In a Southern Wakashan Language

INFORMATION TO USERS

The most advanced technology has been used to photograph and reproduce this manuscript from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

University Microfilms International A Bell & Howell Information Company 300 North Zeeb Road. Ann Arbor. Ml 48106-1346 USA 313/761-4700 800/521-0600

Rethinking noun and verb: An investigation of AUX in a Southern Wakashan language

Renker, An n M., Ph.D. The American University, 1987

UMI 300 N. Zeeb Rd. Ann Aibor, M I 48106

In all cases this material has been filmed in the best possible way from the available copy. Problems encountered with this document have been identified here with a check mark V

1. Glossy photographs or pages.

2. Colored illustrations, paper or print _____

a Photographs with dark background____

4. Illustrations are poor copy______

5. Pages with black marks, not original copy _I

6. Print shows through as there is text on both sides of page.

7. Indistinct, broken or small print on several pages \/

8. Print exceeds margin requirements _____

9. Tightly bound copy with print lost in spine______

10. Computer printout pages with indistinct print.

11. Page(s) ______lacking when material received, and not available from school or author.

12. Page(s) ______seem to be missing in numbering only as text follows.

13. Two pages numbered ______. Text follows.

14. Curling and wrinkled pages v

15. Dissertation contains pages with print at a slant, filmed as received \/

16. Other

University Microfilms International

By Ann M. Renker submitted to the Faculty of the College of Arts and Sciences of The American University in Partial Fulfillment of The Requirements for the Degree of Doctor of Philosophy in Anthropology Signatures of Committee: Cha irman:

•ftt/tL bddjLyn*

Dean of'the College

Date

1987 The American University Washington, D.C. 20016

CEE infeBThw UNIVERSITY LIBRARY

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. RETHINKING NOUN AND VERB: AN INVESTIGATION OF AUX IN A SOUTHERN WAKASHAN LANGUAGE BY Ann M. Renker ABSTRACT The tenuous boundaries of form classes in Northwest Coast languages has been a subject of historical and contemporary linguistic investigation; the subject is especially important in relation to the universal quality of established grammatical categories such as noun and verb. This dissertation asserts that noun and verb are not present in Makah grammar, and the use of these entities in a description of Makah obscures the nature of the language. Instead, the analysis proposes the use of the category AUX as the means to disambiguate form class assignations for Makah. Because AUX is a subject of debate in its own right, the dissertation first confirms the existence of the category in Makah, a Southern Wakashan language, describes the morphology of the category, and then applies the stability of the category to the description of Makah sentences. AUX provides a means to systematically describe Makah sentence variation, discourse topic and

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the rudiments of clausal formation. Additional consequences relate to the universal nature of Alix as a sentential constituent, as well as the universality of the form classes noun and verb.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ACKNOWLEDGEMENTS

Many people contributed to the maintenance of my sanity and/or to my academic growth during the course of my graduate studies. Most prominently, I thank the people of Neah Bay and the Makah Indian Nation for their support and help while I conducted my dissertation fieldwork. I am also thankful that this support continued when I returned to Washington, D.C. to finish the analytic portion of the dissertation. The Makah Cultural and Research Center also deserves special recognition. Working at the facility provided me with the opportunity to learn many important lessons about the philosophy of anthropological field methods. Primarily, my experience at the Makah Cultural and Research Center forced me to define an ethical code which protected the humanistic qualities of anthropology as a discipline while upholding the rights of human beings within a research context. The staff of the Makah Cultural and Research Center, and especially the Makah Language Program staff, merit thanks and recognition in this regard. Helma Ward, Hildred Ides, Meredith Parker, Hamilton Greene, Huey

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Smith, Helen Peterson and John Thomas spent many hours patiently enduring my linguistic questions and insights. In addition, these elders offered me the help I needed to learn to understand and speak Makah. Without these teachers, I would not have been able to complete my research or my analysis. Other Makah people require special thanks. Grieg Arnold gave me a great deal of professional and personal support, as did Maria Parker-Pascua, Meri Flinn,: Keely Parker,- Kirk Wachendorf, and the remaining members of the Makah Language Program staff. Unfortunately, two people who also deserve special mention were killed in an accident years ago. I thank Arly and Kathy Flinn for their part in bringing me to Neah Bay and introduc ing me to the village and its people. In another part of the world, I owe thanks to my dissertation committee, especially Bill Leap, my dissertation advisor. He was instrumental in furthering my interest in American Indian Languages, and introduced me to the idea that linguistic and ethnographic research concerning Tribes should be accountable to. the Tribes themselves. T also thank Hanni Woodbury for initiating my interest in American Indian language study. Other individuals who deserve mention are Donna Longo, Paul Cissna and Steve Gill for providing conversation and opinions regarding my topic and the v

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. dissertation processJ Dale Kinkade and Richard Demers for commentary regarding an early draft of ideas in this dissertation; Mary Theobald and Ann Bannerman for help ing with the tedious process of preparing this document; and the Carson-Barnetts for giving me a family while I lived in Washington, D.C. Last, but not least, I thank my mother, my sister Elizabeth, and my brother Tommy, for their help and encouragement during the last eight years, I also thank my father and dedicate this dissertation to his memory.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE OF CONTENTS

ABSTRACT...... 7...... ii ACKNOWLEDGEMENTS...,...... iv LIST OF TABLES...... ix LIST OF ILLUSTRATIONS...... x INTRODUCTION...... 1 CHAPTER I. THE CONTEMPORARY CONTEXT: THE MODERN RESERVATION AND THE CULTURAL RENAISSANCE; THE MAKAH RESERVATION IN 1986.... 8 The Ozette Excavation...... 20 The Makah Cultural and Research Center...... 26 The Makah Language Program...... 28 II. MAKAH CULTURE.* PRE-CONTACT CLASSIFICATION AND LIFE AND A REVIEW OF THE SOURCES...... 35 Cultural Classification...... 35 Makah Pre-contact Culture...... 44 The Ethnographic Sources...... 54 III. THE MAKAH LANGUAGE...... 58 Linguistic Classification...... 58 The Makah Language in the Literature...... 63 General Phonetics...... 66 Syllable Structure and Phonotactics...... 68 Makah Words, Structures and Syntax...... 77 IV. THE AUX CONTROVERSY.* A LITERATURE REVIEW...... 83 Background: Literature on Kernel Sentence Structure...... 83 AUX as a "Problem” for Kernel Sentence Analysis...... 85 Arguments Favoring AUX as an Independent Sentential Constituent...... 89

vii

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The Argument Against AUX as an Independent Sentential Constituent...... 89 V. AN INSTANTIATION OF AUX IN MAKAH (AUXM) 105 The Analysis! Definitions and Methodology. 105 AUXm Applied to Makah...... 106 The Position of AUX*...... 110 Composition of AUXM ...... 114 The Fixed InternalMOrder of AUXw...... 127 Relationship to Adjacent Elements...... 129 The Properties of AUXM and Generalizations about AUXy...... 130 VI. MORE ON THE MORPHOLOGY AND FUNCTION OF AUXjj...... 134 The Absence of Negation Markers and Aspect Markers in AUXw...... 135 The Making of Discourse Topic in S^...... 143 VII. DESCRIBING THE MAKAH SENTENCE! THE SUPERIORITY OF USING AN AUX ANALYSIS INSTEAD OF A NOUN/VERB FRAMEWORK...... 152 Word Class Ambiguity in Makah...... 153 The Need for an AUX-Centered Analysis of Makah Sentences ...... 158 Makah Stems, Syntax, and AUXj...... 163 Makah Stems and Sentence Variation...... 164

VIII. AUXjj! LINGUISTIC IMPLICATIONS ...... 170 AUX^ and the Makah Language...... 170 AUXJJ and Language Universals...... 175

APPENDIX...... 180 BIBLIOGRAPHY...... 183

viii

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF TABLES

1. Makah Language Proficiency Statistics...... 33

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF ILLUSTRATIONS

Map 1. Map of Makah Territory...... 9

Chart 1. Makah Westcoast Correspondence...... 60 2. Nootkan Pharyngeal Correspondence...... 61 3. Stress in Makah...... 68 4. Makah Consonant Clusters...... 70 5. Phonotactics - Underlying Vowel...... 71 6. Phonotactics - Dsolated Forms...... 72 7. Past Tense Morphophonemic Changes.... 75 8. AUXw in Makah Sentences...... 116 9. Comparative Aspect...... 157 10. AUX and Positional Analysis - 2 Positions 160 11. AUX and Positional Analysis - 3 Positions 161 12. AUX and Positional Analysis - 3 Positions 163

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. INTRODUCTION

The title of this dissertation, "Rethinking Noun and Verb in Makah! An Investigation of AUX in a Wakashan Language," suggests two intersecting issues in this study: 1) the syntactic validity of a con stituent category AUX in Makah, and 2) the utility of employing AUX as a criterion for disambiguating the problematic form classes noun and verb for the southern Wakashan language. Generally speaking, this inquiry applies the inductive methodology developed in the influential Encyclopedia of AUX (Steele et al. 1981) to Makah, to ascertain whether AUX is a fundamental syntactic category in this language. This process termed instantiation, i.e. investigating instances of AUX in a single language, develops a syntactic profile of Makah which will refine its formulaic representation. While the dissertation concentrates specifi cally on the Makah language, universal implications follow. Because Northwest Coast languages exhibit little or no distinction between the traditional English- based form classes known as noun and verb (Kuipers 1968, Jacobsen 1979a, Kinkade 1983), information revealed in this investigation affects the issue of the universality 1

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. of noun and verb. If AUX does disambiguate noun and verb in Makah, the category may serve a similar function in other languages which appear to lack these two form classes. Conversely, the results of this analysis may also support the position that noun and verb are not universal form classes (Kinkade 1983). Universal implications also extend to the composition of the sentence. Contemporary Generative Grammarians write the basic universal sentence in one of two ways: 1, S -* NP AUX VP or 2, S ->■ NP VP where S= sentence, NP= noun phrase and VP= verb phrase. Cl) asserts three essential constituent categories, while (2) recognizes only two essential syntactic categories dominated directly by the sentence. Because the adequacy of any syntactic description relies on the; simplicity of the theory underlying it (Lyons 1968), a controversy regarding the correct placement of AUX in a universal grammar has arisen. The linguists who accept Cl) believe that three constituent categories (NP, AUX, VP) constitute the simplest sentence representation, while proponents of (2) believe only two categories (NP, VP) are directly dominated by the sentence. In the latter case, AUX is accepted to be a subset of VP; consequently, it is not dominated directly by the

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3

sentence CS). Phrase Markers (PM) for these respective orientations illustrate the structural differences between (1) and (2). 3. S NP AUX VP 4. S -*■ NP VP

S S / l \ /\ NP AUX VP NP VP /\ AUX V

Because this dissertation validates the instantiation of AUX in Makah its preparation makes additional data regarding constituency of the category available to the linguistic community. Makah data are particularly important in this regard, because the Wakashan language family is not currently subject to the methodology established in Steele et al. (1981). In fact, the absence of Makah syntactic data extends to the linguistic literature in general. Very few sources utilize sentence-level data (Jacobsen 1979, 1979a; Renker 1985a) because most concentrate on phonology, morphology and historical comparisions. Therefore, an extensive field experience had to accompany this investigation. Although field work is one hallmark of anthropology, the circumstances surrounding my time on the Makah Reservation are critically important to the

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4

preparation of this dissertation in two areas! 1) my perspective on the role of an anthropologist working for a Tribal organization and within a contemporary Tribal environment, and 2) my linguistic elicitation techniques. My initial involvement with the Makah Tribe occurred in 1980 June when the Makah Cultural and Research Center (MCRC) hired me to perform a reservation-wide survey. While living on-site for a month, the Makah Language Program (MLP) of the MCRC asked me to join the staff as a linguist for the 1982-1983 school year. I became the director of the MLP in 1983, and retained this position until I left the reservation in 1985 October to write this dissertation.

During this time on the reservation, my role as a Tribal employee, and then as a person with policy-making responsibilities in the language and culture sphere of the MRC, altered my perceptions about the affect of contempo rary ethnographies on American Indian communities. As a Tribal employee and a resident of Neah Bay for a number of years, I came to believe that ethnographic information in a contemporary time frame could have lasting detrimental effects on the Makah people and on my relationship with them. Makah people view this information as highly personal in nature. In fact, the personal nature of contemporary information is very specialized, because information is

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5

frequently restricted to members of one extended family. So, publication of ethnographic information can violate family confidences and well as confidences which are Tribal in nature, and should remain within the Tribal community. Consequently, the content of the dissertation reflects my concern with the relationship I have with Makah people on the reservation, and my access to certain types of information which came from my role as MLP director and resident in the village. I do not discuss contemporary ethnographic details (outside of the linguistic and cul tural programs at the MCRC) in this dissertation, but concentrate instead on ethnographic information which is already published or available to the academic community. From some points of view, this decision may lead to a restricted ethnographic profile in Chapter 2 \ even so, there is enough data in that chapter to suggest and impart contextual information about Makah culture which furthers the discussion of the Makah language required in the later chapters, My involvement with the MLP and the MCRC also affected my linguistic elicitation techniques. Because I was surrounded by"Makah elders and native speakers on a daily basis in the course of my job, I was able to gather a large body of syntactic data. The constant data-gathering allowed me an opportunity to receive

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6

personal linguistic coaching from a number of speakers, and afforded me the additional opportunity to discuss my findings as my investigation progressed. Because of this situation, I gradually stopped eliciting sentences from native speakers by providing an English sentence and asking for a translation; by the time I left the reservation, I had been giving speakers sentences in Makah and asking for their comments regarding the grammaticality and sophistication of the constructions. All of the sentences used in this dissertation come directly from Makah native speakers, or were initially constructed by me and then checked and accepted by one or more native speakers. I did not use archival u. sentences in this analysis, because the discussion centers on the language as it is spoken on the reserva tion today, not on the people who speak it. A few words about the organization of the document are now in order. The first three chapters deal specifically with the Makah language and culture. Chapter 1 presents a discussion of the contemporary Tribal efforts to safeguard the Makah language and culture through formal programs administered by the MCRC. These programs have drastically altered the academic literature regarding Makah language and culture, so this discussion precedes the literature reviews about Makah culture and Makah language, which appear, respectively in

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 7

Chapters 2 and 3. Once the Makah specific literature is presented, Chapter 4 progresses to the issues surrounding AUX in the linguistic literature. The Chapter discusses arguments for and against the inclusion of AUX on the sentential constituent level, and moves the discussion to Chapter 5. In this chapter, the methodology established in Steele et al. C1981) is applied to Makah in order to validate the existence of AUX in Makah. Since Chapter five does present a positive instantiation of AUX in Makah, the next chapter offers a discussion of the morphology and function of AUX for the southern Wakashan language. Chapter 6 provides additional syntactic insight relating to Makah, Chapter 7 applies AUX to the noun-verb discri mination problem. The final chapter concludes by discussing the implications of the AUX analysis of pre vious chapters on Makah grammar and Universal language studies.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER I

THE CONTEMPORARY CONTEXT: THE MODERN RESERVATION AND THE CULTURAL RENAISSANCE; THE MAKAH RESERVATION IN 1986

The Makah* Reservation occupies 114 sq. km. of land on the Olympic Peninsula of Washington state; its northern border is the Strait of Juan de Fuca, and western border is the Pacific Ocean (Map 1). The 1986 reservation also included a one square mile reser vation around the Ozette archeological site and two prominent offshore islands, Tatoosh (hupa*£akt) and 7 Waadah Cwa?ad?a). Congress restored Ozette to the

While Makah is the federally recognized name of the Tribe residing on the reservation surrounding Cape Flattery, the name is not the one to which the Makah language refers. The Tribe's ancestral name is qwidigta^a*tK, ’People of the Cape'. Makah is a derivation of the Salish word maq’a* , which was the term by which the Clallam Indian people referred to the Cape Flattery people. I use Makah in this dissertation because it is the official documentary name of the Tribe, and the name by which the Tribe is best known. The orthography used to write Makah language is an adaptation of an adaptation of the International Phonetic Alphabet. The original orthographic system for Makah was developed by William Jacobsen and put into community use via the Makah Language Program. Since 1979, the Makah Language Program has made several changes in the Jacobsen orthography, and I use this version of the alphabet in the dissertation. The alphabet is discussed in Chapter Three of the disser^ tation and in Appendix 1.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. without prohibited reproduction Further owner. copyright the of permission with Reproduced Map 1. Map of Territory Map Makah 1. Map PACIFIC OCEAN IDA RES. INDIAN 5— - £ c u W * RESERV. * W u c INDIAN 1 wa^ac e t t e z o MAKAH MAKAH 9 10

tribe in 1970, and the two islands came back to Makah jurisdiction in 1984. The territory which now comprises Makah land differs from the territory claimed aboriginally by the Tribe. The territory claimed by Makah speakers extended east from the Pacific Ocean to the Lyre River, and south from the Strait of Juan de Fuca to Ozette Lake, including the offshore islands Tatoosh and Waadah. Marine territory extended several miles off the coast of the land territory, and included important halibut and sealing banks. The Treaty of Neah Bay, signed in 1855, modified the original claim, and the present reser vation is a small portion of traditional territory CMCRC 1979, Renker and Gunther in press). The boundaries of Makah territory, whether aborig inal or contemporary, lie within the environmental zone termed the Pacific Coastal Forest (Farb 1963:190). The zone, which extends from western Alaska to central California, is identified by a number of shared environmental co.al*ions: temperate climate, abundant moisture in the fprm of lain and ocean winds, and extensive forests composed of towering trees. Sub-zones are recognized by the domination of one or two species of tree within a geographic area. For example, hemlock and spruce trees are plentiful in Alaska. And, while Douglas Fir and cedar predominate in British Columbia

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 11

and Oregon, redwoods abound in northern and central California (Farb 19631190). The entire Olympic Peninsula, including Makah territory, is a unique sub-zone of the Pacific Coastal Forest because the ranges of many trees overlap in the Olympic Forest. In addition, the varieties of trees often attain their greatest respective heights on the Olympic Peninsula, where the largest known Douglas Fir, Red Cedar and Western Hemlock are found (Farb 1963:197). The foundation for the lush forests on the Olympic Peninsula is water. Approximately twelve feet of rain falls on the area annually, and this produces the only temperate rain forest in the United States. Dense with moss, ferns and fallen logs which act as nurseries for seedling trees, the forests on the Olympic Peninsula and the Makah reservation are virtually impenetrable. Without specialized knowledge about locating direction in this environment, foot travellers could easily become hopelessly lost. These generalized environmental points aptly describe the Makah reservation, which receives approximately 90-1 IQ inches of rain annually. The majority of the rain falls during the winter months, which are marked by fierce storms bringing gale force winds to the coast. In contrast, the summers are calm and sunny, though the temperature varies little between the two seasons. In

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 12

1972, the average minimum and maximum temperatures were 46°F and 53°F, respectively, over the course of a year (Phillips and Donaldson 1972:35). While the Makah reservation is environmentally linked to the rest of the Olympic Peninsula, it is physically separated from other communities. The nearest settlement to the reservation is the non-Indian town of Clallam Bay, twenty miles east of the reservation on State Road 112. This road is the paved connection between the reser vation and other small towns on the west end of the peninsula! SR 112 is also the connection between the reservation and Port Angeles, the hub of the Olympic Peninsula’s north shore. Travel between the Makah reservation and Port Angeles, a distance of some seventy miles, is often restricted via SR 112 because of weather conditions, especially in the winter season. Mudslides frequently block the road, forcing travelers to use unpaved logging roads until the slide is cleared. In reality reservation inhabitants make the trip to and from Port Angeles much more often than the reverse. This situation is a result of the fact that Makah people rely on the stores and services which are located in Port Angeles and not on the reservation. Consequently, Makah people drive to Port Angeles regularly and in spite of the weather, while Port Angeles residents often restrict travel to the

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 13

reservation to the summer months. Compared to Port Angeles with its stores, restau rants, service stations, banks, library and a population of about 17,000, the Makah reservation seems small and distant from an outside perspective. Only one centralized village, Neah Bay, exists dn the reservation; the majority of the reservation's 1300 inhabitants live in this village on the northern coast of the reservation. About 800 Makah people live on the contemporary reservation CHall 1983). The remaining non-Indian population is largely composed of the service people and their families who are stationed at the two federal installations on the reservation: a United States Air Force base and a United States Coast Guard base. While the military personnel are served by stores and facilities on the respective bases, Makah people rely on a small number of private businesses for goods. These included one general store, two gas stations and two restaurants in 1986. Numerous motels and tourist establishments dot the main street in Neah Bay, but these are seasonal, and operate on short schedules in the winter, if at all. Health care, preschool education, family services and the like are available to Makah people, and because of the physical distance of the reservation from larger towns on the peninsula, the Makah Tribe is responsible

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 14

for providing and supervising most of the social services required by the reservation population. The most obvious exception is the school system on the reservation. The Neah Bay Elementary and High School are a part of the Cape Flattery School District, and are under the jurisdiction of the Washington State public school system. Government on the reservation rests with the Makah Tribal Council, a five person body elected by the 3 enrolled Makah people on the reservation. The Tribal Council came into being when the Tribe accepted the Indian Reorganization Act and established a constitu tional government. Congress ratified the Makah Tribal Charter in 1937, and this system of government has been in existence since that time. The Makah Tribal Council actively supports the philosophy of Indian self-determination, which became legislative reality with the passage of the Indian Self- Determination and Education Assistance Act of 1975 CP.L, 96-638). This act encouraged Tribes to assume more control of their own respective destinies through education, economic development and the recognition of local priorities ("Cohen 1981:201-202). To this end, the

^Enrolled Makah people are the individuals who have had their Makah ancestry validated and recorded in the Makah Tribal Council office devoted to this task.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 15

Makah Tribal Council encourages local businesses and projects which benefit the reservation. Currently, the natural resources of the reservation are the major income sources for the Makah Tribal Council and for Tribal members CHall 1983). The natural resources on the reservation fuel three dominant industries for Makah people: fishing, logging and tourist trade CHall 1983). Because of the nature of these industries, much of the employment on the reser vation is seasonal and optimal in the summer months. Tourists double the population of Neah Bay from June through August to take advantage of the sport fishing for salmon and halibut. In addition, hikers and campers come to the reservation in the summer to take advantage of the spectacular scenery and the summer weather. While the seasonality of the fishing industry is obviously connected to natural cycles, the tourist industry is affected by the seasonal shifts as well. The previously described harsh winters intimidate many tourists, especially those who do not live on the Olympic Peninsula. Consequently, there is statistically much fewer visitation to the reservation from October to May of each year. The perception of the inaccessibility of the reservation during the winter compounds the outsider view that the reservation is distant, isolated and cut-off from the rest of the world.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 16

In contrast, the Makah people on the reservation believe their ancestral homeland is not as isolated and distant as outsiders perceive. While outsiders consider the unrelenting winters, the amount of rain, the rugged landscape and the distance between Neah Bay and other peninsula settlements as the criteria for judging isolation, Makah people look at internal reservation factors when assessing the perceived isolation of the reservation. While trips outside the reservation may be necessary for material goods and services, Makah people focus on a number of other factors when appreciating the lack of isolation an individual feels within the village. Perhaps the most apparent example of community cohesion for the outside world is the annual Makah Days festival. This three day celebration occurs on the third weekend in August and commemorates the 1924 congressional action which granted American citizenship to American Indian people. First held in 1926, Makah Days features canoe races, footraces, salmon bakes, bone games, a talent show and programs of Makah Tribal and family dances performed by children and adults, respec tively. Preparation for Makah Day begins in early winter, so there is very little time during the year when the community is not involved with this celebration.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 17

Community participation in local sports (high school football and basketball, tournament sports for adults) is well demonstrated by regular attendance, whether the games are in Neah Bay or far away. Supporters do not base participation on the presence of a particular individual or a team; the incentive is the support of Neah Bay and the Makah community as a whole. Additional examples of this community support are contributions of goods and services to special events at the Tribal museum, to fund raisers for school or tribal projects, and to community public relations efforts which may require travel. Other examples of this community solidarity have a Makah cultural dimension. Potlatches, or structured exchanges of information sanctioned by the exhibition of familial ceremonial rights, are frequently held in Neah Bay. Often called "dinners" or "parties" on- the modern reservation, these events reinforce relationships between all of the individuals in the village, and remind Makah people of the history of these relationships. Potlatches can be held to celebrate a joyous event such as a wedding, a birth, a birthday, or the conferral of a Makah name upon an individual. Potlatches are also held following funerals, the rescue of an individual from a deadly situation, or a disaster such as a fire. In the last case, the potlatch serves as a vehicle for

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 18

the collection of money and goods to help the family in trouble. A single extended family sponsors each dinner, though other Makah people make food or provide monetary contributions to support the meal which begins the potlatch. As many as six hundred people may attend the meal, which is served at the Makah Community Hall, a Tribally maintained building, in Neah Bay. After the meal is finished, additional gifts of money and goods are presented to the person or persons for whom the potlatch is held, all in accord with the traditions which surround the Makah potlatch. In spite of the unemployment and financial problems on the reservation, Makah people participate in the monetary and goods exchanges. It is common to see a presentation of money from an individual who is barely

/• solvent financially. The generosity of Makah people at potlatches demonstrates a deep sense of community with, and obligation to, the members of the village. Partici pation in potlatches also exemplifies the desire of the Makah people to perpetuate their ancestral traditions and values. This public reinforcement of Makah values and knowledge at potlatches is a part of life on the present reservation. But for decades, Makah cultural and linguistic practices were monitored and forbidden by

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 19

Indian AgentsJ consequently, many of the traditions had to go underground in order to escape the wrath of the Agency administrators and local missionaries (Renker and Gunther, in press). Deliberate intervention strategies designed to offset the effects of linguistic and cultural oppression did not begin until the 1960s, when individual elders took the responsibility of restoring Makah language and culture through classes in the elementary and high schools on the reservation. For the next decade, limited Makah language and culture classes provided the means for many Makahs to learn about traditional life. This situation changed drastically when the tradi tional Makah village of Ozette emerged from the shroud of mud which buried the village about 500 years ago. The enormity of this excavation fundamentally changed the linguistic and cultural profile of the reservation by providing the impetus for a cultural renaissance which spawned one of the most comprehensive cultural resource management programs operating on an American Indian reservation. Because the full import of this cultural renaissance and cultural resource management effort cannot be adduced without information about the Ozette excavation itself, the next section discusses the archeological effort and its impact on the Makah community. Once this information

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 20

is presented, the discussion proceeds to the state o£ the cultural resource management effort on the contempo rary reservation.

