Household Archaeology on the Northwest Coast 9781789201789

Citation preview

Household Archaeology on the Northwest Coast edited by Elizabeth A. Sobel, D. Ann Trieu Gahr, and Kenneth M. Ames

INTERNATIONAL MONOGRAPHS IN PREHISTORY Archaeological Series 16

© 2006 by International Monographs in Prehistory All rights reserved Printed in the United States of America All rights reserved ISBN-13: 978-1-879621-38-1 ISBN-10: 1-879621-38-X (Paperback) ISBN-13: 978-1-879621-39-8 ISBN-10: 1-879621-39-8 (Hard Cover)

Library of Congress Cataloging-in-Publication Data Household archaeology on the Northwest Coast / edited by Elizabeth A. Sobel, D. Ann Trieu Gahr, and Kenneth M. Ames. p. cm. -- (Archaeological series ; 16) “Volume started as a symposium on household production in Northwest Coast societies, held at the 64th Annual Meeting of the Society for American Archaeology in Chicago, Illinois in 1999”-Pref. ISBN-13: 978-1-879621-38-1 (pbk. : alk. paper) ISBN-10: 1-879621-38-X (pbk. : alk. paper) ISBN-13: 978-1-879621-39-8 (lib. bdg. : alk. paper) ISBN-10: 1-879621-39-8 (lib. bdg. : alk. paper) 1. Indians of North America--Northwest Coast of North America--Dwellings. 2. Indian architecture--Northwest Coast of North America . 3. Indians of North America--Northwest Coast of North America--Antiquities. 4. Excavations (Archaeology)--Northwest Coast of North America . 5. Social archaeology--Northwest Coast of North America . 6. Northwest Coast of North America--Antiquities. I. Sobel, Elizabeth A. II. Gahr, D. Ann Trieu. III. Ames, Kenneth M. IV. Society for American Archaeology. Meeting (1999 : Chicago, Ill.) V. Archaeological series (Ann Arbor, Mich.) ; 16. E78.N79H68 2006 979.5’01--dc22 2005036000

This book is printed on acid-free paper. ∞ International Monographs in Prehistory P.O. Box 1266 Ann Arbor, Michigan 48106-1266 U.S.A.

Table of Contents

List of Contributors ................................................................................................................ v Preface.................................................................................................................................. vii 1. Introduction D. Ann Trieu Gahr, Elizabeth A. Sobel, Kenneth M. Ames ........................................... 1 2. Thinking about Household Archaeology on the Northwest Coast Kenneth M. Ames .......................................................................................................... 16 3. Houses and Domestication on the Northwest Coast Yvonne Marshall ........................................................................................................... 37 4. Architects to Ancestors: The Life Cycle of Plankhouses D. Ann Trieu Gahr ......................................................................................................... 57 5. A Chief’s House Speaks: Communicating Power on the Northern Northwest Coast Gary Coupland .............................................................................................................. 80 6. Temporality in Northwest Coast Households Colin Grier ..................................................................................................................... 97 7. Of a more Temporary Cast: Household Production at the Broken Tops Site David V. Ellis ............................................................................................................... 120 8. The Tsimshian Household through the Contact Period Andrew Martindale ..................................................................................................... 140 9. Household Prestige and Exchange in Northwest Coast Societies: A Case Study from the Lower Columbia River Valley Elizabeth A. Sobel ....................................................................................................... 159 10. Households at Ozette Stephan R. Samuels .................................................................................................... 200 11. Formation Processes of a Lower Columbia River Plankhouse Site Cameron McPherson Smith ........................................................................................ 233 12. Households and Production on the Pacific Coast: The Northwest Coast and California in Comparative Perspective Jeanne E. Arnold ......................................................................................................... 270

List of Contributors Kenneth M. Ames Department of Anthropology Portland State University Portland, Oregon 97207 Jeanne E. Arnold Department of Anthropology and Cotsen Institute of Archaeology University of California Los Angeles, California 90095 Gary Coupland Department of Anthropology University of Toronto Toronto, Ontario, Canada, M5S 1A1 David V. Ellis Archaeological Investigations Northwest, Inc. Portland, Oregon 97236 D. Ann Trieu Gahr Department of Anthropology Southern Illinois University at Carbondale Carbondale, Illinois 62901 Colin Grier Department of Anthropology and Sociology University of British Columbia Vancouver, British Columbia, Canada V6E 2E2 Yvonne Marshall Department of Archaeology University of Southampton Southampton, United Kingdom SO17 1BF Andrew Martindale Department of Anthropology and Sociology University of British Columbia Vancouver, British Columbia, Canada V6E 2E2 Stephan R. Samuels U.S. Bureau of Land Management, Coos Bay District North Bend, Oregon 97459 Cameron McPherson Smith Department of Anthropology Portland State University Portland, Oregon 97207 Elizabeth A. Sobel Department of Anthropology Portland State University Portland, Oregon 97207 v

Preface This volume started as a symposium on household production in Northwest Coast societies, held at the 64th annual meeting of the Society for American Archaeology in Chicago, Illinois in 1999. However, as the symposium and later the book developed, it was impossible to limit the papers to household production. Other significant issues kept coming up including site formation processes, exchange systems, household origins, post-contact change, household social reproduction, and social inequality, to name just a few of these issues. Thus, this volume is about the archaeology of households in general and not just household production on the Northwest Coast. Since the late 1970s, household archaeology has become a central methodology for archaeologists interested in the social, political and economic organization of chiefdoms and states. In 1982 Hayden and Cannon (1982) placed households at the center of cultural and social complexity among hunter-gatherers, and by the late 1980s, household archaeology was a key theoretical and methodological framework for research on the development of permanent social inequality and complexity in Northwest Coast societies. Household archaeology provides a common intellectual framework within which Northwest Coast archaeologists with varied theoretical orientations can work productively on common research questions, as the papers in this volume testify. The symposium in which this volume began was planned, in part, to take stock of Northwest Coast household archaeology after more than a decade of sometimes very difficult work. We wanted to begin answering a number of questions. What have we learned? What do we know that we want our colleagues to know? In what directions should we go now? It was also intended as a forum for some of the best recent household archaeology on the Coast, although we certainly do not have it all here. Finally, we wanted to demonstrate that household archaeology as a methodological and theoretical approach is very productive for understanding the evolution of social complexity, regardless of underlying economy. Household production remains central to this volume. When anthropologists and archaeologists have sought to explain the evolution of complexity on the Northwest Coast, they traditionally explained it as a consequence of the Coast’s once seemingly inexhaustible natural abundance. More recently, scholars have recognized that the Coast’s environment has varied greatly through time and space, thus while it may in aggregate have been abundant, it was not abundant everywhere and there were probably times when it was not at all abundant. However, even if the Coast was always rich, that would not tell us how that richness was translated into food, tools and prestige items. That translation was accomplished, over the past 3,500 years, by large, multi-family households. Despite the Coast’s rich ethnographic literature, we do not have a full picture of how those households worked as units of production and consumption, particularly over several centuries. Only archaeology can contribute new information to that picture, although we can bring our new understanding to the ethnographies and see them in a new light. Inevitably the papers do not cover everything. Our geographic coverage is uneven. We have two papers for the northern Coast, no papers for the central Coast, only one paper for Vancouver Island, and only one paper for that portion of the southern Northwest Coast known as the Gulf of Georgia, which is archaeologically the best known portion of the Northwest Coast. The Lower Columbia River region, often ignored in books on the Northwest Coast, is heavily represented. The absence of central Coast papers resulted when several symposium participants who work in that region were unable to participate in this volume. These individuals are David Huelsbeck, Joanne Greene, and R.G. Matson. Elizabeth A. Sobel D. Ann Trieu Gahr Kenneth M. Ames

vii

Acknowledgements This book is an outgrowth of a symposium at the 64th annual meeting of the Society for American Archaeology held in Chicago, Illinois in 1999. We would like to thank the participants of that symposium including several who were not able to contribute papers to the final volume - R.G. Matson, Brian Hoffman, David Huelsbeck, and Joanne Green. We are grateful to Robert Whallon and our contributors for their patience as this volume wended its way through two dissertations, two weddings, two pregnancies, two adoptions, the arrival of one granddaughter, and all the other exigencies of life. We are especially thankful to Robert Whallon for his early interest and enthusiasm, for shepherding us through the process of producing the book, and ultimately for publishing it. In part, this volume is an outgrowth of Portland State University’s Wapato Valley Archaeology Project (WVAP). All three of us developed our work on Northwest Coast plankhouses in the context of this research project. Consequently, we would like to acknowledge and thank the many people who have supported and assisted the WVAP, particularly the Chinook Tribe, including the Tribal Council and Culture Committee. They have been unfailing in their interest and support. Gary Johnson, the current Tribal Chair, strongly supported the work at Cathlapotle since its inception. Tony Johnson, current chair of the Cultural Committee, has also supported us throughout. We also acknowledge the interest of the Grand Ronde Tribe of Oregon. We appreciate the continuing support of the U.S. Fish and Wildlife Service, particularly archaeologist Anan Raymond, Program Director, and archaeologists Alex Bourdeau, Virginia Parks, and Nick Valentine. We thank the people of Ridgefield, Washington and Scappoose, Oregon, especially Don Meyer. Field research of the WVAP has been supported by grants from the National Science Foundation, Wenner-Gren Foundation, Northwest Research Obsidian Studies Laboratory, Portland State University, Southern Illinois University at Carbondale, University of Michigan, and contracts with the U.S. Fish and Wildlife Service. The excavations were all accomplished by eight field schools and their staffs. Our approach to household archaeology has been stimulated and improved by discussions with many colleagues and mentors including Virginia Butler, David Ellis, Doug Wilson, Cameron Smith, Knut Fladmark, R. Lee Lyman, R. G. Matson, Gary Coupland, Richard Ford, John O’Shea, John Speth, Lee Newsome, and George Crothers. We apologize to those we have missed. The late Wayne Suttles had a major early impact on the research questions and methodology of the WVAP. In ways too complicated to detail, this research is an archaeological outgrowth of Suttles’ dissertation work. For their support and endurance during this and our many academic endeavors, we thank our families including Jim, Xuong, Mariella, Wayne, Jane, Joanna, and the three equally beautiful babies – Ruby, Carly Jo, and Thien.

ix

Introduction D. Ann Trieu Gahr, Elizabeth A. Sobel, Kenneth M. Ames

Houses have long been a primary focus in Northwest Coast anthropology and archaeology. The rich material culture of native Northwest Coast peoples often reached its grandest expression in the architecture and furnishings of plank houses, and in the social, political, and economic dimensions of household life. Ethnohistoric and ethnographic studies of cultures along the entire Coast consistently portray the house as the physical manifestation of the household and of its social rank. Within the community, a house’s location and size often signaled household rank while its interior arrangements reflected status distinctions among household members. Many houses were substantial structures that required skilled construction techniques and large labor investments. They were designed to simultaneously serve as “dwellings …food processing and storage plants … workshops and recreation centers … temple[s] and theater[s] [and] fortresses” (Suttles 1991:214–219). The centrality of the house and household in Northwest Coast anthropological research, both archaeological and ethnological, thus flows naturally from their pivotal role in Northwest Coast societies. Over the past thirty years, houses and households have come to play major roles in archaeological research and theory building on the evolution of Northwest Coast societies. Research into Northwest Coast households raises important theoretical and methodological issues relevant, we think, not only to Northwest Coast studies but also to a broad array of issues including the origins and evolution of social inequality and the range and diversity of hunter-gatherers (e.g. Ames 2004). This introductory chapter first reviews the history and development of household archaeology on the Northwest Coast and then introduces this volume.

the Northwest Coast’s cultural landscape. The earliest journal accounts, such as those of Captain Cook in 1778 (Beaglehole 1955), contain descriptions and pictures of these houses. The pictures range from quick sketches to detailed paintings and engravings. The late nineteenth century produced a remarkable corpus of photographs of house exteriors and interiors. It also produced the first anthropological descriptions of houses and households. Most of our current understanding of Northwest Coast households comes from the vast ethnographic and ethnohistoric literature for the Coast that has accumulated since the late nineteenth century. While ethnographers and other observers recognized that households were the basic units of Northwest Coast economic production, the household was not their primary focus of inquiry. Rather, kin groups and kinship, not households, were among the primary foci of anthropological interest. Most native Northwest Coast households were comprised of extended, corporate groups. These groups controlled estates of corporeal and non-corporeal property. Corporeal property included regalia, canoes, houses, and rights to exploit resources in particular places, among other things. Non-corporeal property included songs, dances, oral traditions, rights to certain spirit helpers, high statuses and the titles and privileges that went with those statuses. It was clear to anthropologists by the late nineteenth and early twentieth centuries that house size, house position in a village, house decoration and patterns of interior use mapped social relationships within and between households. These maps were particularly visible among central and northern groups such as the Kwakwaka’wakw (e.g. Boas 1921), Nuu-chah-nulth (Nootka) (Drucker 1951), and Coast Tsimshian (Boas 1916), and during events such as the potlatch. This brief description masks considerable diversity, fluidity and ambiguity. The relationships among kin-groups, households and estates were not tidy ones. In some cases, a household might contain more than one extended kin group (e.g.

Houses, Households and Production in Northwest Coast Ethnography Before European settlement, large wooden dwellings were among the most visible features of 1

Gahr, Sobel, & Ames Adams 1973); in others a single lineage might span two or three households (e.g. Coupland 1996a). Given residency rules, an extended kin group could have members in several households, the members retaining rights in their common natal household estate. The extended kin-groups themselves varied. On the northern Coast, they were matrilineages, while on the central and southern Coast they were what Jorgensen (1980) terms “patridemes” which are “lineage-like” bilateral descent groups practicing patrilocal residence. They were lineage-like because, among other things, they controlled estates. Affinal relationships could be at least as important as agnatic ties for access to the resources of estates (e.g. Suttles 1960). People, particularly commoners, could shift from household to household. Our intent here is not to delve any deeper into Northwest Coast social organization but to make the point that most Northwest ethnographers conceived Northwest Coast social and economic organization through the lens of kinship. With this emphasis on kin groups, anthropologists viewed households as composed of extended families that were members of extended kin groups. Ethnographers were also interested in Northwest Coast houses, primarily their architecture and construction. Scholars sometimes provided quite detailed (and important) information although, with the passage of time, these details were increasingly reliant on the memories of elderly informants who themselves may not have seen a plank house or lived in one. Anthropologists were also interested in the origins of the plankhouse and the geographic distribution of its regional variants (see Trieu this volume). This distribution was generally thought to be the result of the migrations of ethno-linguistic groups (e.g. Waterman and Greiner 1921). This interest in architecture and construction continues (e.g. Nabokov and Easton 1988; Vastokas 1966) as does interest in the Northwest Coast house as art (e.g. MacDonald 1983). In a departure from much of this work, recent research has looked at local variation in house form. Suttles (1991) links house form among the Coast Salish to household size and organization. He argues that Coast Salish house architecture was designed to accommodate fluctuating household sizes in ways that houses to the north and south could not, linking the organizational flexibility of households to house form. Hajda (1994) shows considerable variation among Chinookan plankhouses during the Early Modern Period1 along the Lower Columbia River while Mackie and Williamson (2003) show equivalent variation in house form

among the houses of a single small Early Modern Nuu-chah-nulth village. This recent research is important because much of our understanding of Northwest Coast culture is based on distillations of the ethnographies into descriptions of what is “typical.” There are basic cultural and social themes that play through the Coast, but like a complex musical composition, these themes are more often played as elaborate variations and even improvisations than as straightforward melodies. Issues of household production, and production generally, were also not important in anthropological accounts of the Coast before 1960. This was largely because of the belief that the Northwest Coast environment was so munificent that the organization of food-getting was not an issue (e.g. Codere 1950). Ethnographies often contain detailed descriptions of aspects of production, including foodgetting techniques and technology, the division of labor, roles of slaves and elites, specialization and so forth, but these are often scattered through the monograph. The first truly coherent, focused description of the organization of production on the Coast was Kalervo Oberg’s dissertation (The Social Economy of the Tlingit Indians), which was completed at the University of Chicago in 1933. However, this work was largely unknown and unused until published in 1973. In it, Oberg analyzed production primarily at the household level. Wayne Suttles’ dissertation (Suttles 1951) also examined production at the household and family levels among Straits Coast Salish. Suttles’ dissertation was not widely available until 19742. A third ethnography, which appeared in 1973 (Adams 1973), focused on households, the household cycle and household recruitment among the Gitksan of northern British Columbia. These three ethnographies, coincidently all published at approximately the same time, are the first ethnographic analyses of household production on the Northwest Coast. These works were followed in 1988 by an edited volume that focused on Northwest economies, which made clear the economic centrality of the household (Mitchell and Donald 1988). Issues of household production and distribution were also implicit in the debate engendered by Suttles’ (1960, 1962, 1968) and Vayda’s (1961) argument that the Northwest Coast’s environment, while rich, was also quite variable, and Northwest Coast social institutions, particularly marriage, were social mechanisms of coping with this variation. Piddocke (1965) carried this argument to a functionalist extreme, drawing upon Boas’ descriptions of the Kwakwaka’wakw numaym or household 2

Introduction and ethnographic records, and anticipated current concerns with houses as archaeological deposits (Samuels this volume; Smith this volume). They also treat the house as an analytical unit by reporting artifacts (although not fauna) by house. However, their primary focus was on the houses as structures, not the households that lived in the houses. Interest in Northwest Coast houses and households as analytical units developed in the late 1960s and early 1970s as a consequence of excavations treating houses as units of analysis and interpretation, and due to growing interest in the evolution of social complexity on the Northwest Coast and among hunter-gatherers. The first excavations focusing on houses as analytical units began simultaneously in 1970, with excavations at the Ozette site on the coast of Washington State (Samuels and Daugherty 1991) and at the Richardson Ranch site (Fladmark 1974) on the Queen Charlotte Islands (a.k.a. Haida Gwaii). Ozette was a Makah village that had been partially buried beneath a catastrophic mudslide in A.D. 1700. The mudslide produced a wet, anaerobic environment that preserved organic matter, including most of three houses. For a variety of reasons (Samuels and Daugherty 1991) these houses became the central analytical units around which the project was organized (see Samuels this volume). Fladmark, on the other hand, elected to test excavate a nineteenth century Haida house, partially to develop an archaeological baseline against which to compare prehistoric components but also to archaeologically explore the impacts of contact on the Haida and their response to it. Among other things, he explored the relationships between the spatial distribution of artifact classes, social status, and the development of a particular specialization, argillite carving, for which the nineteenth century Haida are famous. So far as we are aware, these are the first excavations on the Northwest Coast to explicitly investigate the organization of household production and the relationship between production and rank. In addition to their focus on houses, both projects saw household archaeology (although neither used the term) as the only means to test the famous ethnographic and historic accounts of Northwest Coast culture, society and economy. Ozette predates contact by less than a century, and Richardson Ranch is contemporary with major ethnographic accounts for the northern Northwest Coast. Fladmark was explicit in seeing archaeology as the only way of testing and correcting the ethnographies and generalizations based upon them. This testing

group. This debate (e.g. Drucker and Heizer 1976) drew focused anthropological attention to the Coast’s ecology and economies for the first time. Claude Lévi-Strauss also drew upon Boas’ descriptions of the numaym to formulate his concept of House societies (see Ames, this volume; Marshall, this volume; and Trieu, this volume) published in English in 1982 (Lévi-Strauss 1982). In the House society concept, Lévi-Strauss grappled with the ambiguities of how household members were recruited and held together on the Coast. He emphasized that kinship was not a key element in the organization and persistence of Northwest Coast households, which he labels “Houses.” Rather, the physical house itself and its associated estate were central. Even before the English publication of Lévi-Strauss’ ideas, interest in both Northwest Coast houses and households had shifted from sociocultural anthropologists to archaeologists. The Archaeology of Northwest Coast Houses and Households Until relatively recently, houses were not a common focus of archaeological research or excavations on the Northwest Coast. It should be stressed here, however, that archaeology on the Northwest Coast is largely a post-World War II phenomenon (e.g. Carlson 1990), in contrast with ethnographic research. Initially, because so little archaeology had been done, houses and household production were not high on any list of pressing archaeological research questions. When encountered, houses were usually treated as clusters of features rather than as units of analysis—household archaeology requires the latter. In the Northwest or Cascadia3 generally (coast and interior), houses were of interest primarily as markers of the development of the historic ethnographic pattern; they were indicator fossils of a particular life way. What was important was their presence or absence. There was little or no interest in the households that lived in the houses. De Laguna’s pioneering work in southeast Alaska in the late 1940s and the 1950s is a notable exception (de Laguna et al. 1964), anticipating many aspects of household archaeology as it later developed on the Coast. De Laguna and her associates conducted test and block excavations of houses at the Old Town site in Yakutat Bay (de Laguna 1972). The houses probably date to the late eighteenth and early to midnineteenth centuries. The researchers attempted to reconstruct details of house construction, drawing upon both the excavations and the ethnohistoric 3

Gahr, Sobel, & Ames Subsequent work has explored aspects of household production and consumption (Mitchell and Donald 1988), including specialization and its linkages to inequality (e.g. Ames 1996; Chatters 1989; Grier 2001; Huelsbeck 1988, 1989, 1991; Wessen 1988, 1991), labor and task organization (Ames 1985, 1995, 1996, 2001; Ames and Maschner 1999; Coupland 1985b, 1988a, 1996a; Matson 1992, 1996), the intensification of production (Coupland 1988a; Matson 1985) and issues surrounding the excavation and analysis of Northwest Coast plankhouses (Ames et al. 1992; Matson 2003; Mauger 1991; Samuels 1991, this volume; etc.). Coupland and Banning (1996) expanded the geographic scope of these discussions, editing a volume on the archaeology of “large domestic structures.” Addressing these social and economic issues has also forced archaeologists working on the Coast to disentangle the site formation processes that plankhouses produce. Plankhouse sampling (e.g. Matson and Coupland 1995) and taphonomy have emerged as major issues in Northwest Coast archaeology (e.g. Blukis-Onat 1985; Matson 2003). This ongoing work has usually been done within one of three major theoretical currents regarding the basic nature of the Northwest Coast household (Ames this volume). The most widely applied approach defines the household as a functional unit, an idea that archaeologists first used about three decades ago (Flannery and Winter 1976; Hill 1970; Sheets 1979) and employed more extensively in the 1980s (e.g. Wilk and Rathje 1982). A second approach defines Coast households as residential corporate groups. The third and least developed theoretical current uses Lévi-Strauss’ concept of the House society (1982), conceptualizing the Northwest Coast household as a “House”—a particular type of corporate group that owns intangible and tangible property, and maintains itself over the long-term by transmitting its property across generations within a real or ideal lineage. Most authors in this volume employ one of these three frameworks, explicitly or implicitly. However, Ames points out that the three theoretical frameworks are not mutually exclusive and can be productively integrated. His paper employs the three perspectives in a complementary fashion, perhaps laying the groundwork for increasing integration of these various theoretical frameworks in Northwest Coast household studies.

remains a central role of household archaeology and archaeology more generally on the Coast. However, starting in 1975, the major impetus for Northwest Coast household archaeology was growing interest among archaeologists in explaining Northwest Coast social complexity, including its permanent social inequality (Ames 1981, 1983; Fladmark 1975; Matson 1983; Schalk 1977). At the same time, archaeologists more generally became interested in social complexity among hunter-gatherers (Hayden 1981; Hayden and Cannon 1982; Koyama and Thomas 1981; Price 1981; Price and Brown 1985) and households (Flannery 1976; Wilk and Rathje 1982). These two general trends—complex hunter-gatherers and household archaeology—were persuasively connected by Hayden and Cannon (1982) who argued that the development of closed corporate groups, which included some kinds of households, was fundamental to the evolution of social complexity. It was clear by the early 1980s that the household was the cockpit within which complexity on the Northwest Coast had evolved. Coupland’s dissertation field research was the first explicit application of household archaeology in a Northwest Coast excavation (Coupland 1985a). In the early 1980s, he sampled three of the 10 houses at the Paul Mason site, dated to ca. 1450–950 B.C. Coupland’s primary focus was house size as a proxy measure of relative household status. In historic Northwest Coast communities, particularly on the central and northern Coasts, the physical size of a house was a pretty clear indicator of the household’s relative status. It was a direct measure of the household’s ability to field labor. The houses at Paul Mason are all quite small, leading Coupland to argue that social organization at that time was still egalitarian. Relative house size continues to be used by archaeologists working on the Coast as a measure of relative household prestige and status (Acheson 1991; Archer 2001; Maschner 1992; Sobel 2004). Coupland’s own work continues to emphasize households as the central organizational entities for understanding Northwest Coast social and economic evolution (Coupland 1985b, 1988a, 1988b, 1996a; Coupland and Banning 1996; Coupland et al. 1993, 2003). In 1985 Ames argued that the development of social inequality on the Northwest Coast was shaped by the Domestic Mode of Production (DMP) (Sahlins 1972), focusing on household organization and how the household was integrated into interaction spheres (Ames 1985:157). As Ames discusses in this volume, the DMP was based on theories of peasant household economies.

