Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Prehistoric Northwest Coast art : a stylistic analysis of the archaeological record Holm, Margaret Ann 1990

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

Item Metadata

Download

Media
831-UBC_1990_A8 H64.pdf [ 39.12MB ]
Metadata
JSON: 831-1.0098516.json
JSON-LD: 831-1.0098516-ld.json
RDF/XML (Pretty): 831-1.0098516-rdf.xml
RDF/JSON: 831-1.0098516-rdf.json
Turtle: 831-1.0098516-turtle.txt
N-Triples: 831-1.0098516-rdf-ntriples.txt
Original Record: 831-1.0098516-source.json
Full Text
831-1.0098516-fulltext.txt
Citation
831-1.0098516.ris

Full Text

PREfflSTORIC NORTHWEST COAST ART: A STYLISTIC ANALYSIS OF THE ARCHAEOLOGICAL RECORD by MARGARET ANN HOLM B.A., University of British Columbia, 1980  A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in THE FACULTY OF GRADUATE STUDIES DEPARTMENT OF ANTHROPOLOGY AND SOCIOLOGY  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA April 1990 (c) Margaret Ann Holm, 1990.  5  ?  In  presenting  degree at the  this  thesis in  University of  partial  fulfilment  of  of  department  this thesis for or  by  his  or  requirements  British Columbia, I agree that the  freely available for reference and study. I further copying  the  representatives.  an advanced  Library shall make  it  agree that permission for extensive  scholarly purposes may be her  for  It  is  granted  by the  understood  that  head of copying  my or  publication of this thesis for financial gain shall not be allowed without my written permission.  Department The University of British Columbia Vancouver, Canada  Date  DE-6 (2/88)  frkrWaA'VlO  ii  ABSTRACT  This thesis is a stylistic study of the prehistoric art record from the Northwest Coast of North America. Its purpose is three-fold: to describe the spatial and temporal variation in the stylistic attributes of prehistoric art; to evaluate theories on the evolution of the Northwest Coast art tradition; and to comment on the possible factors behind variation in the prehistoric art record. This study examines stylistic attributes related to representational imagery, concentrating on five variables: decorated forms, carving techniques, design elements, design principles, and motifs. The core sample consists of anthropomorphic and zoomorphic images from dated archaeological contexts; a total of 242 artifacts from 58 sites are examined. The material is presented in chronological order corresponding to the Gulf of Georgia prehistoric cultural sequence. The major finding of this study is that by the end of the Locarno Beach phase or the beginning of the Marpole phase the essential character of the Northwest Coast art style had developed. There are new developments in the late period, but the evidence presented suggests a previously undocumented stylistic continuity from the late Locarno Beach phase to historic Coast Salish art with no decline in quality or productivity. This study indicates that, as far back as the record extends, three-dimensional, naturalistic forms and two-dimensional incising and engraving techniques have equal antiquity. From the Locarno Beach phase onward the flat, engraved style and the three-dimensional sculpture style developed together; the formline concept developed very early out of the raised, positive lines created by deep engraving in antler.  iii  TABLE OF CONTENTS Abstract Table of Contents List of Tables List of Figures  n  iii . . . v vi  1. Introduction  1  2. Studies of Style in Anthropology  5  Defining regional styles and centers of origin The relationship between style and social interaction Stylistic motifs as indicators of social organization and value systems Style variables as indicators of ethnographic cultural patterns Developing a methodology for the analysis of style in prehistoric Northwest Coast art  3. Theories of the Development of Northwest Coast Art Early theorists Recent Archaeological Findings Conclusion  4. Research Methods Research objectives The data sample A definition of terms used in this study  5. An analysis of 4500 years of prehistoric art. 5.1 Art from 4500-3300 B.P  6 9 13 15 20  23 23 31 42  45 45 45 51  .57 57  iv  St. Mungo phase sites in the Fraser Delta Gulf Island Sites . Central and north coast sites Summary 5.2 Art from 3300 to 2400 B.P. Mid-Thompson River region Lower Fraser Region Gulf of Georgia Locarno Beach phase sites Strait of Juan de Fuca North Coast sites Summary 5.3 Art from 2400 to 1500 B.P Marpole phase archaeological sites Puget Sound Sites Lower Fraser Region West Coast of Vancouver Island North Coast Sites Summary 5.4 Art From 1500 to 150 B.P Gulf of Georgia Region Antler Figurines and Whale Bone Clubs Olympic Peninsula Central and North Coast sites The Mid-Fraser and Thompson River Region 6. Conclusion  58 63 69 72 74 75 77 82 97 101 101 107 107 158 173 181 183 188 201 203 229 247 258 270 304  6.1 Summary  304  6.2 Theoretical implications  320  Continuity The relationship between north and south coast art The relationship between the interior and the south coast A Marpole phase artistic climax ? 6.3 Interpretations  320 321 324 326 327  Ethnographic analogy  327  Style as information exchange  331  6.4 Implications for Northwest Coast archaeology  334  Bibliography  336  Appendix I: artifacts in the sample  344  LIST OF TABLES  Table I. Prehistoric Style Attributes 4500-3300 B.P Table II. Prehistoric Style Attributes 3300-2400 B.P Table III. Prehistoric Style Attributes 2400-1500 B.P Table IV. Late Period Human Figurines Table V. Prehistoric Style Attributes 1500-100 B.P Table VI. Stylistic Developments in Prehistoric Northwest Coast Art.  .104  vi  LIST OF FIGURES  Figure 1. Map of sites included in sample  47  Figure 2. Northwest Coast Design Elements  53  Figure 3. Prehistoric Carving Techniques  54  Figure 4. Northwest Coast Design Principles  56  Figure 5. St. Mungo zoomorphic stone fragment  59  Figure 6. Glenrose antler human figurine  61  Figure 7. Pender Canal antler zoomorphic spoon  64  Figure 8. Pender Canal antler owl figure  65  Figure 9. Helen Point antler human figurine  68  Figure 10. St. Mungo antler anthropomorphic carving  70  Figure 11. Central and north coast art  71  Figure 12. Kamloops Indian R. zoomorphic stone maul  76  Figure 13. Milliken stone zoomorphic figures  78  Figure 14. Milliken steatite zoomorphic sculpture  .80  Figure 15. Milliken stone head effigies  81  Figure 16. Musqueam Northeast block engraved design  84  Figure 17. Musqueam Northeast feather motif.  85  Figure 18. Musqueam N.E. anthropomorphic antler spoon  86  Figure 19. Musqueam Northeast zoomorphic antler spoon  87  Figure 20. Undated zoomorphic spoon, Vancouver Island  90  Figure 21. Locarno Beach carvings  91  Figure 22. Locarno Beach carvings  93  Figure 23. Crescent Beach antler bird head carving  94  Figure 24. Northwest Coast mask showing rope edging  96  vii  Figure 25. Hoko River wooden mat creaser  98  Figure 26. Hoko River zoomorphic implements  99  Figure 27. Prince Rupert Harbour stone effigies  102  Figure 28. Marpole zoomorphic antler harpoon points  Ill  Figure 29. Marpole pendants with human face motifs  113  Figure 30. Marpole human and bird motif carvings  114  Figure 31. Marpole bird form carvings  115  Figure 32. Marpole bird form carvings  116  Figure 33. Marpole zoomorphic carvings  118  Figure 34. Marpole antler lightning-snake effigy.  119  Figure 35. Marpole antler browband fragments  121  Figure 36. Marpole block engravings with formlines  122  Figure 37. Marpole steatite seated human-figure bowl  125  Figure 38. Marpole seated human-figure bowls  126  Figure 39. Marpole stone bowls with human heads  128  Figure 40. Marpole "type A" stone zoomorphic bowls  129  Figure 41. Marpole "type B" stone zoomorphic bowls  130  Figure 42. Marpole "type B" stone zoomorphic bowls  131  Figure 43. Marpole stone atlatl weights  133  Figure 44. Marpole miscellaneous decorated objects  135  Figure 45. Beach Grove bone and antler artifacts  138  Figure 46. Beach Grove undated stone human head bowl  139  Figure 47. Whalen Farm stone bowls  142  Figure 48. Whalen Farm antler zoomorphic artifacts  143  Figure 49. Whalen Farm anthropomorphic antler chisel  145  Figure 50. Whalen Farm undated antler artifacts  147  Figure 51. Cattle Point decorated artifacts  149  viii  Figure 52. Gulf of Georgia antler zoomorphic head  151  Figure 53. Garrison antler pendant/browbands  152  Figure 54. Garrison decorated artifacts  154  Figure 55. Undated zoomorphic antler harpoon points  155  Figure 56. False Narrows human face motif pendants  158  Figure 57. False Narrows stone effigies  159  Figure 58. Zoomorphic effigies  161  Figure 59. Musqueam North antler pestle/pendant  163  Figure 60. Musqueam North zoomorphic antler carving  164  Figure 61. Musqueam N.E. antler zoomorphic fragments  166  Figure 62. Musqueam N.E. antler human face pendant  167  Figure 63. Point Grey "type A" zoomorphic stone bowl  169  Figure 64. Port Hammond decorated artifacts  171  Figure 65. Fishtown wooden atlatl (Skagit atlatl)  174  Figure 66. Biederbost wooden bowl  177  Figure 67. Esilao wooden block engraved fragments  179  Figure 68. Lehman zoomorphic maul  180  Figure 69. Prehistoric whale bone clubs  182  Figure 70. Prince Rupert zoomorphic effigies  185  Figure 71. Lachane wooden handle with eye motif  187  Figure 72. Prince Rupert decorated artifacts  189  Figure 73. "Sechelt Image" human figure sculpture  194  Figure 74. Lytton area steatite human figure bowls  195  Figure 75. Maple Bank antler carvings  205  Figure 76. Stselax antler carvings. .  J207  Figure 77. Stselax antler carvings  209  Figure 78. Stselax block engraved design  211  Figure 79. Stselax whale bone club fragment  213  Figure 80. Wooden club from the Fraser River  214  Figure 81. Helen Point antler carvings  216  Figure 82. Gulf Island decorated objects  218  Figure 83. English Camp antler digging stick handle.  220  Figure 84. English Camp antler carvings  221  Figure 85. Victoria area antler carvings  224  Figure 86. Pedder Bay decorated artifacts  225  Figure 87. Undated stone effigies  227  Figure 88. Antler human figurines  232  Figure 89. Victoria area antler human figurines  234  Figure 90. Antler human figurines  235  Figure 91. Antler human figurines  236  Figure 92. Columbia River antler human figurines  238  Figure 93. Columbia River antler human carvings  240  Figure 94. Undated human figurines  241  Figure 95. Whale bone clubs  245  Figure 96. Ozette wooden artifacts  250  Figure 97. Ozette carvings  252  Figure 98. Ozette carving styles  253  Figure 99. Ozette wooden wall screen  257  Figure 100. Kwatna antler carvings  259  Figure 101. Kwatna carvings  261  Figure 102. Garden Island antler comb  263  Figure 103. Negative C carving  -  265  Figure 104. Prince Rupert Harbour carvings  266  Figure 105. Prince Rupert comb with split-U form  268  Figure 106. Prince Rupert bone comb  269  Figure 107. Milliken-Esilao stone carvings  271  Figure 108. Milliken steatite spindle whorl  273  Figure 109. Mid-Fraser River carvings  275  Figure 110. Bell antler bird form comb  276  Figure 111. Bell antler snake-female effigy  279  Figure 112. Mid-Fraser/Thompson zoomorphic carvings  281  Figure 113. Bell anthropomorphic carvings  283  Figure 114. Bell antler anthropomorphic carvings  284  Figure 115. Bell stone zoomorphic maul fragment  286  Figure 116. Seton Lake steatite bowl  288  Figure 117. Mid-Fraser River undated steatite carvings  291  Figure 118. Kamloops area seated human figure bowl  292  Figure 119. Chase Burial wooden mask  295  Figure 120. Chase Burial anthropomorphic carvings  296  Figure 121. Chase Burial antler clubs  297  Figure 122. Chase Burial carvings  299  Figure 123. Chase Burial zoomorphic steatite carvings  300  Figure 124. Chase Burial steatite carvings  301  Figure 125. Coast Salish spindle whorl  323  1  CHAPTER 1. INTRODUCTION  This thesis is a stylistic study of the prehistoric art record from the Northwest 1  Coast of North America. Previous studies have dealt with decorated artifacts from regional prehistoric artifact collections (Borden 1983, Carlson 1983, MacDonald 1983, Stryd 1983) or particular prehistoric art forms, such as stone sculpture (Duff 1956a, 1975, Wingert 1952) and rock art (Lundy 1983, Hill 1974). To date, however, there has been no systematic study of prehistoric art of the Northwest Coast as a whole. The purpose of this study is three-fold.  First and foremost is the description of  spatial and temporal variation in the stylistic attributes of prehistoric art. The model of stylistic development presented in this study can be integrated into existing models of prehistoric cultural development on the coast to increase our understanding of the decorated forms that characterize each cultural phase. Secondly, the results of this study are used to evaluate theories on the evolution of the Northwest Coast art tradition and its regional variants. Thirdly, the study ventures into more speculative territory to comment on the possible factors behind variation in the prehistoric art record. An additional result of this study is to shed light on the hundreds of undated prehistoric Northwest Coast carvings in museum collections throughout the world. A cursory examination of prehistoric art in North America reveals that simple decorative designs are widely distributed over time and space. Representational art is much  1. Art is defined following Boas (1927:10): When the technical treatment has attained a certain standard of excellence, when the control of the processes involved is such that certain typical forms are produced, we call the process art, and however simple the forms may be, they may be judged from the point of view of formal perfection...  2  less common, and the high degree of development of representational motifs distinguishes the Northwest Coast cultural area from other regions of North America.  This study  therefore, concentrates on stylistic attributes related to representational imagery. Excluding nonrepresentational designs from this study does not imply that this type of decoration performs a less important role within a cultural system, but only that it is difficult to evaluate and does not appear to be as sensitive to change as representational art. Since geometric motifs are often associated with representational images this study occasionally comments on the meaning and use of non-representational designs. The core sample for this study consists of anthropomorphic and zoomorphic images from dated archaeological contexts. A total of 242 artifacts from 58 sites are examined in this thesis. The study area, shown in Figure 1, encompasses Northwest Coast archaeological sites from Prince Rupert Harbour in the north to the Lower Columbia River in the south, and inland sites along the Fraser River to Lillooet, and along the Thompson River to Chase. The study includes material from archaeological sites within the Plateau cultural region linked to the representational art tradition of the coast. For historical reasons the Lower Mainland and southern Vancouver Island regions have seen the greatest archaeological activity. This partially accounts for the fact that 155 artifacts constituting 64 % of the sample, come from sites in this region. Fortunately, this region constitutes a distinct historic and prehistoric cultural area.  Mitchell (1971) has  termed this area with its unique prehistoric cultural sequence the "Gulf of Georgia region," and the term is used throughout this study. Although Mitchell uses the same term to characterize the historic aboriginal population of this area, regional anthropologists use the term the "Central Coast Salish" (Suttles 1983, Kew 1981) to describe this cultural unit. The Central Coast Salish are made up of people speaking the Halkomelem, Straits, Squamish, and Nooksack languages of the Coast Salish language family. Because the bulk of the sample is from sites in the Gulf of Georgia area, the model of stylistic development presented in the concluding chapter is specific to this area. Although  3  the sample of prehistoric art from other regions of the coast is small, it does provide an important comparative sample that allows hypotheses on the development of prehistoric art in the western plateau, central coast, and northern coast regions.  It is hoped that as  archaeological work increases the sample of prehistoric art from these peripheral areas, the findings will be tested against the model of stylistic development proposed for the Gulf of Georgia region. The second chapter of this thesis examines the concept of style in anthropology and evaluates different theoretical perspectives that shape the methodologies of stylistic studies. The third chapter is a historical review of literature pertaining to theories on the origins of Northwest Coast art and its regional styles. Both of these chapters are critical reviews of both theories and methodologies put forward by other authors. The third chapter ends with a hypothesis presented by the author: that the prehistoric art record in the Gulf of Georgia area will show a gradual evolution of the decorated forms, design elements, motifs, and carving styles of Central Coast Salish art. After Chapter 4 outlining the methods and terms used in this study, there follows the stylistic analysis of the prehistoric art sample in Chapter 5. The material is presented in chronological order corresponding to the Gulf of Georgia prehistoric cultural sequence which is as follows: the St. Mungo and Mayne phases, from 4300 to 3300 B.P.; the Locarno Beach phase, from 3300 to 2400 B.P.; the Marpole phase, from 2400 to 1500 B.P.; and the late period phases (Stselax, Gulf of Georgia, San Juan, and Developed Coast Salish) from 1500 to 150 B.P. Ethnographic information and examples of art from historic Northwest Coast groups-in particular the Central Coast Salish~are integrated into the stylistic analysis. The stylistic analysis concentrates on five variables: decorated forms (i.e. spoons, pendants), carving techniques, principles of design elements, and motifs. At the end of each of the four temporal sections the findings are reviewed and presented in a table.  The  concluding chapter presents a summary of the stylistic attributes that characterize the four  4  prehistoric cultural periods and integrates this model into a discussion of the development of the Northwest Coast art tradition and the significance behind stylistic variation.  