The Ozette Excavation Makah oral history often speaks of the times "of amazing energy, fabulous hunts, elaborate masks, speeches about the great way of being..." (Colfax 1982) and the creations of Makah artists and craftsmen who joined the supernatural with the real world. But for younger Makah people in the 1950s and 196 0s, there were few material objects which testified to the intensity and complexity of traditional Makah culture. The dearth of pre-contact Makah artifacts was problematic for another reason. Without pre-contact objects, descriptions in the literature of Makah life from this time period had to be based on a limited number of early ethnographies; these presented informa tion that was sometimes contradicted by Makah oral history. The ensuing conflict affected both Makah people and the academic community, and often placed the two groups at odds when the subject of Makah language and culture was at issue. Unfortunately, few Makah artifacts from pre-contact time were available to the Tribe or to academics. The climatic condition of heavy rainfall in the area

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 21

encouraged the deterioration of wooden and woven arti facts, leaving few objects which could reinforce Makah oral history or demonstrate the advanced nature of the pre-contact lifestyle. The Ozette excavation changed this situation forever. From 1970 until 1981, the Makah Tribe and archeologists from Washington State University excavated this traditional winter village which was buried by a catastrophic mudslide about 500 years ago. They recovered 55,000 artifacts, 40,000 structural remains and over 1,000,000 faunal remains (Wessen 1982:17, Huelsbeck 1983 :v) from Ozette, which was occupied year round at the time of the slide. The location of the site and the occurrence of severe mudslides there was common knowledge to Makah people. Oral history mentioned the problem of mudslides at Ozette, a premier whaling station on the Pacific Coast; some elders in Neah Bay even remembered living at Ozette as children, before the village was abandoned in 1917 because compulsory attendance at the Neah Bay school was enforced (Wessen 1982*.17). However, archeologists did not discover the site until 1947 (Kirk 1974:24); initial testing took place in 1955 by Stallard and Denman (Huelsbeck 1983:1), and the first archeological excavation occurred in 1966 and uncovered some 2000 years of continuous

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 22

occupation (Daugherty and Fryxel 1967, Huelsbeck 1983*. 1). Further excavation did not occur until a winter storm in 1970 washed away a large section of the sea wall at the site, exposing wooden artifacts and house planks. This promoted an excavation which uncovered the remains of five longhouses buried by a mudslide 4 approximately 500 years previously (Huelsbeck 198311). The factor which allowed these excacations to yield preserved artifacts is a thick, blue-grey which effectively prevented air from reaching the cultural material. Deterioration was thereby pre vented, and archeologists removed artifacts from the ground in.a remarkable state of preservation. Conservation of these artifacts began immediately, for exposure to air initiated decomposition. Artifact stability was achieved through the use of polyethelene glycol (PEG). This waxy substance infiltrated the molecular structure of the artifacts

This site is not the only valuable Makah site, though certainly the grandest and most visible. The Hoko River Site, the oldest site from western Washington state (MCRC 1985), is actually a complex of two sites, a fishing camp on the banks of the river, and a rockshelter near the mouth of the river. The latter is the only site of its type on the Northwest Coast of America. Like Ozette, Hoko River demonstrates the fact that Makah people ware fishing in this area in the pre-contact period. Radiocarbon dates places the Hoko site complex at some 2800 years b.p. (MCRC 1985).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 23

and replaced the water locked in the organic material when artifacts were immersed in the PEG. The excavation of Ozette and the preservation of the enormous wealth of pre-contact artifacts deeply affected Makah people and anthropologists who studied Northwest Coast culture. Aside from the sheer number of artifacts uncovered for study, the following points summarize the importance of Ozette site for Makah and non-Makah researchers\ 1. Since the slide was a catastrophic one, the items found frozen in time were ready for use by the original owners. Ozette artifacts are not scraps of refuse; these objects were actually in use when they were buried. Archeologists did not have to postulate the physical characteristics of an artifact based on a few scattered remains. 2. Because the slide buried a number of houses, the wide variety of artifacts and house environments paints a picture of pre-contact Makah life. Manufacturing sequences of certain artifacts are established in the archeological collection by the number of tools recovered at different stages in production, thus adding valuable information about Makah technology to the knowledge of material cul ture! Perhaps even more startling is the fact that

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 24

the data reveal status relationships among inhabi tants within one house and between houses (Huelsbeck 1983) . This establishes the fact that status and ranking were important in Makah pre-contact life, and not effects of the contact process. 3. The faunal remains provide data which when analyzed, described and empirically documented the yearly subsistence base of the tribe in pre-contact days; in conjunction with tools, patterns of collection, utilization and disposal become obvious (Gleeson, et al. 1981). 4* Because of the nature of the site and its contents, Makah people participated in every phase of the excavation process. Elders identified artifacts archeologists had never seen. Younger Makahs attended college, excavated with the archeologists, and became archeologists and anthropologists during the eleven years of the dig. While conservation of the artifacts initially proceeded with non-Indian technicians, Makah people again received the training to assume these responsibilities. And finally, because the Tribe decided that these treasures needed to remain on the reservation, the Makah Tribal Council initiated the creation of a Tribal museum to house the artifacts. Tribal members received the training to administer the collection and the facility

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 25

and continue to do so. 5. The artifacts from Ozette provided tangible evidence of the richness and complexity of Makah culture to Tribal members. Young people*s interest in Makah culture and heritage soared, and a reawakening of culture pride sparked many Tribal projects which addressed preserving and maintaining Makah culture in the village. The Makah Cultural and Research Center (MCRC) is the visible symbol which acknowledges and exemplifies the importance of Ozette. The two million dollar facility opened in 1979; 25% of the funding for the establishment came directly from the Tribal treasury. This contribu tion is very substantial when one considers the size of the Tribe. Sustained economic support for this institu tion from the Tribal Council shows the continuing importance of the MCRC for Makah people. Today, the MCRC is the facility responsible for coordinating and administrating linguistic and cultural projects which impact on the Makah Indian Nation. The next section discusses programs operating from the Makah Cultural and Research Center, and relates the programs to the language and cultural restoration efforts on the Makah reservation.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 26

The Makah Cultural and Research Center Because Ozette opened up the richness of pre-contact Makah culture to both Makah people and the academic community, the Makah Tribal Council created an organiza tion to administer the Ozette Collection and to respond to the interests of both groups of interested people. The Makah Cultural and Research Center (MCRC) was designed to administer the care and exhibition of Ozette artifacts and manage anthropological research on the reservation. Changes occurred within the MCRC as the staff became more involved in the management of the cultural resources identified on the reservation*, native speakers, traditional skills, Makah songs and stories, and the like. Instead of simply managing the research of others, the MCRC began to initiate research into linguistic and cultural matters, and to plan and implement strategies for the continuance of cultural resources on the reserva tion. The MCRC also established a set of research guidelines for individuals who wished to conduct investigations on the reservation. The increasing importance of research and cultural resource management led to the development of formal programs within the MCRC which did not directly bear on the exhibition of Ozette artifacts or the conservation efforts directed at the Ozette Collection. In addition to the galleries which exhibit Ozette artifacts, the MCRC

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 27

also houses a research archives which contains a compre hensive collection of Makah oral, photographic and written documents. This collection is critical because available documentary materials are an essential part of the MCRC cultural resource management strategy, and because the nearest repository for Makah documents outside the reservation is about 150 miles away in Seattle. In addition, the Makah Archival Collection provides children and adults in the Makah community with the resource to explore Makah language and culture through the documents of researchers in addition to the oral histories told in individual homes. The formal educational programs at the MCRC are designed to teach visitors to Neah Bay about Makah culture, history and contemporary life, and to "preserve and disseminate ancestral information within the Makah community" (Renker and Arnold: in press). One program in particular, the Makah Language Program CMLP) is described here in detail, because the research for this dissertation was conducted while the author was employed by the Tribe as the.director of the MLP, thus allowing for an immersion in Tribal linguistic issues. This immersion resulted in the collection of a wealth of linguistic data which could not have been gathered during more traditional fieldwork circumstances.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 28

The Makah Language Program The MLP began as the portion of the Ozette-inspired renaissance which concerned itself with reversing the dwindling number of Makah native speakers and the decreasing frequency of the use of the language in Neah Bay. Individuals organized the MLP to reintroduce and reinforce the Makah language in the reservation community, and the program received Tribal sanction in 1978. Three immediate goals were set for the program! 1. The preservation of the Makah language? 2. The restoration of the spoken language within the community living on the reservation? and 3. The education of reservation children to produce students who can compete with non-Indian children while maintaining Tribal heritage and culture. Since 1978, the MLP has worked actively to increase Makah language skills at all age and proficiency levels in the community. The first step, with funds from the National Endowment for the Humanities, NEH, was the publication of two Makah language books? one deals with the Makah alphabet (MLP 1979) and the other with Makah grammar (Jacobsen 1979). Plans for further steps depended on a number of factors, including the following! 1. The number of native speakers on the reservation, 2. The number of people with a limited Makah proficiency and the Makah proficiency level of these individuals.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 29

3. The number of children with some degree of Makah proficiency. The MLP needed these figures before the scope of an educational plan could be developed. In qjder to gather these data, the MCRC commissioned a r^gervation-wide language survey in the summer of 1980. The Makah Lanjfttuge Survey (Renker 1980) utilized a method previously tested on the NortlfgTn . Reservation and . conducted as a needs assessment for the Wisconsin Winne bago. In this survey, one respondent per household answered questions about his language skills, the language skills of his children, and alternatives surrounding language planning efforts. Results showed the following critical points! 1. Since 33% of the adults and 33% of the children exhibited a Makah proficiency, language loss factors had already taken effect and were now leveling off. 2. School involvement with Makah language and culture-- no matter how limited-- increases Makah proficiency. 37% of the school-age children, as compared with 25% of the preschoolers, had some Makah language proficiency. 3. Makah language proficiency has a positive impact on oral language skills. 92% of the children with Makah proficiency were judged by their parents to speak English "very well," as compared to their non proficient counterparts whose 'Very well” ratint was 64% of this population. This assessment was substantiated by a higher level of language arts CAT, (California Achievement Test) scores than reading CAT scores in Makah students. No favoring effect of Makah language instruction reached reading scores. The fact that Makah language proficiency accentuated positive English language ratings became the focus for

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 30

the first successful Title IV-B proposal submitted by the MLP. The Office of Indian Education granted the MLP a three year award in 1981; subsequent proposals have also been funded, the most recent awards from the Educational Services section of Title IV-B monies. Since 1984, when the first grant expired, three one-year awards have come from the Office of Indian Education to continue Makah language instruction in the public school* The growth of the educational' activities of the MLP increased as the technical skills of the Makah staff increased. Large bodies of curriculum materials have entered the MLP files. The Original 1981 plan (MCRC 1981) called for Makah classes in two elementary school grades? by 1986, Makah was taught in the Tribal Headstart program, grades K-8, High school classes and to adults in night classes (MCRC 1986). Though the program has operated at this level of instruction since 1984, teaching and testing materials are refined each year. This need stems from the progres sive nature of Makah language teaching in the school. Lessons were developed in 1981 for beginning and intermediate levels; curricular units needed little specification to grade level at this point. As the MLP began language instruction in more and more classes, the staff developed new curricula to meet the needs of

individual grade levels. Because the MLP now teaches at

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 31

all levels in the Makah educational system, the MLP curriculum must continue to reflect the nature of Makah language instruction (Renker and Arnold! in press). An example follows. Because Headstart instruction hegan in 1984, the Kindergarten curriculum developed for 1984 is obsolete. The obsolescence results from the fact that students entering Kindergarten in the next school year will have had previous language instruction in Headstart, while the Kindergarten curriculum was designed to target children with no prior language instruction. As these Headstart children move through the public school system, new curricula must accommodate earlier language learning. Despite the changes that occur in curriculum because of increased language facility, one primary function of Makah language lessons does not change. Curriculum is designed to incorporate elements of Makah culture, history, reservation geography and science, and Tribal priorities in school lessons. While these lessons are now geared to grade level, the focus of the effort remains the same--enhance the academic skills of the child through Makah language lessons. Fortunately, the success of the MLP educational effort is empirically demonstrated by independent evaluations CLeaP aa

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 32

The following points attest to the success of the MLP educational strategies in the Makah reservation context! 1. Language Arts CAT score increased from an average of 37 NCE's in 1978 Cpre-language classes) to an average of 55 NCE's in 1984. Independent evaluations link this CAT score increase to the presence of Makah language in the school. 2. In 1980, a "performance gap" existed between Makah and non-Makah children in CAT scores. Non-Makah children performed from 6.4 points to a high of 24.4 points better than Makah children. After five years of Makah instruction in the school, the performance gap has been reduced by 83.6%. 3. The drop out rate in the school decreased 2% since the inception of Makah language and culture classes. 4. Makah children now score higher, on the average, in the Mechanics section of the Language Arts CAT text (6 points) and in total Language Arts (2 points) section of the test, than non-Indian children. 5. The 1985 Reservation Household Survey (Renker!1985) indicates MLP success in a number of areas! a. The chart below summarizes the dramatic increase in Makah skills in the five-year period since 1980. The MLP began to teach Makah in the public schools in 1981. b. One hundred percent C100%) of the school-age children who rate as speaking English "very well" have a Makah competency, up from 92% recorded in 1980. Of the children who are rated as speaking no Makah, 0% speak English "very well". These data clearly link Makah competence with better English skills, and support the 1980 link between these two factors. c. Only 38,9% of the preschoolers not enrolled in the Tribal Daycare/Headstart program have a Makah competence, as opposed to 90% of the pre-schoolers who receive MLP instruction in Daycare/Headstart.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 33 TABLE 1.--Makah Language .Proficiency Statistics 1980 1985 +increase -decrease pre-school children 20% 59.3% +39.3% speak Makah pre-school children 33% 53.6% +20.6% understand Makah elementary children 36% 79.6% +43.6% speak Makah elementary children 37% 77.6% +40.6% understand Makah

It is clear that the reservation children react positively to the educational program supplied by the MLP. Through a continuous process of needs assessments and evaluations, the MLP plans to continue this success for the benefit of the Tribal community and the stability of the ancestral language. The MLP is a good example of the investment Makah people make to insure the stability of the Makah language and culture on the reservation. This program is an example of the means that Makah people nov employ to aid in the integration of the old and the new aspects of their culture. Ozette provided the impetus for the Makah community to take steps to manage the cultural resources still available in Neah Bay, by revealing the immense achievements of Makah people of the past. Programs like the MLP at the MCRC insure that contemporary Makah children will learn and understand the language and culture of their ancestors while still possessing the

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 34

skills needed to succeed in American society. This chapter provided part of the cultural milieu which has shaped the Makah language as it is spoken today, and which provided the syntactic data gathered for this dissertation. The next chapter discusses the pre-contact culture of Makah people, and classifies the Tribe within the cultural framework established for Northwest Coast Tribes.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 2

MAKAH CULTURE: PRE-CONTACT CLASSIFICATION AND LIFE AND A REVIEW OF THE SOURCES

Cultural Classification The Makah Culture is a member of the culture area termed the Northwest Coast in Drucker (1955) and Driver (1969). This culture area is geographically circumscribed to the environmentally distinct, thin coastal strip extending from Southeastern Alaska to Northern California (Drucker 1955:1, Driver 1969:20). By definition, a culture area is a geographical area occupied by a number of peoples whose cultures show a significant degree of similarity with each other and at the same time, a significant degree of dissimilarity with the cultures of the peoples of other such areas'(Driver 1969:17). Consequently, the cultures located within the North west Coast culture area share a number of character istic traits and trait complexes, including: 1. emphasis on achieved wealth as measured in property and hereditary rights, 2. complex patterns of social stratification which involved hereditary slavery, 3. a highly developed painting and woodcarving style,

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 36

4. a material culture based on the abundance of the wood resource in the area, especially when related to the absence of other "materials suitable for technological developments" (Drucker 1955:8), and 5; a subsistence pattern based on the utilization of marine resources and have a stable supply of anadromous fish; the combination of these two factors made Northwest Coast cultures surplus - oriented (Driver 1969:20) without an agricultural foundation. Pre-contact and Early Historic Makah culture exhibit all five of the broad culture are character istics in the preceding list. The specific ways in which the Makah culture and others manifest the generalized Northwest Coast areal characteristics provides the mechanism for dissecting the areal group into smaller cultural complexes. In this case, the smaller complex includes the cultures of the native people living on the west coast of Vancouver Island from Cape Cook for Port Renfrew in addition to the Makah people on the Olympic Peninsula (Drucker 1951:3, Arima 1983:1). Called the Nootkan cultural complex, similar traits pertaining to social organization, subsistence patterns and material culture (Drucker 1955:198-201) set these native peoples apart from other subdivisions within the inclusive Northwest Coast areal group. The synonomy of the cultural subdivision is complicated. Discussions centering on the generalized

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 37

subdivision initially used the term Nootkan to refer to the cultural complex (Curtis 1911, Drucker 1951, Drucker 1955) : Confusion resulted from the fact that the term Nootkan was applied equally to the cultural complex and to the group of people living at Nootka Sound on Vancouver Island (Linguistic uses of the term are discussed in 2.1). Some native people and scholars proposed that the cultural group be called West Coast, after the fact that the native people resided on the west coast of Vancouver Island (Renker 1980-1985, Arima 1983:v). The internal groups within the cultural complex could then be designated as northern, central and southern (Arima 1983:2-7), following Drucker (1951) and his use of Northern, Central and Southern Nootkan Tribes. An additional name for this cultural group of native people was proposed in 1980 by the Tribal Council which governs the Vancouver Island people from Cape Cook to Port Renfrew. The Tribal Council proposed that the culturally linked native peoples use the term nut'£atnu,i "Along the mountains," to refer to them selves (Renker 1980-1985). nu*ta*nu*3r refers only to the Vancouver Island members of the cultural complex; Makah people are not included because British and United States’ land settlements made Vancouver Island people Canadian citizens and Makah people American

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 38

citizens based on the geographical separation of the land areas. Following Drucker (1951), this dissertation uses the term Nootkan to refer to the cultural complex uniting the Native people from Cape Cook to Port Renfrew on Vancouver Island with the Makah people on the Olympic Peninsula of the United States. The dissertation also follows the convention established in Drucker (1951) which divides the Nootkan cultural complex into northern, central and southern divisions (4-6); Makah culture is a Southern Nootkan Tribe. The dissertation does not use the term West Coast in cultural discussions, restricting use of that term to linguistic contexts. Drucker (1951) analyzes the uniqueness of Nootkan society in relation to two integrative concepts, "first, that of hereditary rank; and second, the importance of kinship, no matter how distant, as the basis for amicable social relationship" (719). The preoccupation with ranking created a social continuum marked at the top by headmen and their families, and by slaves at the bottom. Between the two fixed ends, Nootkans in the middle social strata could advance or fall back slightly in prestige through shrewd choices in marriage and manipulations in the kinship system (245),

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 39

Primogeniture was the operational criterion working to establish rank within a Nootkan line of descent or extended lineage CDrucker 1951:245). Conse quently, only the eldest son of the eldest son could aspire to the rank of headman, making the number of headman small and fixed. Rank decreased within an extended lineage as did the distance of the individual from the common ancestor. But, as rank within one extended lineage decreased, the opportunities to augment social flexibility through the avenue of kinship increased. Because kinship recognition figured so prominently in the Nootkan perspective, the ambi lateral tendencies in descent reckoning allowed an individual or nuclear family group to choose the most advantageous social situation possible. That is, by affiliating with one extended lineage at one point, and then changing affiliation when the move was advantageous, Nootkan people in the middle of the social continuum could affect their families’ status in an upward direction CDrucker 1951:245). This pattern of ranking accompanied by shift in affiliation based on kinship extended to the resi dential groups within a Nootkan longhouse. Owned by a headman, usually a whaler, longhouses were occupied by several related nuclear families which ranked themselves relative to the longhouse owner. The closest relations

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 40

to the headman, his younger brothers and their families for instance, were bound relatively closely to one longhouse. Families of lower rank were not restricted in choice of residenceJ if a family was dissatisfied with the house headman, or had an opportunity to achieve a better ranking in another longhouse, the family could change residences (Drucker 1951:71). This movement could ultimately affect the headman, who relied on a series of specialized activities and

the cooperation of many laborers in ©Tder effec?..... tively exploit the economic privileges he possessed. "In other words both the chief (or headman) and his tenants knew that the former's effective, performance depended on the assistance of his tenants. If he offended them and treated them ill, they would help him but little, or even move away" (Drucker 1951:454). The loss of these tenants would reduce the number of individuals who could perform activities such as catching and preserving salmon, as well as contributing food and property to potlatches and feasts (Drucker 1951:279). Nootkan social structure also affected subsistence activities, especially male practices associated with specialization of trades and labor. The most prestigious activity, bar none, was whaling (Swan 1869, Waterman 1920, Densmore 1939, Gunther 1942). Restricted

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 41

to a man with the ultimate status via primogeniture in an extended lineage, the whaler also had to be rich with the economic privileges which afforded capital and the ceremonial privileges which afforded status. Only one man in an extended lineage had access to the specialized position at any one time, because success relied on the correct knowledge of and performance of secret whaling rituals in addition to knowledge of the whale hunt itself (Drucker 1951:273). A whaler would only instruct one person, perferably the eldest son, in these ritual matters. Other marine subsistence activities (seal hunting, fishing) were not dependent upon private ritual knowledge, and relied more on the skill of the individual and his diligence to the ceremonial prepa rations for his particular activity (Gunther 1936:115). The pursuit of whales and the reliance on marine exploitation strategies are shared Nootkan cultural traits; consequently, the technology required for these activities also represents cultural similarities among Nootkan peoples. Noted as the best canoe producers and navigators on the Northwest Coast (Waterman 1920:9,; Drucker 1955:73), the Nootkan canoe type is considered the prototype for the area (Waterman 1920:9, Drucker 1955:73). There are seven varieties of canoes within the Nootkan type, ranging from children's canoes for

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 42

play to eight-man whaling canoes to travelling canoes large enough to carry house planks and freight (Waterman 1920:12-14). Numerous sources describe the production of Nootkan canoes (Swan 1896:36-37; Waterman 1920: 16 ff.; Densmore 1939:19-22; Drucker 1951:82-88; MCRC 1979:3-4), and the complex navigational skills required to operate these vessels for food procurement and travel (Waterman 1920:9, 47; MLP curriculum 1981- 1986; Arima 1983:37-38). Nootkan cultures shared ceremonial similarities as well. The #u*kwa*li*, or wolf-ritual, is generally regarded as ’’the most spectacular and important cere monial in Nootkan culture" (Drucker 1951:386). Also referred to as the Shaman's Dance (Drucker 1951), the ceremonial usually occurred during the winter and included the entire population of the village in which the ftlt*kwa*li* was held. Generally sponsored by a single headman in order to mark an event of particular importance, the ceremony was dramatically staged to provide realism for the visit of supernatural wolves sent to kidnap individuals from the village (Drucker 1951, Ernst 1952, Swan 1869). Each person in the village would be "taken" by the wolves at some point, and perhaps at many points, during his lifetime. Each Nootkan group conducted the tfu*kwa'li' with its own unique features. Accounts concentrating on the

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 43

Vancouver Island ceremonies are available in Sapir and Swadesh 01939:134-136; 178-182), Sapir and Swadesh C1955 57-128), Drucker 01951:386-444), and Arima C1983: 152-158). Swan (1869:66-73), Densmore (1939:102-103) and Ernst (1952) discuss the Makah version of the yu’k ^ ’li*, though only the Ernst volume is singularly focused on the ceremony. While each of these sources is academically oriented, another description of the Wolf-Ritual is less scholarly. McCurdy (1961:69-73) presents a description of the ceremony from the perspective of four non-Indian children who stole into the Wolf-Ritual, uninvited, in 1877; the account comes from the diary of one individual involved in this incident. Other Nootkan ceremonial similarities included the c* a'yiq, or healing ceremony (Swan 1869:73-73j Sapir and Swadesh 1939:104-107; Densmore 1939:302-303; Drucker 1951:215-218, 243; Arima 1983:150-159) and the ritual patterns surrounding ceremonies which prepared hunters for the quest (Waterman 1920:38-40; Drucker 1951:169- 170) or enabled ritualists to bring drift whales ashore (Drucker 1951:171, 173;; Arima 1983:23-24). Similari ties extended to the treatment given hunted whales upon reaching the shore, where ritual governed the dis bursement of the whale and placatory treatment to the whale spirits (Swan 1869;21-22J Waterman 1920:45-46J

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 44

Drucker 1951:39, 175ff., 255-256). Other cultural similarities uniting the Nootkans as a cultural group include the shape of the life cycle, relating to birth, puberty, marriage, health and death (Swan 1869, Densmore 1939, Drucker 1955); the type of and framework for potlatches (Drucker 1951, Goodman 1978); and the uses of cedar as the basis for material culture. Now that the relationship between Makah culture and the Nootkan cultural complex have been discussed, the next sections present specific information about Makah culture and history, beginning with a discussion of Makah Pre-Contact Culture.

