 0 7 ) : a grammatical sketch of Washo, which claims to contain the first vocabulary in the language ever to be published. While this claim turns out to be inaccurate (as pointed out by JACOBSEN ( 1 9 6 4 : 10), there is in existence a word list by COLLINS which appeared in SIMPSON ( 1 8 7 6 ) ) , this does not detract from his contribution, as the grammatical analysis is obviously a first and uniquely KROEBER'S ; (c) DIXON ( 1 9 1 0 ) : a grammatical sketch of Chimariko, based on data collected from the last remaining two speakers; (d) KROEBER ( 1 9 1 0 ) : additional data on Chumash languages, consisting of comparative word lists of Obispeno, Ineseno, Barbareno, and Santa Cruz Island (from various sources), and Ventureno (collected by the author); (e) KROEBER (1911a): a grammatical sketch of Eastern Pomo and a very short one of Karok; (f) KROEBER (1911c): a paper on Mohave phonetics; (g) KROEBER - HARRINGTON ( 1 9 1 4 ) : a description of Dieguefio phonetics. Add to this SAPIR'S "Yana texts" (SAPIR 1910), and we are presented in a short period of ten years with some descriptive material beyond mere word lists for all the California families we are concerned with, except Shasta and Achumawi-Atsugewi, and even here some grammatical observations are made in DIXON'S classificatory articles (DIXON 1905, 1906b), though the bulk of his data on these languages has not to my knowledge ever been published. Typological studies include first of all a general typological classification of the languages of California ( D I X O N - KROEBER 1903) followed in rapid succession by studies of specific lexicalsemantic and structural domains, i.e., the dual ( D I X O N 1906a), numerals ( D I X O N - KROEBER 1907), phonetics (KROEBER 1911b), kinship systems (KROEBER 1917). Typological comments on individ-

24

FROM HYPOTHESIS TO (NEAR) PROOF

ual languages within the framework proposed are also found in the predominantly descriptive works mentioned above (specifically KROEBER 1 9 0 4 , 1 9 0 7 ; DIXON

1910).

The linguistic profile of California which thus emerges is worth presenting in some detail. First of all, the Yuman family was not included in the classification on the grounds of its being only marginally Californian and obviously different in structure from other language groups of the state. Upon later examination, Yana also had to be excluded as not fitting any of the types uncovered. With these exceptions, three distinct groups are recognized, on the basis of their sharing the presence or absence of a number of struct^ ural characteristics (plus lexical similarities assumed to be due to borrowing), i.e., phonetics (understood to include such features as gl'ottalization and phonotactic constraints such as complexity of consonant clusters), degree of transparency of the morphophonemics, degree of incorporation of pronominal elements into the verb, existence of syntactic cases (subject, object), existence of material cases, grammatical marking of plurality and reduplication, as well as some less widespread features such as the occurrence of a dual, gender distinctions, sex distinctions, specific types of demonstratives. It is clear that the existence of any of these characteristics as productive structural processes of a language would become readily apparent, even in the course of a fairly short period of field work, especially to investigators familiar with other languages of the area. The positive identification of some of these features in a language can therefore be considered as trustworthy; in the case of negative identification, we may at least conclude that the feature is much less clearly present. The nature of the classification and its restrictions are very clearly stated: It must be clearly understood, however, that the classification that has been attempted deals only with structural resemblances, not with definite genetic relationships; that we are establishing not families, but types of families. When several linguistic stocks have been put into one group, there is no implication that they form one family, in the sense in which this word is accurately used in philology. The classification here

FROM HYPOTHESIS TO (NEAR) PROOF

25

proposed is really one of another order from that used by Powell, for structure and not lexical content is made the basis on which all comparisons are made. (DIXON - KROEBER 1903: 2 - 3 )

The authors thus appear to share POWELL'S conviction that only lexical comparison can provide proof of cognation. The three groups identified, which occupy geographically distinct areas, are assumed to share their common features as a result of a process of diffusion similar to that commonly attested for other cultural elements, and DIXON and KROEBER indeed specify that they find cultural parallels to their linguistic groupings. The three groups are: (a) A Southwestern group, consisting only of Chumash and Salinan, characterized by pronominal incorporation, well developed plural, lack of syntactical cases, use of prepositions instead of case-appositions, and a not very simple phonetic system (DIXON - KROEBER 1903: 18).

(b) A Northwestern group, consisting of Yurok and Wishosk (Wiyot), Athabaskan, and (less typically) Karok and Chimariko, characterized by systematic pronominal incorporation, total lack of a plural, lack of syntactical cases, presence of material cases, and phonetics that are more or less rough and involved (DIXON - KROEBER 1903: 18).

(c) A Central group, consisting of all remaining families, with Maidu considered typical. Among future Hokan languages in this group are Pomo, Esselen, less distinctly Shasta, Achumawi, and possibly Washo. They are characterized by absence of pronominal incorporation, an undeveloped plural, syntactical cases, material cases, and distinct, simple, and soft phonetics ( D I X O N KROEBER 1 9 0 3 : 1 8 ) .

Interesting from the point of view of future conclusions is the fact that those families which were later to be assigned to the Hokan stock are, with the exception of Pomo and Esselen, considered

26

FROM HYPOTHESIS TO (NEAR) PROOF

somewhat divergent from the regular type. Thus Chumash and Salinan are considered a group by themselves, Karok and Chimariko are less clear representatives of the northern group, Washo is only marginally included, and Yana and Yuman are totally excluded, while Shasta and Achumawi are a "somewhat distinct subgroup" (DIXON - KROEBER 1903: 18). The first claim of genetic relationship between some of the POWELL families was made by DIXON (1905) in relating Shasta (for which he recorded four distinct dialects) and Achumawi-Atsugewi (which he clearly identifies as consisting of two main languages) under the label Shasta-Achumawi, substantiating his claim by a comparative wordlist of about twenty items in the six language varieties, and comparing six grammatical elements for Shasta, Achumawi, and Atsugewi. Further details are added a year later (DIXON 1906b). Here he concentrates on the interrelationships of what he considers five major languages: Shasta (plus its dialect Okwanutcu), New River, Konomihu, Achumawi, and Atsugewi. I will summarize the picture of the subgrouping that emerges in the diagram below (Fig. 1), which exhibits shared vocabulary, in

Konomihu

New River

\3 itemsf

^^

Shasta

ι item 5 items Fig. 1

Achumawi

f

At s ., g e w i

FROM HYPOTHESIS TO (NEAR) PROOF

27

percentages, when computed, in number of items otherwise. The diagram intends to show that there is more lexical similarity between Shasta and Achumawi than between Shasta and Atsugewi, that New River is about as divergent from Shasta as is Shasta from Achumawi, and that New River and Achumawi have practically no similarities. As to Konomihu, it is so different that it is "[.. .]a language which was trembling on the verge of independent existence" (DIXON 1906b: 259). Structural similarities also confirm the greater closeness of Achumawi and Atsugewi as opposed to Shasta and, even more so, to New River and Konomihu, though grammatical information for the latter two is almost entirely lacking. In spite of all the divergence, however, DIXON is convinced of the genetic unity of the group. While the basic evidence is lexical (from all areas of the vocabulary), and even a few recurrent sound correspondences are noted, the marked structural similarities are also taken into account, essentially implying a parallelism between the amount of lexical and structural resemblance. Stressing the underlying unity of languages differing so markedly in their observable structure, DIXON goes so far as to predict that new and deeper relationships will be uncovered in the future. More details on subgrouping within established families kept accumulating as well. For Pomo (in addition to vocabularies of all seven languages) BARRETT ( 1 9 0 8 ) sketches the internal relationships as follows:

proposes a subgrouping of the Yuman languages into three groups. Though it is based only on a comparison of the numeral systems, it is probably more in accord with linguistic facts than later proposed subdivisions of the family. HARRINGTON ( 1 9 0 8 )

28

FROM HYPOTHESIS TO (NEAR) PROOF

(A) Eastern Group: Havasupai, Walapai, Tonto (obviously a dialect variant of Yavapai), Yavapai. (B) Central Group: Mohave, Yuma, Maricopa, Diegueno, Cocopa. (C) Lower California Group: Kiliwa (and Santo Tomäs), Cochimi, Waikuri(?). The only question not resolved is the position of Paipai, a language he does not mention. For Chumash, KROEBER ( 1 9 1 0 ) proposes three subgroups, leaving Purisimeno unclassified for lack of data. The following tree diagrams the internal relationships.

Obispeno

Ventureno

Ineseno Barbareno

Island

Finally, in a one-page summary and in a slightly lengthier though sketchy justification DIXON and KROEBER announced, in a total reversal of their previous findings, that they had uncovered new and deep genetic relationships among the languages of California (DIXON - KROEBER 1913a, 1913b), reducing to 12 the number of "families" to which these languages may be assigned, from the initial total of 22 (soon reduced to 21 by DIXON'S Shasta-Achumawi). Notable for our purposes are two new stocks, HOKAN, consisting of Shasta {i.e., Shasta-Achumawi), Chimariko, Pomo, Karok (qualified by "probably" in DIXON - KROEBER 1913a), and Yana (qualified by "possibly" in DIXON - KROEBER 1913a). In DIXON - KROEBER (1913b) the two latter families are included into Hokan on an equal basis with the others. A second suggested stock called ISKOMAN (DIXON - KROEBER 1913b) suggests, but with less conviction, a relationship between Chumash and Salinan (erstwhile southwestern group by the typological approach). Even though Iskoman is considered still dubious, the authors entertain very tentatively the suggestion of ties between it and Hokan. In both

FROM HYPOTHESIS TO (NEAR) PROOF

29

publications, it is clear that the authors are anxious to inform their readers of their new discoveries without delay but cannot find the time to substantiate their conclusions in detail. The 1913 statements stand more as a declaration of faith with the barest amount of demonstration (for Hokan five presumed cognate sets-the words for 'eye', 'tongue', 'water', 'stone', and 'sleep' - with attempted reconstructions, for Iskoman 12 presumed cognates listed side by side). A few tentative general structural characteristics of the Hokan languages are identified: absence of plural in most nouns, extensive suppletion between singular and plural verb stems, verb suffixes indicating plurality, instrumental verb prefixes and local suffixes, affixed pronominal elements (usually prefixes). The prediction is made that Seri and perhaps other unnamed families will eventually be found to be related to Hokan as well. A lucid device of breaking up words into syllables and arranging cognate sets in columns such that the compared elements clearly match is introduced. In the absence of firmly established sound correspondences, this is without a doubt the clearest and most concise way of allowing the reader an insight into the assumptions involved in this type of comparison. An example will demonstrate the point:

Karok Chimariko Shasta Pomo Yana Esselen Yuman original form

'water' a s a ka a tsa ka ha -na a sa -nax a ha a sa a ha

{

This type of reconstruction (covering only five items) is very subjective as no set of correspondences appears more than once in the data, though an inspection of the forms makes the assumption

30

FROM HYPOTHESIS TO (NEAR) PROOF

of cognation quite plausible. Nor did the new claims go without criticism. For example, FRACHTENBERG (1918) finds DIXON'S and KROEBER'S presentation inadequate from the point of view of both methodology and quality of data. He urges publication of the full data on which the conclusions are based and suggests more attention be given to morphological evidence, which he feels is less subject to borrowing than the lexicon. The extent of the reversal in D I X O N ' S and KROEBER'S position is more dramatic than the terse and almost matter-of-fact published statements imply. As late as 1910, D I X O N approached the subject of claims of genetic relationship in the most cautious terms. In the present state of our knowledge of the extent to which borrowing has taken place in California at large, it is difficult to arrive at a definite solution of the question of the relationship of Chimariko with the Shastan family. The extent of the similarity in this case, however, points to the necessity of a thorough investigation of the whole matter of borrowing throughout the state. The question also involves the much wider one of the real limits of genetic relationship, in the need of determining the character and number of agreements which shall be regarded as essential to establish common descent. (DIXON 1910: 339) Even the main announcement of the new discoveries is almost casual: [ . . . ] we undertook some time ago a comparison of more than 200 stem words in all the languages and dialects of California so far as material was available. From the time the material began to be assembled some interesting results as to the character and scope of the borrowing of words commenced to appear; but after analysis of the collected information had progressed beyond a certain point, it became apparent that the only satisfactory explanation of the resemblances between certain languages was genetic relationship. On the basis of these indications the grammatical information extant on the same languages was reexamined, and in every instance was found strongly confirmatory. Lexical and structural similarities coinciding and being relatively abundant, true relationships have been accepted as established. (DIXON - KROEBER 1913a: 225) It is of considerable interest to note that the same year, 1913, also

FROM HYPOTHESIS TO (NEAR) PROOF

31

saw the publication of a theoretical article by KROEBER on the notion of genetic relationship, prompted by criticisms and misunderstandings (by whom is not clear, but emanating from Europe) of the 1903 typological scheme, about which he states: [ . . . ] the principal point of the [1903] essay was the demonstration that languages that were completely diverse in origin [ . . . ] could be similar in structure (KROEBER 1913: 389) because [ . . . ] it would appear that their structural identity must be due to one have [J/C] gradually patterned itself after the other [ . . . ] . (KROEBER 1913: 391)

The notion of 'Sprachbund' made famous by TRUBETZKOY (1939) is clearly foreshadowed here. After discussing the two opposing positions that give greater weight in the demonstration of genetic affinity to grammatical over lexical similarities and vice versa, he clarifies his personal views: The truth probably is the middle one that what is needed to establish positive relationship between languages, is similarity both in grammar and in lexicon. When a tentative classification is required for some practical purpose, as in ethnological or historical study, it is certainly justifiable, in the absence of fuller information, to consider languages as related because their words are regularly similar. On the other hand, to unite them genetically when their words present no resemblance, merely because they seem to employ similar formal procedures, is probably always hazardous and unsound. But everyone must admit a common origin when both form and content are substantially alike. As long as similarity is established only in one respect, there is always the possibility that such resemblance may be due to influence or borrowing. Specific conditions, however, are of the greatest importance. If languages are geographically separated, different in structure, but alike in words, it is at least probable that they are related, especially if there is no evidence of historical contact between them. Their grammars may gradually have diverged considerably in the extension of time since the original separation; but the resemblance in word-stems, if at all large, cannot be accidental, and must be due either to intimate former contact

32

FROM HYPOTHESIS TO (NEAR) PROOF

or to original unity. If two such languages are in geographical juxtaposition, the assumption of relationship is much less reasonable, owing to the increased opportunity for borrowing of terms. If languages in contact are similar in structure but unlike in vocabulary, judgment must also be suspended, because of the likelihood of borrowing or growth of grammatical forms under the stimulus of unconscious imitation. If in such a case the stems of words, when they are finally ascertained, prove to be thoroughly dissimilar, the languages are certainly unrelated in spite of their being of the same morphological type. In short, territorial continuity and contiguity are always factors to be considered. It cannot be doubted that purely morphological or formal factors can and do spread from one language to entirely unrelated ones, in other words, from one linguistic family to another. (KROEBER 1913: 390-1) but Only when the linguistic families of the world, that is to say, the largest ultimate units of speech that can at present be recognized, shall have been approximately defined, will it be possible to realize the degree to which structural convergence and assimilation - and for that matter, borrowing of words - have taken place. (KROEBER 1913: 394) He concludes: There is only one procedure possible for the historian of languages: to advance in the old-fashioned way of studying specific material carefully, practically, broadly, and sensibly, without reference to preconceptions or theorems derived from psychological speculation. (KROEBER 1 9 1 3 : 401)

It is more than probable that KROEBER ( 1 9 1 3 ) was written well before the two 1 9 1 3 articles by D I X O N and KROEBER (which were undoubtedly rushed into print). I have quoted at length from this statement to show that the divergent results of the 1903 and 1913 efforts of D I X O N and KROEBER are not contradictory, but essentially complement each other and are based on deep insights into the various factors which influence language change. Their ability to progress beyond their first assumptions speaks well for the flexibility

FROM HYPOTHESIS TO (NEAR) PROOF

33

of their model of language as the carrier of the historical legacy of peoples. Meanwhile, additional developments pointed to ever more inclusive Hokan connections. The following paragraph appeared in the American Anthropologist. Mr. J.P. Harrington announces that his recent extensive study of the languages of the Yuman and Chumashan stocks of California leads him to believe that these two stocks are genetically related, and that the relationship is traceable even in some of the most minute features of the structure. (HARRINGTON 1913)

It is unfortunate but typical of HARRINGTON'S practice that he never published a justification for his statement. However, KROEBER was (understandably in view of his recent claims) sympathetic, and refers to HARRINGTON'S statement approvingly: As his [Harrington's] studies in recent years have made him the best informed authority on both languages, his verdict must at least be taken seriously. (KROEBER 1915: 287)

No exception can be taken to this statement, as anyone who has consulted some of HARRINGTON'S linguistic material is continuously struck with his deep understanding of the languages he recorded. This, however, is not in itself sufficient to substantiate claims of genetic relationship, an area where HARRINGTON himself has shown his ability to be led astray (e.g., HARRINGTON 1 9 4 3 ) . This comment of KROEBER'S is only a very minor part of an article in which he links Seri and Chontal to Hokan, following an early suggestion of BRINTON'S relating Seri, Chontal, and Yuman (BRINTON 1 8 9 1 ) . The proposal is substantiated by a comparative vocabulary of 35 items in Chontal, Seri, and Mohave (the Yuman language KROEBER knew best) and some corresponding words from other Hokan languages; recurrent sound correspondences are pointed out and discussed. While quite a number of these are probably spurious and several of the sets may not be true cognates, nevertheless the evidence exhibits some rather convincing though few

34

FROM HYPOTHESIS TO (NEAR) PROOF

striking parallels. One might have wished for a maintenance of the clear presentation of the earlier Hokan cognate sets, as correspondences discussed do not always refer to the part of the word one would assume to be compared, particularly in the case of the vowels. With due allowances for the tentative nature of these conclusions, of which KROEBER was quite conscious ("No one is better aware than we of the slenderness of the evidence[...]"-KROEBER 1915: 287), it is clear that the types of relationships being proposed were worthy of serious consideration. Nor should the scantiness of the evidence be allowed to dilute the forcefulness of the claim being made. KROEBER is emphatic on the subject: It has been suggested to me that while there is probably some underlying truth in most of the recent mergings of stocks, the kind of relationship involved may be of a different sort from what has heretofore been regarded as the relationship binding together the members of a linguistic family. I wish to express my absolute opposition to this attitude. If Chontal and Seri are not related just as thoroughly and just as completely to Yuman and Pomo and Chimariko as Omaha is to Dakota or as Cherokee is to Iroquois or as Arapaho is to Delaware, they are not related at all, and the present essay has entirely failed of its purpose. I recognize only one criterion of relationship: reasonably demonstrable genetic unity. Either two languages can be seen to have been originally one, or they cannot be seen to have been one. The evidence may be of such kind and quantity as to leave us in doubt for a time; but there can be no such thing as halfrelationship. Philosophically, the concept of the linguistic family may be of little moment or validity, like the concept of species in biology; but for the organization and practical control of knowledge both these categories are indispensable. And they can be of use only if they stand for something definite and if as categories they are inflexible. (KROEBER

1915:289) t

The very same year, a separate new claim of genetic relationship between a number of languages of Southern Texas and Northeastern Mexico was proposed by SWANTON (1915). He suggests relationships between the Coahuiltecan, Karankawa, Tonkawa, and Atakapa families, all of which, except Atakapa, were later to be considered related to the Hokan languages. The evidence is inconclusive, but the great merit of the article is to have pointed out some

FROM HYPOTHESIS TO (NEAR) PROOF

35

remarkable parallels among these languages, and, most importantly, to have published comparative vocabularies of Tonkawa, Coahuilteco, Karankawa, Comecrudo, Cotoname, and Atakapa, all of which except Tonkawa were by then already extinct. A possible connection of these languages with Muskogean is even hinted at. One more terse announcement by HARRINGTON Mr. John P. Harrington announces that he has found genetic relationship between Washoe and Chumashan. (HARRINGTON 1917)

- brings yet another candidate for Hokan membership into consideration, i.e., Washo, and the stage is set for a grand consolidation of all these suggestions. The final justification of DIXON'S and KROEBER'S claims of genetic relationships in California appeared in 1919. In a very lucid "historical introduction", the authors review their own work as well as their underlying assumptions and graphically describe their increasing doubts about the meaningfulness of their first results and the inevitability of the new conclusions about genetic relationships. They then proceed to demonstrate their results in detail for the Penutian languages, discussing sound correspondences, morphological features, and historical interrelations of the families. For Hokan, however, they limit themselves to reviewing the history of the developments, to officially stating their conviction that their Iskoman (Chumash-Salinan) is part of Hokan, and to predicting, with the support of two comparative vocabularies of Washo (their own and that independently compiled by SAPIR) with other Hokan languages, that this language belongs in the stock as well, as already claimed by HARRINGTON in his linking this language with Chumash. The full justification of the establishment of the basic Hokan stock, however, they consider to have been already performed in a previously published paper SAPIR (1917b), to which we will turn below. Their decision to leave confirmation to someone else (though be it in their opinion in a more masterly way) is to be regretted, as it would have been particularly interesting to compare their treat-

36

FROM HYPOTHESIS TO (NEAR) PROOF

ments of Penutian and Hokan. This is not to detract from SAPIR'S contribution, which no one could deny, but to regret that DIXON'S and KROEBER'S methods could not have been compared with his and thus enriched our understanding of the results obtained. In fact, in his paper on Hokan and Coahuiltecan (SAPIR 1920a), which appeared three years later than his paper on Yana and Hokan but was apparently written earlier and certainly earlier than DIXON - KROEBER (1919), SAPIR (1920a: 280) still assumes that DIXON and KROEBER will publish the full evidence. On this very anticlimactic note ends DIXON'S and KROEBER'S main contribution to the creation of the Hokan stock. A final note on terminology: since they were confronted with assigning labels to their new groupings, DIXON and KROEBER decided to coin terms based on "forms of the numeral 'two' in the families of speech involved" (DIXON KROEBER 1919: 54) (thus Hokan from, e.g., Karok ?dxak, Shasta xuk'\va Chimariko xoku, Atsugewi ho?ki, Yuman (Diegueno) xawak). 3.2 E D W A R D S A P I R