5  CHAPTER 2. STUDIES OF STYLE IN ANTHROPOLOGY  Sackett states succinctly the problem of defining the concept of style in his comment that in anthropology, "style," like the words "structure," and "culture," are "crashing ambiguities of the sort which seem to nourish our thoughts however abstract they may be (1977:369)." The following chapter is an attempt to understand this ambiguous concept, to examine the different ways in which style has been used as a theoretical tool in material culture studies, and to consider which approaches might be used in examining prehistoric art styles of the Northwest Coast. One of the simplest and clearest definitions of style is paraphrased from Sackett as follows: Style concerns a highly specific and characteristic manner of doing something that is peculiar to a specific time and place. Style and function together determine the nature of variability among artifacts (1977:370). Of interest to anthropologists is the "how" and "why" of style: how styles are created and maintained over time and space, and why styles are important to human societies. In archaeology, stylistic studies have concentrated largely on the "how" of style. In places like the American Southwest sophisticated ceramic analyses have been use to attempt to define the relationship between motif distribution and the variables of human social organization. Although anthropology and fine arts have their share of similar studies on motif distribution, other researchers have felt freer to investigate the "why" of style. Out of the literature involving stylistic studies, four different perspectives emerge that have relevance to the present study of Northwest Coast prehistoric art. The first approach is the definition of regional styles and their distinctive traits.  The next approach is often the  second stage of theory development after local styles have been defined, that is to analyze how the spatial distribution of stylistic attributes can indicate patterns of human interaction  6  and social organization. A third approach proposes that stylistic attributes themselves are linked to social organization and value systems, and that art motifs can act as indicators of particular types of social organization. Lastly, since archaeologists are concerned with the evolution of contemporary societies, many studies examine the relationship between decorative styles and ethnicity to evaluate if style can indicate the antiquity of historic cultural practices. A brief overview of these approaches is presented with an analysis of the assumptions each concept works within. The summary evaluates which perspectives might be useful in this regional examination of prehistoric art.  DEFINING REGIONAL STYLES AND CENTERS OF ORIGIN  In its simplest definition "style" refers to an artifacts attributes, outside of its function and physical composition, that denote the preferences of a particular society. These preferences are associated with specific cultures and the time periods they occupy, such as the "Olmec style" of prehistoric Mexico. Historically, many of the first studies by regional ethnologists involved the definition of stylistic regions and their relationships to one another. Solidly researched regional stylistic studies are still valuable contributions to anthropology, furthermore, the material culture of many societies in North America still remain largely undescribed. Ethnologists such as Boas (1927) and Drucker (1955) were among the first to attempt to characterize and understand the origins of Northwest Coast art. (Their ideas are presented in the following chapter). It was not until Bill Holm's (1965) stylistic study of northern Northwest Coast art that a vocabulary and conceptual framework was developed with which to discuss this complex artistic tradition. Holm's definitions of carving styles and design elements have been adopted by most researchers in describing historic and prehistoric art of the coast, although sometimes design elements specific to the northern area are used  7  to describe art from other regions of the coast. Current researchers need to develop terms and concepts to describe these regional variants. As archaeologists and ethnologists describe regional styles, proposals are made about the origin of certain innovations and their diffusion to neighbouring areas. Several studies of Northwest Coast art have focused on the quality and quantity of artistic production in developing hypotheses about the origin, spread and decline of prehistoric art styles. Wingert's (1949) study of Coast Salish sculpture concluded that the essential stylistic character of Coast Salish art and culture developed up-river and inland from the coast. He based this notion on his subjective evaluation that the art styles of inland Salish groups showed greater "boldness and surety of their approach to problems of design and expression (1949:119)." From this stylistic homeland the style spread to other areas where, in Wingert's view, decorated objects did not exhibit the same degree of understanding of stylistic principles. Wingert's stylistic assessments were undoubtedly influenced by the dominant theory of the period that proposed an early coastal migration of Salish people from the interior plateau. Borden also used subjective assessments of artistic quality and productivity in developing his theory that proposed a prehistoric artistic climax in the Marpole period 2400 to 1500 years ago, followed by an artistic decline in the late prehistoric and historic Coast Salish period. More recent studies have shown how difficult it is to make assessments about artistic quality and quantity.  In an examination of productivity among the Coast Salish,  Suttles (1983) hypothesizes that qualitative and quantitative differences between the art of the Salish and of other coastal Northwest Coast groups relate to differences in ritual practices. Among the Salish, the explicit depiction of spirit beings is discouraged, whereas among the Kwakwaka'wakw, masks and dances relating to the spiritual realm are displayed in a highly visible arena involving social position and ceremonial prerogatives.  Suttles  concludes that qualitative and quantitative changes in the prehistoric art record may have  8  more to do with changes in local ritual and social patterns than with the strength, decline or migration of local populations. Other studies have suggested that two or more art styles with distinct functions and aesthetic qualities can exist side-by-side in a community. Stott (1975) has noted different styles and functions of kusiut society and sisaok ceremonial masks among the Nuxalk. Kusiut masks were hastily made and destroyed after their ceremonies, whereas sisaok masks were made with greater care and refinement and were reused during annual ritual occasions.  Jonaitis (1984) has examined the difference between secular and sacred art  among the Tlingit. She proposes that shamanic charms intentionally utilized a crude style as part of the expression of a powerful connection with the spirit world. This style existed in conscious opposition to secular art which emphasized grace and fine execution. These studies by Suttles, Stott and Jonaitis indicate that caution should be used in assessing artistic quality and productivity. Regional studies of art styles are often descriptive and do not explicitly pursue a particular theoretical viewpoint. But anthropologists often uncritically accept the dominant theoretical approaches of the period. Early studies of Northwest Coast art took place while cultural trait distribution and diffusion studies were popular in anthropology.  Research methods and assumptions implicit in the work of early  ethnographers such as Boas and Kroeber remain with us today. In his study of Northwest Coast stone sculpture, Duff (1956) based his conclusions on the origin and development of the seated human-figure bowl on the assumption that stylistic traits with the widest distribution were older than those with localized distributions, and that localities that exhibited the most variation in a particular motif were likely the centres of origin of that motif. As will be discussed in the next section, more recent studies of style present other explanations for the spatial patterning of stylistic attributes.  9  THE RELATIONSHIP BETWEEN STYLE AND SOCIAL INTERACTION  Many early stylistic studies treated diffusion as a natural process whereby innovations spread to surrounding areas. Davis' (1983) review of diffusion research points out that diffusion is a process of donation and adoption, not an answer to explain stylistic patterning. This study and others by Hodder (1979), Conkey (1980), Plog (1983), and Wobst (1977) explore the cultural variables that might affect the distribution of stylistic traits. Their conclusions address factors behind the maintenance of stylistic boundaries, stylistic homogeneity and heterogeneity within communities, and the role of material culture in making statements about ethnic and class affiliation. Working with eleven prehistoric Magdalenian sites in Spain, Conkey (1980) uses stylistic motifs to support speculations that one of the sites was an aggregation area for the local population. She predicted that decorative motifs on bone and antler artifacts at the aggregation site of Altamira would show the greatest stylistic diversity, having examples of stylistic motifs from each of the ten other sites. The other ten sites were predicted to exhibit a greater homogeneity of carving motifs with far fewer motifs per site. Conkey confirmed her hypothesis that aggregation sites exhibit motif heterogeneity while isolated sites exhibit homogeneity. Conkey's approach is appealing but it can only be used if the age and length of occupation of a site is known. As Davis (1983) has pointed out, assemblage heterogeneity can be caused by gradual style changes over time as well as the mixing of styles from contemporaneous groups. Most sites that are occupied for a short length of time exhibit stylistic homogeneity while sites that are occupied for a longer period will show greater stylistic heterogeneity.  The stylistic patterning that Conkey observed in her eleven sites  could also have been the result of a longer occupation period for Altamira. Her study is only applicable when there are a number of sites dated to the same period and occupied for a similar length of time. Conkey's hypothesis that the aggregation site would show greater  10  heterogeneity came out of previously developed ideas about the site and was an attempt to use art styles to support existing theories on intersite relationships. It is generally accepted that similarity in the material cultures of two groups indicates interaction between the two populations and that the frequency of shared stylistic traits is a good indication of the amount of interchange. Hodder (1979) has challenged this notion that the degree of interaction is a sufficient explanation for stylistic similarity. His study of style in decorative and functional objects in western Kenyan tribal groups has shown that stylistic similarity does not necessarily fall off with distance. Hodder found a high degree of stylistic homogeneity within tribal territories, even though there was daily intertribal contact and intertribal marriages. In the border regions between tribal territories the study found both sharply defined style boundaries and heterogeneous areas of stylistic mixing.  Hodder concluded that economic factors were important variables in predicting  areas where styles had well-defined borders. Sharply defined stylistic boundaries occurred in areas where there was fierce intertribal competition over scarce resources. On the other hand, in areas where intertribal economic competition was low there was a gradation of stylistic attributes between the two tribal centres.  Hodder hypothesized that stylistic  differences were more important in maintaining tribal identity in areas of high competition and tension. Hodder has used the results of his ethnographic studies to interpret the spatial patterning of styles in the prehistoric period. He cautions that his assumptions regarding the relationship of stylistic boundaries to economic competition can only be used when the prehistoric record indicates that other variables such as population size, social organization, and economic strategies match those of his ethnographic study area. For example, in a study of Zambian tribal fiefdoms controlled by a ruling class totally different variables accounted for the distribution of decorative styles. Hodder states that his aim is not to develop a theory to explain the material culture patterns of groups with different socio-political structures, rather that styles play a role in defining group identity and that stylistic homogeneity can be  11  a sign of group interdependence and intragroup competition. He believes that this approach allows a better understanding of the human behavior behind changing styles, whereas traditional studies use the amount of shared stylistic traits as passive manifestations of ethnicity or residence patterns. Hodder's study focuses on the variables that can account for the spatial distribution of styles, but his theory leaves aside the question of how style can work in promoting group interdependence and identity in conditions of resource competition. Wobst's (1977) detailed study of style in Yugoslavian ethnic dress is an attempt to explain the adaptive advantage of stylistic variability.  Wobst views stylistic behavior as a strategy of information exchange  where distinct classes of objects are chosen to broadcast the information.  His study  concentrated on isolating the range of messages communicated in ethnic dress styles and assessing the target audiences for the stylistic messages. Dress styles were observed to transmit information about class, ethnicity, religion, and wealth. Wobst theorized that the messages in dress styles helped define expected behavior patterns, making verbal interaction more predictable and less stressful, and allowing less investment in exploratory or competitive interaction. Wobst predicted that the amount of stylistic variability in individual dress would positively correlate with the size of the social network the individual participated in. His observations indicated that clothing that could be seen from the greatest distance, such as headgear and coats, showed the greatest stylistic variability, and that headgear showed the most variation in areas were ethnic groups lived interspersed with one another. Wobst's study confirmed that amongst individuals with small social networks, such as women, dress styles contained fewer messages about ethnic and political affiliation. His observations also confirmed Hodder's findings that stylistic variation was much decreased in areas where a ruling authority (in this case the state government) controlled social interaction. Wobst concluded:  12  ...style is a pleasantly multidimensional and surprisingly dynamic phenomenon. It reacts with great sensitivity to changes in other cultural variables and, of itself, actively supports other cultural processes such as cultural integration and differentiation, boundary maintenance, compliance with norms and enforcing conformity (1977:335). Wobst's theory and methodology would appear to be testable with a variety of material culture studies. No doubt the classes of objects that would be stylistically sensitive to social variables would be different amongst different cultures. In the Northwest Coast canoes, houseposts and house architecture would seem to be highly visible communicators of ethnic identity, whereas head deformation, dress, and ornamentation would more likely address class affiliation and wealth. Wobst's analysis takes place within the realm of living societies where the richness and nuances of all classes of material objects are available to the investigator. As fascinating as his theory is, there would be little data within an archaeological site with which to test his ideas. We must turn to archaeology for a more realistic assessment of the types of research possible in finding the relationship between human behavior and stylistic variation in the prehistoric record. Plog (1983) evaluated the adequacy of four different hypotheses on stylistic variation to explain changes over time in the distribution of ceramic designs in the American Southwest. The inferences tested were that changes in stylistic boundaries and the amount of intersite and intrasite motif homogeneity would relate to one of the following factors: changes in subsistence and settlement patterns, changes in vessel shape and use, ceramic exchange between communities, and changes due to the length of occupation of a site. Plog asserted that Wobst's theory of the role of material culture in information exchange best accounted for the spatial patterning of Southwestern prehistoric ceramics. As population increased and contact between social groups increased between 400 A.D. and 900 A.D., the size of style zones decreased, and homogeneity within the boundaries increased. Plog argues that ceramic designs, which were highly visible due to their use in exchange networks, after 900 A.D. became increasingly important vehicles for communicating group identity.  13  Like Plog, Levi-Strauss proposes that stylistic traits should not be viewed in isolation but are possible expressions of cultural identity. In The Way of the Masks (1982) Levi-Strauss discusses the occurrence of mask and myth complexes among Northwest Coast groups and hypothesizes that as masks spread between neighbouring groups key stylistic elements are intentionally transposed: Contemporary styles do not ignore one another...The originality of each style, therefore, does not preclude borrowing: it stems from a conscious or unconscious wish to declare itself different, to choose from among all the possibilities some that the art of neighbouring people has rejected (1982:144). .Wobst. Plog, Hodder and Levi-Strauss see stylistic traits as involved in the maintenance of cultural identity not only as passive reflections of ethnic preferences but as active messages to neighbouring groups.  STYLISTIC MOTIFS AS INDICATORS OF SOCIAL ORGANIZATION AND VALUE SYSTEMS.  The appeal of the approaches discussed in the previous section lies in many of the initial ideas being generated by ethnographic studies of stylistic behavior, and then being evaluated with data from prehistoric sites that exhibit similar economic, demographic and social variables.  