Makah Pre-Contact Culture Five Makah village existed in pre-contact times: Neah Bay, Biheda, Waatch, Tsoo-yess and Ozette. The first four were located on sites within the 1986 con- tinguous reservation; Ozette*s location is 22 km south of Cape Flattery, Some authors use the geographical distance of Ozette from other Makah villages as evidence supporting a Tribal distinction between Ozette dwellers and Makah people (Riley 1968). The majority regard Ozette as one of five semi-autonomous villages linked by language, kinship and common traditions (Swan 1869, Densmore 1939, Taylor 1974, MCRC 1979,

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 45

Renker and Gunther; in press). The villages were inhabited all year long, and earned the name "winter villages". In spring and summer, Makah speakers moved to summer residences at Archawat, Kyddikabbit, and Tatoosh Island. All of the people of one village did not necessarily go to the same summer site to fish. If a headman owned rights to multiple sites, he could divide the labor resources in his longhouse to accommodate the need for labor groups at different sites. This was especially important during the salmonid season, so that workers could gather the largest surplus possible. This surplus would be dried or smoked for use during the winter, and for trade with other peoples on Vancouver Island, the interior of the Strait of Juan de Fuca, and with Pacific tribes as far south as the Columbia River (Swan 1869:30-31, Taylor 1974:68). All of the resource areas in Makah territory, not simply anadromous resource areas, could be owned by a Makah headman. Areas plentiful in local shellfish or fish could be owned by a headman who would have certain rights as to the disposition of the resource (Drucker 1951:247-257, Swan 1869:54-55), The archeological record at Ozette establishes this fact. Analyses conducted in Wessen (1982) and Huelsbeck (1983) for shellfish and fish remains respectively, indicate that

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 46

longhouses exploited local fish and shellfish resources to different degrees relative to ranked positions within and between the houses. Rank could be determined based on the evidence that variations in the distribution of faunal remains within houses suggest a degree of specialization in tool manufacture and in the exploitation of particular classes of animals. Distributional variations also suggest intra-house and inter-house variation in status (Huelsbeck 19831 141). Ozette data indicated that sea mammals comprised the bulk of the pre-contact Makah diet. Sea mammal remains constituted 95% of the bones found at Ozette CHuelsbeck 1983!23); of these, 94% are from the northern fur seal (Callorhinum ursinus) (Huelsbeck 1983123) and 6% are from a variety of whales (Huelsbeck 1983a*12). These data have added new information in the recon struction of the pre-contact diet of Nootkans in general in Drucker (1951) asserts that the Makah diet was not dependent on either sea mammals in general (36) or on whales in particular (49) . Analysis of the faunal remains of sea mammals and whales at Ozette suggests that the relative amount of food represented by sea mammals other than whale constituted 13.5% of the Makah pre-contact diet at Ozette, while 73.4% of the pre- contact Makah diet at Ozette was composed of whale

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 47

blubber. If whale meat were added to the blubber per centage, the figure would rise to 84.6% (Heulsbeck 1983a: 4 3 ) It remains to be determined if the Makah figures can be applied to Nootkan people on Vancouver Island in relation to pre-contact diet. While both whaling and sealing were economically important vis a vis the Makah food supply, whaling was by far the most prestigious pursuit in pre-contact times and in the historical period before whaling was outlawed in the early 1900s. The pursuit was restricted to the few men who could physically and mentally withstand the rigors of intensive ritualized training and possessed the hereditary access to the position and its ritual know ledge CWaterman 1920:38-40, Gunther 1942, Drucker 1951: 169-170). The whaler owned the canoe and the whaling equipment^ and acted as the sole harpooner in the whaling canoe. CSwan 1869:19-22; Waterman 1920:38-44; Drucker 1951:51-55i MCRC 1979:2-3). He also owned important ceremonial privileges acquired through his hereditary status and his ability to interact with the supernatural to assure a successful hunt. The economic and ceremonial perstige insured the whaler*s place as the headman of his longhouse and as the primary person of importance in the whale hunt.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 48

Whale hunters and their crews travelled miles out to sea in search of grey whales migrating north along the Pacific coast in March and April. Other whales were hunted, but grey whale remains dominante the whale bones assemblage from Ozette (Huelsbeck 1983a). These mammals provided a highly prized oil used to enhance the flavor of dried fish and plant products; Makahs also traded this oil north and south on the Pacific coast. Intensive accounts of the whaling process are available in (Swan 1869, Drucker 1951, MCRC 1979 and Waterman 1920). The Ozette excavation clearly demonstrated the importance of fur sealing in the pre-contact Makah economy. Uniquely Makah, the importance of fur sealing was probably related to the deep waters off Umatilla Reef, some three miles out to sea from Ozette. This area is the only place from Alaska to California where the northern fur seal (Callorhinum ursinus) comes close to land; because fur seals only come onto land in their rookeries, the area off Ozette became the premier sealing station on the coast (Kirk 1974, MCRC 1979). In fact, Makahs are the only Nootkans known to have exploited this resource in aborignial times; Nootkans on Vancouver Island did not hunt fur seals aboriginally (Drucker 1951!46). Unlike whaling, Makah fur sealing was not restricted to men who possessed inherited privileges. Ethnographic

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 49

evidence from the first half of this century indicated that success as a fur sealer depended on knowledge of the game and personal stamina in ritual preparation, leaving the occupation available to any man who possessed these characteristics (Gunther 1936:115). This fact, combined with the relative abundance of the resource in pre-contact times allowed an essentially seasonal pattern to be fundamental to the Makah economy. Because Makah people favored the taste of fur seal to that of other seals in the area, other species were not exploited to their respective limits (MCRC 1979:4). Another important and seasonal food resource was fish of various kinds! halibut, salmon, cod, herring, lingcod, rockfish, greenling and perch (Huelsbeck 1983) . Huelsbeck (1983) discusses the annual vs. seasonal availability of the fish resource, and divided "ethno- graphically important seasonal resources-- salmon, halibut, cod and herring" (109) from the generally available lingcod, rockfish, greenling and perch (108). While the former group is procured primarily in spring, summer and fall, halibut present an exception. Hailbut were accessible to Makah fisherman in the winter months when waters were calm (MCRC 197916), though mature halibut were available only from April to August (Huelsbeck 1983:108).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 50

Like some fish, intertidal resources provided the pre-contact Makah diet with a dependable and varied source of food CMCRC 1979, Wessen 1982). While fishing required weather suitable for canoes and water travel, intertidal resources were available at low tides on a daily basis. Clams, octopus, mussles, and other shelled animals such as chitons, barnacles, and limpets (Swan 1869, MCRC 1979, Wessen 1982) abound in Makah intertidal territory. Studies also document the importance of the raw materials obtained from shellfish as cutting tools, harpoon heads, eating implements and articles of adornment (Swan 1869, Densmore 1939, MCRC 1979, Wessen 1982), though wood-base artifacts clearly dominate the Ozette Collection. Despite the Makah preference for water-based foods and activities CSwan 1869, Densmore 1939, Taylor 1974, Gunther 1936, MCRC 1979), land mammals and birds were hunted to provide limited food and an additional source of raw materials, Gunther (1936) substantiates the Makah preference by noting that Makah hunters drove deer and elk into rivers where they could be clubbed with greater ease. The archeological record also supports ethnographic accounts which emphasize preferences in sea foods despite the accessibility of land animals in the area. Land animal bones are poorly represented in the faunal assemblage from Ozette

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 51

(Hhelsbeck 1983, Wessen 1982). While hunting and fishing activities were restricted to men in pre-contact times, women played an important part in the subsistence economy. Women's activities predominantly involved the gathering portion of subsis tence, and processing the animals killed by men; an exception was the ritualized butchering of whales by men CSwan 1869:21-22, Waterman 1920:45-46, MCRC 1979:2-3). Women processed fish and mammals by cleaning, then smoking or drying the meats. Numerous sources discuss the female activity of plant gathering (Gunther 1945, Densmore 1939, Gill 1983). Men might gather plants necessary for a ritualized activity or in recipes associated with private knowledge, but women conducted the major subsistence activity of gathering plants. Gathered plant foods included a variety of berries, including salmonberries, red and blue huckleberries, blueberries, cranberries, and wild strawberries. Other plant foods included roots, shoots and grasses. Preserving, processing and preparation technologies are thoroughly discussed in Gill (1983) . Plants played a critical part in other areas of Makah life. Dissertations generated by Ozette data review and discuss the emphasis on woods, especially red cedar, in Makah pre-contact material culture; other woods such as yew alder, spruce and hemlock were chosen

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 52

for characteristics different from cedar CFriedman 1975, Gleeson 1980, Gill 1983). The versatility of the cellu lar structure of cedar allowed the material to split straight enough to form house planks or to be beaten soft enough to make diapers and cedar clothing. Friedman (1975) found that the majority of Ozette artifacts were made from red cedar, though other woods were used when the artifact required properties not found in cedar. Yew wood, a dense, heavy wood, was the wood of choice for clubs, whaling harpoon shafts and bark beaters CFriedman 1975, MCRC 1979). The analysis of Ozette artifacts indicates the complexity of Makah woodworking technology. Specific tools and techniques were used for particular types of wood, which was not randomly chosen for a task. Gleeson (1980) indicated that the needs of the project dictated the choice of wood and the techniques used in preparation of the wooden article. Makah basketry also used the versatility of cedar to best advantage. Makah women produced numerous types of baskets; large, open weave frameworks allowed water to escape from intertidal products during gathering activities, while tightly woven baskets could hold water and function as cooking containers (MCRC 1979). While Makah women used cedar most often in basketry, other materials such as cherry bark, cedar root, cattail,

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 53

spruce root, and tule from Ozette were used with the same discretion male carvers showed when choosing their materials (Gill 1983, MCRC 1979). No Ceramic technology developed in pre-contact Makah times; the need for a non-wood based technology like ceramics did not exist given the abundant variety of woods available in Makah forests. However, Makah people were not uncomfortable with the idea of innovation, for Ozette artifacts demonstrate that Makahs used iron and steel carried by the Japanese current when these materials washed upon Makah beaches (MCRC 1979). This description of the pre-contact Makah culture, combined with the information about Nootkan peoples in the first section of the Chapter, provides important informa tion necessary to the understanding of the language. The reliance on the sea, the material culture based on wood and plant materials, and the Nootkan preoccupation with status and inherited rights all are factors which need consideration when evaluating the Makah language. Because the Ozette material removes a great degree of subjectivity from the accounts of Makah life at the time of contact and in the early years of the reservation, Ozette data supercede descriptions of the material cul ture from other sources for this dissertation. The next section briefly discusses the other sources which are generally used in the ethnographic representation of Makah people.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 54

The Ethnographic Sources The earliest written accounts which relate infor mation about Makah people and culture come from the journals of explorers and traders who ventured into the Pacific waters around Cape Flattery and the Olympic Peninsula at the end of the eighteenth century. The journal of John Meares, who sailed around Cape Flattery in 1788, and the journal of Manuel Quimper, who officially "claimed" Neah Bay for the Spanish in 1790, are primary sources used in more comprehensive studies of European contact with the aboriginal people of the Northwest Coast such as Gunther (1972) and Pethick (1980). Additional primary sources, in translation, are collected in Spanish Explorations in the Strait of Juan de Fuca (Wagner 1933). Very little contact, save the casual contact of a few sailing vessels, was to occur between Makah people and non-Indians for the sixty years following the Spanish occupation of Neah Bay (Gunther 1960:17). A spate of ethnographic accounts of Makah life accompanied the occupation of the then Washington Territory by settlers and traders in the middle of the nineteenth century (Hancock 1852), and by individuals connected with the United States' government during this same period (Gibbs 1855, 1877; Swan 1869).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 55

Of these fev documents, the observations of Swan C1859-1866) comprise the first ethnography of Makah people, published in 187Q. As the first schoolteacher appointed to the newly created reservation, Swan's- observations of Makah people at a time of transition are an interesting mixture of detailed material culture descriptions and incomplete social commentary biased by his eastern background and his job as an agent of assimilation for the "civilized" world (Renker and Gunther: in press). Recent studies of Swan as a person ality (Doig 1980, Cole 1985) provide biographical insight of great importance when interpreting the social observations made by Swan in the 1869 publication. Later ethnographic works on Makah people and cul ture were slow to appear. Studies of particular aspects of Makah material culture or life were not published until the first half of the twentieth century, when Waterman (1920) discussed whaling equipment, Densmore (1939) wrote about Makah songs and potlatches, and Gunther (1936, 1945) investigated Makah ethnobiology. Publication of Makah ethnographic studies increased after the mid-point of the century. Researchers published investigative conclusions about Makah accul turation (Miller 1952, Colson 1953), social patterns (Wike 1958, Gunther 1960, Riley 1968) ceremonial and potlatch activities (Ernst 1952, Goodman 1978), Makah

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 56

stories and storytelling (Miller 1952) and the history of the Neah Bay Agency (Gillis 1974) . A comprehensive synthesis of historical and ethnographic Makah culture appeared in Taylor (1974), which originally was an anthropological research document commissioned by the attorneys of the Makah Tribe. Unpublished studies of note include two case studies of Makah acculturation by Whitner. His 1977 study discussed the effects of the imposed school system on the reservation; another study (1981), traced the effects of commercial pelagic seal ing on the economy and life of Makah people. Ethnographic information is also found in the archeological studies from the Ozette village excavation. In exploring the complex links between procurement, processing and utilization of resources in pre-contact Makah economy, Ozette studies describe pre-contact material culture and economy, as well as the relationship between economic resources and pre-contact social structure (Gleeson, et al. 1981). Of particular importance are Friedman on wood use (1975), Mauger on house structure (1978), Gleeson on woodworking (1980),

Wessen on shellfish (1 9 8 2 ), Huelsbeck on fish and mammal use (1983) and on the utilization of whales (1983a), and Gill on plants in pre-contact economy (1983). Daugherty and Friedman (1976) extend the archeological discussion into the realm of Makah art.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 57

A vealth of ethnographic information is also available in the curriculum generated by the Makah Language Program C1981-1986), described in depth in the next chapter.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 3

THE MAKAH LANGUAGE

Linguistic Classification Many studies pay considerable attention to the familial classification of American Indian languages (Gallatin 1836, Powell 1891, Boas 1911, Sapir 1929, Hoijer 1946, Vogelin and Vogelin 1965, Trager 1972, Campbell and Mithun 1979, Leap 1980). Based on similar ities in phonetic, lexical and syntactic systems, language families are accorded differing degrees of autonomy depending on the specific classification. The Wakashan language family, to which Makah belongs, is either regarded as a language isolate fa family with no external relationships) (Boas 1911, Vogelin and Vogelin 1965, Campbell and Mithun 1979) or a family externally related to other language families in the geographic area (Sapir 1929, Swadesh 1953, 1962, Trager 1972) Salish, Chimakuan and Eskimo-Aleut are families with proposed relationships to the Waksahan family. The clarification of the external relationships of Wakashan are subject to further descriptive and compara tive work with languages in the geographical area, while 58

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 59

the internal relations of Wakashan have not been chal lenged since its discovery by Boas in 1889 (Jacobsen 1976b:768). Composed of a northern and a southern branch (also known as the Kwakiutlic and Nootkan branches, respectively), the Wakashan family contains the following languages: Kwakwala (Kwakiutl), Oowekyala (Rivers Inlet), Heiltsuk (Bella Bella), and Kitamat (Haisla) in the northern branch (Lincoln and Rath 1986) and Westcoast (Nootkan), Nitinaht and Makah in the southern branch (Jacobsen 1979b:772). Swadesh (1953:26, 41) theorizes that the northern and southern Wakashan branches diverged some twenty nine centuries ago. Makah appeared as a distinct language within the southern branch approximate ly one thousand years ago (Jabobsen 1979b:776). Because of the distant relationship between the northern and southern branches of the Wakashan family (Lincoln and Rath 1986:2), the discussion of comparative studies related to Makah is restricted to studies con cerning the southern branch of the family. These studies primarily focus on lexical and phonetic correspondences between Westcoast, Nitinaht and Makah (Haas 1969, 1972; Jacobsen 1968, 1969); Haas (1969) and Jacobsen (1968, 1969) examine these language correspondences with the specific intent of relating the languages spoken today to their Proto-Nootkan precursor.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 60

Two examples of these investigations follows. Jacobsen (1968) traces the shift of */m/, *M/, */n/ and */A/ in Proto-Nootkan to /b/ and /d/, respectively, in Makah. While /m/ and /n/ and their glottalized forms are retained in the contemporary phonemic inventory of Westcoast, the resonants and their glottalic forms have become /b/ and /d/ in Makah (3). The following data set illustrates this correspondence.

CHART 1.--Makah Westcoast Correspondence

WESTCOAST MAKAH ENGLISH (5) ma’as ba?as house C6) muwat bukawC deer C7) mu'smu's bu'sbu's cow C8) nfatik baCik bite (9) na?a* da?a* hear, understand 0-0) naSuk daSuk strong ai) rii'ya di’ya Neah Bay CL2) rfaSSi* daSSik become strong

In the first column, Westcoast words exhibit an /m/ or /nf/ that corresponds to /b/ in Makah. Similarly, /n/ or /rf/ in Westcoast corresponds to /d/ in Makah. These correspondences, along with the rare occurrance of /m/ and /n/ in Makah loanwords, led Jacobsen to postulate the existence of glottalized and non-glottalized resonants in the Proto-Nootkan phonemic inventory and to

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 61

argue for the replacement of these phonemes in Makah. Jacobsen also (1 0 6 9 ) describes the presence of the pharyngeals /?/ and /£/ in Westcoast in relation to the presence of /?/ in Nitinaht and the absence of both pharyngeals in Makah. The corpus illustrates the correspondence between /?/ in Westcoast and Nitinaht with /

CHART 2.--Nootkan Pharyngeal Correspondence WESTCOAST NITINAHT MAKAH EN6LISH (13) ?ataxw ?ataxw cfa taw beaver (14) ?a?uk ?a?uk (fa?uk lake (15) ?uyi ?u *y cfu*y medicine (16) wa’it we*?it wa'cfit frog (17) mi*?at bi*?at bi(fa*t sockeye salmon (18) ?a‘nus ?a*lis c^a'lis crane

(19) ’ifia'k ?iCa*k (f^iCa'k rotten (20) fcupta Xu'bitfad Xu'bitfadi' swan C21) ha?u Xu?u* Xu’u* trade, exchange (22) (ft^ati* c?iXati* cfiXati* arrow (23) lapfc- dapX- iapX- fly (action)

The literature which discusses the phonetic relationships between the three southern Wakashan languages also presents phonological and morphological data about the respective languages. (It is of interest

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 62

to note that little data about syntax are available in these comparative studies.) For the bulk of the language-specific information which describes the individual southern Wakashan languages, there are publications that concentrate on one language as opposed to a comparative framework. Sapir first worked with Westcoast in the early twentieth century (1911* 1913, 1915), and went on to present detailed studies of the language with Swadesh (1939, 1955): Swadesh also published Westcoast language studies (1939). Generally, these publications discussed aspects of Westcoast phonology, morphology and syntax, and the joint Sapir-Swadesh publications presented numerous Westcoast texts. Nitinaht studies began in 1931 when Haas and Swadesh undertook the task of describing the language (Turner 1983:9), and together they published a Nininaht text with translation in 1932. Since that time, other Nitinaht texts and analyses were published, including the work of Nitinaht speakers trained in linguistics (Touchie 1977, Thomas 1977, Thomas and Hess 1978, Thomas and Carlson 1979). These studies presented information about the phonological, morphological and syntactic components of Nitinaht. Makah research is not as plentiful as the research addressing issues in Westcoast and Nitinaht. While Makah

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 63

data found its way into some of the Westcoast and Nitinaht studies, particularly the historical ones (Hass 1969, 1972), these are relegated to the phonological and morphological realm. The next section discusses the literature which focuses on the Makah language or presents Makah language data within the context of another study. This discussion not only establishes the relative scarcity of Makah language data before 1968, but provides a foundation for the description of the language in Chapter 3.

The Makah Language in the Literature The first written Makah vocabulary was presented in Swan C1869:93-105). This vocabulary was the only com pilation of Makah words for decades, aside from the inclusion of word lists in Waterman (1920), Densmore (1939), Gunther (1936, 1945) and other ethnographic sources. The Swan list is different from the lists in the ethnographic documents; the 1869 vocabulary includes words from a variety of topical classifications; the ethnographic documents contain word lists which address one topic--the topic of the ethnography. It was not until 1968 that Makah linguistic research began to regularly appear in the literature. The initial studies (Jacobsen 1968, 1969), discussed previously noted phonological and lexical correspondences between Makah

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 64

and other southern Wakashan languages and applied these findings to the historical reconstruction of the proto language of the southern branch of the Wakashan family. For the next ten years, Jacobsen would be the exclusive producer of articles and papers which dealt specifically with the Makah language. The major thrust of his writing fell into two groups: 1) phonology and morphophonemics (1969a, 1971), and 2) morphological patterns (1973, 1978). One paper utilized syntactic data to discuss form classes in Makah in 1976, but was not published until 1979. In other words, in Jacobsen's work as elsewhere Makah syntactic investigation is not as well represented as is Makah phonology, morphology and historical relationships. The first published study concerned with Makah syntax appeared in 1979, when the Makah Language Program commissioned William Jacobsen to draft a series of beginner lessons which would enable non-Makah speakers to learn basic sentence constructions in the language. The final version of the text, First Lessons in Makah (Jacobsen 1979), proved to be too technical for general educational uses, but provided the first description of simple Makah sentences. Additional material prepared by the MLP has not addressed syntax either. Since 1979, the MLP has published four books designed to respective ly teach the Makah alphabet (Makah Language Program

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 65

1979), Makah words in conjunction with stories and legends (Makah language Program 1982a, 1983) and Makah numbers (Makah language Program 1982), Additional curricular materials, videotapes, worksheets, visual aids, flash cards, and the like have been developed to teach Makah in the public school and Headstart program on the reservation, but none of these are syntactic studies, even though they draw on syntactic information. Three additional publications have appeared in the literature since 1983J syntax does not appear in these studies, The first two of these articles discuss Makah ethnobotanical terminology (Gill and Renker 1985) and Makah ethnozoological terminology (Renker and Gill 1985), respectively, Renker and Gill (1986), the third article, presents ideas concerning the most appropriate ways to facilitate the best translations from Makah to English, A H of the academic and educational research con ducted on the Makah language has broadened knowledge about the languageJ however, the brief review of the literature shews the lack of Makah syntactic studies. Other aspects of the language are reasonably well-represented, as compared with sentence-*!eyel research. The next section offers a review of those facts and lays a foundation for the discussion of the AUX phenomena which follows in Chapter 4,

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 66

General Phonetics There are 51 phonetic segments in Makah, 34 con sonants, 10 vowels, and 7 vowel-semivowel combinations (see Appendix 1) (Gill and Renker 1987) . Makah lacks voiced fricative, phonetic and phonemic /r/, and has abundant variations of /k/ and /q/. Appendix 1 shows the fact that certain consonants can be labialized, glottalized or voiced in morphophonemic operations. /m/ and /n/ are rarely attested in any environmentJ their respective presences indicate the appearance of a loan word in Makah. Appendix 1 provides a complete listing of Makah phonetic units as represented by the practical ortho graphy for Makah used on the reservation. The original prototype for the orthography was developed by Jacobsen and described in Jacobsen (1979) J a few modifications have been made by the Makah Language Program since then. The orthography used in this dissertation is the system used by the Makah Language Program, so a few words about the system are in order. The mid-line dot after a vowel indicates a ' difference in the quality (nature of the sound) and the quantity (length of a sound) of a vowel. A above a letter indicates glottalization, and ^ after a consonant indicates labialization. The Makah Language Program version of the Makah orthography transcribes ^ wherever

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 67

the sound appears. The Jacobsen version eliminates the ^ in certain phonetic environments, notably when the consonant k, q, x and their variants appear after /u/ or /u’/» because the labialization is assumed in these contexts (Jacobsen 1969a). The status of this latter feature in the 1980's Makah speech community is not as predictable as Jacobsen suggests, for k, q, x and their variants are not always labialized after /u/ or /u’/• Consequently, the MLP version records * whenever the feature is heard*, this eliminated the confusion which resulted from language learners mis-reading the ortho graphy when they did not know the proposed labialization rule, as well as the confusion which resulted from add ing a feature which was not always present in Makah speech. Makah does not possess tonal accents, though other Wakashan languages to the north exhibit this character istic (Kortland 1975, Rath 1974, Lincoln and Rath 1986). Suprasegmental patterns in Makah are not as complex as other patterns in the language. Stress is dependent on vowel length (signified by the mid-line dot): primary stress falls on the first syllable if the initial vowel is long, and on the second syllable if the vowel is short. In the following corpus, the examples are divided syllabically, not morphemically. Note that the dashes indicate a syllabic division.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 68

The corpus is presented in the following manner the first column in each set presents the dictionary citation form of the Makah word; the second column illustrates the syllabic division of the words in the first column and the third column presents the English equivalent.

CHART 3.--Stress in Makah SET A - First Syllable Stress C 24) su* suk sd*-suk swim C25) hu’iuk hd’iuk dance (26) Su’yu’i Sd*-yu*4 halibut (27) ta*la* td*-la’ money (28) tfi*kpa*4 £i*S-pa-i six

SET B - Second Syllabi e Stress

(29) ?ada*k ?a-da*k• fire (30) Siqu'i 5i-qu*4 frown (31) tiiu’p ti-4u*p octopus (32) ba?as ba-^ds house (33) Cabas ta-bas sweet

Syllable Structure and Phonotactics Makah, like other Southern Wakashan languages, exhibits a distinctive syllable pattern:„an_initial consonant followed by a single vowel with an optional two or three consonants following (Hass 1972, Jacobsen

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 69

1969, Renker and Gill 1985). This means that the sylla bic structure of the Makah word has to meet a series of rigid restraints: 1. No syllable may begin with a vowel. 2. No vowel clusters are attested anywhere in the language. The practical orthography represents this vowel separation with a glottal stop (?) between vowels (ba?as, wa’ad^a, ki*?ak, we?it). 3. No consonant clusters appear at the beginning of a syllable, while they are attested in other environ ments. (One exception to this rule exists, ^a’tpla's, a baby word for ’cute' syllabically divides thusly ?S*tt-pla*s). 4. No contiguous glottal stops are permitted. Under these restraints, the resultant surface exhibit a CV- (consonant-vowel) or CVC- syllabic pattern common to other southern Wakashan languages (Haas 1969, 1972; Jacobsen 1969, 1979; Renker and Gill 1985). The syllabic pattern allows for the addition of a consonant following the CVC-stem, producing a CVCC-construction. Makah makes use of this option in a special way; the language exhibits the use of velar'or post-velar con sonants as stem extenders which intensify the semantic intent of a stem (Haas 1972:83). Renker and Gill (1985: 326) list the following examples for Makah; the first column presents stems and the second, stems with the stem extenders.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 70

(34) Arit- 'spread out' (34a) fcitq- ' explode* (35) but- 'cut' (35a) butq- 'amputate' (36) pit- 'fit together’' (36a) pitq- 'jam together’ w /-> o t - r sit- 'split' (37a) sitX- 'tear' 04 0 0

W cfat- 'heal' (38a) cfatX- 'shrink, shrivel' (39) iit- 'twitch* (39a) iitq- 'twitch involving pain*

Oth er consonant clusters, contiguous consonants, are also possible at the end of syllables (36-40) and at the juncture of syllables (41-45). In the case of clusters occurring at the juncture of syllables, it is important to note that syllabic boundaries do not always coincide with morphemic bounda ries. This proves important when the isolation of constituent morphemes is at issue.

CHART 4.--Makah Consonant Clusters

(40) &rtiiba£§ 'loon' (45) tfa&ktidik 'blackbird' (41) wi?ibaqft 'angry (male)' (46) ?u5u*yak$?alc 'thank you' (42) hupa’ttakt 'island' (47) wiCtqi* 'bald' (43) wjiaq* 'not want to' (48) ha*ha*dq&si*s 'place name* (44) xad?awx$E 'girl' (49) c^ada’lddukub 'finger'

The following general parameters govern consonant clusters: 1. Glottal stops or glottalized consonants rarely begin clusters. 2, Resonants rarely begin clusters.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 71

3. Voiced stops rarely begin clusters. 4. Clusters of /y/ and other consonants are rare. Vocalic phonotactics also contribute to t&e-surface

structure of Makah syllables. Jacobsen (1 9 7 9 ) discusses vocalic phonotactics in terms of two basic rules! 1. No short vowel can appear at the end of a word. Vowels in this environment are dropped. 2. Long vowels appearing in word final environments shorten. Examples (46-48) illustrate the occurrence of the first rule.