A Note On Terminology The reader of this volume needs to be aware of 4

Introduction two sets of terms: those referring to the dwellings within which Northwest Coast households lived, and terms for the households themselves. Northwest Coast specialists generally call the dwellings plankhouses, although the more generic term longhouse is gaining currency among Indian/First Nations peoples, the general public and, to a lesser extent, anthropologists and archaeologists. This is happening, at least in part, we think, because modern ones are used primarily for ceremonial purposes and, in western North America, “longhouse” is widely applied to Native ceremonial structures which vary greatly architecturally, but are usually long houses. Plankhouses are large post and beam structures clad, roofed and sometimes floored with wood planks. Western red cedar (Thuja plicata) was the tree most commonly used, and is famous for the relative ease with which it can be split into thin, wide planks (e.g. 5 cm thick x 100 cm wide by 5–10 meters long). The dwellings themselves varied considerably, as the reader of this volume will learn, in size and construction details, but they had some very basic commonalities: they were square to rectangular with either gabled or sloping “shed” roofs; they were often semi-subterranean to some degree; floors were either planked or earthen (sometimes covered with sand, clay, fine gravels, crushed shell, matting); their interiors contained one or more hearths; and they usually had sleeping platforms or benches aligned along two or more interior walls. The second and more confusing set of terms refers to the group that lives inside these impressive structures. Anthropologists generally recognize that Northwest Coast societies featured several nested residential units. However, no single term for each unit is consistently used, creating confusion and perhaps unnecessary debate. We approach this terminological morass by reviewing the four basic residential units that characterized Northwest Coast societies as indicated by ethnographies, and the labels that anthropologists have applied to these units. The terms village, town, and community are widely applied to the people who lived in the same winter settlement and who “acted as a social unit at least part of each year,” mainly in the context of ceremonial and defensive activities, and less often in the context of subsistence activities (Mitchell and Donald 1988:294; chapters in Suttles 1990a). Some researchers also refer to this unit as a local group (Mitchell and Donald 1988:294). Each winter settlement consisted of one or

more plankhouses. Most researchers generally apply the label household to the social unit that co-resided in each single plankhouse (e.g. Ames and Maschner 1999; Matson and Coupland 1995; Suttles 1990b:464). However, Mitchell and Donald (1988:298) refer to this social unit as the extended household; Oberg (1973) as the house group. It is clear that the household was corporate in nature; it consisted of at least several family units, owned property, featured internal hierarchy, and often had a multi-generational lifespan (Suttles 1990a; papers in this volume). Hence, the household was a residential corporate group, also known as a corporate household. Here, we refer to this unit as both a household and a corporate household, using these labels interchangeably. Several nuclear and/or composite families (the latter involving polygynous unions) resided in each plankhouse. Many anthropologists refer to each such unit simply as a family. However, Mitchell and Donald (1988) label this unit an independent household. Many if not all Northwest Coast societies contained a residential unit that was either coterminous with the household, or nested between the household and the community. This was the kin-based or lineage-like corporate group, and is variously referred to as a lineage (e.g. de Laguna 1990:213; Mitchell and Donald 1988), corporate household (Ames 1995, 2003), House (Ames 1995, 2004; Coupland 1996a; Marshall 2000, this volume; Trieu this volume), house group (e.g. de Laguna 1990:213; Oberg 1973), local group (Drucker 1951; Suttles 1990b:464), deme (Jorgenson 1980) and, in the case of the Kwakwaka’wakw, numaym (Codere 1990). This variation reflects the different ways household members were recruited. On the northern Coast, for example, houses usually contained one or more matrilineages; on the southern Coast, kin recruitment was bilateral. Several of the authors in this volume refer to this unit as a House. While documentary sources clearly indicate the presence of Houses on the central and northern Coast, they do not clarify whether Houses characterized southern Coast societies (see papers in Suttles 1990a). Northwest Coast ethnographic data indicate that historically, some Houses were composed of one corporate household, while other Houses were composed of two or more corporate households (e.g. Codere 1990; Halpin and Seguin 1990:274; Suttles 1990c:464). Thus, in some cases, the corporate household was completely coterminous with the House, while in other cases two or more corporate households composed the House. In the latter case, 5

Gahr, Sobel, & Ames tion, these studies apply a variety of methodological and analytical tools to artifact and feature data. The result is a set of fine-tuned and robust analyses of household material culture, and the first collection of studies focused on the archaeology of Northwest Coast households.

one corporate group (House) contained another (household). This array of terms for each organizational level involves several problems. First, communication about Northwest Coast societies and households is impeded because different researchers use the same term in reference to different residential units. For example, the term local group has been used as a label for both the community and the House, while the term corporate household has been used as a label for both the household and the House. Until scholars agree upon a single set of meanings for each term, or unless each scholar explicitly defines the meaning of his/her terms for residential units, more confusion will prevail than is necessary. Second, documentary sources cannot always clarify the relationship between household and House in some Northwest Coast societies and it is even more difficult to use archaeological data to assess the relationship between household and House (as discussed for the Tsimshian by Coupland [1996a:75]). In particular, it is quite difficult to use archaeological data to determine whether or not a community contained Houses and, if so, whether each household was a partial or complete House. Third, it can be difficult archaeologically to distinguish between houses, and hence between households. This is the case because some sites contain remains of two or more attached structures. At these sites, each multi-structural feature is presumably the remains of either (a) two or more houses that were physically attached to each other, or; (b) one large and internally segmented house (e.g. Greene 1999; Matson 1999, 2003; Sobel, this volume). In these cases, we can usually only hypothesize about whether each multi-structural feature should be regarded as the remains of several households living in contiguous plankhouses, or one household residing in an internally segmented plankhouse.

Theoretical Themes Two recurrent theoretical themes can be traced throughout this volume: 1) hierarchy and households, and 2) temporality and households. Social, Political, and Economic Hierarchy All papers in this collection address in some way the relationship between hierarchy and the household, and many of the papers center on this relationship. This attention to the articulation of households with systems of hierarchy indicates a general consensus among researchers that (1) we still have much to learn concerning the role of the household in the evolution and structure of social, political, and economic hierarchy on the Coast, and (2) archaeological research is a particularly effective means of learning about these things. A number of papers address the relationship between sociopolitical hierarchy and household production. Samuels’ study, expanding on previous analyses of the Ozette site, examines variation in household production between an elite household and a commoner household. His results suggest that the intensity of production was similar in elite and commoner households. However, the elite household invested a higher proportion of its productive resources in woodworking, ethnographically a relatively high prestige activity, whereas the lower status household invested a higher proportion of its productive resources in hunting terrestrial animals, ethnographically a relatively low prestige activity. Ames considers variation in household production in terms of household demography and risk management strategies, which in turn relate to household sociopolitical status. In his simulation of household cycles, larger households, often of fairly high status, were relatively stable and survived for long periods of time while smaller households, often of lower status, were more vulnerable to failure. Ames speculates that a household had to recruit members through birth, adoption, or other means such as slavery in order to maintain a labor force large enough to produce the surplus resources needed to gain and maintain high status as well

This Volume The papers in this volume present the current state of Northwest Coast household archaeology. The household studies gathered here address a range of issues including architecture, domestic cycles, household production and distribution, social hierarchy, the origin and evolution of Northwest Coast settlement patterns, post-contact sociopolitical and economic change as well as the taphonomy of the material remains of household activities. The papers introduce new theoretical angles on the role of the household in social, political, and economic dimensions of Northwest Coast native life. In addi6

Introduction hold prestige in Lower Columbia communities. She concludes that exchange activity was more important while production was less important as a determinant of household prestige at one site than at the other, due in large part to differences in settlement location relative to travel routes and trade centers. Coupland and Trieu place the house dwelling at the center of research on social and political hierarchy. Coupland uses theoretical research on architectural communication and ethnographic data from the northern Northwest Coast to develop a model of the ways that house architecture conveyed information about social, political, and economic status both inwardly, to house residents, and outwardly, to residents of other houses and communities. He then applies this model to the McNichol Creek site, a north Coast archaeological site dated to ca. 1800–1500 B.P. Coupland’s analysis elucidates the link between intra- and interhousehold hierarchies, showing how the symbolic dimensions of house architecture simultaneously functioned within both parts of the status system at McNichol Creek. He argues that within the McNichol Creek community, the size, location, and building materials of House O communicated the wealth, rank, and power of the house owner—the inferred town chief—to residents of other houses and settlements. The architecture of House O also generated and maintained status hierarchy within the household by conveying information about status inwardly, to those living within the house, and thereby structuring the use of space and dayto-day relations within the house. While Coupland’s examination focuses on the symbolic aspects of architecture, Trieu looks at the material aspects of architecture. Trieu uses Lévi-Strauss’s concept of the House society, together with documentary and archaeological data from throughout the Coast, to show how the house was the fundamental piece of property owned by the Northwest Coast house group historically. Central to her analysis is the life cycle of the house structure -- its passage through phases of construction, maintenance, and demise. She argues that the production and intergenerational transmission of the house was material to the reproduction of status hierarchy both within and between households. As a result, she asserts, the nature and timing of house construction, maintenance, and demise were both determinants and products of the social, political, and economic status of a household leader and of the household unit. Two papers consider another issue of long-term interest in the anthropology of households and

as to survive over the long term. Coupland also addresses the dynamic between household status and labor force. Based on his analysis of the McNichol Creek site, he proposes that the architecture of a chief’s house signaled his affluence and power, and thereby served to attract commoners into the household as an additional labor source. Several papers explore the ways status hierarchies relate to unequal control over household production and consumption. Martindale’s archaeological analysis of one Tsimshian household points to differential control over stored subsistence surpluses, with higher status families wielding more control than other families within the household. Ellis provides a contrasting view of the large multifamily household. In his study of the remains of small houses at the Broken Tops site, he found no storage facilities or tools commonly associated with elite activities. He concludes that in the Lower Columbia River region at least, some households consisted entirely of commoners. Moreover, he suggests that these commoner households survived on a hand-to-mouth basis with production directed toward meeting immediate needs rather than producing surplus for winter months. Ellis further speculates that elites had little control over production and consumption by commoner households. Several papers offer a different perspective on the interplay between household economy and status, examining the relative contributions of two processes—exchange and surplus production—in creating and maintaining household status hierarchies. Martindale, applying a Neo-Marxist perspective to archaeological and documentary data from the Tsimshian, argues that sociopolitical hierarchy was strongly bound to household economics, especially to the interplay between household subsistence production, wealth production, and exchange. He attributes the development of a shortlived proto-contract paramount chiefdom among the Tsimshian to the intensification of household wealth economies, a shift in household mobility and exchange patterns, and concomitant changes in relations between households. While the increasing emphasis on wealth production may have facilitated the evolution of a short-lived chiefdom, it also led to the demise of the traditional, extended family household. Sobel examines the interplay between household production and exchange from a different angle. Using obsidian artifact data from multiple house features at each of two Lower Columbia sites, Sobel considers the degree to which household involvement in regional exchange networks, as opposed to household production, affected house7

Gahr, Sobel, & Ames archaeological case study of the Dionisio point site, located on the central Coast. Employing practice theory, he argues that evidence from the Dionisio Point site indicates long-term continuity in the use of space, including the distribution of prestige goods, within the house. Grier suggests that his results reflect the inter-generational transmission of patterns of household organization that in turn facilitated social hierarchy. Smith’s study of the Meier site house feature also provides evidence of long-term continuity in household organization. Likewise, Ames and Trieu are concerned with the inter-generational transmission of aspects of household organization, and how such transmission relates to long-term continuity and change in the structure of ranking. While some house remains reflect long-term continuity, others reflect a history of change or a short lifespan. Martindale’s study of the development of Tsimshian society from the late pre-contact through early post-contact periods explores how changes in household production facilitated a change from traditional extended family households to modern nuclear family households. This fundamental change in household organization was “at the core” of broader social, political, and economic changes, including the demise of the traditional system of sociopolitical hierarchy during the transition from the proto- to post-contact periods. Samuels also examines change over time in household production and its relationship to status systems, finding evidence of change in the elite household but not in commoner households at Ozette. These results may indicate that the elite household attained or maintained its high status through strategic changes in the organization of household production.

hierarchy on the Coast—the importance of elite privilege as opposed to elite managerial roles in the development and structure of ranking. Trieu challenges the predominant current approach to Northwest Coast social hierarchies as systems organized to facilitate elite privilege, suggesting that it is also productive to consider these hierarchies as systems that saddle elites with obligations to construct and maintain multi-generational households. In contrast, Ellis emphasizes elite privilege, using his Lower Columbia data to argue that by virtue of the prerogatives that came with high rank, elite households were better off than commoner households in times of stress. Ellis models a system in which elite households held many privileges, but these privileges were not necessarily gained through exploitation of commoner households, though perhaps by exploitation of commoners and slaves within elite households. Temporality: Evolution and Continuity All archaeologists research temporal phenomena. However, some of the papers in this volume raise and address questions especially pertinent to cultivating a better understanding of evolution and continuity in households. These questions span a range of topics from the origins and evolution of households on the Northwest Coast to the changes that characterized households over the past few centuries in the wake of European colonization. Marshall focuses on long-term evolutionary trends in social organization throughout the Northwest Coast. She integrates Wilson’s concept of “domesticated” societies with Lévi-Strauss’ concept of House societies. Both concepts share the notion of social continuity defined by the built environment and attachment to place. Marshall overviews the precontact evolution of people’s relationships to place and built environments on the Coast. Fundamental to this development is the increasing formalization of settlements and what she refers to as “the emergence of neighbors” and host/guest interactions. Ames, too, seeks to understand the evolution of Northwest Coast households or Houses in the context of interaction at the community or regional level. He proposes that Northwest Coast social and political hierarchies may have co-evolved with the household or House. Another angle of questioning on temporality emerges from analyses of individual houses. Grier provides the most detailed such study, examining intra-household dynamics and inequality through an

Methodological Considerations The studies in this volume address a variety of methodological concerns. Among these concerns, three seem especially important to the productive analysis and interpretation of the archaeological remains of households: 1) methods for studying the site formation processes that create household remains, 2) identifying useful archaeological correlates of the household, and 3) the integration of ethnographic, ethnohistoric, and archaeological household data. Site Formation Processes All the papers in this volume deal with the 8

Introduction in one plankhouse, and thus treat the plankhouse and its remains as the correlate of the household. Following this practice, most contributors to this volume regard the plankhouse as the correlate of the household. Northwest Coast archaeologists are concerned with identifying not only correlates of the household, but also correlates of intra- and inter-household hierarchy. The identification and interpretation of these correlates is integral, both methodologically and theoretically, to nearly all archaeological investigations of the articulation between households and hierarchy in Northwest Coast societies. Based on theoretical considerations, ethnographic models, and/or previous archaeological research, the papers in this volume use several variables as correlates of inter-household status hierarchies, including archaeological evidence of variation in: house size (Coupland, Ellis, Samuels, Sobel); house location within a settlement (Coupland, Marshall, Samuels, Sobel); house construction materials and design (Coupland, Ellis, Trieu); storage facilities (Ellis); presence or absence of a large central hearth (Coupland, Samuels); the frequency, taxonomic diversity, and condition of faunal remains (Coupland, Samuels); and the frequency and diversity of portable prestige goods (Ellis, Sobel). Papers in this volume also suggest that throughout the Northwest Coast, archaeological evidence of variation in house longevity (Ames, Grier, Marshall, Trieu) and house-cleaning activity (Samuels) may be reliable correlates of inter-household status hierarchy. In addition, several papers define and use correlates of intra-household status hierarchy. These correlates include variation in: location of living area within a house (Smith); house construction materials (Coupland); storage facilities (Martindale); the frequency and taxonomic diversity of subsistence remains (Coupland, Martindale); and the frequency and diversity of portable prestige goods (Coupland, Grier, Martindale).

material remains of households, however, papers by Smith, Samuels, and Grier particularly demonstrate how sociopolitical dynamics may be examined in the archaeological record by careful attention to taphonomy. Smith’s paper is instructional, detailing a comprehensive range of processes that can affect house remains. He discusses cultural and natural processes, as well as artifact analytical methods and how they inform archaeological interpretations. Smith emphasizes that prior to interpreting artifact spatial distributions as evidence of social, political and economic variation within and between households, the full range of site formation processes must be considered. Grier’s analysis of long-term continuity in artifact distributions reveals some of the challenges in identifying discrete spatial patterns of artifacts in the deep deposits resulting from long-term occupations. Samuels also explores how site formation processes affect archaeological interpretations of household social and economic dynamics. One of the most interesting parts of his study is the way in which his attention to taphonomy at Ozette reveals that variation in housecleaning practices affected variation between houses in midden development and artifact distributions; the analysis shows that the floor of the elite house was kept clean of debris, possibly because elites often hosted public events such as feasts and dances, whereas the floors of commoner houses were cleaned less thoroughly. Samuels’ study provides a clearcut example of how understanding taphonomic processes informs us about potential pitfalls in the interpretation of archaeological data, as set out by Smith. Archaeological Correlates of the Household An ongoing debate in household archaeology concerns archaeological correlates of the household (Allison 1999:4-5). In part, the debate stems from the lack of a universal correspondence between the household and the co-residential unit. Archaeologists have long recognized that in some parts of the world, the household does not correspond to the co-residential unit. However, when it can be justified, archaeologists usually characterize the household as a co-residential group, since this characterization is well-suited to the spatial and material dimensions of archaeology; whenever we can assume that the household was a co-residential unit, we can use the house and its archaeological remains as a correlate of the household. Most archaeologists working on the Coast apply the label “household” to the multi-family unit that co-resided

Integration of Ethnographical, Ethnohistorical, and Archaeological Data One methodological aspect of every paper in this volume is the integration of documentary (ethnographical and/or ethnohistorical) with archaeological data to explore Northwest Coast society and culture. The temporal and geographic correspondence between documentary and archaeological data sets varies, as some studies involve the historical archaeology of indigenous sites (e.g. Martindale, Smith, Sobel), while others involve the 9

Gahr, Sobel, & Ames finding, Martindale and others had assumed, based largely on documentary sources, that traditional household organization had evolved into nuclear family household organization nearly a century earlier among the Tsimshian. In Trieu’s study of the house life cycle, her examination of ethnographic, historical, and archaeological data from throughout the Coast indicates that although documentary sources contain a fair amount of information about house construction and demise, the archaeological record contains much more information about house maintenance—the middle portion of the house life cycle.

direct historical approach to late pre-contact sites (e.g. Coupland, Ellis, Martindale, Samuels, Smith, Sobel), and still others involve more general forms of ethnographic analogy (e.g. Ames, Trieu). This use of documentary sources in the interpretation of archaeological remains has long characterized archaeological research on the Coast; for decades, Northwest Coast archaeologists have drawn on the rich ethnographic and historical record of Native peoples in the region (e.g. de Laguna 1960). In addition, some archaeologists have investigated house sites in concert with native individuals who aided in the interpretation. Traditionally, archaeologists working on the Northwest Coast rely on ethnographic data to model post-contact societies and cultures, and then use these models in three basic ways: (1) to identify archaeological evidence of behaviors represented in the ethnographic models, (2) to determine how, when, and why the societies and cultures represented in ethnographic models evolved, and (3) as ideas to be tested through archaeological research. To varying degrees, all papers in this volume apply the first approach, and several papers (e.g. Ames, Marshall) apply elements of the second approach. The third approach is implicit in all papers in this volume. In addition, the studies reported here demonstrate some new and inventive ways of integrating ethnographic, historical, and archaeological data. For example, Ames uses ethnographic data to develop a simulation model of the household cycle in a hypothetical Northwest Coast community, and discusses how the simulation results can inform archaeological inferences about past households and communities in the region. In addition, several authors (e.g. Ames, Coupland, Marshall, Trieu) use Northwest Coast ethnographic data to justify applications of general theoretical frameworks to archaeological data from the Coast. Moreover, studies in this book demonstrate that despite the relative wealth of documentary data about Northwest Coast peoples, archaeological data are essential for researching Native houses and households not only in the pre-contact period, but also in the post-contact period. The papers by Martindale and Trieu highlight this critical role of archaeology in research on historical houses and households. Martindale’s analysis shows that despite an abundance of ethnographic and historical information about the Tsimshian people of the early twentieth century, only the archaeological record reveals that traditional household organization persisted until the early twentieth century among some Tsimshian. Previous to this archaeological

Comparative Perspective The concluding paper in this volume is a comparative perspective of household archaeology rendered by Jeanne Arnold. Her long-term research on Chumash households on the Channel Islands of California provides her the opportunity to compare and contrast the material culture of two societies from the West Coast of North America supported by fishing, hunting, and gathering economies. The peoples of the Northwest Coast and the California Channel Islanders have long been lumped as sedentary, socially complex hunting-and-gathering societies in anthropological theorizing. Yet, as Arnold points out, material manifestations of these cultures were dramatically different, especially with regard to houses. The Northwest Coast with its substantial investments in house architecture and the centrality of the house in traditional culture contrasts sharply with the Channel Islanders’ markedly minimal investment in architecture and small households, but well developed plank canoe industry and large scale craft production. Organization of this Volume We found it difficult to place the papers in this volume into separate titled sections because the studies exhibit many topical, theoretical and methodological crosscurrents. Consequently, we have organized the volume so that each study picks up threads of the prior study, continuing the preceding discussion or turning it for view in another light. The first three papers are broad in geographical coverage and theoretical perspective. The following seven papers use specific case studies to develop and highlight insights into major questions confronting Northwest Coast archaeologists. The final paper is again more general in scope, providing a comparative perspective. 10

Introduction Notes

and J. Brown, pp. 155-180. Academic Press, New York. 1995 Chiefly power and household production on the Northwest Coast. In Foundations of Inequality, edited by T. D. Price and G. M. Feinman, pp. 155-187. Plenum Press, New York. 1996 Life in the Big House: household labor and dwelling size on the Northwest Coast. In People Who Lived in Big Houses: Archaeological Perspectives on Large Domestic Structures, edited by G. Coupland and E. B. Banning, pp. 131-150. Monographs in World Prehistory, vol. 27. Prehistory Press, Madison. 2001 Slaves, chiefs and labour on the northern Northwest Coast. World Archaeology 33(1):1-17. 2003 The Northwest Coast. Evolutionary Anthropology 12:19-31. 2004 Supposing hunter-gatherer variability. American Antiquity 69:364-374. Ames, K. M. and H. D. G. Maschner 1999 Peoples of the Northwest Coast: Their Archaeology and Prehistory. Thames and Hudson Ltd., London. Ames, K. M., D. F. Raetz, S. C. Hamilton and C. McAfee 1992 Household archaeology of a southern Northwest Coast plank house. Journal of Field Archaeology 19:275-290. Archer, D. J. W. 2001 Village patterns and the emergence of ranked society in the Prince Rupert Area. In Perspectives on Northern Northwest Coast Prehistory, edited by J. S. Cybulski, pp. 203-223. Mercury Series, Archaeological Survey of Canada Paper 160. Canadian Museum of Civilization, Hull. Beaglehole, J. C. (editor) 1955 The Journals of Captain James Cook on his Voyages of Discovery. Published for the Hakluyt Society. Cambridge University Press, Cambridge Blukis-Onat, A. 1985 The multifunctional use of shellfish remains: from garbage to community engineering. Northwest Anthropological Research Notes 5:6-20. Boas, F. 1916 Tsimshian Ethnography. Thirty-first Annual Report, U.S. Bureau of American Ethnology, 1909-1910. Smithsonian Institution, Washington D. C.