Before entering into the details of SAPIR'S enormous contribution, it may be well to point out that it too had its inception with the Department of Anthropology at the University of California, where he spent one year, 1 9 0 7 - 0 8 , working on the Yana language (MANDELBAUM 1 9 5 8 ) . The results must be assessed in the combined recognition of his insight and diligence and of the obvious excitement which that environment so clearly generated at the time. This is communicated in all of the publications of the period, which have a certain breathlessness, arefilledto the brim with information, and somehow never completely fill all the gaps and demonstrate the claims. Apparently so much that was new and exciting was happening in such a short time, and the number of investigators was so limited, that they could not pause for breath between discoveries. It will become apparent in the course of this survey that the most important work being done today is still concerned with the

FROM HYPOTHESIS TO (NEAR) PROOF

37

evaluation, refining, and precise explication of these early contributions. SAPIR (1917a) welcomed HARRINGTON'S linking of Washo and Chumash, seeing a corroboration of his own conclusions and promising published evidence for the Hokan status of Washo. That same year, under the unassuming title "The position of Yana in the Hokan stock", SAPIR (1917b) assembled more evidence for Hokan than had ever before appeared in print. As pointed out above, so overwhelmed were D I X O N and KROEBER that they considered themselves exonerated from the obligation to present further justification for their Hokan stock. SAPIR'S contribution is so nicely summarized by them that I can do no better than to quote: Dr. Sapir examines nearly 200 stems and elements which Yana shares with the other Hokan languages. Most of these stems are also common to two or several of these languages among themselves. The net result is therefore the demonstration of a mass of evidence on the interrelation of the Hokan languages as such. It merely happens that the starting point of the presentation is Yana. But the list of lexical equivalences is the least weighty content of Dr. Sapir's monograph. In passing, as it were, he formulates phonetic equivalence, indicates the course of the shifts, analyzes forms in one language by the touchstone of forms in others, throws light on morphological developments, and is even able to reconstruct some hypothetical forms of the original undivided Hokan stock. (DIXON - KROEBER 1 9 1 9 : 1 0 3 - 4 )

Language families compared in the study are Karok, ShastaAchumawi, Chimariko, Yana, Pomo, Esselen, Salinan, Chumash, Yuman, Seri and Chontal. His sources are essentially all the materials reviewed above with the addition of Seri material from HEWITT (1898). Among the nearly 200 items compared, roughly three-quarters are radical elements, 16 secondary verbs, 19 local suffixes, 10 other suffixes, and the rest pronouns. The labels refer to the synchronic status of the Yana elements, which always represent the first entry, followed by the comparable items in other

38

FROM HYPOTHESIS TO (NEAR) PROOF

languages. Copious documentation is given whenever available illustrating the item in its variant forms together with other relevant synchronic descriptive detail. Proto-forms for individual families, internally reconstructed ones for isolates, and even some tentative Hokan reconstructions are occasionally proposed. A complete appraisal of each of the suggested parallels is well beyond the scope of this work, and, as far as I know, has not yet been attempted by anyone. It is certainly long overdue; SAPIR'S work has earned either complete uncritical acceptance - not to mention emulation - or disbelief on the grounds of lack of solid descriptive data and methodological rigor. That it deserves better than that, however, is obvious to anyone who has dealt with any of the areas SAPIR pioneered. Nor did he harbor any illusions about the definitiveness of his work: It is only when we have something like an adequate knowledge of all the Hokan languages or dialectic groups that it will be possible to compile a relatively complete comparative Hokan dictionary and to study in detail the linguistic groupings and sub-groupings of these tribes. The material gathered together in the preceding sections enables one to determine (or at least suggest) a number of phonetic laws characteristic of one or other of the Hokan languages. Nevertheless it would be premature to attempt a systematic presentation of Hokan phonology. The evidence is far too scattered and scanty for this purpose. (SAPIR 1917b: 27)

Nevertheless, specific phonological processes are discussed, and are worth summarizing: (a) a frequent correspondence of Chimariko ο (less often u) to a in other Hokan languages; (b) the occurrence of initial vowels in some forms and their absence in other related ones, the alternation existing across languages and also within the same language; (c) frequent syncope of the second vowel of stems, resulting in consonant clusters which may then assimilate to each other. The first is a very specific correspondence which may point to a

FROM HYPOTHESIS TO (NEAR) PROOF

39

Chimariko innovation, but the other two are widespread phonological processes which, if they are to be reconstructed for ProtoHokan, can produce far-reaching changes in the structure of individual languages, and may go far in accounting for the very formidable differences among recorded Hokan languages. A typical entry will illustrate the methodology. I have chosen the set for 'water', since it was already used as an example of D I X O N and KROEBER'S presentation (cf. p. 29), and may provide a useful comparison: 55. Yana ha-na, xa-na "water" Kar. ac, ic "water" (from Hokan *a,xa); ca in ca-ruk "towards river" (literally "water-toward"; from Hokan (a)xa'-, cf. Pomo ui "eye" but compounded yu-, no. 128; parallel to ca-ruk is ma-ruk "away from river," originally "land-toward," cf. Chim. ama "earth") Shas. atsa "water"; New River Shasta ga-ats; Ach. (S) ac; Ats. atssi Chim. äqa (for Chim. q as equivalent to Hokan x, cf. nos. 53, 133), Ska; h-uso-xa "tears" (literally "eye-water") Ν., E., S. E. Pomo xa "water"; C., Ν. E. Pomo ka; S., S. W. Pomo aka Ess. asa-nax "water" Sal. (M) tccC "water" (analysis into t-ccC, according to Dr. Mason, is probable, but bare stem -ca' is not found) Moh. aha "water" Seri ax "water" Chon. aha "water" (SAPIR 1917b: 8) The phenomenon of vowel elision noted above is clearly illustrated by this set. The methodology underlying the decisions on what to compare may be illustrated by a brief quote on the topic of "local suffixes": Yana is characterized by a large number of local verb suffixes, resembling in this respect Karok, Shasta-Achomawi, and Chimariko. On the other hand, it does not possess the local postpositions suffixed to nouns which are found in these languages, also in Pomo and Esselen. To express prepositional relations Yana employs the same method as Salinan and Chumash, that is, the noun is preceded by an independent local noun of

40

FROM HYPOTHESIS TO (NEAR) PROOF

prepositional significance;[...] The close connection between local verb suffixes and local nouns in Yana is at least theoretical ground for suggesting that these suffixes, where they occur in Hokan, are nothing but incorporated local nouns that originally possessed complete independence, as still indicated by Chumash and Salinan. (SAPIR 1917b: 18-9)

The Coahuiltecan languages are included by SAPIR (1920a) in a paper bringing together some 120 items where parallels are found between the Hokan and Coahuiltecan languages. For Hokan, all previously used families are represented from the same original sources except Washo (this paper was obviously written before SAPIR became convinced that Washo had to be included), plus SWANTON'S data for Coahuiltecan (SWANTON 1915), supplemented by Tonkawa forms from GATSCHET (1877a). Comparisons include pronouns, personal nouns (kinship terms),body-part nouns, animals, objects, adjectives, numerals, verbs, adverbs, and grammatical elenients. These headings, which are SAPIR'S, are appropriate only for the English glosses, and no claim is made that the elements compared would be described as such in each language. The purpose I assume is to show that the parallels are to be found in all areas of lexical and semantic structure. Only sets of proposed cognates are listed with no attempt at reconstruction, and no one better than SAPIR was aware of the doubtful status of many of his sets. A certain amount of groping in the dark cannot well be avoided in the pioneer stage of such an attempt as this (SAPIR 1920a: 289).

Alternations of stems with initial vowels with forms where this vowel is absent are again noted, this time in the Coahuiltecan languages. Far from being dismayed by the geographical distance separating these from other Hokan languages, SAPIR is encouraged to note that the intervening area is occupied by Athapaskans, latecomers from the north, and Uto-Aztecans, somewhat earlier intruders from the south. A much better insight into SAPIR'S comparative technique can be gained from his next contribution (SAPIR 1920b), much more narrow and specific, but exposing his orientation very clearly. He

FROM HYPOTHESIS TO (NEAR) PROOF

41

starts from a consideration of the first person plural in Chimariko within the context of the Hokan hypothesis. From DIXON'S description of Chimariko, he gleans all relevant details on the first person (singular and plural) pronominal affixes, subjects them to a detailed process of internal reconstruction on the basis of a number of forms occurring in texts, and compares the units thus arrived at with pronominal elements of Salinan; these show great similarity to them, and even are subject to phonological processes of remarkable parallelism. The linguistic argumentation is impeccable, the illustrative material ample, and the demonstration is highly convincing. The only possible debatable question is the underlying assumption of genetic relationship in the first place, and even if this is rejected some explanation for the resemblances is obviously in order. Commenting in greater depth on MASON'S description of Salinan (MASON 1918), SAPIR (1920C) again singles out aspects of the structure of the language which are relevant to comparative Hokan and restates MASON'S description in this light, even bringing to the fore structural details which MASON did not stress, e.g., instances of final reduplication. SAPIR'S suggestions for improvement of the analysis of Salinan are based on comparative insights from other Hokan languages. They are very typical of his whole approach to historical linguistics. His methodology is grounded in the best of the comparative technique of Indo-European linguistics, boldly applied to cognate languages rather than to earlier (in this case nonexistent) documents. Soon thereafter (SAPIR 1921a) he adds to his previously compiled comparative vocabularies some additional Salinan items with presumed Hokan cognates, and more discussion of some structural features, in particular that Salinan is archaic in its preservation of initial vowels in stems. He also reiterates the Hokan character of Washo, which is revealed more obviously in deep morphological structure than in lexicon. Dealing again with prehistoric hypotheses based on linguistic and geographic considerations, he surmises an early Washo-Pomo-Yana continuum which came to be disrupted by Penutian intrusions. Finally, using as a base his previously proposed Hokan cognate

42

FROM HYPOTHESIS TO (NEAR) PROOF

sets, SAPIR (1925) now matches these with Subtiaba forms, pursuing a suggestion of LEHMANN (1920) (who had earlier (LEHMANN 1915) established the relationship of Subtiaba and Tlappanec) that Subtiaba is related to Washo. Part 1, "Hokan elements in Subtiaba", consists of the by now familiar comparative vocabulary. The practice of listing items under the main grammatical and semantic categories alphabetically by English gloss (already instituted by SAPIR 1920a) makes the material much easier to use than that of the 1917 article (SAPIR 1917b). An interesting result of this presentation is the appearance of multiple entries for the same English gloss, e.g., three entries each for 'day', 'come', 'go', 'sleep', suggesting original sets of semantically related stems. Again, no formal statement is made of sets of regular sound correspondences, although SAPIR was certainly operating with a working hypothesis of what these must be. It should be emphasized that this cannot be considered the result of an oversight. SAPIR never did avoid the postulation of "phonetic laws" if the data permitted such generalization. If correspondences are not given, it is simply because they could not as yet be fully demonstrated (a situation which has hardly improved to this day). Comments on precisely this aspect of SAPIR'S practice by VOEGELIN (1942) in his note "Sapir: insight and rigor" are directly relevant: [ . . . ] I have observed informally that many anthropologists are under the same impression: namely, that because Sapir had mastered and was increasingly contributing to Indo-European linguistics, he must have applied the comparative method in his work with American Indian languages distantly related. This is impossible, for two reasons. [ . . . ] It misses the distinction between the comparative method, a method requiring great rigor in its application, and what we may call for want of a better term "other comparisons," which would include comparisons of similar structural features and categories (it requires a special flair or insight to find these but they are easy enough to see once they have been pointed out); and comparisons of relatively infrequent lexical similarities between scattered languages (in the search for these ingenuity perhaps more than insight is needed).

FROM HYPOTHESIS TO (NEAR) PROOF

43

All of Sapir's work bringing together distantly related languages falls under the general category of "other comparisons," and not of the comparative method. Nor did Sapir ever pretend that the great reduction of language families was accomplished by the comparative method. (VOEGELIN 1 9 4 2 : 3 2 2 )

[··.] But, of course, Sapir made interpretations on the basis of his insightful work: and these resemble the interpretations made on the basis of his comparative work when only one language family was involved (Athabaskan or Uto-Aztecan). Now it was to the latter that Sapir brought all the rigor of the comparative method [ . . . ] (VOEGELIN 1942: 323)

From the items for which tentative reconstructions are provided, it is possible to extract a picture of the phonetic inventory which SAPIR envisaged for Proto-Hokan. There is a well-developed series of plain stops ρ t tc [c] k? and the skeleton of corresponding aspirated and globalized series p' k' f tc' ; a series of spirants s χ χ h; and the voiced resonants m η ?) w I y. Vowels are i a u. On the whole, a fairly simple and plausible system. As far as can be ascertained from the few well-documented examples, he assumes the reflexes of these, when attested at all, to be highly conservative. It is rather through broad recurrent phonological processes that forms are changed through time to achieve the highly disparate synchronic states. Part 2, "Notes on Subtiaba phonology", discusses the types of phonological processes that must be assumed to have operated during the history of Subtiaba in order to account for its idiosyncratic contemporary structure. These fall into two groups: (1) A number of very natural processes, of a type attested time and time again in the historical development of languages, such as: a > e when preceded by /; / > e when followed by α; α ο when followed by w; wa » o; kwa > ku'; a« > w" o r e ; ai > loss of unaccented final and medial vowels. In spite of the generality of these processes, it is startling to note that the full set (if not in the same environments) of these is also attested in the history of Yuman languages (e.g., LANGDON 1970C).

44

FROM HYPOTHESIS TO (NEAR) PROOF

(2) Some rather unusual developments attributable in part to areal tendencies, which would be more than likely to affect such an isolated language. It should be recalled that none of the surrounding languages are assumed to have any connection with Hokan whatsoever. Of interest are two parallel processes (*ak au >- o, and * at > ai > I") brought about, according to SAPIR, by a canonical restriction whereby final consonants are not allowed in Subtiaba. Later work ( R A D I N 1 9 3 2 - 3 3 ) reveals that Tlappanec (and therefore undoubtedly Subtiaba) is a tone language. I suggest that studying the interaction of the processes proposed by SAPIR with tonal phenomena is a profitable avenue of research. Another peculiarity of Subtiaba is the existence of prenasalized stops, some instances of which SAPIR assumes to have developed from plain stops in intervocalic position under the influence of neighboring languages. Part 3, "Notes on Subtiaba and Hokan morphology", is introduced by a statement often quoted out of its context: When one passes from a language to another that is only remotely related to it, say from English to Irish or from Haida to Hupa or from Yana to Salinan, one is overwhelmed at first by the great and obvious differences of grammatical structure. As one probes more deeply, however, significant resemblances are discovered which weigh far more in a genetic sense that the discrepancies that lie on the surface and that so often prove to be merely secondary dialectic developments which yield no very remote historical perspective. In the upshot it may appear, and frequently does appear, that the most important grammatical features of a given language and perhaps the bulk of what is conventionally called its grammar are of little value for the remoter comparison, which may rest largely on submerged features that are of only minor interest to a descriptive analysis. Those who find this a paradox think descriptively rather than historically. It would be an instructive experience in method to compare English grammar with that of the Indo-European language reconstructed by philologists. Whole departments of Indo-European grammar find no analogue in English, while a very large part of what English grammar there is is of such secondary growth as to have no relevance for Indo-European problems. (SAPIR 1925: 4 9 1 - 2 )

45

FROM HYPOTHESIS TO (NEAR) PROOF

The soundness of the argumentation leaves no doubt. It is, however, legitimate to ask how the submerged features SAPIR is looking for can be made to emerge, considering the scarcity of the data at hand and the lack of knowledge of even the more productive morphological processes in several of the languages discussed. The remaining part of Part 3 is devoted to a detailed demonstration of how this is done. There are no magic formulae, no mechanical procedures, but in the hands of an insightful investigator some highly convincing aspects of the morphological structure of ProtoHokan are uncovered. Most interesting is the reconstruction of a set of nominal and verbal prefixes, highly specific in form and content, and systematic in their internal organization. Among these the most abundantly attested are Proto-Hokan *t*m*k*p-

Subtiaba 'nominal, absolutive' 'adjectival' 'intransitive' 'transitive'

> > > >

dmg?

The Subtiaba evidence for the synchronic existence of these elements is full and convincing, even though none of these elements had been recognized by LEHMANN. I cannot resist offering from my own observations the fact that Diegueno (as well as other Yuman languages) has a verbal prefix m- which occurs mainly with stems denoting qualities. Stems of this type usually require an English adjective for their translation (LANGDON 1970b: 83).

No information on the structure of Yuman languages was available to SAPIR at the time. In the final section, "Conclusion: further vistas", SAPIR outlines what he believes to be even wider affinities for Hokan-Coahuiltecan by presenting the full membership of the Hokan-Siouan superstock, which had first been suggested in print in SAPIR (1921b), as follows:

46

FROM HYPOTHESIS TO (NEAR) PROOF

A. Hokan-Coahuiltecan I. Hokan proper 1. Northern Hokan a. Karok; Chimariko; Shasta-Achumawi b. Yana c. Pomo 2. Washo 3. Esselen; Yuman 4. Salinan; Chumash; Seri 5. Chontal II. Subtiaba (and Tlappanec) III. Coahuiltecan: Tonkawa; Coahuilteco-Cotoname-Comecrudo; Karankawa B. Yuki C. Keres D. Atakapa-Tunica-Chitimacha E. Iroquois-Caddoan F. Eastern group I. Natchez-Muskogean; Timucua II. Siouan III. Yuchi In content, this scheme is essentially identical to the HokanSiouan stock as it appears in SAPIR'S summary picture of the linguistic situation in North America (SAPIR 1929a). What is new in this scheme for Hokan-Coahuiltecan is the proposed subgrouping Northern Hokan, which is based in part at least on the distribution of some of the morphological elements discussed above: [ . . . ] Perhaps the most significant point about these absolutive noun affixes is the fact that the old prefix system tends to be best preserved in the south (Salinan, Chumash, Chontal, Subtiaba) and to be wanting or nearly so in the north (Yana, Chimariko, Karok, Shasta-Achomawi; prefixes hardly more than vestigial in Pomo). Esselen and Yuman seem to have a somewhat anomalous geographical position, being of the suffixing type. Washo is to be reckoned as belonging to the southern rather than the northern Hokan type. (SAPIR 1925: 504) As is wont to be the case with subgroupings, this is probably the weakest part of the whole scheme. But not until we know a lot

FROM HYPOTHESIS TO (NEAR) PROOF

47

more about the detailed history of Hokan languages in general, and specifically can identify such diagnostic facts as shared innovations, can we hope to do better. SAPIR'S parting comment is illuminating of his vision: [ . . . ] the real problems of American Indian linguistics have hardly been stated, let alone studied. (SAPIR 1925 : 527)

While this is undoubtedly true, the fact remains that is one of the giant steps in the direction he foresees.

SAPIR'S

work

4 F U R T H E R E X T E N S I O N S OF T H E STOCK

In this chapter, we will consider the effect which SAPIR'S proposals had on the views of those who consider it their business to ponder wide language relationships. First must be mentioned SAPIR'S own Hokan-Siouan (the full membership of which was listed on page 46), of which HokanCoahuiltecan is but a branch and which would link it to languages far to the east, among which are the Siouan and Iroquoian families. While Hokan-Coahuiltecan was substantiated by the series of articles reviewed in section 3.2,1 know of no corresponding evidence for Hokan-Siouan, though SAPIR certainly had such a project in mind: The evidence for this "Hokan-Siouan" construction is naturally morphological rather than lexical, though the lexical bonds that unite Natchez-Muskogian and Hokan, for instance, are by no means negligible. This evidence will be given in due time. (SAPIR 1925: 526)

It remains true, nevertheless, that SAPIR himself had reservations about some aspects of the whole scheme: Such a scheme must not be taken too literally. It is offered merely as a first step toward defining the issue, and it goes without saying that the status of several of these languages may have to be entirely restated. (SAPIR 1 9 2 5 : 5 2 6 )

On the other hand, he also believed that other languages would be found to be related: There is no reason whatever to believe that the "Hokan-Siouan" group as already defined will remain without further adjuncts in Mexico and Central America or perhaps even beyond. (SAPIR 1925 : 527)

FURTHER EXTENSIONS OF THE STOCK

49

The first to take up the challenge was RIVET who, in a paper dedicated to Edward SAPIR (RIVET 1 9 4 2 ) , suggest that Yurumangi, a language of Colombia, represents such an "adjunct". The paper is obviously modeled on SAPIR'S pioneering articles on HokanCoahuiltecan, gives a large number of supposed cognates and discusses morphological similarities. It also points out that Yurumangi is unlike any of the neighboring languages. RIVET'S suggestion must be assessed within the context of his philosophy of linguistic relationship which obviously allowed great daring, since he had earlier (RIVET 1 9 2 6 ) suggested genetic relationship between Malayo-Polynesian and some languages of South America. To my knowledge, no one has accepted the suggestion completely, though VOEGELIN - VOEGELIN ( 1 9 6 4 , 1 9 6 5 ) list Yurumangi as a language isolate possibly affiliated with Hokan. SWADESH, however (in Η AMP et alii 1 9 6 3 ) , states that it is related to Opaye and Chamicura. The status of this language, as that of so many others in South America, remains unconfirmed and will not be further referred to in this monograph. The following year, HARRINGTON (1943) proposes his own extension into South America by promoting Quechua (of Peru, Ecuador, and Bolivia) to membership in the Hokan stock. He was apparently unaware of RIVET'S suggestion concerning Yurumangi since he claims to be the first to find Hokan affinity in South America. The article is difficult to assess. While it purports to follow a SAPiRian approach Following Sapir's discernment that similarity of meaning should have precedence over similarity of sound as a guide in the assembling of forms for comparison [ . . . ] (HARRINGTON 1943: 334)

- and proceeds to perform the usual analysis, complete with internal reconstruction, the results are not very convincing, even though they contain a few striking similarities. SWADESH (1954a) considers Quechua more closely related to Aymara than to Hokan, a suggestion which is later accepted by VOEGELIN - VOEGELIN (1964, 1965).

