Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Plant taxonomic systems and ethnobotany of three contemporary Indian groups of the Pacific Northwest.. Turner, Nancy Jean 1973

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
UBC_1973_A1 T875.pdf [ 56.03MB ]
Metadata
JSON: 1.0107164.json
JSON-LD: 1.0107164+ld.json
RDF/XML (Pretty): 1.0107164.xml
RDF/JSON: 1.0107164+rdf.json
Turtle: 1.0107164+rdf-turtle.txt
N-Triples: 1.0107164+rdf-ntriples.txt
Original Record: 1.0107164 +original-record.json
Full Text
1.0107164.txt
Citation
1.0107164.ris

Full Text

PLANT TAXONOMIC SYSTEMS AND ETHNOBOTANY OF THREE CONTEMPORARY INDIAN GROUPS OF THE PACIFIC NORTHWEST (HAIDA, BELLA COOLA, AND LILLOOET) by . Nancy Jean Turner B.Sc., University of Victoria, 1969 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS 'FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in the Department of Botany We accept this thesis as conforming to the required standard The University of British Columbia September, 1973 In presenting this thesis In partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of pfiJANY - - The University of British Columbia Vancouver 8, Canada •September 14, 1973 - i - ABSTRACT Plant names in three Pacific Northwest Indian languages — Haida (Skidegate and Masset dialects), Bella Coola, and Lillooet (Fraser River "dialect") — were analyzed semantically and taxonomically. A computerized sorting system was developed to handle pertinent infor- mation associated with these names and their corresponding plant types. At the present time, each language contains an average of about 150 generic-level plant names, over 50% of which correspond in a one- to-one fashion with botanical species. Some of the names have no mean- ing other than as plant names, but most are analyzable into smaller units of meaning, reflecting traditional beliefs, utilization, innate characteristics of the plants, or their resemblance to some substance, object, or other plant. Some of the generic terms are obviously bor- rowed from other languages, and a number of taxa can be found in each language which originally applied to indigenous species and have been expanded in recent times to include cultivated or imported counter- parts. Each language contains a few general "life-form" plant names, a number of intermediate taxa — usually unnamed, and in Haida. and Lil- looet, a few specific-level terms. None of the groups has an all- inclusive word for "plant". There are also several specialized gen- eric-level terms in each language, and many general names for parts of plants. Cultural significance of plants correlates positively with the degree of specificity of names applied to them, with the number of specialized terms associated with them, and with the lexical retention of their names in diverging dialects. Linguistic origin, floristic diversity, cultural traits, inter-group contact, and especially the recent acculturation of native peoples into "white" society, are be- lieved to be major factors influencing the character of phytotaxonomic systems of the three study groups. Maps of the study areas are provided, and appendixes are included listing all plant names used in the study, their botanical correspon- dence, and the utilization and cultural significance of the plants in- volved . - ill - CONTENTS Page ABSTRACT 1 ACKNOWLEDGEMENTS ix INTRODUCTORY QUOTATION xi PREFACE xii PLATES xiv INTRODUCTION 1 BACKGROUND TO THE STUDY 7 The Haida Indians 7 The Bella Coola Indians 12 The Lillooet Indians 16 METHODOLOGY 22 Ethnoscience and Componential Analysis 22 Ethnoscience with Respect to the Present Research Program 26 Research Procedures Used in the Study 35 a) Literature research 35 b) Field vegetation surveys 36 c) Consultation with native informants 38 d) Synthesis of data 43 RESULTS 61 Phytotaxa of the Haida, Bella Coola, and Lillooet Indians 61 a) Unique beginner 62 b) Major life-form categories 68 i) Haida life-form categories 72 ii) Bella Coola life-form categories 78 Hi) Lillooet life-form categories 83 - iv - c) Intermediate categories 88 d) Generic categories 102 e) Specific and varietal categories 136 Non-taxonomic Botanical Terminology 147 Synonymy in Nomenclatural Systems 149 Cultural Dimensions of Folk Taxonomic Systems 155 DISCUSSION 171 Historical Development of Folk Phytotaxonomic Systems 171 External Factors Influencing Ethnophytotaxonomies 185 Modern Botanical Taxonomy versus Folk Taxonomic Systems 189 SUMMARY AND CONCLUSIONS 201 GLOSSARY 208 BIBLIOGRAPHY 211 Appendix 1. Native People Contributing Information to the Study, 219 Appendix 2. Practical Orthographic Symbols for the Haida Language (Masset and Skidegate dialects). 229 Appendix 3. Practical Orthographic Symbols for the Bella Coola Language. 233 Appendix 4. Practical Orthographic Symbols for Fraser River Lillooet (Upper Lillooet dialect). 235 Appendix 5. An Alphabetical Listing of Folk Segregates for Plants in Skidegate Haida. 237 Appendix 6. An Alphabetical Listing of Folk Segregates for Plants in Masset Haida. 3ig Appendix 7. An Alphabetical Listing of Folk Segregates for Plants in Bella Coola. 420 Appendix 8, An Alphabetical Listing of Folk Segregates for Plants in Fraser River Lillooet. 497 Appendix 9. General Botanical Terminology in the Skidegate Dialect of Haida. 552 - V - Appendix 10. General Botanical Terminology in the Masset Dialect of Haida. 553 Appendix 11. General Botanical Terminology in the Bella Goola Language. 554 Appendix 12. General Botanical Terminology in the Fraser River Dialect of Lillooet. 556 Appendix 13. Index of Common Names of Plant Species Included in the Study. 5 5 8 - vi - TABLES Table 1. Card design for information directly related to folk segregates for plants. 45 Table 2. Card design for information relating to botanical taxa delimited by the folk segregates. 50 Table 3. Card design for information related to the utilization of plants by Pacific Northwest Indians. 53 Table 4. Keys to codes used in computor sorting system. 57 Table 5. Examples of plant names in Haida containing the life- form markers, xil, lhk'aayii (S) / lhk'aay (M), and tlaas or lhk'amaal 12 (M). 69 Table 6. Haida life-form categories. 79 Table 7. Bella Coola life-form categories. 84 Table 8. Fraser River Lillooet life-form categories. 89 Table 9. Some examples of intermediate taxonomic categories for plants in Haida. 93 Table 10. Some examples of intermediate taxonomic categories for plants in Bella Coola. 96 Table 11. Some examples of intermediate taxonomic categories for plants in Fraser River Lillooet. 99 Table 12. Degree of correspondence of Haida, Bella Coola, and Lillooet plant segregates with botanical species. 104 Table 13. Selected examples of the five species correspondence categories delimited in Table 12. 105 Table 14. Examples of plant taxa originally involving indigenous species, and expanded in historic times to include imported or cultivated counterparts. 109 Table 15. Examples of generic plant names known to have been borrowed from other languages. 112 Table 16. Some examples of unique generic plant names and segments of names. 116 Table 17. Examples of generic plant terms originating in post- contact times. 121 - vii - Table 18. Examples of generic plant names originating from mythology and traditional beliefs; 124 Table 19. Examples of plants named synonymously with objects and materials manufactured from them. 126 Table 20. Examples of plants named after innate species characteristics. 127 Table 21. Examples of plants named after substances or objects they resemble. 131 Table 22. Examples of plants named after other plants. 133 Table 23. Examples of plants having generic names which are "types" for broader taxonomic categories. 135 Table 24. A summary of the nomenelatural criteria applied to plant segregates in Haida, Bella Coola, and Lillooet. 1 3 7 Table 25. Nomenelatural criteria for Hanunoo plants, as denoted by Conklin (1954). 138 Table 26. Examples of named specific plant taxa in Haida and Lillooet. 140 Table 27. Examples of specific taxa which are psychologically valid, but which are not recognized nomenclaturally. 142 Table 28. Examples of specialized "non-taxonomic" botanical terminology in Haida, Bella Coola, and Lillooet. 150 Table 29. Examples of synonymous generic plant names in Haida,. Bella Coola, and Lillooet. 153 Table 30. Cultural status of plants in relation to their nomen- elatural recognition in Haida, Bella Coola, and Lillooet. 158 Table 31. Lexical retention and cultural significance of plants in the Skidegate and Masset dialects of Haida. 161 Table 32. Examples of Skidegate and Masset Haida plant names which are: I. identical, II. cognates, and III. linguistically unrelated. 164 - viii - FIGURES Figure 1. Map of British Columbia Indian groups, linguistic subdivisions. Figure 2. Territory of the Haida Indians in British Columbia (Queen Charlotte Islands). Figure 3. Territory of the Bella Coola Indians, showing the vicinity of the permanent village sites. Figure 4. Territory of the Lillooet Indians, showing linguistic divisions. Figure 5. Diagrammatic representation of a taxonomic hierarchy. Figure 6. A diagrammatic scheme of universal phytotaxonomic category types based on conclusions of Berlin (1971) and Raven, Berlin, and Breedlove (1971). Figure 7. Diagrammatic representation of Haida life-form categories. Figure 8. Diagrammatic representation of Bella Coola life-form categories. Figure 9. Diagrammatic representation of Fraser River Lillooet life-form categories. Figure 10. Suggested historical derivation of the specific segre- gate, 'Haida-apples' for Pyrus fusca (wild crabapple). Figure 11. A graphic portrayal of linguistic divergence of plant names in Skidegate and Masset Haida, showing the rela- tionship between lexical retention and cultural sig- nificance. Figure 12. Graph showing the relationship between the number of linguistically discrete generic-level terms applied to a plant species and cultural significance in Masset and Skidegate Haida, Bella Coola, and Lillooet, Figure 13. The domains and their chiefs in Okanagan Salish mythology. - Ix - ACKNOWLEDGEMENTS :: I have been fortunate in having as my graduate adviser Dr. Roy L. Taylor, Director of the Botanical Garden, University of British Colum- bia. His continuing interest, support, and advise from the beginning of my graduate program to the preparation of this thesis is sincerely appreciated. I would also like to thank the members of my graduate advising committee — Drs. W. Schofield, R. Foreman, and G. Hughes, of the Department of Botany, University of British Columbia, and Professor W. Duff of the Department of Anthropology and Sociology — for their suggestions and comments on the research project and editorial criti- cisms of this thesis. Dr. J. Maze, Department of Botany, and my fath- er, Dr. J. Chapman, also spent considerable time and effort in discuss- ing my research findings with me and providing editorial criticism. This research would not have been possible without assistance from the following specialists: Robert Levine (linguist, Department of Anth- ropology, Columbia University, New York); Dr. Aert Kuipers, Henk Nater, and Jan van Eijk (Department of Linguistics, University of Leiden, Holland); Randy Bouchard and his research assistant, Dorothy Kennedy (B.C. Indian Language Project, Victoria, B.C.); and Stephen Borden (pro- grammer-analyst, Department of Botany, University of British Columbia, Vancouver, B.C.). I am indebted to all of the native people who contributed their time and knowledge to the study — especially George Young of Skidegate, Mrs. Florence Davidson of Masset, Mrs. Margaret Siwallace of Bella Coola, David Moody (late) of Bella Coola, and Sam Mitchell of Fountain (Lillooet). My husband, Robert Turner, accompanied me on all of the field exped- itions., He assisted me in collecting plants, photography, and tape re- cording, and drew the maps appearing in this thesis. He has provided limitless moral suppore throughout the duration of this project. Most of the field research was funded from a National Research Coun- cil of Canada Grant (A5705) to Dr. Taylor. My graduate program was made possible through a series of National Research Council Postgraduate Schol- arships, from July 1970 through October 1973. - xi - "...I am convinced that a likely place to begin one's search for semantic universals which may reflect man's socio-technological dev- elopment is precisely in the area of man's classification of his nat- ural universe. Hence, I personally consider semantic studies of such domains as ethnobotany, ethnozoology, ethnogeography, and the like as representing important research priorities. Here, for once, is a plausible and theoretically significant reason for becoming involved in urgent ethnographic work among vanishing peoples whose apprecia- tion of the natural world comes close to that of man in earliest times." (Berlin 1969). •« - xii - . PREFACE Plant classification of aboriginal groups is not well known, and with each passing decade, a significant loss of information occurs. The following study includes the only known attempt to document native terminological systems for plants in the Pacific Northwest.* Other ethnobotanical studies have been carried out in this region (Steedman 1929; Gunther 1945; Smith 1928; Turner 1972a, 1972b; Turner and Bell 1971, 1973), but these have not included investigations of plant tax- onomic systems as an integral part of the project. The present study was begun in the summer of 1970. Initially, only the Haida Indians of the Queen Charlotte Islands were included in the program. After two summers of field work with Skidegate and Masset Haida, it was decided to expand the study to include first a river- inlet group, the Bella Coola, and secondly an interior group, the Fraser River Lillooet, in order^to add a cultural and vegetational comparative dimension to the program. Field work with these two groups was carried out during the summer and fall of 1972, and in the spring of 1973. General research procedures involved interviewing older members of the native communities, either indoorss employing fresh or dried plant specimens as subject material, or where possible in actual field situa- tions, using living plants. Information on uses of these plants, beliefs * This term, although variable in meaning, is the most appropriate word in the context of the present study. Here, it is defined as the re- gion from northern Oregon to northern British Columbia, and from the Pacific coast east to the Rocky Mountains. - xiii - associated with them, names applied to them, and any ideas about their relationships with other plants, animals, or objects were recorded. Initially an attempt was made to apply componential analysis* techniques of ethnoscience to determine underlying structures of plant taxonomic systems. This method proved to be impractical in terms of the available time of the informants, variability of responses, non- availability of phonological and syntactical information on the lang- uages, and the obvious influence of English folk categories on native thought. Because of these factors, it was felt that the structure and mean- ing of the native plant names themselves, together with the defined conditions of their application and appropriate comments and observa- ations obtained from general informal conversations, would give insights into both aboriginal and post-contact classification systems more ef- fectively and efficiently than would componential analysis. The results and ideas presented in this thesis are hopefully only the beginning of a comprehensive description of the ethnobotany and phytotaxonomy of the Indians of British Columbia. As more information from different language groups and vegetational zones in the Province is collected and analyzed, the data listed here will probably take on new meaning and significance. Meanwhile, they will contribute an addi- tional and significant element to the knowledge of cultures and man's relationship to vegetation in the Pacific Northwest. * In view of the interdisciplinary nature of the study, a selected glossary of specialized terminology is provided at the end. - xiv - PLATES Plates I & II. Florence Davidson of Masset, Queen Charlotte Islands, gathering red cedar bark (Thuja plicata) for weaving (summer 1971 see Appendix 1). Plate III. Ganoderma, one of several types of bracket fungi called 'pilot-biscuit's grandmother' in Skidegate Haida and 'rotten- wood biscuit' in Masset Haida (see Table 17), Plate IV. Sam Mitchell of Fountain (Lillooet) collecting the edible stalks of "Indian rhubarb" or cow parsnip (Heracleum lanatum) (spring 1973). Plate V. Mature plant of Heracleum lanatum. Plate VI. Heracleum stalks, showing male (flowering stalks) and female (leaf stalks) types, as distinguished by Fraser River Lillooet Indians (see Table 28). Both kinds are edible, but are prepared in different ways. Plate VII. Five varieties of Saskatoon berries (Amelanchier alnifolia) as distinguished by Fraser River Lillooet Indians (see Table 26). Top left of picture: "white" variety - spekpek. Top right - "red" variety - swelhkwa7-u7sa7. Center: "rotten" variety - nek'nakw'- ukw'sa7. Bottom left: "sweet" variety - stl'exelus. Bottom right "real Saskatoons" - stsekwm-ul. The photo was taken in June 1973. Plate VIII. The "red" variety of Amelanchier alnifolia (see Plate VII), just before the berries are ripe. When fully ripe, they are dark blue, but they are sweet and juicy even when red. Plate II - xvi - Plate III - xvii - Plate IV Plate II - xix - Plate VIII INTRODUCTION The intellectual capacity of humans for classifying natural objects and even abstract concepts is widely recognized (Tyler 1969). Even in so-called "primitive" societies, in which technology and subsistence are at a comparatively unsophisticated level, the rich diversity of the en- vironment is described in detail by the nomenclatural and classification systems within the culture. Levi-Strauss (1966) has devoted an entire book, The Savage Mind, to the proposition that "savage" societies through- out the world not only have detailed systems of ordering objects and phenomena in their environments, by that these systems, far from being haphazard, are well organized and completely logical when studied on their own premises and in their cultural contexts. Interest in aboriginal taxonomic systems has grown rapidly over the last two decades, and numerous documented descriptions of native classi- fication systems have appeared (cf. Berlin, Breedlove, and Laughlin 1970; Berlin, Breedlove, and Raven 1966; Bright and Bright 1965; Bulmer 1967; Conklin 1954; Diamond 1965;.Frake 1961; Goss 1967; Price 1967). Many such studies have included the classification and nomenclature of local floras by aboriginal peoples. Plants provide a concrete, discrete, and virtually universal semantic domain, and for this reason, are excep- tionally useful subjects for cognitive studies. The accumulation of research data pertaining to individual folk taxonomic systems for plants has inspired the development of a number of generalizations applicable to all ethnophytotaxonomies, and in some cases, to all folk taxonomies. These include a list proposed by Raven, Berlin, and Breedlove (1971) of general characteristics common to folk taxonomic systems, Conklin's (1966) discussion of the differences between folk taxa and the taxonomic groups of biological systematics, and Ber- lin's (1971) series of speculations concerning the growth and develop- ment of ethnobotanical nomenclature and classification systems. None of the data involved in the formulation of these generaliza- tions has originated from the cultures of the Pacific Northwest region. Indeed, the ethnophytotaxonomic studies considered have been largely from tropical or sub-tropical areas in cultures having agricultural economies, such as the Tzeltal-speaking Mayans of southern Mexico (Ber- lin, Breedlove, and Raven 1968), the Hanunoo of the Philippines (Conklin 1954), and the Huichol of northern Jalisco, Mexico (Price 1967). Even the temperate cultures considered — the Ojibwa (Black 1967), Navajo (Wyman and Harris 1941), Hopi (Whiting 1966), and various Californian tribes (Bright and Bright 1965) are almost all of southern temperate distribution, and most have an economy based at least partially on agriculture. The present study considers the plant taxonomic systems of three Pacific Northwest Indian groups, all aboriginally non-agricultural. The first, Haida, is an insular group of the northern Pacific coast (Figure 1). The second, Bella Coola, is a river-inlet group of the central Bri- tish Columbian coast (Figure 1), and the third, Fraser River Lillooet, is a river-oriented culture of the Interior Plateau (Figure 1). The Haida language is apparently of Na-dene stock, while Bella Coola and Lillooet are Salishan languages, and are thus distantly related. Each group is distinct from the others historically, culturally, and vegetationally. Figure 1. Map of British Columbia Indian groups, linguistic subdivisions (after Duff 1964). 