The Early Modern Period (Ames and Maschner 1999) spans the period from first direct contact on the Coast (ca. 1778) to the establishment of reservations in the USA and reserves in Canada, ca. 1855. The term “Early Modern” avoids the various implications of “Historic Period” etc. 2 Suttles’ dissertation was published without his approval as part of Garland Press’ series of published Indian land claims documents. 3 Cascadia is a useful term that includes southeast Alaska, British Columbia west of the Canadian Rockies, Washington State, western Oregon, much of central and eastern Oregon, much of Idaho and western Montana. It essentially includes those regions drained by salmon-bearing rivers. In terms of cultural areas, it includes the Northwest Coast, Intermontane Plateau and southerly portions of the Subarctic (in British Columbia). 1

References Cited Acheson, S. R. 1991 In the Wake of the ya’åats’ xaatgáay [‘Iron People’]: A study of Changing Settlement Strategies among the Kunghit Haida. Ph.D. Dissertation, Oxford University. Adams, J. W. 1973 The Gitksan Potlatch, Population Flux, Resource Ownership and Reciprocity. Holt, Rinehart, and Winston of Canada, Toronto. Allison, P. 1999 Archaeology of Household Activities. Routledge, London Ames, K. M. 1981 The evolution of social ranking on the Northwest Coast of North America. American Antiquity 46:789-805. 1983 Towards a general model of the development of social hierarchies among foragers. In The Evolution of Maritime Cultures on the Northeast and Northwest Coasts of America, edited by R. J. Nash, pp. 173184. Department of Archaeology Publication, No. 11. Simon Fraser University, Burnaby. 1985 Hierarchies, stress and logistical strategies among hunter-gatherers in northwestern North America. In Prehistoric Hunter Gatherers: The Emergence of Cultural Complexity, edited by T. D. Price 11

Gahr, Sobel, & Ames 1921 Ethnology of the Kwakiutl (Based on data collected by George Hunt). 2 pts. In 35th Annual Report of the Bureau of American Ethnology for the Years 1913-1914, pp. 43-1481. Smithsonian Institution, Washington D. C. Carlson, R. L. 1990 History of research. In Northwest Coast, edited by W. Suttles, pp. 107-115. Handbook of North American Indians, Volume 7, W. C. Sturtevant, general editor, Smithsonian Institution, Washington, D.C. Chatters, J. C. 1989 The antiquity of economic differentiation within households in the Puget Sound Region, Northwest Coast. In Households and Communities, edited by S. MacEachern, D. J. W. Archer, and R. D. Garvin, pp. 168-178. Archaeological Association, University of Calgary, Calgary. Codere, H. 1950 Fighting With Property: A Study of Kwakiutl Potlatching and Warfare, 17921930. University of Washington Press, Seattle. 1990 Kwakiutl: traditional culture. In Northwest Coast, edited by W. Suttles, pp. 359-377. Handbook of North American Indians, Volume 7, W. C. Sturtevant, general editor, Smithsonian Institution, Washington, D.C. Coupland, G. 1985a Household variability and status differentiation at Kitselas Canyon. Canadian Journal of Archaeology 9:39-56. 1985b Restricted access, resource control and the evolution of status inequality among hunter-gatherers. In Status, Structure and Stratification: Current Archaeological Reconstructions, edited by M. Thompson, M. T. Garcia and F. J. Kense, pp. 245-252. Archaeological Association of the University of Calgary, Calgary. 1988a Prehistoric economic and social change in the Tsimshian area. Research in Economic Anthropology Supplement 3:211245. 1988b Prehistoric culture change at Kitselas Canyon. Archaeological Survey of Canada Mercury Series 138. Ottawa. 1996a The evolution of multi-family households on the Northwest Coast of North America. In People Who Lived in Big Houses: Archaeological Perspectives on 12

Large Domestic Structures, edited by G. Coupland and E. B. Banning, pp. 121-130. Monographs in World Prehistory, No. 27. Prehistory Press, Madison. 1996b The early prehistoric occupation of Kitselas Canyon. In Early Human Occupation in British Columbia, edited by R.L. Carlson and L. Dalla Bona, pp. 159-166. University of British Columbia Press, Vancouver. Coupland, G. and E. B. Banning (editors) 1996 People Who Lived in Big Houses: Archaeological Perspectives on Large Domestic Structures. Monographs in World Prehistory, No. 27. Prehistory Press, Madison. Coupland, G., G. Bissel and S. King 1993 Prehistoric subsistence and seasonality at Prince Rupert Harbour: evidence from the McNichol Creek Site. Canadian Journal of Archaeology 17:59-73. Coupland, G., R. H. Colten, and R. Case 2003 Preliminary analysis of socioeconomic organization at the McNichol Creek Site, British Columbia. In Emerging from the Mist, edited by R. G. Matson, G. Coupland and Q. Mackie, pp. 152-169. UBC Press, Vancouver. de Laguna, F. 1960 The Story of a Tlingit Community: A Problem in the Relationship between Archaeological, Ethnological and Historical Methods. Bureau of American Ethnology Bulletin, No. 172. U. S. Government Printing Office, Washington D. C. 1972 Under Mount St. Elias: The History and Culture of the Yakutat Tlingit. Smithsonian Contributions to Anthropology, Vol. 7. (in three parts) U. S. Government Printing Office, Washington D. C. 1990 Tlingit. In Northwest Coast, edited by W. Suttles, pp. 203-228. Handbook of North American Indians, Volume 7, W. C. Sturtevant, general editor, Smithsonian Institution, Washington, D.C. de Laguna, F., F. A. Riddell, D. F. McGeein, K. S. Lane, J. A. Freed, and C. Osborne 1964 Archaeology of the Yakutat Bay Area, Alaska. Bureau of American Ethnology Bulletin 192. Washington, D.C. Drucker, P. 1951 The Northern and Central Nootkan Tribes. Bureau of American Ethnology Bulletin 144. The Smithsonian Institution, Washington D. C.

Introduction Drucker, P. and R. F. Heizer 1967 To Make My Name Good: A Re-examination of the Southern Kwakiutl Potlatch. University of California Press, Berkeley. Fladmark, K. R. 1973 The Richardson Ranch Site: a 19th century Haida house. In Historical Archaeology on Northwestern North America, edited by R. M. Getty and K. R. Fladmark, pp. 53-95. Archaeological Association, University of Calgary, Calgary. 1975 A Paleoecological Model for Northwest Coast Prehistory. Archaeological Survey of Canada Paper, Mercury Series, No. 43, Ottawa. Flannery, K. V. (editor) 1976 The Early Mesoamerican Village. Academic Press, New York. Flannery, K. V. and M. C. Winter 1976 Analyzing household activities. In The Early Mesoamerican Village, edited by K. V. Flannery. Academic Press, New York Greene, J. 1999 From seasonality to status: an emerging picture of a Coast Salish household based on a floor midden analysis. Paper presented at the 64th Annual Meeting of the Society for American Archaeology, Seattle. Grier, C. 2001 The Social Economy of a Prehistoric Northwest Coast Plank House. Unpublished Ph.D. Dissertation, Arizona State University. Hajda, Y. 1994 Notes on Indian houses of the Wapato Valley. Northwest Anthropological Research Notes 28(2):177-188. Halpin, M. and M. Seguin 1990 Tsimshian peoples: Southern Tsimshian, Coast Tsimshian, Nisga, and Gitksan. In Northwest Coast, edited by W. Suttles, pp. 267-284. Handbook of North American Indians, Volume 7, W. C. Sturtevant, general editor, Smithsonian Institution, Washington, D.C. Hayden, B. 1981 Research and development in the Stone Age: technological transitions among hunter-gatherers. Current Anthropology 22:519-548. Hayden, B. and A. Cannon 1982 The corporate group as an archaeological 13

unit. Journal of Anthropological Archaeology 1:132-158.

Hill, J. N. 1970 Broken K Pueblo: Prehistoric Social Organization in the American Southwest. University of Arizona Press, Tucson. Huelsbeck, D. R. 1988 The surplus economy of the Northwest Coast. Research in Economic Anthropology, Supplement 3, 149-177. 1989 Food consumption, resource exploitation and relationships with and between households at Ozette. In Households and Communities, edited by S. MacEachern, D. J. W. Archer and R. D. Garvin, pp. 157-166. Archaeological Association, University of Calgary, Calgary. 1991 Mammals and fish in the subsistence economy of Ozette. In Ozette Archaeological Project Research Reports, Volume II, Fauna, edited by S. R. Samuels, pp. 17-92. Washington State University, Department of Anthropology Reports of Investigations 66, National Park Service, Pacific Northwest Regional Office, Seattle, Pullman, WA. Jorgensen, J. G. 1980 Western Indians: Comparative Environments, Languages, and Cultures of 172 Western American Indian Tribes. W.H. Freeman, San Francisco. Koyama, S. and D. H. Thomas (editors) 1981 Affluent Foragers: Pacific Coasts East and West, Senri Ethnological Series No. 9. National Museum of Ethnology, Osaka. Lévi-Strauss, C. 1982 The Way of the Masks (Trans. by S. Modelski). University of Washington Press, Seattle. MacDonald, G. F. 1983 Haida Monumental Art: Villages of the Queen Charlotte Islands. University of Washington Press, Seattle. Mackie, A. P. and L. Williamson 2003 Nuu-chah-nulth houses: structural remains and cultural depressions on southwest Vancouver Island. In Emerging from the Mist: Studies in Northwest Coast Culture History, edited by R. G. Matson, G. Coupland, and Q. Mackie, pp. 105-151. University of Washington Press, Seattle. Marshall, Y. 2000 Transformations of Nuu-chah-nulth houses. In Beyond Kinship: Social and

Gahr, Sobel, & Ames Matson, R. G. and G. Coupland 1995 The Prehistory of the Northwest Coast. Academic Press, Orlando. Mauger, J. E. 1991 Shed-roof houses at Ozette and in a regional perspective. In Ozette Archaeological Project Research Reports, Volume I: House Structure and Floor Midden, edited by Stephan R. Samuels, pp. 29-174. Washington State University, Department of Anthropology Reports of Investigations 63. National Park Service, Pacific Northwest Regional Office, Pullman, WA. Mitchell, D. and L. Donald 1988 Archaeology and the study of Northwest Coast economies. In Prehistoric Economies of the Northwest Coast, pp. 293-351. Research in Economic Anthropology, Supplement 3. JAI Press, Greenwich. Nabokov, P. and R. Easton 1988 Native American Architecture. Oxford University Press, Oxford. Oberg, K. 1973 The Social Economy of the Tlingit Indians. University of Washington Press, Seattle. Piddocke, S. 1965 The potlatch system of the Southern Kwakiutl: a new perspective. Southwestern Journal of Anthropology 21:244264. Price, D. T. 1981 Complexity in “non-complex” societies. In Archaeological Approaches to the Study of Complexity, edited by S.E. van der Leeuw, pp. 55-99. Universiteit van Amsterdam, Amsterdam. Price, T. D. and J. A. Brown (editors) 1985 Prehistoric Hunter-Gatherers: The Emergence of Cultural Complexity. Academic Press, Orlando. Sahlins, M. 1972 Stone Age Economics. Aldine, Chicago. Samuels, S. R. 1991 Patterns in Ozette floor middens: reflections of social units, In Ozette Archaeological Project Research Reports, Volume 1, House Structure and Floor Midden, edited by S. R. Samuels, pp. 175-284. Washington State University, Department of Anthropology Reports of Investigations 63. National Park Service, Pacific Northwest Regional Office, Pullman, WA.

Material Reproduction in House Societies, edited by R. A. Joyce and S. D. Gillespie, pp. 73-102. University of Pennsylvania Press, Philadelphia. Maschner, H. D. G. 1992 The origins of hunter-gatherer sedentism and political complexity: a case study from the northern Northwest Coast. Unpublished Ph.D. Dissertation, University of California, Santa Barbara. Matson, R. G. 1983 Intensification and the development of cultural complexity: the Northwest versus the Northeast Coast. In The Evolution of Maritime Cultures on the Northeast and Northwest Coasts of America, edited by R. J. Nash, pp. 124-148. Department of Archaeology Publication, No. 11. Simon Fraser University, Burnaby. 1985 The relationship between sedentism and status inequalities among huntergatherers. In Status, Structure and Stratification: Current Archaeological Reconstructions, edited by M. Thompson, M. T. Garcia and F. J. Kense, pp. 245-252. Archaeological Association of the University of Calgary, Calgary. 1992 The evolution of Northwest Coast subsistence. In Long-term Subsistence Change in Prehistoric North America, edited by D. E. Croes, R. A. Hakins and B. L. Isaac, pp. 367-430. Research in Economic Anthropology, Supplement 6, JAI Press, Inc., Greenwich. 1996 Households as economic organization: a comparison between large houses on the Northwest Coast and in the Southwest. In People Who Lived in Big Houses: Archaeological Perspectives on Large Domestic Structures, edited by G. Coupland and E. Banning, pp. 107-120. Monographs in World Archaeology No. 27, Prehistory Press, Madison, Wisconsin. 1999 Coast Salish shed-roof houses and households: lessons from Shingle Point, B.C. Paper presented at the 64th Annual Meeting of the Society for American Archaeology, Seattle. 2003 The Coast Salish house: lessons from Shingle Point, Valdes Island, British Columbia. In Emerging from the Mist, edited by R. G. Matson, G. Coupland and Q. Mackie, pp. 76-104. UBC Press, Vancouver. 14

Introduction Samuels, S. R. and R. D. Daugherty 1991 Introduction to the Ozette Archaeological Project. In Ozette Archaeological Project Research Reports, Volume 1, House Structure and Floor Midden. Washington State University, Department of Anthropology Reports of Investigations 63. National Park Service Northwest Regional Office and Washington State University, Pullman, WA. Schalk, R. F. 1977 The structure of an anadromous fish resource. In For Theory Building in Archaeology, edited by L. R. Binford, pp. 207-249. Academic Press, Orlando. Sheets, P.D. 1979 Maya recovery from volcanic disasters: Ilopango and Ceran. Archaeology 32:3243. Sobel, E. A. 2004 Social Complexity and Corporate Households on the Southern Northwest Coast. Unpublished Ph.D. Dissertation, Department of Anthropology, University of Michigan, Ann Arbor Suttles, W. 1951 Economic life of the Coast Salish of Haro and Rosario Straits. Ph.D. Dissertation, University of Washington. 1960 Affinal ties, subsistence and prestige among the Coast Salish. American Anthropologist 62:296-305. 1962 Variation in habitat and culture on the Northwest Coast. Paper presented at the Proceedings of the 34th International Congress of Americanists, Vienna, Austria. 1968 Coping with abundance: subsistence on the Northwest Coast. In Man the Hunter, edited by R. B. Lee and I. DeVore, pp. 5668. Aldine, Chicago. 1990b Introduction. In Northwest Coast, edited by W. Suttles, pp. 1-15. Handbook of North American Indians, Volume 7, W. C. Sturtevant, general editor, Smithsonian Institution, Washington, D.C.

15

1990c Central Coast Salish. In Northwest Coast, edited by W. Suttles, pp. 453-475. Handbook of North American Indians, Volume 7, W. C. Sturtevant, general editor, Smithsonian Institution, Washington, D.C. 1991 The shed-roof house. In A Time of Gathering: Native Heritage in Washington State, edited by R. K. Wright, pp. 212-222. Burke Museum, Seattle. Suttles, W. (editor) 1990a Northwest Coast, Handbook of North American Indians, Volume 7, W. C. Sturtevant, general editor, Smithsonian Institution, Washington, D.C. Vastokas, J. M. 1966 Architecture of the Northwest Coast Indians. Unpublished Ph.D. Dissertation, Columbia University, New York. Vayda, A. P. 1961 A re-examination of Northwest Coast economic systems. Transactions of the New York Academy of Sciences, Series 2 23(7):618-624. Waterman, T. T. and R. Greiner 1921 Indian Houses of Puget Sound. Indian Notes and Monographs. Museum of the American Indian, Heye Foundation, New York. Wessen, G. 1988 The use of shellfish resources on the Northwest Coast: the View from Ozette. Research in Economic Anthropology, Supplement 3:293-351. JAI Press, Greenwich. 1991 The utilization of whales at Ozette. In Ozette Archaeological Project Research Reports, Volume II, Fauna, edited by S. R. Samuels, pp. 265-391. Washington State University, Department of Anthropology Reports of Investigations 66, National Park Service, Pacific Northwest Regional Office, Seattle, Pullman, WA. Wilk, R. R. and W. L. Rathje 1982 Household archaeology. In Archaeology of the Household: Building a Prehistory of Domestic Life, R.R. Wilk and W. Rathje, eds. American Behavioral Scientist 25(6):631-640.

Thinking about Household Archaeology on the Northwest Coast Kenneth M. Ames

In the 1970s, Northwest Coast archaeology was riveted by the extraordinary excavations at Ozette, where archaeologists were able to recover portions of entire houses that had been catastrophically sealed by mudslides. These excavations confirmed many aspects of early European portrayals of the interiors of Northwest Coast houses and demonstrated the utility and importance of the house as a central unit of archaeological methodology and analysis. In 1982, Wilk and Rathje published their edited issue of the American Behavioral Scientist that introduced household archaeology as a concept and method (Wilk and Rathje 1982a). In 1985, Gary Coupland applied the concept of household archaeology to the Northwest Coast, specifically to the Paul Mason site in northern British Columbia (Coupland 1985). Also, in 1985, I published a paper in which I argued that Sahlins’ concept of a domestic mode of production was a framework within which to discuss changes in labor organization and production on the Northwest Coast and among complex huntergatherers generally (Ames 1985). Subsequent years have seen an increasing number of papers on household archaeology on the Northwest Coast, including papers in edited volumes (e.g., Coupland and Banning 1996; Isaac 1988; MacEachern et al. 1989; Matson 2003) and journals (Ames et al. 1992, Coupland et al. 1993) and field projects explicitly designed as household archaeology projects (e.g., Ames et al. 1992, 1999; Coupland 1985, 1993; Grier this volume; Losey 2002; Matson 2003). A range of topics has been investigated, including the role of household leaders in production, specialization, and household control of resources, among others. However, during this productive period, there has been little explicit theory building about households and the archaeology of households. It may be no accident that Hendon’s review of household archaeology in the Annual Review of Anthropology (Hendon 1996) cites only one paper dealing with the Northwest Coast (Moss 1993), and that was a paper on the division on labor, not explicitly about households. She does, however, 16

offer a criticism of household archaeology in general which may be applicable to our work on the Northwest Coast: Interest in the household should not become merely a convenient justification for excavation of houses … The functional orientation of most current research in household archaeology does not provide a compelling reason for why the domestic group should matter to our reconstructions of the past, especially since much of this research has not even done a particularly good job of establishing what households do. (Hendon 1996:55)

While I am far more sympathetic to “functional” approaches to household archaeology than is Hendon (e.g., Ames 1996), I would maintain that despite our efforts of the past 20 years, and even with the available ethnographic record (e.g., Mitchell and Donald 1988), we do not really know what Northwest Coast households did and how they did it. I would also suggest that, if we want to contribute to a broader, cross-cultural and evolutionary understanding of households, we need to develop explicit theory about households on the Northwest Coast and elsewhere. Such theory need not be developed de novo. A considerable body of theory and ethnographic observations about houses and households has developed over the past century or more, but most particularly during the last 20 to 30 years (see below). Most, if not virtually all, of this theory has been developed and tested through research on peasant farmers. One issue to be addressed here is the extent to which this theory can be applied to hunter-gatherers, such as those on the Northwest Coast. It has become increasingly evident since about 1980 that large populations, permanent settlements, specialization in production and social inequality, among other traits, evolve among hunter-gatherers as well as among agriculturalists, although the message has yet to completely penetrate anthropology or archaeology (i.e. Pauketat 1996). Households also clearly evolve among hunter-

gatherers as well as farmers. Indeed, the appearance of permanent and/or substantial dwellings in the archaeological record is often one of the major changes in the archaeological record of ancient hunter-gatherers, and is sometimes accompanied by, or presages, other significant economic, social, and cultural changes. In Japan, for example, pottery has long been used to mark the Jomon period. However, the appearance of small pithouse settlements is what most clearly signals the evolution of Jomon lifeways. Subsequent Jomon history is marked, in part, by increasing variation in both settlement and house sizes. In Southwest Asia, small but substantial houses are among several innovations marking the abrupt shift from the Late Paleolithic Kebaran culture to the Natufian (Bar–Yosef 1998). The presence and absence of houses may indicate shifts from between foraging and collecting (e.g., Habu 1996), or increasing levels of sedentism (e.g., Ames 1991) and even herald the evolution of social inequality (Ames and Maschner 1999), or not, as the case may be. In Wyoming, substantial numbers of pit dwellings are present in the record between about 4800 and 3300 B.C., and then become comparatively rare (Larson 1997). On the Columbia Plateau, relatively sedentary, long–lived households develop by c. 5000 B.C., and become plentiful for a time after 3200 B.C., but there is no evidence of other profound social or economic changes for another 2,500 years or so (Ames 2000). If households are significant actors in social, economic and political evolution, as many believe (Gillespie 2000), than we also need to account for cases in which households do not play that role. A number of researchers, for example, have argued that the roots of permanent inequality, of ranking and stratification, lie, at least in part, in household production and actions (e.g., Boone 2000; Diehl 2000; Gillespie 2000; Pauketat 1996). However, an issue of long standing is whether the household economy, the domestic economy, must be transformed into a political economy that transcends the household for this to occur (Sahlins 1972). Arnold (1996) and I (Ames 1995) have both argued that such a transformation is required to turn the variety of pathways to power described, for example by Hayden (1995), into ranking among hunter-gatherers. I do not wish to pursue that point here, but what is clear is that large–scale political economies, polities, commonly do evolve among farmers, and do not among hunter-gatherers. Rare exceptions may include the Calusa of Florida (Marquardt 1999), and perhaps the Chumash of southern 17

California (Arnold 2001). There were certainly no formal polities on the Northwest Coast during the Modern Period (Ames and Maschner 1999), although chiefs did operate in regional interaction spheres (Adams 1973). Why don’t polities commonly develop among complex hunter-gatherers? Indeed, why do (archaeologically) clear-cut status distinctions not evolve more commonly among hunter-gatherers? The Jomon period, for example, spans at least 12,000 years, and is marked in some areas by high population densities, large settlements, and an intensive subsistence economy (Imamura 1996). There is ongoing debate, however, about whether ranking developed among Jomon peoples or not (e.g., Kobayashi 1992), let alone polities. While it is difficult to estimate Jomon populations, it is likely that population densities were high enough in some parts of Japan for small polities to evolve. They certainly were along parts of the Northwest Coast (Ames 1995), and yet polities did not develop, at least permanently. Most archaeologists would probably argue that the energetics of hunting and gathering economies simply cannot support large-scale polities. However, this does not explain why small-scale ones do not often develop. It may also be that there is something about the structure of hunter-gatherer economies or social organization that prevents the widespread development of extra-household political institutions. Tringham (2000) has recently made the intriguing suggestion that household production itself limited political development in Bronze Age Europe, preventing the urbanization and state development that occurred in nearby Southwest Asia. The basic argument of this paper is that these questions, and questions developed below, cannot really be addressed until we gain a much better understanding of hunter-gatherer household economies. Households and Houses The growth of household archaeology on the coast parallels developments elsewhere in archaeology in particular and in anthropology in general. Household studies in anthropology arose in the 1970s as part of a focus on micro-scale social, political, ecological and economic processes (e.g., Ellen 1982; Yanagisako 1979). These studies were conducted in Western industrial contexts as well as in so-called third world societies. One of the major concerns of these latter studies has been peasant farmers. That work stresses a “bottom-up” approach

Kenneth M. Ames

Thinking about Household Archaeology on the Northwest Coast

to a range of questions. Generally, household studies are seen as complimentary to what might be termed a more top-down, macro-scale approach to the same issues (e.g., Fricke 1986, Maclachlan 1987). This research has strong ecological foundations; Wilk even entitles his recent book on the topic Household Ecology. For Wilk, (following Fricke 1986): [The] household is the logical level of analysis for human ecological studies. Here the individual patterns of choice and strategic behavior can be placed within larger social structures and economic–ecological contexts. Societies adapt in only the most abstract sense of the word, but households adapt in concrete and observable ways. …[T]he rational adaptive actions of individuals and households can lead to irrational and destructive consequences at the level of larger systems such as communities and ecosystems. This suggests that households are a crucial link between the micro– and macroscale of human systems. (Wilk 1997:31)

Netting (1993) adds an almost Darwinian sensibility to this approach in his analysis of smallholding farmers, arguing that this particular form of household will develop many times independently under conditions of intense agricultural production. The wide temporal and spatial distributions of smallholders are, then, a form of convergent evolution. Netting further argues that small householders will develop in societies markedly different in their overall social and political organizations. This version of household studies underpins the discussions in this and previous papers (Ames 1995, 1996) and has its ultimate roots in Sahlins’ concept of a domestic mode of production (Ames 1985; Sahlins 1972). It stresses production and consumption, strategic behavior, decision-making and the evaluation of decisions against consequences (e.g., Fricke 1986). An alternative line of development in household studies has its roots in Lévi-Strauss’ concept of Sociétés à maisons: House societies.1 According to Lévi–Strauss, the House in these societies is “a corporate body holding an estate made up of both material and non-material wealth, which perpetuates itself through the transmission of its name, its goods and its titles down a real and imaginary line, considered legitimate as long as this continuity can express itself in the language of kinship or of affinity and, most often, both (Lévi-Strauss 1982, 174).” Lévi-Strauss’ concept is based in part on his understanding of the Kwakwaka’wakw (formerly Kwakiutl) numayam (Lévi-Strauss 1982), and 18