50

FURTHER EXTENSIONS OF THE STOCK

have found no serious reference in later literature IOHARRINGTON'S suggestion just reviewed. Perhaps more serious and certainly better received was the proposal by GREENBERG - SWADESH ( 1 9 5 3 ) that Jicaque, a language of Honduras, is related to Hokan. It must be emphasized, however, that our knowledge of Jicaque is quite minimal; CONZEMJUS ( 1 9 2 1 - 2 3 ) , LEHMANN ( 1 9 2 0 ) , and HÄGEN ( 1 9 4 3 ) give nothing beyond word lists and very sketchy grammatical notes. The methodology of GREENBERG and SWADESH is that of lexicostatistics based on cognate counts; in the case of poorly recorded languages, at best only distantly related, the assumption of cognation must perforce be based on very superficial inspection, and an evaluation of results must take these reservations into account. Jicaque is nevertheless included on the map of Hokan-Coahuiltecan languages appended to this volume to reflect the fact that VOEGELIN - VOEGELIN ( 1 9 6 4 , 1965) include it in their Hokan phylum and their 1 9 6 6 map reflects this decision (VOEGELIN - VOEGELIN 1966). One remark must be made with regard to both the HARRINGTON and the GREENBERG and SWADESH proposals, namely that they use the term Hokan to mean 'Hokan-Siouan'. Thus HARRINGTON includes Choctaw (a language of the Muskogean family) in his comparison, and GREENBERG and SWADESH freely use data from the Gulf languages, i.e., Muskogean, Natchez, Tunica, Chitimacha, and Atakapa. In a footnote to his published bibliography of Hokan-Coahuiltecan languages, BRIGHT (1955) reveals his conviction that Naolan (in Tamaulipas, Mexico) is related to Hokan-Coahuiltecan, on the basis of his perusal of the material in WEITLANER (1948), in spite of the latter author's preference for a link with Uto-Aztecan. Once more, the suggestion is interesting and additional study is in order. The language was reported near extinction. SWADESH (1967a) gives some evidence for affinity of Naolan to Tonkawa, of which more below. In a series of articles ( 1 9 6 3 - 1 9 6 8 ) , GURSKY makes various suggestions for extending the Hokan network. These are of two kinds: I

FURTHER EXTENSIONS OF THE STOCK

51

(1) the addition of individual extinct languages to HokanCoahuiltecan, i.e., Quinigua of Northeastern Mexico and Waicuri, of southern Baja California, Mexico. The data available for both languages are again scanty and unanalyzed and many of the compared items may be spurious cognates (of which GURSKY is well aware). At least some data on these languages can now be found in more readily available publications as a result. (2) More ambitious is the second type of conjecture, where GURSKY links Hokan to Algonkian-Gulf, extrapolating from HAAS ( 1 9 5 8 , 1959, 1960, 1 9 6 7 ) who found evidence linking Tonkawa and the Gulf languages to Algonkian. What HAAS attempted to demonstrate (e.g., HAAS 1 9 6 0 ) is stated very clearly: (1) to validate the Algonkian-Ritwan connection, (2) to show that the possibility of an Algonkian-Mosan affiliation merits further investigation, (3) to show that the Gulf languages and Tonkawa are also related to Algonkian, and (4) to suggest that all these languages are probably related to one another. (HAAS 1960: 988-9) And, on the status of Tonkawa specifically, she concludes: We are forced to conclude that the final word about the relationships of Tonkawa has not yet been said. Progress has been made recently, however, by attempts to make new comparisons not only within Sapir's Hokan-Coahuiltecan but more particularly outside of it. It is also clear that the problem of classifying Tonkawa is bound up with the problem of classifying the languages of North America in toto. Tonkawa remains, then, a discrete language with no known close relatives. It is not quite as lonely as Basque or even, perhaps, as Zuni. But the clouds which hover over its past linguistic connections are not dispersing as readily as we might have hoped. The isolated languages of the world still remain one of linguistics' greatest challenges. It is important to recognize Tonkawa as one of these challenges [ . . . ] (HAAS 1967: 318). The conservative conclusion to be drawn from all this is that, if Tonkawa is related to Hokan-Coahuiltecan, it is in a very distant and as yet unclear manner.

52

FURTHER EXTENSIONS OF THE STOCK

using lexicostatistics as a methodology and the notion of "chain" relationships as a model, provides a totally different picture of the affiliations of the languages we are concerned with. Thus (SWADESH 1967a, 1967b) he sees Subtiaba-Tlappanec as equally related to Hokan-Coahuiltecan and to Manguean, and Coahuiltecan fits into a Yumapakwan network including among others Washo, Esselen, Salinan, Chumashan, Yuman, Seri, Pakwan (Coahuiltec, Comecrudo, Cotoname, Quinigua, Karankawa, and Tequistlatec). Tonkawa, which he considers separate from Coahuiltecan on the stock level, belongs here too, and a closer affinity of Tonkawa with Naolan is postulated. The whole network ultimately has connections with Mayan and Uto-Aztecan and is called the Macro-Mayan Phylum. Recently, LANDAR (1968) has attempted to link Hokan to Carib using Karankawa as the intermediary. Finally, I have been told of a suggestion by RENSCH (1969) that Tlappanec is Otomanguean and that this may constitute a bridge between Hokan and Otomanguean. In short, suggestions abound and the challenge cannot be ignored. Further work is obviously imperative and we may look forward to interesting results. Perhaps, when all is said and done, we will be somewhat closer to dealing with the question of what constitutes proof of genetic relationship. SWADESH,

5 TOWARD PROOF A N D E S T A B L I S H M E N T

5.1 T H R O U G H T H E

FORTIES

In the meantime, decriptive data were accumulating, including first SAPIR'S own extensive work on the structure of Yana (SAPIR 1909, 1910, 1916, 1918, 1922, 1923, 1929b), the Hokan language he knew best and from which he drew the most insightful leads to deep comparisons. Other investigators were also beginning to show interest in these languages, some for purely linguistic purposes, others only incidentally within the framework of ethnographic research. The contributions of a synchronic nature not mentioned previously and extending through the 1940's are reviewed in this section. Kinship terminologies are compiled by GIFFORD (1922) for all of California; for Hokan, they include: Shasta; Achumawi; Atsugewi; Northern Yana and Yahi; Karok; Chimariko; all seven Pomo languages; Esselen; Antoniano and Migueleno Salinan; Obispeno, Ineseno, Barbareno and Island Chumash; Northern and Southern Diegueno; Yuma; Mohave; Cocopa. Kinship terms for southern Mexico, including Chontal, are contributed by de ANGULO (1925b). Texts are added for Karok (HARRINGTON 1930,1932; de ANGULO - FREELAND 1931) and for Eastern Pomo (de ANGULO 1927). Interest in Achumawi evokes contributions of grammatical information and texts (de ANGULO 1 9 2 6 ; de ANGULO - FREELAND 1930), phonetic description (ULDALL 1 9 3 2 - 3 3 ) , and discussion of dialects (VOEGELIN 1946). A short controversy flares when MERRIAM ( 1 9 2 6 , 1930), interesting himself in the Shasta-Achumawi territory, claims to have discovered a new Shasta language and is sharply

54 rebuked by

TOWARD PROOF AND ESTABLISHMENT

DIXON ( 1 9 3 1 ) .

Atsugewi kinship terms are found in

GARTH ( 1 9 4 4 ) .

Washo receives a new grammatical description and texts (DANGas well as ethnographic notes (LOWIE 1 9 3 9 ) . The more remote languages of Mexico become slightly more accessible through de ANGULO - FREELAND (1925), who contribute grammatical notes, vocabulary, and short texts for the "mountain" dialect of Chontal, a welcome supplement to an early short word list by BRINTON (1892) and the less readily accessible description of BELMAR (1900, 1905). De ANGULO (1925a) asserts the Hokan affinity of Chontal and promises later substantiation, but to my knowledge this proof never appeared. Grammatical notes, vocabularies, and short texts on Tlappanec come from RADIN (1932-33) and SCHULTZE-JENA (1938), the latter quite negatively reviewed by RADIN (1940). WEITLANER - WEITLANER DE JOHNSON (1943) compare their own word lists from two dialects of Tlappanec with some of LEHMANN'S early ones on Subtiaba. SWANTON (1940) publishes all data available to him on Coahuiltecan languages. Some data on Sen are included in KROEBER'S ethnographic monograph (KROEBER 1931); Cochimi, Yuma, and Cocopa word lists are also appended for comparison with Seri. The work of HOIJER (1933, 1946b, 1949a, 1949b) on Tonkawa is of major importance. He gives us a grammar, a restatement of the grammar in the form of a structural sketch, a discussion of syntactic suffixes and anaphoric particles, and a dictionary. Since the language has since become extinct, we must BERG 1 9 2 2 , 1 9 2 7 )

[ . . . ] express our gratitude to Harry Hoijer for having saved for us the most important body of material we have, or will ever have, on this language. (HAAS 1967: 318)

No better tribute can be paid to this work than to notice the amount of interest expressed recently in the position of this language (and reviewed in the preceding chapter), though it is ironic that Tonkawa has become the language whose Hokan-Coahuiltecan affiliation has most strongly been doubted. In that respect, HOIJER'S scepticism

TOWARD PROOF AND ESTABLISHMENT

55

with regard to the broader claims of genetic relationship (e.g., HOIJER 1946a: 10) appears justified. A good deal of additional data was accumulating about several of the Yuman languages. First, linguistic forms of various degrees of accuracy are interspersed in ethnographic descriptions of Diegueno (WATERMAN 1 9 1 0 , SPIER 1 9 2 3 ) , Kamia - a dialect of Diegueno - (GIFFORD 1 9 3 1 ) , Akwa'ala (Paipai) (GIFFORD - LOWIE 1 9 2 8 ) , Kiliwa (MEIGS 1 9 3 9 ) , and Yavapai "(GIFFORD 1 9 3 2 , 1 9 3 6 ) . Of greater linguistic interest are SPIER'S contributions, consisting of a short text in Havasupai and parallel texts and vocabularies in Maricopa and Havasupai (SPIER 1 9 2 4 , 1 9 4 6 ) . This is the first textual material in any Yuman language (of more than Lord's Prayer length) to appear in print, and, in spite of a lack of analysis and systematic notation, can be of use to a reader familiar with Yuman languages. The main contribution to Yuman scholarship, however, is the work of HALPERN (1942,1946a, 1946b, 1947) on the Yuma language, comprising a full linguistic description of kinship terms, a detailed account of the structure of the language, and a short restatement of the longer grammar. The accuracy of his data, the wealth of examples, the listing of exceptions, and the deep insights into the underlying structure of the language all contribute to making this work the indispensable basic reference for Yuman. The excellence of this description has not been widely acknowledged for a variety of reasons, among which we may note the fact that the more detailed account was published as a series of articles (a form of scholarship not often reviewed), that the shorter sketch appeared in a volume (HOIJER et alii 1946) containing twelve other parallel endeavors by such famous Americanists as BLOOMFIELD, HOIJER, HAAS, etc., that a description in the tradition of SAPIR was not in fashion in the forties, and that the casual observer may miss the deeper insights often humbly concealed among the wealth of detail. Finally, MASSEY (1949) attempts to clarify the aboriginal linguistic and tribal constitution of the Baja California peninsula. The situation was obviously complex and is further complicated by the

56

TOWARD PROOF AND ESTABLISHMENT

fact that all information must perforce be obtained from early documents - all native groups of the area south of the Kiliwa are extinct. Most languages have died without any recorded data, but secondary information reporting lack of mutual intelligibility indicates a good deal of linguistic diversity. In the area south of Cochimi - for which some linguistic data are available in the form of word lists and which is either Yuman or closely related - the scanty data in existence indicate that if the languages are related to Hokan-Coahuiltecan, the proof may never be forthcoming. MASSEY'S survey is the most accessible summary of a difficult area.

5.2 T H E S Y M P O S I U M ON A M E R I C A N

INDIAN

L A N G U A G E S AND "THE S U R V E Y "

As we have seen, the thirties and forties do not concern themselves much with Hokan-Coahuiltecan as such. They rather see the slow consolidation of the basic data available to researchers, the two main events being the publication of full-scale descriptions of Tonkawa and Yuma. The intensity of the early decades of the century, culminating in SAPIR'S vast classificatory scheme seems to have left the Americanist world stunned, but gratefully accepting the new order out of chaos. With the beginning of the fifties, a new trend begins, once more with a focus at the University of California, Berkeley. At a Symposium on American Indian Languages held in conjunction with the Linguistic Institute at Berkeley in the summer of 1951, and organized by Professor Murray Β. ΕΜΕΝΕ AU, the participants chose as the theme for their discussions to sum up the past of this field of linguistic scholarship and point the way

for its future (CHRETIEN et alii

1954: 1).

In a volume issued in 1954 publishing six of the papers presented at this Symposium, the editors ( C . D . CHRETIEN, M.S. BEELER, M . B . EMENEAU, and M . R . HAAS) emphasize the "programmatic"

TOWARD PROOF AND ESTABLISHMENT

57

nature of the contributions and spell out the pressing needs for future research and the order of priorities: The needs just stressed are perhaps, for North America, at their most urgent on the Pacific Coast and not least so in California. The HokanSiouan superstock received detailed comparative treatment in three of the Symposium papers, two of which are published here. The California part of this superstock was long ago surveyed, but the mere number of languages concerned, in this stock as well as in the other stocks spoken in the state, has prevented the descriptive work in California from being brought to a successful conclusion. The salvaging of all possible data, of all stocks, must be a lively concern of linguistic scholars at the University of California during the next generation, [emphasis mine] At the end of that time practically all the California languages now still spoken will be extinct or, at least, will no longer be "going concerns". Some of them are already on the verge of extinction, being carried by only a few speakers, or even by only one. These must be recorded (for the last or only time) at the earliest possible moment. The Board of Editors of the Series and the Group in Linguistics, University of California, Berkeley, regard work on the California languages as peculiarly their responsibility [ . . . ] (CHR£TIEN et alii 1954 : 2)

For Hokan-Coahuiltecan, we may recall that, in spite of the feverish activity of the early days, a truly detailed analysis of the structure of the languages in question had been performed only by SAPIR on Yana, and even here no full-fledged grammar had appeared. The only other California Hokan language adequately described by then was Yuma, straddling the California-Arizona border. The concern of the editors of the Symposium papers is clearly justified. The implementation of their program will concern us in the remaining sections of this monograph. Rarely is a programmatic statement rewarded with the success we are about to record. The reasons are many. Obviously, the time was ripe, linguistics was reaching recognition as an academic discipline, often severing the ties which had kept it a sub-discipline of anthropology but retaining the anthropological orientation of its origins. New theoretical ideas and a refinement of methodology were crying out to be tested and applied, and it is not surprising that many a graduate student

58

T O W A R D PROOF A N D ESTABLISHMENT

answered the call of the field. But most important was the foresight which led not only to the specification of a clear and urgent program, but to the establishment of a source of funds which would support the field expenses. So it came about that the year 1953 saw not only the formation of the new Department of Linguistics at the University of California, Berkeley, but also the establishment of the "Survey of California Indian languages", a fund which has supported a vast amount of field work since. Its scope has recently been expanded beyond California and it is now known as the "Survey of Californian and other Indian languages".3 Coming back to the Proceedings of the Symposium ( C H R E T I E N et alii 1954), we note that the only two papers dealing with a specific language family concern aspects of the Hokan-Coahuiltecan group. In a first step toward rectifying the fact that Although thirty years have now elapsed since Sapir's earliest suggestion concerning this superstock [Hokan-Siouan], no serious attempt has been made in the interval either to prove or to disprove the validity of his theory. ( H A A S 1954: 57) H A A S considers a single set of forms, the words for 'water' in the Hokan-Coahuiltecan languages, proposes a Hokan-Coahuiltecan reconstruction, and then demonstrates that it can be meaningfully compared to her previously ( H A A S 1 9 5 1 ) proposed reconstructions of the same word for Proto-Gulf and Proto-Siouan, in order to show that

the whole problem of the interrelationships of the Hokan-Siouan languages definitely merits further investigation ( H A A S 1954: 57). 3

Among the publications reviewed in this monograph, those of the following authors are based on field work at least partially supported by the "Survey": APPLEGATE, BEELER, BRIGHT, James and Judith CRAWFORD, GREKOFF, HAAS, JACOBSEN, LANGDON, MCLENDON, MIXCO, MOSHINSKY, OLMSTED, OSWALT, SHATERIAN, SILVER, TAL MY, VIHMAN.

TOWARD PROOF AND ESTABLISHMENT

59

She proceeds to reconstruct first Proto-Coahuiltecan, then ProtoSubtiaba-Tlappanec and Proto-Hokan, and finally by comparing these three to reconstruct Proto-Hokan-Coahuiltecan. The word for 'water', we may recall, was a favorite of previous investigators as well, and we have given their results in previous sections of this monograph. It is therefore of interest to also list the set of forms and the reconstruction of HAAS for the sake of comparison. 'water' Comecrudo: ax Cotoname: ax Tonkawa: ^a-x Proto-Coahuiltecan: *ax Subtiaba: i'ya Tlappanec: i'ya ? Proto-Subtiaba-Tlappanec: *ixya, possibly from *axyi Karok: ?a\s 'water'; sa- 'toward the river' Chimariko: aqa 'water' Shasta: ?äcca < * ?äca Achomawi: as Atsugewi: acsi (atssi) (< *aci cf. Shasta) Yana: xa-na Pomo: (N, E, SE) xa; (C, NE) ka; (S, SW) aka Washo: asa 'to urinate' ( < 'to water' ?) Esselen: asa-nax 'water' Salinan: tSa ? (< t-+~$a ??) Yuman: ?αχά Seri: ax Tequistlatecan (Chontal): aha Proto-Hokan: *axa (SAPIR), or possibly *axi Proto-Hokan-Coahuiltecan: *axi, or possibly * axil ~ *ixa

60

TOWARD PROOF AND ESTABLISHMENT

Notable by its absence is Karankawa, which HAAS considers on the basis of the evidence for this set, to be a better candidate for comparison with Proto-Gulf. VOEGELIN - VOEGELIN ( 1 9 6 6 ) list Karankawa as a totally unaffiliated language isolate. In the 1954 volume, BRIGHT, using SAPIR'S hypothesized sub-group Northern Hokan (a), i.e., Karok, Chimariko, Shasta-Achumawi, as a convenient set of languages to consider in view of his own newly collected data on two of them (Karok and Shasta), proposes sets of sound correspondences of the apical sonorant phonemes of the Northern Hokan (a) languages (BRIGHT 1954: 64).

There are three sets, reconstructed as *n, */, and *«, the latter a purely arbitrary symbol to differentiate it from *n, with which it shares reflexes in Chimariko and Shasta. Over twenty sets of words (not complete for all languages) are exhibited to substantiate the correspondences, and the evidence for each, though not overwhelming, looks quite solid. The situation is complicated by what BRIGHT simply calls "additional correspondences" (BRIGHT 1 9 5 4 : 65) in words which certainly appear to be cognates but where these same phonological types are differently distributed in the languages in question. He aptly points out that r and η participate in morphophonemic alternations which cannot be ignored. He further postulates a reconstructed stop system with contrastive sets in labial, apical, and velar position, consisting of plain, aspirated, and glottalized phonemes. The word for 'two' is reconstructed as *xaku (from Karok ?äxxak and its combining form xak-, Chimariko xoku, Shasta xiikkwa?, Achumawi hag, and Atsugewi hokki). Comparing these two reconstructions (for 'water' and 'two'), we note they both contain a reconstructed *x, but that the reflexes are not the same in the languages considered. The additional assumption must be made that the vowel *i which HAAS reconstructs is responsible for the different reflexes of *Λ: in the forms for 'water' in Karok, Shasta, Achumawi, and Atsugewi. While fronting under the influence of a front vowel is a well-attested

TOWARD PROOF AND ESTABLISHMENT

61

process, additional instances of the same sets would be welcome to allow the convincing grouping of the two sets of reflexes under a single reconstruction. But such are the facts of comparative Hokan and the main reason why progress has been so slow. Exactly ten years after the Symposium, at the 1961 meeting of the Southwestern Anthropological Association, a session was devoted to Californian linguistics. As an outgrowth of this session, there appeared another collection of papers (BRIGHT 1964) entitled Studies in Californian linguistics reporting on work initiated in the intervening decade. The volume is appropriately dedicated to Professor EMENEAU, whose leading role in this florescence was noted above. The extent of the progress is summarized by the editor: The result has been that, of the 27 languages listed in 1951 as requiring investigation, 17 have now been studied by competent field workers. Since 1951, teachers and students of the University of California have published over 30 articles and monographs on the languages of the state. (BRIGHT 1964: viii)

Those results which are relevant to Hokan-Coahuiltecan will be reported on in the next sections, which bring our survey to the present. If the year 1964 seems predominant in the contributions, this is because, of the 19 articles in BRIGHT (1964), nine are concerned with aspects of the Hokan-Coahuiltecan story.