CHART 5.--Phonotactics - Underlying Vowel with ADDITIONAL UNDERLYING SPOKEN FORM SUFFIXES VOWEL (5 ) ha?ub ’food' ha?ubasis a (51) kawad 'killer kawadibadaX i whale' (52) kibta’l 'horse' kibta’lasis a

Illustrating the second rule is more difficult, as data in the following chart suggest. Column one contains an underlying form with English equivalent. Column two contains the form in isolation from Jacobsen (1979) and column three contains the word form as currently attested in the Makah speech community according to Makah Language Program data.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 72

CHART 6 .--Phonotactics - Dsolated Forms FORM IN FORM IN UNDERLYING ISOLATION ISOLATION FORM CJacobsen 1979) (MLP 1985)

(53) ta'la* 'money' ta'la ta'la* (54) twi£i’ya* ’land' kwt£i*ya kwiti*ya* (55) lu*lapi * 'hand' lu*lapi lu*lapi* (56) SiCi’i* 'tooth' £i£i7i Citi’i* (57) ' ear' jfijfi?i j/ipfi’i* (58) qali7i* ' eye' qali?i qali?i*

According to the comparison, Jacobsen's second rule is no longer in operation in the Makah community; conse quently, no further reference is made to it in the present analysis. Other sound changes found in Makah (additions, deletions and replacements) can be traced to morphophonemic processes in Makah grammar. Morphophonernes are "a class of phonemes which occur within a particular set of grammatical environments" (Crystal 1980:233), and therefore, "(are) tightly related to the morphological, or word-forming level of language" (Bolinger 1975:77). While many languages have morphophonemic processes, the southern Wakashan languages have long been noted for their morphophonemic complexity; this complexity has been described in the literature in a variety of ways. Boas

(1911) described the morphophonemic system for Kwakiutl as

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 73

a system of hardening and softening suffixes; Sapir and Swadesh (1939) described hardening, softening and glottalizing suffixes for Westcoast. In his earlier writings about Makah, Jacobsen (1969, 1969a, 1971, 1973) followed this tradition and described suffixes in terms of their phonetic tendencies. By 1979, he modified the presentation, and set up two morphophonemes represented symbolically by ° and _V respec tively. Each symbol was associated with a set of phonetic processes affecting the immediately preceding phonetic element. Appendix 2 lists the changes associated with these two morphophonemes (Jacobsen 1979). As the data in the appendix show, the more complex morphophoneme of the two is °. (In fact, it may be the most complex of all Makah morphophonemes.) When a ° is placed before a suffix, as in /-°iq/, the symbol indicates a change in the sound preceding the symbol. For example, in the word huktu*p ’bird', the /p/ becomes /b/ when combined with the morphophonemically influenced suffix /-°iq/. \ the second morphophoneme, also indicates a variety of changes in the sound preceding the symbol. The phonetic interaction here shares some behaviors with the morphophoneme mentioned previously, but also demon strates some singular effects. For example, using huktu*p, /p/ becomes /jf/ as opposed to the /p/ to /b/

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 74

change associated with °-, while /c/ becomes /

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 75

CHART 7.--Past Tense Morphophonemic Changes MORPHOPHONEMIC VOCALIC CONSONANTAL SUFFIX ENVIRONMENT ENVIRONMENT

C59) -Bit- tfupa’bits tfupSikits ’past tense' 'I pointed* 'I pointed' C60) -Qa*k- kupa*qa*ks tfup5i*a*ks 'confirmational 'Did I point* 'Did I point' interrogative'

Notice that the B and Q remain when the stem ends in a vowel (lfapa*- 'point continuously*) and disappear when the stem ends in a consonant (kuplSi*- 'point suddenly or briefly'). These morphophonemic changes support the preferred CV- and CVC- patterns discussed earlier in this section. The last series of morphophonemic processes applies to vowels, and also supports the CV- and CVC- syllabic structure in Makah. These morphophonemic rules govern vowel insertion, loss and assimilation, and are described in Jacobsen (1971) . A brief example to summarize the facts is cited here. Consider the following examples with the stem kutX- 'drum* (61) kutxu'yak 'drum* (lit. instrument for drumming) (62) kutXkutX 'drum' (REPETITIVE) In the first example, a copy of the stem vowel (u) plus length (u*) is inserted between the stem kutx- and the instrumental suffix-yak, because the initial consonant

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 76

of yak is a glottalic consonant. In contrast, this vowel insertion rule does not apply when a non-glottalic consonant is involved. kutkkutX shows no vowel inser tion because the initial consonant /k/ is a non- glottalic consonant. Unlike the other cases, the vocalic morphophonemes are predicted by the membership of the stem-initial or suffix-initial consonant in any of three consonantal groups I glottalic, non-glottalic and labialized con sonants. Therefore, these morphophonemes are not indicated orthographically. These vocalic processes, consonantal processes and syllabic constraints are critical to a morphological appreciation of Makah, because the rules and restric tions which apply on the phonetic level of the language do not coincide with Makah the rules governing morphology. That is, the sound changes necessary for the production of correct Makah words often make it difficult to dissect a Makah word into constituent morphemes. Unless an investigator is acquainted with the phonological detail of the language, an analysis might mistake a surface level vowel or consonant as a discriminator of a morpheme rather than a result of epenthesis or assimilation. In a language where key grammatical markers may consist of one or more sound

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 77

segments, such an evaluation could be extremely serious.

Makah Words, Structures and Syntax The literature concerning the Makah language says little about the morphological and syntactic structures of this language. Consequently, much of the disserta tion focuses on these issues, making a summary of those findings inappropriate at this time. For these reasons, this section provides a general introduction to the basic facts about Makah words and sentences, and draws on examples from the literature, as available, and on materials collected during the author*s fieldwork, to illustrate each of these claims. Each Makah word consists of a stem and a string of morphemes which can contain one or more segments. Makah utilizes suffixation to the virtual exclusion of pre fixation as the primary morphological process; Makah stems are safely defined as the primary or initial morpheme in a sequence of morphemes. The following set of data illustrates this definition by using the stem #i%- ’red* to show the amount of combinative flexibility a primary morpheme has within the Makah framework which allows for expansion via strings of suffixed morphemes, C63) tfiki’b ’woodpecker* Cliterally, red thing) tfiX - i* - ba red -epenthetic - thing

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 78

(64) tfikuk ’red’ (the color, literally, the state of red) tfiX - uk red - durative C65) tfiXi/’iit ’red nose’ JfiX - i* - ’iita red -epenthetic - nose locative (66) tfiXi*?aq*bap ’Pacific yew' (literally, red inside plant) tfiX - i* - *aq* - bap red - epenthetic - inside locative - plant suffix (67) tfitfi’Xadii ’carrot* (literally, red along the length)

Reduplication, shown in (67) and (68), is another fundamental morphological process, and figures prominent ly in the formation of repetitive and iterative aspects (Jacobsen 1971) plurals, neologisms (Jacobsen 1980) and a type of word termed "looks like" words by Makah speakers. This construction is characterized by the pre sence of the suffix -kuK/ -ku#w , literally meaning "looks like", (69-73) provide examples of this

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 79

construction. C$9) c?ic?iyupl&itf Spaghetti* Cliterally, looks like intestines J cfiyup- * intestines') C7Q) qaqavaSJ?u>fw ’raspberry* Cliterally, looks like salmonberries; qawaS- 'salmonberry') C7X) di

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 80

a person, an object or an event. Since additional information about the morphological system in Makah is discussed in the later analytic chapters, the chapter now procedes to a presentation about Makah sentences. Makah sentences can be structurally grouped into two categories (Renker 1985a):

1. Group 1 Sentences CGIS) - sentences with only one sentence element. (74) babuyatfe*?iss "I will work (at some point).” (75) cfidi'ktfeis "I am imitating a dog.” (76) tfi’?ak?ui "They walked (somewhere)." (77) tfisuw "It is white."

2. Group 2 Sentences (G2S) - sentences with numerous sentence elements, each bounded by open junctures, (78) babuyaw ia’Xuwiq "The man works." (79) ?ut?eis cfidi’liq "I'm imitating a dog." (80) XaXad?ais jftisak "I am very tired." (81) /isuw cfidi’liq "The dog is white."

Makah is generally classified along with languages which exhibit the sentential pattern VSO (verb-subject- object) (Jacobsen 1979a). In these instances, VSO refers to sentences which generally have translations placing English verbs in initial position. The data in (74-81) show that this assignation is confusing for Makah, which does not necessarily place a translational verb in first position in a sentence. In (67.) for example,

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 81

the Makah word cfidi**-, which corresponds to the English noun, dog, begins the Makah sentence. XaXad?a*-, corresponding to the English adverb, ’very,* is in initial position in (80), while the word corresponding to the English adjective, white, begins the sentence in (81). In short, words which correspond to English verbs, nouns, adjectives and adverbs can all begin Makah sentences. This situation poses a problem in writing a grammar of Makah. Formulating linguistic rules with relation to English translations in classifications such as VSO. Relying on English translational terms in Makah description produces a heterogeneous representation which focuses on how Makah deviates from an English standard. To represent Makah in its own terms, analytic tools and processes must recognize the integrity of Makah structural components and patterns, and use Makah units in describing grammatical expressions. Unless an English semantic bias is removed from Makah language description, the rules which govern Makah sentence Structure are obscured under an English cloak. This dissertation ultimately addresses this issue, and uses AUX to show the highly regular nature of Makah syntax without reference to English foTm classes like noun and verb, or to English structure like a VSO translation. Because AUX figures so prominently in the

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 82

analysis in this study, a review of the literature on the category follows in the next chapter.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 4

THE AUX CONTROVERSY: A LITERATURE REVIEW

Background: Literature on Kernel Sentence Structure Like the literature addressing the formal proper ties of linguistic universals, the literature addressing the formalized conception of the most basic components of sentence is relatively recent addition to linguistic science. Nevertheless, the issue explored in these pages--the nature of the phenomenon called a kernel sentence Cor basic sentence, simple sentence and the like)— has become fundamental to modern linguistics; regardless of theoretical orientation, some form of the kernel sentence is always associated with a discussion of basic sentence properties and their representation in grammatical structure. Two different representations of this kernel structure are currently being employed in the literature. The first, called Phrase Structure analysis CPS analysis), posits a kernel of three components, Noun Phrase CNP), Auxilliary CAUX), and Verb Phrase CVP)• The second, called the Main Verb analysis CMV analysis), posits only two kernel components, NP and VP. Conse quently, the place of AUX as a component of the kernel 83

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 84

sentence has recently been debated in the literature. Historically, the PS analysis follows the formulaic representation of the sentence kernel proposed in Chomsky (1957), where the following PS rules appeared: (82) S NP AUX VP (where S=sentence, NP=noun phrase, VP=verb phrase) C83) NP (Det)N (84) VP V (NP) (85) AUX Tense (Modal (have + en) (be + ing) This formulaic representation signifies that the sentence is composed of three constituent, or primary categories. There are no simpler categories than NP, AUX, and VP; the omission of any of these categories in the PS representation produced an inadequate description of the sentence. Chomsky (1965), and elsewhere, refined the generative model proposed in Syntactic Structures (1957), but has not altered the phrase structure rule. fox the sentence. The second theoretical position did not emerge until 1967, when Ross (1967) offered an alternative structure for the kernel. This analysis recognized only two constituent categories, and suggests the following formulaic representation of the sentence: (86) S NP VP (87) NP * (Det) N (88) VP * AUX V (NP)

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. * 85

According to this analysis, AUX is not a sentential con stituent, but a subset of the constituent VP. According ly, MV analysis posits subcategorization rules to relate AUX within VP, that is, formally recognizes AUX to be a subset of VP, rather than a constitutent on an equal level with VP. Interestingly enough, the controversy which surrounds the formulaic representation of the kernal- sentence has not been an issue of interest for most linguists. Many follow the lead established in Chomsky (1957), while others follow Ross (1967). As Lyons states: Under either of the two definitions, the kernal- sentences of a language are intended to consti tute a subset of what would traditionally be described as simple sentences (1977:469).

AUX as a >fProblem,t for Kernal Sentence Analysis Unfortunately, those linguists interested in AUX cannot be as neutral as Lyons suggests. Interestingly enough, the interest in AUX made apparent in the literature stemmed from studies by those concerned with English syntax, thus establishing English as the language at the center of the AUX controversy. The theoretical linguists interested in the issue of AUX constituency continued to investigate the category in English (Emonds 1970, Akmajian and Wasow 1975, Pullum and Wilson 1977),

and paid special attention to the ordering of the

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 86

elements contained in the English AUX. Not all scholars chose to investigate the consti tuency of AUX using English as the language of investigation. In her groundbreaking article "The Category AUX as a Language Universal" (1978a), Steele proposed that the complex nature of the AUX in English obscured salient features of the category which would appear through a cross-linguistic comparison. Consequently, she tested twenty of the world's languages (based on a representative sample of language families) to ascertain if a category corresponding to the English AUX existed for each language. Based on this investi gation, AUX did appear in the languages of the sample, and "certain essential characteristics of the AUX (were) established within this cross-linguistic study" (Steele 1978a:7). Not surprisingly, the publication of this article marks the turning point in the AUX controversy. The typology (which emerged from her analysis of AUX in non-European languages in addition to English) established the boundaries and the composition of AUX across parameters of genetic affiliation, word order preference and ergativity. Now analytic treatments of the subject had a more explicitly stated universal scope, and the characteristics which could now be associated with the category provided linguists with an opportunity

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 87

to investigate AUX without the conceptual baggage which accompanied the investigation of the category in English. Steele (1978a) also proposed another idea which concerned AUX and its constituency. AUX might be easier to recognize in the surface structure of some languages as opposed to others C37), thus necessitating the formal analysis of a wide variety of the world’s languages before a definite conclusion regarding the universal properties of AUX constituency could be reached. To this end, Akmajian, Steele and Wasow (1979) published an analysis of LuiseSo, a Uto-Aztecan language, in an article which also featured a unique analysis of English; both languages were used in order to substantiate the argument which called for the inclusion of AUX on the sentential constituent level. The respective analyses made the additional point that AUX in each language shared certain universal qualities, in addition to language-specific characteristics. A flurry of rejoinders followed Akmajian, Steele

and Wasow (1979). LaPointe (1980) and Pullum (1 9 8 1 ) refuted the findings which supported the inclusion of AUX on the constituent level as opposed to the verbal subset level. While both discussions questioned the Akmajian, Steele and Wasow position regarding the English AUX, only Pullum (1981) also questioned the AUX analysis for Luiseno, and called for an objective method

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 88

of equating AUX from language to language. Pullum's contention was that the AUX elements proposed for each respective language did not constitute a "natural class" (436), and that no framework to establish equivalence was present in the Luiseno-English analysis in Akmajian, Steele and Wasow (1979). To answer this concern, to present additional detailed AUX analyses for previously untested languages and to refute the Pullum (1981) analysis of Luiseno, Steele and a number of colleagues collaborated to publish The Encyclopedia of AUX (1981) . In addition to a detailed investigation of AUX in Luiseno, Demers pre sented an analysis of AUX in Lummi, and Jelinek describes AUX in Egyptian Arabic. On a theoretical level, Steele et al (1981) develops a methodology for instantiating (i.e. establishing an instance of) AUX in any given language. Described and tested in Chapter 4, this methodology is the product of a quest for categorical equivalence and the isolation of AUX in particular languages. In order to fulfill both of these requirements, the AUX methodology establishes both definitional similarities and non-definitional similari ties surrounding AUX. In other words, the elements which fit the definitional model of AUX must also exhibit a series of non-definitional similarities if the category, truly is cross-linguistically equivalent (12-13),

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 89

: Kaisse responds to these assertions in a 1984 review of Steele et al. (1981). Citing a lack of detailed investigations (there were four languages under study in the volume), and an editorial procedure which did not coordinate all of the contributions to the volume, Kaisse (19841928) echoes the sentiment of Pullum regarding the constituency of AUX, "As to other languages, progress can be made at this stage only through the presentation of arguments.. .for a wider range of typologically distinct languages" (19811461). With this review of the AUX literature this dissertation returns to the original point of controversy, that is, the number of constituent categories in the kernel orientations.^

Arguments Favoring AUX as an Independent Sentential Constituent The discussion of the AUX literature established the focus of the AUX controversy! is the category a sentential constituent? Proponents of the constituency of AUX conceptualize the category as a sentential node distinct from the VP. Arguments supporting this posi tion revolve around two issues!

^Only the arguments which accept and reject AUX constituency are discussed in this dissertation. Because the question of implicational universals arises after a category has been instantiated, this question is not explored in this study which focuses on the presence of AUX in a single language, Makah.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 90

1. The need to include the category in the grammars of particular languages. This issue concerns the syntactic uniqueness of AUX in the grammar of a language. 2. The need to include the category in Universal Grammar This issue concerns the equivalence of language- specific AUX categories across language boundaries.

For proponents of AUX constituency, the Universal sentence or the PS representation of the sentence for a particular grammar looks like the phrase marker illustrated in (89):

(89) S

NP AUX VP

The phrase marker illustrates the perspective that AUX is directly dominated by the sentence (S), and as such, must be uniquely described along with NP and VP if the grammar of a language is to be adequately represented. AUX proponents argue that AUX is necessary in the descrip tion of single languages when recurring morphological sequences marking notional categories like modality exhibit syntactic, semantic and phonological properties unique from other categories in a language. Elements marking sentence modality then become the definitional criterion for the process of instantiation, which

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 91

empirically establishes the validity of AUX in a single language by isolating non-definitional similarities accompanying the AUX sequence. Therefore, the identification of a set of language- internal categories has interesting consequences only if the members of the set of categories share certain non-definitional properties of the set of the categories that the definition identifies (Steele et al. 1981:6). Steele's analysis showed that there are four general properties which accompany the specification of a sequence of elements subject to testing for consis tency: position, composition, internal order and relationship to adjacent elements. In the case of AUX, the sequence of elements subject to the constituent testing is comprised of the elements marking tense, modality and other notional categories specified by the language under investigation (This point is elaborated upon in the next chapter.) And, in the case of AUX, when the sequence of elements which fits the definitional criteria also conforms to five generalized parameters established through AUX investigation (i.e. the non- definitional properties of AUX), the category is instantiated for a particular language. This instantiation process uses the syntactic uniqueness of the AUX sequence as the motivating argument supporting the inclusion of AUX as a sentential con stituent. If the AUX is a "locus of phonological,

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 92

semantic, and syntactic properties" CSteele et al. 1981! 25) and the syntax of a language must refer to the elemental sequence in order to produce grammatical constructions (Akmajian, Steele and VTasow 1979!8) the category is a constituent. If the category were a subset of VP, then the elements in AUX would not be syntactically unique; other elements in the language, such as verbal inflections, would also be governed by the syntactic rules which govern the AUX sequence. Akmajian, Steele and Wasow (1979) and Steele et al, (1981) use the following data from Luiseno, a Uto- Aztecan language to illustrate these observations! (90) noo n heyi-q "I am digging" r 1SCP6 dig- present (91) wunaalum pum heyi-wan "They are digging" They 3plCP dig- present (92) noo n il heyi-qujS "I was digging" I 1SCP past dig-past

(93) wunaalum m il heyi-quj{ "They were dig They 3plCP past dig-past ging."

For clarity and convenience, the clitic sequence under discussion is highlighted by underlining the cor responding elements in the English morphemic translation. Consequently, Makah words are not underlined in ename- rated sentences, while Makah units are underlined in textual reference. Abbreviations used in the English morphemic translation are found and defined in the glossary in Appendix 3.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 93

(94) noo xu n po heyi "I dig." I modal IsCP future dig (95) wunaalum xu m po heyi "They dig." They modal 5plCP future dig (9Q-95) are simple declarative sentences in Luiseffo. In each sentence in the corpus, there is an element which corresponds to an English personal pronoun (noo "I", and wunaalum "they"). Adjoining this element is a clitic sequence which marks modality, subject, and tense (in this order) in a succession of slots in the sequence. A % third element, also inflected for tense, occurs after an open juncture following the clitic sequence. This last element corresponds to the English concept of the verb in a sentence. But, in looking at (90-95) it becomes obvious that the clitic sequence does not appear in the same form in each of the sentences. Modality and tense are not obligatorily marked by a particular morpheme in the spoken realization of the sentence. In fact, the only obligatory morpheme in the clitic sequence in the corpus is the second element, which indicates subject marking. This does not mean that the optional nature of morpheme inclusion in the first and third slots of the clitic sequence is devoid of meaning. If the first slot is occupied in a 0 (which translates to the absence of a morpheme in the spoken form), the modality of the

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 94

sentence is indicative. Likewise, a 0 in the third slot also conveys a meaning, present tense in this case. Using the 0 provides an additional way to highlight underlying AUX components in the written corpus, and bring the regularity of thedata to a more obvious level.

(96) noo 0-n-0 heyi-q "I am digging." I ind-lsCP-present dig-present (97) wunaalum 0-pum-0 heyi-q "They are digging" They ind-3plCP-present dig-present (98) noo 0-n-il heyi-q "T was digging." I ind-lsCP-past dig-past (99) wunaalum 0-m-il heyi-q "They were digging." They ind-3plCP-past dig-past (100) noo xu-n-po heyi "I should dig." I modal-IsCP-future dig (101) wunaalum xu-m-po heyi "They should dig." They modal-3plCP-future dig

The sentences show a three position particle sequence follows the initial sentence element. In (96- 97), the particle sequence is 0 - subject marking - 0; in (98-99) the particle sequence is written 0 - subject marking - tense; and (100-101) the particle sequence is written modal-subject marking - tense. Thus, the overall representation of this particle sequence is (1Q2|) (Modal) Subject Marking (Tense).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 95

The clitic sequence residing in the second senten tial position meets the definition of AUX as "a set of elements marking tense and/or modality" CSteele 1978a, ASW 1979, Steele et al. 1981). So, the representation of AUX^ CAUX in Luiseno) is also the form AUX^ * (Modal) Subject Marking (Tense)

When Steele analyzed Luiseno data to relate AUX^ to the four general properties which accompany a clitic or elemental sequence being tested for constituency, she discovered the following in relation to the position, composition, internal order and relationship to adjacent elements of AUX^. AUX^ must occur in second sentential position, adjacent to the sentential position which indicates the pronominal nature of the subject. In addition, the order of the elements within AUX^ is fixed in the order indicated in (102). These findings support the syntactic uniqueness of the AUX^ sequence because the syntactic constraints operating on AUX^ do not operate on other portions of the Luiseno sentence. Only the AUX^ sequence is obligatorily constrained to second sentential position, and the internal order of AUX^ must be preserved if the sentence is to remain grammatical. If the AUX^ sequence shifts to another place in the Luiseno sentence, or if the order of the elements within AUX^ shift relative to one another, the resulting construction is

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 96

ungrammatical. Another generalization concerns the non-definitional notional categories also marked in AUX! "subject marking, subject agreement, aspect, question marking, emphasis, evidential, object marking, object agreement, and negation" (Steele et al. 1981!159) as opposed to the definitional criteria of tense and modality. In the limited Luiseno corpus (90-101), subject marking is a non-definitional notional category marked in the particle sequence. Indeed, this is the only obligatory particle appearing in the surface structure of a grammatical Luiseno sentence. Finally, (90-101) show that the parti cle sequence is attached to an adjacent element. Because the particle sequence exhibits both definitional and non-definitional properties of AUX, it can empirically be termed AUX^ and regarded as a senten

tial constituent equivalent to other AUXjj. This instantiation process hinges on three essential arguments by those scholars who support the inclusion of AUX on the constituent level! 1, Elements under VP are syntactically differentiated from elements under AUX. AUX elements represent a unique nexus of phonological, semantic and syntactic properties not predictable by any other rules in the language. 2. Categories like AUX are equivalent across language boundaries because they rely on definitional and

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 97

non-definitional criteria for identification. Because of this equivalence, AUX is not only a language-speci fic phenomenon; it affects the inventory for the universal sentence as well. 3. AUX has not been seen as a part of the universal inventory because the category is so complex in English, the language most associated with theoretical syntax. Since many of the world's languages have now been entered into the debate through the work in steele Cl978a), analysis shows that the category is more readily identifiable in some languages than others. The Luiseno corpus presents sentences in which the regularity of the particle sequence AUX^ is obvious.