3a Hopefully the data presented here on plant taxonomic systems of these groups will contribute a new dimension to the general study of cogni- tive systems, particularly as they relate to ethnophytotaxonomies, One advantage of research in ethnophytotaxonomic systems of the Pacific Northwest is that the flora in this region is well studied (cf. Henry 1915; Hitchcock et al. 1955 - 1969; Calder and Taylor 1968). The comparison of folk taxa with current botanical taxonomic categor- ies is greatly facilitated in areas that have thoroughly described floras. Modern phytotaxa, when well known, can be utilized as a "tran- slation medium" for comparison of two or more folk taxonomic systems. The present phylogenetic system, as the most universal of all nomen- elatural and classification systems for plants, and the most completely documented and regulated, serves as the only available standard against which various folk taxonomies can be described and contrasted. A disadvantage to the study of ethnophytotaxonomies in the Pacific Northwest is the recent rapid loss of language and cultural information amongst native peoples, a direct result of their acculturation into western society. None.of the Indian people involved in the study was completely monolingual, although all of them learned English only as a second language in school. Interviews for this project were conducted in English, sometimes with the help of another member of the family as a partial interpreter. The high degree of acculturation of the study groups was one of the main factors involved in the lack of success in applying componential analysis procedures in the program. Formal semantic methodology, including componential analysis techniques, have been successfully ap- plied in many of the more rigorous studies of folk taxonomic systems (Tyler 1969) , but attempts to use them in the present study proved impractical and produced inconclusive results, at least partially be- cause of a pronounced but immeasurable influence of "white" cognitive systems on native thought. As a result, a more informal type of inter- viewing was adopted. The results of these interviews were combined with an analysis of the content and conditions of application of the aboriginal plant terms themselves. From all indications, these terms, as basic lexical components of a language, seem less subject to varia- tion over time than cultural ideas and opinions about inter-relation- ships between plants (cf. Bright and Bright 1965). In the Pacific Northwest, as in other regions, it is essential to consider ethnobotanical information on the uses and roles of plants in a society as both influencing and reflecting classification of plants. Thus, collecting data on the cultural importance of plants has been an integral part of the present study. Ethnobotanical data are significant in their own right as resource materials for many different fields of study (cf. Schultes 1960; Turner and Bell 1971), and for purposes of this project, the ethnobotanical information accumulated for each of the three groups has been organized for publication as a discrete unit. Sorting and summarizing the immense variety of data relating to aboriginal plant names, botanical taxa, and cultural information about plants was accomplished in this project by means of a computer. The coding system and the sorting program used were designed specifically to handle these data and similar types of data for other Pacific North-west groups. To my knowledge, this particular approach to the analysis of aboriginal plant names and ethnobotanical data for the purpose of discerning and summarizing folk taxonomic relationships has not been attempted elsewhere. It has a number of advantages, and may prove useful on a wider scale. BACKGROUND TO THE STUDY In this section, pertinent information on the history, culture, language, geography, and vegetation of the three study groups is pre- sented. The Haida Indians The Haida formerly occupied about 20* permanent villages around the coast of the Queen Charlotte Islands and the Prince of Wales Island group in Alaska (see Figure 2). ** Their aboriginal population is es- timated at 7,000 to 10,000 (Jenness 1934). By 1915 the Haida popula- tion on the Queen Charlottes had been reduced to under 600 individuals (Duff 1964). Those from the southern half of the Islands had assembled at the village of Skidegate, and are now recognized as the speakership of the Skidegate dialect. *** People from the northern villages had congregated at Masset, and now comprise the speakership of the Masset dialect. As of 1970, the Haida (excluding the Kaigani people of Alaska) numbered 1367: 1015 at Masset, and 352 at Skidegate (Department of Indian * Estimates of the number of village sites run as high as 39 (Harrison 1895), but for any given time, the figure of 20 is probably more realistic. ** According to available information, the migration of Haida people to Alaska was relatively recent, having taken place about A.D. 1750 (Swan- ton 1911). The Kaigani dialect, spoken by Alaskan Haida, is very simi- lar to the Masset dialect. The Kaigani Haida are not considered in the present study. *** The Skidegate dialect, even today, is apparently a partially artificial grouping, since different Haida speakers at Skidegate display major phonological and grammatical differences in their speech (Robert Levine Columbia University, New York, personal communication). For purposes of this study, the Skidegate people will be considered a dialectic unit Figure 2. Territory of the Haida Indians in British Columbia (Queen Charlotte Islands). 8a HAIDA INDIAN T E R R I T O R Y (QUEEN CHARLOTTE Is.) 10 20 30 «0, 'Ml. •d^o g'o ^ =wm K m-Affairs and Northern Development 1970). The linguistic affiliations of the Haida are not well understood. Like the Tlingit Indians of Alaska and the Athapaskan peoples of the northern interior of British Columbia and Alaska, their language is suggested to be an isolate of the Na-dene phylum of languages (Driver 1961). Similarly, the territorial origins of the Haida are not known. Archaeological studies indicate that the Queen Charlotte Islands have been occupied for at least 8,000 years, presumably by at least some of the ancestors of the present day Haida population (Fladmark 1970), It has been suggested that the first people travelled to the Queen Char- lottes during the Pleistocene, at the time of a glacial maximum, over an exposed section of sea floor. * Fladmark (1970) points out that a drop of only 250 feet in sea level would connect the Charlottes to the off-shore islands and mainland of Alaska. Heusser (1960) suggested this route to explain the presence of caribou on the Queen Charlotte Islands. The Haida Indians belong to the northern province of the Northwest Coast Cultural Area. This sub-unit also includes the Tlingit and Tsimshian cultures, and marginally those of the northern divisions of the Kwakiutl (Drucker 1955) (see Figure 1). These groups are charac- terized by a number of cultural traits, including a matrilineal social organization, with exogamous moieties, forming the basis of crest * Archaeological evidence suggests that at this time, the technology for construction of ocean-going canoes had not yet been developed (Flad- mark 1973, public lecture on "The Prehistory of the Queen Charlotte Islandc11 - 10 - ownership, inheritance, life-cycle rituals, and social functions. The Haida, and to a lesser extent, the other groups in the northern sub- unit, are widely known for their outstanding sculptural and graphic artforms, based on stylistic representations of natural objects. Dru- cker (1955) lists other features characterizing the northern sub-unit. The Haida people were coast dwellers. Their economy centered around the ocean, beaches, river-mouths, and lowland forests of the Queen Charlottes. They rarely ventured into the mountainous interior of the Islands or into the extensive muskegs of Graham Island, except to hunt waterfowl or pick berries. Their villages were situated in the wet subzone of the Coastal Western Hemlock Biogeoclimatic Zone (Krajina 1970), which extends throughout the lower elevations of the Islands. Through seasonal migrations and inter-village contacts, the Haida encountered a variety of plant community types within this zone. Most notable are: marine and intertidal algal communities, maritime com- munities (including shingle and sand beaches, rocks and cliffs, and salt marshes), bog and swamp communities, fresh water aquatic com- munities, and forest communities (including sand-dune forest, meadow forest, and closed forest) (described in Calder and Taylor 1968). Several upland forest and montane communities also occur on the Islands (see Calder and Taylor 1968), but because of their lack of contact with upland areas, the Haida people were generally unfamiliar with montane flora. Even before before the coming of the white man, the Haida apparent- ly had frequent contact with other Indian groups. They crossed over to - 1 1 - the mainland every spring to obtain eulachon grease from the Nass River Tsimshian, in exchange for canoes, carved chests, sea-otter skins, dried herring eggs on kelp, and dried Porphyra.* They also traded with the Tlingit for Chilkat blankets, copper, mountain-goat horn, and mountain-sheep horn (Drucker 1950). In post-contact times, these trading expeditions increased in frequency, and potatoes, turnips, and other garden vegetables were added to the list of items traded by the Haida. The dominant tree species of the lowland forests of the Queen Charlottes are all conifers: Tsuga heterophylla (western hemlock), Picea sitchensis (Sitka spruce), and Thuja plicata (western red cedar). All of these attain considerable stature in mature forests, and all were important economic species to the Haida. In the upland forests, of the Mountain Hemlock Biogeoclimatic Zone (Krajina 1970), Tsuga merten- siana (mountian hemlock) becomes increasingly prevalent (Calder and Taylor 1968). Pinus contorta (lodgepole pine) and Chamaecyparis nootka— tensis (yellow cedar) are dominant species of the lowland muskegs, and Taxus brevifolia (western yew) occurs sporadically in the forested areas. Alnus rubra (red alder) is the only abundant deciduous tree on the Islands. It commonly grows in burned or disturbed sites, and has undoubtedly increased in frequency since the advent of logging. Pyrus fusca (wild crabapple) and Alnus crispa ssp. sinuata (Sitka alder) also occur in many areas. * An index of common names of plants mentioned in this thesis is in- cluded at the end. - 1 2 - A number of tree species are conspicuously absent from the Queen Charlotte flora. These include Pseudotsuga menziesii (Douglas fir), Abies amabills (amabilis fir), A., grandis (grand fir) , A. laslbcarpa (subalpine fir), Pinus monticola (white pine), Populus trichocarpa (black cottonwood), P. tremuloides (trembling aspen), Betula papy- rifera (paper birch), Acer macrophyllum (broad-leaved maple), A. glabrum (Rocky Mountain maple), Prunus emarginata (bitter cherry), and Rhamnus purshiana (cascara). Most of these species do occur in the territory of the Bella Coola Indians, and form a major differentiating feature between the types of vegetation encountered by these two groups. The Bella Coola Indians The Bella Coola people once lived in numerous scattered villages along the Bella Coola, Kimsquit, and Kwatna Rivers, and the upper reaches of Dean and Burke Channels (see Figure 3). Mcllwraith (1948) lists about 20 villages in this area which were occupied around the time of Mackenzie.'s journey to the coast in 1793. Various other sites are known to have been occupied at the same period at least on a temp- orary basis (Hobler 1970). The most concentrated aboriginal population was apparently in the Bella Coola Valley (Hobler 1970). In pre-contact times, the Bella Coola population probably exceeded 3,000 individuals, but by 1929, this figure had declined to 250, mostly due to disease epidemics (Mcllwraith 1948; Duff 1964). As of 1970, there were 597 people in this group, occupying a single village, Bella Coola, near the mouth of the Bella Coola River (Department of Indian Affairs and Northern Development 1970). . Figure 3. Territory of the Bella Coola Indians, showing the vicinity of the permanent village sites. (The actual range of the Bella Coola extended over a significantly wider area.) 13a - 1 4 - The Bella Coola are an Isolated