5.3 R E C E N T D E V E L O P M E N T S I N E A C H F A M I L Y

book-length description of Karok (BRIGHT 1957) represents the first fruits of Survey-sponsored field work. In a preface to the volume, the Editors of the University of California Publications in Linguistics hail it as a model of the minimal treatment they hope each remaining Californian language will receive and the goal of "a complete grammar, texts, and a full lexicon" (BRIGHT 1957: v) remains the main one field workers sponsored by the Survey have BRIGHT'S

62

TOWARD PROOF AND ESTABLISHMENT

set for themselves since that time. Also resulting from the same period of field work on Karok are BRIGHT'S two articles (BRIGHT 1952, 1967b) dealing with aspects of borrowing, as well as, at one remove, a restatement of BRIGHT'S description of Karok vowels, by HAMP (1958). Most important, however, are the increasing number of comparative studies (reviewed below) which rely on BRIGHT'S Karok data. It is heartening to report that Shasta (as good as extinct at this writing) received a detailed description by SILVER (1966b). Earlier, a Shasta vocabulary had been provided by BRIGHT - OLMSTED (1959). The extensive and as yet unpublished material collected by SILVER will constitute the major source of information on this language. Chimariko, though extinct, is not irrevocably lost. There are field notes of SAPIR'S in existence and also a vast body of field notes by the indefatigable HARRINGTON. A description of Chimariko based on the HARRINGTON material is in progress by GREKOFF, and several unpublished chapters (GREKOFF s.a.) dealing with phonology are on file with the Department of Linguistics of the University of California, Berkeley. Considerable field work has been done on languages of thePalaihnihan family, the main contributor being OLMSTED, to whom we owe descriptive articles on Atsugewi (OLMSTED 1958a, 1961), an Achumawi dictionary (OLMSTED 1966), notes on the non-reciprocal intelligibility between the two languages (OLMSTED 1954), and finally a reconstruction of Proto-Palaihnihan phonology (OLMSTED 1964). The latter is an important work in the history of Hokan-Coahuiltecan since it represents the first systematic application of the comparative method to one of the families involved. It proposes reconstructed phonological segments with data on the number of instances of the correspondence described, and also provides a lexicon of 205 items (consisting of full words, stems, and affixes), complete with reconstructions and cross-reference to the specific correspondences in each item. OLMSTED (1964: 11) disclaims that his "treatment is final or complete", but does claim that he is

TOWARD PROOF AND ESTABLISHMENT

63

reconstructing the phonemes of Proto-Palaihnihan. His methodology is rigorous; it is a classic example of the rigid application of the comparative method. One cannot help but wonder, however, whether the full inventory of 28 consonants and 28 vowels represents contrastive phonological units of Proto-Palaihnihan, particularly in the case of the vowels, where OLMSTED'S own description for the daughter languages recognizes no more than 12. Additional work using internal reconstruction is an obvious next step to be hoped for. An early Atsugewi word list collected by KROEBER in 1 9 0 0 was not published till 1 9 5 8 (KROEBER 1 9 5 8 ) . Supplementing SAPIR'S published work on Yana, the Yana dictionary compiled by SWADESH on the basis of SAPIR'S Yana notes (SAPIR - SWADESH 1960) contributes a grammatical sketch, dialect notes, and Yana-Hokan notes, and lists items in all dialects with cross-reference to SAPIR'S published texts. Its value for comparative Hokan studies is beginning to emerge as new studies using its data are appearing regularly. Intensive fieldwork on several of the Pomo languages has already resulted in a number of publications (see below), several contributions are in press, and the vast quantity of recorded data promises to afford a detailed knowledge of the structure of these languages. Southwestern Pomo (Kashaya), through the able efforts of OSWALT, is the Pomo language best described to date. We have a grammar (OSWALT 1961), texts (OSWALT 1964C), and shorter articles (OSWALT 1958, I960, 1971a, 1971b). One more article (WORTH 1960) deals with the origin of a borrowed word in Southwestern Pomo. Eastern Pomo is a close second; the field work of MCLENDON is bearing fruit in the form of a grammar (MCLENDON 1966), and a study of Spanish loans (MCLENDON 1969); more is obviously to come. A description of Southeastern Pomo (MOSHINSKY 1970b) has just been completed, and current field work on Northern Pomo by VIHMAN will ultimately lead to equivalent descriptive work. Less extensive field work but ranging over all Pomo languages was undertaken by HALPERN, who reports (HALPERN 1962, 1964) on his findings and specifically proposes his

64

TOWARD PROOF AND ESTABLISHMENT

own view of the internal relationships, in which Southeastern Pomo is most divergent, then Eastern Pomo, while the otherfivelanguages form a Russian River group, in which the precise subgrouping is not clear. Remarkable resemblances between Pomo and Yuman (HALPERN has worked on both) are pointed out. The peculiarly Hokan characteristic of elusive first-syllable vowels is examined in some detail for the Pomo languages by GREKOFF (1964), who discovers several general patterns of vowel elision in different syllable types. Southwestern and Central Pomo are compared by OSWALT (1964a), and Kashaya (Southwestern Pomo) is found to have retained some archaic features lost in Central, such as firstsyllable vowels, whereas Central has innovated a contrastive high tone. OSWALT also addresses himself to the internal relationships of the Pomo languages (OSWALT 1964b), his conclusions differing somewhat from HALPERN'S in that he sees Kashaya, Southern, and Central most closely related to each other in a Southern group, which, in turn, combined with Northern, constitutes the Western branch; the three eastern languages do not combine in any meaningful way. The main disagreement with HALPERN is the position of Northeastern, which HALPERN had classified with OSWALT'S Western branch. In addition, the paper offers in print the 100-word diagnostic list for all Pomo languages, illustrating in full the data on which the classification was based. Recurrent sound correspondences for all the Pomo languages are given preliminary treatment in WEBB et alii (1965-66) and more fully in WEBB (1971). WEBB, a student of OLMSTED, models her historical work on his, applies the comparative method, discusses sound correspondences, gives a comparative vocabulary complete with reconstructions for 239 items, and, in addition, proposes some timedepth calculations on the basis of cognate counts (the greatest distance suggesting Proto-Pomo times is approximately 1700 years ago) and yet another attempt at subgrouping which shares with HALPERN a Russian River group, but considers Southeastern and Eastern as constituting a separate subgroup called Lake Pomo. The problem is obviously still unresolved, but additional work on

TOWARD PROOF AND ESTABLISHMENT

65

the reconstruction of Proto-Pomo is in press and we may look forward in the very near future to a sophisticated understanding of the structure of Proto-Pomo. An admirable description of Washo (JACOBSEN 1964) will hopefully be supplemented in future publications by the same author so that the benefits of his extensive field work on this language may accrue to all. Washo texts were published posthumously under the name of LOWIE (1963) as a supplement to his ethnographic monograph (LOWIE 1939). Bibliographic sources are reviewed by d'AzEVEDO - PRICE (1963) for all works dealing with the Washo tribe, and the linguistic literature is covered by JACOBSEN (1966). A restatement of the role of reduplication in the language based on JACOBSEN'S description is proposed by WINTER (1970a), in an attempt to unravel the historical origin of reduplication in Washo. Salinan and Esselen fare much less well, the only recent addition to our knowledge being the publication of Mission Indian vocabularies by HEIZER ( 1 9 5 2 ) . JACOBSEN has Salinan material which he collected from some of the last remaining speakers; it is hoped he will make it available in the future. Field notes on Salinan and Esselen by the ubiquitous HARRINGTON may some day prove amenable to analysis, though to my knowledge no one is attempting this at the moment. The vocabularies published in HEIZER (1952, 1955) also contain Chumash data. In addition, the philological investigations of BEELER and his students based on the analysis of early documents in Chumash languages, coupled with the unexpected opportunity he encountered to work directly with the last remaining speaker of Barbareno Chumash, are producing results such that the structure of Chumash languages no longer need remain a mystery. On Ventureno, we have a treatment of numerals (BEELER 1964) and the publication and analysis of the Confesionario of Father Jos£ Senän, dating from the early years of the nineteenth century (BEELER 1967). For Barbareno, there is an article on sibilant harmony (BEELER 1970a) and much forthcoming. Unpublished field notes of HARRINGTON'S are being analyzed by APPLEGATE and others, and

66

TOWARD PROOF AND ESTABLISHMENT

preliminary results are already appearing, i.e., a paper on vowel harmony (APPLEGATE 1971), with more to come. Following a very lean period immediately after HALPERN'S description of Yuma, our knowledge of Yuman languages is now increasing rapidly. From the starting point of WINTER'S first impressions (WINTER 1957), and BIGGS' study of intelligibility among the Yuman languages (BIGGS 1957), we encounter next a formal analysis of Yuman kinship systems (VOORHEES 1959) and MIERAU'S notes on Yavapai-Apache bilingualism (MIERAU 1963). We then progress to monograph-length descriptions of specific languages, i.e., Havasupai (SEIDEN 1963), Walapai - a bonanza, two descriptions (REDDEN 1966, WINTER 1966), Paipai (JOEL 1966), Cocopa (CRAWFORD 1966), the Mesa Grande dialect of Diegueno (LANGDON 1970b), Kiliwa (Mixco 1971), Yavapai (SHATERIAN 1971). In addition, there are recent still unpublished extensive notes on Mohave (Judith CRAWFORD) and Maricopa (Barry ALPHER). All surviving Yuman languages are thus receiving close attention. Comparative word lists of Yuman dialects in various locations in northern Baja California are found in ROBLES (1965), and a sketch of Diegueno phonemics (the Inaja dialect) in BRIGHT (1965b). More anthropologically oriented studies based on linguistic data are FRISCH - SCHUTZ (1967), ROMNEY (1967), FRISCH (1968), ALMSTEDT (1968), and there is a delightful study of Cocopa baby talk by CRAWFORD (1970a). Back in 1943, KROEBER had proposed a classification of the Yuman languages (KROEBER 1943) based on comparative vocabularies mostly collected by himself and GIFFORD, but in some cases rather sketchy. In addition to publishing the comparative vocabulary, he gives phonetic and lexical justification for four subgroups, and a sketch of probable Proto-Yuman phonetics. The suggested groupings are: Delta group: Cocopa, Kahwan, Halyikwamai (all but Cocopa now extinct) River group: Yuma, Maricopa, Mohave

TOWARD PROOF AND ESTABLISHMENT

67

Arizona group: Walapai, Yavapai, Havasupai California group: Diegueno, Kamia, Paipai, Kiliwa The geographic bias of the classification is clear from the labels and, in the case of the California group, obviously not linguistically accurate, since KROEBER himself discusses similarities between Paipai and his Arizona group, and notes that Kiliwa is divergent. It is precisely this aspect of the classification that JOEL (1964) reexamines on the basis of her own extensive recent data on Paipai, Kiliwa, and adjacent Diegueno dialects. Her groupings differ from KROEBER'S mainly in first setting Kiliwa apart from all other languages and then setting up Paipai as a separate subgroup on an equal footing with KROEBER'S other four. This is more in keeping with the linguistic facts. She also substantiates her conclusion by exhibiting a chart of nine phonetic correspondences. In a monograph proposing a first approximation of the Proto-Yuman consonant system, WARES (1968) gives yet another internal classification of the Yuman languages. Noting once more (like KROEBER and JOEL) the linguistic similarities between the languages of Arizona and Paipai of Baja California, he takes the formal step of putting them together in a single subgroup which he calls somewhat inappropriately Northern Yuman; his Central Yuman is KROEBER and JOEL'S River group (Yuma, Maricopa, Mohave), his Delta-California combines Cocopa and Diegueno, and Kiliwa is a separate fourth group. The last word on the subject is obviously yet to come. Further steps in the reconstruction of Proto-Yuman are LANGDON (1970a), reviewing and slightly amending WARES' consonantal system, a reconstruction of the demonstrative system (LANGDON 1968), and a discussion of the phenomenon of sound symbolism in various Yuman languages (LANGDON 1971). A number of forthcoming contributions deal with the subject as well, and it is clear that we will have in the near future a fairly detailed view of Proto-Yuman. Our survey would not be complete without acknowledging the contribution of another institution, the Summer Institute of Linguistics, whose members keep adding to our knowledge of

68

TOWARD PROOF AND ESTABLISHMENT

many languages in far-off lands. In the Hokan story, much work on the extant languages of Mexico is being conducted under their auspices and is reviewed below.4 The extensive work of Edward and Mary MOSER on Seri is finally beginning to appear in print, and it is heartening to report (E. MOSER 1961, 1963a) a thesis on "Number in Seri verbs" - an extremely complex topic-and a text; a dictionary ( M O S E R - M O S E R 1961) and an article on phonological patterns ( M O S E R - M O S E R 1965). A large number of Seri lexical items are also included in their ethnographic contributions (E. MOSER 1963b, M . MOSER 1964, B O W E N MOSER 1968). The philological investigations of TROIKE into the early manuscripts containing the only material we will ever have on Coahuilteco make this available to linguists in a form maximally useful for comparative purposes (TROIKE 1959, 1961, 1963, 1967b). Ongoing work on Oaxaca Chontal is now reported in a number of publications drawing on the deep knowledge of the language acquired by WATERHOUSE, who contributes descriptive articles (WATERHOUSE - MORRISON 1950; WATERHOUSE 1949a, 1957), a monograph-length description (WATERHOUSE 1962), a short sketch (WATERHOUSE 1967), a dictionary (WATERHOUSE - PARROTT 1970), a description of the kinship system (WATERHOUSE - MERRIFIELD 1968), observations on bilingualism (WATERHOUSE 1949b), and a discussion of some problems of Proto-Chontal (WATERHOUSE 1969). TURNER describes the highland dialect (TURNER 1967b, 1968) and reconstructs the Proto-Chontal phonemic system (TURNER 1969). Just published is a dictionary of Highland Chontal (TURNER TURNER 1971). Kinship terminology is also analyzed by OLMSTED (1958b) and LANDAR (1960); OLMSTED (1967) discusses ceremonial texts. Since the language is still spoken by a large number of people, it will remain an invaluable source of new material. Nothing new can be reported for extinct Subtiaba, but Tlappanec 4

Most of the publications of the following authors originated under the auspices of the Summer Institute of Linguistics: MOSER and MOSER, W A T E R HOUSE and coworkers, T U R N E R .

TOWARD PROOF AND ESTABLISHMENT

69

is still spoken and is being extensively studied by Juan H.V. and Mildred LEMLEY. SO far nothing but a very short article (LEMLEY 1955) has appeared, but much is forthcoming. The availability of descriptive materials on this language in the near future will be an event of the utmost importance in helping to determine its status within Hokan. A reinterpretation of a Tonkawa paradigm from HOIJER'S work by HYMES (1967) confirms the continuing interest in this language and illustrates the type of study which can be undertaken on the basis of such detailed descriptive materials, even after no more native speakers can be found.

5.4 C R O S S - F A M I L Y

COMPARISONS

Comparative work at a deeper level and in the direction of ProtoHokan(-Coahuiltecan) is also accumulating rapidly, as the availability of better and more extensive synchronic data increases. Using the technique of glottochronologic counts, BRIGHT (1956) discusses the results achieved in this manner by applying it to a subset of Hokan-Coahuiltecan languages, namely Seri, Chontal, Subtiaba-Tlappanec, Jicaque, and Comecrudo. Wherever possible he uses good modern data, and he shows that, in the cases where his calculations cover the same pairs of languages as were treated by GREENBERG - SWADESH (1953), his results differ somewhat from theirs and should be considered more accurate. In general, his results tend to support SAPIR'S Salinan-Seri subgroup, rather than a closer connection between Seri and Yuman; on the other hand, his figures reveal no particularly close relationship between Comecrudo and Tonkawa, adding to the general dissatisfaction with SAPIR'S Coahuiltecan subgroup. A number of investigators chose to approach the comparative problems by selecting two subgroups for intensive comparison. This often amounts to comparing two language isolates, since many a Hokan family has only one known member. These essentially

70

TOWARD PROOF AND ESTABLISHMENT

binary comparisons of course do not propose to reconstruct Proto-Hokan, nor do they necessarily imply a closer relationship between the languages considered than between any one of them and other Hokan languages. They do, however, add to the growing body of comparative material available, and, since in each case the investigator uses one language with which he is intimately familiar, the comparisons contain insights well beyond those which only the use of published material would permit. Anticipating his monograph on Proto-Palaihnihan phonology, OLMSTED, in a series of three articles (OLMSTED 1 9 5 6 , 1 9 5 7 , 1 9 5 9 ) , discusses the stop correspondences exhibited in sets of words in the Palaihnihan languages and Shasta, in an attempt to test on modern data the hypothesis that Shasta-Achumawi does indeed constitute a subgroup within Hokan. The comparison also draws on other Hokan languages to establish what if any similarities are uniquely shared by Shasta and Palaihnihan. He concludes that the data do not warrant the subgrouping, since while Atsugewi and Achumawi clearly exhibit shared innovations fully justifying their grouping into the Palaihnihan family, no special relationship emerges linking Palaihnihan to Shasta more than to any other Northern Hokan language. Shasta thus remains for OLMSTED a language isolate of the same order as Karok. Combining his intimate knowledge of Washo with the information in BRIGHT'S description of KarokJ(BRIGHT 1 9 5 7 ) , JACOBSEN ( 1 9 5 8 ) compares a variety of elements of both languages, and develops an interesting methodology of grading sound correspondences by their relative probabilities. It involves considerations of the probability of the sets in which they are exhibited being cognate, as well as of the number of instances of the correspondences. The results are encouraging particularly because of the high proportion of compared morphemes which are of basic importance in the morphology of the two languages (JACOBSEN 1958: 195).

Somewhat similar in approach are comparisons of Shasta and Karok (SILVER 1 9 6 4 ) , Eastern Pomo and Yana (MCLENDON 1 9 6 4 ) ,

TOWARD PROOF AND ESTABLISHMENT

71

and Yana and Karok (HAAS 1964). They contribute to the store of suggested cognates illustrated by the most recent data, and reveal the sound correspondences which emerge. All this sets the stage for the major task of full-fledged pan-Hokan-Coahuiltecan comparisons, and leaves little doubt that Karok, Shasta, Palaihnihan, Pomo, Washo, and Yana are indeed ultimately related. In addition, the HAAS article concludes with an excursus into California Hokan and exhibits tables listing the words for 'ear' and 'navel' in California Hokan languages. Reconstructions are proposed together with a summary of peculiarly Hokan structural characteristics, including vowel syncope (noted earlier by SAPIR), the fact that retained vowels and consonants tend to be conservative, and that metathesis is fairly common. On the other hand, the problem of the intersection of Hokan with other California families is beautifully demonstrated and one can only concur fully in the conclusion that The linguistic prehistory of California remains one of the most challenging to be found anywhere in the world. (HAAS 1964: 84)

In another important article, "Shasta and Proto-Hokan" (HAAS 1963), she proposes reconstructions for nine Proto-Hokan words, here as in her 1964 article exhibited in charts which summarize all her assumptions about the historical developments of the words in question. The words are listed in such a way that corresponding segments all appear in one column so there can be no doubt as to what is being compared, and additional devices identify cases where she assumes assimilation, and in what direction. Intermediate reconstructions for each family are also given when appropriate. The most original contribution of the paper is the recognition that certain long vowels of Shasta have resulted from the contraction of a Proto-Hokan segment which may be symbolized as *VmV (HAAS 1963: 42).

She urges Hokanists to look for other trends of this type in the history of other Hokan languages, and to engage in internal

72

TOWARD PROOF AND ESTABLISHMENT

reconstruction. She proposes that some of these phenomena are of fairly recent date and, in fact, for the Shasta case, she proposes Proto-Shasta reconstructions of the shape *VmV. She then brings up once more the question of subgrouping involving specifically the relationship of Shasta to Palaihnihan. She points out that some of her Proto-Shasta reconstructions are closer to her Proto-Palaihnihan ones than to any others, and considers this encouraging evidence to reopen the question of Shasta-Palaihnihan closeness, though she finds no evidence to substantiate any of the other Hokan subgroupings proposed previously. OLMSTED (1965b) questions both her conclusions and her methodology by pointing out that to reconstruct a Proto-Shasta sequence *VmV is in some instances not justified internally in Shasta. He defends his own Proto-Palaihnihan reconstructions as preferable to HAAS' and points out some cases of closer resemblances between Shasta and other Hokan languages. A new look at the material given suggests (to me) the following observations. Whether one accepts HAAS' or OLMSTED'S reconstructions depends indeed on theoretical considerations and specificity of proposed reconstructions. It does not, however, appear to change in any marked degree the number of forms which seem more similar between Shasta and Palaihnihan, although the specific lists are notidentical. A very rough look at both approaches reveals to my subjective inspection perhaps three items supporting the Shasta-Achumawi hypothesis, two where Yuman and Shasta are more similar, and one where Shasta agrees better with Yana. Clearly the whole question needs to be kept open (as all investigators agree), and exchanges like the one reported above will help keep the topic alive and subject everyone's assumptions to critical evaluation. A very sophisticated look at the syntactic feature of "switchreference" (i.e., the overt marking of a verb in one clause as having the same or different subject as that in another clause) leads JACOBSEN ( 1 9 6 7 ) to recognize its existence in not only several Hokan languages (Washo, Kashaya Pomo, Tonkawa) but also in a wide variety of other languages. He does not consider the evidence

TOWARD PROOF AND ESTABLISHMENT

73

sufficient to represent proof of genetic relationship among the Hokan languages, though it is of course not excluded. A rather negative comparison, this time between Tonkawa and Coahuilteco (TROIKE 1967a), concludes that evidence for genetic relationship is not revealed by the method of structural comparison he undertakes. This essentially substantiates the results of the lexical counts of BRIGHT (1956) and the considerations of HAAS (1967) on the relations of Tonkawa to the effect that there is certainly no special relationship between Tonkawa and Coahuiltecan. Much more negative are the results of TURNER (1967a), who, after comparing aspects of the structure of Seri and Chontal concludes that these two languages are not genetically related. His comparative cognate count (using essentially the basic 100-word list) comes to eight percent shared vocabulary as opposed to the 21 percent of BRIGHT (1956). One cannot deny that TURNER'S data are the best that are available at present. On the other hand, his criteria for cognation are obviously very stringent. His data unambiguously confirm what we have known for a long time: these two languages are quite different from each other. However, as pointed out by BRIGHT (1970a) in response, it is a priori impossible to ever prove that two languages are actually unrelated, and, furthermore, the data may not be quite as non-comparable as TURNER proposes. A note on the methodology of these two comparisons is appropriate. Comparison on typological lines of specific aspects of the synchronic structure of two languages not very closely related is not the way to reveal deep genetic affinity, as SAPIR so well pointed out in 1925. 5.5 P R O T O - C U L T U R E A N D H O M E L A N D

From time to time, attempts have been made to infer from linguistic evidence some aspects of the cultural history and population movements of Hokan-speaking peoples. We have already noted SAPIR'S hypothesis that a Hokan-Coahuiltecan geographical continuum was broken up by incursions from Uto-Aztecan and

74

TOWARD PROOF AND ESTABLISHMENT

Athapaskan groups (cf. p. 40) and Webb's calculations of ProtoPomo times (cf. p. 64). The lexicostatistic counts of GREENBERG SWADESH (1953) arrive at a figure of 5,000 years before the present for the period when Hokan-Coahuiltecan was a single language which, within a few centuries, divided into a chain of dialects from which the present-day languages eventually emerged. In a critical review of lexicostatistical results proposed for North America, KROEBER (1955) considers the results of GREENBERG - SWADESH (1953) and SWADESH (1954b), where percentages are translated into centuries of divergence, and retabulates the results for HokanCoahuiltecan, with times of separation ranging from 34 centuries (between Jicaque and Chontal) to 55 centuries (between Washo and Comecrudo). Combining the time-depth results with a consideration of the geographical distribution of the various families, KROEBER proposes an entity called Pacific Hokan, which includes all of Hokan-Coahuiltecan (even Jicaque) but neither Washo nor Coahuiltecan. California must have been the point of origin, and, at one time, some "explosive disruption" (KROEBER 1955: 102) caused the movements which resulted in the present spread over 3,000 miles, at a time when the individual languages were already fully differentiated. Considering the southernmost outliers into Pacific Mexico-Honduras-Nicaragua, he argues that their distribution could have been the result of a single migration with individual groups dropping off on the way. He warns of the dangers of trying to interpret archaeological findings as corroboration of linguistic results on the very sensible grounds that artifacts contain no clue whatsoever to the linguistic affiliation of their makers. While fully aware of these dangers, TAYLOR ( 1 9 6 1 ) nevertheless takes on the job of interpreting simultaneously linguistic, geographical, cultural, and archaeological data to present a "working hypothesis" of Western North American prehistory, in which Hokan-Coahuiltecan plays a crucial part, as all agree to its great antiquity. His model of the historical developments can be summarized as follows: earlier than 1 0 , 0 0 0 years ago, Hokaltecan speakers moved into the Great Basin, where they adopted a Desert