The studies discussed below attempt to find ethnographic support for  previously developed ideas on the link between design motifs and psychological states internalized by individuals living under different forms of social organization. Each works with the assumption that motif shapes and design compositions have latent meanings addressing universal human concerns. Fischer (1961) tested the notion that art styles reflect general attitudes which are characteristic of societies at different levels of social complexity. For example, egalitarian  14  societies would exhibit simple, symmetrical design elements surrounded by empty space, while complex societies with class differentiation would exhibit unlike design elements integrated into tight patterns and enclosed by borders. A second set of hypotheses tested psycho-sexual explanations of motif shapes: i.e. straight lines thought to represent masculine qualities were predicted to be associated with patrilocal communities. Twenty-nine groups from Murdock's (1967) "World Ethnographic Sample" were used to test twenty hypotheses that linked motif styles with variables related to social organization. Fischer asserts that the five hypotheses that were supported at statistically significant levels lend support to his theory. Several other studies have proposed similar correlations between social organization and art styles. Dressier and Robbins (1975) have found that the complexity of designs in ancient Greek vase paintings increased during periods of greater social stratification. A study by Wolfe (1967) used statistical tests to support the hypothesis that high artistic development was associated with matrilineal societies. Lastly, Adams (1973) related textile designs styles with social organization in East Sumba, Indonesia. Leaving aside questions about the statistical methods used and the possibility of geographical sample bias-criticisms which have been leveled at Fischer and Wolfe (Dittmer 1969)~the main weakness of these studies is insufficient explanation of how or why motif shapes are linked to human behavior or thought patterns.  Wolfe may be correct in his  assessment that matrilineal societies foster greater achievements in the plastic arts, but his statistical tests do not validate his hypothesis that art develops in societies were men feel socially divided and need to express their isolation. Similarly, Fischer's study does not present a convincing case of his assertion that an egalitarian community's shared sense of geographic isolation and interdependence within the village results in simple-shaped decorative motifs. More serious flaws in these studies are: firstly, the highly subjective evaluation of such variables as "well-developed" versus "poorly-developed" art styles, and "complex" versus "simple" designs; and secondly, the investigators' assumptions about human psychological responses to cultural variables. Fischer's study used the evaluations of  15  a single psychologist to assess the latent meanings behind motif shapes. One also suspects that among societies with several art styles in different media only one of the styles would be highlighted while the others would be ignored or downplayed. The Northwest Coast, for example, has complex curvilinear two-dimensional designs and simple, interlocking geometric basketry motifs, among other styles. Finally, there are no evaluations of the conclusions and test-implications using detailed ethnographies or living societies. Many of these authors are searching for alternate explanations of stylistic behavior in opposition to diffusionist explanations of the distribution of stylistic traits. Levi-Strauss (1963) began his study on the split-representation style of art as a polemic against historicaldiffusionist studies of material culture.  He investigated possible similarities between a  number of prehistoric and contemporary societies sharing the split-representation style of design organization.  Levi-Strauss found several aspects of social structure and ritual  practice shared between the six groups exhibiting the split-representation style of art. LeviStrauss proposed internal psychological connections between these cultures that accounted for the similar stylistic and social variables and predicted that any society exhibiting the split-representation style would be found to share these same cultural elements. As in the previous studies he equated a demonstration of the correlation of style and social structure with an affirmation of his psychological interpretations of the meaning of that style. As well, several of the stylistic and social correspondences presented were weak and were not shared by all six societies in the study.  STYLE VARIABLES AS INDICATORS OF ETHNOGRAPHIC CULTURAL PATTERNS  As archaeology has changed its focus from identifying temporal and spatial variation in artifact styles to understanding processes of cultural development and change, stylistic studies of the past decade have also paid increasing attention to the same concerns.  16  One of the goals of regional archaeological research is to identity the temporal depth of aboriginal cultural patterns. The studies that follow each use art styles to help define the prehistoric evidence for ethnographic patterns of human behavior. McGhee's (1977) study of the symbolic attributes of materials used by prehistoric Thule craftsmen posits that ethnographic information can support observations about the significance of stylistic variation in the prehistoric record. In examining a Thule assemblage from the eastern Arctic McGhee observed that certain classes of artifacts were made of ivory while others were made of antler.  In testing his observations with material from five  temporally and geographically distinct Thule sites McGhee discovered that ivory and seamammal bone were associated with objects used during the winter period, sea-mammal and bird hunting equipment, women's sewing tools and ornaments, and figurines of a mythological bird-woman.  Since ivory is less common, difficult to carve, and presents no  functional explanation for its selection McGhee postulated that ivory might have symbolic associations with certain activities.  McGhee studied ethnographies from a variety of  Canadian Inuit cultures and discovered that ivory and antler formed two opposing and distinct conceptual categories. Ivory was associated with the sea, sea-mammals, winter, life on the pack ice, and women; while antler was associated with the land, caribou, summer, and men. These conceptual oppositions were evinced in taboos against using ivory and seamammal products, or antler and caribou raw materials for any activities in the opposite category. McGhee suggests that the same conceptual oppositions are in evidence in artifact assemblages belonging to the Thule culture. Since the Thule are the prehistoric ancestors of the contemporary Inuit this study provides strong evidence of similar symbolic concepts influencing behavior and artifact styles one thousand years ago. McGhee, like Fischer, Levi-Strauss, and others cited previously, looked for cognitive concepts shared by individuals within a culture to explain stylistic variation.  But in  McGhee's study the conceptual categories of opposition were first observed in assemblages of objects then compared to information from the living cultures to see if there were  17  correlations.  The symbolic categories are seen as very ethnospecific and the objects  themselves-outside of their Inuit context-are not indicators of any particular psychological state or form of social organization. In the studies discussed previously in this section the stylistic attribute itself is assumed to be related to social organization and diagnostic of particular social and ritual forms. In McGhee's study, the spatial patterning of stylistic variables is seen as a possible source of information about prehistoric culture, but it can only reflect the temporal continuity of ethnographic patterns of thought and behavior. Stryd's (1983) examination of prehistoric art from the British Columbia interior shows a similar methodology to McGhee's study. As well as noting the materials used for making different types of objects, Stryd compared classes of artifacts that were chosen for embellishment, and the quality of design execution in the different classes of decorated objects.  His findings indicated that the quality of design execution showed consistent  variation between distinct classes of objects. Roughly executed incisings were most common on objects of personal ornamentation while steatite objects exhibited the best workmanship. Certain motifs were not only associated with specific objects but the placement of motifs on the artifacts also showed consistency. Stryd searched ethnological literature pertaining to the area for possible information on the stylistic patterning observed on the artifacts. Long parallel lines-like those incised on women's digging-stick handles-were said to represent snake tracks and/or earth trenches.  Descriptions of guardian spirit rituals suggested that these motifs also  symbolized the supernatural power of a guardian spirit aiding the user of the tool. Stryd concludes that similarities between the prehistoric record and the ethnological record allow archaeologists to develop hypotheses on how the ideational realm relates to the economic realm, and if art motifs might indicate the importance of certain prehistoric economic strategies. The strength of the assumptions or hypotheses developed from these types of studies depend upon the quality of the ethnographic information available.  Many site reports  18  contain suggestions on the time depth of aboriginal patterns when there is not the ethnographic or archaeological support to substantiate the claims made. Carlson (1987) has suggested that there is evidence at the Pender Island site of a 3500 year old time depth for shamanism, the memorial potlatch, masks, and other Northwest Coast culture ritual complexes. Carlson presents archaeological evidence at the Pender Canal site suggesting prehistoric ritual feeding of the dead and breaking and scattering of items associated with human burials.  Carlson cites these practices as evidence for memorial potlatches.  Ethnographic descriptions of the potlatch describe wealth destruction and the feeding of guests, while ritual feeding of the dead during funerals is also described. I do not believe that evidence of prehistoric ritual feeding of the dead should be equated with feeding potlatch guests, nor burnt and scattered prehistoric burial goods with wealth destruction at potlatches (which some researchers suggest may be a historic anomaly).  Although  archaeological evidence of potlatching behavior may eventually come to light, at this time the material culture patterns that could suggest potlatching behavior have not been thoroughly researched.  In many cases ethnographic material culture studies that might aid the  researcher-such as the significance of artistic motifs-simply do not exist. Burley's monograph on the prehistoric Marpole culture type contains the most recent evaluation of the antiquity of certain Northwest Coast ritual practices (1980). Where Carlson cites ritual practices from a wide range of groups on the coast, Burley appropriately, uses ethnographic information from the Coast Salish area only. He assesses ritual concepts and practices among the Salish as falling into several categories: spiritual power available to the layman and to the shaman, daily acts of purification, major rites of passage, and winter ceremonies. He points out that linking prehistoric art objects and styles to specific ritual practices is difficult. A particular zoomorphic rendering such as a salmon, "might represent a spirit power...could be meant to affect the outcome of a fishing expedition; or perhaps, it is related to the first salmon ceremony (1980:67)." Despite the cautious approach in making interpretations about the significance of prehistoric art Burley states, "...specific inferences  19  of Marpole ceremonial and ritualistic practices may be drawn. They are based on a common presence of idio-technic items with Coast Salish culture (1980:68)." Marpole phase objects associated with ceremonial activities include drinking tubes used by puberty initiates, and scallop shell rattles used in the sxwayxwey dance. Burley's inferences are more convincing because they are supported by ethnographic information from the same locality as the prehistoric sites under investigation. But inferences such as those made by Carlson and Burley are based on the assumption that stylistic attributes retain the same meanings and social contexts over time. Croes (1985) study of the Hoko River site in Washington State also addresses the notion that stylistic attributes can indicate the antiquity of behavioral choices that are specific to ethnic groups.  His examination of weaving techniques used in prehistoric  Northwest Coast basketry has revealed distinct stylistic variations among sites that share Locarno Beach phase characteristics. These 2500 year-old basketry styles correlate with the boundaries of present day Northwest Coast language groups. His findings have led him to question the notion that the Locarno Beach phase represents a single ethnic group.  In  previous studies outlining a sequence of prehistoric cultural development on the Northwest Coast archaeologists have focused primarily on changing styles of artifacts relating to the food quest, describing change in terms of evolving economic strategies that have created new forms of social organization.  These prehistoric developmental stages outlined for the  southern coast have been viewed as representing cultures, as reflected in the term "cultural phases." Croes asserts that stylistic variables in basketry techniques are better indicators of ethnicity than variability in stone and bone artifacts. Croes suggests that these economic stages should be viewed as economic strategies rather than cultural stages, shared among . several distinct ethnic groups.  20  DEVELOPING A METHODOLOGY FOR THE ANALYSIS OF STYLE IN PREHISTORIC NORTHWEST COAST ART  The studies reviewed above all attempt to explain the dynamics behind stylistic variability.  Boas, Wingert, Borden, and Duff attempt to relate the spatial distribution of  styles to the diffusion of stylistic variables from centres of origin. Stott and Jonaitis explain the presence of two or more styles in a society as a function of different ritual contexts or roles played by each style. Conkey links the distribution of motifs at prehistoric sites to the number of different groups visiting a site and the degree of interaction between groups. Fischer, Levi-Strauss, and Wolfe correlate motifs and design principles with different forms of social organization and the psychological states they engender. Hodder, Wobst,  Plog,  and  Suttles  associate stylistic behavior  with  the  communication of group identity under different conditions such as the size of social networks, degree of intergroup competition, and form of political organization.  Lastly  McGhee, Stryd, Carlson, and Burley associate stylistic choices with ritual complexes and symbolic concepts specific to local ethnic groups.  Croes has discovered specific craft  techniques that correlate with regional ethnic boundaries. In selecting approaches that might be used in the present study of prehistoric Northwest Coast art both the adequacies of the theories and applicability to the data available must be considered. Only one of the approaches summarized above-the linking of motif shapes to forms of social organization and associated psycho-emotional states-is rejected as an avenue of enquiry for the present study. Reservations about the selective use of ethnographic data and the interpretation of the results have been noted in the previous discussion. The limitations of the data available for the present study restricts the evaluation or testing of other approaches discussed in this chapter.  There are three outstanding  21  deficiencies in the data relating to prehistoric Northwest Coast art. Firstly, it is likely that most prehistoric art was made of wood which has rarely been preserved.  Secondly,  radiocarbon dates can provide only rough approximations of the ages of artifacts, and thirdly, in most cases the context relating to the artifact-its placement within a midden, house structure or grave-was not recorded. Plog's study of Southwestern ceramic design reveals the differences in the supportive data available for the Northwest Coast in comparison to his studies in the Southwest. Tree ring dating and diagnostic pottery styles allow refined temporal definitions of sites in the south, while in the Northwest Coast, components cannot usually be dated to a precision of greater than several hundred or often several thousand years. Further, most sites in the Southwest were only occupied for a few decades in contrast to the millennium or two occupation of Northwest Coast sites. In the south the dry climate and adobe architecture ensures good preservation of living areas, while on the Northwest Coast the house structures that might provide demographic information are poorly preserved in the wet, acid soils. Due to the problems noted above it would be difficult to assess the possible prehistoric centres of origin and diffusion of stylistic traits, or the relative quantity and quality of art from different localities as Duff and Borden have attempted. Assessing the variety of motifs at different sites as a clue to possible aggregation sites and the amount of interaction between groups would be difficult due to the poor definition of length of site occupation. There are two remaining perspectives that have potential for interpreting the dynamics behind stylistic variation in prehistoric art of the Northwest Coast. The first approach links stylistic homogeneity and heterogeneity to the expression of identity or membership.  Hypotheses by Wobst, Hodder, and Plog on increased intergroup stylistic  differences and intragroup homogeneity can be fruitfully examined in the context of the Northwest Coast where archaeologists are searching for the earliest evidence of elite groups  22  with inherited economic and social privileges and regional differences which might reflect language-cultural divisions between prehistoric coastal populations. Wobst's approach in defining artifact classes which exhibit the most stylistic variation and then investigating the possible messages being communicated might also be examined in the context of the Northwest Coast where messages relating to rank, class, wealth, and ritual prerogative might be expressed in the decoration of objects. A significant difference between Wobst's study of Yogoslavian ethnic groups and the Northwest Coast cultural area is that the assertion of class, rank or ritual prerogative by members of the same village would be much harder to detect in the prehistoric record than intergroup differences among ethnic groups as in the Yogoslavian study. Lastly, studies suggesting that stylistic choices in the prehistoric record may relate to the historic cultural practices of aboriginal groups must be considered and carefully evaluated.  