enclave of Salish speakers in a Kwakiutl speaking region (see Figure 1). Their exact affiliations with other Salishan groups are still unknown (see Jorgensen 1969). They may have originally migrated northward along the coast from the lower mainland of British Columbia, or across the Coast Mountains from the Interior Plateau. Alternately, they may have at one time had a continuous distribution with other Salish groups, having been subsequent- ly isolated by the intrusion of the Kwakiutl and Carrier peoples. The length of time the Bella Coola have occupied the area is also unknown. C-^ datings at one archaeological excavation site indicate the presence of humans in the area about 9,000 B.P. (P.M. Hobler, archae- ologist, Simon Fraser University, Vancouver, B.C., personal communica- tion) , but the Bella Coola probably did not arrive until considerably later. Archaeological work in the Bella Coola area is continuing, and ultimately a more complete chronology of human habitation will be es- tablished. The Bella Coola are a uni-dialectic group, although people from Kimsquit and probably from other outlying areas as well, show slight differences in their speech compared to those from Bella Coola (Henk Nater, linguist, University of Leiden, Leiden, Holland, personal com- munication) . The Salishan origin of the Bella Coola is reflected in part by their amorphous, informal social organization. Generally, however, Bella Coola cultural traits, particularly their material culture and mythology, show a remarkable similarity to those of the neighbouring - 1 5 - Kwakiutl peoples, indicating a high degree of social interchange and cultural "borrowing". Thus, the Bella Coola are included in the middle, or Wakashan-speaking province of the Northwest Coast Cultural Area (Drucker 1955). The resources of the local rivers and inlets were crucial to the subsistence of the Bella Coola people, but their resource base was broader than a local one. Within the year, they travelled over a wide territorial and elevational range hunting mountain-goat and other mam- mals, gathering berries, cedar bark, and various other plant products, and trading with adjacent Indian groups, including the Northern and Southern Kwakiutl on the coast, and the Carrier and/or Chilcotin * peoples of the interior (Mcllwraith 1948; Margaret Siwallace, Bella Coola, B.C., personal communication). As with the Haida, the Bella Coola village sites are situated in the Coastal Western Hemlock Biogeoclimatic Zone (Krajina 1970), but their travels brought them in contact with a number of other vegetation zones, including the Mountain Hemlock Zone, the Engelmann Spruce - Subalpine Fir Zone, and the Caribou Aspen - Lodgepole Pine - Douglas-fir Zone (described in Krajina 1970). In terms of subsistence, the most important community types to the Bella Coola were the well vegetated estuarine flats, such as those at Bella Coola, and the various types * Morice (1925) states that the references of other ethnographers, such as Harlan Smith, to contacts between the Bella Coola and Carrier peoples were mistaken, and that actually the Chilcotin people, not the Carrier, were involved in contacts with the coast. Mcllwraith (1948) and Goldman (1941), on the other hand, cite many instances of inter- action between Carrier and Bella Coola peoples. of forest communities. These have not been studied floristically or ecologically in any detail. The peat bog or muskeg community, so pre- valent on the Queen Charlottes, is almost entirely lacking from the Bella Coola area; some of the most common plants of this community, such as Kalmla polifolia and Vaecinium oxycoccus, are unknown to the Bella Coola people, at least at the present time. Common tree species of the Bella Coola area are: Tsuga heterophylla (western hemlock) , T_. mertensiana (mountain hemlock), Ficea sitchensis (Sitka spruce), Thuja plicata (western red cedar), Pseudotsuga menziesii (Douglas fir), Abies amabilis (amabilis fir), Chamaecyparis nootkatensls (yellow cedar), Pinus contorta (lodgepole pine), Alnus rubra (red alder), A. crispa ssp. sinuata (Sitka alder), Populus trlchocarpa (black cotton- wood), Acer glabrum (Rocky Mountian maple), and Pyrus fusca (wild crab- apple). Other species, less frequent, but nevertheless present are: Taxus brevifolia (western yew), Abies lasiocarpa (subalpine fir), Picea engelmannii (Engelmann spruce), Populus tremuloides (trembling aspen), and Prunus emarginata (bitter cherry). Notably absent from the Bella Coola flora are: Abies grandls (grand fir), Pinus monticola (white pine), Cornus nuttallii (Pacific flowering dogwood), Arbutus menziesii (Pacific madrone), and Acer macrophyllum (broad-leaved maple), whose ranges do not extend as far north as Bella Coola. The Lillooet Indians Lillooet peoples are categorized into two dialectic groups: Lower Lillooet and Upper Lillooet. These are differentiated not only - 1 7 - linguistically, but culturally, geographically, and ecologically. They are further divided into four smaller divisions, or "bands"* (Teit 1906). The Lower Lillooet group includes the Lillooet River band, form- erly occupying eight villages at Douglas and along the Lower Lillooet River, and the Pemberton band, formerly occupying five villages at Lillooet Lake and Pemberton Meadows (Mount Currie) (see Figure 4). Lower Lillooet people are presently concentrated at the villages of Douglas, Skookum Ghuck, Samahquam, and Mount Currie.** The Upper Lillooet group consists of the Lake band, formerly occupying six villages around Anderson and Seton Lakes, and the Fraser River band, formerly occupying six villages along the Fraser River from just below the present town of Lillooet to below the mouth of Pavilion Creek (Figure 4). Present Upper Lillooet settlements** include Anderson Lake, Seton Lake, Cayoose Creek, Lillooet, Bridge River, and Fountain. The present study involves only the Fraser River band of Upper Lillooet; the major informant is from Fountain. Early estimates suggest that Lillooet peoples may have numbered 4,000 in precontact days (Teit 1906). By 1903, this number had been reduced by smallpox epidemics and famines to just over 1,100 ~ about 500 in the Lower Lillooet group and about 650 in Upper Lillooet (Teit 1906). In 1970, the Lillooet people numbered 2,494: 1,321 Lower Lillooet and 1,173 Upper Lillooet (Department of Indian Affairs and Northern Development 1970), * Teit's (1906) use of the term "band" is more general than the present day concept of "band" as a single Indian village unit. ** These are "bands" in the modern context. - 1 8 - Figure 4, Territory of the Lillooet Indians, showing linguistic divisions. 1 8 a - 1 9 - The Lillooet language is related to the Thompson, Shuswap, and Okanagan languages of British Columbia. All of these are classified in the interior division of the Salish language family (Driver 1961), as seen in Figure 1. Recent archaeological research in the Lytton- Lillooet area has given some insights into the pre-history of this region (Sanger 1969). After Pleistocene glaciation, the first people to enter the Lillooet area, as early as 9,000 B.P., were migrants from the area now included in the State of Washington. Their identifiable traits are known as the Lochnore complex, and are characterized by leaf-shaped projectile points, macroblades, edge-battered cobbles, and an absence of micro- blades (Sanger 1969). About 7,000 years ago, these people were appar- ently displaced by another group from the central interior of British Columbia. The new occupation, termed the Nesikep Tradition, is char- acterized by a more advanced stone-chipping technology with microblades. The Nesikep Tradition passed through a number of distinct periods, and can be viewed as ultimately evolving into the Interior Salish cultures of the historic period (Sanger 1969; Stryd and Hills 1972)„ The cultural features of the Fraser River Lillooet are similar to those of other Interior Salish groups. Together, these groups form a rather loosely defined unit known as the Plateau Culture Area, named after the Columbia plateaus (Driver 1961). General features of this unit include: a loosely structured social organization, without emphasis on rank or class; the use of semi-subterranean winter dwellings, called pithouses, for extended family groups; and a hunting-gathering economy, with emphasis on fishing anadromous salmon. - 20 - The Plateau cultures, particularly in the peripheral groups, show many cultural traits of adjacent culture areas (Driver 1961). The Lower Lillooet, for example, are actually culturally transitional be- tween the Northwest Coast and Plateau Culture Areas. The Fraser River Lillooet are a definite unit of the Plateau Culture Area, both cultur- ally and geographically, but even this group had indirect access to coastal cultures through frequent trade contacts with the Lower Lillooet. This latter group used to make annual journeys to the Fraser River area in late summer, to exchange goods such as dentalia and other shells, cedar bark, wood of yew, vine maple, and yellow cedar, hazelnuts, dried huckleberries, goat-hair blankets, and fish oil, for interior products, such as Indian hemp (Apocynum spp.), Salix exigua bark twine, Erythronlum grandiflorum bulbs, dried Saskatoon berries, soapberries, and choke cherries, dried meat, fat, and animal skins (Teit 1906). Three Biogeoclimatic Zones are distinguished in the Fraser River Lillooet territory: the Ponderosa Pine - Bunchgrass Zone of lower eleva- tions (below about 2,000 feet); the Interior Douglas-fir Zone of middle elevations (approximately 2,000 to 4,500 feet); and the Engelmann Spruce - Subalpine Fir Zone of elevations above about 4,500 feet (Krajina 1970). The permanent winter dwellings of the Fraser River Lillooet were usually located in the Ponderosa Pine - Bunchgrass Zone. The topography of the Fraser River Lillooet territory is extremely variable. Stryd and Hills (1972) divide the area into three major physiographic units: the rocky canyon floor of the Fraser River, fre- quented for salmon fishing; the sloping terraces above the River, where most of the winter pithouse dwelling sites and modern settlements are - 21 - located; and the hills and mountains above the river valley, the usual sites of hunting and root-gathering activities. Numerous plant communities occur within these biogeoclimatic zones and physiographic units, each modified by soil texture, slope, exposure, and available moisture. In terms of the Fraser River Lillooet economy, the most important communities were those of the high mountain slopes and valleys, where large quantities of "Indian potatoes" (Glaytonia lanceolata and Erythronium grandiflorum) were dug annually, and the dry river terraces, where several types of berries (such as Amelanchier alnifolia, Crataegus douglasii, and Prunus virginiana) and "roots" (e.g. Balsamorrhiza sagittata) were gathered. Common tree species in the vicinity of the Lillooet-Fountain area are: Pinus ponderosa (ponderosa pine), P^. contorta (lodgepole pine), Pseudotsuga menziesii var. glauca (interior Douglas fir), Populus tremu- loldes (trembling aspen), jP. trichocarpa (black cottonwood), Betula