Northwest Coast societies, at least in the nineteenth century, fit his definition of House societies (e.g., Marshall 2000, this volume). However, most discussions of House societies focus on the ethnography of farmers in South Asia and the western Pacific (papers in Carsten and Hugh-Jones 1995 and Joyce and Gillespie 2000). This particular approach has taken a couple of directions at least, including what Carsten and Hugh-Jones (1995) term an “anthropology of architecture” which for them will be quite similar to an anthropology of the body, and, while difficult to summarize succinctly, is not strongly materialist, instead emphasizing semantic systems (Sandstrom 2000). A somewhat different, more materialist approach is developed in Joyce and Gillespie (2000) in a recent edited volume. While it is well beyond the scope of this paper to review this literature, it has some points relevant to this paper and this volume. Houses live longer than any of the humans that comprise them. This long life span is central to most discussions of Houses. Gillespie puts it this way: “(T)he operation of the house over time (is) the result of strategic decision making whose goal is to reproduce a corporate body linked to a perpetual estate” (Gillespie 2000:33). The corporate decision-making is generally seen as grounded in what Sandstrom (2000, 57) terms the “down to earth factors of how people make a living.” This echoes much of the household literature cited above (e.g., Fricke 1986). The estate may include both corporeal and non-corporeal property (e.g., land and songs). Houses are manifested by houses and the people who occupy them. The material house is the organizing focus of the House, not rules of kinship or affiliation. The houses may or may not be the large, imposing structures of the Northwest Coast or European aristocracy, but the material house is still central. House membership may be fluid through time, although the house may always be “full’ (Marshall 2000). Recruitment becomes a crucial strategy for House reproduction (e.g., Adams 1973). Finally, House societies appear generally to be hierarchical, although there is debate about this point. They are thought to be most common among so-called “middle range” societies, or societies where class divisions are not fully or completely evolved. Houses are often ranked; and they sometimes only exist in the higher ranks of societies. These factors raise three evolutionary questions: How to account for the evolution of House societies (assuming such things exist)? How to

Thinking about Household Archaeology on the Northwest Coast found on the coast; 2) those in which families occupy separate structures, but these structures are physically close to each other in patterned ways, as in compounds, and 3) large amorphous residential corporate groups, such as the neighborhood or barrio. The first two types are essentially isomorphic with households. Their own research was on the latter form of corporate group among Maya peasant farmers in Highland Guatemala In archaeology, in general, household studies played an early role in the development of ethnoarchaeology (e.g., Hayden and Cannon 1982; Kent 1984; Kramer 1982). But their major role has been in research on the evolution of agricultural economies and civilization in Mesoamerica (e.g., Flannery 1976) and more recently in Mesopotamia (e.g., Wattenmaker 1998) and the southeastern USA (e.g., Rogers and Smith 1995, Muller 1997). In Flannery’s The Mesoamerican Village, the household is both a methodological focus and a means of establishing a phenomenological scale for archaeological data. Flannery begins with households, then moves on to communities and then to regions and so on. Flannery’s work thus has much in common with household studies in socio-cultural anthropology and in other social sciences. The household is central, but is one among several scales, and analysis plays these scales against each other. Be that as may be, the main thrust of household studies has never been hunter-gatherers of any variety, and the form of household archaeology adopted by researchers on the coast has very strong ties to studies of peasant farmers. The concept of House societies has also not played an important role in archaeology beyond the archaeology papers (Kirch 2000; Marshall 2000; Tringham 2000) in Joyce and Gillespie’s edited volume of House societies (Joyce and Gillespie 2000). The concept is implicit in papers by Coupland (Coupland 1988, Coupland 1996) and myself (Ames 1995) although neither Coupland nor I frame our discussion in those terms, nor do we cite Lévi–Strauss’ founding publications on the topic (e.g., Lévi–Strauss 1982) In sum, households, and in some cases Houses, evolve among hunter-gatherers. They may be central to the development of social and economic complexity, both as the context in which complexity evolves and perhaps in limiting or shaping the directions such evolution may take. While we know a great deal about households and Houses among farmers, we know very little about them among hunter-gatherers. I agree with Sandstrom (2000), that the “down to earth factors of how people make a living” is key to understanding that evolution.

account for their persistence? And, lastly, how to account for their role, if any, in the evolution of social hierarchy? I would, however, echo Gillespie (2000) and argue against seeing House societies as yet another intermediate stage between egalitarian societies and states. I also believe that the answers to these questions rest in the kind of approaches advocated by Wilk, Netting, Fricke and others. Household and House studies thus share some overlap, although their respective practitioners do not often or extensively cite each other. However, households and Houses are conceptually quite distinct. A House can include one or more households. Many households may not persist beyond the deaths of their founders, while Houses, by definition, do. Houses are generally larger, with many members, while household sizes can vary from a nuclear family and up. Despite these differences, the concepts are heuristically useful, because, as will be seen below, they allow us to frame interesting evolutionary questions. The concept of House societies shares some features with Hayden and Cannon’s closed corporate group (Hayden and Cannon 1982). While they do not actually define what a corporate group is, they stress several features including long–term durability. Corporate groups last longer than the life spans of individual members. They strongly affect members’ lives: “Residential corporate groups are much more closed and exert a pervasive influence on all aspects of individuals’ lives, including their marriage, their post marital residence, their economic production, their feasting and celebrations, and their pastimes and pleasures “(Hayden and Cannon 1982, 135). Finally, corporate groups control crucial resources, particularly land, echoing the concept of the House estate. As with household and House theory, the corporate group is seen as an analytical level between the household and the community. Hayden and Cannon also see corporate groups as playing central roles in the evolution of social and political complexity, a view strongly held among many workers on the Northwest Coast. This parallels some discussions that Houses may be crucial actors in the evolution of social inequality, at least in some circumstances. However, for them, the corporate group is closed in terms of recruitment, while Houses are open—kinship may be the terminology of membership, but people affiliate themselves with houses for strategic reasons (Gillespie 2000). Hayden and Cannon (1982) specify three types of corporate groups: 1) those where several families live together in the same structure, such as are 19

Kenneth M. Ames Household Archaeology, Houses, and Households

is an aspect of underproduction that Sahlins describes as part of the Domestic Mode of Production. Alternatively, do they make decisions to maximize returns that may involve increased levels of risk. Here risk refers to both the likelihood and the cost of subsistence failure. In this instance, surpluses are produced which are “invested” for long-term purposes. For the Northwest Coast, this issue translates to what were the basic goals guiding the management of Northwest Coast household economies. This is a different way of approaching risk than is common to hunter-gatherer studies in which definitions of risk are usually drawn from evolutionary ecology and optimal foraging theory. In these approaches, risk refers to variation in outcomes, and the scale of risk is the individual forager. At the household level, this is risk in terms of variation in outcome and in terms of the costs of failure (Bamforth and Bleed 1997). The scale of risk is the household, and decisions are made at that scale. Crucial to managing risk at the household scale is the household cycle. Central to all studies of the peasant household is the notion of a household cycle, or a household life cycle (Gallant 1991). The household’s life cycle, at its simplest, spans the period from its founding, generally by a young, childless couple, through its “death” with the deaths of its final members. The circumstances of the household—its needs, its ability to field labor, its ability to respond to crises—will vary according to the age and makeup of the household at particular times. The amount of food (caloric demands) a household consisting of a man and woman in their early twenties will require will be quite different than those of a household with four children aged 3 to 10. The labor the latter household can field will be quite different than the labor the same household fields when those children are aged 10 to 17. Tringham (1991) has applied the concept of the household cycle to understand the site formation processes of house construction and abandonment in the Neolithic of the Balkans. The concept can perhaps be applied in a similar way to sites on the Columbia Plateau and in Japan, for example, where there may be a great many pithouses, but they appear not to have been all occupied simultaneously. Initially, at a residential locality, new houses may be dug and roofed at the beginning of a household, and abandoned at its death. At some point, old houses may be reoccupied by new households, and then abandoned. The household cycle could produce a pattern of occupation, abandonment, reoccupa-

As observed above, household archaeology generally has its roots in the archaeology and anthropology of small-scale farmers, particularly of peasants (e.g., Blanton 1994, Netting 1993). There are very strong links between ethnographic studies of peasants and the archaeology of households, particularly in Mesoamerica (e.g., Wilk 1997, Wilk and Ashmore 1988) where household archaeology began (Wilk and Rathje 1982a). Research issues central to this approach arise from the articulations among the following factors (e.g., Fricke 1986, Netting 1993, Wilk 1997): • The demography of households, including their domestic cycle, or life cycle, household biological reproduction and recruitment and how these affect the organization and reproduction of labor; • The strategies employed for managing risk; • The organization of households, including division of labor by age and sex, specialization, household leadership and inequality within households; • Systems of land tenure; • The ecology of farming, including but not limited to technology; • The ways in which the household articulates with larger scale social and economic systems, including the mobility of household members to and from the household. The rest of this paper looks at three of these: household demography, risk management, and articulation with larger scale systems. These topics are discussed because they are directly relevant to the evolutionary questions asked above: why and how do households and House societies evolve? What accounts for their long-term persistence? And what, if any role, do they play in the development of permanent social inequality? Risk and the household cycle An underlying and controversial problem in studies of peasant farmers that is relevant here has to do with the aims of household production: are these household economies basically risk aversive, in that household members “consistently and predictably select production strategies which enable them to lessen the risk of production failure” (Gallant 1991, 7). In other words, do they habitually elect to minimize returns to maximize safety? This 20

Thinking about Household Archaeology on the Northwest Coast of becoming extinct because of “the variability of birth and death rates” (Adams 1973:30), i.e. the household cycle. Because of this, the Gitksan made considerable efforts to ensure the persistence of Houses and of lineages. Adams’ monograph is, in fact, an exploration of the implications of variable household size and household vulnerability for Gitksan social organization. The mean size of a Gitksan House in 1920 was 20 members. This mean, however, provides a false sense of the true distribution of household sizes (Fig. 1). The actual size is skewed to the small end of the distribution. The lineages within these houses were quite small, and from Adams’ discussion, are essentially extended families of no more than three generations time depth. The mean number of such lineages/house in 1920 was three (Fig. 2). While these data are for the twentieth century, they are applicable to the premodern period as well, at least in terms of basic household demography of these households. The range of Gitksan household sizes in 1920 is similar to household size estimates I made several years ago (Ames 1996) for historic and premodern houses along the Northwest Coast, based on floor area (Figs. 3 and 4).4 The houses in my sample span the last 3,000 years. Since the households are generally the same size, then they should be subject to very similar demographic problems. These figures suggest that many Northwest Coast households were not so large as to be immune to the domestic cycle and to a cycle of vulnerability. If premodern household sizes had the same size distribution as the Gitksan households in Adams’ sample, then the bulk of households were exposed to the risk of household failure. Adams even suggests that Gitksan households were not economically or socially viable below a certain population threshold, and so once below that level, would spiral to extinction. If that is the case, then a crucial aspect of managing a household and a House would be maintaining or increasing its population to avoid extinction. There is a generally strong relationship between household (and house) size and status on at least the central and northern coasts in the nineteenth century: large Houses lived in physically large structures, and large households generally held higher status than smaller households did. In part, this may be a direct reflection of labor: a larger household has more labor to produce the goods that can be converted to wealth and status. However, a larger household is also more likely to be relatively immune to the household cycle and to demographic vulnerability. Hence, all things be-

tion, and abandonment resulting in a very complex residential site, but without a great many families being involved. However, the utility of the concept goes far beyond this. Standard demographic issues must be factored in, including life expectancy, fertility, fecundity, morbidity, etc. Though these issues are difficult to tackle archaeologically, it is possible to do so, as Muller’s (1997) recent work on the household cycle among Mississippian farmers demonstrates. For Muller, the central statistic for Mississippian households is the producer/consumer ratio—how many producers there are in a household to the number of consumers. This ratio will change through the household’s life cycle. The number of producers and consumers will increase for a time. As children grow, their contribution will exceed their consumption, but as children leave home, and as the founding couple age, the numbers of producers will fall more quickly than the number of consumers. Gallant (1991:102) defines what he terms “the household vulnerability cycle.” This cycle includes those periods when the household is particularly exposed to subsistence risk because of the interplay between the household’s demographic cycle and the demographic effects of the subsistence economy, i.e. the ratio between consumers and producers. A related issue is that of optimum household size, or, how big a household is big enough. A problem in applying the notion of household cycle to the Northwest Coast is that models of the cycle are based on households composed of single, small families: a couple, their children, perhaps an elderly dependent or two, and maybe an adult sibling. Northwest Coast households were much larger: with multiple families, and sometimes more than 100 members. Such household sizes might suggest that Northwest Coast households were large enough that they were buffered against the cycle’s effects. Thus, while one family had no children, or several small ones, another might have several strapping teenagers, anxious to work at household tasks. While this may be plausible, we do not know it. We do not really know the demography of the Northwest Coast household. However, Northwest Coast households may have been subject to a cycle of vulnerability. Data developed by John Adams for Gitksan households suggests household extinction was a very real possibility. Adams’ data are for the period between 1920 and 1967. As will be seen, however, these data are relevant to the premodern period. Gitksan houses, and the lineages they contained, appear to have been in regular danger 21

Kenneth M. Ames

Fig. 1. Gitksan House Sizes in 1920.

Fig. 2. The Number of Lineages/House.

22

Thinking about Household Archaeology on the Northwest Coast

Fig. 3: Archaeological Estimates of Household Populations for three ethnographically–known Villages.

Fig. 4. Estimated Household Populations for Premodern Houses.

23

Kenneth M. Ames assuming a 15.5% rate of lineage extinction/50 years, and three lineages/House (Adams 1973) for 100 lineages and 33 Houses. While stressing that this is an exercise, or “thought experiment,” it has some useful implications. The first is that it is theoretically possible for Houses to persist for a millennium. In the model, there are three lineages and one House left at the end of 1,000 years. Interestingly, when running the model with three houses, as at Keatley Creek, two houses are left after 250 years, and one after 300 years. The third house persists another 250 years, but that is an artifact of the model (through rounding numbers up). It essentially becomes extinct 500 years after the model’s starting point. The common life span for a single House in this model is 200 to 250 years. These comments are not offered to dispute Hayden et al.’s empirical conclusions, but to suggest that for these Houses to last for a millennium, recruitment is the overriding issue, particularly with small numbers of houses where stochastic processes such as accidents will be important in their survival. The second implication follows directly from the last. The shape of the curve clearly shows that to maintain a certain number of Houses—a clear goal in Gitksan society at least—recruitment is a constant issue. This is supported by Adams’ study. Clearly, some “recruitment” would occur by maintaining birth rates higher than death rates. However, presumably Adams’ extinction rate of 15.5%/47 years includes births. His population data for the Gitksan for this period shows a net increase of 170 people between 1921 and 1949. Thus, in the absence of rapid population growth, recruitment through births is probably insufficient to stave off lineage, and perhaps House, failure. Among the Gitksan, for the period of Adams’ study, adoption was the primary formal way for lineages to gain new members. This implication supports the statements of Gillespie (2000) and others about Houses, that while kinship is the idiom of membership and recruitment, Houses work to attract and hold members any way they can. The curve, coupled with Adams’ population data, also suggests that long-term persistence of Houses (beyond statistical luck) may require recruitment beyond the immediate community. The model includes 33 houses, perhaps three to 10 villages. Moving populations around among those villages would forestall failure among some Houses and accelerate it for others, and this is likely to be what occurred (cf. Adams 1973). However, this does not necessarily preclude long-term extinctions among Houses and an overall reduction in the number of Houses. One form

ing equal, a large household is also more likely to persist through time as a House, and to maintain its estate, than is a small household.2 There is some evidence to suggest that some households in the Pacific Northwest may have been quite long lived. Hayden et al. (1996) have claimed that three households at Keatley Creek, in south–central British Columbia, persisted for at least a millennium. Households and perhaps Houses may have persisted for 400 years at two sites on the Lower Columbia River, the Meier site (Ames et al. 1992) and Cathlapotle (Ames et al. 1999). In three excavated houses at these sites, we observed houses timbers being replaced in almost exactly the same places for the entire dated use life of the structures. Marshall (2000) has made similar observations for Nuu-chah-nuulth houses on Vancouver Island, arguing that this is evidence for the existence of Houses. Cybulski (1993) and I have also suggested that corporate groups, or lineages, at the Boardwalk site, in Prince Rupert Harbor, northern British Columbia, may have existed for 400 to 600 years (Ames 1996). Given what we know about the household cycle, a critical question becomes how is this apparent durability achieved? Hayden et al.’s claim is of particular interest to this discussion. They argue that residents of three of the excavated pithouses (Houses 1, 5 and 7) at Keatley Creek, in the British Columbia interior, were members of residential corporate groups that had different territories and access to different lithic sources over periods in excess of a millennium. These households would therefore certainly qualify as Houses by any of the definitions in that literature. Hayden (Hayden 1997, Hayden et al. 1996) estimates the population of House 7 at approximately 45 people (at 2m2/person). This frankly seems high to me, but, regardless, it is still too small to be reproductively viable over a long time period (Wobst 1974). Thus, if Hayden et al. (1996) are correct, these households were able to successfully replenish themselves, recruiting new members over 1,000 years. In this regard, it is worth noting that Adams found that 15.5% of Gitksan lineages extant in 1920 had died out by 1967 (Adams 1973). It is important to note that Adams did not speak of Houses dying out, he discussed lineages. In fact, he comments on the efforts made by other Houses to prevent a House from becoming extinct. In any case, I do not know if the 15.5% is a “typical” attrition rate or not, but it is a useful number to think with. I used it to model the attrition of Gitksan lineages and Houses across a millennium (Fig. 5), 24

Thinking about Household Archaeology on the Northwest Coast of extra-community recruitment on the coast was slavery (Ames 2001) Formation of new Houses would also counteract the overall effects of House extinction. A number of authors, myself included (Ames 1995), have noted that house chiefs had little power to control the actions of their followers, that people on the Northwest Coast could “vote with their feet.” Stearns (1984) suggests that this freedom was the ultimate limit on the power and authority of Northwest Coast chiefs. However, this may not have been as straightforward as it appears, and the failure rate of new Houses was probably quite high. They plausibly had few members, and, as a consequence, were more likely to fall victim to the house cycle than larger, established Houses. While in some parts of the coast, they may have had access to productive resource localities when they formed (Maschner 1992), readily available resource locations are likely to have been the property of failed

Houses, and therefore perhaps ecologically more marginal than the estates of established houses. It seems possible, therefore, that over the long run, Northwest Coast societies consisted of a core of large, successful Houses, with well-established estates, that generally persisted for long periods of time and a second group of smaller, less stable Houses that lasted only a few generations or a century or two. Clearly the core group would include the highest status Houses (e.g., Donald and Mitchell 1975). This core group would not be immune, however, to failure. Houses could be extirpated through warfare, a demographic glitch, or a run of bad luck or poor management. As a consequence, some of the more marginal Houses could seize the opportunity and move into the core. The examples cited above (Meier, Cathlapotle, Boardwalk, Keatley Creek) represent successful Houses and households, assuming the interpretations of them are correct. They were successful in

Fig. 5. Rates of Lineage and Household Extinction based on the Gitksan case (Adams 1973).

25

Kenneth M. Ames to management. However, it is also plausible that a large house would pursue a mixed strategy. Archaeologists generally accept that Northwest Coast households were large because the household economy was based on complex simultaneous tasks (e.g., Ames and Maschner 1999) that sometimes needed to be carried out in short periods of time. However, these relatively large households also permitted labor to be deployed in both high risk and risk aversive tasks. Whaling on the coast is one example. Whaling was a high risk (high variance), but very high return occupation. The cost of failure, however, was probably generally quite low (no one would starve if the whaling crews returned empty handed). Most household subsistence activities could focus on tasks that were both low risk (low variance) and low cost of failure, with some individuals, generally high status ones, pursuing the high-risk ventures. High-risk, high return strategies could benefit a high status household by producing surplus food that could be used to create economic and social debt and status advertising. Diehl (2000) and Boone (2000) both predict, using modeling derived from evolutionary ecology, that in variable environments, establishing patron-client relationships among households will benefit both patron and clients, leading to competition among patrons for clients and among clients for patrons. Boone predicts that surplus production will be used as “advertising” in this competition. Intense competition will place a high value on high risk, high return strategies. Thus faunal remains and other indicators of economic activity associated with houses are, to some degree, likely to reflect long-term decision making based on individual Houses’ circumstances as opposed to long-term, coast-wide or even regional economic trends (e.g., Maschner 1992). It is clear that households and Houses failed continuously and eventually all did so. Some of this failure was probably due to bad luck, combinations of the household cycle and other circumstances that even the best management could not survive. In any case, another key to long-term persistence may be how households articulated with larger scale entities, such as communities and interaction spheres.

that they persisted for centuries although they all ultimately failed. The Meier and Cathlapotle households failed because of introduced epidemics, such as smallpox (Boyd 1985), and contact with Europeans. The Keatley Creek and Boardwalk households failed for other reasons (e.g., Hayden and Ryder 1991). However, the point here is that the archaeological record of households and Houses may largely be a record of success, of Houses that for lengthy periods employed successful strategies to recruit members and manage risk. Some households probably did not persist long enough to be archaeologically visible while others may have lasted two or three generations, and be visible as a short-lived structure. Archaeologically, these latter will be hard to distinguish from settlement pattern shifts. There may be cases, however, where the archaeological record includes houses with both long and short use lives, particularly where they are architecturally identical. In this case, the short use lives may reflect short-lived households or Houses. At the beginning of this section, I observed that one of the issues in studying peasant households is the degree and under what circumstances they are risk aversive or risk seeking. These are polar strategies with quite different outcomes. Risk aversive strategies essentially maintain low, but predictable, rates of return in subsistence and economic pursuits, while risk seeking strategies may result in high rates of return, but also high rates of failure. Risk aversive strategies are those that seek to minimize variance in returns, while risk seeking strategies aim to take advantage of variance in order to get high returns. Peasants are generally thought to pursue risk aversive strategies. A second, somewhat related concept to risk is “cost of failure.” Variation in outcome may remain the same through time, but the costs of failure can vary according to circumstances. To take whaling as an example, whalers may harvest a whale every tenth trip more or less (risk—variance in outcome), but in most circumstances, that failure may have little overall impact on the household, but under some conditions (prolonged famine), success or failure may be critical. It seems plausible that risk management strategies would vary, depending on where a House was in its history. A small, new or struggling House might employ risk seeking strategies, gambling that a big return would attract new members through a number of ways, while a large, established House might be risk aversive, taking a cautious approach

Articulation with larger scale social and economic entities By definition, peasants are members of polities. Many ancient hunter-gatherers, including those on the Northwest Coast, were not. However, 26

Thinking about Household Archaeology on the Northwest Coast example, differences in house size and the public display of titles on poles and painted houses on at least the northern and central coast during the Modern period. Additionally, our earliest firm evidence of rectangular houses on the coast, the Paul Mason site, is also our earliest firm evidence for linear villages, the distinctive spatial form of Early Modern Northwest Coast villages (Ames and Maschner 1999). Further, the community is implicit in most archaeological studies of households and status on the coast. We generally sample at the community (settlement) level. Except for projects aimed explicitly at sampling individual houses, most projects focus on the settlement, and our results only really make sense when they can be comparatively analyzed at that or larger scales. Finally, some communities, or at least settlements, on the coast appear to have been as stable, if not more stable, than the households that comprised them. I have elsewhere (Ames 1998) suggested that Northwest Coast economies are to be understood at two levels above the site or settlement: the locality and the region (or interaction sphere). Some resources, for example eulachon on the northern Coast, are “regional” resources in the sense that a region was dependent on them while local resources were those exploited locally, but which played no regional role. The definition of “locality” is fuzzy, but is conceptually important to understanding Northwest Coast households and Houses. This importance can be shown with a brief case study: the Boardwalk site and Prince Rupert Harbor in northern British Columbia (Fig. 6). Boardwalk is one of 12 excavated sites in Prince Rupert Harbor. Ten of these sites were excavated as part of the Canadian National Museum of Civilization’s North Coast Prehistory Project (NCPP). The other two, including McNichol Creek (see Coupland, this volume), have been excavated more recently. Of the 10 NCPP sites, I analyzed the artifacts recovered from nine, eight of which are in Prince Rupert Harbor (Ames 2005). As part of the NCPP effort, 280 burials were excavated at eight sites (Cybulski 1993, 2001). Boardwalk is on Digby Island, across Prince Rupert Harbor from the city of Prince Rupert itself, which is on Kaien Island (Fig. 6). Boardwalk is a large midden, about 220 m x 50 m, averaging 2 m. in depth. It is immediately adjacent to the Parizeau Point site, a somewhat smaller midden. A third major midden in the vicinity is the Dodge Island site, a deep midden on an island in Dodge Cove next to Boardwalk. Another major residential

central to the concept of the household economy as it has developed is how the household interacts with larger social, economic, and political entities. Northwest Coast household studies have tended to focus on the household, rather than on its role in larger social, economic, and social systems. Northwest Coast archaeologists have looked at larger scale social and economic fields (sensu Ames 1991; Madden 1983; Welsch and Terrell 1989) including interaction spheres (Ames and Maschner 1999; MacDonald 1969; Suttles 1990) and obsidian networks (Carlson 1994). These fields are generally regional in scope. However, it seems clear from the archaeological record that there may be at least one or two levels intermediate between the household and the interaction sphere (Ames 1998, Ames and Maschner 1999). The community may be one such level. Chang (1968) proposed the community as the primary social entity that archaeologists study, with the archaeological settlement as its physical manifestation (Chang 1968) and the concept has seen consistent use in studies of the Neolithic (e.g., Kuijt 2000) and the development of states (e.g., Schwartz and Falconer 1994). Varien (1999) productively used the concept of community in his analysis of mobility patterns in the Mesa Verde region of the Southwest, arguing that the mobility patterns of households and communities may be different. He defines households similarly to Wilk, Rathje, Netting and others cited above. For Varien, “[A] community consists of many households that live close together to one another, have regular face–to–face interactions, and share the use of local social and natural resources” (Varian 1999:19). He suggests that communities seldom exceed 2,000 people and will usually be smaller. He stresses that while communities included multiple households, the crucial actors are individuals. The obvious communities on the Northwest Coast are villages and towns, although some single dwellings on the coast were large enough to house an entire community (Ames 1996). The ethnographic record of the coast suggests that these communities were primarily aggregations of Houses: Houses had estates, not larger communities, although communities (villages and towns) sometimes controlled territories for certain purposes. Houses were generally the highest political and economic entity. Consequently, we have thought little about the role of the community and how households interact in communities. However, the community was the stage or theater for the tangible expression of the relative status and prestige of Houses, through, for 27

Kenneth M. Ames

Fig. 6. Map showing location of Prince Rupert Harbor and the Boardwalk Site (used courtesy of Canadian Museum of Civilization).