Unfortunately,  the  supporting ethnographic information  detailing  the  significance of motifs and decorated forms among the Salish is poor. It would therefore be unwise to attempt to comment on the significance or meaning of particular motifs and decorated forms as they occur in the prehistoric record. But it may be possible to discover significant associations between prehistoric stylistic variables in decorated objects that match historic cultural practices or preferences, as in Croes discovery of continuity between historic and prehistoric basketry techniques in Makah sites. In conclusion, this study can evaluate stylistic change in each prehistoric cultural phase and reflect on how the changes may relate to changing forms of social and economic organization and the evolution of present day cultural practices.  23  CHAPTER 3. THEORIES OF THE DEVELOPMENT OF PREHISTORIC NORTHWEST COAST ART  This chapter presents a survey of literature on prehistoric art in British Columbia in a search for hypotheses which might be profitably examined using the data collected for this thesis. This is a chronological review, allowing the origin of certain ideas to be traced and evaluated.  EARLY THEORISTS  In Primitive Art (1927), Franz Boas devoted a chapter to the analysis of Northwest Coast art.  Boas noted that in certain parts of the coast, art played an important role in  validating the title to rank and privileges. He debated whether the use of animals as crest figures had amplified the importance of art, or conversely, if the concept of painted symbols had enriched the totemic concept.  He postulated (1927:281) that each influence had  enhanced the other: ...the particular symbolic development of art would not have occurred if the totemic ideas had been absent...we are dealing with the gradual intrusion of ever fuller animal motives into a well established conventionalized art...on the other hand it seems quite certain that the exuberance of totemic form has been stimulated by the value given to the art form. Boas went on to observe that totemic or crest-related art was more highly developed on the north coast than the south. He was the first to propose that this northern art style had spread south and was exhibited in its weakest form among the Coast Salish who retained an older, simpler style:  24  In the south, there is clear evidence of the late exuberant development of the totemic, or perhaps better, crest idea...The multiplicity of forms among the numerous small divisions of the Kwakiutl and the sporadic appearance of animal forms among the adjoining Salish are ample proof of these relations. The style has undoubtedly its home in northern British Columbia and southern Alaska. The manufactures of the tribes of Vancouver Island show a far more extended use of geometrical ornamentation than those of the northern tribes. I am under the impression that these are survivals of an older style (1927:281). Boas found evidence for his notion of the retention of an older style on the south coast in the geometric motifs on utilitarian items such as combs and boxes manufactured by the Kwakwaka'wakw, in whalebone clubs produced by the Nuu-cha-nulth, and in the art of the Coast Salish in the Gulf of Georgia. Boas' comments can be restated in the following hypothesis: Coast Salish art and some aspects of Nuu-cha-nulth and Kwakwaka'wakw art represent a survival of an older style of art on the coast. Amongst the Haida, Tsimshian, and Tlingit, where concepts of crest ownership and inherited rank and privileges are most highly developed, "the fullest style of a richer ornamentation" (1927:288) developed. According to Boas (1927:288) features of this symbolic style include a tendency to ornament the whole body of an object, avoiding a plain background, the interlocking of figures, and the use of certain standardized decorative motifs. "The further south we go the more meager becomes the vestiges of the symbolic style (1927:295)." Boas' analysis of Northwest Coast art was first presented in 1897. In the ensuing five years, Boas headed the Jesup North Pacific Expedition, which conducted archaeological and ethnological investigations in British Columbia and Alaska. As expedition head and also editor of the manuscripts presenting the expedition's findings, Boas had the opportunity to further reflect on his ideas about art and how they fit into his overall understanding of the cultures of the coast. Boas and expedition archaeologist Harlan Smith cited evidence of a prehistoric migration of interior peoples to the coast. They believed that analysis of human remains in coastal shell middens showed two physical types present in components representing early stratas of human occupation. Artifact analysis seemed to support their  25  claim of a migration from the interior to the coast: while interior archaeological sites showed a uniformity from early to late levels of occupation, coastal middens revealed interiorassociated traits in early levels.  The discovery of artifacts with simple, geometric  ornamentation in early layers of several coastal sites was taken to confirm the association with the province's interior plateau, since Plateau ethnographic art is characterized by geometric rather than representational art. Smith stated (1903:190): It is probable that at an early time a migration took place from the interior to the coast and Vancouver Island. This migration carried the art of stonechipping, pipes, and decorative art, to the coast. Seventy years later, physical anthropologist Jerome Cybulski (1973) found no support for the notion of two skeletal types. The differences were accounted for by the practice of infant skull deformation. In the years following the Jesup expedition the notion of a migration of interior peoples to the south coast, and a north to south diffusion of a northern art style exerted a strong influence upon the next generation of anthropologists, including Philip Drucker, Marion Smith, Wilson Duff, and Charles Borden. Philip Drucker in particular integrated Boas' ideas in Indians of the Northwest Coast (1955).  Art of the Northwest Coast was characterized by two styles, one which  emphasized conventionalized treatment of subject matter and two-dimensional design while the other stressed three-dimensional forms with realistic treatment of animal and human subjects (1955:161). The Haida, Tsimshian and Tlingit exhibited the former style while the Wakashan (West Coast and Kwakwaka'wakw) exhibited the latter style. He viewed Coast Salish art as a simplified derivative of the Wakashan style. Based on an examination of prehistoric stone sculpture from the Lower Fraser River Drucker postulated (1955:178): ...these ancient objects fit the artistic traditions which the historic Coast Salish derived from their Kwakiutl and Nootka neighbours. This archaeological material thus fits the hypothesis just suggested: that the Wakashan style, and its Salish derivatives, may have been the old form. Gradually the ancestors of the historic Tlingit, Haida and Tsimshian modified that basic pattern into the subtle, more symmetrical and also more static and rigidly standardized style...  26  In a major study of Salish sculpture Paul Wingert (1949:1) also viewed Salish art as exemplifying an earlier coastal style: The comparatively simple culture of this area has produced less complicated and diversified wood carvings than the more northern regions, and the lack of the mid-nineteenth century florescence which developed in the north under the stimulus of metal tools makes Salish sculpture more representative of earlier basic styles. Unlike Drucker who saw Salish art as a derivative of Wakashan sculpture, Wingert saw the influence coming from the other direction or at least being equal (1949:121), The association of Kwakiutl and Nootka potlatch and other figures with carvings of the northern Salish strain indicate that, if they were not at one time influenced by the older Salish tradition, they at least shared with them their basic sculptural elements. After a comparative analysis of sculpture from both coastal and riverine Salish groups in British Columbia and Washington State, Wingert defined major "style centres" and related marginal styles for the Salish area. Style centres exhibited a bold, deft handling of motifs and carving styles while the marginal areas exhibited elaboration or deterioration of essential style features. Wingert postulated that the essential features of the Salish style were more prominent in inland, up-river areas, which led him to hypothesize that Salish art had developed amongst inland rather than coastal Salish groups. He noted, "An important fact favoring an earlier date for these up-river objects is the boldness and the surety of their approach to problems of design and expression...(1949:119)."  Unlike Boas and Smith, who  postulated that early prehistoric art was geometric, Wingert noted two fundamentally different traditions-one geometric and the other naturalistic~and stated, "both of these traditions are ancient, and may in fact derive from separate centers of origin. It is tempting to place one of these near or in the Puget Sound area and the other up the Fraser (1949:119)." Wingert postulated that ethnographic art of the Duwamish, Quinault and Lillooet represented the old style (though neither the Duwamish nor the Quinault could be considered up-river or inland groups).  27  Boas', Drucker's and Wingert's reconstructions of the evolution of Northwest Coast art reflect the theoretical interest of the early part of this century in reconstructing routes of migration and diffusion to account for the distribution of cultural traits. The reconstructions required subjective assessments on which motifs and styles were strongly versus weakly developed, or whether the art work was crudely versus finely executed.  While it is  reasonable to assume that northern art and its expression in crest motifs may have influenced central coast groups such as the Nuxalk and Kwakwaka'wakw while having little or no effect on Salish groups, there is no reason to conclude that the Salish style in its apparent naturalism and use of geometric motifs is necessarily older, weaker, less developed or representative of a prehistoric style once used throughout the coast. Prehistoric sites on the south coast do reveal an ancient art tradition which has links to Salish ethnographic art. In a round-about way some of the hypotheses put forward by the early theorists may turn out to be valid, but the theoretical notions out of which the ideas were generated are inadequate. Similar inadequacies are found amongst a group of writers who postulate stylistic connections between the Northwest Coast, Asia and Meso-America (not to mention the South Pacific and South America), and suggest that the Northwest Coast style is derived from an Asian antecedent.  Studies by Coe (1972), Inverarity (1972), and Schuster (1951) present  examples of similar motifs and principles of composition shared by diverse cultures in Asia, the Northwest Coast of North America, and Central and South America.  What their  presentations lack are hypotheses accounting for the distributions of motifs over this wide geographic areas.  In the absence of any explanation for motif similarity, and with the  suspicion that wherever one looked in the world one could find motifs that reoccur over vast distances, it is difficult to treat this research seriously. It is easy to agree with Inverarity (1972:784) when he states that his research method doesn't prove anything:  28  What I really have tried to do in this paper is merely show that forms, motifs, arrangements and so forth can be compared and found in common. This does not prove anything but when I find such an abundance of qualitative relationships I believe each of us must speculate more and more. Upon careful analysis of the evidence presented by Inverarity, many of the similarities he postulates appear trivial, or at best, tenuous. Is it so remarkable that a 19th Century Haida bowl and an undated ancient Chinese sculpture from Hsin-Yang, Honan both depict an animal holding a rigid stick in the front teeth? Out of twenty-three similar art elements found in both Asia and the Northwest Coast of America, twenty of these elements are unremarkable similarities such as faces with protruding tongues, squatting postures, and combined bird and snake motifs. To their  credit, the proponents of Pan-Pacific art styles challenge other  anthropologists to refute their claims or help account for the similarities.  In the article  "Split representation in the art of Asia and America" (1963) Levi-Strauss proposed that the similarities in values held by two cultures would account for the manifestation of similar art styles. His theory negated the need to search for the mechanisms of transmission between geographically separated peoples.  (Levi-Strauss perspective on style was discussed 1  previously in Chapter 2). Levi-Stauss' ideas are not far removed in spirit from Boas' when the former remarked that Northwest Coast art is an integral part of the expression of fundamental cultural values, such as the importance of inherited rank and privileges. But Boas' historical method of tracing the paths of diffusion of ritual complexes and the migration of prehistoric populations is a methodological world away from the structuralist notions of Levi-Strauss.  Interestingly, another Northwest Coast anthropologist, Wilson  Duff, would struggle to mediate between the particularist view of Northwest Coast anthropology inherited from Boas, Drucker and others and the exciting opportunity for interpreting art styles and forms offered by French Structuralism. A Northwest Coast ethnologist and archaeologist, Duff conducted a study of prehistoric stone sculpture. Prehistoric Stone Sculpture of the Fraser River and Gulf of Georgia (1955) remains the most comprehensive work on British Columbia's prehistoric art  29  to date. In the absence of firm dates for the archaeological data, Duff was forced to fall back on historical methods to reconstruct stylistic developments in prehistoric art.  But Duffs  reconstructions stand out amongst those of other scholars who have tackled the same subject. Firstly, Duff synthesized all of the available data on Northwest Coast archaeology, linguistics, and ethnology and applied it to his sample of stone sculpture. His carefully crafted hypotheses were formulated following extensive investigation and evaluation of alternative arguments. Secondly, his analysis did not end with the examination of stylistic traits and their relationship to art in other parts of the coast. His conclusions presented a well-developed model of craft specialization with supportive linguistic and archaeological data that attempts to explain the dynamics behind stylistic development and change on the Northwest Coast. Duff stated (1956a:115): Recent discussions of the subject [of Northwest Coast prehistory] have been concerned more with tracing the origins of individual elements of the culture than with describing their repatterning and elaboration into the complexes that make this culture area distinctive...A fully developed complex is more than the sum of its elements, and the origin of a complex is not fully explained by tracing the origin of its elements. Duff proposed the following scenario (1956a:93) for the development of stone sculptured bowls in the Gulf of Georgia and Fraser River areas. Out of a wide-spread use of animal-form bowls executed in a variety of styles and mineral types throughout the Gulf of Georgia, Lower Fraser and Columbia River areas, there evolved a specialized soapstone sculpture depicting a human holding a shallow bowl. While the Lower Fraser River was an early centre of sculptural activity, as the carving complex evolved the interest in carving spread up-river until the Lytton area became the centre of production. Eventually the seated human-figure bowls reached  their most elaborate development further up-river near  Lillooet. The stone sculpture complex developed within the Marpole archaeological horizon, beginning 2000 years ago, climaxing sometime after 1000 years ago, and continuing into historic times. The use of human-figure bowls and stone bowls (1956a:56) by Coast Salish shamans suggests that the climax of the sculpture complex paralleled an elaboration of  30  shamanism and ritualism, and perhaps a general elaboration of culture. The distribution of human figure bowls suggests that at one time there may have been a single linguistic and cultural region extending along the Fraser River from its mouth to Lillooet. The climax of the stone sculpture complex correlates with the increased use of ground stone during the Marpole period. As the sculptural complex developed there was increased use of stylistic traits typical of ethnographic Northwest Coast wood sculpture. The stylistic traits appear to have little in common with Coast Salish wood sculpture, suggesting that the stone sculpture style represents an early style that was in use throughout the Northwest Coast. Ethnographic Salish sculpture may represent a later decline or change in style. The lack of stone sculpture in late prehistoric sites suggests that wood rather than stone became the popular artistic medium. In later years Duff published several other examinations of Northwest Coast art and stone sculpture (1975, 1981) where he concentrated on examining the meaning of these prehistoric images.  