papyrifera (paper birch), and Acer glabrum (Rocky Mountain maple). At higher elevations, Chamaecyparis nootkatensis (yellow cedar), Abies lasiocarpa (subalpine fir), Picea engelmannii (Engelmann spruce), Picea glauca (white spruce), Pinus albicaulis (white-bark pine), Alnus crispa ssp. slnuata (Sitka alder), and A. incana (mountain alder) are found. A number of coastal trees occur in the mountains to the west of Lillooet, and are recognized by the Fraser River Lillooet. These include Pinus monticola (white pine), Thuj a pllcata (western red cedar), Taxus brevi- folia (western yew), Alnus rubra (red alder), Acer macrophyllum (broad- leaved maple), Prunus emarginata (bitter cherry), Pyrus fusca (wild crabapple), and Rhamnus purshiana (cascara). - 22 - METHODOLOGY ' The methodology applied in the present study was influenced and directed by methods used by other researchers in the description of folk taxonomic systems. The standard type of methodology associated with cognitive studies is outlined and discussed in the first part of this section. Following this is a discussion of the methodology in the context of the present project. Finally, the procedures ultimately adopted in the study are outlined in detail. Ethnoscience and Componential Analysis Many descriptions of folk taxonomic systems have been "...nonsystem- atic, incomplete, and anecdotal" (Berlin, Breedlove, and Raven 1966). In recent years, however, there has been a trend towards increased formal- ization in the collection and analysis of ethnosemantic data. A new ethnographic school has developed in the past two decades, variously known as ethnoscience, ethnosemantics, or cognitive anthropology,* whose basic tenets include the systematic collection and formal analysis of cognitive and semantic terminology.. The central method within the field of ethnoscience Is known as componential analysis.** This method was first described in relation to cognitive systems by Goodenough (1956) in the periodical Language (v. 32, no. 1). Also appearing in the. same number of Language was a paper * Other synonyms for these terms include: ethnographic semantics, linguistic ethnography, and folk science (Eglin 1972; Berlin 1968; Werner 1967). ** This term is frequently regarded as another synonym of ethnoscience (Werner 1967; Berlin 1969), but I feel that Eglin's (1972) descrip- tion of it as a method rather than a discipline is more appropriate. - 2 3 - by Lounsbury (1956) in which componential analysis was applied to the study of a kinship system. Since that time, adherents of the ethno- scientific school have produced as ever-growing body of semantic and folk taxonomic studies, based on componential analysis and other formal methods of analysis. The goals of ethnoscience and componential analysis are commend- able. "The problem is to define the taxonomic system itself — that is, to explicate the rules by which users of the terms group various social and genealogical characteristics into concepts" (Wallace 1962), or, more generally, to discover "how people construe their world of experience from the way they talk about it" Frake 1962). Ethnoscientific procedures can be detailed as follows: 1) an inventory is made of terminology within a given semantic domain; 2) information is assembled on each linguistic form as a semantic class of objects; 3) when possible, the classificatory dimensions imposed upon the field by native linguistic usage are isolated; 4) through a series of culturally appropriate questions,- semantic distinctions (components) are established which apportion the terms into sets and sub-sets, such that every item in the domain is dis- tinguished from every other item by at least one component, and is at the same time related to every other item by inclusion at some level in a broader taxonomic category; and 5) a classification is erected based on the successive-inclusion and exclusion of each defined item within the domain (Lounsbury 1963; Burling 1964; Berlin 1968). The procedures outlined are accomplished through interviews with preferably a large number of native speakers. In order that there be no cultural bias or misunderstandings on the part of the ethnographer, the interviews should be conducted entirely in the language of the - 2 4 - native informant (Conklin 1962; Werner 1967), and care should be taken not to bias the informant's responses by allusions to other taxonomic systems familiar to the researcher. In terms of ethnophytotaxonomic research, this means that the interviewer must make a special effort not to imply equivalence between folk phytotaxa and botanical taxa, even when it is convenient to do so. To obtain an authentic description of a classification system in another culture, one must never incorporate assumptions or implications about the nature of the system into the elicitation process. Thus, to ask a question, "What kind of a tree is that (x)?" without first estab- lishing the informant's definition of 'tree' and his assurance that x is a kind of tree, would immediately render the informant's response invalid. Metzger and Williams (1966), Price (1967), and Frake (1964) describe a program of elicitation based on successive or linked questions and responses which, at least theoretically, eliminates bias introduced by the questioning process. Ideally, this program allows the interviewer to begin with any given item or segregate within a domain and position it vertically and horizontally within the taxonomic hierarchy of the do- main. Thus, beginning with a described lexeme, x, in a hypothetical classification system (see Figure 5), one can progress downwards through X Figure 5. Diagrammatic representation of a taxonomic hierarchy. - 2 5 - the taxonomic hierarchy by asking, "What kinds of x are there?" Given answers x^, x2, x3 ••»xn» each differentiated from the other by at least one character, one can proceed through a similar line of questioning to discover the various sub-categories of x, (namely x, and x„,). When 1 la lb the lower taxa have been explored and described to their limits, one can return to the first item, x, and define the more general taxa of the system by asking, "What is x a kind of?" Given answer X, one can then expand the system horizontally to include y and z by the question, "What other kinds of X are there?" Theoretically, this type of progressive elicitation can be applied in exploring and describing any taxonomic system. The above example is simplified to an extreme. In practise, folk taxonomic systems are more complex, irregular, and indefinite than the example implies (Conklin 1962). Checking the validity of the derived taxonomic structure can be accomplished by repeating the construction of the hierarchy from several different starting points, thus providing cross-referencing for each item. The system can also be tested by formulating questions which are indi- cated to be inappropriate by the nature of the derived hierarchy. For example, in the hierarchy illustrated above, one could ask, "Is y a kind of x?" or "Is z a kind of y^?" Positive responses to these questions would obviously demonstrate some irregularity in the system as it is constructed. Numerous cognitive systems in many different cultures have been investigated using ethnoscience techniques. The most thoroughly ex- plored domain is that of kinship (cf. Lounsbury 1964; Conklin 1964; Wallace and Atkins 1960; Romney and D'Andrade 1964), where even "Yankee" - 26 - terminology has been subjected to analysis as a test case (Goodenough 1965). Other terminological systems which have been described include: numeral classifiers (Berlin 1968) ; firewood (Metzger and Williams 1966); terms of personal reference (Metzger and Williams 1962); curers (Metzger and Williams 1963a); weddings (Metzger and Williams 1963b); agriculture, betel chewing, pottery, verbal play, colour, water (see Con'klin 1962 for references); law (Black and Metzger 1965); spiritual- ist church language (Zaretsky 1969); and medicine (Werner 1967). Part- icularly relevant to the present study are the ethnoscientific descrip- tions of ethnobiotaxonomies (Berlin, Breedlove, and Raven 1966; Black 1967; Bulmer 1967, 1970; Bulmer and Tyler 1968; Conklin 1954; Diamond 1965; Bright and Bright 1965). Ethnoscience with Respect to the Present Research Program Superficially, ethnoscience and componential analysis appear to provide an ideal theoretical and methodological framework for investi- gating and describing the ethnophytotaxonomies of the Haida, Bella Coola, and Lillooet Indians. However, attempts to apply ethnomethodology to the study of cognitive systems for plants in these groups were generally unsuccessful in producing meaningful or conclusive results. Formal analysis was ultimately abandoned as a technique, although certain procedures of the described methodology were retained. The reasons for this are discussed in the following paragraphs. Even proponents of ethnoscience are aware of a number of methodol- ogical and theoretical limitations of the discipline. One serious problem is that componential analysis, even when properly conducted, - 27 - does not automatically yield a single "true analysis" or description of a semantic system. Instead, there is a "virtually infinite number of ways a lexical set can be componentially divided" (Colby 1966; see also Wallace and Atkins 1960; Burling 1964; Goodenough 1965). In other words, several different models of semantic structure of a term- inological system can be established, each of which can accurately account for the lexical items within the system. Even the original researchers in componential analysis admit that "... the criteria by which one chooses one model over another, however, remain to be deter- mined" (Goodenough 1965). Thus, the "psychological validity" of systems derived by compon- ential analysis is subject to question. At least some ethnoscience critics have charged that any single analysis offered as the taxonomic system of a given domain in a culture is necessarily based on the ar- bitrary exclusion of large bodies of relevant data, and cannot in any way be taken as a complete or conclusive statement (Schneider 1969; Burling 1964; Eglin 1972). From my own experiences in attempting to apply formal analysis to discern ethnophytotaxonomies, I found it diffi- cult to obtain consistent responses among several different informants, or even from individual informants from one day to the next. For example, to each of the six Haida informants (four at Skidegate, two at Masset), I showed a needled branch of the botanical species Picea sitchensis, and asked, "What is this?" The answer was always given, "kaayt" in Skidegate (S), or "kiiyt" in Masset (M). A conversation would - 2 8 - follow, in which kaayt * would be described in more detail, and comments on its size and abundance on the Queen Charlotte Islands would be made. The unvarying answer to the question, "How many kinds of kaayt are there?" was always, "There is only one kind — that's it." At this point, however, agreement ceased. Several of the informants volunteered the information that the word kaayt meant "tree" in English, so I attempted to determine how closely the taxon corresponded to our own folk taxon, "tree". I asked in turn about each of the terms I had al- ready elicited for "trees", including k'aang (Tsuga spp.