TOWARD PROOF AND ESTABLISHMENT

75

culture. After some time, they expanded westward into California, and southward through the Southwest and into Northern Mexico. By about 1 0 , 0 0 0 years ago, they occupied a continuous territory from the Northern Great Basin to Northeastern Mexico; later, but before 5 , 0 0 0 years ago, some groups had started to move westward into California and eventually spread southward to account for the Pacific Hokan groups. After Hokan establishment in California, Penutians moved in, probably from the North, and may represent the cause of the explosive disruption postulated by KROEBER ( 1 9 5 5 ) . TAYLOR finds the Uto-Aztecan situation harder to explain, but of less antiquity than the Penutian, with northern highland origins, spreading into Mexico some 5 , 0 0 0 to 6 , 0 0 0 years ago. MILLER ( 1 9 6 6 : 8 7 - 9 1 ) and JACOBSEN ( 1 9 6 6 : 2 6 0 ) cogently criticize TAYLOR'S proposal. Less ambitious, but perhaps of more interest because dealing with shallower time depths and better data are a few papers dealing with the prehistory of a specific family. BAUMHOFF - OLMSTED ( 1 9 6 3 , 1 9 6 4 ) find correlations between their glottochronological counts (for which they use very conservative criteria), i.e., 31 to 35 centuries of divergence between Achumawi and Atsugewi (or almost as much as KROEBER'S estimate for Jicaque and Chontal!) and archaeological evidence. Both lexicostatistics and archaeology are obviously subject to widely different interpretations in various hands, and their use must be assessed with due caution. On a purely linguistic basis, L A W ( 1 9 6 1 ) surmises some of the cultural traits of Proto-Yuman speakers on the basis of comparative vocabularies. The quality of the paper is marred by some serious errors in transcription which can be attributed to the author's unfamiliarity with the languages in question, but which fortunately do not have much effect on the overall conclusions. He proposes that the speakers of Proto-Yuman lived in an area similar to their present location, and were concerned with gathering wild plants, hunting animals, and the development of agriculture. The main area was probably the lower Colorado River. Bow and arrow were the weapons. Tools included the grindstone, mortar, and pestle;

76

TOWARD PROOF AND ESTABLISHMENT

gourd rattles, dreams, dance, flutes, and shamans were important aspects of the religious life. The time is suggested as about 1,500 years before the present. There is nothing here that is startling, but what there is corroborates what is known from ethnographic studies and validates the technique used, which is basically not novel, but provides interesting methodological devices for classifying cognates into different sets graded for their relative validity for the purposes of reconstructing the items they denote. Commenting on her classification of the Yuman languages, JOEL ( 1 9 6 4 ) suggests that resemblances between Arizona Upland Yuman (Havasupai, Walapai, Yavapai) and Paipai are due to peripheral retention rather than to more recent contact. She sees the linguistic diversity among the Colorado River Yumans as a concomitant of cultural specialization, and among the Diegueno as a result of outside (non-Yuman) influences. WINTER ( 1 9 6 7 ) , on the other hand, chooses the alternate hypothesis and suggests a fairly recent migration of the Paipai. SHATERIAN ( 1 9 7 1 ) , with whom I agree, states that the linguistic facts available at the moment are neutral with respect to the two hypotheses. The marked resemblances between Arizona Upland Yuman and Paipai thus remain unexplained. For Pomo, OSWALT (1964b) postulates that the early Pomo lived on the shores of Clear Lake in four branches (Western, Eastern, Southeastern, and Northeastern), which then expanded to their present locations. The maximum time separating Pomo languages he considers greater than that between the Germanic or the Romance languages, but certainly less than that between the major branches of Indo-European. HALPERN (1964) places his ProtoRussian River group (cf. p. 64) between the Russian River valley and Clear Lake, an area from which they then expanded in a fanlike migration to the north, west, and south. These tentative suggestions point to many problems which linguistic considerations may help resolve. As our linguistic insights increase we may look forward to further light on these problems as well.

6 THE FUTURE

6.1 T H E F I R S T C O N F E R E N C E ON H O K A N L A N G U A G E S

In the spring of 1 9 6 7 , Shirley SILVER spent a month at the University of California, San Diego, where Margaret LANGDON is a faculty member in the Linguistics Department. The opportunity to discuss Hokan problems at length uncovered a common frustration at the lack of availability of Hokan work in progress and the rarity of opportunities for Hokanists to meet. In this setting originated the idea of organizing a conference devoted uniquely to Hokan languages, with particular emphasis on historical considerations, to provide a forum for discussion of important problems. The response to the suggestion was enthusiastic, funds were generously granted by the National Science Foundation, and, in April, 1970, Hokanists gathered for the first time on the campus of the University of California, San Diego, to participate in the First Conference on Hokan Languages. They came from as far as Germany, Southern Mexico, and New York City to discuss papers they had previously circulated, to share data, to take stock of progress made, and to identitfy common problems. The Proceedings of the Conference (which are not available in print at this writing) will be incorporated into a volume entitled Hokan studies (LANGDON SILVER to appear), which contains revised versions of many of the papers read at the conference. It consists of 24 articles devoted to synchronic and diachronic problems, and will constitute the largest addition to the Hokan literature in a single year. An annotated table of contents of this volume will serve as a summary of the contributions. These items are listed in the bibliography under the

78

THE FUTURE

authors' names and as unpublished manuscripts, but with the date 1970, when they were presented at the Conference.5 6.1.1 Pomo

Robert L. OSWALT, "Comparative verb morphology of Pomo" (OSWALT 1970ms), after a brief outline of the Proto-Pomo sound system, reconstructs 20 "instrumental prefixes" of Proto-Pomo, the only class of productive verbal prefixes in the Pomo languages. These are morphemes of the shape CV (where the vowels can only be *a *i *u). He also reconstructs 54 verbal suffixes (inflectional, locative, modal, syntactic). A complete picture of Proto-Pomo verb morphology emerges. He then brings the morphological evidence to bear on the subgrouping problem, with results which differ slightly from his previous subgrouping (OSWALT 1964b) based on lexical evidence, and are close to one of HALPERN'S suggestions (HALPERN 1964). The difference affects only the status of Northeastern Pomo, which is lexically the most divergent Pomo language. This discrepancy is explained by the geographical isolation of that language. Sally MCLENDON, "The Proto-Pomo pronominal system" ( M C L E N D O N 1970a ms), reconstructs both surface and underlying forms of pronouns in Proto-Pomo. The results are summarized in charts, rules are given for relating the underlying Proto-Pomo forms to the surface forms and for relating the latter to the daughter forms, and full derivations of individual items are given. The 6

In addition to the papers summarized in this section, there were papers read at the Conference which could not, for various reasons, be included in the published proceedings. They are: George Grekoff, "The structure and function of the nominalized clause in Chimariko" (GREKOFF 1970ms); Eric P . Hamp, "Methods and goals in comparative Hokan studies" (HAMP 1970ms); Edwin Kozlowsky, "Havasupai syntax" (KOZLOWSKY 1970ms); David Olmsted, "The contribution of Jeremiah Curtin [to Atsugewi]" (OLMSTED 1970ms); Leonard Talmy, "The Atsugewi instrumental prefix system" (TALMY 1970ms). On the other hand, three of the papers included in the volume (APPLEGATE 1970ms, J. G. CRAWFORD 1970ms, MOSER-MOSER 1970ms) were not read at the Conference.

THE FUTURE

79

system consists of a first person stem *ha ?äw and a second person stem * ?a-md, which are inflected for possessive and object case in singular and plural. There is also a third person stem *hamiya· with masculine and feminine variants, both inflected for possessive and object case. Third person plurals exist in the daughter languages but cannot be reconstructed with assurance. Julius MOSHINSKY, "Historical Pomo phonology" (MOSHINSKY 1970a ms), reconstructs first the prosodic system of Proto-Pomo (the first stem syllable of the word receives the stress). He then proposes the phonemic inventory of Proto-Pomo and provides a set of ordered rules which allow the derivation of the daughter language forms from the proto-forms. Finally, he illustrates his discussion with a comparative Pomo lexicon complete with reconstructions. He finds that segmental changes can be characterized as very general processes, such as "stop fronting, spirantization, glottal increment deletion/insertion, vowel and consonant deletion, and aspirate dissimilation." Eero VIHMAN, "On pitch accent in Northern Pomo" (VIHMAN 1970ms), considers the recurrent pitch contours which are associated with various syllable types and subjects them to instrumental analysis. He concludes pitch is a predictable feature of lexical stress in Northern Pomo. 6.1.2 Yuman "Yavapai [+sonorant] segments" (SHATERIAN 1970ms), discusses in detail some phonetic processes in Yavapai and their interactions in the phonology of the language. These are of a type encountered in other Yuman languages as well. They involve in particular the behavior of the unstressed vowel a and the segment h and its effect on surrounding segments. The language distinguishes three degrees of vowel length, a phenomenon also attested in another Yuman language, Paipai. Martha B . KENDALL and William L . COLEMAN, "Directions in the study of Upland Yuman" (KENDALL - COLEMAN 1970ms), discuss A.V.

SHATERIAN,

80

THE FUTURE

problems of the same nature as those in SHATERIAN (above) for Walapai, Yavapai, Havasupai, including stress placement in simple stems and compounds, a insertion and/or deletion, the status of glottal stop; they present alternate ways of handling these phenomena. James E. REDDEN, "Walapai syntax: a preliminary statement" (REDDEN 1970ms), describes "the major phrase and sentence structure patterns of Walapai", including complex sentences. The feature of switch-reference, which has been uncovered in several other Yuman languages, is discussed, but REDDEN believes the evidence of Walapai to be such that it contradicts that of other Yuman languages in this respect. Leanne HINTON and Margaret LANGDON, "Object-subject pronominal prefixes in La Huerta Diegueno" (LANGDON - HINTON 1970ms), describe a set of personal verb prefixes for a dialect of Diegueno which differs in some important ways from other dialects of the same language as well as from other Yuman languages. The facts become clearer when one recognizes the application of several phonological rules peculiar to this dialect including a glottal stop metathesis rule. The Proto-Yuman object-subject prefix system is then reconstructed on the basis of data from Walapai, Havasupai, Paipai, Cocopa, Yuma, and Mesa Grande and La Huerta Diegueno, and is seen to consist of three morphological elements (plus a significant zero) which combine with each other in various ways. An attempt is made to infer an earlier Pre-ProtoYuman system to account for the development of the ProtoYuman system. Margaret LANGDON, "The Proto-Yuman vowel system" (LANGDON 1970c ms), proceeds first to perform internal reconstruction of the stressed vowel system of the verb stems of Yuma proper, using the data of HALPERN (1942, 1946-47), where a complex system of vocalic ablaut relates non-plural to plural stems. With appropriate phonological rules, this system turns out to consist of only three contrastive vowels (both long and short), and evidence is presented to indicate that this same vocalic system can be reconstructed for Proto-Yuman.

THE FUTURE

81

Μ. Mixco, "Historical implications of some Kiliwa phonological rules" (Mixco 1970ms), brings to bear on the Proto-Yuman problem some facts of Kiliwa which other investigators had not been able to use in view of the lack of descriptive material on the language. Mixco's field work among the Kiliwa provides the data. A process of sound symbolic consonantal alternations is described, as well as a process of symbolic devoicing of sonorants, the latter also operative in other areas of the language. In this context, he discusses related phenomena in other Yuman languages, particularly Diegueno and Cocopa, and proposes that they have a common origin. He also suggests that palatalized consonants developed under the effect of a following high vowel now lost in most languages, a natural enough development, which he, however, attributes to an earlier *u, rather than */, on the basis of evidence from Kiliwa. Rudolph C. TROIKE, "The linguistic classification of Cochimi" (TROIKE 1970ms), presents some new data uncovered by the author on this extinct language, which allows a calculation of cognates with other Yuman languages. Two varieties of Cochimi are included and are found to have less similarity with each other than one of the two has with Proto-Yuman. Some grammatical observations are offered which strengthen the affinity with Yuman, but TROIKE concludes that Cochimi must (though it is closer to Yuman than to other Hokan families) be considered to constitute a family related to Yuman at a somewhat deeper level than the other Yuman languages. Werner WINTER, "Switch-reference in Yuman languages" (WINTER 1970ms), reviews all the syntactic evidence available for Yuman languages with respect to the feature of "switch-reference", and demonstrates that it can be reconstructed for Proto-Yuman and that a morpheme *-m carried the function of "different subject" marker. He does, however, share JACOBSEN'S doubts about the Hokan nature of this feature (JACOBSEN 1967), and points out that it is attested in Indo-European (as the ablative absolute), where the evidence points to independent development in various languages.

82

THE FUTURE

6.1.3 Other Hokan languages James M. CRAWFORD, "A Comparison of Chimariko and Yuman" (CRAWFORD 1970b ms), hypothesizes a new subgrouping within Hokan, namely a special affinity between Chimariko (for which he uses HARRINGTON'S unpublished data) and Yuman. The paper provides a comparative vocabulary of 134 items and tabulates sound correspondences between Chimariko and Proto-Yuman (the best attested sets are in the area of labial and velar consonants). He concludes that the evidence is strong for genetic relationship between Chimariko and Yuman and finds that, within Yuman, the Arizona languages (and Yavapai in particular) show more archaic features in their vocabulary than other Yuman languages. Shirley SILVER, "Comparative Hokan and the Northern Hokan languages" (SILVER 1970ms), attributes the intractability of Hokan material to a type of change which she calls "morphemization", i.e., a process whereby originally distinct morphemes in set strings become petrified and can no longer be analyzed. Traces of this process may, however, be found in irregular deviant forms of a language, and, on the other hand, the results of this process may lead the investigator to false comparisons. Only by intensive use of internal reconstruction at all possible levels can the extent of this process be uncovered. Several sets of forms illustrate the problems involved, and the suggestion is made that some of the globalized and aspirated consonants may have originated in this process of morphemization. William H. JACOBSEN, Jr., "Observations on the Yana stop series in relationship to problems of comparative Hokan phonology" (JACOBSEN 1970ms), cogently reviews the basic problems of Hokan comparative studies and, paralleling a study he made of the Washo stop series (in which he concluded that internal reconstruction gives evidence for the earlier non-contrastiveness of voiced and voiceless stops, though the glottalized series remains separate), engages in a detailed study of the distribution and behavior of Yana stops and concludes that the aspirated series also

THE FUTURE

83

originates from secondary developments in the history of Yana. Comparing these results with some of the other findings presented at the Hokan Conference, he concludes that there is evidence to suggest that Proto-Hokan had no more than two stop series, namely plain and glottalized. Based on elegant scholarship and full documentation, this is a highly convincing demonstration. Bruce E . NEVIN, "Transformational analysis of some 'grammatical morphemes' in Yana" (NEVIN 1970ms), discusses some aspects of Yana syntax in which various sentence types are related transformationally. This is of potential comparative interest, since it consists essentially of a form of syntactic internal reconstruction. M.S. BEELER, "Barbareno Chumash grammar: a farrago" (BEELER 1970b ms), is nothing less than a sketch of the structure of a Chumash language, based on extensive modern data collected by the author from the last speaker of the language. Some highlights : an elaborate consonant system with pervasive glottalization (affecting even continuants and sibilants), stops with an added constrast of aspiration, vowels including a very un-Hokan high central i'. Some instances of aspirated consonants are the result of underlying geminate clusters, many morphemes have alternants with elided second-syllable vowels, words are subject to sibilant harmony. There are prefixed, suffixed, and free forms of pronouns (dual forms exist). Nouns have not only possessive prefixes, but, alternately, an article, which in turn may be preceeded by locational or demonstrative prefixes. The distributive plural of nouns is formed by initial reduplication accompanied by glottalization and falling tone on the final syllable. This paper is a major contribution to future comparative studies. Richard B . APPLEGATE, "Reduplication in Chumash" (APPLEGATE 1970ms), singles out one of the features of Chumash languages and its detailed operation in Ineseno. There are many distinct types of reduplication, and their description presents some very interesting problems in that the process appears to be of low level phonologically, though complex morphological restrictions control its operation.

84

THE FUTURE

Edward and Mary MOSER, "Seri noun pluralization classes" (MOSER - MOSER 1970ms), present extensive data on this extremely complex area of Seri structure. The earlier study on verb pluralization in the same language (MOSER 1961) revealed a similar though even more elaborate system. Some interesting processes of plural formation are: morphophonemic alternations between χ and /, some vowel harmony, single forms marked for plural simultaneously by an infix and a suffix or by a stem change and a suffix, forms with both individual and collective plurals. Paul R . TURNER, "Pluralization of nouns in Seri and Chontal" (TURNER 1970ms), pursues his earlier hypothesis (TURNER 1967a) that Seri and Chontal are not genetically related by juxtaposing data on noun pluralization in the two languages, and succeeds once more in showing that the two languages are considerably different in their processes of plural formation. Judith G. CRAWFORD, "Seri and Yuman" (J. G. CRAWFORD 1970ms), presents evidence linking Seri and Yuman "in hopes of dispelling doubts as to the Hokan affinity of Seri". She lists 227 sets gleaned from the Seri dictionary ( M O S E R - M O S E R 1961) and various modern sources for Yuman languages (including her own field notes on Mohave, Maricopa, Havasupai, and Yavapai). Sound correspondences are listed with a number of instances and the results allow the conclusion that Yuman and Seri are definitely related, though no special claim of subgrouping can be made. Viola Grace WATERHOUSE, "Another look at Chontal and Hokan" (WATERHOUSE 1970ms), brings to bear on the problem a vast amount of comparative data, first inter-Chontal comparisons amounting to a summary of the historical phonology of Chontal. Challenging TURNER'S methodology in his contention that Seri and Chontal are not related, she selects areas of the vocabulary he used in his negative comparison, namely the kinship terms and numerals, and demonstrates that these are quite different even within Chontal (where no more than dialectal variation is recognized). She then proceeds to comparisons beyond Chontal and adds appropriate Chontal items to sets given by earlier investigators,

THE FUTURE

85

first to Yuman cognates, then to Seri forms, then to the other Hokan languages. Without committing herself to LANDAR'S hypothesis (LANDAR 1968), she also, however, adds Chontal material which might support a link with Karankawa. Finally, she compares some Tlappanec and Chontal forms. 6.1.4 Wider perspectives Mary R. HAAS, "The Northern California linguistic area" (HAAS 1970b ms), reminds the assembled Hokanists that their problems are not tied uniquely to Hokan languages, but that areal diffusion features enter into the complicated picture they are trying to unravel. She exemplifies her point with a discussion of some features of Northern California languages, i.e., phonetics with a tendency to three series of stops and affricates and to the presence of back velar consonants and voiceless spirantal and affricated /-sounds (all Northwest Coast areal features). Retroflex apical stops are a southern trait with northernmost representation in Chimariko. Notable is the strangeness of Karok with an extremely simple consonant system in the proximity of languages with fairly complex systems. Does this point to an archaic Hokan feature? Consonant symbolism (cf. HAAS 1970a) also gives evidence of being a diffused trait. Numeral systems and pronominal forms are considered and found to show similarities which cross genetic boundaries. "This makes the task of determining the validity of the various alleged Hokan languages [ . . . ] all the more difficult." We all agree. William BRIGHT, "The First Hokan Conference: conclusions" (BRIGHT 1970b ms), while noting the great advances made both descriptively and comparatively, heeds HAAS' warning and suggests that the historical framework in which our efforts belong cannot fit the model of a family tree alone. "And faced with languages as diverse as those we call Hokan, we may need to discard the idea of a Hokan family as a neatly bounded entity [ . . . ] . The task of Hokanists would then be not to put languages into genetic pigeonholes, but rather to trace the prehistory of linguistic sharing."

86

THE FUTURE

6.2 T A K I N G

STOCK

We may conclude that Hokan-Coahuiltecan as conceived by SAPIR has withstood the test of time fairly well. Seriously disputed as to membership are only Tonkawa and Karankawa, and a new member, Jicaque, is likely. Less clear is the validity of SAPIR'S subgroupings, which, however, have no serious competitors to replace them, except the very cautious view that no two of the families grouped under the label Hokan-Coahuiltecan can be shown as yet to be closer to each other than to the others. This is encouraging since it makes possible new hypotheses which too close an acceptance of any single scheme would dampen. Suggestions in that direction abound, and I would like to add my own: a possible Chumash-Seri-Chontal (Southern Hokan?) subgroup, suggested by some of the newly described features of Chumash, sharing such features as a nominal affix whose function is somewhat like an article (clear in Chumash, called "limiter" in Chontal by TURNER, petrified in Seri but possibly recoverable in some singular/plural alternations and tentatively associated with the shapes I in Chumash, I and I in Chontal, and χ and I in Seri), a process of pluralization involving final glottalization (regular in Chumash, sporadic in Chontal, not attested in Seri, which has no globalized consonants and where ? itself is apparently of restricted distribution) and processes of pluralization linking Seri and Chontal, such as replacement, suffixation, infixation, shift of stress, and loss of consonant (TURNER 1970). Most encouraging is the growing number of insightful descriptions of individual languages making possible the reconstruction of such proto-languages as Proto-Pomo, Proto-Palaihnihan, ProtoYuman (the work for all three is well on its way) and that of earlier stages of other groups by internal reconstruction (pre-Washo, preYana, and pre-Shasta are emerging). While a full demonstration of the validity of the Hokan-Coahuiltecan hypothesis is not yet a reality, there is a growing sense of excitement as convergent results are reported.