Duff wrote (1981:223): "I think that the explanation of meaning of  Northwest Coast art has lagged behind the explanation of form..." Duff saw the seated human-figure bowls as representing a culmination of both form and meaning (1975:50), yet his previous analysis of all the ethnological and archaeological data relating to the bowls had not allowed him to proceed very far in uncovering the significance of the stone images. Finally, two decades after his 1956 publication, the ideas of Levi-Strauss allowed Duff to push further into the realm of meaning examining universal metaphors that might be addressed in the sculptures' forms. Hence, Duff later explored (1975) the possible themes relating to reproduction, death and sexuality expressed in prehistoric sculptures such as "The Sechelt Image". As appealing as French Structuralism is for generating ideas to explain the meaning behind sculptural forms, on its own this methodology does not lead to fruitful hypotheses to apply to prehistoric societies. A structuralist methodology concentrates on the formulation of models explaining the structural similarity between social organization and the latent  31  content of art, language and myth. It is a qualitative, interpretive approach which at its worst is always self-fulfilling. At its best, in the hands of a painstaking researcher such as Duff with his comprehensive grasp of the quantitative information on prehistoric stone sculpture, structuralism can be used to generate interpretations where none are forthcoming from the objective data. There is much to be gained in using Duffs method of integrating ethnological, archaeological and linguistic information as presented in his 1956 study of stone sculpture. His hypotheses reconstructing the sequence of development in stone bowl forms can now be reevaluated in the light of archaeological data from the past fifteen years. Unfortunately, although a number of recently discovered zoomorphic bowls have associated dates, no seated human-figure bowls have yet come to light under an archaeologist's trowel. Where Duffs work can be most fruitfully reexamined is in the relationship he proposes between the stone bowl complex and other examples of prehistoric art from the Gulf of Georgia and Fraser River regions. A number of antler carvings from sites of all ages have been unearthed which are examined in this thesis, and will allow a review of the possible antecedents of stone sculpture in British Columbia.  RECENT ARCHAEOLOGICAL FINDINGS  The next stage of conjecture about Northwest Coast art took place in the 1970's when archaeologists could evaluate their findings with the help of dates obtained from radiocarbon analysis and a more refined understanding of the prehistoric cultural sequence. Many archaeologists were stimulated to evaluate the decorated objects from regional sites for a 1976 symposium at Simon Fraser University entitled "The prehistory of Northwest Coast Indian art." A publication from this symposium was edited by archaeologist Roy Carlson (1983). Carlson, like Duff, approached the study of prehistoric art as an inquiry into the  32  origins of those systems of belief which engendered Northwest Coast art motifs and carving styles.  The premise presented by Carlson (1983) is that Northwest Coast art is an  expression of spiritual power based on encounters with the supernatural. Spiritual beliefs and their associated ritual complexes fall into four categories: guardian spirits, shamanic spirits, secret society spirits and crest spirits.  Using ethnographic information Carlson  defines the classes of objects that are associated with each category of spiritual power, then looks at the prehistoric record to find the earliest examples of these objects. For example, Carlson states (1983:198): The lightning snake, a powerful spirit for success in whaling, is depicted on the ethnographic harpoon valves. By analogy the prehistoric example shown [from a Marpole phase site] of a wolf incised on the end of a point for a leister or fish spear should represent the guardian spirit of the owner. Although Carlson is unable to discern images relating to family crests which portray inherited privileges, he does cite evidence that the depiction of guardian spirits may go back four thousand years, shamanic spirits may relate to fish effigies from forty-five hundred years ago, and secret societies may be indicated by miniature model masks produced twentyfive hundred years ago. A more recent paper (Carlson 1987) suggests evidence for a 3500 year old antiquity for the memorial potlatch, shamanism, the use of masks, and artistic craft specialization. Carlson admits the inherent difficulty in relating ethnographic information to the prehistoric period, but he under-estimates the differences in cultural practices among distinct coastal groups.  For example, the four categories of spiritual beliefs should be  carefully evaluated for relevance to the Salish area. Cleansing rituals which are a prominent aspect of Coast Salish spiritualism may not fit into this scheme. Although Carlson cites masks as evidence for secret societies, masks are not exclusively associated with secret societies, and secret societies are not a prominent feature of Coast Salish religion. To make inferences about the antiquity of these social complexes in the Salish area based on small mask-like effigies seems inappropriate.  33  Carlson's methodology can lead the researcher into speculative territory, but he presents interesting hypotheses worthy of careful consideration. As Lee Anne Wilson (1986) states in her paper on interpreting prehistoric Southern Cult iconography, "some interpretations are better than none as long as we see them as inferences not facts." Carlson contributes a useful model in which sets of objects related by their symbolic function are discussed together. This offers an alternative to the standard practice of archaeologists in describing objects grouped by their physical composition: antler, stone, wood, etc. Carlson's conclusions are based on material from the southern coast. Although there is not enough dated material to state much about the development of the northern style of art, Carlson remarks that the limited data suggest it was "full blown" at least eight hundred years ago.  Considering the antiquity of some elements of the southern style,  Carlson postulates (1983:204), "...that classic northern style resulted in part from the interaction created by the influence of southern art on an originally simple rendering of life forms on the Northern Coast." The next archaeologist to be considered is Charles Borden, who first outlined the cultural development sequence for the Lower Fraser River and Lower Mainland regions. In Carlson's 1983 publication Borden provides an important synthesis of prehistoric art from sites in southwestern British Columbia from the mouth of the Fraser River to the beginning of the Fraser Canyon at Yale (corresponding to the eastern boundary of the Halkomelem). In general, Borden's theories on prehistoric art are not as well integrated into his model of prehistoric cultural development as one might expect from someone who proposed the cultural sequence. Borden himself stated (1983:131) that although art might shed light on problems in the prehistoric cultural sequence, his paper concentrated on description and chronology. Despite the limitations of his scope, several important hypotheses were put forward. Firstly, Borden provided evidence for the continuity of certain motifs and styles over a three thousand year period. Depiction of ribs, vertebrae, angled-eye forms, and "feather-like"  34  designs occur from Locarno Beach and Baldwin phase components to art of the present day. This continuity led him to propose a sequence of the development, climax and decline of prehistoric artistic activity in the Salish area which formed his major hypothesis: The Developmental Period: from approximately 5500 to 3050 B.P. includes the Eayam Phase from the Esilao site on the Lower Fraser River near Yale, and the St. Mungo Phase in the Fraser Delta. The Climax Period: from 3050 to 1600 B.P. includes the Skamel and Baldwin phases in the Lower Fraser area, and the Locarno Beach and Marpole phases in the Fraser Delta. The Post Climax Period: from 1600 to 750 B.P. includes the Emery Phase in the Lower Fraser and Whalen II Phase in the Fraser Delta. The Late Period: from 750 to 150 B.P. includes the Esilao Phase in the Lower Fraser and the Stselax Phase from the Fraser Delta. Borden's work in outlining the cultural phases of the Fraser Delta and Lower Fraser River is to be commended for relating the decorated objects to the larger artifact assemblages. His article contains a careful description of art items with their associated dates and contexts, but as previously mentioned, it fails to integrate the art evidence into the existing models of cultural development in the Fraser River area. Why are Marpole and Baldwin phases periods of artistic climax?  Why are later periods less productive? Borden  does not address the cultural dynamics explaining his proposed chronology. Unfortunately the data are unsupported by a coherent theory to explain the development and devolution of artistic activity and on their own constitute too small a sample to be convincing. For instance, when he comments on the paucity of Emery Phase decorated objects in comparison to the previous Baldwin Phase, he is comparing three objects from the Emery component to eight object from the Baldwin component. Borden (1983:155) admitted that the lack of artistic evidence in sites from the later periods might be due to factors of preservation and sampling: This dearth of evidence may in part be attributed to inadequate sampling of critical components, to the vagaries of preservation, and especially to the paucity of excavations at crucial sites in the region between the Fraser Canyon and the Delta.  35  Borden also recognized that wood likely became an important medium of artistic activity in the Marpole Phase since all the tools needed for woodworking are present in assemblages of this period. Therefore, it is puzzling that Borden did not consider the possibility that the absence of stone and antler artwork from Post Climax and Late Period sites reflects a use of wood rather than a decline in artistic activity. Wayne Suttles has commented (1983:87) that quantitative and qualitative changes in the prehistoric art record may indicate a change in ritual emphasis, rather than dramatic cultural change. Part of Borden's reluctance to consider that evidence of artistic activity would be scarce in the Late Period may have been his lack of understanding of ethnographic Coast Salish art.  Salish art was not well-understood or even well-represented in standard  reference books on Northwest Coast art.  Perhaps Borden was influenced by Drucker's  theories when he stated that in the Nineteenth Century Salish artists had been heavily influenced by the "classic art" of their northern neighbours. Borden does not show evidence of comparing Coast Salish art with his prehistoric sample. Several prehistoric objects from Fraser region sites that he compares to Kwakwaka'wakw ethnographic art items (1983:141, 147) lack convincing similarities. As well, there are several instances where Salish examples of similar motifs or carving styles are overlooked.  For example, an intriguing wooden  fragment from the Skamel Phase associated with dates of 2080 ± 130 B.P. and 2000 ± 120 B.P. bears a striking resemblance to the carving style of several Coast Salish spindle whorls. Similarly, a bone fragment from the Whalen site at Point Roberts (1983:Fig. 8:30c) which Borden states (1983:160), "has no close parallel in any other Lower Fraser period" is similar to feather motifs on Coast Salish spindle whorls. Borden's attitude to Coast Salish art caused him to attribute several finely executed objects, which were from undated or problematic excavations, to the historic period rather than the Late Period.  He argued that masterful execution and true composition were  features of the historic period. For example, he suggested that the deep, sharply-cut lines on a zoomorphic blanket pin of antler (1983:Fig. 8:34a) could only be made with steel tools of the  36  historic period, even though Smith illustrates very similar antler carvings in prehistoric Lillooet burial sites (1899:158 Fig. 114 & 115). When Borden's site records are carefully reviewed, one can discover other decorated objects from problematic circumstances which Borden assigned with confidence to Locarno Beach or Marpole phases. Like most others before him, Borden was occasionally guilty of erring on the side of caution when the data didn't fit his theory and dismissing irregularities if the data supported his ideas. (A stone spindle whorl and bear head from the Milliken site, and two zoomorphic spoons from the Musqueam site were all found sloughed off excavation walls but Borden treated each case differently.) A second hypothesis presented in Borden's (1983:131) examination of prehistoric art is that decorative artifacts that belong to the same time period but are stylistically distinct indicate "the simultaneous presence on the Lower Fraser region of different cultures or subcultures." Borden wanted to present evidence of the distinctiveness of his proposed Whalen II cultural phase appearing at the demise of the Marpole Phase. One of the characteristics attributed to the Whalen II phase was a lack of personal ornaments and stone sculpture in the assemblage. Borden neglected to consider that much of the art in his entire sample was found with burials and since the Marpole site contained hundreds of burials, the sample of stone and antler sculpture was indeed rich. The Whalen site contained only a small number of decorative objects, but significantly, they were associated with a burial. These finely executed human and bird forms can be compared to the best examples of art uncovered at the Marpole site. Whalen II is now considered by most archaeologists to be part of the Marpole Phase (Burley 1980:38). Despite problems with Borden's attempt to define two subcultures living in the Lower Mainland approximately fifteen hundred years ago, his hypothesis regarding the presence of two art styles representing two different cultures remains worthy of consideration. Even though Borden developed his chronology of prehistoric art styles with a very small sample, and underestimated the relationship to ethnographic Coast Salish art, his  37  paper firmly established Northwest Coast art as a local development with strong prehistoric roots. Borden demonstrated the three thousand year antiquity of several motifs typical of (although not exclusive to) Northwest Coast art. Design concepts such as the proportion of head to body, and eye to face, are found in the earliest carvings. The depiction of ribs, backbone and vertebrae are found on several objects that are from two to three thousand years old. Borden identified a U-form Northwest Coast design element on a spoon estimated to be approximately twenty-six hundred years old. Borden stated (1983:143), "Clearly, some of the art forms of the Locarno Beach phase reflect in significant ways the general spirit, concepts and styles of later Northwest Coast art," and in the Marpole Phase, "a fine sense of form, a pride in craftsmanship and an obvious delight in representing the creatures of their environment and the supernatural beings of their myths." In his eagerness to discover deep roots for Northwest Coast ceremonialism Borden occasionally made too much of scant evidence. A Baldwin phase, small stone bear-head with deeply incised lines on the snout and skull, found at a site near Yale (1983:Fig. 8:8b) was interpreted as a muzzle suggestive of ritualism involving young captive bears (Borden was likely influenced by Hallowell's 1933 publication on bear ceremonialism in the Old and New World.) While Duff proposed that the stone bowl carving complex dated from approximately 2000 to 1000 B.P. with some use of stone bowls for shamanic ritual during historic times, Borden proposed that the stone bowl complex occurred from 2300 to 1750 B.P. A third perspective has been put forward by archaeologist Arnoud Stryd (1983) in his 1976 conference paper, "Prehistoric Mobile Art from the Mid-Fraser and Thompson River Areas." Stryd's careful analysis of prehistoric art from Interior Salish Plateau sites led him to propose that Plateau sculpture had begun by 2000 B.P. with the decoration and carving of large stone implements such as mauls. Smaller carved figures in steatite and antler appear in site assemblages by 1500 B.P., for example a zoomorphic steatite bowl is associated with a date of 1220 ± 85 B.P. Stryd guesses that seated human figure bowls date to 800 ± 200 B.P.  38  He concludes (1983:176), "...it appears that Plateau carving including the steatite carving complex probably had its origin in first the decoration and then the carving of large stone implements including mauls and mortars or large non-steatite bowls." Stryd also observes that, in his sample, carvings are primarily recovered in burial contexts but are not age or sex specific. One of the most interesting burials containing numerous zoomorphic and anthropomorphic carved images is that of an infant. The carvings are executed in several distinct styles, countering Borden's hypothesis that different carving styles would indicate the presence of contemporary but distinct cultural groups. Stryd also notes that the absence of carvings in the Nicola Valley and Williams Lake area may mark the boundary of prehistoric Plateau sculpture. Like Borden's findings for the Lower Fraser and Fraser Delta regions, Stryd views ethnographic Plateau art as a degeneration from a late prehistoric developmental climax. As we have seen, many anthropologists and art historians consider art from the northern coast of British Columbia the most refined and sophisticated regional variant of Northwest Coast art. They would agree with George MacDonald's statement (1983:99), "it is particularly the tribes of the northern coast, the Haida, Tsimshian and Tlingit, who produced the classic expression of the Northwest art tradition." What has often resulted from this same admiration is an assumption that the core design features that distinguish the Northwest Coast style originated in the north and spread south, resulting in a less refined execution of essential style elements among the Coast Salish. MacDonald's examination of northern prehistoric art (1983) reveals that there is not enough evidence to support this assumption. Sites in Prince Rupert Harbour and along the Skeena River constitute a small but essential sample of decorated objects from the north coast. Unfortunately, very few radiocarbon dates accompany the artifacts from these sites, so in most cases the dates given are estimates. Despite these limitations, MacDonald was able to roughly outline a picture of artistic development for the north region. In his view, all of the major elements of Northwest  39  Coast art were in place 1000 years ago. Before this time decorated stone and bone artifacts were stylistically simple and rare. The preservation of wood and basketry items at several wet sites shows that a woodworking and fibre technology similar to the present day existed 2000 years ago. MacDonald postulates that wood was a medium of artistic expression in this period and that the basic elements of Northwest Coast art will someday be traced at least as far back as three to four thousand years (1983:117). An anthropomorphic carving on a chisel dating to approximately 3500 B.P. supports the idea that "finely finished carvings were probably being made, although none have survived" (1983:101). MacDonald states his conviction that the key economic and settlement features of coastal culture are evident in the archaeological record in the Early Period dating from 5000 to 3500 B.P., but symbolic modes of graphic expression appear to be absent in this period. This leads him to postulate that art was not yet bound into ceremonial expressions of group identity: We can only infer that the accumulation of historical and mythological traditions by the corporate lineages of northern coastal villages was approaching the threshold where graphic symbols of corporate identity became meaningful. Implicit here is the assumption that graphic symbolism expressed in art works, requires a base of shared cognitive modes, belief systems, etc. which must develop to a certain point, perhaps over several millennia, before it can be meaningfully expressed in art works (1983:101). Although MacDonald goes on to give a detailed description and time frame for the decorated artifacts in his sample, he leaves a trail of provocative and unclear statements. Does he believe that the first expressions of art found in the archaeological record are graphic symbols of corporate lineage identity?  