The Arguments Against AUX as an Independent Sentential Constituent There are four essential arguments presented by scholars who categorize AUX as a subset of VP; these arguments constitute the view termed the Main Verb CMV) analysis, which revolves around three issues: 1, The AUX node should be abolished, 2, AUX is a special subset of verbs, 3, An element within a proposed AUX sequence can be a main verb, CPullum 198l;435) The phrase marker associated with the MV position represents the constituency of the sentence in C103)•

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 98

Cl03) S

NP • VP / \ AUX V

When compared to the PS representation of the sentence in (89), the phrase marker in (103) shows only two sentential constituents, NP and VP, as opposed to three constituents in the phrase marker in (89) . The only difference is the location of AUX in the respective representations. While (89) includes AUX under direct domination of the S, (103) includes AUX under the domination of VP, not S. For MV proponents, AUX is a verbal subset, not a sentential constituent. MV supporters argue that the verbal nature, as opposed to independent nature, of AUX is supported by the fact that certain syntactic transformations operate on both verbs and elements within the AUX, (Very little MV analytic material is available for non-Western languages; English examples are used because they are the clearest examples of the MV argument.) Pullum and Wilson (1977:742-746) use the following argument to demonstrate their contention that AUX is a verbal subset,

AUX elements cannot comprise a syntactically distinct category unless these elements act independently in a syntactic frame of reference. The transformational series

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 99

known as Affix-Hopping described in detail in the Akmajian and Heny (1980) introductory text, An Introduc tion to the Principles of Transformational Syntax) operates in English and affects elements which meet the definitional criteria for AUX in English (104) AUXg CModal) (have+ en) (be + ing) (Chomsky 1957:112) as well as the elements which are verbs (V) in English. V follows the (be+ en) component; (105) illustrates this structure, called VP for this argument. (105) VP -*■ (Modal) (have + en) (be + ing) V Before proceeding through the Pullum and Wilson (1977) argument, a word about the Affix-Hopping rule is in order. This rule operates to move the affix ineach paranthetical slot to the end of the element to the immediate right of the affix. For the construction in (105), the substitution of 'eat' for V, and the applica tion of the Affix-Hopping rule produces 'have been eating'. The -en has moved to attach to the element immediately to the right, giving be + -en, or 'been'. The -ing has moved to the immediate right and attached to V, 'eat' in this case, giving 'eating'. Pullum and Wilson argue that the structure of the Affix-Hopping rule negates the syntactic uniqueness of AUXj, because the rule operates on the elements in (104) and the V. In effect, the Affix-Hopping rule operates

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 100

on (105), a VP which includes both AUXg and V elements. This point is further clarified by the addition of another construction affected by the Affix-Hopping rule. Because the (be + en) component in (104) is optionally included (as indicated by the parentheses), the English VP can also take the form in (106) VP (have + en) V. Again substituting 'eat' for V, 'have eaten’ is produced when the Affix-Hopping rule, operates. This rule still operated on (106), in spite of the fact that the terminal element V, and not a part of AUXg, takes -6n (Pullum and Wilson 1977:743), which is an AUXg element by definition. Pullum and Wilson argue that the Affix-Hopping rule does not operate on an exclusive syntactic category called AUXj,. The Affix-Hopping rule operates on elements defined in (104) as AUXg and the V, which together comprise a more inclusive grammatical category VP. For AUXg to possess syntactic integrity in English, unique syntactic rules should apply to AUXg and to VP, respectively. According to the 1977 analysis, there is no syntactic evidence to support AUX constituency; there is only evidence to support the verbal nature of AUXg. The lack of support for the syntactic uniqueness of AUXg in the Pullum and Wilson analysis extends to the question regarding the inclusion of AUX in the universal grammar. Because AUXP was not fouricT to be a necessary

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 101

or motivated part of English grammar by Pullum and Wilson, the contention remains that AUX may not be necessary in the grammars of other specific languages; consequently, the MV analysis supports the two constituent universal grammar suggested by the phrase marker in (103). A second argument supporting the inclusion of AUX as a verbal subset stems from the general linguistic principle that the simpler structure of two proposed structures is preferable, all other things being equal. A two constituent grammar, rather than a three consistu- ent grammar, would then be more desirable from an analytic standpoint (Pullum and Wilson 1977:783). Although these two arguments, the verbal nature of AUX and the simplicity of the two constituent grammar, were refuted in Akmajian, Steele and Wasow (1979), Pullum responded in 1981 and supported the MV position by offering a refined analysis and two more arguments against AUX constituency. Pullum discussed two problems with the 1979 analysis raised by Akmajian, Steele and Wasow (ASW). One major argument concerned the definition of AUX used in the 1979 ASW analysis. Pullum viewed the definition, which relied on the isolation of modal elements as the salient criterion, as constituting a class of elements entirely too broad for investigation. In effect, the choice of modal elements as the definitional criterion for the

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 102

surface existence of AUX insured the isolation of the category because "far too many thing meet the definition” (Pullum 1981:437). The other argument initially appeared in Pullum and Wilson (1977) but was refined in Pullum (1981). Accord ing to Pullum, AUX by definition cannot be a universal category; semantic input from the broad modal class makes each proposed AUX a unique category. Because the modal category tends to be subject to culturally specific priorities (time reckoning, temporal contour, truth value, and the like), "no unified phenomenon exists linking such diverse phenomena as the English 'modal' verbs and the question or hearsay particles of Luiseno" (1981:438). A final objection to AUX constituency was summarized in' Kaisse (1984), when the author questioned the inte grity of an AUX string based on criteria such as the fixed order of the elements within AUX or the phonological interactions which occur between adjacent elements within an AUX component. Kaisse cited Zwicky (1977), Kaisse (1981, 1984) and Piillum (1981) to argue that proposed AUX criteria such as fixed order of elements and mutual phonological interaction of elements are properties of clitics and of strings of clitics (1984J925). Her argument would eliminate the analytic need for an AUX component by attributing properties within the proposed

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 103

AUX constituent to the clitics that mark tense and/or modality in a language. By this argument, reference to clitic features, as opposed to the properties of a con stituent category AUX, explains the phenomena associated with the elements marking tense, modality and the like. Steele responded to the clitic feature arguments of the earlier sources cited in Kaisse (1984) in her discussion of Luiseno in Steele et al. C1981). In response to the point which relegates the phonological and syntactic properties of the AUX sequence in Luiseno to clitic features, Steele raised the counterargument which called for evidence relating the AUX sequence to the boundaries of another syntactic constituent. Since the evidence could not be produced for Luiseno, Steele argued that "if there is no evidence that the members of the particle sequence come from some other constituents) --and there is not--then there is no reason to consider some level where ^:he evidence adduced in the Akmajian, Steele and Wasow article does not hold" 01981:119). The MV arguments against AUX sentential constituency can be briefly summed in four points: 1. AUX is not a syntactically unique category. It is better described by rules which include AUX under the sentential node of VP. 2, A two constituent grammar constitutes a simplified grammar, and is therefore desireable.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 104

3, AUX is, by definition, a broad category which includes too many potential elements to be useful. 4. The semantic nature of the AUX definition brings together too many disparate elements across language boundaries, and therefore, does not con stitute a natural class. From the presentations in the last two sections of this chapter, it is apparent that proponents and opponents of AUX constituency do not agree on a number of fundamental theoretical issues regarding the struc ture of the sentence in its most basic form, or on the methodology one must use when attempting to test the construct of a syntactic constituent.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 5

AN INSTANTIATION OF AUX IN MAKAH (AUX^

The Analysis*. Definitions and Methodology Steele et al. C1981) define AUX as C107) ...those constituents which may contain only a specified Ci*e. fixed and small) set of elements marking tense and/or modality. This definition operates within a "given set of language- internal analyses" which accept the following assumptions as 1. Categorical equivalence exists across grammatical boundaries of respective languages. 2. Phonological, syntactic and select semantic criteria which exist on the surface level of language provide adequate data to establish sentential constituency. 3. For any language N (LN), there exists an AUX.. which contributes data in tne consideration of a universal AUX (AUXy). Final determination of the nature of a fixed universal inventory relies on AUX instantiations for many more of the world's languages. Steele's process for instantiating AUX^ follows this progression! 1, Start with the definition of AUXn in Steele et al. C l 9 8 1 ) u 2. Isolate the definitional criteria of AUXy in 1Q5

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 106

3. Identify the elements in Lj,j which share definitional similarities with AUXy. Term these elements AUX^. 4. Identify the properties of AUX™ relating to the following parameters! the position of AUXN in S™, the composition of AUX^, the internal order of the elements syntactically constrained to AUX™, and the relationship of AUX^ to adjacent elements in the Sjj. 5. Evaluate the linguistic interest of the properties of AUX™ in relation to the non-definitional criteria currently proposed for AUXy.

According to this analysis, AUX^ (AUX in Makah) is

said to be instantiated if the set of elements, AUXjj, shares both definitional and non-definitional similari

ties with AUXy. The remainder of this chapter follows

the Steele et al. procedure to explore AUXy properties

in Makah and validate AUXjj.

AUX^ Applied to Makah

Each Makah sentence (Sj^) contains a stem (often inflected for aspect) or word and a clitic sequence! the latter is obligatorily attached to and follows the former. Thus it is the second sentential element. Additional sentence elements follow an open juncture and can be represented as #n#J sentence elements expand to the right of the first surface element. Sentences (108-121) illustrate a representative cross section of the most common enclitic particles in Makah. All sentences contain the stem kup- ’point" inflected for the momentaneous aspect -Sik-, so the effect of the clitic sequences present in the respective

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 107

sentences are readily apparent. Each example presents three lines of data. The first line is a representation of the spoken Makah sentence. The second line is a mor phemic representation of the spoken Makah sentence, and the third presents a morpheme-by-morpheme translation in English. Cl08) lftipSiks I point, tfup- Sik-s point - mom- Is Cl09) tfupSikits I pointed, tfup- Sik- Bit - s point - mom- past- Is C110) tfupS’eyiks I will pointCdefinitely) . tfup - sik- eyik - s point - mom- definite future - l£ Clll) l&ipSitfe*iss I intend to point Cat some tfup - Sik- V ’is - s point - mom - indefinite future- Is C112) IfupSikwa'tsi* I am told I point, tfup - Sik - wa*t - si* point - mom - quot-1s-absolutive C113) tfupSiksiSi* I point Cas a response). Hup - Sik - si - Si* point - mom- la - responsive Cl 14) l£upSika*ks Do I point? tfup - Sik -Qa*k - s point - mom - confirmational int -Is C115) tfupSikqeysi*k Iwish I could point, tfup - Sik - qey - s-i*k point - mom - cond. - Is - habitual

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 108

(116) l&ipSikike* Point, (as a direction or Kip - 51* - ’1 - ke- advice) point - mom - imp - advisorial (117) tfupS?ak * Point! (as a command) tfup - Sik - *ak - *i point - mom - temporal proximity - imp (118) JfupSikV’^isaks 1 will point (at some time). Jfup - Sik - 'e*?is - *ak - s point - mom - indefinite future - temporal proxi mity- - Is CH9D l&ipS’akits I pointed (just now) . tfup - Sik - *ak - Bit - s point - mom - temporal proximity - past - Is

(120) lAipSikita’ks Did I point? tfup - Sik - Bit -Qa'k - s point - mom - past - confirmational int - Is

(121) lfupSiksi*cuXw I point at you. ldup - Sik - si* - cuXw point - mom - lsub/2obj Certain morphological elements common to other languages and often assumed to be essential for any grammatical system are lacking in this corpus. Makah does not inflect sentences for the present tense by using a particular morpheme which indicates this tense. Likewise, the language does not use a morpheme to mark indicative modality. (108), (109), and (114) illustrate this point.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 109

Cl08) tfup$i*s I point. Kup - $i* - s point - mom - Is C109) l&ipSi* I pointed. lAip - M * - Bit - s point * mom - past - Is (114) ]£upSi*a*ks Do I point? tfup - Si* - Qa*k - s point - mom - confirmational int - l£ (108) references both present tense and indicative modality, as compared to (109) which references past tense and indicative modality. (114) also presents a present tense utterance, but poses a question using the confirmational interrogative modality. The presence of underlying -Bit- in (109) changes the reference of (108) to tpast tense, while underlying -Qa‘k- in (114) changes the reference of (108) from indicative modality to confirmational int. modality. No other structural changes in these sentences take place when inflection for tense and/or modality occurs. With the presentation of these sentences from the corpus, it is possible to complete steps one through three of the instantiation process. The definition of

AUXjj specifies that a fixed and small number of elements which mark tense and modality comprise the center of AUX for any L^. Therefore, for Makah, the elements which meet the definitional criteria of AUXy are the elements

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 110

described above that mark these two categories for Makah. The clitic sequence under consideration in Makah marks both tense and modality, the analysis to this point has established

C122) Clitic Sequence + (Tense); (Modality) Subject Marker in which tense and modality are optionally marked in surface structure, while subject marking is mandatory. Because the definition in (107) allows for a category X to be called AUX if X marks tense and/or modality, sufficient evidence exists to label the clitic sequence governed under (122) as AUX^,

The Position of AUXjj

Now the analysis proceeds to an investigation of the non-definitional criteria of AUX^! the position of AUXj| in S^j, the composition of AUX^, the internal order of the elements in AUX^, and the relationship of AUX^ to adjacent elements in V AUXjf occurs in one, unambiguous position in a

specific Ljj according to Steele et al. ^Syntactic con

stituency relies on this fact, for AUXjj must be a "clustering of semantic, syntactic, morphological and phonological properties" which must "associate members of the string and distinguish them as a unit from the other parts of the sentence" (Steele et al. 1981!4, 25).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ill

In other words, AUX^ must be syntactically and distributionally distinct from any other constituent (C^) in

AUXjj is positionally distinct from other categories in SM ; it is restricted to second sentential position. This point becomes clear when sentences (108-121) are compared to the ungrammatical sentences (123-128), below! (123) *kup - Bit - Sik - s point - past - mom - Is (124) *kup - s - Bit - £ik point - Is - past - mom (125) *Bit - s -kup - Sik past - Is - point - mom (126) *kup - $ik - Qa’k - Bit - s point - mom - confirmational int - past - Is (127) *s - kup - Sik Is - point - mom (128) * V? i s - *ak - kup - Sik indefinite future - temporal proximity - point - mom (108-121) display grammatical utterances containing 7 two positions! initial and final. A category X occupies the initial position and the regularly occurring clitic sequence now called AUX^ occupies the final position. This pattern does not vary in (108-121). For each

7 This representation supercedes the Makah sentential representation in Gill and Renker (1987).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 112

with one surface element, AUXjj occurs after X. No variations of elements in X (stems and words) with any of the tense/modal elements in AUX^ will produce grammatical constructions.

In (123-128) AUXji has been moved within the Sj|, either shifting sentential position completely (125, 127, 128) or disturbing the expected internal order of the sequence by moving morphological units within AUXj^ from one position to another (123, 124, 126). The position of AUX^ appears to be constant in utterances with on surface unit. In order to establish the syntactic uniqueness of AUX^, it must retain the same characteristics when more than one surface element occurs in S^j. Analysis of the sentences (129-135) shows how this can be done! (129) wiki's kupSik I point (NEG) wiki* -s. kup - Sik not - Is point - mom ASPECT (130) Sabuks kupSik I can point fcabuk - Is kup - Sik ability - Is point - mom — ASPECT (131) yubuks kupSik I can’t point yubuk - s kup - Sik inability - Is point - mom ~ ASPECT

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 113

(132) fcabukita’ks tfupSik Can I point or Am X able tabuft-Bit-Qa' ks l&ip-Si* t0 point Cp e ™ i « i 0S f ABILITY-PAST - confirmational-ls point-mom TENSE int ASPECT

(133) Cabukita'ks UupSik Was I able to point? or fcabuk-Bit-Qa’k-s tfup-$i* Could 1 Point? ABILITY-PAST - confirmational-ls point mom TENSE int ASPECT

(134) ?usu*bas IftxpSik I need to/have to/ ?u-sub(a)-s lfop-sik should point. ref stem-need-ls point-mom ASPECT (135) kupMJrsu*bas I need to/have to/should tfup-5i*-sub(a)-s point point-mom - need - Is ASPECT In (129-135) the S^ now have two surface elementsJ underlying structure now contains first, second and third positions. The previously used position terms, initial and final, must be altered to correspond in circumstances such as (108-121). Initial position (occupied by category X) becomes position 1, and final position (occupied by AUX^) becomes position 2. Position 3 O (occupied by category Y) is the #n# to the right of the open juncture bounding position 2. Making these adjustments, the critical differences between sentences (108-121) and (129-135) now becomes

8 This representation supercedes Gill and Renker (1987).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 114

clearT note how #upSi*- appears in first surface position in (2) and third surface position in (129-135) even though AUX^ remains in second sentential position in both underlying sequences. As shown in (136-138), ffupSift- cannot appear in second position and produce grammatical constructions. Only AUX^ can occur in second position, and only in this second position in simple Makah sen tences. (136) *tabu4-tfup$i* ability - point (mom) (137) *yubui-l£up5ik inability - point (mom)

(138) * ?usub -l£up5 i k need - point (mom) According to these data, there are constraints which

apply to AUXjj but not to any other syntactic elements in the S^. This second position constraint identifies AUX^ as a unique grammatical element in the language.

Composition of AUX^

Next, the analysis must concentrate on the composi tion of AUXj^. By definition, the analysis must look for a "set of elements marking the notional categories of tense and/or modality" (Steele et al. 1981!4) as well as other elements which are syntactically relegated to second position in underlying structure.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 115

AUX^ is a regularly occurring, syntactic element composed of seven distributionally exclusive morphologi cal slots. Each slot corresponds to an identified unit

in AUXjj structure, as illustrated in Chart 8. Each slot

is now described, in its order of occurrence in AUXjj, in the following sub-sections.

CSlot 1) INDEFINITE TENSE9 Only two morphological choices can occupy slot It the clitic - *e*?is-, which indicates an indefinite time in the future, and the 0, which conveys the information that time is definitely established, within reason. The nature of this slot, like others to come, does provide for an 'either-or' reference distinction for each S^. Either the time of an event or action is reasonably well known, or it isn't. In the case of - *e’?is-, the uncertainty or indefiniteness applies to future actions, because one assumes the time of an action happening in the present or past is definitely known to a speaker. Should the speaker not be personally familiar with the

g In order to highlight the categories within AUX^ in textual reference, this dissertation presents the names of AUX slots and markings within AUX slots in upper case letters. This convention is necessary in this disserta tion because underlined references and references in quotations carry linguistic meanings. Underlined references in text refer to Makah morphenes or wordsJ underlined elements in data sets indicate AUX. Single quotation marks are used to indicate an English gloss in text.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 116

CHART S.— AUXj^ in Makah Sentences

ENGLISH MAXAH Stem Aspect Indefinite Temporal Dei Modal Subj. Obj Soeake Tense Proximity Tense Marker Marker Intent

els'1) I point HlnSiks Idip Sik 19 0 0 0 -si 0 0 (140) I point (now) HipS’aks Hip Sik 0 ’ak 0 0 -si 0 0 C141) I pointed HipSikits Hip Sik 0 'ak -Bit- 0 -si 0 0 (142) I just pointed HipS7akits kbp Sik 0 'ak -Bit- 0 -si 0 0 (143) I will point HipSik ’e‘7isaks kbp Sik ’e'7is ’ak 0 0 -si 0 0 (144) I will point HipSk7eyiks kbp Sik 0 0 ’eyik 0 -si 0 0 (14S) I will point Hipsik V ’iss Idip Sik V 7is 0 0 0 -si 0 0 (146) I point HipSikwatsi* kbp Sik 0 0 0 wa.t -si 0 0 (I'a told) (147) Yeah. I point HipSiksiSi’ kbp Sik 0 0 0 0 -si 0 -Si. (148) Do I point HipSika'ks kbp Sik 0 0 0 Oa’k -si 0 0 (14?) Do I point now? HipS’aka'Ics kbp Sik 0 'ak 0 Qa’k -si 0 0 (ISO) Did I point? HipSikita'ks kbp Sik 0 0 -Bit- Qa*k -si 0 0 (151) Did I iust Hip57akita'ks kbp Sik 0 'ak -Bit- Oa’k -si 0 0 point?

(152) Point Hip57ak ’ kbp Sik 0 ’ak 0 0 i 0 0 (153) Go and point HipS7akike" kbp Sik 0 ’ak 0 0 a 0 ke. now.

(154) Go and point HipSiftke* Hip Sik 0 0 0 0 1 0 ke. (155) Did I just HipShka ‘ ksi "cux" kup Sik 0 'ak 0 Qa'k -sit cuk- 0 point? (156) I wish I HipSikqeysi'k kbp Sik 0 0 0 -qey- -si 0 -i*k could point.

(157) I might point. HipSikqeys kbp Sik 0 0 0 -qey- -si 0 0 (158) Do you point. HipSilic kbp Sik 0 0 0 0 ■°ic 0 0 (159) You pointed. HipSik7ue kbp Sik 0 0 0 -°ic 0 0 (160) You will point. HipSike‘7isic kbp Sik V 7is 0 0 0 -°ic 0 0 (161) Point at you. Hip5iksi'cu3c“5i' Hip Sik 0 0 0 0 -si? cu*“ 0 (162) Yeah, I point HipSiksi'cuJt'Si' kbp Sik 0 0 0 0 -si? cuk” -Si. at you.

(163) I pointed at HipSikitsi’cuX” kbp Sik 0 0 -Bit- 0 -si? cuk" you.

(164) Will I point? Hip5i/e'7isa'ks kbp Sik V 7is 0 0 -Qa'k -s 0 0 (165) Will I point? Hip57eyika'ks Hip Sik 0 0 - ’eyik-• -Qa'k -s 0 0

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 117

past/present action or event, a modal clitic such as -wa’t- "quotative" signifys this facts, not - *e,?is-. While PAST and PRESENT markers reside in the DEFINITE TENSE slot Cslot 3) a FUTURE DEFINITE TENSE, - ’eyik- also occupies slot 3. The fact that both INDEFINITE and DEFINITE TENSE markers exist in Makah is illustrated in the following sentence pairs. Sentences (143-144) on Chart 8 also reflect this same distinction. (166) a. dudu*ke*7iss MI will sing (at some point)" or dudu k e is sc n]c^ intend ^ teppe^,,to sing (but who knows sing - indefinite future - Is b. dudu'keyiks "I will sing (at a specified dudu'k - ’eyik - s t5me) ■" sing - definite future - Is (167) a. ha'^utfe'?iss "I will eat (at some point)" or TiaTui- - "I intend to eat (but who knows " e 1S ' s what will happen)." eat - indefinite future - l£ b. ha?uk5eyiks "I will eat (at a specified ha?u - uk - ’eyik - s time). eat - dur definite future - Is The significance of this definite/indefinite distinction, and the need for accurate analysis of it, will become clearer as the analysis progresses. The conclusion that - ’e*?is- and - ’eyik- are distinc tive is supported by the distributionally exclusive domains of each clitic. While both are unable to appear with the PAST marker -Bit-, other morphemic evidence

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 118

supports the distinctiveness of - *e,?is- from - ’eyik-. This evidence relates to the TEMPORAL PROXIMITY marker - *ak-. - *ak- follows - *e*^is- in a grammatical (143), cannot occur in a with - *eyik-, and precedes the PAST marker -Bit- (142,151) in a grammatical SM . - *e*?is- and the other two tense markers cannot belong to the same morphological series because of this difference regarding combination with - *ak-. From a morphological standpoint, the importance of the distinction can now be addressed. - *e‘?is- and - ’eyik- occupy distinct places within the AUXM sequence, as evidenced by the combinatory differences with - *ak-j this distinctiveness calls for the two elements to be

treated as occupants of two distinct slots within AUXjj, rather than as morphemes in complimentary distribution within a single slot. - *eyik- and -Bit- belong in an

AUXjj category which marks time reference definitely known to the speaker. - *e,?is- is restricted to a separate

category in AUXjj which indicates a time reference indefinitely known to a speaker. Jacobsen (1979) describes this contrast between definite/indefinite reference. However, he translates - ye*?is-* to mean "primarily that one will go (or come) somewhere before doing something, but since travelling necessarily takes a certain amount of time, the meaning seems often to be that of a more distant future" (31).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 119

I submit for reasons just presented, that the situation actually works in reverse; primary distinction is one regarding the definite/indefinite nature of time reckon ing in the statement, and because travelling is always risky in a temporal sense, - *e’is- is used for travelling contexts.

(Slot 2) TEMPORAL PROXIMITY (TP) - This slot involves

a choice between two elements! - *ak- and -0-. -0- indicates that the sentence reference is one of the tense specified, either PAST (-Bit-) or PRESENT (-0-). However, if - *ak- occupies the TEMPORAL PROXIMITY slot, AUXM indicates an action or event which occurs as closely to the specified tense as possible. Consider, in addition to (139-142) on Chart 8, the following sentences! (168) a, dudu’ks I sing, dudu’k - s sing ^ l£ b, dudu’liaks I am singing now. dudu’k - 'a* - s sing - temporal - Is proximity (169) a. dudu’kits I sang, dudu’k - Bit -_s sing - past - Is b, dudu’tfakits I sang (just now), dudu’k - 'ak - Bit - s sing - temporal proximity - past - Is

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 120

In each pair, the addition o£ - 'aft- in the AUX^ sequence adds a degree of immediacy, either in the PRESENT or the PAST; hence the name of the slot. Jacobsen (1979) termed - ?aft- the immediacy suffix, but that label has been changed to TEMPORAL PROXIMITY to accommodate the change which occurs when - 'aft- combines with - 'e*?is- in AUX^ structure. In this case, the sentence containing - 'g* ?isaft- describes an action or event as a progression which is now nearing completion. Thus, (170) ia'XukSiJfe'^isal He will become a man. ifca'Xuk - Sift - 'e*?is - 'aft - °i man - mom - indefinite future - temporal - 3s proximity C171) siqa£itfe*?isal It is almost cooked. siqa - Sift - 'e'7is - 'aft - °i cook - mom * indefinite future - temporal - 3s proximity (172) #i*daqSi#e*?isal It is going to get foggy. tfi'daq - Sift - V ’is - 'aft - °i fog - mom - indefinite future - temporal - 3s proximity The - 'aft-* element also figures prominently in the impera tive construction. The morphology of this Makah construction is well described in Jacobsen (1979b), analyzing the components of imperatives within the AUX^ framework allows for better comparison and contrast with other Makah syntactic forms. In an imperative construction in Makah, slot 2 contains - 'aft- and slot 5 (Subject) contains - 'i-, which

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 121

can be inflected for plurality, - *iC-. The intent of the command is modified by inclusion of one clitic from slot 7 (Speaker Intent or Intentive)J -ke* indicates that the speaker intends the command to be taken as advice or a direction. Sample sentences appear as (153-154) on Chart 8. The formation of clauses which translate to the ENGLISH "when" clause contain the final instance in which - *ak- appears in slot 2 of AUX^. Since the morphological discussion in this section concerns G-jS and not sentences with multiple distinct surface elements, this function of - *ak- will be discussed in Chapter 6.

(Slot 3) DEFINITE TENSE - This component of AUX^ can be occupied by three mutually exclusive elements! 0, indicat ing PRESENT, -Bit-,1 indicating PAST and - *eyik-, indicat ing DEFINITE FUTURE (135, 141, 144 respectively) on Chart 8. While PRESENT and PAST can be preceded by -ak-, DEFINITE FUTURE cannot. Interestingly enough, only -*eyik- can contract with the preceding momentaneous aspect marker (Jacobsen 1979), because the morphological slots between the aspect marker (not in the AUX^) and the slot containing - feyik- (slot 3) are empty. (144, 165) on Chart 8 illustrate this contention, contrasted with (140, 142;jL49,15I-154j), where the - *ak- contracts with the momentaneous aspect marker.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 122

(Slot 4) MODALITY The four most common MODAL morphemes are illustrated in Chart 8! -0- indicating INDICATIVE, -wa*t- indicating QUOTATIVE, -Qa*k- indicating YES/NO QUESTION (or the confirmational interrogative), and -qey-, indicating the CONDITIONAL. Other MODAL markings do occur, such as -Qi*k-, INFORMATIONAL INTERROGATIVE*, -X-, SUBORDINATE, and -pi*t~, INFERENTIAL. A few other MODAL markers are detailed in Jacobsen (1973)J they are extremely rare in my corpus and are therefore not discussed here. All of the MODAL clitics in slot 5 can combine with the PAST marker -Bit- in the preceding slot of AUX^. - *eyik- is somewhat more restricted in its combination with MODAL CLITICS, and is only commonly associated with the -0- and the -Qa*k- in the MODAL series. - *e*?is-, the INDEFINITE FUTURE marker from slot 1, is also generally restricted to co-occur with the two MODAL markers -0- and -Qa*k-, but can be followed by -qey- in certain clause formations which will be discussed in Chapter 6. None of these MODAL particles are themselves inflected for person or number. Those inflections are carried on the subject and object markers which occupy the two remaining slots respectively, in the AUXj^. The lack of inflection on these MODAL clitics, aside from morphophonemic changes associated with certain subject

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 123

or object markers, is one criterion which establishes the distinctiveness of slot 5 from slot 6. Many of the MODAL elements residing in slot 4 of AUXjyj were originally described as part of the pronominal series in Makah (Jacobsen 1973). By isolating the MODAL markers from the morphemes which indicate SUBJECT and OBJECT reference, this analysis provides a better means for describing the syntactic regularity which exists within the variations possible in a S^. In addition, the separation of elements within slots 5, 6 and 7 in AUXM facilitates a clearer and simpler description of the combinatory rules which govern the formation of S^.