THE FUTURE

87

In addition to the continuing field work on languages still spoken and the consolidation of comparative results, there is a renewal of interest in the detailed analysis of notes left by earlier recorders of languages now extinct - most importantly the voluminous data of J.P. HARRINGTON - which promises to rescue from oblivion such languages as Chimariko, several Chumash languages, possibly Esselen. More emphasis on syntax through the framework of transformational grammar is bound to broaden the basis for comparative studies, and, in turn, the comparative efforts of Hokanists cannot fail to contribute to the methodology of historical linguistics insofar as problems begin to be solved which demand more subtle tools than the comparative method. A summary of what the convergent feelings of Hokanists foreshadow is as follows. Proto-Hokan probably had a rather simple sound system, with very little trace of the more marked categories exhibited by many of the attested languages. Contrasts involving plain versus aspirated and perhaps even glottalized consonants may well turn out to be accountable as independent developments; voiceless sonorants are already accounted for as innovations in Pomo, Yuman, and Washo. Vowels may not have been more than three with a probable length contrast-Proto-Yuman has such a system. In the few available good cognate sets, the persisting elements appear to be essentially conservative. The great diversity of the daughter languages, it seems, must be accounted for by repeated processes of loss of vowels leading to subsequent loss and change of consonants (particularly in the laryngeal area), with resulting lexical items where little remains that is truly comparable. Typical Hokan morphemes must have been short (monosyllabic). "Morphemization" is a direct concomitant of the type of historical processes hypothesized, and the reconstitution of underlying morpheme strings becomes a prerequisite for adequate reconstruction. The serious confrontation of the many problems to be handled is also leading to a renewed interest in and reappraisal of SAPIR'S fundamental contributions. After a long time, the avenues of research he so forcefully outlined are beginning to be explored. It is hoped that

88

THE FUTURE

the present survey will be a step in the direction of synthesis and clarification of issues. It is offered in tribute to the insight of Edward SAPIR, whose ideas have pervaded and will continue to pervade Hokan-Coahuiltecan studies.

7

BIBLIOGRAPHY

The following bibliography does not claim to be an exhaustive listing of all works which have a bearing on the Hokan-Coahuiltecan group or on the individual languages involved. It lists rather all works which I have consulted, or, in a very few cases, attempted to consult during the preparation of this monograph, including some more general items which treat matters of methodology or are of particular interest for other reasons. I have also included some publications which are mostly ethnographic in nature, but which were consulted to ascertain whether they contained statements of linguistic interest. The ethnographic coverage is, of course, far from complete and may be supplemented by consulting MURDOCK'S ( 1 9 6 0 - 6 1 ) Bibliography .On the linguistic side, I have tried to be exhaustive in listing all items dealing with comparative or classificatory matters relating to Hokan-Coahuiltecan, as well as modern descriptive works. Early descriptive work (before the twentieth century) is not covered systematically, but references to it may be found in the more recent items. In order to make the bibliography easier to use, I have prepared a topical Index listing for each heading the relevant contributions identified by name of author (alphabetically) and year of publication (chronologically). Since some papers are relevant to more than one topic, the same item may appear in the Index more than once. The main bibliography is arranged alphabetically by author; when more than one publication by the same author is cited, the arrangement under each author is chronological. Some items not directly referred to in the body of this paper but containing material of special interest are preceded by an asterisk (*). Items which have no direct bearing on the main topic are marked by

90

BIBLIOGRAPHY

7.1 A L P H A B E T I C A L L I S T I N G B Y A U T H O R Almstedt, Ruth F. 1968 "Diegueno Tree: an ecological approach to a linguistic problem", IJ AL 34: 9-15. *1970ms "Bibliography of the Diegueno Indians", in: Lifeways of the Diegueno (Ed.: Emma Lou Davis et alii) (Department of Anthropology, San Diego State College), [mimeographed; to appear] Applegate, Richard B. 1970ms "Modes of reduplication in Ineseno Chumash", in LANGDON-SILVER (1970). [to appear] 1971 "Vowel harmony in Chumash", Berkeley papers in linguistics 1: 3-12. Baegert, Johann Jakob 1773 Nachrichten von der amerikanischen Halbinsel Californien (Mannheim: Löffler). [translated as: M.M. Brandenburg - Carl L. Baumann, Observations in Lower California (Berkeley-Los Angeles: University of California Press 1952). [chapter 10 has notes on Waicuri ( = Waikuru)] Barrett, S.A. 1908 "The ethno-geography of the Pomo and neighboring Indians", University of California publications in American archaeology and ethnology 6:1-332. [vocabularies of Pomo languages, pp. 56-8] Bartholomew, Doris A. * 1 9 6 9 Review of HYMES-BITTLE (1967), Lingua 2 3 : 8 5 - 6 . Baumhoff, Μ.A. - David L. Olmsted 1963 "Palaihnihan: radiocarbon support for glottochronology", AmA 65: 278-84.

1964 "Notes on Palaihnihan culture history: glottochronology and archaeo l o g y " , i n BRIGHT ( 1 9 6 4 : 1 - 1 2 ) .

Beeler, Madison S. •1955 Review of HEIZER (1952, 1955), Lg 31: 165-9. 1 9 6 4 "Venturefio numerals", in BRIGHT ( 1 9 6 4 : 1 3 - 8 ) . 1967 The Venturefio Confesionario of Jose Senän, O.F.M. (=UCPL 47) (Berkeley - Los Angeles: University of California Press). 1970a "Sibilant harmony in Chumash", UAL 36: 14-7. 1970b ms "Barbareno Chumash grammar: a farrago", in LANGDON-SILVER (1970). [to appear] Belmar, Francisco 1900 El estudio de el Chontal (Oaxaca). [notes on grammar, phrases, vocabulary, and stories] 1905 Lenguas indigenas de Mexico; familia Mixteco-Zapoteca y sus relactones con el Otomi; familia Zoque-Mixo; Chontal; Huave y Mexicano (Mexico: Imprenta particular), [contains descriptive material on Chontal]

BIBLIOGRAPHY

91

Biggs, Bruce 1957 "Testing intelligibility among Yuman languages", IJ AL 23: 57-62. Boas, Franz-John Wesley Powell 1966 Franz Boas, Introduction to [his] Handbook of American Indian languages - John Wesley Powell, "Indian linguistic families of America north of Mexico" (Ed.: Preston Holder) (Lincoln, Neb.: University of Nebraska Press), [cf. the first editions: Franz Boas, Handbook of American Indian languages (= Smithsonian Institution, Bureau of American ethnology, bulletin 40.1) (Washington: Government printing office); POWELL (1891)] Bowen, Thomas - Edward Moser 1968 "Seri pottery", The Kiva 33: 89-132. Bright, Jane O. - William Bright *1965 "Semantic structures in Northwestern California and the SapirWhorf hypothesis", in: Formal semantic analysis (Ed.: E.A. Hammel) (=AmA 67.5.2) (Menasha, Wise.: American Anthropological Association), pp. 249-258. [uses Karok data among others] Bright, William 1952 "Linguistic innovations in Karok", IJ AL 18: 53-62. [borrowings, new usages, new formations, and their effect on Karok structure] 1954 "Some Northern Hokan relationships; a preliminary report", in CHRETIEN et alii ( 1 9 5 4 : 6 3 - 7 ) . 1955 "Bibliography of the Hokan-Coahuiltecan languages", UAL 21: 276-85. 1956 "Glottochronologic counts of Hokaltecan material", Lg 32: 42-8. 1957 The Karok language {= UCPL 13) (Berkeley-Los Angeles: University of California Press). "I960 "Animals of acculturation in the California Indian languages", UCPL 4: 215—46. [includes data from all California Hokan language families] *1965a Review of OLMSTED (1964), Lg 41: 175-8. 1965b "A field guide to Southern California languages", Annual report, archaeological survey, Department of Anthropology, University of California, Los Angeles, 7: 389-408. [contains a sketch of Diegueno phonemics; pp. 403-6] * 1967a "Inventory of descriptive materials", in M C Q U O W N (1967: 9-62). [bibliography of descriptive material on Middle American languages] 1967b "Karok nrnkkay < Scottish McKay", Names 15: 235-6. 1970a "On linguistic relatedness", IJAL 36: 288-90. 1970bms "The First Hokan Conference: conclusions", in LANGDON - SILVER (1970). [to appear] Bright, William (ed.) 1964 Studies in Californian linguistics (= UCPL 34) (Berkeley-Los Angeles: University of California Press), [contains BAUMHOFF-OLMSTED, BEELER, GREKOFF, HAAS, HALPERN, JOEL, LAMB, M C L E N D O N , OSWALT, and SILVER, all 1 9 6 4 . Also has an excellent bibliography on Californian

92

BIBLIOGRAPHY

linguistics (pp. 217-38) on which the present bibliography is modeled] Bright, William - David L. Olmsted 1959 "A Shasta vocabulary", Papers of the Kroeber Anthropological Society 20: 1-55. Brinton, Daniel G. 1891 The American race: a linguistic classification and ethnographic description of the native tribes of North and South America (New York: N.D.C. Hodges), [connects Seri and Yuman to Chontal] 1892 "Chontales and Popolucas: a contribution to Mexican ethnography", in: Proceedings of the eighth International Congress of Americanists, 1890 (Paris: Leroux 1892), pp. 556-64. [contains short Chontal word list] Castetter, E.F. - W.H. Bell *]951 Yuman Indian agriculture (Albuquerque: University of New Mexico Press), [plant names in Mohave, Cocopa, Yuma, Maricopa] Chafe, Wallace L. *1962 "Estimates regarding the present speakers of North American Indian languages," IJ AL 28: 162-71. [for Hokan lists Achumawi, Atsugewi, Chumash, Cocopa, Diegueno, Havasupai, Karok, Maricopa, Mohave, Pomo (all except Eastern), Salinan, Shasta, Walapai, Washo, Yavapai, Yuma] Chretien, C.D. - M.S. Beeler - M.B. Emeneau - M.R. Haas (Eds.) 1954 Papers from the Symposium on American Indian Linguistics held at Berkeley July 7, 1951 ( = UCPL 10) (Berkeley-Los Angeles: University of California Press). Conzemius, E. 1921-23 "TheJicaques of Honduras", IJ AL 2: 163-70. [vocabulary and a few grammatical observations] Crawford, James 1966ms The Cocopa language (University of California, Berkeley). [Ph.D. dissertation] 1970a "Cocopa baby talk", IJAL 36: 9-13. 1970bms"A comparison of Chimariko and Yuman", in LANGDON-SILVER (1970). [to appear] Crawford, Judith G. 1970ms "Seri and Yuman", in LANGDON-SILVER (1970). [to appear] Croft, Kenneth *1948 "A guide to some material on extinct North American Indian languages", IJAL 14: 260-8. [contains items on Chumash (p. 265), Karankawa and Salinan (p. 266), Yana (p. 268)] Dangberg, Grace 1922 "The Washo language", Nevada Historical Society papers 3: 145-52. 1927 "Washo texts", University of California publications in American archaeology and ethnology 22: 391-443.

BIBLIOGRAPHY

93

Davis, I. *1961 "The native languages of America: a survey of recent studies", Phonetica 7: 40-63. [contains bibliographic references for the years 1951-60] d'Azevedo, Warren L. - John A. Price 1963 "An annotated bibliography of Washo sources", University of Utah anthropological publications 67: 153-201. de Angulo, Jaime 1925a "The linguistic tangle of Oaxaca", Lg 1: 96-102. 1925b "Kinship terms in some languages of Southern Mexico", AmA 27: 103-7. [includes Chontal] 1926 "Two parallel modes of conjugation in the Pit River language", AmA 28: 273-4. 1927 "Texte en langue Pomo de Californie", JSAm 19: 129-44. [Eastern Pomo] de Angulo, Jaime - L.S. Freeland 1925 "The Chontal language (dialect of Tequixistlan)", Anthropos 20: 1032-52. 1930 "The Achumawi language", IJ AL 6: 77-120. [grammatical sketch and texts] 1931 "Karok texts", IJ AL 6: 194-226. Dedrick, J. *1958 "Panorama de la lingiiistica de Baja California", in: Memoria del Primer Congreso de Historia Regional (Mexicali, B. C.) 1 (Mexico, D.F.: Manuel Casas), pp. 181-194. [general survey of Yuman languages of Baja California, not based on personal observation] Diebold, A. Richard, Jr. *1959 Review of B R I G H T (1957), Lg 35: 388-94. *1961 Review of SAPIR-SWADESH (1960), AmA 63: 878-9. Dixon, Roland B. 1905 "The Shasta-Achomawi: a new linguistic stock with four new dialects", AmA 7: 213-7. 1906a "The pronominal dual in the languages of California", in: Boas anniversary volume (New York: G.E. Stechert & Co), pp. 80-84. 1906b "Linguistic relationships within the Shasta-Achumawi Stock", in: Proceedings of the 15th International Congress of Americanists (Quebec: Dussault-Proulx), pp. 255-63. 1910 "The Chimariko Indians and language", University of California publications in American archaeology and ethnology 5: 295-380. 1931 "Dr. Merriam's 'Τΐό-höm-tah'-hoi' ", AmA 33: 264-7. Dixon, Roland B. - Alfred L. Kroeber 1903 "The native languages of California", AmA 5: 1-26. 1907 "Numeral systems of the languages of California", AmA 9: 663-90. 1913a "Relationship of the Indian languages of California", Science 37: 225. 1913b "New linguistic families in California", AmA 15: 647-55. 1919 "Linguistic families of California", University of California publications in American archaeology and ethnology 16: 47-118.

94

BIBLIOGRAPHY

Elmendorf, W.W. 1965

"Some problems in the regrouping of Powell units", The Canadian journal of linguistics 10: 93-107. Fernändez de Miranda, Maria Teresa *1967

" I n v e n t o r y of classificatory m a t e r i a l s " , i n MCQUOWN (1967: 63-78).

[bibliography of material pertaining to classification of Middle American languages - specifically pp. 64-5] Frachtenberg, L.J. 1918 "Comparative studies of Takelman, Kalapuyan, and Chinookan lexicography, a preliminary paper", IJAL 1: 175-82. Frisch, Jack A. 1968 "Maricopa foods: a native taxonomic system", IJAL 34: 16-20. Frisch, Jack A. - Noel W. Schutz, Jr. 1967 "Componential analysis and semantic reconstruction: the ProtoCentral-Yuman kinship system", Ethnology 6: 272-93. Garth, Thomas R., Jr. 1944 "Kinship terminology, marriage practices, and behavior toward kin among the Atsugewi", AmA 46: 348-61. Gatschet, Albert S. *1876 Analytical report upon Indian dialects spoken in Southern California, Nevada, and on the Lower Colorado river, & c., & c., based upon vocabularies collected by the expeditions for geographical surveys west of the 100th meridian, Lieut. Geo. M. Wheeler, Corps of Engineers, US. Army in charge (Washington: Government Printing Office). [ = extract from "Appendix JJ" of the Annual report of the Chief of Engineers for 1876; short comparative word list and grammatical notes on Mohave, Walapai, and Diegueno] 1877a "Die Sprache der Tonkawas", ZEthn 9: 64-73. 1877b "Der Yuma-Sprachstamm", ZEthn 9 : 341-50, 365-418. *1882 "Indian languages of the Pacific states and territories and of the pueblos of New Mexico", Magazine of American history 8: 254-63. [Chimariko constitutes a family by itself; Washo shows analogy with Shoshone in phonetics] 1883 "Der Yuma-Sprachstamm; zweiter Artikel", ZEthn 15: 123-47. 1886 "Der Yuma-Sprachstamm 3", ZEthn 18: 97-122. 1892 "Der Yuma-Sprachstamm 4", ZEthn 24: 1-18. [extensive vocabularies of Yuman languages, as well as Seri, with some comparative remarks] 1900

"The Waikuru, Seri, and Yuma languages", Science 12: 556-8.

Gifford, E. W. 1922 "California kinship terminologies", University of California publications in American archaeology and ethnology 18: 1-285. 1931 The Kamia of Imperial Valley (= Bureau of American ethnology, bulletin 97) (Washington: U.S. Government Printing Office), [lexical items]

BIBLIOGRAPHY

95

1932 "The Southeastern Yavapai", University of California publications in American archaeology and ethnology 29: 177-252. [some lexical items in the language] 1936 "Northeastern and Western Yavapai", University of California publications in American archaeology and ethnology 34: 247-354. [linguistic forms throughout the text] Gifford, E.W.-R.H. Lowie 1928 "Notes on the Akwa'ala Indians of Lower California", University of California publications in American archaeology and ethnology 23: 339-52. [some lexical items and a comparative word list for Paipai, Kiliwa, Cocopa] Goldstein, W.M. *1948 Review of SPIER (1946), AmA 5 0 : 108-10. Greenberg, Joseph H. * 1960 "The general classification of Central and South American languages", in: Men and Cultures, selected papers of the fifth International Congress of Anthropological and Ethnological Sciences (Ed.: Anthony F. C. Wallace) (Philadelphia: University of Philadelphia Press), pp. 791—4. [his Hokan includes Jicaque; Urumangi is listed as probably Hokan] Greenberg, Joseph H. - Morris Swadesh 1953 "Jicaque as a Hokan language", UAL 19: 216-22. Grekoff, George 1964 "A note on comparative Pomo", in BRIGHT (1964: 67-72). 1970ms "The structure and function of the nominalized clause in Chimariko". [paper prepared for the First Conference on Hokan Languages] s.a. Unpublished ms. on Chimariko phonology. Gursky, Karl-Heinz 1963a Die Stellung der Sprachen von Nordost-Mexico und Süd-Texas (= Abhandlungen der Völkerkundlichen Arbeitsgemeinschaft 4) (Nortorf, Holstein: U. Johannsen). 1963b "Algonkian and the languages of Southern Texas" AnL 5.9: 17-21. 1963c "Die Stellung der Waicuri-Sprache", Nachrichtenblatt der Völkerkundlichen Arbeitsgemeinschaft 6: 171-86. 1964 "The linguistic position of the Quinigua Indians", IJ AL 30: 325-7. 1965a "Ein lexikalischer Vergleich der Algonkin-Golf- und Hoka-SubtiabaSprachen", Orbis 14: 160-215. 1965b Das Proto-Hoka-Wort für 'Kaninchen' (= Abhandlungen der völkerkundlichen Arbeitsgemeinschaft 9) (Nortorf, Holstein: U. Johannsen). 1966a "Ein Vergleich der grammatikalischen Morpheme der Golf-Sprachen und der Hoka-Subtiaba-Sprachen", Orbis 15: 511-37. 1966b "On the historical position of Waikuri", IJ AL 32: 41-5. 1968 "Gulf and Hokan-Subtiaba: new lexical parallels", UAL 34: 21^1. Haas, Mary R. *1936a Review of HOIJER (1933), IJ AL 9: 122-4. * 1936b Review of HOIJER (1933), AmA 38: 115-6.

96

BIBLIOGRAPHY

1951 "The Proto-Gulf word for Water (with notes on Siouan-Yuchi)", IJAL 17: 71-9. 1954 "The Proto-Hokan-Coahuiltecan word for 'water' ", in CHRETIEN et alii (1954: 57-62). 1958 "A new linguistic relationship in North America: Algonkian and the Gulf languages", SJA 14: 231-64. 1959 "Tonkawa and Algonkian", AnL 1.2: 1-6. 1960 "Some genetic affiliations of Algonkian", in: Culture in history; essays in honor of Paul Radin (Ed.: S. Diamond) (New York: Columbia University Press), pp. 977-92. 1963 "Shasta and Proto-Hokan", Lg 39: 40-59. 1964 "California Hokan", in BRIGHT (1964: 73-87). 1966 "Historical linguistics and the genetic relationship of languages", in: Current trends in linguistics (Ed.: T.A. Sebeok) 3: Theoretical foundations (The Hague-Paris: Mouton), pp. 113-53. 1967

"On the relations of Tonkawa", in HYMES-BITTLE (1967 : 310-20).

1970a "Diminutive symbolism in Northern California: a study in diffusion", in: Languages and cultures of Western North America (Ed.: Earl H. Swanton, Jr.) (Pocatello: The Idaho State University Press), pp. 86-96. 1970b ms "The Northern California linguistic area", in LANGDON-SILVER (1970). [to appear] Hagen, V.M. von 1943 The Jicaque (Torrupan) Indians of Honduras (= Museum of the American Indian, Heye Foundation, Indian notes and monographs 53) (New York: Museum of the American Indian, Heye Foundation). Halpern, A.M. 1942 "Yuma kinship terms", AmA 44 : 425-41. 1946a "Yuma", in HOUER et alii (1946: 249-88). 1946b "Yuma 1, 2, 3", IJ AL 12: 25-33, 147-51, 204-12. 1947 "Yuma 4, 5, 6", IJAL 13: 18-30, 92-107, 147-66. 1962 "The reconstruction of Proto-Pomo". [tape recording of a lecture delivered at the University of California, Berkeley, May 3,1962] 1964 " A report on a survey of Pomo languages", in BRIGHT (1964: 8 8 - 9 3 ) . Hamp, Eric P. 1958 "Karok syllables", IJAL 24: 240-1. *1960 "Selected summary bibliography of language classification", SIL 15: 29-45. [for Hokan, pp. 34-5] *1966 "On two California grammars", IJAL 32: 176-88. [includes review of BRIGHT (1957)]

1970ms "On methods and goals in comparative Hokan studies", [paper prepared for the First Conference on Hokan Languages] Hamp, Eric P. et alii 1963 "On aboriginal languages of Latin America", CAnthr 4:317-9.