Is all Northwest Coast art related to the  expression of corporate lineages, or is this true only in the north coast? Other archaeologists who have worked in the north coast disagree with his assertion that developed economic and settlement patterns are in place during the Early Period and would argue that the evidence for a corporate identity of kinship groups does not appear until the latter half of the Middle Period dating from 3500 to 2500 B.P. (Gary Copeland pers. comm.). Does MacDonald deduce  40  that there were corporate groups because he found art in early period sites? His argument is not well enough developed to answer these questions. Also unclear is MacDonald's statement that the major elements of Northwest Coast art are in place in the north coast by 3000 B.P. Although he qualifies this by stating that he is referring to the first occurrences of relevant details of style, motif and iconography, it is still difficult to evaluate his assertion, since he doesn't define what these relevant details are. He does discuss the occurrence of what he terms "core features" of backbone, ribs, eyes and mouths, and cites several prehistoric links to ethnographic art styles such as West Coast style whale bone clubs. In a concluding general discussion of Tsimshian cosmology and north coast art MacDonald states that the important elements of northern art are the emphasis on skeletal structures which relate to regeneration and access to food, and on sense organs which relate to the communication between animals and man via shamanic spirit power. The major weakness of this discussion remains a vague delineation between the archaeological evidence, MacDonald's own hypotheses regarding meaning in Tsimshian art, and his concepts of the elements that constitute "classic" Northwest Coast art. For an example of one of the provocative statements that are scattered throughout his paper, MacDonald asserts, without providing evidence, that pecked and ground zoomorphic stone art may be a technological-artistic complex from the Maritime cultures of north Asia (1983:114). Similarly, simple bone pendants with knobs are thought to represent ancestors hung upside down in the manner of Thule human figure pendants (1983:104). These are interesting assertions that deserve a more careful discussion. MacDonald's paper is a valuable synthesis of the prehistoric art record for the northern coast. It contributes a vitally needed sample to compare with the art record for sites on the southern coast. Although MacDonald doesn't go into a comparison of the two areas, it is immediately evident that the trends he outlines follow developments in sites in the Lower Mainland and Gulf of Georgia areas: rib and backbone motifs are found in the earliest examples of zoomorphic art, finely carved pieces with crescent and T-shaped design  41  devices are in evidence by at least 1100 B.P. Unfortunately, no evidence has been uncovered that clarifies the development of the characteristics that make north coast art distinct.  It  remains to be discovered how complex compositions utilizing formlines and standardized design motifs arose. Similarly, there is a poor understanding of how artistic developments parallel socio-economic developments in the prehistoric period. Much more ethnographic information needs to be analysed to understand the role of art in expressing identity, territoriality, and inheritance in northern society before we attempt to extrapolate this information from the prehistoric record. For the most part, archaeologists understandably are more comfortable commenting on their own site-specific data, than developing all-encompassing hypotheses on the evolution of the Northwest Coast art. It is appropriate, therefore, to end this discussion with a hypothesis put forward by Bill Holm in "Form in Northwest Coast Art" (1983). As an art historian, Holm developed his ideas on prehistoric art through an analysis of style in ethnographic art of the coast. Holm makes the point that although most of the art of the coast is three-dimensional, the conception of form which underlies this tradition is twodimensional. Through his analysis of the formal arrangements of design among different groups on the coast Holm develops a scenario to explain various regional styles. Holm proposes that two-dimensional silhouette figures with incised features are characteristic of an early stratum of art in all parts of the coast. Whether geometric or representational in design, the incised details came to represent negative spaces within the composition. While this early style was developing it also diverged into a number of regional styles each of which retained certain common concepts of form and motif from the early period such as: the raised positive/recessed negative space delineation, crescent and Tshaped space reliefs, and skeletal motifs such as ribs and joints. In the north, the use of paint led to a sophisticated formline system, while in the south a simple, naturalistic sculpture developed. At a later period, the two styles converged leading to formlines which overlaid and modified the sculptural forms of totem poles, dishes and other three-  42  dimensional objects. Unfortunately, although several aspects of Holm's hypothesis on the evolution of Northwest Coast art can be tested with the data presented in this thesis, the prehistoric record is inadequate in the central and north coast area where one might find evidence for the merging of the painted formline system with the sculptural emphasis of the south. Holm has carefully selected objects from as far south as the Columbia River and as far north as the Skeena River to show how each region has evolved a distinct style. His perspective allows each region its unique style without presenting any one area as dominant, or "classic". The Coast Salish area which has been characterized as being influenced by the north (Borden), Interior (Smith), or west (Drucker, Carlson) is allowed its stylistic integrity by Holm: There's no way these Salish pieces can be watered down northern designs. The concept of the placement of these reliefs is too sure and too knowledgeable, too perfect...So this is not a watered down or backwater copy of the northern formline (1983:36). Holm states (1986:37) his lack of understanding of how the prehistoric seated human figure bowls fit in to the prehistoric picture, "with their deeply rounded sculptural forms, heavy, arched eyebrows over eyes on a full orb, distinct eyelid lines, flaring nostrils, modelled cheek structures and projecting mouth with full lips" they are unlike other Salish carvings.  CONCLUSION  As this overview of the literature on prehistoric Northwest Coast art has revealed, theories on the evolution of prehistoric art have been influenced by the prevailing theoretical trends of each decade, and the gradual increase in knowledge regarding the prehistoric cultural sequence and the ethnographic art of the coast. Many researchers divide coastal art into two stylistic regions and attempt to explain the differentiation between the two regions.  43  Boas proposed that a "symbolic style" evolved among the Haida, Tsimshian, and Tlingit in complement with their social system.  The symbolic style spread slowly south to the  Kwakwaka'wakw and lastly to the Coast Salish. The southern Salish retain an older style once common along the Coast.  Drucker characterized the northern style as three-  dimensional versus a two-dimensional style among the Wakashan with Salish art as imitative of the Wakashan style. Wingert also saw Salish art as retentive of an early style. All three authors were influenced by the theory that the Coast Salish were recent migrants from the interior Plateau and retaining elements of a simpler, geometric art. Another group of authors allude to connections between the Pacific Coasts of North America and Asia.  Coe, Inverarity and MacDonald all find similarities  between  ethnographic and prehistoric art from the Northwest Coast and ethnographic art from Asia, although no one has attempted a systematic examination of the prehistoric art records in the two areas. Borden, Duff, and Stryd find evidence of continuity between prehistoric art and ethnographic art in the Salish area. All three also propose a prehistoric climax of the design and production of stone sculpture in the Lower Fraser and Thompson river regions, with ethnographic Salish art representing a degeneration from the climax period approximately one thousand years ago.  Duff and Holm propose development sequences for aspects of  prehistoric art based on their analyses of form in ethnographic Northwest Coast art. Carlson and MacDonald attempted to address the cultural dynamics behind changes in the art record in their regional samples. Both agree that the essential elements of Northwest Coast art can be found in the prehistoric art record of eight hundred to one thousand years ago. Anthropologists, and more surprisingly, archaeologists have not been reticent to discuss meaning in prehistoric art.  Hypotheses which suggest a continuity between the  beliefs and rituals of historic and prehistoric occupants of the coast have been put forward by Duff, Borden, Burley, Carlson, and MacDonald. In all of the previous studies the authors all worked with small regional samples.  Furthermore, even though the largest sample of  44  prehistoric art comes from the south coast, most researchers used northern or central coast ethnographic art as a comparative sample.  The present thesis can test some of the  hypotheses put forward with a larger sample, much more temporal information, and a more refined understanding of south coast art. As will be discussed in the following chapter there are serious gaps in the prehistoric art record. The only region that has adequate data is the central Coast Salish area encompassing the Lower Fraser River, Gulf of Georgia and Vancouver Island regions. Most archaeologists working in this area support a hypothesis of in situ development of subsistence and settlement patterns from the beginning of the Marpole phase and likely the Locarno Beach phase culminating in the ethnographic cultural patterns of the Coast Salish (Mitchell 1971, Matson 1981). This study can test the hypothesis that the prehistoric art record will similarly show a gradual evolution of the motifs, design elements, carving techniques, and decorated forms found in ethnographic Coast Salish art. The methodological difficulties involved in testing this simple hypothesis will be discussed in the following chapter.  45  CHAPTER 4. RESEARCH METHODS  RESEARCH  OBJECTIVES  The primary goal of this study is to present a model of stylistic development for prehistoric art of the Central Coast Salish region of the Northwest Coast culture area, and to relate the model to the sequence of prehistoric cultural development for the Gulf of Georgia. Secondly, the results are used to evaluate hypotheses on the evolution of Northwest Coast art and its regional variants. Thirdly, interpretations of the significance of stylistic variation in the prehistoric record are presented in the concluding chapter.  THE DATA SAMPLE  In attempting to define stylistic development and change in art one must have a set of variables which can characterize the attributes of a particular time period and region. Simple geometric designs (such as dots enclosed by circles, or zigzag lines) are common throughout North America in many regions and time periods, while the representational art tradition is the dominant feature of Northwest Coast art.  For this reason, artifacts  embellished solely with non-representational geometric designs are not included in this examination of stylistic change in prehistoric Northwest Coast art. The sample chosen for this study consists of prehistoric objects whose forms have been modified to depict human or animal subjects. The sample includes objects made of stone, bone, antler, shell, and wood. It contains such diverse objects as a flat pebble abraded to form rudimentary limbs, an antler comb with an elegantly carved bird design, and a bone disc incised with a grinning, bearded human face.  46  Two important variables for selection of the sample are locality and age.  Each  artifact in the sample has an archaeological site provenience and comes from an assemblage that has been dated.  In most cases, site assemblages have been dated using estimates  obtained from radiocarbon analysis and comparisons with collections from well-documented archaeological sites. Where possible, the original field notes, site reports, and radiocarbon dates were scrutinized to obtain accurate evaluations of the artifacts in this sample.  Sites included in the sample  Figure 1 shows a map of the Northwest Coast with archaeological sites mentioned in this study identified by their local names and archaeological designations. The study sample consists of 155 artifacts from 31 sites within the Gulf of Georgia prehistoric region, and 88 artifacts from 27 sites in adjacent regions, distributed as follows:  Gulf of Georgia Lower Mainland Southern Vancouver I. & Gulf Is Strait of Juan de Fuca, Puget Sound, & Columbia River Lower Fraser and Thompson Rivers Central and North Coast  155 116 39  Total sample  21 38 29 243  No dated anthropomorphic or zoomorphic decorated material was available from sites occurring north of the Skeena River, south of the Columbia River, in the Johnstone and Queen Charlotte Strait region, and in the Queen Charlotte Islands. The total of 243 objects includes a small number of undated artifacts considered and compared with the better dated objects. These artifacts are included in the study because of their similarity to dated objects and their geographic provenience. Included in the undated material are surface finds from archaeological sites, undated burial items, objects donated to  47  Figure 1. Map of sites included in sample. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 11. 18. 19.  DcRt 15 Cadboro Bay DcRt 12 Maple Bank DcRv 1 Pedder Bay DeRt 2 Pender Canal DfRS 3/DgRs 14 Whalen Farm DfRu 8 Helen P o i n t DfRu 13 Montague Harbour DgRr 1 C r e s c e n t Beach DgRr 2 S t . Mungo Cannery DgRr 6 Glenrose Cannery DgRs 1 Beach Grove DgRw 4 F a l s e Narrows DhRs 1 Marpole DhRt 2 S t s e l a x DhRt 3 Musqueam K. DhRt 4 Musqueam N.E. DhRt 5 P o i n t Grey DhRt 6 Locarno Beach P o r t Hammond 45SJ1 C a t t l e Point 45SJ24 E n g l i s h Camp 45SJ25 Garrison 45SJ100-105 S u c i a I s l a n d DjRi 3 M i l l i k e n Esilao DjRi DjSp 1 Yuquot EdRk 1 Texas Creek EdRk 6 McPhee EdRk 8 Lehman EeQw 1 Chase B u r i a l EeRb 10 Kamloops I.R. EeRk 4 B e l l EeRi 21 Seton Lake E l S x 1 Namu Kwatna s i t e s : FaSu 1 A x e t i FaSu 2 N u t l i t l i q u o t l a n k FaSu 10 A n u t c i x P r i n c e Rupert s i t e s : GbTo 18 Dodge I s l a n d GbTo 23 Garden I s l a n d GbTo 31 Boardwalk GbTo 33 Lachane GbTo 34 K i t a n d a c h GbTo 36 Baldwin 45CA24 O z e t t e 45CA213 Hoko R i v e r 45SK33 Fishtown 45SK59 Conway 45SN100 B i e d e r b o s t Sauvies I s l a n d The D a l l e s - D e s c h u t e s s i t e s  48  museums by private collectors, and artifacts from sites with undated assemblages. The uncertain provenience of these artifacts is always noted and they have not been included in the tables and summaries of stylistic attributes.  Problems with the sample  To undertake this analysis requires a recognition of certain inadequacies in the data available for this study. First of all, although wood and fiber formed a major portion of the material culture of people living in the Pacific Northwest, the climate and soils of this area prevent preservation of these materials. Apart from one unique site, only six other decorated wooden objects have been recovered from  prehistoric contexts.  Luckily, a complete  inventory of wooden and fibre objects which is thought to date from 200 to 400 years old has been recovered from the Ozette site. Although this site is of too recent age to tell us much about developmental stages in Northwest Coast art, it provides invaluable data on the quantity and stylistic variety of wooden art forms from several house floors within a single village. Another problem with the data used in this study is the poor definition of dates for many of the artifacts. Archaeologists are often forced to define assemblages in broad five hundred to fifteen hundred year time periods due to a lack of datable material. This makes it difficult to compare artifacts from different sites which may include one object dated from 1650 to 650 years old, and another dated from 2000 to 1500 years old. The artifacts could be the same age or could differ by fifteen hundred years. A related problem is the lack of published site reports for most of the archaeological sites pertinent to this study.  Information in unpublished student reports, preliminary  summer field reports, sketches in notebooks, and even hastily written notes on brown paper bags had to be sorted through and evaluated. Although there are radiocarbon dates for most sites in this study one is rarely certain how these dates relate to the artifacts used in this  49  sample.  