(Slot 5) SUBJECT MARKER - In order for an AUXM to be complete in a simple sentence, a SUBJECT MARKER is obligatory. Unlike the other elements which appear in the AUXj^ sequence, a SUBJECT MARKER alone is required for the AUX^ sequence to create a simple sentence when attached to a Makah stem. In more complex constructions, it is the lack of a SUBJECT MARKER in the AUX^ sequence which identifies the existence of a clause in Makah, While this phenomenon is discussed in greater detail in Chapter 7, it is important to note that the appearance of a SUBJECT MARKER in AUXjj is obligatory in simple sentences; in complex sentences, an AUX^ without a SUBJECT MARKER indicates a clausal formation.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 124

The morphology of Makah SUBJECT MARKER is exhaus tively discussed in Jacobsen (1973). This analysis is easily amplified to demonstrate that the definite/ indefinite distinction which figures so prominently in the AUXj^ sequence is reflected in the SUBJECT MARKER slot. Jacobsen (1973) presents two series of endings which he terms indicative and non-indicative. Indicative endings occur with modalities marking definite con ditions! indicative statements, responses to questions,

and the like (see 139-144, 146), respectively, on Chart 8 while non-indicative endings, shown for first person on Chart 8 with a final long i (i*), co-occur with indefinite situations such as CONDITIONAL statements (156) and QUOTATIVE utterances (146). Note just one of the contrasts between these items! the indicative (or definite) series marks the INFEREN TIAL EVIDENTIAL,10 -pi’t- (a statement by a speaker

Regarding the marking of evidential reference in Makah, Jacobsen writes, "As this category is semantically characterized, it will often fail to be a structurally homogeneous one..." (1982!1). The analysis in this dis sertation submits a contrasting opinion* The evidential modal suffixes (-wa*t-, -pi*t-, and -cfad-) do present a structurally homogeneous class occurring within AUXw, and as such, are linked with the sentence formation process associated with AUXw. Other elements classified as evidential markers m Jacobsen (1982), such as the suf fixes -Kuk ’M - and -ckwi* - are not a part of the sentence formation process, and occur before AUXw in sentential constructions. In addition, the suffixes, which indicate 'looks like’ and 'remains of,' respectively,

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 125

based on physical evidence), while the non-indicative Cor indefinite) series marks the QUOTATIVE, -wa*t- (a statement by a speaker based on verbal evidence from someone other than the speaker). Events experienced by the speaker are definiteJ events experienced by anyone other than the speaker are indefinite experiences for the speaker. C173) bika’pi'd It is raining. (Based on TSiVa* °i physical evidence seen - pi z i Qr experienced by the rain - inferential - speaker) . (174) bika’wa*d It is raining. (Based bika’-wa't - °i on ver^al evidence from someone else). rain - quotative - 3s This distinction provides further evidence for a definite/indefinite distinction in Makah grammar.

Slot 6 OBJECT MARKER - The morphology and use of OBJECT MARKER is detailed in Jacobsen (1973). His analysis shows that reflexive OBJECT MARKERS do not exist in Makah, and that third person singular and plural OBJECT MARKERS do not exist on the surface structure of the language. Instead, OBJECT MARKERS which do exist in Makah surface structure are relagated to combinations which involve second person SUBJECT and first person OBJECT, and the

are not evidential markers in a semantic sense, because they are used to describe a physical state or appearance of an object, rather than indicating the source of truth or definiteness of a speaker's assertion.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 126

reverse. Consequently, the AUX^ sequence posits an underlying 0 in circumstances where a third person singular or plural OBJECT exists. In fact, all Makah sentences without an explicitly stated OBJECT in surface structure do possess a -0- in underlying structure. The following sentences help to illustrate this point! C175) ha?uks I eat. ha?u - uk - s eat - dur - l£ (175a) ha?uks I eat (something). ha?u - uk - s - 0 eat - dur - ls_ - 3sohj (175a) presents a more thorough morphological description of a simple Makah sentence, which understands there to be a "something," "someone," or "someplace" in the OBJECT function at all times. Though only certain OBJECT categories are represented in surface structure in AUX^, the OBJECT category is always present in underlying structure! this fact is accounted for in the AUXj^ sequence, because -0- fills the underlying slot and accounts for the absence of an OBJECT MARKER in surface structure. This point is further clarified by Makah speakers. When speakers are asked to translate (175), they generally provide the English translation in (175a) as opposed to (175). This corroborates the analysis in Jacobsen (1973),

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 127

which indicates that the absence of third person singular and plural markers is a regular feature in Makah, while OBJECT MARKERS in other person and number categories follow the SUBJECT MARKERS.

CSlot 7) SPEAKER INTENTION - This category contains clitics which clarify the intention of the speaker making an utterance. For example -i*k tells the hearer that the speaker intends the action or event discussed in the preceding AUXj| slots to be viewed as HABITUALJ -Si* indicates that the statement made is in response to a question, rather than an unsolicited statement (RESPONSIVE). -ke*, discussed previously, indicates that a command given is intended as advice or direction (ADVISORIAL). Each of these clitics must occur on the right-hand boundary of the AUX^ structure. None are inflected in any way.

The Fixed Internal Order of AUX^

The previous discussion presented an AUXj^ sequence composed of seven morphological slots in the order’,

(176) AUXm --- CINDEFINITE TENSE) (TEMPORAL PROXIMITY) (DEFINITE TENSE) (MODALITY) SUBJECT MARKER (OBJECT MARKER) (SPEAKER INTENTION) Next, the analysis has to explore the internal order of the AUX^ sequence as written in (176). A previous section demonstrated that the entirety of the sequence cannot move

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 128

from its position in the underlying structure of the S^. Further restrictions which preclude internal shifting of elements from one morphological slot to another. The AUXj^ sequence illustrated in (176) exhibits a fixed internal order which does not vary, despite the syntactic constraints which can affect other sentential position occupants. The sentences (123-128) also demonstrated that sentence ungrammaticality results when the elements in AUX^ are shifted in relation to one another. Chart 8 provides 27 sentences which illustrate the fixed nature of the clitics in AUX^. Note how each of the seven morphological slots in AUX^, represented by the vertical columns, can be occupied by one and only one element from a number of choices assigned to each respective slot alone. The presence or absence of marked forms in each respective slot, as well as the combination of clitics chosen for an utterance, convey differences in meaning. (Morphological specifics and rules for combinations were discussed earlier in the chapter.) Data in the internal order of AUX^ is fixed in two respects! (1) within each morphological slot, because combinations of clitics in an utterance will affect meaning, and (2) between the distributionally exclusive morphological slots which appear in surface forms.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 129

Neither of these orders is predictable based on informa tion regarding other syntactic rules in the language. Even the verbal morphology rules which provide for three stem subclasses (see p. 165) give no clues to the AUX^ constraints. This fact is particularly interesting because the subclass stem membership does predict syntactic constraints on two other sentential positions in Makah (see Chapter 7) . AUXj| is governed by an independent set of combinatory rules which governs the fixed order in (176). In this sense, AUX^ satisfys another of the non-definitional requirements needed for any AUX to fulfill AUX instantiation.

Relationship to Adjacent Elements The point has already been made that AUX^ is always attached to the first sentential element, in spite of the number of surface elements in a simple S^. The point is made here once again, in order to follow the method ology of Steele et al. (1981), which requires that AUX^ be evaluated in relation to the non-definitional characteristic proposed for AUXy. In AUX^, the sequence is attached to one and only one element in a SM , the first sentential element. This investigation has established certain critical

facts about the internal order of AUXjj! 1, The order of the clitics in AUX^ is fixed and

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 130

immutable. No evidence of preferential order exists in the Makah speech community today. If the clitics appear in any order other than C176), the result is an ungrammatical sentence. 2. The clitics which appear in the AUX^ sequence are restricted to one and only one morphological slot in the construction. No clitic may appear in more than one slot within a single utterance, and no evidence exists to show that a single clitic may have member* ship in more than one morphological slot. 3. The integrity of each morphological slot is maintained by a 0 which occupies the slot if no other clitic is chosen for inclusion. This phenomenon finds justification in the unmarked nature of both present tense and indicative modality in the Sjj. CSee Chart 8) .

The Properties of AUXM and Generalizations about AUX^j

The question now becomes one of a relationship of

AUXji to the other AUXjj posited for other individual languages. The AUXM is interesting linguistically if the properties it exhibits coincide with generalizations about AUX in other languages or if it offers information which does not conform to previous standards. Recall that Steele et al. C1981) list five generalizations about the respective sets of clitics which meet the definitional requirements for AUXjj. 1. AUX™ occurs in one of three sentential or underlying positions. While the position may change with each individual language N, the sentential position of AUX is fixed in each language N. 2. AUXj. contains elements marking certain notional types sucn as tense and modality. 3. AUX™ consists of elements arranged in a fixed internal order,

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 131

4. The fixed internal order of the elements in AUXN is not predictable by other language specific proper ties, that is, AUXjj is syntactically distinct.

5. AUXn is generally attached to another element in Ljj. As discussion has shown, AUX^ illustrates a number of these non-definitional criteria. AUX^ is restricted to second sentential position, consists of the notional categories in (176) arranged in a fixed internal order. This internal order relates specifically to AUX^, and does not predict the behavior of other morphemic sequences in the language. Additionally, the clitic sequence AUX^ is attached to the element which occupies syntactic position 1 in S^. Another generalization about AUX^ in relation to

AUXjj concerns the notional categories marked in AUX^.

While AUXjj potentially marks "elements indicating subject marking, subject agreement, aspect, question marking, emphasis, evidential, object marking, object agreement, and negation" (Steele et al 1981!142), AUX^ obligatorily includes only SUBJECT MARKERS in surface structure. OBJECT MARKER is also included in AUX^ though object agreement is not. This exclusion is explained because of the absence of inflected markings for gender in the third person objectj agreement is accomplished by adding proper names or specific subjects in position 3 following the open juncture to the right of AUX^,

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 132

Question marking, whether CONFIRMATIONAL INTERROGA TIVE or INFORMATIONAL INTERROGATIVE, is another category

present in AUX^. Because AUXjj contains the references for both possible types of questions, (yes-no questions and wh-questions respectively) there are no ways to create questions in Makah without making use of AUX^. The evidential category is another important category marked in AUXj^. Because one must indicate definiteness/indefiniteness in using the Makah language, the evidential category functions to specify the degree of definiteness or indefiniteness on the part of the speaker for observations. As with question marking, references to the evidential distinctions cannot be made outside of the AUX^ sequence. In presenting the morphology of the AUX^ sequence, this chapter has provided a means to describe key points which figure in the construction of grammatical simple sentences in Makah. AUX^ morphology as described in this chapter allows the definite/indefinite distinction in Makah to emerge, and offers an analysis which systematically explains the relation between this distinction and morphemic placement of items such - *e*?is- and - *eyik-. This morphological description also demonstrates the regularity which surrounds the elements in the AUX^ sequence, allowing for the formula

tion of rules which can predict sentence grammaticality

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 133

in terms of AUX^. The next chapter expands upon these points, and offers a discussion of some functions of AUXj^ in Makah grammar while clarifying the exclusion of certain notional types from AUX^.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 6

MORE ON THE MORPHOLOGY AND FUNCTION OF AUXM

Chapter 5 isolated a syntactically unique clitic sequence which marks tense and modality in the SM « This sequence meets all of the requirements established in Steele et al. (1981) for AUX validation! 1. AUXw meets the definitional criteria for instanti ation because it marks tense and modality. 2. AUXj^ meets the non-definitional criteria relating to position, composition, internal order and adjacency to another element in the sentence.

3. AUXjj occurs in second sentential position. 4. AUXj^ consist of clitics arranged in a fixed internal order. 5. AUXj» is a clitic sequence not predictable by other language specific properties. 6. AUXj^. is attached to the first element in the Makah sentence. For these reasons, the clitic sequence in Makah with the form! (176) (INDEFINITE TENSE) (TEMPORAL PROXIMITY) (DEFINITE TENSE) (MODALITY) SUBJECT MARKER (OBJECT MARKER) (SPEAKER INTENTION) may be termed AUX^, a language specific example of the universal category AUXy under the methodology cited in Chapter 5. 134

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 135

But, the process of instantiation relied on an assumption of certain morphological and syntactic facts to establish the recurring morphological phenomenon

called AUXjj. This chapter offers additional details about the morphemic and syntactic features which constrain a clitic to placement within AUX^. In addition, the chapter discusses two categories, NEGATION and ASPECT

which are not found in the AUXjj sequence described in C176)J this absence is of interest because the two categories are found within other language-specific examples of AUX.

The Absence of Negation Markers and Aspect Markers in AU5^------

The previous chapter discussed the seven morphologi cal slots which comprise AUX^, the section did not discuss the exclusion of two categories common to Makah sentences from AUX^. These categories, NEGATION and ASPECT, are present in the AUX component of other languages which have been subject to the instantiation process (Steele et al. 1981U59); both are listed amont the notional types which can be marked in an AUX^! AUX„ may include, as well, elements indicating subject marking, subject agreement, question, evidential, emphasis, aspect, object marking, object agreement, and negation (Steele et al. 1981!156). Here then, are two instances where AUXy potential properties do not occur in AUXj|. And, since ASPECT

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 136

appears so prominently in the sentences listed on Chart 8, a discussion which explicitly justifies the exclusion

of these two notional types from AUXjj is in order.

Negation and AUX^ An interesting fact concerning Makah negative sentences becomes apparent when examining the following sentence pairs! (177) a. fcabuds kupSik I am able to point, £abud -s kup - $ik I can point. ability -Is point - mom b. yubuds UupSik I am not able to point, 1 can't yubud - s kup - 5ik point. inability-Is point-mom (178) a. l&ipSikbe'qks I want to eat. tfup - kik- be*qk-s poiiit - mom- desire -Is b. widaqks tfupkik I don't want to eat. widaqk-s l?up-1&ik not want -l£ point-mom. (179) a. t?upuks I am happy. tfupuk - s happy - Is b , wiki * s fupuk I am not happy. wiki * - s fupuk not-ls happy (180) a, wi’yas tfupkik I never point. wiya*-s IfupSik never -Is point-nom

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 137

b. wiya’s Cabui tfupSi* I am never wiya* -s Cabui kup-5i* able to point. never - Is ability point-mom c. wiya’be’qks lfupSik I never want wiya’ - be’qk - s lAip-$i* to eat. never - want -Is point-mom In (177a-b), the position and negative connotations of the sentences come from the Ca- / yu- distinction in the stems in sentential position 1. This Ca-/yu- opposi tion is identifiable elsewhere in the language as well; for example, CaCabaX 'correct, proper' and yuyubak 'wrong, improper, not correct*. Another type of negative reference appears in (178- 179); these sentences illustrate variances in the general stem, wi-/wjk-, which expresses the idea conveyed by 'not* in English. In effect, with the exception of stems like those in (177 a-b), wi-/wik- can be used to negate any positive sentence in Makah, as long as this negative stem assumes position one in the Makah sentence. Finally, sentences in (180) present the stem wiya*- 'never* as the NEGATIVE marker, this element is also restricted to first position in sentence structure. NEGATION in all three instances is marked in posi tion 1. AUX^, it will be recalled, is a second position phenomenon in Makah; hence, without saying anything more on this issue, it is clear that NEGATIVE is not an AUX- related phenomenon in this language.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 138

ASPECT and AUXM Like with NEGATION, this analysis shows that ASPECT is not included in AUXjj. ASPECT involves the inflection of Makah steins for "duration or type of activity" CCrystal 1980*. 34) through morpheme placement and a variety of morphological processes! vowel lengthening, suffixation, and reduplication, primarily. This section

asserts that ASPECT does not reside within AUXjj. Rather, ASPECT and the surface forms which result form the combination of ASPECT and a base or stem are a Makah word-formation phenomenon, rather than a sentence formation phenomenon associated with AUX^. Indications of the boundary separating ASPECT and AUX^ are attested in the sentences in (181-182). (181) a. Uup'Siksubas I need to point fop - H * - subCa) -s 2 U S ? ) ° r point - mom-need -Is (MOMENTANEOUS ASPECT) b. l&ipa’subas I need to point continuously. kup - a* - sub(a) - s (CONTINUOUS point - cont - need - l£ ASPECT) c. kupu'lfii'psubas I need to point Kup -u*- tfu’p - sub (a) -s (REPETITIVE ASPECT) point - epen - red -need - l£ (182) a. ha?uksubas I need to eat. ha?u - uk - sub(a) -s (DURATIVE ASPECT) eat - dur - need - Is

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 139

b. ha?ukSiftsubas I need to start ha?u - uk - Si* - sub(a) - s eating. CINCEPTIVE eat - dur - mom - need - Is ASPECT) c. ha?ukwa*psubas I need to feed ha?u - uk - a^p - sub(a) - s someone. CCAUSATIVE eat - dur - caus - need - Is ASPECT) According to (176), the clitic sequence termed AUX^ contains elements expressing tense, modality and other notional categories. Syntactic proscriptions assign AUX^ obligatorily to second position in the S^. As shown in (181) and (182), the respective ASPECT suffixes -Sift- MOMENTANEOUS, -a’- CONTINUOUS, -ukCw)-/-ak- DURATIVE, -a"p- CAUSATIVE, -uk»Sift- INCEPTIVE (from a combination of the DURATIVE and the MOMENTANEOUS), and the REPETITIVE formation kupu*#u*p- are not adjacent to the AUXj^j sequence, which consists of -s-, the 1st person singular marker in (181) and (182). Therefore ASPECT can not occur within the scope of the AUX^ sequence by virtue of the AUX instantiation criteria. In order for

ASPECT markers to fall within AUXjj sequence, Makah sentences would need to constrain ASPECT markers to begin the AUXj^ sequence in the second position in the SM . (183) and (184) invalidate the possibility of a second sentential position constraint for Makah ASPECT formations.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 140

(183) a. ?usubas tfupSi* I need to point briefly or suddenly. (MOMENTANEOUS) ?u - sub(a) - s tfup - Sik ref stem - need - Is point - mom b. ?usubas tfupa’ I need to point continuously. (CONTINUOUS) ?u- sub(a) - s l£up - a* ref stem - need - Is point - cont c. ?usubas £upu*l&i*p I need to point repeated' ly. (REPETITIVE) ?u - sub(a) - s l£up- u* - tfu’p ref stem - need - l£ point - epen - reduplication (184) a. ?usubas ha?uk I need to eat. (DURATIVE) ?u - sub(a) - s ha?u - uk ref stem - need - l£ eat - dur b. ?usubas ha7uk£ik I need to start eating. (INCEPTIVE) ?u - sub(a) - s ha?u - uk - M k ref stem - need - l£ eat - dur - mom. c. ?usubas ha?ukwa^p I need to feed someone. (CAUSATIVE) ?u - sub(a) - s ha?u - uk - axp ref - need - Is eat - dur - caus Sentence (181) and (182) are Makah sentences with one surface element; (183) and (184), respectively, show Makah sentences with the same semantic content as (181) and (182), expressed with two surface elements insteadof one. (The rules surrounding the incorpora tion process which produces the variants (181) and (183), as well as (182) and (184) are fully described in the

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 141

next chapter). In (181-182), sentences with one surface element, ASPECT is not adjacent to AUXj^J -sub(a) -

separates ASPECT from AUXjj. In (183-184), sentences with more than one surface element, ASPECT markers are not adjacent to the element in sentence position 1, and therefore, are not adjacent to the AUXjj in any way. The ASPECT markers in (183-184) are attached to the stems Ifap- and ha?u-, respectively, and occupy the third

surface position in the Sjj. Earlier, the analysis showed that AUXj^ always occurs in second position in the

SjjJ by that analysis, finding ASPECT markers in position 3 in a simple sentence automatically eliminates any reason for considering ASPECT to be an AUX component in Makah. The importance of this argument was anticipated in earlier studies of Nootkan languages. The discussion of suffix classes in the Sapir and Swadesh classic, Nootka Texts (1939), for example, excludes ASPECT from a group of markers which resemble AUX as instantiated in this analys is! Suffixes fall into two classes!! (1) formative suffixes, which are added to a theme (i.e. a stem or a derivative form) and which either make a derivative theme or indicate the aspect; (2) incremental suffixes, which are added only to formally complete words and which express notions of time, voice, mode, person, and number of subject and object, and other notions, (236) The definition of incremental suffixes in Nootkan

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 142

(Westcoast) closely resembles the clitic sequence AUX^j in content and order of marked elements, while aspectual inflections are clearly seen as comprising a distinct group. While no mention of the syntactic relations between these suffix groups appears in Sapir and Swadesh (1939) and (1955), this analysis submits that the use of the category AUX in other Nootkan languages may help dis ambiguate passages like the following Certain stem suffixes tend to be used only after otherwise complete words, usually of durative or momentaneous aspect, and in this respect approach the word suffixes, but their aspect inflection identifies them as stem suffixes CSwadesh 1939* 80). This passage, like the one which precedes it, talks about suffixes which are restricted to complete words and stems which are restricted to stems or a derivation of a stem. Formalizing and utilizing the AUX sequence in Makah language description allows for certain elements to be described in a more explicitly functional sense. The elements which comprise the AUX^ are necessary for grammatical sentence formation; in effect, the AUX^ sequence is what creates a sentence from a word or an inflected stem.'*'® It is impossible to have a sentence in Makah without an AUXj^ sequence attached to

^Demers and Jelinek propose a similar argument for Salish (Berners, Personal Communication;1986) .

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 143

the first element in the S^. This fact radically differentiates AUXj^ elements from other affixes in the language? no elements outside of those within AUX^ can create a grammatical sentence from a word or a stem in Makah. While further research will clarify the remaining classification of non-stem

elements in Makah, the use of an AUXjj construct allows one of these groups to be functionally described.

The Making of Discourse Topic in

One of the fundamental discriminators which

isolates AUXjj is the restriction which places a complete (one with subject marking) AUX^ in second position in Makah simple sentences. In complex sentences, while the underlying constraint is still the same, AUXj| may appear immediately to the left of any phonological or morphemic surface juncture, but without subject marking. As C185), (186) and (187) show, the location and the components which comprise AUX^ in a specific surface structure determine the clausal structure and conse quently, the assignation of meaning to the sentence. Recognizing AUX^ within complex sentences provides regularity to the process of assigning meaning to a sentence. (185a-b) illustrate this claim*

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 144

(185) a. ?usubqeys cfidi’liq. I might need a dog. ?u - sub(a) - qey - s cfidi** - °iq ref - need - conditional - l£ dog - art b. ?usubas tfidi'kqey I might need a dog. ?u * sub(a) -s tfidi.'k- qey ref- need - l£ dog - conditional While the generalized English translation of these two sentences seems equivalent, Makah speakers will trans late the two sentences differently, and will also provide directions which distinguish the context in which each sentence is most appropriate. (182a) exhibits the surface form -qeys, which indicates PRESENT TENSE, CONDITIONAL MODALITY, and first person singular . SUBJECT (rendered 0-qey-s morphemically). The sequence appears in the prescribed second surface position before the first surface juncture. This is different from the other sentence (185b) . Here the complete AUX^j contains DEFINITE PRESENT TENSE, INDICATIVE MODALITY^, 1 singular SUBJECT (- 0 - 0 - s) and occurs in second sentential position? but an additional AUXj^ clitic appears before the next surface juncture in the form of -qey-, indicating CONDITIONAL MODALITY. Speakers

translate (1 85a) to mean fI might need something? probably a dog.' is translated very differently? 'I might need something? it might be a dog.' In (185a), the conditionality of the sentence focuses on the need rather than on the object which is

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 145

needed. This sentence would be used in a situation like the following: two people are discussing the fact that one of the conversants spends a great deal of time alone in an isolated house. The other speaker indicates that it would be a good idea for the isolated individual to take steps to alleviate some of this isolation. The isolated party recognizes the concern of the first, but is not convinced that any measures should be taken, or need to be taken. Hence, the position of AUXj^ places the indefinite portion of the sentence on the need by attaching the MODAL clitic -qey- to the initial surface element. The opposite situation exists in (185b). Here, the speaker definitely recognizes the need; indefiniteness centers on what will fulfill the need. A speaker offered the following situation to exemplify (185b) . Two people are discussing the fact that person A needs a pet. The discussion centers around the best type of pet for person A. Cl85b) recognizes the fact that a pet is needed, but whether a dog will fit this need is not definitely established in the speaker's mind. Therefore,

the MODAL clitic of AUXjj is attached to cfidi *k 'dog', to identify this component as the indefinite element of the sentence. This analysis submits that the appearance of an AUX^j in any place other than to the immediate left of the

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 146

first surface juncture indicates as in the example, that the grammatical construction is a complex sentence. Variation in placement of the elements which constitute AlIXj^ in the surface structures will account for other semantic variations in Makah clauses. Consider these two SM '.

C186) a. I&ip5ifrsi*k dudu’tf^akqey tfup-Si*- s-i’k dudu'k- *afc-qey point - mom -Is -hab sing-temporal-cond proximity

xad^awiq. i point when the woman xad?ak-°iq sings. woman-art

b. l

xad7awiq T'Jhen I point, the woman xad?ak-°iq sings. woman-art

In (186a), AUXj^ contains g-g-s-i^k (PRESENT TENSE, INDICATIVE MODALITY:, Is SUBJECT and HABITUAL SPEAKER INTENTION) in the initial clause, and - *a*qey (IMMEDIATE TEMPORAL PROXIMITY., CONDITIONAL MODALITY) in the adjunct clause. The situation is essentially reversed in (186b),

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 147

except the SUBJECT MARKER remains consistently in the initial AUX^. The object of the adjunct clause is specified by the proximity of the xad7awiq 'woman’ to the adjunct AUXjj. In terms of AUX^ composition, the com bination of the clitics - *aft- and -qey- consistently correspond to the English translation of "when". Makah Constructions which correspond to "if-then" clauses in ENGLISH also exhibit similar clausal contrasts in AUX*. markers and combinations of AUX^ clitics.