**), ts'alh (Pinus contorta) , ts'uu (Thuja plicate) , s_gaalhaan (Chamaecyparis nootka- tensis), Ihgilt (Taxus brevifolia). kal (Alnus spp.), and k'Snlhel (Pyrus fusca), "Is this a kind of kaayt?" Those who did not give the English translation of kaayt as "tree" gave a negative response for each of the terms listed. Those who did translate kaayt as "tree", although they had just said that there was only one kind of kaayt (whose features, when described, corresponded to those of the botanical species, Picea sitchensis), gave affirmative answers for the first one, the first two, the first four, or the first five terms,.all of which are "coniferous" species. In some cases, the sixth term, kal (primarily Alnus rubra), was included in the concept of "tree". One informant agreed that k'anlhel "must be" a kind of kaayt, but when asked later, disagreed. All informants were ambivalent about including kal as a kind * For the rest of this discussion, Skidegate names only are used, although Masset people were also involved in the study. ** As far as I have been able to determine, the equilibration of Haida terms with these botanical taxa is an accurate estimate of their semantic range. - 2 9 - of kaayt, sometimes agreeing to it, and sometimes not. When the ques- tion was again asked, "what are the kinds of kaayt?", regardless of the answers to the questions about the other terms, each informant would emphatically and sometimes impatiently state, "There is only one kind of kaayt!" This process was repeated several times with each person, and when the opportunity arose, I would ask the same thing when outside, pointing out the various kinds of "trees". I would again receive answers which were seemingly contradictory. I concluded that the term kaayt is at some stage of semantic evolution between a "generic" term and a "life-form" term (see Berlin 1971). The same degree of response varia- bility was obtained in many other discussions about other types of plants including specimens of lichens, ferns, and marine algae. Jones (1971) found a similar situation when working with twelve informants on the English folk classification of evergreen trees. Burl- ing (1964) implied the same degree of variation for English folk phyto- taxa in general. Price (1967) was completely unsuccessful .in applying this elicitation technique to Huichol phytotaxonomy. The problem of response variation is compounded in a transitional culture such as Haida, where virtually all speakers are bilingual. It is impossible to discern the extent to which ideas of the Haida about the inter-relationships between plants have been influenced by English folk categories. For example, if the Haida term kaayt is transitional between a generic and a life-form label, it is probable that there is a strong tendency now to equate its life-form status with that of "tree", whether or not it was evolving towards an equivalence with "tree" before white contact. - 3 0 - Some investigators have maintained that rigorous componential analysis, by its very definition, does not allow comparison of cogni- tive systems between two or more cultures, since as soon as the seman- tic elements of a given culture are translated into terms of another culture, they lose their discrete and essential nature. This situation is comparable to one in chemistry, where a compound being subjected to analysis changes its structure as a result of the conditions imposed by the analytic process. Only the strictest of interpretations of compon- ential analysis yields such, a barrier to cross-cultural studies; indeed, some of the classic componential analyses (Lounsbury 1956; Goodenough 1956) rely heavily on cross-cultural kinship descriptions (Colby 1966). Nevertheless, the conflict between "anthropologists who stress rigorous descriptive ethnography Cof a particular culture — i.e., 'ethnoscience1 1 and anthropologists who emphasize comparative studies" has been con- sidered as a very real concern (introductory remarks, Colby 1966). Closely related to this conflict is an argument centered on the basic goal of componential analysis — to seek.out "what is inside people". The so-called "inside men", the staunch supporters of com- ponential analysis, believe that the only meaningful and accurate de- scription of a native's universe is that attained by investigating the very thought processes of the natives themselves. The "outside men", on the other hand, believe that, "It is not necessary that the dimen- sions or principles of the anthropologist's model be expressed by informants in a direct fashion or even that of their model's as given verbally or by other means, have some correspondence in their principles or dimensions with those of the analyst... If his concern is the accurate - 3 1 - and economical description of native behavior, or, further, of human behavior, selection of some particular native model and its translation may indeed be undesirable" (Hammel 1966). In fact, Hammel (1966) suggests that "...a good analysis by and 'inside' man and a good analysis by an 'outside' man are likely to be equivalent and only redundantly different, if not identical." These two questions — the degree of inter-cultural comparison allowed by componential analysis, and the necessity or even desirability for an "inside view" of taxonomic perception — have been important considerations in the methodological approach of this study. They are both theoretical questions, and have been debated at length in semantic literature (cf. Colby 1966; Tyler 1969), but from a practical view, in terms of the present study, a less structured analysis allowing some means of inter-group comparison was felt to be desirable. An incidental criticism of componential analysis Is that it does not account for unlabelled folk segregates or "covert" categories, which were originally suggested to be as significant in native*thought as the normally recognized monolexemically labelled folk taxa within a given semantic domain (Berlin, Breedlove, and Raven 1968). This criti- cism can also be directed at the methodological approach adopted in the present study, since it is directly involved with terminological systems. However, Berlin himself, in a later paper (1970), implied that covert categories, by the very fact of being unlabelled, cannot be considered to be as culturally important as labelled taxa, and are in fact highly variable and of short duration in folk taxonomies. A consideration which to my knowledge has not been discussed in - 3 2 - ethnoscience literature, is the practical matter of the time and effort componential analysis requires from the informants. This was a parti- cularly critical problem in the present study. All academic research is necessarily limited by time and available resources. In studies such as this one, where consultation with informants plays a major role, the number of available informants and the time and energy they are able to contribute to the study is of primary concern. Particularly in the case of the Haida, but also in Pacific Northwest groups generally, the only informants capable of answering the questions about plants and plant terminology required for this type of study are members of the oldest generation. These people were cooperative and enthusiastic, but were always extremely busy with affairs involving their friends and families, and interviewing time was usually "squeezed in" whenever they had a few hours of free time. Even during these periods, there were con- stant interruptions, which made elicitation difficult. Also, being elderly, they became tired easily, and I had to be careful to allow frequent breaks and not to let the sessions continue beyond about two hours. ~ All of the informants were willing to answer questions relating to the names and uses of plants and any other information they could think of concerning the plants, and with few exceptions, they stressed accuracy above all other factors.* Additionally, since they were bilingual, they were able to give glosses for aboriginal botanical terms fairly readily, * Many told me at one time or another, "I don't want to say that because it might be the wrong thing, and everyone will think I am crazy." It was a point of honour and reputation to tell nothing but what was known to be true. - 3 3 - and with a high degree of accuracy, as I could discern by cross- checking with different people. Thus, I was successful in obtaining native terms for different kinds of plants, in defining the extent and conditions of their actual application, in approximating their meanings by obtaining English equivalents, and in determining the cultural importance of different kinds of plants. However, all of my attempts to apply the formal questioning procedures of componential analysis were met with impatience and irritation on the part of the informants. Xt was not a question of lack of interest or capability, or of unwillingness to cooperate, but rather, I believe, of a true inability to provide definite answers to the questions, namely because such definite answers do not exist, at least within the present cognitive system for plants. This apparent vagueness of semantic distinctions for plant taxa will be discussed in a later section of the paper, but essentially, it has resulted firstly in the response variation described earlier, and secondly, in the unwillingness of the informants to be "pinned down" by specific questions relating to the inter-relationship of plant categories. It is probable that some of the difficulties in elicitation I en- countered in attempting to use componential analysis in describing Haida phytotaxonomy would have been eliminated if I had conducted the study in the Haida language, but this was impractical in view of the - 34 - scope of the study, the lack of knowledge of Haida linguistics,* and my own lack of linguistic experience. Furthermore, since I was working directly with bilingual informants and actual communication with them was not a problem, I felt that the advantages of learning the language were substantially outweighed by the technical difficulties involved. A more serious problem, resulting from the bilingualism of the Haida, was the already mentioned probable alteration of original Haida taxonomic categories by English folk taxonomic concepts. The extent of such interference could never be determined fully, and componential analysis, rather than indicating and crystallizing past semantic ideas about plants, served only to emphasize the complexities of the present phytotaxonomic system. I finally concluded that if it were possible to describe original Haida phytotaxonomy under present conditions, an approach other than componential analysis should be attempted. Further, I reasoned that the actual Haida plant names, still virtually unchanged after 70 years of rapid acculturation,** would provide the only valid key to the Haida * In 1972, Robert Levine, a doctoral student from Columbia University, New York, began an extensive study of Haida grammar, but at the