28

Thinking about Household Archaeology on the Northwest Coast site of interest here is Lachane, which is located on Kaien Island immediately across the harbor from Dodge Cove. The earliest well dated occupation in the harbor occurred around 3800 B.C.. Lachane appears to have been first occupied c 3000 to 3400 B.C. and Boardwalk c. 2800 B.C.5 Boardwalk achieved its largest extent by about 2000 B.C.. The site’s topography suggests it was at least a two–row linear village (see Ames and Maschner 1999). The site has two broad platforms separated by a low ridge of shell midden. A higher midden ridge is behind the rear platform. These platforms or terraces are thought to have been for house rows. There are two visible house depressions on Boardwalk, at the rear of the site, which may date to c. 500 B.C.. Lachane also appears to have had platforms and back ridges, although the site’s surface topography had been altered before excavation and it has now been destroyed. These sites—all shell middens—are similar in many ways, including in the artifacts, features, and faunal remains. However, they all differ from each other in a variety of ways and at different scales, Boardwalk the most markedly. For example, of the nine sites I analyzed, Boardwalk consistently has more complete or unbroken tools/m3, more tool taxa/m3, but fewer tools/taxa. In other words, its artifact assemblages are consistently the most diverse in the harbor. One of Boardwalk’s analytical units contained the highest densities of decorated and zoomorphic objects and the greatest diversity of such objects of the 26 analytical units employed in my study (Ames 2005). This same analytical unit has the highest densities of items

of personal adornments per cubic meter as well. Boardwalk lacks certain heavy-duty wood working tools that are found in abundance at the neighboring Parizeau Point site. It is the only site to have classes of artifacts in non-burial midden deposits that occur elsewhere in the harbor exclusively in burial contexts. In sum, Boardwalk contains the richest occupational deposits of any in my analysis. It also has the highest densities of burials/m3 in its cemeteries. The back ridges of the residential sites dating before c. A.D. 600 were used as cemetery areas. The earliest such burial is dated to c. 1560 B.C. at a site called Kitandach. The earliest at Lachane date to c. 1530 B.C. and at Boardwalk to just before 900 B.C.. The use of shell middens for cemeteries was common on the coast before about 1,000 years ago or so. The harbor-wide burial program included internment in graves with no grave goods (about 87%) and burials with grave goods (c.13%). The sites differ markedly among themselves in the grave goods present (Table 1; Ames and Maschner 1999). The available data suggest the presence of ascribed ranking of individuals in the harbor by c. 900 B.C., if not actually several centuries earlier. They also strongly suggest contemporary ranking of households or lineages, as reflected by inter-cemetery differences, and of residential sites, as reflected by inter-settlement differences. The evidence also indicates that this pattern was stable over at least 1,500 years, if not longer. In terms of households, and the development of household organization on the northern coast, the appearance of ranking in the harbor is contemporary with the earliest, well–documented village on

Table 1. Distribution of grave goods by site in Prince Rupert Harbor (Ames 2005). BALDWIN: LABRET, SLATE MIRROR, BIRD BONE TUBE, COBBLE TOOLS, QUARTZ FLAKES BOARDWALK: LABRETS, COPPER, SHELL, AMBER, ZOOMORPHIC PENDANT, CANINES, BONE POINTS/DAGGERS, ANKLE BRACELETS, GROUND SLATE POINTS/DAGGERS, BONE POINT PENDANT, OTHER BONE, COBBLE TOOLS, ABRADERS, TROPHY SKULLS, SEA OTTER TEETH, OCHRE. DODGE ISLAND: LABRET, SLATE POINT/DAGGER, OBSIDIAN FLAKE, WORKED BONE. GARDEN ISLAND: LABRET, WORKED BONE LACHANE: AWLS, HARPOON/ PENDANT, SLATE POINT/PENDANT, BONE TOOLS, COBBLE TOOLS

29

Kenneth M. Ames and features, imposing considerable analytical costs (e.g., Samuels this volume, Smith this volume). Along the Lower Columbia River, for example, crews under my direction have taken a decade to extensively sample three houses, one at the Meier site (Ames et al. 1992, Smith this volume) and two at Cathlapotle (Ames et al. 1999). The Meier house is approximately 30 x 14 m, while the smaller of the sampled Cathlapotle houses is 20 x 10m and the larger is 69 x 10 m. It will take many years to develop a large excavated sample of such structures from many parts of the coast. However, empirical overload is only one part of the difficulty, and not the primary one. A lack of theory is the primary problem. There is no theory about hunter-gatherer households. One can search Kelly’s magisterial book on hunter-gatherers for such a discussion in vain (Kelly 1995). It is therefore our opportunity to develop such theory. The theory does not need to be made from whole cloth. I have briefly reviewed two major streams of theory in this paper: household theory (e.g., Fricke 1986; Netting 1993; Wilk 1997) and House theory (e.g., Carsten and Hugh-Jones 1995; etc.). As different as these approaches can seem to be, they have some common, or at least somewhat overlapping, assumptions about households and Houses: • They are central, even fundamental, institutions for their societies; • They are links (either in reality or analytically) between micro and macro social levels (the individual person and broader institutions); • They are important social/historical actors—or they are real units of adaptation (depending on one’s theoretical predilections); • Houses (and perhaps households) as social actors (or adaptive actors) are significant for understanding the evolution of permanent social inequality and for social and economic change. These assumptions explicitly or implicitly underlay most of the household archaeology done on the Northwest Coast. However, they need to be treated more as testable hypotheses, rather than as assumptions or assertions framing research. Additionally, they do not, in and of themselves, tell us much about how and why households actually accomplish those things. Nor do they answer the three questions posed earlier in this paper: • How to account for the evolution of House societies (assuming such things exist)?

the northern coast, the Paul Mason site. This site contains evidence of 12 rectangular structures, 10 of which form two rows facing the Skeena River in its Kitselas Canyon. These rectangular structures are probably evidence of households on this portion of the Northwest Coast by c.1500 B.C.. Coupland (1985, 1988) has suggested that because these 10 structures display no significant difference in floor area, they indicate that ranking had not yet developed at the time this village was occupied. Archer (2001) mapped and dated house depression in Prince Rupert Harbor on the surfaces of 14 middens in Prince Rupert Harbor and came to conclusions similar to Coupland’s. Archer argues that ranking developed in the harbor around A.D. 100. However, the data sketched here demonstrate the opposite: that ranking existed in Prince Rupert Harbor a millennium earlier. There appears to have been at least three levels of ranking: the level of individual, of cemeteries, and of settlements or communities. These data raise the distinct possibility that the form of Northwest Coast household that we investigating, and discussing in this volume, evolved at the same time as did the ranking system.3 It is possible that the two are the opposite sides of the same organizational coin. In other words, ranking did not evolve out of this particular form of household on the coast, but that the two coevolved from some previous form of organization. Further, they evolved in a broader context than just the household. One plausible part of this context is long-term inter-house competition for members, clients and patrons. Summary and Conclusions This paper rests on the belief that despite 15 years of productive research and an increasing number of publications, Northwest Coast archaeologists still know relatively little about Northwest Coast households, particularly about how and why they operated as they did. This is despite the centrality of households in some theories about the evolution of social inequality on the coast (e.g., Ames and Maschner 1999). Some of our relative ignorance is a consequence of the excavation and analytical scales and demands imposed by household archaeology itself. Other archaeologists (e.g., Coupland 1996, Hayden 1997) have also commented on this. Northwest Coast houses are big and costly in time and money to excavate. They are generally complex as archaeological deposits and often rich in artifacts, ecofacts 30

Thinking about Household Archaeology on the Northwest Coast • •

How to account for their persistence? How to account for their role, if any, in the evolution of social hierarchy. Turning to the first question, it seems clear that households and perhaps Houses evolve in a wide range of social and economic contexts. Netting (1993) stresses that small households, what he calls householders, will evolve under conditions of intensive agricultural production, particularly where landholdings are small. Larger households are more likely, according to him; to evolve under what might be termed extensive agricultural practices, as in, for example, swiddening. Wilk and Rathje (1982b) argued that large households evolve where “economic opportunities” are diverse or scattered, either in space or time (also see Watanabe 1968) and require simultaneous task organization, while small households evolve under conditions where linear task organization is more effective. What is important here is not basic economic form but such things as whether the economy is land and labor intensive or extensive and how tasks are distributed in time and space. Houses clearly evolve among both hunter-gatherers and farmers, although at present the Northwest Coast represents the only well documented ethnographic example of the former. However, Natufian burial practices (e.g., Byrd and Monahan 1995) may suggest the presence of Houses among those ancient hunter-gatherers, and Hayden’s work suggests their presence at Keatley Creek. It seems clear the issues of property and resource control and transference central to Houses crosscut economies. The concept of House as an entity to control and transmit property is quite close to Woodburn’s concept of delayed–return foragers (Woodburn 1980). The second question is about persistence: how do Houses, and long-lived households among hunter-gatherers, persist for long-periods of time. It seems quite likely that many, if not all, households on the Northwest over the past 3,000 years or so were affected by their demographic cycle, despite the relatively large size of households and Houses on the coast. Recruitment of new members was no doubt an on-going problem for all Houses, but particularly the smaller ones. In fact, I think we can assume that ongoing recruitment was a basic parameter of household management on the coast. However, archaeological data from the coast, as well the model developed here, suggests that very long-term persistence of some Houses is plausible. The model suggests that while the live span of a single Gitksan House can be expected to be about 31

250 years, that out of 100 Houses, three can be expected to last 1000 years. Boyd (1985) estimates the population of the northern Northwest Coast at the beginning of the nineteenth century at approximately 45,000 people. If the average House had 20 members, there were some 2,250 Houses on the northern coast at that time. The model developed here suggests that of that number, about 418 would be left after 500 years and 78 after 1,000 years. The actual number would be the result of good management and good luck. The point here is that long-term persistence of Houses is possible. On the other hand, 2,172 Houses would have failed during that millennium, and their membership redistributed to old and new Houses. I suggested that one consequence of this circumstance is that Northwest Coast societies were probably two–tier affairs, with a core of successful, long-lived Houses, and a second group of more recent Houses. I also suggested that the odds of survival were generally stacked against the latter. Long-term House success would depend, in part, on managing the demographic cycles of households and Houses and on managing risk, risk both in term of variability in the outcome of resource procurement but also in terms of the costs of failure, which would vary according to, among other things, the household cycle. While in most analyses of risk, the unit of analysis is the individual, here the household becomes the level of risk taking. Household management could be risk aversive, risk seeking or a mix of both. Large, successful Houses probably pursued a mixed strategy in order to generate the economic surpluses they needed. Many observers assume that peasant households will generally be risk aversive. However, there seems no pressing theoretical reason to make that assumption about Northwest Coast or forager households. Finally, the last question asked was about the role of Houses in the evolution of social inequality. One of the implications of the discussion of House longevity is that long-lived Houses would accumulate higher prestige and rank simply by virtue of their longevity, which would probably be a consequence in part of their long-term success in recruiting members to sustain themselves. Modeling (Boone 2000; Diehl 2000) also suggests that patron-client relations contribute to household survival in variable environments, creating inequality. However, the archaeological data presented for northern British Columbia at least raises the possibility that Houses on the Northwest Coast evolved at the same time as linear villages and the earliest form of the ranking system. There-

Kenneth M. Ames References Cited

fore, Houses, I would suggest, were not actors in the initial evolution of inequality of the coast, but were a part and a consequence of that evolution (see Ames and Maschner 1999). It also seems clear that this development had strong extra-household dimensions, at the level of the community and the locality or sub region. This discussion leaves unanswered the question as to why polities did not precipitate out of these larger–scale dynamics on the Northwest or among other complex hunter-gatherers. I suspect the answer to that question will emerge as we learn more about how Northwest Coast households and Houses did what they did.

Adams, J. W. 1973 The Gitksan Potlatch, Population Flux, Resource Ownership and Reciprocity. Holt, Rinehart, and Winston of Canada, Toronto. Ames, K. M. 1985 Hierarchies, Stress and Logistical Strategies among Hunter-Gatherers in Northwestern North America. In Prehistoric Hunter Gatherers, The Emergence of Cultural Complexity, edited by T. D. Price and J. Brown, pp. 155-180. Academic Press, New York. 1991 Sedentism, a Temporal Shift or a Transitional Change in Hunter-Gatherer Mobility Strategies. In Between Bands and States: Sedentism, Subsistence, and Interaction in Small Scale Societies, edited by S. Gregg, pp. 108-133. Southern Illinois University Press, Carbondale. 1991 The Archaeology of the Longue Durée: Temporal and Spatial Scale in the Evolution of Social Complexity on the Southern Northwest Coast. Antiquity 65:935-945. 1995 Chiefly Power and Household Production on the Northwest Coast. In Foundations of Social Inequality, edited by T. D. Price and G.M. Feinman, pp. 155-181. Plenum Press, New York. 1996 Life in the Big House: Household Labor and Dwelling Size on the Northwest Coast. In People Who Lived in Big Houses: Archaeological Perspectives on Large Domestic Structures, edited by G. Coupland and E. B. Banning, pp. 131-150. Monographs in World Prehistory, no. 27. Prehistory Press, Madison. 1998 Economic Prehistory of the North British Columbia Coast. Arctic Anthropology 35(1):68-87. 2000 Kennewick Man: Cultural Affiliation Report, Chapter 2, Review of the Archaeological Data. USDOI, National Park Service. 2001 Slaves, chiefs and labour on the northern Northwest Coast. World Archaeology 33(1):1-17.. 2005 The North Coast Prehistory Project Excavations in Prince Rupert Harbour, British Columbia: The Artifacts. British Archaeological Reports: International

Notes House, “H” capitalized (and italicized at the beginning of a sentence), refers to the House in Lévi-Strauss’ sense, whereas house, lowercase “h”, refers to the dwelling. 2 Missing from the archaeological estimates are houses with more than 25 people, of which there are a number in Adams’ data set. This could be due to 1) estimates based on floor area underestimate household sizes, or 2) the very large houses are a development of the Modern period. However, the absence of large houses from the estimates does not materially affect the arguments in this paper. The point here is merely that premodern and modern households were generally the same size. 3 Coupland et al. (2001:243) suggest that the long-term success of Houses was due to human agency, to “the strategic use of social mechanism by opportunistic individuals like Ligeex [see Martindale this volume] and Ts’baasa. In the case of Ligeex and the Gispaxlo’ots, high rank was achieved through trade marriage and warfare.” This success was achieved, as Martindale shows, during the fur trade era. However, it is likely that before contact individuals like Ligeex tried to expand the scope of their power and influence. 4 It is possible that there much earlier occupations—c. 8000 b.p.—at Boardwalk. However, that possibility remains only a possibility at present and is not relevant to this discussion in any case. Therefore, I use the standard dates for initial occupation at the site. 5 It is possible that ranking predates these households. There are suggestions of status differentiation in burials excavated at the Blue Jackets Creek site on the Queen Charlotte Islands. These burials date to between 2250 B.C. and 3850 B.C. (Ames 2001). 1

32

Thinking about Household Archaeology on the Northwest Coast Series 1342. John and Erica Hedges Ltd., Oxford. Ames, K. M. and H. D. G. Maschner 1999 Peoples of the Northwest Coast: Their Archaeology and Prehistory. Thames and Hudson Ltd., London. Ames, K. M. , D.F. Raetz, S. C. Hamilton and C. McAfee 1992 Household archaeology of a southern Northwest coast plank house. Journal of Field Archaeology, 19(3):275-290 Ames, K.M., C.M. Smith, W.L. Cornett, E.A. Sobel, S.C. Hamilton, J. Wolf and D. Raetz 1999 Archaeological Investigations at 45CL1, Cathlapotle (1991–1996), Ridgefield Wildlife Refuge, Clark County, Washington: A Preliminary Report. US Department of Interior, Fish and Wildlife Service, Region 1. Cultural Resource Series Number 1. Copies available, Portland State University, Portland, Or. Archer, D. J. W. 2001 Village Patterns and the Emergence of Ranked Society in the Prince Rupert Area. In Perspectives on Northern Northwest Coast Prehistory, edited by J. S. Cybulski, pp. 203-223. Mercury Series, Archaeological Survey of Canada Paper 160. Canadian Museum of Civilization, Hull. Arnold, J. E. 1996 Organizational Transformations: Power and Labor among Complex Hunter-Gatherers and other Intermediate Societies. In Emergent Complexity: The Evolution of Intermediate Societies, edited by J.E. Arnold, pp. 59-73. International Monographs in Prehistory, Ann Arbor. Arnold, J. (editor) 2001 The Origins of a Pacific Coast Chiefdom. University of Utah Press, Salt Lake City. Bamforth, D. B. and P. Bleed 1997 Technology, Flaked Stone Technology and Risk. In Rediscovering Darwin: Evolutionary Theory in Archaeological Explanation, edited by C. M. Barton and G. A. Clark, pp. 109-140. Archaeological Papers of the American Anthropological Society No. 7. Bar-Yosef, O. 1998 The Natufian Culture in the Levant: Threshold to the Origins of Agriculture. Evolutionary Anthropology 6(5):159177.

Blanton, R. E. 1994 Houses and Households: A Comparative Study. Plenum Press, New York. Boone, J. L. 2000 Status Signaling, Social Power, and Lineage Survival. In Hierarchies in Action: Cui Bono?, edited by M. W. Diehl, pp. 84112. Center for Archaeological Investigations Occasional Paper No. 27. Southern Illinois University, Carbondale. Boyd, R. T. 1985 The Introduction of infectious Diseases among the Indians of the Pacific Northwest, 1774-1874. Ph.D. Dissertation, University of Washington. Microfilms, Ann Arbor. Byrd, B. F. and C. M. Monahan 1995 Death, Mortuary Ritual and Natufian Social Structure. Journal of Anthropological Archaeology 14:251-287. Carlson, R. L. 1994 Trade and Exchange in Prehistoric British Columbia. In Prehistoric Exchange Systems in North America, edited by T. G. Baugh and J. E. Ericson, pp. 307-361. Plenum Press, New York. Carsten, J. and S. Hugh-Jones (editors) 1995 About the House: Lévi-Strauss and Beyond. University of Cambridge Press, Cambridge, England. 1995 Introduction. In About the House: LéviStrauss and Beyond, edited by J. Carsten and S. Hugh-Jones. University of Cambridge Press, Cambridge. Chang, K. C. 1968 Toward a Science of Prehistoric Society. In Settlement Archaeology, edited by K. C. Chang, pp. 1-9. National Press Books, Palo Alto. Coupland, G. 1985 Household variability and status differentiation at Kitselas Canyon. Canadian Journal of Archaeology 9(1):39-56. 1988 Prehistoric Economic and Social Change in the Tsimshian Area. Research in Economic Anthropology (Supplement) 3:211-245. 1996 The Evolution of Multi-Family Households on the Northwest Coast of North America. In People Who Lived in Big Houses: Archaeological Perspectives on Large Domestic Structures, edited by G. Coupland and E. B. Banning, pp. 121-130. Monographs in World Prehistory, No. 27. Prehistory Press, Madison. 33

Kenneth M. Ames Gallant, T. W. 1991 Risk and Survival in Ancient Greece. Stanford University Press, Stanford. Gillespie, S. D. 2000 Lévi-Strauss: Maison and Société a Maisons. In Beyond Kinship: Social and Material Reproduction in House Societies, edited by R. A. Joyce and S. D. Gillespie, pp. 22-52. University of Pennsylvania Press, Philadelphia. Goody, J. (editor) 1958 The Developmental Cycle in Domestic Groups. University of Cambridge Press, Cambridge, England. Habu, J. 1996 Subsistence-settlement systems and intersite variability in the Moroiso phase of the early Jomon period of Japan. Ph.D. Dissertation, McGill University. Hayden, B. 1995 Pathways to Power: Principles for Creating Socioeconomic Inequalities. In Foundations of Social Inequality, edited by T. Douglas Price and Gary M. Feinman, pp. 15-86. Plenum, Inc, New York. 1997 The Pithouses of Keatley Creek. Harcourt Brace College Publishers, Fort Worth. Hayden, B. and A. Cannon 1982 The Corporate Group as an Archaeological Unit. Journal of Anthropological Archaeology 1:132-158. Hayden, B., G. A. Reinhardt, D. Holmberg and D. Crellin 1996 Space Per Capita and the Optimal Size of Housepits. In People Who Lived in Big Houses: Archaeological Perspectives on Large Domestic Structures, edited by G. Coupland and E. B. Banning, pp. 151-163. Monographs in World Prehistory, No. 27. Prehistory Press, Madison. Hayden, B. and J. M. Ryder 1991 Prehistoric Cultural Collapse in the Lillooet Area. American Antiquity 56:5065. Hendon, J. A. 1996 Archaeological Approaches to the Organization of Domestic Labor: Household Practices and Domestic Relations. Annual Reviews in Anthropology. vol. 25. Annual Reviews Inc., Palo Alto. Imamura, K. 1996 Prehistoric Japan, New Perspectives on Insular East Asia. University of Hawaii Press, Honolulu.

Coupland, G. and E. G. Banning (editors) 1996 People Who Lived in Big Houses: Archaeological Perspectives on Large Domestic Structures. Prehistory Press, Madison. Coupland, G., G. Bissel and S. King 1993 Prehistoric Subsistence and Seasonality at Prince Rupert Harbour: Evidence from the McNichol Creek Site. Canadian Journal of Archaeology 17:59-73. Cybulski, J. S. 1993 A Greenville Burial Ground: Human Remains in British Columbia Coast Prehistory. Archaeological Survey of Canada, Canadian Museum of Civilization, Ottawa. 2001 Human Biological Relationships for the Northern Northwest Coast. In Perspectives on Northern Northwest Coast Prehistory, edited by J. S. Cybulski, pp. 107-144. Mercury Series, Archaeological Survey of Canada Paper 160. Canadian Museum of Civilization, Hull. Diehl, M. W. 2000 Some Thoughts on the Study of Hierarchies. In Hierarchies in Action: Cui Bono?, edited by M. W. Diehl, pp. 11-30. Center for Archaeological Investigations Occasional Paper No. 27. Southern Illinois University, Carbondale. Donald, L. and D. H. Mitchell 1975 Some correlates of local group rank among the Southern Kwakiutl. Ethnology 14(3):325-346. Doyel, D. E. (editor) 1987 The Hohokam Village: Site Structure and Organization. AAAS Publication 8715, Southwestern and Rocky Mountain Division of the American Association for the Advancement of Science, Glenwood Springs, CO Ellen, R. 1982 Environment, Subsistence and System, The Ecology of Small-Scale Social Formations. Cambridge University Press, Cambridge. Flannery, K. V. (editor) 1976 The Early Mesoamerican Village. Academic Press, New York. Fricke, T. E. 1986 Himalayan Households, Tamang Demography and Domestic Processes. Studies in Cultural Anthropology 11. UMI Research Press, Ann Arbor. 34

Thinking about Household Archaeology on the Northwest Coast logical Association of the University of Calgary, Calgary. Maclachlan, M. D. (editor) 1987 Household Economies and Their Transformations. University Press of America, Lanham. Madden, M. 1983 Social Network Systems amongst HunterGatherers Considered within Southern Norway. In Hunter-Gatherer Economy in Prehistory, pp. 191-200. University of Cambridge Press, Cambridge. Marquardt, William H. 1999 The Emergence and Demise of the Calusa. In Societies in Eclipse: Eastern North America at the Dawn of European Colonization, edited by R. Mainfort D. Brose, and C.W. Cowan. Smithsonian Institution, Washington D.C. Marshall, Y. 2000 Transformations of Nuu-chah-nulth Houses. In Beyond Kinship: Social and Material Reproduction in House Societies, edited by R. A. Joyce and S. D. Gillespie, pp. 73-102. University of Pennsylvania Press, Philadelphia. Maschner, H. D. G. 1992 The Origins of Hunter-Gatherer Sedentism and Political Complexity: A Case Study from the Northern Northwest Coast. PhD Dissertation, University of California Santa Barbara. Matson, R.G. 2003 The Coast Salish House: Lessons from Shingle Point,Valdes Island, British Columbia. In Emerging from the Mist, edited by R.G. Matson, G. Coupland and Q. Mackie, pp 76-104. UBC Press, Vancouver Mitchell, D. and L. Donald 1988 Archaeology and the Study of Northwest Coast Economies. In Prehistoric Economies of the Northwest Coast, pp. 293-351. Research in Economic Anthropology. Supplement 3. JAI Press, Inc., Greenwich. Moss, M. L. 1993 Shellfish, Gender and status on the Northwest Coast: Reconciling Archaeological, Ethnographic, and Ethnohistoric Records of the Tlingit. American Anthropologist 95:631-652. Muller, J. 1997 Mississippian Political Economy. Interdisciplinary Contributions to Archaeology. Plenum, New York.