(187) a. dudu-k ^e ’^isqey xad7awiq dudu*kw- *e,?is-qey xad7ak-°iq sing-ind. future -cond woman-art

da7a*7uX If the woman will da 7 a*?uXw- 'eyik-s s^-nS» ^ will listen. listen-def. future -Is

b, dudu’^e'^isqeys da7a,7uX"ey7i* dudu‘kw- ’e'Hs-qey-s da?a*7uXw- *e)fik-:0i sing -ind. future-cond -Is listen-def. future 3s

xad?awiq If I will sing, the xad7ak-°iq woman wil1 listen* woman-art

(188) a. dudu'k^qey xad7awiq da7a'7uXweyiks dudu'kw-qey xad7ak-°iq da7a ’7uXw- ’eyik-s sing-cond woman-art listen-def. future -Is If the woman sings, I will listen.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. If I sing, tiie woman will listen. b. dudu*kwqeys da?a*7uXwey7i^ xad7awiq dudu*kw-qey-s da^a’?uXw- *eyik-°i xad7ak-°iq sing-cond listen-def. future-3s woman-art

In order to appreciate the contrasts in AUX^ sequences in clausal constructions, recall the AUXj^ sequence in its entirety,

(189) AUXjj -* (INDEFINITE TENSE) (TEMPORAL PROXIMITY) (DEFINITE TENSE) (MODALITY) SUBJECT MARKER (OBJECT MARKER) (SPEAKER INTENTION) and notice that only a SUBJECT MARKER is necessary for the AUXj^i sequence to create a grammatical sentence. While the other elements within AUX^ indicate information about the definiteness/indefiniteness, time, and the like about a SM , it is the obligatory SUBJECT MARKER that makes AUX^ create a sentence. Without the obligatory SUBJECT MARKER;, the AUX^ sequence does not create a sentence when it is combined with a Makah word or stem. Rather, it creates an utterance which translates as a clause or an incomplete sentence in English. For example, (187a) contains two distinct AUX^ sequences. The first is attached to the stem dudu*kw-, and takes the surface form - *e,7isqey~, which lacks a SUBJECT MARKER. The second AUXj^ sequence has the sur face form - ’eyiks, which contains the Is marker this sequence is attached to the Makah stem da7a*7uXw-. The second AUX^ sequence, with the obligatory SUBJECT

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 149

MARKER, transforms the meaning of the stem da?a*?ukw - to the DEFINITE FUTURE meaning ’I will understand*; this is a complete sentence in Makah. On the other hand,

the first AUXj^j sequence which lacks the obligatory SUBJECT MARKER does not transform the stem dudu*kw- to an utterance which speakers consider to be a complete sentence. dudu*]fre*?isqey, without a SUBJECT MARKER, is not a complete sentence, and its meaning is trans lated as 'if someone sings* or 'should someone sing*. (In keeping with the second surface position constraint of AUX^, (187a) is also constructed as da?a *?u£weyiks dudu* ?isqey xad?awiq.)

(187b) presents a sentence with two AUX^ sequences that contain the obligatory SUBJECT MARKER. The first AUXjj sequence, - *e*?isqeys is attached to the stem dudu'kw-, thus translating as a complete sentence 'I might sing at some point*. But, in the context of another AUXjg sequence - ?*Oy?i* - (attached to the stem da?a*?u%*-) which also contains an obligatory SUBJECT MARKER, the translation becomes 'If I will sing' for the dudu*kKe*?isqeys construction. Comparison of (187a) and (187b) reveals an interest-* ing point about the formation of clause topic and the function of complete vs. incomplete AUXj| sequences. In (187a) the subject of the initial clause is xad?awiq,

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 150

while the subject of the initial clause in (187b) is Is. The subject of the clause in 187a) can be indicated by a word in an adjunct position to the clause, a situation which is impossible in (187b) . In this instance, the Is subject of the clause can be indicated only by a SUBJECT MARKER at the proper point in the AUX^ sequence; no independent element which indicates Is can be substituted for Is in the AUXj^ in the way that xad?awiq functions to indicate subject reference in (187a). In effect, the occurrence of two AUX^ sequence with obligatory SUBJECT MARKER is a consequence of the lack of independent reference to Is in the adjunct position to the clausal construction. (188a) corresponds to (187a), and (188b) corresponds to (187b). While the entire description of Makah clause morphology and syntax is beyond the scope of this dissertation, the functional role AUX^ plays in con structing clauses and marking the subject of the clause emerges. The composition of the AUX^ sequence figures prominently in the assignation of meaning to a clausal construction, and the specification of clausal subject results from the appearance of independent words in the adjunct position to the AUX^ sequence, with the exception of Is subjects. The use of AUX^ as a descriptive category in Makah permits the regularity in semantic and

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 151

syntactic renditions of yet another aspect of Makah grammar to emerge.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 7

DESCRIBING THE MAKAH SENTENCE! THE SUPERIORITY OF USING AN AUX ANALYSIS INSTEAD OF A NOUN/VERB FRAMEWORK

Previous chapters in this dissertation instantiated

AUXjj, described AUX^, and discussed the function of AUXjj in Makah grammar. This chapter illustrates another analytic use of the category. Makah, like other North west Coast languages, is often cited as an example of a language with "unusually weak differentiation of parts of speech, and, in particular, with no differentiation between nouns and verbs" (Jacobsen 1979a!83). This analysis agrees with that characterization, but takes the argument a step further! by this analysis, the distinction between noun and verb is irrelevant once a well-described AUX^ appears in the grammar. Since AUXj^ functions as the "anchor (of the sentence) around which the other sentential elements float" (Steele 1978a!39), its syntactic regularity provides a data-centered and objective nonphilosophical definition of the other syntactic elements in the Makah sentence. To support this claim, Chapter Seven first traces the question of noun-verb differentiation in Makah as 152

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 153

discussed in the literature, and concludes that the ambiguity is the result of English-based philosophical distinctions! The next section demonstrates how includ

ing AUXjj in the description of the Makah sentence erases the need for form classes which rely on criteria not related to sentence position. The chapter continues by presenting an analysis which relates syntactic variation to subcategorization rules which restrict the occurrence of Makah formsJ again demonstrating the extraneous nature of the noun-verb distinction for Makah.

Word Class Ambiguity in Makah The idea that Makah word classes are ambiguous and overlapping (in terms of the previous Jacobsen citation) stems from the supposition which equates the phrase "word classes" which the binary distinction noun and verb. When linguistic investigation primarily concern itself with Indo-European languages, little doubt existed as to the universal nature of the form classes noun and verb, or of the applicability of these form classes in the description of any language. The notion that a sentence could be described in philosophical terms, e.g. "something to talk about" and "something to be said about it" (Sapir 1921!119), dominated the primary view of the discipline.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 154

The validity of this notion began to be challenged when American Indian languages became a focus for American linguistic inquiry. These languages rely heavily on intricate systems of surface-level, cross referencing morphology in sentence formation; it was no longer easy to assign an individual word or other meaningful unit to a specialized form class based on a philosophical meaning or translation. In short, a noun could no longer be defined solely as a name word, and a verb could no longer exclusively be defined as an action or a state of being. As a result, some linguists began to question the universal applicability of the form classes entirely, for example, Bloomfield (1933120). Others continued to struggle with linguistic representations which mapped, more or less successfully, the form classes of noun and verb onto segments of linguistic utterances. Analytic problems resulted when the categories of noun and verb did not correspond to syntactic or semantic categories in a non-Indo-European language. The lack of correspon dence was often interpreted as a form-class ambiguity and the task of the analysis came to center on finding a means to distinguish nouns and verbs in these languages. For example, the now infamous "discovery process" arose once more linguists began to find that the designations noun and verb were irrelevant for many languages (Whorf 1956;

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 9-10, Sapir and Swadesh 1939:235-236), Swadesh 1939!78). However, according to Kinkade (1983126), the irrelevance of noun and verb as grammatical categories did not receive published attention as a singular issue until Kuipers (1968). This article questioned the need for a noun-verb distinction in the description of Squamish, a Salishan language. Numerous analyses have addressed and confirmed Kuipers* hypothesis for other Salishan languages (e.g. Kinkade 1983, Demers and Jelinek 1985). Similar treatments of noun-verb distinction in Makah are not, as of yet, available. The most comprehensive discussion of noun and verb as grammatical categories in Makah is found in Jacobsen (1979a). His analysis accepts the idea that noun and verb as form classes are relevant for Makah, and he distinguishes the two classes based on a mixture of English semantics and Makah morphology. For example, Jacobsen CL979a:ilO-lll) established a series of "undoubted verbs," nouns, and "apparent adjectives" as predicates in Makah. Examination of his corpus reveals that membership of a stem in a Makah form class has been determined by the form class of its English translation. Stems translated by Jacobsen to include an English infinitive are placed within the form class of "undoubted verb". Stems without a translated infinitive fall into the category of noun if the stem corresponds to an English noun, and into the category of

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 156

adjective if the stem corresponds to the English adjec tive. The problem with this scheme is discussed in Renker and Gill (1986). This article maintains that Makah stems must be translated as semantically neutral to eliminate the unnecessary verbal quality that the infinitival translation imparts to the stems (Renker and Gill 1986135). So Makah stems such as bi*-, we?i£-, and lftip- should not be translated as 'to rain', 'to sleep' and 'to point*, respectively, but as 'rain*, 'sleep', and 'point'. Once the semantic implications which accompany the English infinitive are removed from the translation, the stems bik-, we?i£-,and flip- become structurally equivalent to the stems Jacobsen (1979a)

designated as nouns! cfidi*k- 'dog*, ?ada*k> 'fire' and babafrid- 'white person'J and as adjectives! ?i,?i*%w 'big', ?akyiq- 'many* and ?at- 'thick'. All of these stems can be predicated, in spite of the form class of the English translation. Now, in fact, there is no distinction in Makah grammar between a "verb" to rain, and a "noun” rain. In Makah, the stem bift- refers to a generic concept of rain which carries no indication of events or substancesJ a Makah stem carries no noun or verb distinction, regardless of an English translation. Confusion over Jacobsen's form classes "noun and verb" also results when noun status is assigned to Makah

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 157

forms limited to inflection for durative aspect, and verb status to forms inflected for the momentaneous aspect (Jacobsen 1979aI114-115, 140-141). The following corpus illustrates the problem with this classification.

CHART 9.--Comparative Aspect STEM WITH DURATIVE ASPECT (190) bu*- 'burn' (193) bu#?ak 'state of burning' (191) caxw- 'roll, rotate* (194) caxwukw 'state of (192) jfatq- 'belongings' building* 'things to pack (195) ]fatqukw 'belong ings'

WITH MOMENTANEOUS ASPECT (196) bu*£i* 'begin to burn' (197) caxwsi* 'roll once' (198) jfatqti* 'pack to leave'

Stems (190-192) can combine with both the durative suffix and with “the momentaneous suffix. The resultant con structions (193-198) can all be predicated, while the stems (190-192) cannot be predicated without an aspectual inflection. The distinction between (190-192) and (193- 198) does not reside in form class designation, because the durative constructions and the momentaneous constructions do not differ in their syntactic context. Attributing noun and verb labels to (193-195) and (196-

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 158

198) respectively, again obscures the uniformity shared by all of the stems mentioned in the discussion to this point! each stem can be predicated, regardless of the form class to which the English translation belongs. In short, the distinction between noun and verb which applies to English words does not apply to Makah stems. Imposing the distinction on stems introduces an unnecessary set of semantic relations into the descrip tion of the Makah sentence. Moreover, because the concepts of noun and verb are not indigenous to Makah grammar, their use sets up classes within Makah structure which are ambiguous and overlapping. All this compli cates the description of the Makah sentence needlessly. However, there is an alternative to the terms noun and verb when describing Makah sentences. The category AUX^, established in Chapter 5 and described in Chapter 6, provides a syntactically predicatble constant around which other categories in the Makah sentence can now be described. The next sections discuss the use of AUX^ in Makah sentence representation, and demonstrate the superiority of this analysis to the traditional noun and verb description.

The Need for an AUX-Centered Analysis of Makah Sentences Previously, the Makah sentence was described as having a minimum of two positions, with surface bounded

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 159

elements (#n#) expanding to the right of these positions. In a sentence with one surface element, two positions, initial and final, exist; in a sentence with more than one surface element, there are a minimum of three posi tions, corresponding to initial and secondary in the first surface element, with additional positions numbered consecutively to the right with each new surface element.

AUXjj always occupies position 2^ in the deep struc ture of a simple sentence and in its surface structure, regardless of the stem occupying position 1. As just argued, stems are semantically neutral where grammatical

function is concerned, allowing a complete AUXjj the function of creating a sentence when it is attached to a stem. This restricts a distribution in word classes to those constructions which can combine with AUX^ and those which cannot. Only a few elements in Makaft can never appear in position 1 in a Makah sentence; these elements include interjections, certain demonstratives, and proper names * And, because AUX^ obligatorily follows position 1, these rare elements are not found in a syntactic context which requires AUX^ to be attached to these elements. So, these items must always accompany material which is grammatically acceptable in positions 1 and 2, respec-: tively, if they are to occur in a syntactic context.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 160

For the remainder of Makah stems and affixes, AUX^ attaches to the element in position 1 and formis a sentence by doing so. Here are some examples of Makah sentences with position 1 and position 2 occupied.

CHART 10.--AUX and. Positional Analysis - 2 Positions ENGLISH POSITION 1 POSITION 2 SENTENCE TRANSLATION

(199) Sapac- + -°i tfapacf 'It is a canoe* ’canoe' (200) ba?as- + -°i ba?as 'It is a house* ' house' o. (201) tfi*?ak + - X ^i'?aw 'It walks, he 'walk' walks' (202) ha?uk- + -°i ha?uw 'It eats, He 'eat' eats'

(203) cfidi* k- + -°i cfidi *1 'It is a dog' •dog' (204) babuyak- + -°i babuyaw 'It works, He 'work' works' (205) ta'la’- + -°i ta*le*? 'It is money.' 'money* (206) ?i’?i’Xw-- + -°i ?i.?i^Xw 'It is big'. •big' 0. (207) #isukw- + - 1 ??isuw 'It is white’. 'white' 0. (208) xad?ak- + - 1 xad?aw 'It is a woman' 'woman' In this corpus, AUXM is represented by -°i, the third person singular SUBJECT MARKER. In each of the stems (199-208), -°i combines with the position 1 stem to

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 161

produce a predication of the most basic sort. The resul tant Makah construction has nothing to do with the parts of speech accorded to the English translations of the component parts. (207), with a translated "adjective" in position 1 is identical in structure to (208), with a translated "noun" in that position. AUX^ can attach to any stem and produce a Makah sentence irrespective of the grammatical classes of noun and verb. The AUX-centered analysis works equally as well in describing the structure of Makah sentences with more than one surface element, and consequently, three posi tions for a simple sentence. In Chart 11, (209-211) illustrate the positional framework when three positions are occupied.

CHART 11.--AUX and Positional Analysis - 3 Positions Makah Sentence/ POSITION 1 POSITION 2 POSITION 3 English Translation (209) tfi,?ak -°i xad7ak -°iq & * 7aw xad7awiq "The woman walls." (210) 7i,7i*X" --°i ba7as -°iq 7i*7i*xw ba7asiq "The house is big (211) ba7as -°i 7i*7i*X" -°iq ba7as ?i‘?i*Xw "The big one is a house." Again, there is no need to refer to noun and verb when describing the structure of the Makah sentence; the positional framework provided by reference to the syntac tic category AUX accomplishes this end.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 162

Additional uses o£ the AUX-centered analysis are found in a corpus which illustrates two types of syntactic variation in Makah! C212) ?u ’3XuX cfa'sqSitf * Somebody knocks.f ?u ’sXuX - °i cfa'sq- Si* somebody - 3£ knock - mom C213) cfa'sqXil ’u'SXuX * Somebody knocks.’ cfa’sq - Si* - °i ’u ’SXuX knock - mom - 3£ somebody C214) ?u ’da’ks ta’la* ’I have money.’ ?u’- da’ks ta’la* ref. stem - have- Is money C215) ta’la’da'ks ’I have money.’ ta’la' - da*k - s money- have- Is Sentences in each pair indicate a specific point of varia tion in Makah. (212-213) show how two major surface elements may invert without an accompanying change in the semantic properties of the utterance. The structures in the other pair, (214-215), are related through an incorpo ration process which moves a unique surface element in (214) within the first surface element, resulting in (215); again, there is no change in the semantic properties of the utterance. Arranging (212-215) in terms of the positions under the AUX-centered analysis of sentence forms shows why sentences in each pair need to be treated as grammatical variants, not distinct syntactic sequences. Note that (212-215) correspond to sentences in Chart 12;

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 163

the sentences in the latter are designated by the addi tion of (a) to the number of the sentences (212-215).

CHART 12.--AUX and Positional Analysis - 3 Positions

POSITION 1 POSITION 2-AUX POSITION 3

C212a) ?u ’SXuX -°i c?a*s$sik somebody 3s knocks (312a) (fa’sqSik -°i ’u'SXu* knocks 3s somebody (214a) ?u ’da*k -s ta'la’ have Is money

C215a) ta*la'da’k -s 0 have money Is

AUXj functions as the grammatical discriminator in both pairs of sentences because of the unique positional boundary operating on the category. When variation in both sentences is expressed in terms of position and the AUXj| second positional constraint, the regularity under lying the variation types in the corpus become clear, regularity extends within and between the contrasting pairs, according to this analysis.

Makah Stems, Syntax, and AUX^

The positional categories proposed in the previous section allow any Makah sentences to be classified into one of two categories!

1. Group 1 Sentences (G1S)- sentences with only one surface element. These sentences are described as

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 164

S=X - AU&., where X =Position 1 and AUX„ = Position 2. An example is (215). 2. Group .2 Sentences (G2S)- sentences with numerous sur face elements, each surface element bounded by an open juncture. These sentences are described as X - AUX« - Yi* .Ytj, where X = Position 1, AUXM = Position 2, Y = Position 3, Y2 = Position 4, and so on until Yn = the terminal surface element to the right. Examples are (212-2143 . Examples of G2S with more than four posi tions, written as X - AUX^ -Y1 - Y2 are rare in the data, (2l2a-215a) demonstrate two possible variations in Makah syntax which are best explained using the sentence groupings and AUX^. One type of sentence variation occurs within the G2S group. In this type of sentence (212a- 213a) the elements X and Y, positions 1 and 3 respectively, are inverted. However, the position and structure of AUX^ remains unchanged. (214a-215a) demonstrate a second type of Makah sen tence variation. In this case, a G1S may alternately be constructed as a G2S. The Y in (214a) is incorporated into the X position, and the element 7u- is deleted. Now (215a) becomes a G1S construction because the Y (Position 3) is no longer present in the surface structure. Again, AUX^ remains unaffected by these changes; both types of vari ation are governed by subcategorization rules which specify the positions in sentence structure which a Makah stem can occupy.

Makah Stems and Sentence Variation In Makah, stems fall into one of three mutually exclu sive categories as determined by the pattern of a stem's

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 165

location in positions X or Y in Makah sentence structure: none of these steins ever appear in AUX^, which occupies position 2.

Cl) Stems in category A (Stem.) occur in position 1 (X) or position 3 (Y), with either an incorporate pre-stem element or a stem from category C preceding stem^. (2) Stems in category.B (Stemg) occur only in position 1. (3) Stems in category C (Stern^) occur in position 1 (X), position 3 (Y), or expanded posit such as 4 (y,), and can occur in complimentary distribution with pre-stem elements which precede Stem^. Examples of each sub-class of stems are now presented to exemplify the sub-categorization restrictions which accompany stem membership. To date, nine stems which satisfy the Stem^ criteria have been identified! (216) -ba’c? 'talking about' (221) -sub- 'need' (217) -c?ak- 'going to' (222) -t?ei- 'imitate* (218) -cfak- 'cook' (223) -i*ks- 'consume' (219) -da*k- 'possess* (224) -]Jai- 'smell' (220) -kwi’i- 'make'

The following examples illustrate the complimentary distribution of one pre-stem element (?u-) with Stem^ in the environment preceding Stern^! (225) ?u'ba*cfus ta'la* "I’m talking about money.” (225a) ta*la*ba*Ais (226) ?u ’c?aks di’ya* "I'm going to Neah Bay." (226a) di'ya'c^aks-

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 166

(227) 7u*(?aks kwiSu* "I’m cooking pork." (227a) kwiSu*

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 167

(234) wiiaqk- 'not want' (237) Cabui- 'be able' (235) hi’yu’- 'desist* (238) yubui- 'be unable' (236) wi ’ya’- 'never' (239) wiki’- 'not' Regardless of the type of sentence in which they appear stems like (234-239) must occupy.the initial surface position. (240-242) provide examples of Stem^ in G2S and (243-245) illustrate Stem^ in G1S. (240) wiiaqks kupSik. 'I don't want to point (briefly).’ (241) wi*ya*s kupSik^ 'I never point,' (242) tabuis kupSik. 'I am able to point.' (243) wi*ya*s. 'I never (do).’ (244) yubuis. * I can't (do something)^' (245) wiki * s. ' I'm not ( something') : * Stenig constructions are more commonly found in G2S forms. No variation of StemB is possible, even in Y expands past position 3. The final sub-class of Makah stems, Stem^,, is the most flexible sub-class within the positonal framework. Stem^ encompasses all stems not in Stem^ or Stem^, making this sub-class the largest one in Makah. (246-250) illustrate Stemg members (some are translated with aspect inflection, but remember that aspect is outside of the frame of reference of the classification system)I (246) babuyak- 'work' (247) ha’uk- ’eat' (248) kupMk- 'point (suddenly)'

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 168

(249) tfi/’ak- 'walk* (250) we?it- 'sleep* By adding AUXjf to these stems, G1S occur (246a-250a) . (246a) babuyaks 'I work' (247a) ha?uks 'I eat' C248a) tfupSiJts .'I point’ (249a) #i*?aks 'I walk' (250a) we?ifcs *1 sleep' G2S can occur when Stem^ (216-224) or Stem^ (234- 239) occupy XJ Stemc occupies Y in these instances. Stemg can also occur in both X and Y, not possible for either of the other stem groups. In this instance, the Stemg often functions as the subject of the Makah sentence. (246b) babuyaw xad?awiq. 'The woman works* (247b) ha?uw xad?awiq 'The woman eats' (248b) l&iplil xad^awiq 'The woman points’ C249b) #i*7 aw xad7awiq 'The woman walks' (250b) we7il? xad?awiq 'The woman sleeps' (xad7ak) 'woman' is a Stemg because it can occur in the incorporated position with a Stem^, as in xad?akleis 'I'm imitating a woman). One final set of positional constraints affects . variation in sentence form. This constraint does not impact StemB, which by definition must always reside in X. Instead, Stem^ and Stemg receive further differentiation by constraints operating on placement in Y. Since Y

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 169

expands to the right, the positional framework must account for restrictions affecting stems within Y. While both Stem^ and Stemg can occupy position 3 (Y), only Stemg can occupy Y expansions to position 4 and further right. This constraint is related to the incorporation phenomenon discussed in (225-233a). C251) wi'ya’s ta’la'ba'c* 'I never talk about money* (251)a. wi’ya's ?u ’ba*c? ta’la’ *1 never talk about money* *(251)b. wi’ya's ta*la* ?u*ba’

to contrast (251b) and (252b)\ thus demonstrating the syntactic contrainst explained in the previous paragraph. In this group of G2S, Stem^ can appear only in positions 3 of Y. If the construction expands further, only Stemg can move with the construction (251a, 252a). Ungrammatical sentences result when Stern^ is placed in position 4 in Y (251b, 252b).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 8

Al%: LINGUISTIC IMPLICATIONS

The discussion offered in this dissertation centered on the issue of AUX and the constituency of the category in Makah. After establishing that AUXj^ existed, the discussion proceded to describe the morphology of the category and its function in Makah grammar. Finally, in proposing that AUX played a critical part in the descrip- • tion of the Makah sentence, the dissertation used the category to describe Makah sentence structure without reference to the traditional categories of noun and verb. These points of interest, generally speaking, have linguistic consequences on two levelst the Makah language level and the Universal Language level. The remainder of this concluding chapter summarizes the relationship between these respective levels and the AUX analysis con ducted in the previous chapters. In addition, the conclusion offers a perspective which integrates AUX^ information with a larger linguistic sphere of inquiry.

AUX^ and the Makah Language

The primary consequence of the AUX^ discussion at the Makah language level relates to Makah syntax. The 170

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 171

structure of the clitic sequence AUX^, and the morphology of the sequence, had not been described in the literature until this study. While it had been mentioned that modal and pronominal endings in Makah occur on the end of the first word in a Makah sentence (Jacobsen 19791132), pre vious studies mentioned these endings without describing the structure or morphology of the entire clitic sequence. This dissertation isolated and described the clitic sequence in Makah, thus adding valuable information about the formation of Makah sentences; this is especially important because AUX^ is critical to the formation of grammatical Makah sentences. In addition, the analysis has shown that an AUX- centered description of the Makah sentence provided a means to explore the structure of certain types of clausal formations. Positing that the structure of AUX^ allows for the creation of either a clause or a complete sentence (depending on its placement and the clitics contained in the particular sequence) established a framework for further systematic study of the relation between clauses and sentences in Makah. The analysis also benefits the grammatical description of Makah. Because AUX^ is integral to the sentence formation process, and potentially clause derivation as well, the regularity of the category works to simplify Makah grammar. The regularity of AUX^ allows the analysis

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 172

to explain sentence variation as a function of sub-cate gorization rules which constrain some Makah stems to one position, or allow others flexibility within a sentence. Because AUX^ cannot, under any circumstances, move from its specified position in the Makah sentence, the AUX analysis provides a stable point from which variation can be assessed and described in a systematic representation. AUX^ analysis affects another aspect of Makah grammar. The analysis allows Makah grammar to be written without reference to the designations of noun and verb. The utility of this development arises from the elimination of the potential semantic baggage noun and verb carry when applied to form classes of some non-Western languages. Consequently, using AUX^ as the discriminator of form classes within Makah allows the data to be interpreted without a translational bias. But, these consequences relating to the Makah language level can be used to argue for the universal constituency of AUX, and again the position accepted by those who reject the idea of universal AUX constituency. In recalling the arguments proposed by the MV theorists, two points are immediately challenged by the AUX^j instantiation. Argument 1 against AUX constituency proposes that AUX is a subset of VP and not a unique syntactic category. The AUX^ instantiation throws suspicion on this perspective because the analysis has eliminated the need to refer to

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 173

a surface category verb in a grammatical description, or a verb-based constituent on this level as well. By virtue of this analysis, there are three unique syntactic cate gories to which a Makah grammar must refer; AUX^, X, and Y. There is no evidence to support the contention that either X or Y must be related to standard categories such as NP or VP, so it is not productive to argue that AUX^ is a subset of VP when VP does not appear to exist in Makah. The next argument against AUX constituency uses the grammatical simplification principal to discredit the utility of including an AUX in grammatical description. The AUX^ analysis demonstrates that Makah grammar is best represented by a three constituent grammar which includes AUXjj, so an argument which contends that the elimination of the category AUX simplifies a grammar is not valid for Makah. In fact, since the elimination of AUX appears to be dependant on the establishment of the subset relation ship between AUX and VP, the simplification argument loses credibility in a language like Makah which is not best described in terms of NP and VP.