time of this research, no linguists were actively involved with the Haida language in British Columbia. Two linguists, Dr. Joseph Kess, Univ- ersity of Victoria, and Dr. G. Bursill-Hall, Simon Fraser University, had done past field work in Haida and were helpful in making their tapes available to me, and in offering advice on Haida phonetics. An- other linguist, Dr. Michael Krauss, of the University of Alaska, is engaged in studies of the Kaigani dialect of Haida, and has offered some help in transcribing tapes. Salishan linguist, Randy Bouchard, and especially Robert Levine have recently given me a great deal of assis- tance recording Haida phonetics. ** I was able to determine the stability of Haida plant names by checking them with terms recorded at the turn of the century by Newcombe (1897 - 1906) and Swanton (1905a, 1908). - 35 - phytotaxonomic system. From this point, my investigations were directed towards the isolation and description of features of Indian plant term- inology which might indicate both nomenclatural practices and direct or indirect grouping of plants into taxonomic categories. The specific procedures involved in these investigations are outlined in the follow- ing section. Research Procedures Used in the Study For each of the three study groups — Haida, Bella Coola, and Fraser River Lillooet — research procedures can be subdivided into the follow- ing general categories: literature research and collection of background information; field vegetation surveys; consultation with native infor- mants; and synthesis of research data. a) Literature research The accumulation of background information pertinent to the study has been a continuing process. It has involved a review of literature on ethnoscience and componential analysis, as discussed in the previous section, and of ethnological and vegetational literature relevant to each of the three groups. Ethnological materials were reviewed for the most part before field work had commenced. They allowed many insights into the cultures of these groups and in some cases provided a preliminary discussion of some of the problems I would encounter in attempting to describe plant taxonomic systems. For example, the complexity of Bella Coola phytotaxonomy is sug- gested in the following statement by Mcllwraith (1948): - 36 - "In regard to plants, a difficulty lies in the fact that Bella Coola nomenclature is not always strictly botanical, Two or more distinct ferns, for example, may be grouped together on account of their similar use as food, and one name applied to them indiscriminately. Conversely, different terms are sometimes given to various parts of the same tree, the roots, the bark, the leaves, etc." Of particular value were the works of Swanton (1905a & b, 1908, 1911) and Newcombe (unpublished notes, 1897 - 1906) for Haida, Mcllwraith (1948) and Smith (1928)* for Bella Coola, and Teit (1906) for Lillooet. Botanical references include the floras of Calder and Taylor (1968), Hitchcock et al. (1955 - 1969), and to a lesser extent, Henry (1915). These were consulted throughout the study. Unfortunately, none of these covers the Bella Coola region, and as a result, some of the plants from Bella Coola were particularly difficult to identify botanically. One paper, "Speculations of the Growth of Ethnobotanical Nomen- clature" (Berlin 1971), proved to be exceptionally useful in the later stages of this research, by providing a directional focus for describing and explaining some of the observed characteristics and trends in the terminological systems studied. Other papers by Berlin,- Breedlove, and Raven (1966, 1968, no date), Berlin, Breedlove, and Laughlin (1970), and Raven, Berlin, and Breedlove (1971) have also been helpful in this regard. b) Field vegetation surveys At each of the three locations included in the study, an effort was A number of Smith's manuscripts (Smith 1920-22a, b, c, d, & e) are available at the National Archives of Canada in Ottawa, and would undoubtedly have proven useful, but were not known to me until the summer of 1973. - 37 - made to compile an inventory of the flora encountered in the area, particularly those species noted by the Indian people themselves to be of some significance. In Bella Coola and Lillooet most of the in- ventories were carried out during field expeditions with informants, while on the Queen Charlottes, vegetational survey work was accomp- lished during the course of the first summer of field work. Whenever possible, herbarium specimens of these plants were pre- pared, although I was limited both in time and collecting materials. Drying the plant specimens proved to be the most serious problem, both on the Queen Charlottes and at Bella Coola. On cool, wet days, the presses were set on an oil stove, and most plants dried within two days, but during hot spells, this method produced intolerably high temperatures in our living quarters, and had to be abandoned. Space limitations permitted us only a few hundred sheets of pressing card- board, so that the combination of poor drying facilities and lack of cardboard placed severe restrictions on our collections. Photographs of plants were also taken on many occasions during the field work. Some of the plants described by the informants could not be located in the field, and their identification could only be approximated from descriptions. In some cases, these descriptions were detailed, allow- ing an almost positive botanical designation, while in other cases, usually when a type of plant had not been encountered directly by the informant, the descriptions were too vague to allow even remote specu- lation about its botanical characteristics. I had to emphasize to myself constantly the necessity of not - 38 - harbouring preconceptions about the botanical status of aboriginal plant segregates. In this regard, having a botanical background was a distinct disadvantage, since it was extremely difficult not to immed- iately perceive native categories in terms of botanical species. How- ever, since every growing plant or plant specimen referred to in the study was botanically identifiable as a species or even subspecies, plant species names could be used without bias as labels for the types of plants involved in the study. The semantic range of a particular Indian plant taxon could then be defined in terms of the range of botanical species and subspecies shown by all plant specimens desig- nated by a single folk segregate. Not surprisingly, in a majority of cases, the folk segregates did in fact show a one-to-one correspondence with botanical species (see RESULTS). As mentioned earlier, botanical species also provide a "translation medium" for describing the semantic ranges of folk segregates in different Indian languages. Thus, famil- iarity with modern nomenclature and classification of plant species, while detrimental in maintaining a completely unbiased perspective in folk taxonomic studies, is useful in labelling and cataloguing folk segregates, and is also essential in comparative semantic studies. c) Consultation with native informants Interviewing has been the most enjoyable, and also the most challen- ging part of this project. A list of the native people consulted in the study is given in Appendix 1. As mentioned earlier, the informants are all elderly members of the Indian communities and all are bilingual to some extent. Most of the interviews took place at the informants' homes. They continued over several to many sessions, depending on the - 39 - extent of the informant's knowledge about plants. In each group, two or three people provided a major part of the information, although oth- ers contributed significant details. Whenever possible, the informants were taken on field trips to areas surrounding the villages, in order to observe plants in their natural habitats. At Lillooet, for example, all of the preliminary interviewing was done in the field. More commonly, fresh plant specimens would be collected on the morn- ing before an interview, and in the afternoon each plant would be dis- cussed with the informant to learn details of its Indian name, its relationship to other plants, seasonal variation, growth features, habitat, and cultural significance. Care was taken not to refer to the English common names of the plants unless they were already known by the informant. Collections of fresh plants were supplemented with her- barium specimens, or even photographs, of plants not readily available. Verifying the Indian plant names and information was accomplished by asking the informant about the same plant in a later session or through discussions with other informants. The first technique was especially effective in the Haida study, where I was able to repeat the interviewing process over two consecutive years. I did find a few inconsistencies from one year to the next, but for the most part, the data collected over the two years were remarkably consistent. Historical records of plant names and terms previously collected by other field workers were also useful in corroborating my own data. In many cases, the plants under discussion had not been seen or talked about by the informants for a long time — sometimes for as long - 40 - as 50 years. Under these circumstances, the names and characteristics of the plants did not always come readily to mind. Often, additional details about a plant would be remembered after several sessions of talking and thinking about plants, or after consultations with friends or relatives. Sometimes, a plant was recognized only after some previously elicited details about it were furnished, such as its name or use. This was done only as a last resort, when it was obvious that the informant would not remember anything without assistance. At times, an informant would specifically ask what another informant had said about a plant. When told, he would often volunteer supplementary information. I do not feel that providing this kind of stimulus produces biased or false re- sults, since each informant was sincere in his attempts to tell only "the truth". Information volunteered by myself or others resulted in a negative or non-committal response as often as a confirmation. Realis- tically, the "suggestion" technique provided a significant amount of data to which I would not otherwise have had access. All of the informants were pleased at being able to refresh their memories on "Indian plants", and all made very positive comments about having this information recorded and written down.