BIBLIOGRAPHY

97

Harrington, John P. 1908 "A Yuma account of origins", Journal of American folklore 21: 324-48. 1913 [Announcement of relation of Yuman and Chumashan,] AmA 15: 716. 1917 [Announcement of relation of Washo and Chumashan,] AmA 19:154. 1930 "Karuk texts", UAL 6: 121-61. 1932 Karuk Indian myths (= Bureau of American ethnology, bulletin 107) (Washington: U.S. Government Printing Office). 1943 "Hokan discovered in South America", Journal of the Washington academy of sciences 33: 334-44. Hayes, A.S. *1954 "Field procedures while working with Diegueno", IJ AL 20: 185-94. [no linguistic data] Heizer, R.F. *1966 Languages, territories, and names of California Indian tribes (BerkeleyLos Angeles: University of California Press). Heizer, R.F. (ed.) 1952 "The Mission Indian vocabularies of Alphonse Pinart", University of California anthropological records 15: 1-84. [Salinan, Esselen, and Chumash data] 1955 "The Mission Indian vocabularies of H.W. Henshaw", University of California anthropological records 15: 85-202. [Chumash data] Henshaw, H.W. 1890 "A new linguistic family in California", AmA [old series] 3: 45-9. Herndndez, Fortunato * 1 9 0 2 Las razas indigenas de Sonora y la Guerra del Yaqui (M6xico: J. de Elizade). [Seri word lists, etc., from unpublished manuscripts of M C G E E , PINART, e t c . ; p p . 2 3 7 - 6 9 ]

Hewitt, J.N.B. 1898 "Comparative lexicography", in: W.J. McGee, "The Seri Indians", Bureau of American ethnology report 17: 9-344. [vocabularies of Yuman languages, presented for comparison with Seri; pp. 299-344] Hicks, Frederic *1963ms Ecological aspects of aboriginal culture in the Western Yuman area (University of California, Los Angeles). [Ph.D. dissertation; includes appendix of Yuman plant names] Hinton, Thomas B.-Roger C. Owen *1957 "Some surviving Yuman groups in Northern Baja California", America indigena 17: 87-101. Hockett, Charles F. *1948 Review of HOLIER et alii (1946), Lg 24: 183-8. Hoijer, Harry 1933 "Tonkawa, an Indian language of Texas", in: Handbook of American Indian languages 3 (Ed.: F. Boas) (New York: J. J. Augustin), pp. 1-148. [cf. HOIJER (1946b)]

98

BIBLIOGRAPHY

*1941

"Methods in the classification of American Indian languages," in: Language, culture, and personality; essays in memory of Edward Sapir (Ed.: L. Spier-Α. Hallowell-S. Newman) (Menasha, Wise.: Sapir Memorial Publication Fund), pp. 3-14. 1946a Introduction to HOIJER et alii (1946: 9-29). 1946b "Tonkawa", in HOIJER et alii (1946: 289-311). [a shorter version of HOIJER (1933)]

*1947 1949a 1949b *1954 Hoijer, 1946

Hoijer, *1965

Review of SPIER (1946), Journal of American folklore 60: 433-4. "An analytical dictionary of the Tonkawa language", UCPL 5: 1-74. "Tonkawa syntactic suffixes and anaphoric particles", SJA 5: 37-55. "Some problems of American Indian linguistic research", in CHRETIEN et alii (1954: 3-12). Harry et alii Linguistic structures of Native America (= Viking Fund publications in anthropology 6)(New York: Wenner-Gren Foundation for Anthropological Research). Harry - Eric P. Hamp - William Bright "Contributions to a bibliography of comparative Amerindian", IJ AL 31: 346-53. [for Hokan see pp. 350-1]

Hoyo, Eugenio del 1960 "Vocablos de la lengua quinigua," Humanitas, Anuario del Centre de Estudios Humanisticos, Universidad de Nuevo Leon 1.1: 489-515. [fragments of an extinct language of north-east Mexico, possibly Coahuiltecan; cf. TROIKE (1967b); English abstract in IJAL 27 (1961): 251] 1965 El cuadernillo de la lengua de los indios Pajalates (1732), por Fray Gabriel de Vergara, y El confesonario [sic] de indios en lengua Coahuilteca (= Publicaciones del Institute Technolögico y de Estudios Superlores de Monterrey, Serie Historia 3) (Monterrey, Mexico: Instituto Technolögico y de Estudios Superiores). [fragmentary data from the eighteenth century on Coahuilteco and another language of northeast Mexico] Hymes, Dell H. *1956 Review of BRIGHT (1954) and HAAS (1954), Lg 32: 585-8. •1958 Review of BRIGHT (1957), AmA 60: 612-3. 1959 "Genetic classification: retrospect and prospect", AnL 1.2: 50-66. 1967

"Interpretation of a T o n k a w a paradigm", in HYMES-BITTLE (1967: 264-78).

Hymes, Dell Η. - William E. Bittle (eds.) *1967 Studies in Southwestern ethnolinguistics; meaning and history in the languages of the American Southwest (The Hague-Paris: Mouton). Jacobsen, William H., Jr. 1958 "Washo and Karok: an approach to comparative Hokan", IJAL 24: 195-212. 1964ms Agrammarof the Washo language (University of California, Berkeley). [Ph.D. dissertation]

BIBLIOGRAPHY

99

1966 "Washo linguistic studies", in: The current status of anthropobgical research in the Great Basin, 1964 (Ed.: Warren L. d'Azevedo et alii) (= Desert research institute, technical report series S-H, social sciences andhumanitiespublication 1) (Reno, Nevada: Desert Research Institute, University of Nevada), pp. 113-36,259-64. 1 9 6 7 "Switch-reference in Hokan-Coahuiltecan", in HYMES - BITTLE ( 1 9 6 7 : 238-63).

1970ms "Observations on theYana stop series in relationship to problems of comparative Hokan phonology", in L A N G D O N - SILVER (1970). [to appear] Joel, Judith 1 9 6 4 "Classification of the Yuman languages" in B R I G H T ( 1 9 6 4 : 9 9 - 1 0 5 ) . 1966ms Paipai phonology and morphology (University of California, Los Angeles). [Ph.D. dissertation] Johnson, Frederick *1940 "The linguistic map of Mexico and Central America", in: The Maya and their neighbors (Ed.: C.L. Hay et alii) (New York: D. AppletonCentury), pp. 88-114. [details of geographical distribution; classification agrees with M A S O N (1940)] Kendall, Martha B. - William L. Coleman 1970ms "Directions in the study of Upland Yuman", in LANGDON - SILVER (1970). [to appear] Kozlowski, Edwin 1970ms "Havasupai syntax", [paper prepared for the First Conference on Hokan Languages] Kroeber, Alfred L. 1904 "The languages of the coast of California south of San Francisco", University of California publications in American archaeology and ethnology 2: 29-80. *1905 "Supposed Shoshoneans in Lower California", Am A 7 : 570-2. [dispells the mistaken notion of Uto-Aztecan languages spoken in Lower California; confirms they are Yuman; see LEÖN (1903)] 1907 "The Washo language of East Central California and Nevada", University of California publications in American archaeology and ethnology 4: 251-317. 1910 "The Chumash and Costanoan languages", University of California publications in American archaeology and ethnology 9: 237-71. 1911a "The languages of the Coast of California north of San Francisco Bay", University of California publications in American archaeology and ethnology 9: 273-435. 1911b "Phonetic constituents of the native languages of California", University of Californiapublications in American archaeology and ethnology 10: 1-12. 1911c "Phonetic elements of the Mohave language", University of California publications in American archaeology and ethnology 10: 45-96. 1913 "The determination of linguistic relationship", Anthropos 8: 389-401.

100

BIBLIOGRAPHY

1915 "Serian, Tequistlatecan, and Hokan", University of California publications in American archaeology and ethnology 11: 279-90. *1916 "California place-names of Indian origin", University of California publications in American archaeology and ethnology 12: 31-69. 1917 "California kinship systems", University of California publications in American archaeology and ethnology 12: 340-96. [includes kinship terms for Mohave, Washo, Eastern Pomo, Karok, Chimariko, Salinan, Chumash] *1925 Handbook of the Indians of California (= Bureau of American ethnology, bulletin 78) (Washington: U.S. Government Printing Office), [facsimile reproduction: (Berkeley, California: California Book Company, Ltd. 1953); identifies languages by genetic affiliation; some forms listed in the text] 1931 The Seri (= Southwest museum papers 6) (Los Angeles, California: Southwest Museum). [Cochimi, Yuma, and Cocopa word lists, presented for comparison with Seri] 1943 "Classification of the Yuman languages", UCPL 1: 21-40. 1955 "Linguistic time depth results so far and their meaning", IJ AL 21: 91-104. 1958 "An Atsugewi word list", IJ AL 24: 213-4. [a word list collected in 1900] Kroeber, Alfred L. - John P. Harrington 1914 "Phonetic elements of the Diegueno language", University of California publications in American archaeology and ethnology 11: 177-88. [also gives a Dieguefio-Mohave comparative vocabulary and lists some of the sound correspondences] Landar, Herbert 1960 "Semantic components of Tequistlatec kinship", IJAL 26: 72-5. 1968 "The Karankawa invasion of Texas", IJAL 34: 242-58. Langdon, Margaret 1968 "The Proto-Yuman demonstrative system", Folia linguistica 2.1-2: 61-81.

1970a Review of WARES (1968), Lg 46: 533-44. 1970b A grammar of Diegueno: the Mesa Grande dialect (= UCPL 66) (Berkeley-Los Angeles: University of California Press). 1970c ms "The Proto-Yuman vowel system", in LANGDON - SILVER (1970). [to appear] 1971 "Sound symbolism in Yuman languages", in SAWYER (1971:149-73). Langdon, Margaret - Leanne Hinton 1970ms "Object-subject pronominal prefixes in La Huerta Diegueno", in LANGDON - SILVER (1970). [to appear] Langdon, Margaret - Shirley Silver (eds.) 1970 Hokan studies: papers from the First Conference on Hokan Languages held in San Diego, California, April 1970 (The Hague: Mouton). [to appear]

BIBLIOGRAPHY

101

Law, H.W. 1961 "A reconstructed proto-culture derived from some Yuman vocabularies", AnL 3-4: 45-57. Lehmann, Walter 1915 "Über die Stellung und Verwandtschaft der Subtiaba-Sprache", ZEthn 47: 1-34. [recognizes Subtiaba-Tlappanec as a language family by very convincing comparative wordlists; one can see why people have stated these to be dialects of the same language] 1920 Zentral-Amerika 1: Die Sprachen Zentral-Amerikas in ihren Beziehungen zueinander sowie zu Süd-Amerika und Mexiko (Berlin: D. Reimer). [Jicaque, pp. 654-68 (vocabulary); Subtiaba, pp. 910-69 (grammatical sketch and vocabulary); Tlappanec, pp. 969-73. The last two are stated to be related, and a relation to Hokan languages of California is suggested; pp. 973-8] Lemley, Juan H.V. 1955 "The Tlapaneco root -AK-", El Mexico Antiguo 8: 279-82. Leön, Nicoläs *1903 "Los Comanches el dialecto cahuillo de la Baja California", Anales del Museo Nacional de Mexico la. epoca 7: 263-78. [contains supposed "Cahuilla" vocabulary from Ensenada, actually Yuman; cf. KROEBER (1905)] Leskien, A. °1876 Die Declination im Slavisch-Litauischen und Germanischen (= Preisschriften der Jablonowskf sehen Gesellschaft 19) (Leipzig: Hirzel). Lo'riot, J. * 1964 "A selected bibliography of comparative American Indian linguistics", IJ AL 30: 62-80. [for Hokan, see pp. 72-4] Loukotka, Cestmir *1968 Classification of South American Indian languages (= Latin American Center, University of California, Los Angeles, reference series 7) (Los Angeles: Latin American Center, University of California). Lowie, R.H. 1939 "Ethnographic notes on the Washo", University of California publications in American archaeology and ethnology 36: 301-52. [states Washo is either isolated stock or aberrant Hokan group] 1963 "Washo texts", AnL 5-7: 1-30. [essentially a supplement to the 1939 monograph, published posthumously by his wife] Mandelbaum, David G. (ed.) 1958 Selected writings of Edward Sapir in language, culture, and personality (Berkeley - Los Angeles: University of California Press), [contains reprints of SAPIR (1929a, 1929b), a biography, and a bibliography of SAPIR]

Mason, John A. 1918 "The language of the Salinan Indians", University of California publications in American archaeology and ethnology 14: 1-154. [Antoni-

102

BIBLIOGRAPHY

ano and Migueleno dialects both recorded, all earlier information on the language listed; fairly detailed description with texts, comparative dialectal lexicon] *1940 "The native languages of Middle America", in: The Maya and their neighbors (Ed.: C.L. Hay et alii) (New York: D. Appleton-Century), pp. 52-87. Massey, W.C. 1949 "Tribes and languages of Baja California", SJA 5: 272-307. McLendon, Sally 1964 "Northern Hokan (B) and ( Q : a comparison of Eastern Pomo and Yana", in B R I G H T (1964: 126-44). 1966ms The Eastern Pomo language (University of California, Berkeley). [Ph.D. dissertation] 1969 "Spanish loan words in Eastern Pomo", RomPh 23: 39-53. 1970a ms "The Proto-Pomo pronominal system", in LANGDON-SILVER (1970). [to appear] * 1970b ms "The reconstruction of Proto-Pomo". [forthcoming] McQuown, Norman * 1955a Review of B R I G H T (1954) and H A A S (1954), UAL 21: 73-7. *1955b "The indigenous languages of Latin America", AmA 57: 501-70. [follows MASON (1940): although he cites GREENBERG - SWADESH (1953), he does not include Jicaque in Hokaltecan] McQuown, Norman (ed.) *1967 Handbook of Middle American Indians 5: Linguistics (Austin: University of Texas Press). Meigs, P., Ill 1939 The Kiliwa Indians of Lower California (= Ibero-Americana 15) (Los Angeles: University of California Press). Merriam, C.H. 1926 The classification and distribution of the Pit River Indian tribes of California (= Smithsonian miscellaneous collections 78.3) (Washington: Smithsonian Institution), [notes the nine dialects of Achumawi and two of Atsugewi; gives comparative Achumawi-Atsugewi word list] 1930 "The New River Indians Ήό-höm-tah'-hoi", AmA 32: 280-93. Mierau, E. 1963 "Concerning Yavapai-Apache bilingualism", IJ AL 29:1-3. [reports on phonetic differences between native Apache and Apache as second language for native speakers of Yavapai; lists phonemes of Yavapai] Miller, Wick R. 1966 "Anthropological linguistics in the Great Basin", in: The current status of anthropological research in the Great Basin, 1964 (Ed.: Warren L. d'Azevedo) (Reno, Nevada: Desert Research Institute), pp. 75-112.

BIBLIOGRAPHY

103

Mixco, Μ. 1970ms "Historical implications of some Kiliwa phonological rules" in LANGDON - SILVER (1970). [to appear] 1971ms Kiliwa grammar (University of California, Berkeley). [Ph.D. dissertation] Moser, Edward 1961ms Number in Seri verbs (University of Pennsylvania). [M. A. thesis] 1963a "The two brothers who went away angry: a Seri text", Tlalocan 4: 157-60. 1963b "Seri bands", The Kiva 28: 14-27. Moser, Edward - Mary B. Moser 1961 Vocabulario Seri (= Serie de vocabularios indigenas Mariano Silva y Aceves 5) (Mexico: Instituto Lingüistico de Verano). [English abstract in IJAL 28 (1962): 203] 1965 "Consonant vowel balance in Seri (Hokan) syllables", Linguistics 16: 50-67. 1970ms "Seri noun pluralization classes", in LANGDON - SILVER (1970). [to appear] Moser, Mary B. 1964 "Seri blue", The Kiva 30: 27-32. Moshinsky, Julius 1970a ms "Historical Pomo phonology", in LANGDON - SILVER (1970). [to appear] 1970b ms Southeastern Pomo grammar (University of California, Berkeley). [Ph.D. dissertation] Murdock, G.P. *1960-61 Ethnographic bibliography of North America (New Haven: Human Relations Area Files), [also lists linguistic items. Of interest for Hokan are Karok, p. 79; Achumawi, pp. 85-6; Chimariko, p. 86; Pomo, pp. 90-2; Salinan and Esselen, p. 92; Shasta, pp. 92-3; Yana, p. 95; Chumash, pp. 100-2; Cochimi, p. 102; Diegueno, pp. 102-3; Kamia, pp. 103-4; Seri, p. 105; Washo, pp. 115-6; Coahuilteco, p. 294; Tonkawa, p. 295; Cocopa, pp. 311-2; Halchidhoma, p. 314; Havasupai, p. 314; Maricopa, p. 326; Mohave, pp. 327-9; Walapai, pp. 352-3; Yavapai, p. 353; Yuma, pp. 353-4] Nevin, Bruce E. 1970ms "Transformational analysis of some 'grammatical morphemes' in Yana", in LANGDON - SILVER (1970). [to appear] Newman, Stanley *1954 "American Indian linguistics in the Southwest", AmA 56: 626-44. [p. 632 summarizes the status of Yuman studies at that time] •1955 Review of HEIZER (1955), AmA 57: 1338-9. •1961 Review of SAPIR - SWADESH (1960), IJAL 27: 267-8. •1965 Review of BRIGHT (1964), IJAL 31: 361-3. Olmsted, David L. 1954 "Achumawi-Atsugewi non-reciprocal intelligibility", IJAL 20: 181-4. 1956 "Palaihnihan and Shasta 1: labial stops", Lg 32: 73-7.

104

BIBLIOGRAPHY

1957 1958a 1958b 1959 1961 1964

"Palaihnihan and Shasta 2: apical stops", Lg 33: 136-8. "Atsugewi phonology", IJAL 24: 215-20. "Tequistlatecan kinship terminology", SJA 14: 449-53. "Palaihnihan and Shasta 3: dorsal stops", Lg 35: 637-44. "Atsugewi morphology 1: verb inflection", IJAL 27: 91-113. A history of Palaihnihan phonology (= UCPL 35) (Berkeley - Los Angeles: University of California Press). *1965a Review of OSWALT (1964), AmA 67: 1018. 1965b "Phonemic change and subgrouping: some Hokan data", Lg 41: 303-7. 1966 Achumawi dictionary (= UCPL 45) (Berkely-Los Angeles: University of California Press), [contains a life and bibliography of Jaime de Angulo] 1967 "Tequistlatec ceremonies and the analysis of stereotypy", in HYMESBITTLE (1967: 6 8 - 8 8 ) .

1970ms "The contribution of Jeremiah Curtin [to Atsugewi]". [paper prepared for the First Conference on Hokan Languages] Oswalt, Robert L. 1958 "Russian loan words in Southwestern Pomo", IJAL 24: 245-7. 1960 "Gualala", Names 8: 57-8. [place name derived from Southwestern Pomo] 1961ms A Kashaya grammar (University of California, Berkeley). [Ph.D. dissertation] 1964a " A comparative study of two Pomo languages", in BRIGHT (1964: 149-62). [compares Southwestern and Central Pomo] 1964b "The internal relationships of the Pomo family of languages", ACIAm 35.2: 413-21. [and five-page 100-word diagnostic list in the Pomo languages] 1964c Kashaya texts (= UCPL 36) (Berkeley - Los Angeles: University of California Press). 1970ms "Comparative verb morphology of Pomo", in LANGDON - SILVER (1970). [to appear] 1971a "Inanimate imitatives in Pomo", in SAWYER (1971: 175-90). 1971b "The case of the broken bottle", IJAL 37: 48-9. Powell, John Wesley 1891 "Indian linguistic families of America North of Mexico", Bureau of American ethnology, report 7 : 1 - 1 4 2 . [reprinted in BOAS-POWELL ( 1 9 6 6 : 8 3 - 2 1 8 ) ; cf. THOMAS - SWANTON (1911)]

Powers, Stephen *1877 Tribes of California (= Contributions to North American ethnology 3) (Washington: Government Printing Office), [appendix contains word lists for Karok, Chimariko, Pomo, Chumash, Achumawi, Shasta] Preston, W.D. *1947 Review of HOIJER et alii (1946), IJAL 13: 59-66. Radin, P. 1919 "The genetic relationship of the North American Indian languages",

BIBLIOGRAPHY

105

University of California publications in American archaeology and ethnology 14: 489-502. [on typological-morphological grounds proposes "the genetic relationship of all Indian languages of North America"] 1932-33 "Notes on the Tlappanecan language of Guerrero", IJ AL 8:45-72. 1940 "Notes on Schultze-Jena's Tlappanec", Boletin Bibliografico de Antropologia Americana 4 : 7 0 - 4 . [cf. SCHULTZE-JENA (1938)] Redden, James E. 1966 "Walapai 1: phonology", "Walapai 2: morphology", UAL 32:1-16, 141-63. 1970ms "Walapai syntax: a preliminary statement", in LANGDON - SILVER (1970). [to appear] Rensch, Calvin 1969ms "Classification of the Otomanguean languages and the position of Tlappanec". [in press] Rivet, Paul 1926 "Les Malayo-Polynesiens en Am6rique", JSAm 18:141-278. 1942 "Un dialecte hoka Columbien, le yurumangi", JSAm 34: 1-59. Robins, R.H. *1966 Review of WATERHOUSE (1962), Lingua 16: 9 1 - 3 . Robles, Carlos U. 1965 "Investigacion lingüistica sobre los grupos indigenas del Estado de Baja California", Anales del Institute Nacional de Antropologia e Historia (Mexico) 17: 275-301. [contains 100-word comparative word lists for Kiliwa, Paipai, Diegueno, Ko'al Diegueno, and La Huerta Diegueno] Romney, A. Kimball 1967 "Internal reconstruction of Yuman kinship terminology", in HYMESBRRRLE (1967: 3 7 9 - 8 6 ) .

Sapir, Edward 1909 "Characteristic features of Yana", AmA 11: 110. [abstract of paper read by Sapir] 1910 "Yana texts", University of California publications in American archaeology and ethnology 9: 1-235. [with Yana myths collected by R.B. Dixon] *1911 Review of DIXON (1910), AmA 13: 141-3. 1916 "Terms of relationship and the levirate", AmA 18: 327-37. [Yahi kinship terms] 1917a "The status of Washo", AmA 19: 449-50. 1917b "The position of Yana in the Hokan stock", University of California publications in American archaeology and ethnology 13: 1-34. 1918 "Yana terms of relationship", University of California publications in American archaeology and ethnology 13: 153-73. 1920a "The Hokan and Coahuiltecan languages", IJ AL 1: 280-90. 1920b "A note on the first person plural in Chimariko", IJ AL 1: 291-94. 1920c Review of MASON (1918), IJAL 1: 305-9.

106

BIBLIOGRAPHY

1921a "A supplementary note on Salinan and Washo", IJ AL 2: 68-72. 1921b "A bird's-eye view of American languages north of Mexico", Science [new series] 54:408. 1922 "The fundamental elements of Northern Yana", University of California publications in American archaeology and ethnology 13: 214-34. 1923 "Text analyses of three Yana dialects", University of California publications in American archaeology and ethnology 20: 263-94. 1925 "The Hokan affinity of Subtiaba in Nicaragua", AmA 27 : 402-35, 491-527. 1929a "Central and North American languages", Encyclopaedia Britannicali (London - New York: Encyclopaedia Britannica) 5: 138-41. [reprinted in MANDELBAUM (1958: 169-78)] 1929b "Male and female forms of speech in Yana", Donum natalicium Schrijnen; verzamelingvanopstellen door oud-leerlingen en bevriende vakgenooten opgedragen aan Mgr. Prof. Dr. Jos. Schrijnen bij gelegenheid van zijn zestigsten verjaardag 3 Mei 1929(Red.: St. W.J. Teeuwen et alii) (Nijmegen - Utrecht: Dekker - van de Vegt), pp. 79-85. [reprinted in MANDELBAUM (1958: 206-12)] Sapir, Edward - Leslie Spier 1943 "Notes on the culture of the Yana", Anthropological records 3: 239-97. [contains many lexical items] Sapir, Edward - Morris Swadesh 1960 "Yana Dictionary", UCPL 22: 1-267. Sawyer, Jesse (ed.) *1971 Studies in American Indian languages (= UCPL 65) (Berkeley - Los Angeles: University of California Press), [papers dedicated to Mary R. Haas] Schuller, Rudolf *1931 "Bibliography of American linguistics 1926-1928", 1JAL 6: 69-75. [has a few items relating to Hokan] Schultze-Jena, Leonhard 1938 Indiana 3: Bei den Azteken, Mixteken und Ttapaneken der Sierra Madre del Sur von Mexiko (Jena: Gustav Fischer), [grammatical notes, texts, and vocabulary for Tlappanec; cf. R A D I N ( 1 9 4 0 ) ] Sebeok, Theodore • 1 9 4 4 Review of KROEBER ( 1 9 4 3 ) , UAL 1 0 : 2 1 3 - 4 . Seiden, William 1963ms Havasupai phonology and morphology (Indiana University). [Ph.D. dissertation] Shaterian, Α. V. 1970ms "Yavapai [+sonorant] segments", in LANGDON - SILVER (1970). [to appear] 1971ms Yavapai phonology Silver, Shirley 1964 "Shasta and Karok: a binary comparison", in BRIGHT (1964: 170-81) *1966a Review of OLMSTED (1964), IJAL 32: 210-2.