As well, since much of the site data remain unanalysed the accompanying  contextual information which might have added to this study has been lost. For the reasons discussed above the research objectives and interpretations of stylistic variability have been shaped somewhat by the limitations of the data. As is evident in the geographic breakdown of the sample, the Lower Fraser River region is the only wellrepresented area in this study. More archaeological recovery has occurred in this region because of historic human settlement in the area, the proximity of universities and museums sponsoring archaeological investigations, and increasing urban development leading to salvage archaeological work. Ironically, the southwest part of the province which has the richest record of prehistoric art, has a poor record of ethnological research which might have aided this study. The art of the north and central Northwest Coast which has attracted most of the attention from anthropologists, conversely, has the poorest prehistoric art record. Since there is only one region with an adequate sample representing different periods in prehistory this thesis is only able to present a model of stylistic development for prehistoric art of southwestern British Columbia. This region corresponds to the Central Coast Salish cultural area occupied by Halkomelem and Straits Salish groups. Data from other regions although scant, add significantly to this study and its conclusions.  Recording and Analysing  Information about each artifact in the study was recorded on a data sheet. Drawings were made of each item, and in most cases black and white photographs were taken and attached to each data sheet. (A few items were on public display and could not be removed for close study or photography.) Only illustrations and replicas were available for decorated artifacts from northern and central coast sites (Prince Rupert, Bella Bella and Kwatna localities).  50  Artifacts are referred to in the text by their figure numbers. The catalogue number, site origin, dimensions, museum provenience, and stylistic attributes for each artifact are listed in Appendix I. In the analysis and discussion contained in Chapter 5 the sample is divided according to age and region of origin so that motifs and carving styles can be compared temporally and geographically. Although it would have been preferable to divide the sample into 500-year periods, the poor temporal definition of most site assemblages precluded this method. Most artifacts come from components with one or two radiocarbon dates and a general cultural phase assignment. For this reason, the analysis is divided into four sections corresponding to the prehistoric cultural sequence for the Gulf of Georgia region. They are as follows: St. Mungo and Mayne phases Locarno Beach phase Marpole phase Late period phases  4300 to 3300 B.P. 3300 to 2400 B.P. .2400 to 1500 B.P. 1500 to 150 B.P.  The late period is comprised of four local phases: Stselax, Gulf of Georgia, San Juan, and Developed Coast Salish. Material from other regions of the Northwest Coast and Plateau are also divided into the four temporal units outlined above. Although regional prehistoric cultural sequences differ from that of the Gulf of Georgia, they share developmental similarities that allow the material to be divided into the four sections without sacrificing internal coherence. All dates and ages are discussed in terms of the age or date before the present, referred to as B.P., rather than B.C. or A.D. The analysis of the prehistoric art record concentrates on five stylistic variables: carving techniques, decorated forms, design elements, principles of design, and motifs. These variables are discussed in the following section.  51  A DEFINITION OF TERMS USED IN THIS STUDY  Geographic and cultural terms: Northwest Coast cultural area: an area from Yakutat Bay in Alaska to Cape Mendocino in northern California within which aboriginal groups shared common economic and cultural patterns. Ethnographic information: information collected by anthropologists on the traditional cultures of Native peoples. West Coast Indians: formerly referred to as Nootka Indians, comprised of the Nuuchah-nulth, Dididath and Makah groups. Central Coast: archaeology sites falling within the territories of the Nuxalk, Heiltsuk, and Haisla native groups. North Coast: archaeological sites falling within the territories of the Tsimshian, Tlingit, and Haida native groups. Gulf Islands: islands in the Strait of Georgia from the San Juan Island in the south to Gabriola Island in the north. Lower Mainland: the Fraser Delta and Fraser Valley extending to Yale. Lower Fraser River region: from Yale to Texas Creek on the Fraser River. Mid-Fraser and Thompson River region: the area from Lillooet on the Fraser River to Chase on the Thompson River.  Stylistic Terms:  Bill Holm (1965) first developed a vocabulary for describing Northwest Coast art in his study of design used by northern coastal groups.  Many of his terms apply to two-  dimensional design of the Kwakwaka'wakw, Haida, Tsimshian and Tlingit and have been used inappropriately in describing the art of other areas. Kew (1979) has added to this Northwest Coast art vocabulary by providing much needed definitions for Coast Salish two  52  and three-dimensional art. I have attempted to select and define terms used in previous stylistic studies such those just mentioned, and only when necessary add new words. decorated form: refers to the type of artifact that has been embellished with anthropomorphic or zoomorphic designs, e.g. barbed harpoon point, stone bowl, pendant. effigy: an artifact with an unknown function or use exhibiting an anthropomorphic or zoomorphic image. figurine: an artifact shaped to present a full human figure. design element: a standardized carved or painted shape widely used in Northwest Coast art. Design elements are either positive design features or negative shapes that help define positive elements in a composition. Figure 2 illustrates design elements referred to in this study. motif: a representational subject or theme found in many decorated objects such as "the rib and backbone" motif or the "seated humanfigure"motif.  The following terms refer to carving techniques some of which are illustrated in Figure 3: sawing: to shape features in stone with a sawing motion. pecking: to shape features in stone by striking with another stone. abrading: to shape features in stone by rubbing or scraping the object against a stone. modelling: defining features with gentle contours that create rounded, sculptured forms. relief: the projection of sculptedfiguresand forms from a flat, background surface. high relief: forms are almost fully-lifted from their background and are modelled in three dimensions. low relief: a small portion of the volume of a form is raised from the background surface. incising: defining features with a thin, carved line. The line is a positive feature. engraving: defining features by cutting away portions of a surface. An engraved line reveals an area that has been cut out and is a negative feature. two-dimensional block engraving: cutting away background to leave a raised primary form. The cut-away area is a negative feature.  63 Figure 2. Northwest Coast Design Elements.  Negative C  Wedge  Chevron  ^  Elongated Wedge E  Eyelid Form  U Form  \  T Shape  ^  Crescent y>  ^>-dAl  Ovoid  54  Figure 3. Prehistoric Carving Techniques.  55  The following terms refer to principles of design illustrated in Figure 4: formline: a continuous positive line delineating the primary features in a design. A formline may be a painted line or a raised line created by engraving the surrounding space. Formlines often swell and diminish as they interact with other elements in a composition. form surface: a connecting uniform surface whose perimeter is marked by an engraved line or by a conjunction with another lower surface. The primary form surface stands out in precise relief. There may also be secondary and tertiary form surfaces. Two-dimensional block engraving is used to create complex designs with several form surfaces.  57  CHAPTER 5. AN ANALYSIS OF 4500 YEARS OF PREHISTORIC ART  5.1 ART FROM 4500-3300 B.P.  The first evidence of artistic activity on the Northwest Coast appears in the middle period of the prehistoric cultural sequence. As in other parts of the New World, this middle period is characterized by a shift from a generalized hunting economy to one exploiting the potential of local resources. In the Fraser Delta this middle period is known as the St. Mungo phase. During this phase many aspects of the Northwest Coast subsistence base, including salmon fishing, sea-mammal hunting and shell-fish gathering, are added to the former reliance on land mammals. The St. Mungo phase has a range of new bone and antler artifact forms such as harpoon points and simple bone pendants (Matson 1981). Although wedges and chisels of this period suggest a well-developed woodworking technology there is presently no evidence for large houses or permanent villages. Although only two anthropomorphic or zoomorphic decorated objects have been found in St. Mungo phase sites, several incised slate and antler objects attest to an interest in incised and pierced decoration. Similarly aged sites on the Gulf Islands assigned to the Mayne phase reveal a rich artistic tradition. Similarities between the Mayne phase on the Gulf Islands," the Eayam phase south of the Fraser Canyon, and the St. Mungo phase led Borden to unite the three phases into the Charles Culture (Borden 1975).  Decorated  artifacts also appear in similarly aged prehistoric deposits of the central and northern coast. Period III at Prince Rupert Harbour is also considered broadly similar to the St. Mungo phase.  58  St. Mungo phase sites in the Fraser Delta  St. Mungo Cannery The St. Mungo Cannery site, DgRr 2, is situated on the south arm of the Fraser River 20 kilometres from the river mouth. When the site was first occupied, prior to the formation of the Fraser Delta, it was on the westernmost point of land. It is currently within the territory of the Kwantlen Halkomelem-speaking Coast Salish. The site contains three temporal units, the oldest of which is a St. Mungo phase component containing the one of the oldest representational image from the Pacific Northwest.  Figure 5 shows the small  fragment of coarse-grained stone from a component dated between 3300-4480 B.P. (Ham 1985:114). The edges of the flat stone have been abraded to form a head and body of what was likely once the full outline of an animal. The eye is a drilled hole while the other features have been incised. The mouth and lines running down the chin from the mouth have been deeply incised as has a curved line which sets off the head from the body. Although a face has been incised on both sides of the object, one side received more attention and features a deeply incised eyebrow and faintly marked crescents on the cheek. The workmanship is rough and uneven but some attempt has been made to make the lines smoothly curving and deep. The animal's identity is difficult to discern although the definition of the head suggests a mammal, salamander or lizard. A cautious guess is that the crescent-shaped lines on the cheeks and the defined eyebrow could refer to the costal grooves and prominent parotoid gland of several common salamander species (Carl 1943). If this is the case then this artifact represents the earliest depiction of this creature which is found in prehistoric art of later periods, Coast Salish art, and Salish mythology. The second possibility is that this image is not strictly representational but is stylized in a manner similar to examples in Northwest Coast art. In this case the curved line above the eye may represent an eyebrow  Figure 5. St. Mungo zoomorphic stone fragment.  60  and the crescents on the cheek may be decorative elements. This interpretation would indicate great antiquity for two common artistic conventions of ethnographic Northwest Coast art-giving animals human-like features such as eyebrows and defining the cheek and mouth margin with crescent shapes (as in Holm 1987:Fig. 40). Although these suggestions are intriguing, confirmation of these interpretations will have to wait for further examples of zoomorphic art from this period to be uncovered. Ham (1985:121) cites this carving as evidence of a complex society with art and religious artifacts during the St. Mungo phase. But the assemblage dated to this phase-representing a temporary fishing site-is too small to support these assertions. An undated carving found at the site is shown in Figure 10 and is discussed later in this section.  Glenrose Cannery One kilometre upstream from the St. Mungo Cannery site is the Glenrose Cannery site, DgRr 6, where another decorated object of great antiquity was discovered. Figure 6 shows a human image carved from elk antler found in the lower levels of the St. Mungo Component which is dated from 4300 to 3300 B.P. (Matson et al. 1976:183). Holes at the back of the carving indicate that it might have had been worn as a pendant, but a deep channel at the bottom of the figure also indicates this object's use as a haft for a chisel. Elk antler is common in Fraser Delta assemblages from the Old Cordilleran cultural phase dating to approximately 8000 B.P. to the Marpole phase dating to 1100 B.P. This artifact is the earliest antler sculpture found to date in British Columbia. Although this little figure is three-dimensional, its features have been defined by engraving, incising and sawing which are two-dimensional carving techniques. The effect created by these techniques is one of sharply defined features set off by abrupt changes in the surface plane. The eyes and mouth are engraved ovals with a faint incised line defining the lips. The margins around the eyebrows and nose have been cut away to form a raised surface above the plane of the face  61  Figure 6. Glenrose antler human figurine.  62  and set off from the hairline.  A top-knot rising from the crown (a hair style worn by  Northwest Coast men [Barnett 1955:73]) deep cheek furrows, and a prominent chin or beard suggest that this is a male figure.  If a beaver-tooth was used as a blade for this chisel, the  tooth would curve up from the figure's torso similar to a penis. If this was the case it would confirm the figure's masculine identity. Flexed arms held on the chest have been formed by carving an outline shape then incising the definition of hands and fingers. What is most impressive about this piece is its subtle symmetry achieved in the repeating V-shapes in the eyebrows, hairline, beard, and arms. It is a gracefully composed and carefully executed composition that suggests that the carver was familiar with antler as an artistic medium. At first glance this carving might appear foreign to those familiar with Northwest Coast sculpture, but upon closer examination there are many points of similarity.  The  continuous raised surface forming the eyebrows and the nose can be found in Salish art, and the three defined planes of the face marked by deep engraving are also reminiscent of Coast Salish block engraving. As well, the equal proportions given to the figure's head and trunk are common to carvings of the coast. Individually, no one of these compositional features is unique to the Northwest Coast, for example, the raised and joined nose and forehead can be found in art of the Western arctic. But on the whole, the presentation of the features, the pose, and the proportions are well within the familiar realm of Northwest Coast attributes. It has been suggested that the figure's slanted eyes suggest an Asian or Eskimo-Aleut racial identity, but humans are not depicted with slanted eyes in art from these other regions.  63  Gulf Island Sites  Pender Canal The Pender Canal site, DeRt 2, is a large shell midden and former Straits Salish Saanich village on Pender Island. This site was first tested by archaeologists in 1957-8 (field notes by Duff, Kew and Sendey, Royal British Columbia Museum) during which a zoomorphic antler spoon and fragment of a carved bird were uncovered. Recent Simon Fraser University excavations from 1984-86, under Carlson, have uncovered several more carvings and added more precise site information to our understanding of the earlier discoveries. The results of the recent excavations have not been published but early reports confirm that this site will contribute significantly to our understanding of prehistoric art and burial practices. All of the decorated items come from a section of the site designated "Mound 1" containing over 160 burials dating from 5000 to 3000 years old (Carlson 1987:8). Carlson has suggested that the midden matrix may have been deposited for burial purposes (Carlson pers. comm. 1988). At least seven anthropomorphic and zoomorphic antler spoons and several small sculptures have been uncovered in Mound 1.  Two of the spoons are  associated with radiocarbon dates of approximately 3600 B.P. (Carlson 1986), and the others are thought to date to between 3200-3700 B.P. (pers. comm.). These finds may belong to either the Locarno Beach phase or the earlier Mayne phase. They have been included in this section since the two securely dated spoons are from burials dating to the Mayne phase. The spoons exhibit similar motifs, principles of composition, and carving techniques and are found associated with human burials. Although the carvings from the most recent Simon Fraser University excavation were not available for this study, an examination of the two earlier finds will illustrate some of the features common to sculptures from this site. The scoop-shaped elk antler spoon in Figure 7, and the owl figure in Figure 8 were uncovered in the Mound 1 section of the Pender Canal site in 1958 along with the skeleton of an adult  65  Figure 8. Pender Canal antler owl figure.  66  male (1958 accession notes). Although the carvings in Figure 7 and 8 have not been dated, their position in Mound 1 and stylistic similarity to the dated spoons strongly suggests a similar age. The carving on the handle of the spoon in Figure 7 depicts an animal with short legs and a broad, flat head. The base of the legs have broken off the spoon handle and the hind portion of the animal's body is missing. The animal's features have been defined by both sculptural modelling and engraving. The shape of the head, snout and eyes are gently contoured, while the mouth is an incised line.  