Isaac, B. L. (editor) 1988 Prehistoric Economies of the Pacific Northwest. JAI Press, Inc., Greenwich. Joyce, R. A. and S. D. Gillespie (editors) 2000 Beyond Kinship: Social and Material Reproduction in House Societies. University of Pennsylvania Press, Philadelphia. Kelly, R. L. 1995 The Foraging Spectrum: Diversity in Hunter-Gatherer Lifeways. Smithsonian Institution Press, Washington D. C. Kent, S. 1984 Analyzing Activity Areas:An Ethnoarchaeological Study of the Use of Space. University of New Mexico Press, Albuquerque. Kirch, P.V. 2000 Temples as “Holy Houses”: The Transformation of Ritual Architecture in Traditional Polynesian Societies. In Beyond Kinship: Social and Material Reproduction in House Societies, edited by R. A. Joyce and S. D. Gillespie, pp. 103- 115. University of Pennsylvania Press, Philadelphia. Kobayashi, T. 1992 Patterns and Levels of Social Complexity in Jomon Japan. In Pacific Northeast Asia in Prehistory, edited by C. M. Aikens and S. N. Rhee, pp. 91-98. Washington State University Press, Pullman. Kramer, C. 1982 Village Ethnoarchaeology: Rural Iran in Archaeological Perspective. Academic Press, New York. Kuijt, I. (editor) 2000 Life in Neolithic Farming Communities: Social Organization, Identity, and Differentiation. Kluwer Academic/Plenum Publishers, New York. Larson, M. L. 1997 Housepits and Mobile Hunter-Gatherers: A Consideration of the Wyoming Evidence. Plains Anthropologist 42(161):353-369. Lévi-Strauss, C. 1982 The Way of the Masks (Trans. by S. Modelski). University of Washington Press, Seattle. MacDonald, G. F. 1969 Preliminary culture sequence from the Coast Tsimshian area, British Columbia. Northwest Anthropological Research Notes 3(2):240-254. MacEachern, S., D. J. W. Archer and R. D. Garvin (editors) 1989 Households and Communities. Archaeo35

Kenneth M. Ames Varien, M. D. 1999 Sedentism and Mobility in a Social Landscape. University of Arizona Press, Tucson. Watanabe, H. 1968 Subsistence and Ecology of Northern Food Gatherers with Special Reference to the Ainu. In Man the Hunter, edited by R. B. Lee and I. DeVore, pp. 69-77. Aldine Publishing Co., Chicago. Wattenmaker, P. 1998 Household and State in Upper Mesopotamia. Smithsonian Institution Press, Washington D.C. Welsch, R. L. and J. E. Terrell 1998 Material Culture, Social Fields, and Social Boundaries on the Sepik Coast of New Guinea. In The Archaeology of Social Boundaries, edited by M. T. Stark, pp. 51-77. Smithsonian Institution Press, Washington DC. Wilk, R. R. 1997 Household Ecology: Economic Change and Domestic Life Among the Kekchi Maya. Northern Illinois University Press, DeKalb. Wilk, R. R. and W. Ashmore (editors) 1988 Household and Community in the Mesoamerican Past. University of New Mexico Press, Albuquerque. Wilk, R. R. and W. L. Rathje (editors) 1982a Archaeology of the Household: Building a Prehistory of Domestic Life. 25. American Behavioral Scientist 25(6). Wilk, R.R. and W.L. Rathje 1982b Household Archaeology. American Behavioral Scientist 25(6):613-616. Wobst, M. 1974 Boundary Conditions for Paleolithic social systems: a simulation approach. American Antiquity 39:147-179. Woodburn, J. 1980 Hunter-Gatherers Today and Reconstructing the Past. In Soviet and Western Anthropology, edited by E. Gellner, pp. 94-118. Columbia University Press, New York. Yanagisako, S. J. 1979 Family and Household, the Analysis of Domestic Groups, pp. 161-205. Annual Review of Anthropology. vol. 8. Annual Reviews, Inc. Palo Alto.

Netting, R. M. 1993 Smallholders, Householders: Farm Families and the Ecology of Intensive, Sustainable Agriculture. Stanford University Press, Stanford, CA. Pauketat, T. R. 1996 The Foundations of Inequality in a Simulated Shan Community. Journal of Anthropological Archaeology 15(3):219-236. Rogers, J. D. and B. Smith (editors) 1995 Mississippian Communities and Households. University of Alabama Press, Tuscaloosa. Sahlins, M. 1972 Stone Age Economics. Aldine, Chicago. Sandstrom, A. R. 2000 Toponymic Groups and House Organization: The Nahuas of Northern Veracruz, Mexico. In Beyond Kinship: Social and Material Reproduction in House Societies, edited by R. A. Joyce and S. D. Gillespie, pp. 53-72. University of Pennsylvania Press, Philadelphia. Schwartz, G. M. and S. E. Falconer (editors) 1994 Archaeological Views of the Countryside: Village Communities in Early Complex Societies. Smithsonian Institution, Washington D.C. Stearns, M. L. 1984 Succession to Chiefship in Haida Society. In The Tsimshian and Their Neighbors of the North Pacific Coast, edited by J. Miller and C. M. Eastman, pp. 190-219. University of Washington Press, Seattle. Suttles, W. 1990 Introduction. In Handbook of North American Indians, Vol. 7: The Northwest Coast, edited by W. Suttles, pp. 1-15. The Smithsonian Institution, Washington, DC. Tringham, R. E. 1991 Households with Faces: The Challenge of Gender in Prehistoric Architectural Remains. In Engendering Archaeology: Women and Prehistory, edited by J. M. Gero and M. W. Conkey, pp. 93-131. Basil Blackwell Ltd, Oxford. 2000 The Continuous House: The View from the Deep Past. In Beyond Kinship: Social and Material Reproduction in House Societies, edited by R. A. Joyce and S. D. Gillespie, pp. 115-134. University of Pennsylvania Press, Philadelphia.

36

Houses And Domestication On The Northwest Coast Yvonne Marshall

similar type, without, however, losing the memory of its origin” (Lévi-Strauss 1982:164-165). Kinship and descent are not abandoned but their influence on the way social groups are defined is argued to follow from, rather than determine, people’s choices as to place of residence. Resulting groups may describe themselves by taking a “collective name derived from that of a founder” or they may use the name of their place of origin, or they may “adopt an honorific name such as ‘The Rich’ or ‘The Great’”(Lévi-Strauss (1982:165). But whichever kind of name is chosen, all groups are materially constituted by co-residence. A House consists of the people who choose to live together within a specific dwelling. Place, generally, and residence, in particular, holds the casting vote in deciding who is and who is not a member of any specific group at any point in time. Although his model was built primarily on the Kwakwaka’wakw whom he knew well from Boas’ work, Lévi-Strauss intended his model to apply to the Northwest coast as a whole, from the Yurok of California north to the Tlingit of Alaska (LéviStrauss 1982:171-173). He argued that even among the northern groups where in recent centuries, if not earlier, kinship has been matrilineal rather than cognatic, group membership remained negotiable. Speaking of these northern groups Lévi-Strauss (1982:171) asks:

In previous papers I have argued that the societies of the Northwest Coast, both prior to and following European contact, fit Lévi-Strauss’s concept of House societies well, and that the House model offers archaeologists a valuable analytical tool for understanding continuity and change (Marshall 1989, 2000a, b). In this paper I suggest that the explanatory power of the House model is much enhanced when used in conjunction with Peter Wilson’s (1988) concept of domesticated societies. Together these theories offer archaeologists a new way of both conceptualizing and investigating the development of Northwest Coast societies. House Societies Lévi-Strauss’s aim in developing the concept of House-based social organization was to make sense of societies that do not fit anthropological kinship models (Carsten and Hugh-Jones 1995; Joyce and Gillespie 2000; Lévi-Strauss 1982). Because cognatic societies trace descent through both the female and male lines, kinship links do not define exclusive social groups and people are potentially members of several different social groups. The Kwakwaka’wakw are a cognatic society. Despite repeated attempts, Franz Boas was never able to fit them into established kinship classifications and he eventually abandoned these efforts, using instead the Kwakwaka’wakw’s own terms and definitions for social groups (1897; 1920; 1921). After Boas’ death, Lévi-Strauss returned to the problem of cognatic kinship. He proposed that in some cognatic societies, place rather than kinship was the primary principle defining sociopolitical groups (Lévi-Strauss 1982; Gillespie 2000a, b). In reference to the Kwakwaka’wakw and Salish peoples he argued: “In both cases, the basic units of social structure seem shaped by a supposed descent from a mythic ancestor who built his home in a definite place, even if this community later left its ancestral land to unite with other communities of a

How can very strictly formulated rules of descent and succession be applied so loosely? The question does not arise in the case of the Kwakiutl, the Nootka, and the Bella Bella, who fully exploit…the flexibility of their cognatic system, and can thus disguise all sorts of socio-political manoeuvres under the veneer of kinship. By contrast, the rules of the Tsimshian, the Haida, and the Tlingit seem at first too rigid to allow one to shift from one plane to the other; furthermore, the role of kinship proper is more limited, and consequently, the combinations inspired by other motives can be seen more

37

Yvonne Marshall openly. In both cases, the local life inextricably meshes the ties that result from political and economic history…with the ties that are based on real or supposed genealogies.

The key here is the slippage or mutability between place, kinship, and economic or political interests. Membership in a social group is never finally resolved because it is always subject to shifts in the interests of both individuals and groups. Kinship, place and politics can each be judiciously invoked to include or exclude individuals and groups according to context or event. Changes are always possible and take place regularly. As Lévi-Strauss (1982:176) puts it, this allows people the “freedom to disguise social or political maneuvers under the mantle of kinship.” The roles of place and residence are therefore critical in House societies. It is by reference to place, both in the form of a group origin place and in the sense of taking up residence in a particular place or House, that kinship, economics and politics can be made to stand for each other, thus allowing membership in groups to remain fluid. By taking up residence within a House structure a person or group enacts and lives the set of kin, political and economic relationships they make claim to. Selected social connections are made real because they are materially realized in everyday life. In the same way responsibility for group stability and continuity is also invested in place rather than in kinship and descent. Houses in both material and social senses are, in principle, perpetual establishments. However, it is in the visible permanence of a physical house structure that claims for social continuity are grounded, so self-conscious efforts are made to disguise social and political shifts behind the material façade of permanence provided by the House dwelling structure (Gillespie 2000a:12-14; Marshall 2000a, 2000b). In summary, House societies consist of sociopolitical groups that define themselves on the basis of place, particularly residence within a dwelling, while maintaining open flexible kinship systems and fluid membership. Economics follows from residence. In contrast, theories based in the concept of household generally work from a definition of the household as an economic unit in which members share or divide labor and resources within the group. In addition to households and House, another concept popular with archaeologists is the corporate group. It is defined on the basis that the group owns property, such as land, rights and privileges in common (see Ames this volume). The key difference between Houses and households or 38

corporate groups lies in what is privileged as the defining and therefore most stable element identifying socio-political groups. Whereas households and corporate groups are defined by and locate continuity in economics and property, Houses are defined by and locate continuity in place. Domesticated Societies This privileging of place in House societies dovetails closely with ideas developed by Peter Wilson (1988). Wilson reworks the concept of domestication into a process that humans apply to themselves through the creation of built environments. In doing so he also shifts our attention away from kinship and economics to place, arguing that the most important transformation to have taken place in human history was not the shift from a hunter/gatherer economy to agriculture, but the shift from open living to life in a built environment; from open to domesticated societies (Wilson 1988:24-25). Domesticated societies build substantial dwellings and are defined by the distinctive ways in which they locate themselves in, and construct around them, the physical and social landscapes in which they reside. The hunter/gatherer pins ideas and emotions onto the world as it exists: the landscape is turned into a mythical topographical map, a grid of ancestor tracks and sacred sites…. A construction is put upon the landscape rather than the landscape undergoing a reconstruction, as is the case among sedentary peoples, who impose houses, villages, and gardens on the landscape, often in place of natural landmarks. Where nomads read or even find cosmological features in an already existing landscape, villagers tend to represent and model cosmic ideas in the structures they build. (Wilson 1988:50)

Wilson’s point is that domestication changes the relationship between people and places and in doing so transforms the way people relate to each other. “The anchoring of a person that comes with domestication results in the identification of person with location and location with person, a merging of person and place which differs from the idea of focus that specifies the relation between hunter/gatherer and location” (Wilson 1988:71). Non-domesticated or open societies build, at most, only flimsy temporary shelters and they arrange their living space into an unbounded focal area. There is no perimeter or boundary - only a focus or center. In contrast, domesticated living spaces

Houses And Domestication On The Northwest Coast scends an individual’s lifetime is a distinctive feature of domesticated societies. Consequently, once “domestication becomes the condition for living” so too, “burial becomes a condition of death” (Wilson 1988:129). Thus domesticated and House-based social groups make themselves material, visible and perpetual in both life and death by displaying themselves in imposing architectonic constructions. These can include substantial dwellings, cemeteries and tombs, platforms arranged with feast foods, and high mounds of formally stacked gifts. The concepts of House societies and domesticated societies are closely related but not the same. While domesticated societies may be organized into Houses, they need not be. However, a House society is by definition a domesticated society. Wilson’s notion of domestication therefore incorporates the concept of House societies (Wilson 1988:75) to the extent that House societies may be considered one form of domesticated society. Although House societies display the key characteristics of domesticated societies they also depend on maintaining flexibility in the way social groups are constituted, a feature Wilson considers characteristic of open rather than domesticated societies. House societies are then a distinctive form of domesticated society. Significantly, it is the same seemingly contradictory feature, namely flexible group composition, that both gave rise to the concept of House societies and sets House societies apart within the Wilson’s wider concept of domestication.

have clearly demarcated edges or boundaries which divide groups off from one another. This shift to living in a built environment, with its myriad of fixed boundaries, had profound social implications. As Wilson (1988:61) puts it “architecture is a materialization of structure” so once architecture is adopted “as a permanent feature of life” people begin to live, literally and materially, within that structure. He goes on to say: The most far-reaching difference between open and domesticated societies…is that in the former the sense of structure and constraint is tacit, subjective, personal, and focused, whereas in the latter it is explicit, embodied, objective, and externally bounded. The source of this difference, its origin, lies in the adoption of architecture as the permanent living environment. (Wilson 1988:77-78)

In open societies, newcomers and visitors are unproblematically integrated into the everyday life of the group they seek to join. There are no special rules of etiquette to distinguish host from guest. In contrast, by arranging people into a series of sensorially separated, clearly demarcated spatial domains, domestication fosters the development of a formal host/guest relationship and creates conditions conducive to the advent of hospitality (Wilson 1988:97). In short, domesticated living creates neighbors; groups of people who live in close physical proximity but who are divided from one another by physical and social boundaries. The adoption of a set of formally structured host/guest interactions allows individuals and groups to move across those boundaries in unthreatening, socially appropriate ways. It also creates conditions that foster the controlled exposure or display of the inside of a dwelling to outside persons or groups through events such as feasts or exchanges (Wilson 1988:92). While those who live within each dwelling coalesce into tight social groups by living together and by describing themselves in the language of kinship, each group is increasingly cut off from, even alienated from other groups living in close proximity but within other houses. Regulation of relationships between people living in different houses, and neighbors in particular, therefore starts to require formal social channels (Wilson 1988:168-169). Lévi-Strauss and Wilson also argue that in House and domesticated societies built environments, particularly houses, are key vehicles for social continuity. Social groups are constituted by and provided with continuity through reference to the built environment. This need for social groups to assert continuity, even permanence, which tran-

Houses and Domestication in Northwest Coast Prehistory In anthropological theory, the Northwest Coast has commonly played the role of misfit or exception and kinship theory is just one example of this. Another example is economic theory. The addition of “complex” to hunter/gatherer or “affluent” to forager allows the Northwest Coast to be accommodated within the foundational hunter/gatherer versus agriculturalist economic dichotomy (Ames 1994; Mackie 1998). Similarly, as Ames (this volume) points out, ideas from household archaeology are used by Northwest Coast archaeologists, but the transposing of theory developed in reference to peasant societies to the Northwest Coast context is problematic. In contrast, the concepts of House and domesticated societies have arisen out of attempts to make sense of the specificities of Northwest Coast societies. This is explicitly the case with House societies and is implied in the way Wilson uses Northwest Coast societies as illustrative examples 39

Yvonne Marshall What follows is a brief sketch of Northwest Coast prehistory rethought in terms of the ways place has structured social life. This review draws heavily on Ames and Maschner’s (1999) thematic overview of Northwest Coast prehistory and develops several ideas initially suggested there.

of the domestication process. Neither of the two concepts is a qualified version of a theory developed to fit some other time or place, and in their emphasis on understanding how space structures social relations these theories address themselves directly to issues at the heart of Northwest Coast prehistory. As presently told the prehistory of the Northwest Coast is a story about stone tool typologies. A series of temporally and regionally specific cultures are defined on the basis of stone working techniques, e.g., Pebble Tool Tradition, Microblade Tradition and vast numbers of traditions defined by various kinds of projectile points (see, Carlson and Dalla Bona 1996; Matson and Coupland 1995; Suttles 1990). These traditions or culture types are tied into what are argued to be a series of key economic shifts: the intensification of shellfishing beginning around 5500 B.P., the intensification of salmon fishing and the development of food storage techniques around 3500 B.P., and the development of whaling on the west coast of Vancouver Island and probably the Queen Charlotte Islands by 1000 B.P. At the core of this story is a presumption of one-way, linear, causation: environmental change leads to economic and/or technological change which in turn leads to social and/or political change. Most commonly the argument goes: climatic change forces people to invent new artifacts in order to harvest resources differently or more intensively and this leads to the emergence of social inequality and ranking. In this formulation social and political systems are assumed to be stable and to change only in response external factors. They are not considered inherently dynamic and therefore not capable of internal, socially or politically directed change. In this formulation social and political systems are assumed to be stable and to change only in response to external forces. The theories of House and domesticated societies bring with them very different assumptions about the nature of sociopolitical continuity and change. They presume that social systems are inherently dynamic and that both continuity and change are features of their everyday operation. Change is understood to be intrinsic to social systems and not a handmaiden to external forces. This means that both continuity and change must be actively investigated and understood. In rethinking Northwest Coast prehistory using the concepts of House and domesticated societies we not only shift our focus from economics and technology to place, we also shift our starting assumptions about the nature of sociopolitical continuity and change.

Archaic Period: First colonization to 5500 B.P. The established view of the early occupants of the Northwest Coast is that they lived in “lowdensity, highly mobile groups” and had “conservative, foraging lifestyles, making seasonal use of coastal and inland resources” (Matson and Coupland 1995:301). But is this really the case? Ames and Maschner (1999:155-6) make the interesting observation that pithouses are of very great antiquity in Northeast Asia, with the earliest examples in Siberia dating back at least 20,000 years. By the time the Americas were colonized, probably between 15,000 and 10,000 B.P., pithouses were in widespread use throughout Northeast Asia. Some of the earliest houses found in the Americas are also pithouses with the oldest known example located at the site of Anangula in the Aleutian Islands. It is dated to around 8,000 B.P. and was a structure about 1 m deep and 5 m in diameter. People who construct such large, substantial dwellings are, by Wilson’s definition, domesticated, an interpretation supported by the very early occurrence in North America of both interment and cremation burials, including cemeteries. The oldest example is the cremation burial of at least five individuals at the Marmes rockshelter, which is located on the southern Columbia Plateau and dates to around 10,000 B.P. While it is clear that people living on the Plateau at this time were very mobile, these burial practices are indicative of domestication in Wilson’s terms (Wilson 1988:123), and archaeologists routinely interpret them as evidence for “partial or full sedentism, territoriality or the presence of corporate groups with long-lasting ties to particular places” (Ames and Maschner 1999:184). Further evidence pointing to early domesticated societies comes from faunal indicators of seasonal occupation at the site of Namu. Cannon (1996) argues that even during the earliest period of occupation people were definitely present at Namu in October, November, March and June, and were almost certainly there throughout the period from October to June. He then points out: The implication is that Namu was a winter village site from as early as 8000 years ago, and

40

Houses And Domestication On The Northwest Coast Middle Pacific Period: 3500 B.P. to 2000/1500 B.P.

an essentially sedentary pattern of settlement had been established by that time. The further implication is that at least part of the subsistence economy was based on winter storage of salmon. This pattern of subsistence and settlement was maintained for a period of some two thousand years before the major accumulation of shell midden deposits began. (Cannon 1996:107)

During the Middle Pacific Period the formalization of space within settlements intensified further. Villages were now laid out in formal rows of closely spaced houses, commonly facing the shoreline of a beach or river. Ames and Maschner (1999:158-159) call these “linear villages.” The most intensively investigated examples are the Paul Mason and Boardwalk sites, in the Prince Rupert Harbor area. Both sites had two, possibly three, rows of relatively small, rectangular frame houses varying in size from about 4 m by 5 m to 5 m by 10 m. The rear row of houses at the Boardwalk site had an associated cemetery that remained in use for some 700 years. Large midden cemeteries, like the one at the Boardwalk site, became common during the Middle Pacific Period. They included large numbers of individuals buried with a wide range of grave goods. A two-row pithouse village dating to this period was identified at the Katz site on the lower Fraser River and on the southern coast a very large rectangular plank house measuring around 20 m by 6 m was excavated at the Palmrose site. This house was rebuilt several times and was in use for over 1,000 years. It is clear from this evidence that people throughout the coast were developing increasingly “long-lived ties to particular places” (Ames and Maschner 1999:160). As these attachments grew, the spatial arrangement of settlements and the demarcation of living spaces within settlements became increasingly formal. Implied in this process is the emergence of neighbors – groups separated by social and material boundaries but living in close proximity to each other (Wilson 1988:97). Social separation of groups is reinforced by clearly delineated material boundaries in the form of dwelling walls and by the formal arrangement of these dwellings relative to one another. In these circumstances, crossing of the social and material boundaries erected between dwellings becomes problematic and the informal absorption of newcomers and visitors into the everyday social fabric of group life within dwelling is replaced with the structured interactions of host/guest rituals.

These are only few small glimpses, but they do suggest that the first people to enter the Americas and settle on the Northwest Coast brought with them domesticated attitudes to place and landscape. Their domesticated lifeways did not emerge in response to environmental changes on the Coast; they came with them and were then transformed over time by people’s experiences of living there. Early Pacific Period: 5500 B.P. to 3500 B.P. During this period pithouses become increasingly common in the Plateau area and Ames and Maschner (1999:156) propose “an early period of village formation in the interior” called Cascadia Village I. They further suggest that the large shell midden sites that began forming on the coast around this time were probably also part of this pattern of village formation. The oldest substantial structure identified on the Northwest Coast is a rectangular dwelling located at Hatzic Rock. It measures approximately 8 m by 10 m and dates to around 4500 B.P. A large structure was also identified at the Maurer site and a possible post-and-beam structure is suggested for the St. Mungo Cannery site (Ames and Maschner 1999:156). In addition, large cemeteries dated to the Early Pacific Period have been excavated from shell middens at the sites of Blue Jackets Creek, Namu and Pender Canal (Ames and Maschner 1999:186). Many individuals were buried with grave goods; some of which, such as the carved spoons recovered with burials at Pender Canal, were very elaborate (see Ames and Maschner 1999:104). By 5500 B.P. early patterns of domesticated life were changing. They were becoming more formalized in the sense that they were being written into the spatial patterning of dwellings, settlements and burials in increasingly explicit, material ways. While the evidence is much too scanty to know with certainty, it is possible that this shift toward more explicitly demarcated spatial boundaries and places accompanied the formation of early Houses.