BIBLIOGRAPHY

107

1966b The Shasta language (University of California, Berkeley). [Ph.D. dissertation] 1970ms "Comparative Hokan and the Northern Hokan languages", in LANGDON - SILVER (1970). [to appear] Simpson, J.H. 1876 Report of explorations across the Great Basin of the territory of Utah in 1859, Engineer Department, U.S. Army (Washington: Government Printing Office). [Washo word list, pp. 467-74] Spier, Leslie 1923 "Southern Diegueno customs", University of California publications in American archaeology and ethnology 20: 297-358. 1924 "Havasupai texts", IJAL 3: 109-16. 1946 Comparative vocabularies and parallel texts in two Yuman languages of Arizona (= University of New Mexico publications in anthropology 2) [Havasupai and Maricopa] (Albuquerque: University of New Mexico Press). Swadesh, Morris *1938 "Bibliography of American Indian linguistics 1936-37", Lg 14:318-23. [poor years for Hokan-Coahuiltecan, only HAAS (1936)] 1954a "Perspective and problems of Amerindian comparative linguistics", Word 10 : 306-32. 1954b "Time depths of American linguistic groupings", AmA 56: 361-4. *1963 Review of WATERHOUSE (1962), AmA 65: 1197-8. 1967a "Linguistic classification in the Southwest", in HYMES-BITTLE(1967: 281-309). [specifically pp. 289-90, 295], 1967b "Lexicostatistic classification", in MCQUOWN (1967: 79-115). Swanton, John R. 1915 "Linguistic position of the tribes of Southern Texas and Northeastern Mexico", AmA 17: 17-40. [cf. TROIKE (1963)] 1940 Linguistic material from the tribes of Southern Texas and Northeastern Mexico (= Bureau of American ethnology, bulletin 127) (Washington: U.S. Government Printing Office). Talmy, Leonard 1970ms "The Atsugewi instrumental-prefix system", [paper prepared for the First Conference on Hokan Languages] Tax, S. *1960 "Aboriginal languages of Latin America", CAnthr 1: 431-6. Taylor, W.W. 1961 "Archaeology and language in Western North America", American antiquity 27: 71-81.

Teeter, Karl V. * 1970 Review of BRIGHT (1964), Lingua 24: 407-9. Thomas, Cyrus - John R. Swanton * 1911 Indian languages of Mexico and Central America and their geographical distribution (= Bureau of American ethnology, bulletin 44) (Washington: Government Printing Office), [extension to the south of

108

BIBLIOGRAPHY

POWELL (1891); lists sources; has map] Troike, Rudolph C. 1959ms A descriptive phonology and\ morphology of Coahuilteco (University of Texas). [Ph.D. dissertation] 1961 "A Nahuatl loan-word in Coahuilteco", UAL 27: 172-5. 1963 "A contribution to Coahuilteco lexicography", IJ AL 2 9 : 2 9 5 - 9 . [elaboration and correction of SWANTON (1915); he states Tonkawa is now linked to Algonkian] 1967a " A structural comparison of Tonkawa and Coahuilteco", in HYMESBITTLE (1967: 321-2). 1967b Review of HOYO (1960, 1965), IJ AL 33: 78-82. [lists all identifiable forms of Coahuilteco found in this work, with English glosses] 1970ms "The linguistic classification of Cochimi", in LANGDON - SILVER (1970). [to appear] Trubetzkoy, N.S. °1939 "Gedanken über das Indogermanenproblem", Acta linguistica 1: 81-9. Turner, Paul R. 1967a "Serf and Chontal (Tequistlateco)", IJ AL 33: 235-9. 1967b "Highland Chontal phonemics", AnL 9.4: 26-32. 1968 "Highland Chontal clause syntagmemes", Linguistics 38: 77-83. 1969 "Proto-Chontal phonemes", UAL 35: 34-7. 1970ms "Pluralization of nouns in Seri and Chontal", in LANGDON - SILVER (1970). [to appear] Turner, Paul - Shirley Turner 1971 Chontal to Spanish-EnglishjSpanish to Chontal dictionary (Tucson: The University of Arizona Press). Uldall, Η. T. 1932-33 "Sketch of Achumawi phonetics", IJAL 8: 73-7. Velten, Η. V. *1947 Review of SPIER (1946), IJAL 13: 187-9. Vihman, Eero 1970ms "On pitch and accent in Northern Pomo", in LANGDON - SILVER (1970). [to appear] Voegelin, C. F. *1941 "North American Indian languages still spoken and their genetic relationships", in: Language, culture andpersonality; essays in memory of Edward Sapir (Ed.: L. Spier-A. Hallowell - S . Newman) (Menasha, Wise.: Sapir Memorial Publication Fund), pp. 15^40. [his group 6 is SAPIR'S Hokan-Siouan] 1942 "Sapir: insight and rigor", AmA 44: 322-4. 1946 "Notes on Klamath-Modoc and Achumawi dialects", IJAL 12: 96-101. Voegelin, C. F. - F. M. Voegelin 1964 "Languages of the world: Native America fascicle one", AnL 6.6. 1965 "Languages of the world: Native America fascicle two", AnL 7.1.

BIBLIOGRAPHY 1966

109

Mapof North American Indian languages, compiled for the American Ethnological Society.

Voegelin, C.F. - F.M. Voegelin - Noel W. Schutz, Jr. *1967 "The language situation in Arizona as part of the Southwest culture area", in

HYMES-BITTLE (1967: 403-51). [especially section

G:

"The exclusively Southwestern Yuman family", pp. 419-28] Voorhees, J.E. 1959ms The formal analysis and comparison of Yuman kinship systems (Stanford University). [M.A. thesis] Wares, A.C. 1968 A comparative study of Yuman consonantism (= Janua linguarum, series practica 57) (The Hague-Paris: Mouton). Waterhouse, Viola 1949a "Oaxaca Chontal: sentence types and text analysis", El Mexico antiguo 7: 299-314. 1949b "Learning a second language first", I J AL 15: 106-9. 1957 "Two Oaxaca Chontal words", I J AL 23: 244-5. 1962 The grammatical structure of Oaxaca Chontal (= Publications of Indiana University Research Center in anthropology, folklore, and linguistics 19, =IJAL 28.2.2) (Bloomington: Indiana University). 1967

"Huamelultec Chontal", in MCQUOWN (1967: 349-67).

1969 "Oaxaca Chontal in reference to Proto-Chontal", IJ AL 35: 231-3. 1970ms "Another look at Chontal and Hokan", in LANGDON - SILVER (1970). [to appear] Waterhouse, Viola - William R. Merrifield 1968 "Coastal Chontal of Oaxaca kinship", Ethnology 7: 190-5. Waterhouse, Viola - M. Morrison 1950 "Chontal phonemes", UAL 16: 35-9. Waterhouse, Viola - Muriel Parrott 1970 Diccionario de la Sierra Chontal (Cherän, Michoacän, M6xico: [no publisher]). Waterman, Τ. T. 1910 "The religious practices of the Dieguefio Indians", University of California publications in American archaeology and ethnology 8: 271-359. [has some words in the language; makes observations on Diegueno dialects] Webb, Nancy M. 1971 A statement of some phonological correspondences among the Pomo languages (= I J AL 37.2, = Memoir 26) (Baltimore, Md.: Indiana University), [an unusually large number of errors (typographical?) reduce the usefulness of the work] Webb, Nancy M. - Carolyn F. Wall - Louise Tanous 1965-66 "Consonant and vowel correspondences among the Pomo langua ges", SIL 18: 59-70. Weitlaner, Roberto J. 1948 "Un idioma desconocido del Norte de Mexico", in: Actes du 28*

110

BIBLIOGRAPHY

congris international des Americanistes, Paris 1947 (Paris: Societ6 des Americanistes), pp. 205-27. Weitlaner, Roberto J. - Irmgard Weitlaner de Johnson 1943 "Acatlany Hueycantenango, Guerrero", El Mixico antiguo6:140-202. Winter, Werner 1957 "Yuman languages 1: first impressions", IJAL 23: 18-23. 1966 "Yuman languages 2: Wolf's Son - a Walapai text", IJAL 32:17-40. 1967 "The identity of the Paipai (Akwa'ala)", in H Y M E S - B I T T L E (1967: 372-8). 1970a "Reduplication in Washo: a restatement", IJAL 36: 190-8. 1970b ms "Switch reference in Yuman languages",in L A N G D O N - SILVER (1970). [to appear] Worth, D.S. 1960 "Russian kniga, Southwestern Pomo kalikak", IJAL 26: 62-6.

7.2 I N D E X TO T H E B I B L I O G R A P H Y 7.2.1 Bibliographic information: Almstedt 1970 (Diegueno); Bright 1955 (Hokan-Coahuiltecan), 1957 (Karok), 1964 (California), 1967a (Middle America); Crawford 1966 (Cocopa); Croft 1948 (extinct North American languages); Davis 1961 (literature for 1951-60); d'Azevedo - Price 1963 (Washo); Fernändez de Miranda 1967 (Middle America); Hamp 1960 (language classification); Hoijer et alii 1965 (comparative Amerindian); Jacobsen 1964 (Washo), 1966 (Washo); Langdon 1970b (Diegueno); Loriot 1964 (comparative American Indian linguistics); Mandelbaum 1958 (Sapir); McLendon 1966(Eastern Pomo); Murdock 1960-61 (ethnography of North America); Olmsted 1966 (de Angulo); Oswalt 1961 (Southwestern Pomo), 1964c (Pomo languages); Powell 1891 (linguistic sources for classification); Schuller 1931 (American Indian linguistics 1926-28); Silver 1966b (Shasta); Swadesh 1938 (American Indian linguistics 1936-37); Troike 1959 (Coahuilteco); Waterhouse 1962 (Chontal). 7.2.2 Languages Hokan and Hokan-Coahuiltecan: Bright 1956, 1970a, 1970b; Brinton 1891; Crawford, J.G. 1970; Crawford, J.M. 1970b; Dixon - Kroeber 1913a, 1913b, 1919; Greenberg - Swadesh 1953; Haas 1954, 1963, 1964; Hamp 1970; Harrington 1913, 1917; Hymes 1956; Jacobsen 1958, 1967, 1970; Kroeber 1915; Langdon - Silver (to appear); Olmsted 1965b; Sapir 1917b, 1920a, 1921a, 1925; Silver 1970; Waterhouse 1970. Northern Hokan: Bright 1954; McLendon 1964; Silver 1964, 1970. Extensions of the stock: Bright 1955 (Naolan); Greenberg - Swadesh 1953 (Jicaque); Gursky 1963-68 (Algonkian-Gulf, Quinigua, Waikuru);

BIBLIOGRAPHY

111

Harrington 1943 (Quechua); Landar 1968 (Carib); Rensch 1969 (Otomanguean); Rivet 1942 (Yurumangi); Sapir 1921b, 1925, 1929a (HokanSiouan); Swadesh 1967a (Macro-Mayan phylum). Karok Classification and comparison: Jacobsen 19S8; Silver 1964. Descriptive: Bright 1952, 1957; Bright - Bright 1965; de Angulo - Freeland 1931; Hamp 1958, 1966; Harrington 1930, 1932; Kroeber 1911a. Chimariko Classification and comparison: Dixon 1910; Sapir 1920b. Descriptive: Dixon 1910; Grekoff s.a., 1970; Sapir 1920b. Shasta-Achumawi: Dixon 1905, 1906b; Haas 1963; Olmsted 1956, 1957, 1959, 1965b. Shasta Classification and comparison: Dixon 1931; Haas 1963; Merriam 1930; Olmsted 1956,1957, 1959; Silver 1964. Descriptive: Bright - Olmsted 1959; Silver 1966b. Achumawi-Atsugewi (Palaihnihan): Baumhoff - Olmsted 1963, 1964; Olmsted 1956,1957,1959. Classification and comparison: Merriam 1926; Olmsted 1954, 1964; Voegelin 1946. Achumawi (descriptive): de Angulo 1926; de Angulo - Freeland 1930; Olmsted 1966, 1970; Uldall 1932-33. Atsugewi(descriptive): Garth 1944; Kroeber 1958; Olmsted 1958a, 1961,1970; Talmy 1970. Yarn Classification and comparison: Jacobsen 1970; McLendon 1964; Sapir 1917b; Sapir - Swadesh 1960. Descriptive: Nevin 1970; Sapir 1909, 1910, 1916, 1918, 1922, 1923, 1929b; Sapir - Spier 1943; Sapir - Swadesh 1960. Pomo Classification and comparison: Barrett 1908; Grekoff 1964;Halpern 1962,1964; McLendon 1964 (Eastern), 1970a, 1970b; Moshinsky 1970a; Oswalt 1964a, 1964b, 1970, 1971a; Webb 1971; Webb - Wall - Tanous 1965-66. Descriptive: Barrett 1908; de Angulo 1927 (Eastern); Kroeber 1911a (Eastern); McLendon 1966 (Eastern), 1969 (Eastern); Moshinsky 1970b (Southeastern); Oswalt 1958 (Southwestern), 1960 (Southwestern), 1961 (Southwestern), 1964c (Southwestern), 1971b; Vihman 1970 (Northern); Worth 1960 (Southwestern).

112

BIBLIOGRAPHY

Washo Classification and comparison: Jacobsen 1958, 1966: Sapir 1917a, 1921a. Descriptive: Dangberg 1922, 1927; Jacobsen 1964; Kroeber 1907; Lowie 1963; Simpson 1876; Winter 1970a. Esselett Classification and comparison: Henshaw 1890. Descriptive: Heizer 1952; Kroeber 1904. Salinem Classification and comparison: Sapir 1921a. Descriptive: Heizer 1952; Kroeber 1904; Mason 1918 (Antoniano and Miguelefio). Chumash Classification and comparison: Beeler 1964,1970a. Descriptive: Applegate 1970 (Ineseno), 1971; Beeler 1964 (Venturefio), 1967 (Venturefio), 1970a (Barbarefio), 1970b (Barbareno); Heizer 1952, 1955; Kroeber 1904 (Ineseno), 1910. Yuman Classification and comparison: Biggs 1957; Frisch - Schutz 1967; Gatschet 1876, 1877b, 1883, 1886, 1892, 1900; Gifford - Lowie 1928; Harrington 1908; Hewitt 1898; Joel, 1964; Kendall - Coleman 1970; Kroeber 1905, 1931, 1943; Kroeber - Harrington 1914; Langdon 1968, 1970a, 1970c, 1971; Langdon - Hinton 1970; Law 1961; Massey 1949; Mixco 1970; Robles 1965; Romney 1967; Spier 1946 (Havasupai, Maricopa); Troike 1970; Voegelin - Voegelin - Schutz 1967; Voorhees 1959; Wares 1968; Winter 1957, 1967 (Paipai), 1970b. Descriptive: Almstedt 1968 (Dieguefio); Bright 1965b (Dieguefio); Castetter Bell 1951 (Mohave, Cocopa, Yuma, Maricopa); Crawford 1966(Cocopa), 1970a (Cocopa); Dedrick 1958 (Yuman dialects of Baja California); Frisch 1968 (Maricopa); Gifford 1931 (Dieguefio), 1932 (Yavapai), 1936 (Yavapai); Gifford - Lowie 1928 (Paipai); Halpern 1942 (Yuma), 1946a (Yuma), 1946b (Yuma), 1947 (Yuma); Hayes 1954 (Diegueno); Hicks 1963; Joel 1966 (Paipai); Kozlowski 1970 (Havasupai); Kroeber 1911c (Mohave); Kroeber - Harrington 1914 (Dieguefio); Langdon 1970b (Dieguefio); Langdon - Hinton 1970 (Dieguefio); Le6n 1903; Meigs 1939 (Kiliwa); Mierau 1963 (Yavapai); Mixco 1971 (Kiliwa); Redden 1966 (Walapai), 1970 (Walapai); Robles 1965 (Kiliwa, Paipai, Diegueno); Seiden 1963 (Havasupai); Shaterian 1970 (Yavapai), 1971 (Yavapai); Spier 1923 (Diegueno), 1924 (Havasupai); Troike 1970 (Cochimi); Voorhees 1959; Waterman 1910 (Diegueno); Winter 1966 (Walapai). Seri Classification and comparison: Gatschet 1900; Hewitt 1898; Kroeber 1915, 1931; Turner 1967a, 1970.

BIBLIOGRAPHY

113

Descriptive: Bowen-Moser 1968;" Hernändez 1902; Hewitt 1898; Kroeber 1931; E. Moser 1961, 1963a, 1963b; Moser - Moser 1961, 1965, 1970; Μ. Moser 1964. Coahuiltecan Classification and comparison: Gursky 1963a; Sapir 1920a; Swanton 1915, 1940; Weitlaner 1948. Descriptive: Hoyo 1960 (Quinigua), 1965; Swanton 1940. Coahuilteco Classification and comparison: Troike 1963, 1967a. Descriptive: Troike 1959, 1961, 1963, 1967b. Tonkawa Classification and comparison: Haas 1959, 1967; Troike 1967a. Descriptive: Gatschet 1877a; Hoijer 1933, 1946b, 1949a, 1949b; Hymes 1967. Chorttal (Tequistlatecan) Classification and comparison: de Angulo 1925a; Kroeber 1915; Turner 1967a, 1970; Waterhouse 1969,1970. Descriptive: Belmar 1900,1905; Brinton 1892; de Angulo 1925b; de Angulo Freeland 1925; Landar 1960; Olmsted 1958b, 1967; Turner 1967b, 1968, 1969; Turner - Turner 1971; Waterhouse 1949a, 1949b, 1957,1962, 1967; Waterhouse - Merrifield 1968; Waterhouse - Morrison 1950; Waterhouse Parrott 1970. Subtiaba-Tlappanec Classification and comparison: Lehmann 1915, 1920; Weitlaner - Weitlaner de Johnson 1943. Subtiaba Classification and comparison: Sapir 1925. Descriptive: Lehmann 1920. Tiappattec (descriptive): Lehmann 1920; Lemley 1955; Radin 1932-33,1940; Schultze-Jena 1938. Jicaque (descriptive): Conzemius 1921-23; Lehmann 1920; Hagen 1943. 7.2.3 Methodology, theory, general reviews: Elmendorf 1965; Frachtenberg 1918; Gatschet 1882; Greenberg 1960; Haas 1966, 1970a, 1970b; Hamp et alii 1963; Heizer 1966; Hoijer 1941,1946a, 1954; Hymes 1959; Kroeber 1913,1955; Loukotka 1968; Mason 1940; McQuown 1955b; Powell 1891; Radin 1919; Sapir 1921b, 1929a; Swadesh 1954a, 1954b, 1967a, 1967b; Tax 1960; Thomas - Swanton 1911; Trubetzkoy 1939; Voegelin 1941, 1942; Voegelin - Voegelin 1964,1965,1966; Voegelin - Voegelin - Schutz 1967.

114

BIBLIOGRAPHY

7.2.4 Prehistory andglottochronology: Baumhoff - Olmsted 1963, 1964; Bright 1956; Greenberg - Swadesh 1953; Halpern 1964; Jacobsen 1966; Joel 1964; Kroeber 1955; Law 1961; Miller 1966; Oswalt 1964b; Swadesh 1954b; Taylor 1961; Winter 1967. 7.2.5 Ethnography: Baegert 1773; Barrett 1908 (Pomo); Bowen - Moser 1968 (Sen); Castetter - Bell 1951 (Yuman); Conzemius 1921-23 (Jicaque); Dixon 1910 (Chimariko); Gifford 1931 (Kamia, i.e., Diegueno), 1932 (Yavapai), 1936(Yavapai); Hicks 1963 (Western Yuman); Hinton-Owen 1957 (Yuman); Kroeber 1925 (California), 1931 (Sen); Lowie 1939 (Washo); Massey 1949 (Baja California); Meigs 1939 (Kiliwa); E. Moser 1963b (Sen); M. Moser 1964 (Seri); Sapir - Spier 1943 (Yana); Spier 1923 (Diegueno); Hägen 1943 (Jicaque); Waterhouse - Merrifield 1968 (Chontal); Waterman 1910 (Diegueno). 7.2.6 General: Boas-Powell 1966; Bright 1964; Bright-Bright 1965; Chafe 1962; Dixon 1906a; Dixon - Kroeber 1903,1907; Gifford 1922; Haas 1951, 1958,1959,1960,1970a, 1970b; Johnson 1940; Kroeber 1911b, 1916,1917; Powers 1877; Rivet 1926. 7.2.7 Acculturation: Bright 1952, 1960, 1967b; McLendon 1969; Oswalt 1958, I960, 1971b; Troike 1961; Worth 1960. 7.2.8 Anthologies: Bright 1964; Chretien et alii 1954; Hymes - Bittle 1967; Langdon - Silver (to appear); Mandelbaum 1958; McQuown 1967; Sawyer 1971. 7.2.9 Reviews: Bartholomew 1969; Beeler 1955; Bright 1965a; Diebold 1959, 1961; Goldstein 1948; Haas 1936a, 1936b; Hockett 1948; Hoijer 1947; Hymes 1956, 1958; Langdon 1970a; McQuown 1955a; Newman 1955, 1961,1965; Olmsted 1965a; Preston 1947; Robins 1966; Sapir 1911,1920c; Sebeok 1944; Silver 1966a; Swadesh 1963; Teeter 1970; Troike 1967b; Velten 1947.