* The informants Were * For each of the groups where I have done field work, I have provided the informants and the local Band Councils with copies of the infor- mation I obtained on the names and uses of plants, organized in a popularized format, and accompanied by tapes of the plant names pro- nounced by the informants. These have been greatly appreciated, and, as I have heard indirectly, are a source of considerable pride to the people who originally provided the information. - 41 - given an hourly payment of about $2.50,* but in no case was monetary gain a primary motivation for providing information. Each of the Indian languages in the study has a unique inventory of consonants, vowels, and phonetic sequences. Many of the sounds in these languages do not occur in English speech. Accurate transcriptions of the plant names would have been impossible without the help and advice of several linguists presently involved in field work in these areas. They include: Robert Levine, doctoral student at Columbia Univ- ersity, New York, who is currently working on the Skidegate dialect of Haida: Dr. Aert Kuipers, Professor of Linguistics at the University of Leiden, Holland, a Salish language specialist; two of Dr. Kuipers' students, Henk Nater and Jan van Eijk, working on Bella Coola and the Fraser River "dialect" of Lillooet respectively; and Randy Bouchard, of the B.C. Indian Language Project, Victoria, B.C., who has provided general assistance throughout this project, particularly with the Salishan languages. During the course of their own field work, these linguists were able to transcribe all of the plant names directly from the native informants, either in conjunction with my elicitation sessions, or independently. Each of them spent considerable time and energy on my behalf, not only in the original transcription process, but also in subsequent checking and revision of the plant names. Undoubtedly, further corrections and additions *-This was more or less standard payment for field work in linguistics and anthropology in the Pacific Northwest, although there has been a recent trend towards higher rates. Some field workers pay over $5.00 per hour at present. - 42 - to the plant names will be necessary over the course of the next few years, but I feel confident in presenting the data as linguistically accurate and complete within the practical time limits of this study. The orthographies used for writing Indian plant names in this study have been adopted from a series of practical writing systems developed and described by Randy Bouchard with the assistance of a number of native language specialists (cf. Bouchard 1970, 1971, 1972). Practical orthographies are currently being used by native peoples in several Salish groups for recording their own languages. In the case of Haida, a number of modifications to the system originally described by Bouchard have been made, with advice from Robert Levine (cf. Levine 1973). Descriptive keys to the pronunciation of the orthographic sym- bols in Haida, Bella Coola, and Fraser River Lillooet are given in Appendixes 2, 3, and 4 respectively. Tape recording, while not a substitute for actual speech, proved valuable in the study, both as a note-taking device during elicitation sessions, and in providing a permanent record of the plant names in the three languages. In the latter capacity, tapes are useful for checking or confirming linguistic transcriptions, but in most instances, absolute final decisions on phonetic designations can be made only in the field. In the present study, a Uher 4400 stereo recorder was used. Copies of all of the tapes made in conjunction with the study have been filed with the B.C. Indian Language Project, sponsered by Randy Bouchard, and with the National Museum of Canada in Ottawa. A number of informants were distinctly apprehensive of the tape recorder, and some of the Haida - 43 - people requested that I not use it. In these cases, I asked other informants to repeat any new or unique names given by these people in order to have a complete taped record of the plant names. d) Synthesis of data The cognitive data resulting from the elicitation process take two forms: unstructured statements and opinions about relationships between plants, derived from informal conversations; and, a series of actual native names applied to different kinds of plants. The first type of information is generally variable and incomplete. It was con- sidered in formulating and influencing ultimate conclusions about folk categories in the three groups, but was too irregular to be regarded as anything other than supplementary information. The series of plant names, on the other hand, show constancy and stability, and in the context of other types of related data, can be construed as a discrete set of verifiable information, to which quantitative sorting techniques can be applied, and from which trends and generalities can be realized. These names formed the major resource data in this study. Various factors must be considered in conjunction with the Indian plant names, including details of their source and terms of application; their origin, when known; and their semantic range. Also relevant are data on the habitats, distributions, and botanical status of the kinds of plants referred to by the folk segregates, and information com. >.rning the cultural significance of these plants. Consideration of these parameters, within the context of three unique cultural and linguistic groups (one of which is further divided into two - 44 - dialectic units) requires the multi-dimensional sorting of a vast quantity of diverse data. To meet this requirement, a coding system was designed for computer- ized sorting of the various details associated with the folk segregates. Three separate sets of standard Fortran data cards were employed to accommodate these details. The first set includes the folk segregates themselves and information relating directly to them, such as language and dialect, descriptive characteristics of the terms, assumed origin, and botanical equivalence (see Table 1). The second set contains infor- mation relating to the various botanical taxa delimited by the folk segregates (see Table 2). The third set includes information on the cultural significance of the plants in each of the linguistic groups (see Table 3)i Keys to the various codes in the system are provided in Table 4. The card layouts, coding systems, and sorting program were designed so that additions and alterations can be made at any time, as new infor- mation becomes available. This means that the comparative base for this type of study can be expanded to include as many different linguistic and dialectic groups in the Pacific Northwest as can be adequately de- scribed botanically. - 45 - Table 1. Card design for information directly related to folk segre- gates for plants. Field Column Description of number number information 1 1 CARD SET LABEL - All cards in this set are labelled "A" in this column. 2 2-3 LANGUAGE - Each different language group in the study is given a separate code (see Table 4). 3 4 DIALECT - When more than one dialect is represented in a language, letter codes are used when necessary to refer to a specific dialect (see Table 4). 4 5- 7 PLANT SEGREGATE NUMBER (right-adjusted) - This number is unique for each different term applied to a part of a specific plant, a plant taxon, or a group of taxa within a given language and dialect. 5 8-43 FOLK SEGREGATE - The native terms for different kinds of plants are written in a modified practical ortho- graphy (see Appendixes 2, 3, and 4). Underlining is indicated in the computor printout by a slash follow- ing a letter (e.g., k = k/), and accent marks are printed as an asterisk *. English glosses for the terms are also given when known. 44 CONTINUING INFORMATION - When the folk segregate and its English gloss is too long to fit the alloted number of columns in one card, a "CM is placed here, and a second card is used to continue the information. In this case, the first four fields are repeated on the second card, as identification. 6 45 PART OF PLANT REFERRED TO BY SEGREGATE - Specifies appli- cation of the term, either for a specific part of a plant (e.g., fruit, bark, or cambium) or for a speci- fic growth stage or state (e.g., immature or prepared for use). (See Table 4 for specific codes used.) 7 46-47 REFERENCE SOURCE - Each different source of folk plant segregates is indicated by a unique code (see Table 4). 8 48-67 DESCRIPTIVE CHARACTERISTICS OF THE PLANT SEGREGATES (20 subfields). - 46 - Field Column Description of number number information 8a 48 8b 49 8c 50 Uniqueness - Terms which have no meaning other,than as plant segregates, or terms including segments which have no meaning (according to native informants) are indicated by a "1" placed in this column. Generic term polysemous with or partially inclusive of a term of broad taxonomic standing (e.g. plant, berry, tree, or grass) - indicated by a code (see Table 4). Compound lexeme with a specifying modifier (e.g. 'Haida- apples', 'real-saskatoons') - indicated by a "1" placed in this column. 8d 51 Sex differentiation indicated - If sexual status is specified or implied in a folk segregate, an M (male) F (female), or N (neuter) is placed in this column. 8e 52 8f 53 Colour - If specified or implied in a segregate, an ap- propriate symbol is given (see Table 4). Texture - If a term indicates or implies texture (e.g. 'soft', 'rough', 'sticky'), a "1" is placed in this column. 8§ 54 Shape or growth form - If a segregate indicates or im- plies the shape or growth form of a plant (e.g. 'climbing-plant', 'bunched-up*, 'strings-along'), a "1" is placed in this column. 8h 55 Taste, smell, touch, or sound - If a plant term refers to any of these factors (e.g. 'sweet-berry', 'smells- nice', 'burning'), a "1" is placed in this column, 8i 56 Reference to anatomical feature (e.g. 'rain's-navel', 'goose-tongue', or 'man-foot') - indicated by a "1" in this column. 8J 57 Comparison of plant to some substance or object (other than to an anatomical feature) (e.g. 'crow's-lace', 'tree-biscuit', 'Raven's-canoe') - indicated by a "1" in this column. 8k 58 Other quality of plant indicated by name - If the term includes some other descriptive feature not covered in Fields 8e to 8j (e.g. location within a plant: 'inside'; or state of plant: prepared'), a "1" is placed in this column. - 47 - Field Column Description of number number information 81 59 Plant name includes reference to habitat — appropriate code is placed in this column (see Table 4). 8m 60 Plant name includes reference to use - An "H" indicates use by humans (e.g. 'bow-tree', 'rope-plant'), and an "A" indicates use by an animal (e.g. 'Raven's- canoe', 'grizzly's-highbush-cranberries'). 8n 61 Term indicates some action or process associated with the plant (e.g. 'bustling'. 'hold-in-the-mouth', 'buy- ing') - A "1" is placed in this column. 8o 62 Name indicates some human attribute of plant - An "R" denotes use of a kinship term (e.g. 'pilot-biscuit's- grandmother', 'tobacco-mother'), and an "H" indicates some other human attributes (e.g. 'child', 'thief'). 8p 63 Name indicates association of plant with an animal - appropriate code is placed in this column (see Table 4). 8q 64 Name indicates association of plant with a supernatural being (e.g. 'hermaphrodite-plant', 'ground-ghost') - A "1" is placed in this column. 8r 65 Name Indicates association of plant with a natural phen- omenon or astronomical feature (e.g. 'rain-leaves', 'forest-cumulus-cloud') - A "1" is placed in this column. 8s 66 Name includes segregate for a plant - An "S" indicates inclusion of another segregate applied to the same kind of plant (e.g. in Haida, Ihk'iit-lhk'aayii 'lhk'iit-branches' contains lhk'iit,