Late Pacific Period: 2000/1500 B.P. to Historic Period By the Late Pacific Period linear villages consisting of rows of rectangular plank houses are widespread throughout the Northwest Coast. 41

Yvonne Marshall These are Houses in the Lévi-Straussian sense. In the north, most villages continue to consist of a single row of small, quite regularly sized houses. However, two and three row villages become more common and there is increasing differentiation in house size with an occasional house of exceptional size appearing near the center of some villages. To the south, houses continued to be built in a vast array of sizes although house width was far less variable than length. Width seldom exceeded 15 m but length could reach 200 m or more (Ames and Maschner 1999:163). In both the north and south settlements were strongly linear spaces. However, houses in northern villages were more formally divided off from one another and were internally subdivided into smaller, materially bounded spaces by larger numbers of house walls. On the Plateau living space was never arranged in this linear manner. Pithouses remained common and villages continued to be arranged in informal clusters. A well-documented example is the Keatley Creek site on the Fraser River (Hayden et. al. 1996). Disposal of the dead changed dramatically in the Late Pacific. Compared with the Middle Pacific, few burials exist for this period, and it is assumed that the historic practice of leaving wrapped bodies exposed in secluded locations became the preferred method of disposal. The maintenance of long-term ties to specific settlement places, already a well established practice of considerable antiquity, was further consolidated during the Late Pacific Period. For example, along the west coast of Vancouver Island all of the villages where settlement is known for the Late Pacific Period were occupied continuously into the historic period (McMillan 1999:126). Continuous occupation of this kind was now the rule rather than the exception throughout the Northwest Coast and it is rare to find a village of this period that is briefly occupied and then abandoned. It is perhaps significant that as differentiation among social groups became more explicitly written into the spatial structure of settlements, people stopped expressing social continuity through the repeated internment of the dead within the specially designated space of cemeteries. It appears that as the formalization of living space intensified, living space became the primary focus of social continuity, and the practice of expressing social continuity by juxtaposing living people with the places where ancestors were buried was abandoned. This represents a fundamental shift in the way relationships between people and place were expressed. Alongside these developments, it is easy to see how the controlled exposure of Houses 42

in ritual feasts or potlatches, as ways of making power visible, might escalate into what Wilson (1988) calls the surrealities of power. Summary I argue that the people who initially colonized the Northwest Coast were domesticated in the sense proposed by Peter Wilson. With its rich natural resources the Northwest Coast was a place favorable to the sedentary or semi-sedentary lifeways of the early colonists. Over time they accommodated themselves to the new lands and developed new ways of relating both to each other and to the varied and variable environmental setting in which they lived. Viewed in this way the prehistory of the Northwest Coast peoples becomes a story about the development, refinement and differentiation of a unique kind of domesticated lifeway. As Aubrey Cannon has argued, no single technological, environmental, or cultural change was necessary or sufficient to bring about the Northwest Coast pattern of intensive storage-based production and permanent village settlement. The implication is that regardless of the patterns of economic and cultural transition evident at specific site localities, there were no critical thresholds in Northwest Coast economic prehistory. Developments did not lead in a linear fashion toward the ethnographically recorded pattern of subsistence and settlement. The basic framework was established at a very early date. Subsequent developments in Northwest Coast prehistory will have to be explained in more specific social and historical terms; they cannot be explained with simple reference to technological innovation, cultural migration, or environmental change. (Cannon 1996:110)

Regional Variation in the Late Pacific Period The final section of this paper looks more closely at regional variation within the general patterns of domestication identified above. As already discussed, one of the most striking differences between the north, central and southern parts of the Northwest Coast at the time of European contact was the degree of flexibility built into the indigenous social systems. The Houses of the northern peoples, in both sociopolitical and material senses, were more rigidly delineated than those further south (Ames and Maschner 1999:163). The northern matrilineal descent systems produced

Houses And Domestication On The Northwest Coast beach berms. As at Yuquot each successive house floor, firepit and posthole was located directly above the earlier one (McKenzie 1974). Unfortunately, dates for these floors remain unclear (McMillian 1999:88). At the Ozette wet site five protohistoric houses were excavated (Samuels 1991; this volume). All demonstrate the same concern with maintaining precise placements for firepits and house structures. The most dramatic example is the direct superimposition of House 2 over the earlier House 5. A further unique form of evidence for Houses is the Yuquot Whalers’ Shrine, a ritual structure that was curated, rebuilt and added to over an extended period spanning the late prehistoric and historic periods (Jonaitis 1999). As the above summary shows, excavation evidence is important for establishing the antiquity of Houses and for detailing their size, structure and internal organization. However, to understand how Houses were arranged within settlements and how they fitted into wider settlement patterns, spatial evidence in the form of site size, definition and distribution is needed. Vast quantities of survey data are stored in site record files. While this information is difficult to interpret within an artifact based culture historical framework, it is very open to analyses based on place. If Houses formed the basic unit of settlement, we can expect that over time settlements will increase and decrease in both physical size and population, in whole House units. As a result, settlement size will tend to be modular. In addition, if Houses act as sites of social continuity both the locations they occupy and the area they enclose will show consistency despite the fact that Houses will inevitably move through periods of demographic growth and decline (see Ames this volume for discussion). This curation of House placements should generate coherent surface feature patterning. Variation in this patterning can provide insight into the antiquity and nature of historically recorded regional variation in settlement organization and the degree to which Houses were clearly defined in spatial terms during the Late Pacific Period. The west coast of Vancouver Island and the southern Queen Charlotte Islands are areas remote from modern urban settlement and consist of open coast, maritime environments dominated by steep, rocky shorelines and open exposed beaches (Fig. 1). In such environments locations with protected canoe access are limited, so the number of places where villages can be located are also limited. In these circumstances, relatively complete inventories of late prehistoric habitation sites with intact surface

closely defined social groups who occupied fixed perimeter, gable roof dwellings with floors dug below ground level. Houses built in this style were relatively small and had fixed dimensions. Further south, there was a preference for patrilineal and cognatic descent systems, which generated more fluid social groups. These groups occupied shed roof dwellings of widely divergent sizes that could easily be adjusted in size by adding or removing sections at either end of the structure (see Senday 1977 for example). To investigate the antiquity of these regional differences a comparative analysis of late prehistoric settlement patterns on the west coast of Vancouver Island, in the southern Queen Charlotte Islands, and along the lower Skeena and Nass rivers is presented. The data for Vancouver Island and the Queen Charlottes are analyzed first, and they are then briefly compared to the settlement patterns of the Tsimshian as reconstructed by Martindale (this volume). Excavated settlement evidence from the Queen Charlotte Islands is extremely sparse. A possible hearth dating to circa 9300 B.P. was defined at the Arrow Creek 2 site (Fedje et al 1996:139). More solid evidence for early domestication was recovered at Blue Jackets Creek where 14 midden burials, which included 25 individuals, were found within shell midden deposits (Severs 1974). They probably date to between 4900 B.P. and 4000 B.P. Hearths, pits, living floors and postholes were also identified but no house structures were defined (Acheson 1998:27). The only excavated House is Richardson Ranch site, which dates to the early historic period (Fladmark 1973). These data provide only the briefest glimpse into the antiquity of Houses and domestication on the Queen Charlotte Islands. Much more information is available for the west coast of Vancouver Island. The earliest evidence for domesticated societies comes from the period 4000 B.P.-2500 B.P. It includes the remains of a substantial structure from Shoemaker Bay I and stone or whalebone cairn burials from Shoemaker Bay I, Ch’uumat’a and Little Beach (McMillan 1999:118). Whalebone cairn burials have been also been found at Quattishe in Quatsino Sound (Maas et. al. 2000). The oldest evidence for Houses comes from the village of Yuquot where 18 semi-rectangular, rock rimmed firepits dated to 2300-1900 B.P. were identified (Dewhirst 1980:50). They formed two clusters of superimposed firepits indicating the presence of large permanent house structures with curated placements for firepits. The dry midden excavations at Ozette identified four sequential house floors, stratigraphically separated by sandy 43

Yvonne Marshall

Fig. 1. Map showing key sites named in the text and the Nootka Sound and southern Queen Charlotte Islands survey areas.

44

Houses And Domestication On The Northwest Coast break occurs around 3,000 m2. Among sites under 3,000 m2 in area, very few fall into the upper part of the range - most are extremely small (Fig. 2). Overall, breaks and modal points in the distribution define three size classes: Large sites of more than 3,000 m2; Medium to Small sites measuring 3,000 m2 - 200 m2 in area; Very Small sites of less the 200 m2 in area. Two strong patterns emerge from the comparison. Firstly, the expected tendency for site size to be multi-modal is confirmed. Interestingly, the size classes for each region are remarkably similar as breaks and modal points in the distribution of site sizes occur at similar points in both the Nootka Sound and the southern Queen Charlotte Islands. This parallel patterning raises the possibility that thresholds in the growth of settlements and optimum settlement sizes may show consistencies across the whole of the Northwest Coast. The second pattern concerns the frequency with which sites fall into each size class. This pattern is totally different between the two regions, as highlighted by comparing the percentage of sites in each size class (Table 1). For Nootka Sound the graph approaches a normal distribution, but for the southern Queen Charlotte Islands it is heavily skewed toward the small end (Fig. 3). The key differences are the absence of Very Large sites, the rare occurrence of Medium size sites, and the vast numbers of Very Small sites in the southern Queen Charlotte Islands compared to Nootka Sound. In summary, although the same thresholds in settlement size are common to the two regions, processes directing the number of settlements that grow to a particular size appear to be quite different.

features can be compiled for analysis. Site Size Archaeological surveys in Nootka Sound, recorded 92 habitation sites (Marshall 1992, 1993). This survey built on the earlier work of Folan (1972). Information on size and surface features could not be obtained for 17 sites and they are excluded from analysis. For the remaining 75 sites, area is approximated for two sites and 16 sites had surface modifications that destroyed some features (Marshall 1993). Site area is a calculation of the extent of archaeological deposits and is not simply the product of length and width. Those parts of a site that had surface evidence of historic structures but no subsurface cultural deposits are excluded from the area calculations. Figure 2 shows the distribution of site areas in Nootka Sound. There are three striking patterns. Firstly, there is an enormous size range from a tiny rock shelter of less than 10 m2 to the huge settlement of Yuquot at 24,000 m2. Secondly, while small sites are both small and numerous, the largest sites are very large and very rare. Thirdly, site size distribution reveals two major gaps in the distribution creating three groups. The first group consists of Yuquot which is twice the size of the next largest site and thus exceptional in size by any standards. The second group includes the 12 sites that exceed 4,000 m2 in area. Within this group there is some clustering at 6,000 to 7,000 m2. The third group consists of the 62 sites that range in size from 10 to 3,000 m2. Within this third group there are three modal points that break the range into three subgroups: 2,200 m2, 1,200 m2, and 200-400 m2 (Fig. 2). In total, the identified gaps and modal points define five size classes: Very Large sites which include only Yuquot; Large sites measuring 3,000-13,000 m2; Medium sites measuring 1,000-3,000 m2; Small sites measuring 1,000-200 m2; and Very Small sites of less than 200 m2. In the southern Queen Charlotte Islands 99 habitation sites were recorded in an archaeological survey conducted by Acheson (1998). To make this data directly comparable with the Nootka Sound data, those parts of the two largest sites which had surface evidence for historic occupation but no subsurface cultural deposits were excluded from the site area calculations. The absolute size range in the southern Queen Charlotte Islands is considerably less than in Nootka Sound with the largest site measuring 8,350 m2. Sites over 3,000 m2 in area are also less numerous. As at Nootka Sound a clear

Surface Features Examination of the surface features associated with each site size class in the two regions reveals further differences. In Nootka Sound, four categories of surface features were identified: terraces, back midden ridges, house depressions, and small ridges (Figs. 4 and 5). All these features are argued to have formed in association with structures. The more complex and spatially coherent the surface features, the more formal and permanent the structures with which they were associated (Marshall 1993). Terraces are highly variable in length but are remarkably consistent in width, varying only from 10 m to 15 m, suggesting they were formed to accommodate structures of fairly standard width. Back midden ridges mark the rear extent of the sites on which they occur. They ran along 45

Yvonne Marshall

Fig. 2. Distribution of site sizes by area in Nootka Sound and the Southern Queen Charlotte Islands: distribution of all site areas (top); frequency of site areas in 1000 m2 size ranges; distribution of site areas under 3000 m2; sites under 3000 m2 by 200 m2 size ranges. 46

Houses And Domestication On The Northwest Coast

Fig. 3. Percentage of sites in each size class for Nootka Sound and the Southern Queen Charlotte Islands. 47

Yvonne Marshall LEGEND MN

Large Tree

Cliff, Bluff, Steep Slope Top of Slope

x

0

10m

DjSp24

Stump

450 m2

Bottom of Slope

Vegetation Edge

Low Scarp

Grass, Marsh

House Depression/ Ridge Feature

Rock

Back Midden Ridge Feature

Pebble Beach

Exposed Midden

Stream

Midden Exposed in Tree Throw

High Tide

Edge of Cultural Deposit

Water, Sea

Rock

Path

0

MN

DjSm4

10m

Rock

Rock

Approximate Edge of Cultural Deposit MN

1,250 m2

0

10m

DkSo2

Rock Outcrop

450 m2

Edge of Cultural Material

Rock Rock Outcrop Outcrop

Toilet

MN

Area of Timbers

10m

0

DjSo3

850 m2

MN

0

10m

DjSo4 2,150 m2

Fig. 4. Site maps of three Small sites (DjSp24, DkSo2, DjSo3) and two Medium sites (DjSm4, DjSo4) from Nootka Sound showing the variety and complexity of the surface features identified. 48

Houses And Domestication On The Northwest Coast

Salal and Dense Undergrowth

Beach Flat

MN

0

20m

DjSq1

12,500 m2

MN

0

Rock Outcrop

20m

DkSp17

Approximate Edge of Cultural Deposit

9,250 m2

Possible Toilet

Rock Outcrop Rock Outcrop Rock Outcrop

Rock Outcrop

Rock Outcrop

Beach Flat

MN

20m

0

DjSq2

6,000 m2

Fig. 5. Site Maps of three Large sites from Nootka Sound showing the variety and complexity of surface features. See Fig. 4 for legend. 49

Yvonne Marshall the back of the houses and accumulated from rubbish disposed out of the backs of dwellings. House depressions consist of low ridges, which define the perimeters of areas (depressions) resulting from substantial, permanent dwellings (Marshall 1993, 2000a:79). These ridges may have been deliberately constructed around the base of dwellings or formed in the course of occupation from discarded boiling stones and hearth rocks. The small ridges are more variable in composition, and consist of earth, stone and sometimes shell. Some may be remnant fragments of old house depressions, and others may have formed in association with smaller, less permanent structures. On all sites in Nootka Sound, the coherence of surface features is remarkable. Even on small sites with only one or two terraces or small ridges, the features are well defined and spatially coherent (Fig. 4, see site maps in Marshall 1992; 1993). On larger sites surface features increase in number and complexity, and it is apparent that each feature is in part defined by, and accommodated to, neighboring features. This coherent spatial patterning corroborates and strengthens the excavation data indicating the curation and long term maintenance of the locations of all kinds of structures, particularly those associated with house depressions.

Relationships between surface features and site size are summarized in Table 2. Terraces and small ridges are the most commonly recorded surface features and they are found on sites of all size classes, except Very Small (Figs. 4 and 5). On Medium and Large sites terraces are found alongside other kinds of features. Small ridges are particularly numerous on Large sites where they are found in association with well-defined house depressions. Ten sites have house depressions. These house depression features are usually grouped and they occur most frequently on Large sites. House depressions and small ridges are shown in historic photographs of Yuquot, but all these features are now destroyed. Six Medium and Large sites have back midden ridges. No category of surface feature is exclusive to any site size class, but surface features increase in both frequency and complexity as site size increases. The wide range of surface features found on sites in Nootka Sound produces a gradient in the degree of formality identified in the spatial structuring of settlements. Very Small sites are distinguished by the virtual absence of surface features. Small sites commonly have terraces and occasionally have small ridges. Medium sites also have terraces and small ridges, but also an occasional back midden ridge or house depression is present. Over one-third

Table 1. Habitation Site Size Classes in Nootka Sound and the Southern Queen Charlotte Islands Nootka Sound (N=75)

Number

%

Number

%

< 200

12

16

58

59

201-1,000

21

28

21

21

1,001 - 3,000

29

39

12

12

3,001 - 15,000

12

16

8

8

15,000 +

1

1

--

--

Size Class (m2) Very small Small Medium Large

Very Large

S. Queen Charlotte Islands (N=99)

50

Houses And Domestication On The Northwest Coast have more than one kind of feature. It is only on Large sites that back midden ridges and house depressions become common, and multiple types of surface feature are the norm. The house depressions are commonly arranged in orderly rows, but some intercutting or disrupted features are present on all sites suggesting that although structures are permanent there remains some degree of flexibility in their dimensions or composition. At Yuquot, the single Very Large site, only terrace features are present today but a range of features including house depressions were recorded historically (Marshall 2000a, 2000b). Wider analysis of settlement patterns in other parts of the West Coast of Vancouver Island has established that the patterns identified in Nootka Sound are generally characteristic of the wider region (Marshall 1993). Using a combination of site size and surface features a four-tier settlement pattern is suggested for Nootka Sound: 1. Very rare, Very Large, highly structured settlements making up around 1-2% of settlements. 2. A limited number of Large, formally structured settlements, with permanent structures of flexible dimensions, making up approximately 20% of settlements;

3. Numerous, less formally structured, Medium to Small settlements, with permanent or semi-permanent structures of flexible dimensions, making up approximately 60-70% of settlements; 4. A limited number of Very Small, informal camps, with no substantial structures, making up approximately 20% of settlements. In the southern Queen Charlotte Islands only one category of surface feature was identified: house depressions (Fig. 6). There is no gradient of surface features - only the presence or absence of house depressions. House depressions are found on almost all Medium and Large settlements, but very rarely on Small sites and never on Very Small sites (Table 3). They are very formally arranged and there is no evidence of the intercutting or disruption of one feature by another, as found in Nootka Sound. This means that sites are either very formally structured, or have no evident formal spatial structure. Due to lack of data it is not yet possible to assess whether the patterns identified in the southern Queens Charlotte Islands are characteristic of the Queen Charlotte Islands generally. For the southern Queen Charlotte Islands a two-tier settlement pattern is suggested: 1. A limited number of Medium to Large, very

Table 2. Frequency of surface features by site size class for Nootka Sound % with two + kinds of features

Size Class

Total Sites

T

SR

HD

BMR

% with features

Very small

12

2

--

--

--

17

0

Small

21

12

2

--

--

62

5

Medium

28 (1)

19

12

5

2

93

36

Large

9 (3)

8

8

5

4

100

100

1

1

yes

yes

?

100

100

71 (4)

42

23

11

6

72

28

Very Large Total

Notes: T = terrace, SR = small ridge, HD = house depression, BMR = back midden ridge. Four sites listed in parenthesis are excluded from the % calculations due to destruction of their surface features.

51

Yvonne Marshall

Contour Interval - 10 cm

House 1 480

510

450

470 460

43 0 0 44 0 45 0 46

House 2

House 3

350

50 0

460 490 470

480

House 4

460 450 440 430

0

10m

16 15

Cobble/gravel Beach

14 13

12

11

10

9

8

7

6 4

3

1

2

Key: House timbers Fire altered rock Excavation unit Tree

0

50m

Fig. 6. Site Maps of FaTt 9 (above) and FbTs 4 (below), Large sites in the southern Queen Charlotte Islands. Redrawn from Acheson 1998:154 [Map 9], 166 [Map10].

52

Houses And Domestication On The Northwest Coast tributaries of the Skeena River. At these sites very large plank houses with fixed dimensions measuring about 10 m by 15 m were also constructed. At the Pascelay Site, GbTH-4, there were two of these houses on a site with an area of about 2,000 m2. In the classification established above such sites would be considered Medium size with highly structured surface features. This pattern is different from Nootka Sound and southern Queen Charlotte Islands, but shares elements in common with both areas. Most late prehistoric Tsimshian settlements consisted of formally structured Medium and Large sites. As in Nootka Sound, the core Tsimshian settlement was Medium sized. However, Tsimshian sites were also like those of the southern Queen Charlotte Islands, in that surface features on sites consist primarily of house depressions. Unstructured camps, or resource extraction sites, which are a major component of settlement in the southern Queen Charlotte Islands and a small but nevertheless significant component of Nootka Sound settlement, only appeared among the Tsimshian in the post-contact period (Martindale this volume).

formally structured, settlements, with permanent dwellings of fixed proportions, making up approximately 20% of settlements. 2. Very numerous, Small to Very Small, informal camps, with no substantial structures, making up approximately 80% of settlements. Comparison with Late Prehistoric Settlement Patterns Among the Tsimshian Martindale (this volume) has examined the way Tsimshian settlement has changed from the pre-contact to post-contact periods. For the late prehistoric period, approximately 500 B.P.-A.D.1787, he suggests a two-tier pattern of settlements. Local group, winter villages located on the coast. These were few in number and consisted of 10 to 20 very large houses of fixed dimensions. In the above classification they would be considered Large, possibly even Very Large, very formally structured, settlements. More numerous lineage-based summer sites located on low elevation riverside terraces along

Table 3. Correlation of Surface Features in Nootka Sound and the Southern Queen Charlotte Islands.

Size Class

Very small

Small

Medium

Large

Very large

Length (m)

Area (m2)

Nootka Sound Features

Queen Charlotte Islands Features

Usually none, commonly rock shelters, caves.

Mostly rock shelters, caves.

Terracing, occasional small ridges.

Very occasional house depressions

1,001 - 3,000

Terracing, small ridges, occasional midden ridge, or house depression.

House depressions

201 - 350

3,001 - 15,000

House depressions, back midden ridges, small ridges, terraces.

House depressions.

350 +

15,000 +

Same as for Large.

No sites.