The simplification argument does not apply to AUXjj for another reason. To date, the AUX^ analysis has provided the means to uncover and describe the regularity which underlies Makah grammar, thus allowing this disser tation to explain such previously puzzling phenomena such as sentence variation, the ambiguity of Makah form classes

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 174

and the apparent heterogeneity of semantic classifications like the evidential category. Without reference to AUX^, these language phenomena appear disparate and difficult to explain. Including AUX^ in a Makah grammar i£ the simplification; discounting the category complicates the description of Makah and obscures the rules which govern grammatical Makah constructions. The instantiation of AUX^ is relevant in discounting another of the arguments which object to AUX constituency; this argument asserts that it is the properties of the clitic elements, as discussed in Zwicky and Pullum (1983) and Zwicky (1985), which produce the non-definitional properties of the AUX sequence such as position and fixed

internal order. While the elements in AUXjj are easily classified as clitics, the properties of clitics do not account for the semantic and syntactic integrity of the entire AUX^ sequence, as demonstrated in Chapters 5, 6, and 7 by the syntactic constraints operating on the sequence. Clitic properties also do not explain the syntactic rules which describe the relationship between a complete AUXj| (a sequence which contains a subject marker) and sentence formation, and between an incomplete AUX^ (a sequence without a subject marker) and clause formation. In addition, the properties of clitics do not provide any evidence that the clitics themselves originate within another sentential constituent. By discounting AUXj| as a

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 175

sentential constituent based on the clitic property argu ment, there would need to be some evidence which pointed to the origin of the clitic elements in a constituent other than an AUX. In Makah, there is no evidence which points to the origin of the AUX^ sequence within another of the constituent categories in Makah, namely X or Y. To the contrary, the dissertation demonstrated that each category in Makah operated under unique syntactic con straints J there was no evidence to support the contention that AUX^ is somehow derived from either X or Y. To this point, this section has addressed three issues which are raised by theorists when discounting the applic ability of AUX as a sentential constituent. The next section relates AUX^ to the remaining arguments against the constituency of AUX when discussing the implications of this dissertation for universal language study.

AUX^ and Language Universals

In relation to the study of language universals, the data presented in this dissertation affects a number of concepts. Most visibly, this analysis presents data and a discussion which show noun and verb to be irrelevant in the description of Makah when AUX becomes the focus of the grammar. By extension, this conclusion implies that the universality of the form classes noun and verb is suspect because the form classes do not appear in Makah.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 176

Rather, the AUX^ analysis lends credence to the view point that noun and verb, along with other form class designations, are optional grammatical categories which reside in a universal inventory. While some languages may best be described in relation to form classes noun and verb, it is not a sufficient condition to expect all languages to include noun and verb. Noun and verb then, are better described as available, not mandatory universal classes. This optionality relating to form classes extends to the constituents of the universal sentence on a phrase structure level. If noun and verb are not mandatory grammatical categories for all languages, PS rules which include the constructs noun phrase (NP) and verb phrase CVP) may not be relevant for a language either. In the case of Makah, the dissertation has shown that sentences can be structurally represented in terms of X, Y and AUX^ rather than a semantically biased NP and VP. Consequent ly, the sentence representations proposed in this dissertation may be applicable to other languages with a doubtful noun-verb form class designation and therefore a doubtful NP and VP structural representation. And, relating to the constituency of AUX, the dissertation has demonstrated that AUX is indeed a consti tuent category in Makah grammar. Consequently, the data support the contention that AUX cannot be eliminated from

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 177

the universal inventory of constituent categories. Like NP and VP, AUX is, at^ least, an available constituent category in universal grammar, because it may be analyci- cally viable in some languages and not others. On the other hand, further research with the methodology of Steele et al. (1981), could establish that AUX is the single mandatory category in universal grammar. This last statement brings the discussion in the chapter to the final two arguments which are used to reject the constituency of AUX on the universal language level. One argument interprets AUX to contain too many options relating to the notional categories marked in the sequence from language to language! subject marking, subject agreement, aspect, question marking, emphasis, evidential, object marking, object agreement and negation. The AUX^ analysis submits that the notional categories marked in the Makah sequence are all connected by a semantic thread which revolves around the indication of the definite or indefinite in a Makah sentence or clause. This conclusion follows the conclusion in Steele et al. (1981) which discusses the function of AUX in relation to certainty or uncertainty in a sentence; each of the numerous categories which can be present in a language- specific AUX sequence contributes information that aids in the establishment of the truth value of an utterance.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 178

The final argument against AUX constituency also questions the logic which links notional categories across the boundaries which separate the natural languages of the world. This argument asserts that individual linguistic and cultural perceptions surround criteria like tense and modality, in addition to the other notional categories mentioned aboveJ there cannot be equivalence between individual AUX sequences in individual languages. Again, this dissertation implies that a distinction between definiteness/indefiniteness, certainty/ uncertainty, and the like is a critical part of any sentence in any language. If the AUX sequence is viewed as the ''purveyor of the truth value" of a sentence, there is room for languages to include specific notional categories which impact on the cultural value of certainty, and the informal tion which is needed in order to adequately judge or conyey that certainty, in individual languages. The notional categories in question do not appear unconnected in this respect. The AUX^ analysis added valuable insight here, because the data supported the need to further examination of the link between AUX and its role as the "truth indicator" in a sentence. An additional implication of the dissertation research concerns the Makah Language Program on the Makah reserva tion. Till this point, curriculum designed to teach syntax was difficult to develop because there was no

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 179

information which could account for Makah sentence vari ation, for the construction of clauses in Makah, and for the structure and ordering of the elements which created a Makah sentence. The analysis in this dissertation provides a foundation from which Makah syntactic curriculum can be developed. Because the analysis provides a syste matic means to evaluate the AUX^ sequence, curriculum can now be generated to convey the critical points which surround the fixed order and internal composition of the sequence, and therefore, the structure of sentences and clauses in the language. In conclusion, the analysis offered in this disserta tion has addressed a problem which had implications on both the single language level and the universal language level. The instantiation of AUX^ in this document has provided a means to simplify and clarify Makah grammar, and to enhance the knowledge about the constituents of the universal sentence.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX 1

MAKAH ORTHOGRAPHIC SYSTEM

la. Consonants

Voiceleas stops Voiced Voiceless Voiced Voiced and silicates stops fricatives rseoaanta/ nasals* semivowels plain gloUalised b ilab ial P * b m dental t i d n sibilant e I S { t S y la te ra l A A’ 1 I velar k i X rounded L« velar i * x w w postvelar q 4 A rounded poetvelar q * 4 - *• g lo tta l ? h a t» .______. a . * Very m In Makah, limited to loan words.

lb . Vowels

'f r o n t Central Back Back unrounded unrounded unrounded high i* U' i U mid a o- e o low e* a*

1c. Vowel - semivowel combinations aw as in njjw ay as in hike ey as in cake iw as in fgw iy no English equivalent, similar to geee! oy as in bov

180

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX 2

MAKAH MORPHOPHONEMES

°indicates that the following changes take place in the sound preceeding the °. p h t d A - 1

k/kw + v

q *4 qw -*■ cfw

c -*■ d

c ♦ e s, s, x, xw, x, xw, and 1 remain the same, vowel -*• ’ vowel

With all remaining consonants, 0 indicates the addition of a ’ after the preceeding sound. * indicates that the following sound changes take place in the sound preceeding the *.

p * kw -*• t -► tf vowel -*■ ’vowel

q -*• c -*■ d q w -*■ cf" Z + Z k ♦ tf A •+ A

181

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX 3

ABBREVIATIONS

Ind £ut Indefinite future Is First person singular De£ fut Definite future dur durative aspect mom momentaneous aspect TP temporal proximity 3s Third person singular lsb/2obj First person subject acting on a second person object Conf int Confirmational interrogative epen epenthethic cont continuous aspect red reduplication rep repetitive aspect ref stem referential stem art article cs clitic sequence

CP clitic pronoun hab habitual

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. BIBLIOGRAPHY

Akmajian, Adrian and Frank Heny 1980 An Introduction to the Principles of Trans formational Syntax. Cambridge! MIT Press. Akmajian, Adrian, Susan Steele and Thomas Wasow 1979 The Category AUX in Universal Grammar. Linguistic Inquiry 10!l-64. Akmajian, Adrian and Thomas Wasow 1975 The Constituent Structure of VP and AUX and the Position of the Verb BE. Linguistic Analysis 11205- 245. Arima, Eugene 1983 The West Coast (Nootka) People. British Columbia Provincial Museum Special Publication No. 6. Bloomfield, Leonard 1933 Language. New York! Henry Holt and Company. Boas, Franz 1911 Introduction to the Handbook, of American Indian Languages. Bureau of American Ethnology 4011-83. Bolinger, Dwight 1975 Aspects of Language, 2nd edition. New York! Harcourt, Brace and Jovanovich. Campbell, Lyle and Marianne Mithun, eds. 1979 The Languages of Native America! Historical and and Comparative Assessment. Austin! University of Texas Press. Chomsky, Noam 1957 Syntactic Structures. The Hague! Mouton. Chomsky, Noam 1965 Aspects of the Theory of Syntax. Cambridge! MIT Press. Cohen, Felix 1982! Felix S. Cohen's Handbook of Federal Indian Law. Charlottesville, VA! Michie Bobs-Merrill.

183

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 184

Cole, Douglas 1985 Captured Heritage! The Scramble for Northwest Coast Artifacts. Seattle! University of Washington Press. Colfax, Gregory 1982 Introduction to the Ninth Printing of The Indians of Cape Flattery by James G. Swan. Seattle! Shorey Publications. Colson, Elizabeth 1953 The Makah Indians! A Study of an Indian Tribe in Modern American Society. Minneapolis! University of Minnesota Press. Crystal, David 1980 A First Dictionary of Linguistics and Phonetics. Boulder, CO! Westview Press. Daugherty, Richard and Janet Friedman 1976 An Introduction to Ozette Art in Indian Art Traditions of the Northwest Coast. Roy Carlson, ed. pp. 183-195. Burnaby, British Columbia! Archeological Press. Daugherty, Richard D. and Roald Fryxell 1967 Archeological, Geochronological, and Ecological Investigations of the Ozette Village Site Complex on the Northwest Coast of Washington! II. Manuscript on file, Department of Anthropology, Washington State University. Demers, Richard and Eloise Jelinek 1985 Constraints in Arguments in Lummi in Proceed ings Languages, pp. 169-188. Vancouver! University of British Columbia. Densmore, Frances 1939 Nootka and Quileute Music. Bureau of American Ethnology Bulletin 124. Doig, Ivan 1980 Winter Brothers! A Season at the Edge of America. New York! Harcourt, Brace and Jovanovich Driver, Harold E. 1969 Indians of North America, 2nd edition. Chicago! University of Chicago Press.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 185

Drucker, Phillip 1951 The Northern and Central Nootkan Tribes. Bureau of American Ethnology Bulletin 144. Drucker, Phillip 1955 Indians of the Northwest Coast. Garden City, NY! Natural History Press. Emonds, J. 1970 A Transformational Approach to English Syntax. New York! Academic Press. Ernst, Alice Henson 1952 The Wolf Ritual of the Northwest Coast. Eugene! University of Oregon Press. Farb, Peter 1980 The Face of North America. New York! Harper and Row. Friedman, Janet 1975 The Prehistoric Uses of Wood at the Ozette Archeological Site. Unpublished Ph.D. dissertation, Department of Anthropology, Washington State University, Pullman. Gallatin, Albert 1836 A Synopsis of the Indian Tribes within the United States East of the Rocky Mountains and in the British and Russian Possessions of North America. Transactions of the American Antiquarian Society 2! 305-306. Gibbs, George 1855 Transcript of Journal of Proceedings, Treaty of Neah Bay. Manuscript filed in Letters Received, Bureau of Indian Affairs, Treaty File, Washington W537. Gibbs, George 1877 Tribes of Western Washington and Northwestern Oregon. Contributions of North American Ethnology 1! 157-361. Gill, Steven J. 1983 The Ethnobotany of the Makah and Ozette Peoples. Unpublished Ph.D. dissertation, Department of Botany and Anthropology, Washington State University, Pullman.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 186

Gill, Steven J. and Ann M. Renker 1984 How the Makah Name Plants. Paper presented at the 7th Ethnobiology Conference. Manuscript on file in Makah Archives, Neah Bay, WA. Gill, Steven J. and Ann M. Renker 1985 Makah Botanical Nomenclature--An Analysis of the 20th Salish and Neighboring Languages Conference, pp. 85-97. Vancouver! University of British Columbia. Gill, Steven J. and Ann M. Renker 1987 An Overview of Makah! From Phoneme to Text. Amerindia. Gillis, Alix 1974 A History of the Neah Bay Agency. Coast Salish and Western Washington Indians 3!93-115. Gleeson, Paul 1980 Ozette Woodworking Technology. Unpublished Ph.D. dissertation, Department of Anthropology, Washington State University, Pullman. Gleeson, Paul et al. 1981 Ozette Project Research Design. Manuscript on file, Laboratory of Archeology and History, Washington State University. Goodman, Linda Jean 1978 This is My Song! The Role of Song as Symbol in Makah Life. Unpublished Ph.D. dissertation, Depart ment of Anthropology. Washington State University, Pullman. Gunther, Erna 1936 A Preliminary Report on the Zoological presented to Alfred Droeber. R. Lowie, ed, pp. 105-118. Berkeley! University of California Press. Gunther, Erna 1942 Reminiscences of a Whaler's Wife. Pacific Northwest Quarterly 3!65-69. Gunther, Erna 1945 Ethnobotany of Western Washington. University of Washington Publications in Anthropology 10(1) . Gunther, Erna I960 Makah Marriage Patterns and Population Stability. Internationale Ainerikanistenkbngress 34! 538-545.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 187

Gunther, Erna 1973 Indian Life on the Northwest Coast of North America CAs Seen by the Early Explorers and Fur Traders during the Last Decades of the Eighteenth Century). Chicago! University of Chicago Press. Haas, Mary 1969 Internal Reconstruction of the Nootka-Nitinaht Pronominal Suffixes. IJAL 35!108-124. Haas, Mary 1972 The Structure of Stems and Roots in Nootka- Nitinaht. IJAL 38!83-92. Hall, Carl 1983 Makah Indian Reservation Statistical Information Guide. Manuscript on file in the Makah Tribal Council Planning Office, Neah Bay, FA. Hancock, Samuel 1927 The Narrative of Samual Hancock, 1845-1860. New York! Robert McBride and Company. Huelsbeck, David 1983 Mammals and Fish in the Subsistence Economy of Ozette. Unpublished Ph.D. dissertation, Department of Anthropology, Washington State University, Pullman. Huelsbeck, David 1983a The Utilization of Whales at Ozette. Manuscript on file at the Makah Cultural and Research Center, Neah Bay, WA. Jacobsen, William, H. 1968 Traces of Glottalized Resonants in Makah. Paper presented at the 43rd Meeting of the Linguistic Society of America. Manuscript on file in British Columbia Provincial Museum, Victoria, B.C. Jacobsen, William H. 1969 Origin of Nootka Pharyngeals. IJal 35!125-153. Jacobsen, William H. 1969a Labialization in Nootkan. Paper presented at the 4th Salish and Neighboring Languages Conference. Manuscript on file in the British Columbia Provin cial Museum, Victoria, B.C.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 188

Jacobsen, William H. 1971 Makah Vowel Insertion and Loss. Paper presented at the 6th Salish and Neighboring Languages Con ference. Manuscript on file in the British Columbia Provincial Museum, Victoria, B.C. Jacobsen, William H. 1973 The Patterns of Makah Pronouns. Paper presented at the 8th Salish and Neighboring Languages Con ference. Manuscript on file in the British Columbia Provincial Museum, Victoria, B.C. Jacobsen, William H. 1978 Makah Vocative Vocalism. Paper presented at the 13th Salish and Neighboring Languages Conference. Manuscript on file in the British Columbia Provincial Museum, Victoria, B.C. Jacobsen, William H. 1979 First Lessons in Makah. Forks, WA! Olympic Graphic Arts. Jacobsen, William H. 1979a Noun and Verb in Nootkan in The Victoria Conference on Northwestern Languages. Barbara S. Efrat, ed. pp. 83-113. British Columbia Provincial Museum Heritage Record No. 4. Jacobsen, William H. 1979b Wakashan Comparative Studies in Languages of Native America. Chafe and Mithun, eds. pp. 766-791. Austin! University of Texas Press. Jacobsen, William H. 1980 Metaphors in Makah Neologisms. In Proceedings of the 6th Annual Meeting of the Berkeley Linguistics Society. Bruce Caron, ed. pp. 166-179. Berkeley! Berkeley Linguistic Society. Jacobsen, William H. 1982 Heterogeneity of Evidentials in Makah Structure in Evidentiality! The Linguistic Coding of Episem- ology. Chafe, ed. pp. 1-41. Norwood,N.J.! Ablex. ' Kaisse, Ellen 1984 Review of An Encyclopedia of AUX, by Susan Steele et al. Language 60!924-928. ! Kinkade, Dale 1983 Salish Evidence Against the Universality of "Noun’ and 'Verb'. Lingua 60!25-40.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 189

Kirk, Ruth 1974 Hunters of the Whale. New York! Harcourt, Brace and Jovanovich. Kortland, F.H.H. 1975 Tones in Wakashan, Linguistics 146131-34. Kuipers, Aert H. 1968 The Categories Verb-Noun and Transitive- Intransitive in English and Squamish. Lingua 211610- 626. LaPointe, Steven G. 1980 A Note oh Akmajian, Steele and Wasow's Treatment of Certain Verb Complement Types. Linguistic Inquiry ll!770-786. Leap, William and Paul Cissna 1984 Evaluation of the 1983-1984 Makah Language Program. Manuscript on file in the Makah Archives, Neah Bay, WA. Leap, William and Paul Cissna 1985 Evaluation of the 1984-1985 Makah Language Program. Manuscript on file in the Makah Archives, Neah Bay, WA. Lincoln, Neville and John Rath 1986 Phonology, Dictionary and Listing of Roots and Lexical Derivatives of the Haisla Language of Kitlope and Kitimaat, B.C. Canadian Ethnology Service Paper No. 103. Lyons, John 1968 Introduction to Theoretical Linguistics. Cambridge! Cambridge University Press. Makah Cultural and Research Center 1979 Museum Exhibit Leaflet. Tacoma!iWestern Media. Makah Cultural and Research Center 1981 Proposal to the Department of Education, Office of Indian Education. Manuscript on file in Makah Archives, Neah Bay, WA. Makah Cultural and Research Center 1986 Proposal to the Department of Education, Office of Indian Education. Manuscript on file in Makah Archives, Neah Bay, WA.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 190

Makah Language Program 1979 Makah Alphabet Book. Forks, WA! Olympic Graphic Arts. Makah Language Program 1981-1986 Makah Language Curriculum. Neah Bay, WA! Makah Cultural and Research Center. Makah Language Program 1982 Makah Counting Book. Forks, WA! Olympic Graphic Arts, Makah Language Program 1982a Q '"e'tl* Coloring Book. Forks, WA! Olympic Graphic Arts. Makah Language Program 1983 Raven and Crow. Forks, WA! Olympic Graphic Arts. Mauger, Jeff 1978 Shed Roof Houses at the Ozette Archeological Site. Unpublished Ph.D. dissertation, Department of Anthropology, Washington State University, Pullman. McCurdy, James 1981 Indian Days at Neah Bay. 2nd edition. Seattle! Seattle Historical Society. Miller, Beatrice 1952 The Makah in Transition. Pacific Northwest Quarterly 43!262-272. Miller, Robert 1952 Situation and Sequence in the Study of Folklore. Journal of American Folklore 65!29-48. Pethick, Derek 1980 The Nootka Connection! Europe and the Northwest Coast, 1790-1795. Vancouver! Douglas and McIntyre Limited. Phillips, Earl and Wallace Donaldson 1972 Weather and Climate for the Clallam, Gray's Harbor, Jefferson, Pacific and Wahkiakum Counties. #013708, Cooperative Extension Service, College of Agriculture, Washington State University. Manuscript on file in College of Agriculture, Washington State University, Pullman.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 191

Powell, J.W. 1851 Indian Linguistic Families of America North of Mexico. Bureau of American Ethnology Bulletin 7. Pullum, Geoffrey 1981 Evidence Against the "AUX" Node in Luiseno and English. Linguistic Inquiry 121435-463. Pullum, Geoffrey and Diedre Wilson 1977 Autonomous Syntax and the Analysis of Auxi liaries. Language 531741-788. Rath, John C. 1974 On the Phonological Description of the Heiltsuk Language in Dutch Contributions to the 9th Conference on Salish and Neighboring Languages, pp. 39-75. Leiden. Renker, Ann M. 1980 Makah Language Survey. Neah Bay, WA! Makah Cultural and Research Center. Renker, Ann M. 1980-1985 Field Notes from Neah Bay. Renker, Ann M. 1985 Makah Language Survey II. Neah Bay, WA1 Makah Cultural and Research Center. Renker, Ann M. 1985a Sentence Inversion in Makah!. Evidence for an AUX? Paper presented at the 38th Northwest Anthro pology Conference, Ellensburg, WA. Renker, Ann M. and Greig Arnold In press The Makah Cultural and Research Center! Education and Cultural Resource Management. William Leap, ed. Renker, Ann M. and Steven J. Gill 1985 The Salient Features of Makah Ethnozoological Nomenclature in Proceedings of the 20th Salish and Neighboring Languages Conference, pp. 325-337. Vancouver! University of British Columbia. Renker, Ann M. and Steven J. Gill 1986 Some Thoughts on Translating Makah to English in Proceedings of the 21st Salish and Neighboring Languages Conference, pp. 35-37. Seattle! University of Washington.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 192

Renker, Ann M. and Erna Gunther In press The Makah. To appear in The Handbook of North American Indians, Northwest Coast Volume. Washington, D.C.l Smithsonian Institution. Riley, Carroll 1968 Makah Political and Social Organization. Ethnohistory 15157-95. Ross, John 1967 Auxiliaries as Main Verbs. Studies in Philoso phical Linguistics, Series 1. W. Todd, ed. Evanston, 111.* Great Expectation Press. Sapir, Edward 1911 Some Aspects of Nootka Language and Culture. American Anthropologist 13115-28. Sapir, Edward 1913 A Girl's Puberty Ceremony among the Nootka Indians, Proceedings and Transactions of the Royal Society of Canada 2167-80. Sapir, Edward 1915 Abnormal Types of Speech in Nootka. Ottawal Government Printing Bureau. Sapir, Edward 1929 Central and North American Languages. Encyclo pedia Britannica, 14th'ed, 51138-141. Sapir, Edward and Morris Swadesh 1939 Nootka Texts. New York! AMS Press. Sapir, Edward and Morris Swadesh 1955 Native Accounts of Nootka Ethnography. New York! AMS Press. Steele, Susan 1978 Word Order Variation! A Typological Study in Universals of Human Language, J. Greenberg, C. Ferguson, and Edith Moravicsik, eds. Vol IV, pp. 587- 623. Stanford! Stanford University Press. Steele, Susan 1978a The Category AUX as a Language Universal in Universals of Human Language, J. Greenberg, C. Ferguson, and Edith MoravcsiK, eds. Vol IV, pp. 9-45. Stanford! Stanford University Press.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 193

Steele, Susan et al. 1981 An Encyclopedia of AUX. Cambridge: MIT Press. Swadesh, Morris 1939 Nootka Internal Syntax. International Journal of American Linguistics 9177-102. Swadesh, Morris 1953 Mosan I! A Problem of Remote Common Origin. IJAL 19!26-44. Swadesh, Morris 1962 Linguistic Relations Across the Bering Strait. American Anthropologist 64:1262-1291, Swan, James G. 1869 The Indians of Cape Flattery, at the Entrance of the Strait of Juan de Fuca, Washington Territory. Smithsonian Contributions to Knowledge 16. Washington, D.C.: Smithsonian Institution. Taylor, Herbert C. 1974 Anthropological Investigation of the Makah Indians. Coast Salish and Western Washington Indians 3:27-89. Thomas, John 1977 Nootkan Band Names. Paper presented at the 13th Salish and Neighboring Languages Conference. Manu script on file at the British Columbia Provincial Museum, Victoria, British Columbia. Thomas, John and Barry Carson 1979 The Nitinaht Inferential. Anthropological Linguistics 21:317-327. Thomas, John and Thomas Hess 1978 Introduction to Nitinaht Language and Culture. Manuscript on file at the Department of Linguistics, University of Victoria, Victoria, British Columbia. Touchie, Bernice 1977 Stealing Daylight. International Journal of American Linguistics, Native American Text Series, 2C3) *.69-67. Trager, George L. 1972 Language and Languages. San Francisco’ Chandler Publishing Co.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 194

Turner, Nancy et al. 1983 Ethnobotany of the Nitinaht Indians of Vancouver Island. British Columbia Provincial Museum Occasional Papers Series, No. 24. Voegelin, C.F. and F.M. Voegelin 1965 Classification of American Indian Languages. Anthropological Linguistics 7*121-150. Wagner, H.R. 1933 Spanish Explorations in the Strait of Juan de Fuca. Santa Ana. Waterman, Thomas 1920 The Whaling Equipment of the Makah Indians. University of Washington Publications in Anthropology 1C2). Seattle! University of Washington Press. Wessen, Gary 1982 Shell Midden as Cultural Deposits! A Case Study from Ozette. Unpublished Ph.D. dissertation, Department of Anthropology, Washington State Univer sity, Pullman. Whitner, Robert 1977 Culture Confict in the Agency School! An Introduction to a Case Study, The Neah Bay Reser vation, 1861-1896. Manuscript on file in Makah Archives, Neah Bay, WA. Whitner, Robert 1981 Makah Commercial Sealing 186?-1897! A Study in Acculturation and Conflict. Manuscript on file in Makah Archives, Neah Bay, WA. Whorf, Benjamin 1956 The Relation of Habitual Thought to Language in Language, Thought and Reality! Selected Writings of Benjamin Lee Whorf. John Carroll, ed. pp. 134-159. Cambridge! MIT Press. Wike, Joyce 1958 Social Stratification among the Nootka. Ethnohistory 51219-241, Zwicky, Arnold 1977 On Clitics. Bloomington! Indiana University Linguistics Club. Zwicky, Arnold 1985 Clitics and Particles. Language 611283-305.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Zwicky, Arnold and Geoffrey Pullum 1983 Cliticization vs. Inflection! English n't. Language 59!502-513.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.