< 21

< 201

21 - 60

201 - 1,000

61 - 200

53

Yvonne Marshall Coast prehistory. These theories shift our attention away from economics and technology to place, and change our starting assumptions about the nature of continuity and change in social life. The argument is not that we should abandon economic and technological approaches and concentrate solely on analyses based in place. Rather, I have tried to show in this paper that approaches based around concepts of place and space produce different insights. Economic and technological explanations have always dominated Northwest Coast archaeology and this domination has produced an unnecessarily narrow conversation about how we understand the past of this region. Analyses located in place are a way of broadening the conversation and of focusing attention on the people rather than the objects they produced. Such analyses also allow us to make productive use of a range of spatial data that is not easily incorporated into prehistories based on changing economies and artifact types. Informed by the concepts of Houses and domesticated societies, I briefly reviewed the prehistory of the Northwest Coast focusing on evidence for houses and burials. The review raised the possibility that all people living on the Northwest Coast, even the earliest, were in some ways domesticated. People did not always live in the open, unbounded settlements typical of most hunter-gatherers. Instead, they often chose to remain at selected settlements for extended periods, to build substantial dwellings and to bury their dead in cemeteries. These are all domesticated behaviors indicative of people establishing and maintaining ongoing ties to specific places. This view of Northwest Coast prehistory suggests the highly domesticated, hunter-gatherer societies of the Late Pacific and Historic Periods were not brought into existence by sudden changes in climate, rising sea levels or the invention of new technologies. Rather they developed from the incipient seeds of social domestication brought to the Northwest Coast by the people who came and settled there. In the final section of the paper I used some of the vast, but largely neglected, body of spatial data collected during site surveys to examine the internal structuring of settlements and wider settlement patterns in several parts of the Northwest Coast with a view to establishing the antiquity of House based sociopolitical organization. Although regional differences within the Northwest Coast were very apparent, results from all areas examined highlighted the pervasiveness of houses as a key element structuring both settlements and settlement pattern by at least the Late Pacific

Summary Comparisons of late prehistoric settlement structure in three regions of the Northwest Coast have identified a number of interesting patterns. First, the site size classes established from the multi-modal distribution of site size in Nootka Sound reappeared in the southern Queen Charlotte Islands data, and may also be a feature of Tsimshian site size distribution. The persistence of these site size classes suggests the existence of Northwest Coast wide thresholds that constrain settlement growth to a limited number of optimum sizes. The identified site classes may therefore provide a basic model for wider comparisons of settlement patterns across both time and space throughout the Northwest Coast. Second, there is wide variation across the Northwest Coast in the number of sites that fall into each size class. This suggests that while there may be a limited number of optimum settlement sizes, factors governing the number of sites that grow to any particular size are very different in different parts of the Northwest Coast. Third, there is considerable variation in the range of surface features identified. Based on the evidence analyzed here, northern sites demonstrate a limited range of surface features so that variation is largely limited to the presence or absence of intact, formally laid out house depressions. In contrast, a wide range of surface features is found on sites to the south and these features commonly intercut each other. Fourth, there are consistent correlations between site size and number and complexity of surface features in that larger sites have larger numbers of surface features that are also more formally arranged. However, in the north formal surface patterning is only found on Large and some Medium sized sites whereas in the south even Small sites commonly show coherent patterning of surface features. The implication is that in the north, when spatial organization of settlements is present, it is very formal and this occurs on only a limited number of sites. In the south, the spatial ordering of settlements is more pervasive and is found on almost all sites, even small ones, but it is generally less formal. Conclusion When brought together, Lévi-Strauss’s concept of House societies and Peter Wilson’s concept of human self-domestication through the creation of a built environment offer Northwest Coast archaeologists a new way of thinking about Northwest 54

Houses And Domestication On The Northwest Coast Fedje, D. W., A. P. Mackie, J. B. McSporran and B. Wilson 1996 Early Period Archaeology in Gwaii Haanas: Results from the 1993 Field Program. In Early Human Occupation in British Columbia, edited by Roy L.Carlson and L.Dalla Bona, pp.133-150. UBC Press, Vancouver. Fladmark, K. R. 1973 The Richardson Ranch Site: A 19th century Haida house. In Historical Archaeology in Northwestern North America, edited by R.M. Getty and Knut R. Fladmark, pp.5395. Universisty of Calgary Archaeological Association, Calgary. Folan, William J. 1972 The Community, Settlement and Subsistence Patterns of the Nootka Sound Area: A Diachronic Model. Unpublished Ph.D. dissertation, Southern Illinois University, Carbondale. Gillespie, S. D. 2000a Beyond Kinship: An Introduction. In Beyond Kinship, edited by R. A. Joyce and Susan D. Gillespie, pp.1-21. University of Pennsylvania Press, Philadelphia. 2000b Lévi-Strauss: Maison and Société à Maisons. In Beyond Kinship, edited by Rosemary A. Joyce and Susan D. Gillespie, pp.22-52. University of Pennsylvania Press, Philadelphia. Hayden, B., G. A. Reinhardt, D. Holmberg and D. Crellin 1996 Space per Capita and the Optimal Size of Housepits. In People Who Lived in Big Houses: Archaeological Perspectives on Large Domestic Structures, edited by Gary Coupland and E.B.Banning, pp.151164. Monographs in World Prehistory 27, Prehistory Press, Madison. Jonaitis, A. 1999 The Yuquot Whalers’ Shrine. University of Washington Press, Seattle. Joyce, R. A. and S. D. Gillespie (editors) 2000 Beyond Kinship. University of Pennsylvania Press, Philadelphia. Lévi-Strauss, C. 1982 The Way of the Masks. Translated from the French by Sylvia Modelski. University of Washington Press, Seattle. Maas, A., R. Cahill and A. Williams 2000 Archaeological Excavation and Inventory of Quattishe EDSV-4: Results of the 1999 Field Season. Unpublished Preliminary

Period, suggesting houses, if not Houses, were by this time of central importance to the construction of sociopolitical groups. References Cited Ames, K. M. 1994 The Northwest Coast: Complex HunterGatherers, Ecology, and Social Evolution. Annual Reviews of Anthropology 23:209229. Ames, K.. M. and H. D. G. Maschner 1999 Peoples of the Northwest Coast. Thames and Hudson, London. Acheson, S. 1998 In the Wake of the Ya’aats’xaatgaay [`Iron People’]: A Study of Changing Settlement Strategies among the Kunghit Haida. BAR International Series 711, Oxford. Boas, F. 1897 The Social Organization and the Secret Societies of the Kwakiutl Indians. Report of the U.S. National Museum for 1895:311738. Washington, D.C. 1920 The Social Organization of the Kwakiutl. American Anthropologist 22:111-126. 1921 Ethnology of the Kwakiutl. Bureau of American Ethnology Thirty-fifth Annual Report, Parts 1 and 2. Cannon, A. 1996 The Early Namu Archaeofauna. In Early Human Occupation in British Columbia, edited by Roy L. Carlson and L. Dalla Bona, pp.83-102. UBC Press, Vancouver. Carlson, R. L. and L. Dalla Bona (editors) 1996 Early Human Occupation in British Columbia. UBC Press, Vancouver. Carsten, J. and S. Hugh-Jones (editors) 1995 About the House - Lévi-Strauss and Beyond. Cambridge University Press, Cambridge. Coupland, G., and E. B. Banning (editors) 1996 People Who Lived in Big Houses: Archaeological Perspectives on Large Domestic Structures. Monographs in World Archaeology No 27. Prehistory Press, Madison. Dewhirst, J. 1980 The Indigenous Archaeology of Yuquot, a Nootkan Outside Village. In The Yuquot Project, Volume 1, edited by William J. Folan and John Dewhirst. History and Archaeology 39. Parks Canada, Ottawa. 55

Yvonne Marshall Report to the Quatsino First Nation, Coal Harbour, British Columbia and the BC Heritage Trust Foundation, Victoria. Marshall, Y. M. 1989 The House in Northwest Coast, NuuChah-Nulth, Society: The Material Structure of Political Action. In Households and Communities, edited by Scott MacEachern, David J.W. Archer and Richard D. Garvin, pp.15-21. Proceedings of the 21st Annual Conference of the Archaeological Association of Calgary, University of Calgary Archaeological Association, Calgary. 1992 Mowachaht/Muchalaht Archaeology Project, Final Report. Unpublished report prepared for the British Columbia Heritage Trust, Mowachaht/Muchalat Band and the British Columbia Archaeology Branch, Victoria. 1993 A Political History of the Nuu-ChahNulth People: A Case Study of the Mowachaht and Muchalaht Tribes. PhD Dissertation, Department of Archaeology, Simon Fraser University, Burnaby. 2000a Transformations of Nuu-chah-nulth Houses. In Beyond Kinship, edited by Rosemary A. Joyce and Susan D.Gillespie, pp.73-102. University of Pennsylvania Press, Philadelphia. 2000b The Changing Art and Architecture of Potlatch Houses at Yuquot. In Nuuchah-nulth Voices, Histories, Objects and Journeys, edited by Alan L.Hoover, pp.107-130. Royal British Columbia Museum, Victoria. Matson, R. G. and G. Coupland 1995 The Prehistory of the Northwest Coast. Academic Press, London. Mackie, Q. 1998 The Archaeology of Fjordland Archipelagos: Mobility Networks, Social Practice and the Built Environment. Unpublished

56

Ph.D. dissertation, University of Southampton. McKenzie K. H. 1974 Ozette Prehistory – Prelude. Unpublished M.A. Thesis, Department of Archaeology, University of Calgary. McMillian, A. D. 1999 Since the Time of the Transformers: The Ancient Heritage of the Nuu-chah-nulth, Ditidaht and Makah. UBC Press, Vancouver. Samuels, S. R. 1991 Patterns in Ozette Floor Middens: Reflections of Social Units. In Ozette Archaeological Project Research Reports, Volume 1: House Structure and Floor Midden, edited by Stephan R. Samuels, pp. 175270. Reports of Investigations No. 63. Department of Anthropology, Washington State University, Pullman. Samuels, S. R. (editor) 1994 Ozette Archaeological Project Research Reports, Vol 2: Fauna. Reports of Investigations 66. Pullman: Department of Anthropology, Washington State University; Seattle: National Park Service. Sendey, J. 1977 The Nootkan Indian - A Pictorial. Alberni Valley Museum, Port Alberni, British Columbia. Severs, P. 1974 Archaeological Investigations at Blue Jackets Creek, FlUa4, Queen Charlotte Islands, British Columbia, 1974. Canadian Archaeological Association Bulletin 6:163-205. Suttles, W. (editor) 1990 Handbook of the North American Indians Volume 7: Northwest Coast. Smithsonian Institution, Washington Wilson, P. J. 1988 The Domestication of the Human Species. Yale University Press, New Haven.

From Architects to Ancestors: The Life Cycle of Plank Houses D. Ann Trieu Gahr

For more than 4,000 years peoples of the Northwest Coast built and lived in wood plank houses (Ames and Maschner 1999). In the oral histories and myths all beings—salmon, raven, the moon, and people—dwelt in houses.

the structure that, over time, would have demanded maintenance. The cumulative picture gained from examining the life cycle of a plank house—from construction, through maintenance, and finally demise—is that these structures entailed substantial economic investments. The question I consider here is how did such substantial investments of materials and labor in house structures impact social organization? Ethnographic and historic evidence demonstrate a strong link between the costs of house ownership and social organization. This link was most directly described by Oberg who asked “Why should a group like the Tlingit house-group be a house-owning body? He answered his question by stating that

Houses were the physical manifestation of the household and its social rank; they were theater and stage for social and spiritual rituals, but they were also shelter in the Northwest’s dank climate; they were food-processing factories,…and they were objects of enormous effort and great skill. (Ames and Maschner 1999:14)

The centrality of houses in native Northwest Coast societies underlies the research of all the papers in this volume and all of them use archaeological evidence to show how these households functioned in the past. In this paper I focus on houses as “objects of enormous effort and great skill” rather than the activities that took place within house walls. Ames et al. (1992) undertook some of the first investigations to examine the economics of building and maintaining a plank house based on archaeological evidence. They estimated that a single house at the Meier site (35C05) may have used between half a million to a million board feet of lumber to construct and maintain the structure over its 400 year history. This point is instructive when one considers that the modern single-family house requires 10,000 to 12,000 board feet of lumber (Ames et al. 1992). Ames (1996) later expanded on this research and applied similar estimates of materials to a number of archaeologically recovered rectangular house remains, presumably plank houses, from throughout the Northwest Coast. In this paper, I pick up the thread of inquiry raised in these papers and look more closely at factors affecting the economics of plank house construction and maintenance over time, that is, over the life cycle of a plank house. I use forestry and wood technology research to provide more detail about construction costs in terms of materials and labor and also to expand our understanding of threats to

[t]here is a strong connection with the difficulties involved in making a house out of wood. The decoration of the posts, plates, and moldings also required much time and skill. Thus the heavy costs involved in the making of a house made it more economical for a group of close relatives, like brothers and their sister’s sons, to band together to accumulate enough wealth to build a house. (Oberg 1973:62)

These house-owning groups characterized native Northwest Coast social organization and long eluded precise anthropological definition as they did not follow standard kinship rules (Lévi-Strauss 1982). Rather than kinship, residence in the house was the pivot of social organization (Marshall 2000). Lévi-Strauss (1982) identified this type of social organization as “société á maison” or “house society.” Hereafter, I refer to the House in the LéviStraussian sense with a capital “H” and refer to the house dwelling using a lower case “h.” The organizing principal of House societies is that the House groups share a “commitment to a corpus of [H]ouse property” that is the foundation of unity and the source of continuity as groups “socially reproduce themselves by the actions involved with the preservation of their joint property” (Gillespie 2000:3). On the Northwest Coast the corpus of House prop57

D. Ann Trieu Gahr that might dilute the cohesion of the House group. Social hierarchies functioned to transmit property that was not easily divisible such as houses, titles, names, and songs. In Northwest Coast societies transmissions of property were accompanied by ceremonies that historicized the transfers of names, dances, songs, and oral histories. Individuals whose identities were intertwined with property and place were assigned roles as stewards of social histories, traditions, and knowledge, and they embodied the ideals of their society. These were the roles of House chiefs. To ensure that House chiefs fulfilled their duties to the House’s estate of tangible and intangible property, these stewards were given privileged access to ancestors and history or other modes of social memory. That these privileges were used to benefit House leaders in the context of day-to-day interactions is without doubt. However, in this paper, I bypass the rights and privileges of elite status that have been the focus of much anthropological attention, and instead examine the material obligations involved in owning a plank house on the Northwest Coast in the era before mechanized lumber production and woodworking technology. Archaeological investigations have provided physical evidence of house construction, repairs, and histories of long and short term house tenure. These histories of house architecture provide a unique perspective rarely ascertained in ethnographies or in the narrow time frames of Northwest Coast written histories. The material remains of house structures can yield valuable insights into the cohesiveness of a social group through that group’s labor and material investments in a single object. Some archaeological examples reveal remarkable efforts to preserve continuity over prolonged time periods, on the order of generations and occasionally centuries (Ames et al. 1992; Connolly 1992; Grier this volume; Marshall 2000 this volume; Smith this volume). These examples are testament to resilience in the face of historical change and of shared commitment through time to a common purpose (i.e., a structure that embodies group identity). Other archaeological examples provide evidence of abandonment of places and houses (e.g., Ellis this volume), perhaps a record of demise reflecting the failure of members to sustain obligations to the House group. Individual members may have continued to survive and thrive by joining and taking on commitments in other houses, but the unique entity of that particular house and House group died. The remainder of this paper presents a closer look at the life cycle of plank houses—from con-

erty included house structures and decoration; and ceremonial furnishings such as costumes, masks, and feasting dishes; intangible property such as house crests, names, histories, songs and dances; and land, such as hunting territories, sealing rocks, bird rookeries, shellfish beds, sections of rivers, trade route passes, and other productive resource areas (e.g., Ames and Maschner 1999; Drucker 1951; Mitchell and Donald 1988; Richardson 1982). This property was held in common with respect to House members. Drucker described property rights as follows: The basic concept was that all the members of the unit shared in the joint right to these prerogatives, as they are often termed, but that the chief of the lineage was the custodian of both the intangible rights and of the lands and material possessions. The lineage chief was in this respect similar to the executor, in our own culture, of a large estate who manages its various enterprises for the heirs. (1951:118)

Drucker’s “unit” is a corporate body synonymous with Lévi-Strauss’s House. Those who held heritable rights in the House estate were called “elites,” “chiefs,” “house masters” as they are called among the Tlingit (Duff 1973), or “housekeepers” as referred to among the Gitksan (Adams 1973). Other House members who had few heritable rights were referred to as commoners. Throughout the life cycle of a plank house the House entity, or corporate identity, was created and renewed through the actions of co-residents. Construction of the house, the major property holding of a House group, involved corporate action that concomitantly produced the entity of the House and materially manifested the identity of the House group. Maintenance of the house structure and events such as the winter ceremonies where House histories or “narratives of kinship and descent” (Marshall 2000) took center stage served to renew corporate or House identity. H[h]ouse structures had life expectancies exceeding those of their architects, builders, and original occupants and thus depended on institutions that could field sufficient labor and materials to maintain and reconstruct the house over time. That is, survival of the house structure depended on commitments in perpetuity. The most efficient and effective way to ensure this commitment was through organized, hierarchical, transmission of property rights and obligations. The House entity melded social, political, and economic interests of members, whose efforts were continually directed toward preserving their joint associations and excluding associations 58

From Architects to Ancestors: The Life Cycle of Plank Houses struction through maintenance and finally demise, showing that house survival depended on commitments of labor and materials in perpetuity. I propose that these commitments in perpetuity are key to understanding the hierarchical social structure of native Northwest Coast societies.

this volume). In addition to floor treatment, some houses had excavated floors or “house pits.” Among the Tsimshian, Haida and Tlingit, elaborate houses had multiple level, terrace-like floors as described by Coupland (this volume) and in Chief Wí:ha’s house described below and shown in Figure 2. Floor features and other elements of basic house design are often revealed in archaeological excavations of house remains: for example, by rectangular house features, post molds and sometimes plank molds. Beyond the basic house design there was considerable variation in house size and architectural detail. For example, a large plank house on the lower Fraser River was described by Simon Fraser in 1798 as measuring almost 200 m long by 18 m wide (Ames and Maschner 1999), but other houses, such as house remains found at Broken Tops (Ellis this volume) and the seven plank house remains at the Clahclellah site (Sobel this volume) averaged around 8 m by 10 m. Unique architectural features such as carved posts and painted screens distinguished different houses. Some houses were named after these special architectural features. House form also varied seasonally. While some groups owned more than one substantial house frame, allowing them to move residences seasonally by transporting planks between the frames, others used light weight, temporary mat lodges during warmer parts of the year when people were more mobile. In this paper, I am concerned with the more substantial permanent houses generally occupied during the winter months, sometimes referred to as “winter houses,” as these houses represented major economic investments.

Life Cycle of Plank Houses The life cycle of the house structure provides a framework to examine the economic obligations entailed in owning a plank house. In the construction phase section I will look at procurement, choice of building materials, and labor. Under the maintenance section I will evaluate different risks facing these structures and the impact of those risks on maintenance requirements. And, in the demise section, the phase that concludes the life cycle, I consider different ways that houses fail. Before launching into the particulars of the house life cycle, I will first briefly describe native Northwest Coast house plank house architecture. Architecture of Plank Houses Northwest Coast plank houses were wood post-and-beam frame structures built on a square or rectangular floor plan with split wood planks for walls and roof. Suttles (1990) classified five architectural house types based on roof support systems and geographic variation in architecture. Roof support systems, wall systems, and floor plans varied geographically although there was stylistic overlap between areas (Fig. 1). Gable roof systems were most common throughout the region, except for the Coast Salish area where the single slope shed roof prevailed. In other areas, shed roofs were used for lighter-weight temporary summer houses, but not for permanent winter houses. Wall systems varied in the orientation of planks (Fig. 1). In many areas, vertical planks were set directly on the ground or in shallow wall trenches (e.g., Smith this volume), however on the northern Northwest Coast, horizontal base and header boards held vertical planks in place. In the central area where shed roof houses were prevalent wall planks were typically oriented horizontally. Floor design also varied, although this variation was less correlated with geography. Throughout the region packed dirt floors were most common, though there are scattered reports of clay floors (Byram 2003 a, 2003b; Coupland this volume; Newman 1959; Smith this volume). Some houses had wood plank floors covering a corridor of large pits (e.g., Smith

Construction Phase Throughout the Northwest Coast historic and ethnographic literature there are accounts of the high cost and extended lengths of time required for house construction. Preparation for building occupied the greatest length of time while houseraising, the most visible of house building activities, occupied a much shorter period of time. Preparation for house building involved procuring construction materials and fabricating structural elements such as posts, beams, and planks. In addition, house builders had to accumulate supplies of feasting food and gifts to be given to people from other houses who assisted in construction and those who witnessed at the house raising ceremony. Boas reported that some Kwakwaka’wakw house groups worked for over a three-year period to accumulate sufficient supplies for house building (cited in 59

Figure 1. Variation in native Northwest Coast house architecture styles.

D. Ann Trieu Gahr

60

Figure 2. Neí:wons (Monster House) estimated measurements derived from Blackman 1972.

From Architects to Ancestors: The Life Cycle of Plank Houses

61

D. Ann Trieu Gahr Nabokov and Easton 1989). Albert Niblack, on the Alaska Survey from 1885 to 1887, described the construction of a house by a Tlingit group at Fort Simpson in Tsimshian territory over the course of several years.

sufficient to warrant a new house, the yitsati [house master] came out through the ceremonial opening in the screen before the back wall and announced to the gathered clansmen that a new house was to be built. Next the Wolf people, or members of the opposite phratry, were informed to hold themselves in readiness for the coming task. The Frog house people then began to accumulate food supplies; more salmon, eulachon oil, and berries were stored than usual; blankets and money were accumulated by trade so that a year later the Wolf people began work. (Oberg 1973:81)

The cutting and roughing out of the timbers in the forest, the launching of these and towing to the village, the carving of the totemic column and supports for the huge beams or rafters, and their final erection into the framework of a house, all required not only the expenditure of much time and labor, but a very extensive outlay of wealth. (Niblack 1970[1887]:306)

In another description of house building, Swan observed that

The protracted time periods required for procuring construction materials and surplus food and gifts to reward labor contrasted markedly with the relatively short period of time for house-raising, even though house-raising events were usually accompanied by feasting, speeches, and gift giving. While considerable efforts were directed toward accumulating surpluses to be used for feasting, in this paper I will concentrate on the building materials and labor tied with house construction.

[t]he self-denial of comforts and even necessaries exercised for many years in the accumulation of property by a man and wife is very remarkable, but in their estimation is amply repaid on the occasion of the distribution of the same and the erection of a decorative column.… (Swan, Publications, Manuscripts, Notes, Letters, and Collection in National Museum, Port Townsend, Washington Territory, n.d., cited in Niblack 1970[1887]:308)

Building Materials

The importance of initial preparations preceding the erection of the building is illustrated in Oberg’s account of the circumstances of the seventh Frog house of the Ganaxtedi clan of Klukwan Tlingit. In about 1908,

In the abundantly forested Northwest Coast, building materials were surprisingly costly. Western red cedar was the wood of choice for building plank houses. It was also favored for canoes, bentwood boxes, carved bowls and figurines as well as many other necessities of daily life. In the northern reaches of the region beyond the range of western red cedar other wood resources were also used for house construction, especially Sitka spruce and western hemlock (Emmons and de Laguna 1991). (Table 1 presents the vernacular and botanical

[w]hen the old house threatened to become inhabitable, the old men of the house began to talk about building a new one. The matter was later fully discussed, not only by the people of the Frog house but by the whole clan division of the village. When it was finally agreed that the resources of the Frog house people were

Table 1. List of native taxa referred to in the text. Scientific Nomenclature

Common Name

Cupressaceae Callitropsis nootkatensis, formerly Chamaecyparis nootkatensis Chamaecyparis lawsoniana Thuja plicata

Cypress Family Alaska cedar, yellow cedar

Pinaceae Picea sitchensis Pseudotsuga menziesii Tsuga heterophylla

Pine Family Sitka spruce Douglas-fir, Oregon pine Western hemlock

Port Orford cedar Western red cedar

62

From Architects to Ancestors: The Life Cycle of Plank Houses western red cedar is variable with some wood being prone to extreme drying defects. This condition usually occurs in wood originating from very wet heartwood (Panshin and de Zeeuw 1980:340). Volumetric shrinkage is a quantitative measure of wood drying behavior based on the ratio of shrinkage in wood from a green state to an oven-dry condition (12% moisture content). The greater the percentage of volumetric shrinkage, the more likely a wood is to suffer from drying-related defects. While wood from the cedar family has less volumetric shrinkage than that from the pine family, within the cedar family western red cedar experiences a markedly lower percentage of volumetric shrinkage than other species (Table 2). Even careful selection of wood species for specific woodworking applications does not eliminate problems associated with drying; consequently, Northwest Coast woodworkers developed a variety of techniques for handling wood defects such as straightening warped planks (Stewart 1984). A third ideal attribute of western red cedar described by William Reid is that the wood is “soft, but of a wonderful firmness.” The Forest Products Laboratory, generally concerned with timber industry needs, does not report specifically on wood qualities prized by wood workers using hand tools. However, two comparative analogues, specific gravity and hardness of wood, may be used to discern ease of working wood with hand tools. Specific gravity, a measure of wood density, provides an estimate of the ease of working wood with hand tools. Generally, the lower the specific gravity of a wood, the more easily it can be cut (USDA, FPL-FS 1987:3-18). Both western red cedar and Sitka spruce are noted for having soft wood, but the specific gravity of western red cedar ranks significantly lower than Sitka spruce (Table 2). The second comparative analog for ease of working is the hardness attribute. Hardness is usually regarded as a measure of the suitability of a wood species for flooring or other applications involving impact. The hardness test is conducted by driving a 0.444 inch ball into the wood to a depth of one-half of the ball’s diameter (Panshin and deZeeuw 1980). As in other tests, western red cedar is markedly softer than Sitka spruce, the other native wood species noted for ease of working (Table 2). Reid also commented on the light weight of the wood. Western red cedar is the lightest weight of native conifer species (Table 2). Overall, western red cedar scores the highest in qualities valued by wood workers – ease of working with hand tools and few drying defects. Additionally, it is one of the most durable

nomenclature for the various wood taxa referred to in this paper). All these timber species possess wood properties suited for planking and carving. However, properties inherent in western red cedar wood made it preferred over all other wood. Haida carver William Reid described the woodworking qualities of western red cedar in these words: The wood is soft, but of a wonderful firmness and, in a good tree, so straight-grained it will split true and clean into forty foot planks, four inches thick and three feet wide with scarcely a knot. Across the grain it cuts clean and precise. It is light in weight and beautiful in color, reddish brown when new, silvery grey when old. It is permeated with natural oils that make it one of the longest lasting of all woods, even in the damp of the Northwest Coast climate.… (Reid 1984:8-9)

These favorable woodworking qualities identified by Reid are confirmed in quantitative tests of wood mechanical properties. These tests, conducted by the Wood Products Laboratory of the U.S. Department of Agriculture, involve a series of trials applying different forces or stresses to clear-grained wood to determine variance in behavior between species. The stress tests produce changes or distortions in the shape and size of the wood to the point where distortion becomes unrecoverable or to the point of failure (United States Department of Agriculture, Forest Products Laboratory, Forest Service [USDA, FPL-FS] 1987). One of the most desirable mechanical properties for Northwest Coast woodworking applications, especially planking, is the splitting characteristic of wood. All conifer timber species used by native Northwest Coast woodworkers split under low loads. However, western red cedar and Port Orford cedar require less than 150 lbs per inch [1034 kpa per 2.54 cm] of width to split green wood. The other selected species require at least 150 to 250 lbs per inch [1034 to 1723 kpa per 2.54 cm] of width to split green wood (Table 2). While this difference in ratings may seem negligible today with the common use of metal or power tools, the differences in force required to split wood are more significant when using hand tools, especially those without metal cutting edges. Hence, it is no surprise that the splitting characteristic of western red cedar was highly valued, especially for plank manufacture. A second important wood property is its vulnerability to drying defects. In general, woods from the cedar family are less prone to drying defects than those from the pine family (Table 2). However, 63

Cedar Family

64

6.80% 12.40% 11.10%

1,5 2 2 2

Western red cedar

Coastal Douglas-fir

Western hemlock

Sitka spruce

11.50%

10.10%

1

Port Orford cedar

9.20%

1

Volumetric Shrinkage

Alaska cedar

Taxa

Drying Defects

150-250

150-250

150-250