The division between the head and body is marked by an abrupt  constriction. Three pairs of ribs are deeply engraved and join a spinal column formed by two raised ridges with a deeply engraved channel in between. Raised ridges also define four engraved circles which likely mark the limb joints. The legs are two-dimensional and appear to have been formed by cutting out sections of the handle. The proportions, bulbous snout, lack of ears, and accentuated rib-lines suggest that this figure represents a salamander which is a common species to the Gulf Islands (Carl 1943). Associated with the spoon is a fragment of an unknown antler object depicting an owl shown in Figure 8. The contours of the body have been delicately modelled to form a beak, ear tufts, feather facial disc, and tail. A sharply defined neck constriction accentuates the face and shoulders of the bird. Barely visible incised lines on the facial disc and tail are likely meant to describe feathers. The large ear tufts suggest that this sculpture represents a Great Horned Owl which is common throughout British Columbia.  When the two  sculptures are placed side by side they appear stylistically distinct~the salamander with its stylized, deeply engraved ornamentation versus the owl's more naturalistic sculpturing. Comparable stylistic features include a mixture of modelling, incising and engraving techniques, and a sharply constricted neck setting off the head from the body. In comparing these two sculptures with recent discoveries from the same site, some exciting similarities are evident. The raised ridges that define the leg joints and ribs of the salamander occur in several of the newly uncovered antler sculptures. The raised circular  67  ridges also define eyes and nostrils in the recently found carvings. Round and oval discs carved out of soapstone, shell, and stone have been found in association with some of these recent sculptures suggesting their use as inlays fitting into these raised areas.  The  combination of sculptural modelling of the face and body with stylized, deeply engraved features depicting eyes, nostrils, ribs, and spinal columns also characterize several of the newly acquired sculptures. Two other horned owls are depicted in these sculptures as well as several humans and birds.  Several sculptures depict two or three figures: smaller  animals protruding out of the mouth of larger mammals, and human figures with birds on top of their heads. The bird and fish figures are all attached to human or mammal figures so it is possible that the owl in Figure 8 was once part of a larger multi-figure sculpture.  Helen Point Related to the Pender Island finds is a sculpture from the Helen Point site, DfRu 8 on nearby Mayne Island. The site is on the territory of the Tsartlip Saanich Coast Salish. The small antler carving of a human, shown in Figure 9 was uncovered by B.C. Provincial Museum archaeologists in undated deposits in 1966 (unpub. A.S.A.B. report).  McMurdo  (1970) and Carlson (1970) conducted more extensive work at the site and proposed three prehistoric cultural occupations of the site. The oldest component Helen Point I, is thought to date from 5000 to 3000 years old. It is subdivided into Helen Point la dating to the Mayne phase and lb with Mayne and Locarno Beach phase items. The sculpture in Figure 9 was found in deposits corresponding to Helen Point lb (McMurdo 1970:147, unpub. A.S.A.B. report).  Similar to the sculptures from the Pender Canal site, the Helen Point antler  figurines may date to either the Mayne or Locarno Beach phase. The carving style and form of this figure closely relate to the carvings from Pender Canal. There is the definition of features by abrupt raised and lowered planes, the deeply engraved holes for eyes, ears, earspools and hip joints which may have held inlays, and the  Figure 9. Helen Point antler human figurine.  69  sudden constriction of the neck. The posture, earspool hole and carving style closely mirrors one of the recent but unpublished finds in the SFU dig at Pender Canal. Although a portion of the figure has split away, indentations in the back indicate that a spinal column was depicted with notched lines. As well, a missing section has been broken from the top of the head. The tightly flexed posture, and exaggerated clavicle suggest a skeleton or emaciated figure. An undated object from the St. Mungo Cannery site, shown in Figure 10, exhibits several characteristics common to the Gulf Island sculptures. Raised ridges circle the deeply engraved eyes of a human-like face and suggest that inlays once rested in the eye sockets. The features are defined by deep engraving lifting the mouth margin, eyes, and joined nose and forehead from the plane of the face. At the back of the head projects a small unmodified antler tine which Gay Calvert (1970:66) suggests represents the dorsal fin of a whale. Although not found in situ I believe the style of this carving suggests that it belongs to the St. Mungo component of the site.  Central and north coast sites  An antler haft with a human effigy found at a Prince Rupert Harbour site is illustrated in Figure 11a. A lack of datable material prevents precise dating of the site's components, but this deer antler chisel is estimated to date to roughly 3500 B.P. (MacDonald 1983:103). The Prince Rupert carving is too worn and roughly executed to allow much comparison with the Glenrose site example. Engraved slanted eye margins and a simple line for the mouth can be discerned. It is interesting to note that in such a small sample of early prehistoric art there are two antler hafts depicting human figures. A single recovery of a whale bone pendant from the Namu site, ElSx 1, near Bella Bella completes our knowledge of the early art record from the central coast area. Estimated  Figure 10. St. Mungo antler anthropomorphic carving.  Figure 11. Central and north coast art. a) Prince Rupert anthropomorphic antler haft; b) Namu antler fish pendant (Carlson 1983:Figure 7:1).  72  to date to approximately 3500 B.P. (Carlson 1983:121, Figure 2:1), the undecorated pendant, shown in Figure l i b , has the outline shape of a fish. Although Carlson hypothesizes that the pendant form may represent the spirit helper of its owner, Burley (1980:67) has pointed out several other possible interpretations offish effigies (see Chapter 2).  Summary: art from 4500 to 3300 years old  Table I summarizes the stylistic attributes found in the small sample of art dating from approximately 4500 to 3300 years old.  There are clear similarities between the  carvings from the Pender Canal site and the Helen Point site, in particular the engraved circular joint markings in Figures 7 and 9. The antler carvings illustrated in Figures 6, 7, and 9 show a general similarity in the techniques of carving and delineating representational forms including a sharp delineation of features and abruptly stepped surface planes. Although the round antler tines appear to make these carvings three-dimensional, the carving techniques of engraving and incising are two-dimensional. This two-dimensional approach to carving in the round is also characteristic of developed Northwest Coast art, and may be partially attributed to the reductive nature of antler and wood carving. Aside from the general similarities mentioned above there are no distinctive attributes that link the carvings from the Fraser Delta and the Gulf Islands. Taking the sample size into account this should not be surprising. The significant aspect of the prehistoric art record of this period is the many parallels with Northwest Coast art. The engraving in Figures 7 and 9, and in the recent finds from the Pender Island site is reminiscent of the deep engraving in Coast Salish spindle whorls. The "salamander" spoon in Figure 7 bears a remarkable resemblance to horn spoons from Columbia River Chinook groups (for an example see Holm 1987:29). Other links  Table I. Prehistoric Style Attributes, 4500-3300 B.P.  i DECORATED FORMS harpoon points human face pendants bird form pendants hafts and handles small pestle/pendants "slave killer" clubs whale bone clubs stone bowls stone fish/whale forms atlatl weights browbands scratchers/blanket pins comb stone effigy human figurine mat creaser spoon maul CARVING TECHNIQUES incising engraving block engraving inlaying modelling pecking sawing cut-out antler DESIGN PRINCIPLES Formline Formsurface DESIGN ELEMENTS crescents negative C-shape T-shape U-form split U-form wedges Salish eye chevrons joint ovals/circles MOTIFS human animal  bird fish skull lightning-snake rib and backbone bear wolf/dog feather  n=7:l  -  +  +  + -  + +  -  +  +  +  +  74  with later periods include: the pendant in the shape of a fish from the Namu site which may be an ancestral form to the fish effigies common in later prehistoric periods, and the rib and backbone motif which is a predominant motif throughout the coast in later prehistoric and historic periods. The skeletal features on the human form from the Helen Point are similar to the emaciated features on stone human-figure bowls from the Gulf of Georgia and Lower Fraser regions. The Pender Canal site provides a wealth of decorated material that once analysed will likely complete the evidence to show that a definable style of art with particular motifs, carving techniques and design elements existed in the Gulf of Georgia region from approximately 3600 to 3000 years ago. The evidence presented in this study is only able to suggest that this might be the case.  5.2 ART FROM 3300 TO 2400 B.P.  The next part of the sample under examination is material that is dated from 3300 to 2400 years old from the Locarno Beach prehistoric cultural phase in the Gulf of Georgia and Strait of Juan de Fuca regions, the Baldwin phase in the Lower Fraser River region, the Shuswap horizon in the Mid-Fraser and Thompson River region, and the Middle period in North Coast sites at Prince Rupert Harbour. Although assemblage changes between the preceding St. Mungo phase and the succeeding Locarno Beach phase are well-documented, the dynamics behind these changes are not fully understood. Sites identified as belonging to the Locarno Beach phase occur in the Fraser Delta, Gulf Islands, southern Vancouver Island and the Olympic Peninsula. The same dependence on a mixed marine-oriented economy is suggested by faunal remains from this period (Mitchell 1971:51). Archaeologists think that the changes in this phase reflect technological innovations allowing increased exploitation of marine resources (Mitchell 1971).  Assemblages show high quality workmanship in ground stone items including  75  labrets, ear spools, slate points, and adze blades. New artifact forms also include toggling harpoon valves and mircroblades. Although early researchers hypothesized that ground stone items and toggling harpoons suggested Eskimo-Aleutian influences arriving on the south coast during this period (Borden 1951), later researchers (Mitchell 1971) suggest a continuity between the Locarno Beach phase and the preceding phases of the south coast. Art evidence from this phase is scarce and most comes from sites dated to the very end of the period near the onset of the succeeding Marpole phase (although it should be noted that antler sculpture from the Pender Canal and Helen Point site may belong to the early Locarno Beach phase). Since decorated artifacts from the interior plateau and Lower Fraser pre-date examples from the coast, the discussion begins with the material from the Lower Fraser and Thompson River regions of the Province.  Mid-Thompson River region  Kamloops Indian Reserve A zoomorphic stone maul fragment from the Kamloops Indian Reserve site, EeRb 10, is illustrated in Figure 12. It is associated with dates of 2950 ± 120 and 3000 ± 400 B.P. from an early Shuswap horizon component in a pit house site (Richards and Rousseau 1982). This predates the occurrence of mauls in coastal sites by approximately 1000 years, therefore, the association of the maul with these early dates should be treated with caution. The proximal or handle end of the granitic stone maul has been pecked and sawn to form a downturned mouth and round, bulbous eyes suggestive of a turtle or frog. The maul may be the oldest example of sculptured stone in both the Plateau and Northwest Coast. The simple features on the maul do not provide much stylistic information with which to compare this  76  Figure 12. Kamloops Indian Reserve zoomorphic stone maul.(Richards and Rousseau 1987:Figure 17p).  77  object with other sculpture in stone.  But the find suggests that the Mid-Fraser and  Thompson River region may have a sculptural tradition dating back three thousand years.  The Lower Fraser Region  Milliken The Milliken site, DjRi 3, is situated on the east bank of the Fraser River near Yale. Although the site is more than 150 kilometres from the Pacific Ocean it is within the border of the coastal habitat zone and was occupied by Coast Salish Sto:lo groups. The area remains an important salmon fishing location for the Sto:lo. Milliken was thought to have been used over a 9000 year period as a late-summer fishing site.  Unfortunately this  important site remains only partially reported. Nine zoomorphic stone objects are associated with the Baldwin phase at Milliken dating from 2800 to 2360 B.P. (Borden 1975:62). A siltstone fish effigy, shown in Figure 13a, has been shaped by sawing a notched body outline and mouth. The eye has been drilled and on one side a line has been sawn that extends from the eye halfway down the body. The head shape with the prominent lower jaw suggests the depiction of a migrating salmon. Figures 13b-d illustrate three of out five small effigies from the same component made of a similar soft sedimentary stone. Crude outline shapes have been ground and sawn and simple features have been added by drilling and incising. Figure 13b shows a carved stone fragment of an unidentified animal. Two parallel lines run along the midline of the body and short perpendicular lines on the side likely represent a rib and backbone motif. The zoomorphic features on Figure 13c are barely discernable, but there is a simple drilled eye, and the rudimentary sawn form of a neck, mouth and forelimb. Figure 13d with the  78  Figure 13. Milliken stone zoomorphic figures, a, siltstone fish; b-d, zoomorphic effigies.  79  crude outline of a head, body and forelimb, also features two parallel lines running along the mid-line of the body. Figures 13b and 13d each exhibit an anomalous line behind the eye. A single sculpture in the round found in the lowest depths of the Baldwin component is illustrated in Figure 14. This unidentifiable animal also features an incised, sectioned backbone and rib markings. A gaping mouth, neck, and the stumps of forelimbs have been sawn, while circular eyes have been incised. This small sculpture, together with the stone maul from the Kamloops site, confirms that crude representational sculptures in stone date to approximately 3000 B.P. Borden placed two other sculptures from the Milliken site within the Baldwin phase, but the provenience of both objects is problematic. While the examples above were found deep within the Baldwin component, the next two where found within the topsoil stratum which in some portions of the site contained burials from a later period. Figure 15a is a soapstone effigy which likely represents the head of a mammal. Its features have been sawn, drilled and engraved and include ears, nostrils and teeth.  Borden made much of this  sculpture (1983:139) claiming that the sharp definition of the snout and forehead represented a muzzled bear used in bear ceremonialism. No doubt Borden was looking for confirmation of the circumpolar diffusion of bear ritualism which was a popular theory in the first half of this century (Boas 1933). Of interest is the fact that the carvers tool markings are clearly visible. Although some of the shaping has been done by sawing and incising it has also been carved or whittled down~a technique that is only possible with wood, antler, or soapstone.  Unfortunately, although this object could easily fit within the Baldwin  assemblage and would represent the first example of zoomorphic soapstone carving, we cannot place it with certainty in this phase. Figure 15b also comes from the topsoil stratum. It likely represents a human skull with the triangular nasal hole as a distinctive feature. Not all of the characteristics fit this description: the eyes are lenticular instead of hollowed, a notched lip margin, cheek markings, a heavy brow, and bulbed crown are indicated. I agree with Borden's suggestion  Figure 14. Milliken steatite zoomorphic sculpture-  Figure 15. Milliken stone head effigies, a, steatite bear head; b, human skull.  82  (1983:139) that this image might represent a mask since the back of the head is hollowed out, but it cannot definitely be placed within the Baldwin assemblage. Almost all of the sculptures that have been examined from the Baldwin component of the Milliken site are roughly conceived and executed images that could have been completed in an hour or less. Nevertheless, there are stylistic similarities among them that suggest a common use. Five of the seven found deep within the site depict a spinal column and ribs, and all are palm-size or less. A purely speculative interpretation is that these are "good luck" charms made by local inhabitants coming to the locality to fish. The images may depict the fish and other animals such as birds, seals and bears that would congregate in the area during the spawning season. The rough execution and number of images found in the site may indicate that they were made and discarded within a short period of time. The depiction of the backbone and ribs could indicate that special handling of the skeletal remains of prey species~as documented in ethnographic accounts of hunting and fishing rituals-was a feature of the Baldwin cultural period. These interpretations must wait for further discovery of similarly aged components along the Lower Fraser River.  Gulf of Georgia Locarno Beach phase sites  Locarno Beach phase sites are dated from approximately 3300 to 2200 B.P. These dates overlap with cultural phases that pre-date and post-date Locarno Beach. It is possible that some of the material discussed in the previous section under art from 4500 to 3300 years old belongs to the Locarno Beach phase since the boundary between the Mayne and Locarno Beach phase components on the Gulf Islands is not clear. The present sample contains material from late Locarno Beach phase assemblages dating after 2750 B.P.  83  Musqueam Northeast Musqueam Northeast, DhRt 4, situated at the mouth of the Fraser River facing into Georgia Strait is a midden site with Locarno Beach and Marpole phase components. The site is within the boundaries of the Halkomelem Coast Salish village of Musqueam. The Locarno Beach phase component yielded dates of 2970 ± 85 and 2550 ± 85 B.P. (Borden and Archer 1975). Unless otherwise noted all of the following material is associated with the Zone A2 date of 2550 B.P. Portions of the site were waterlogged and contained basketry and fibre objects but no decorated wooden artifacts were found. The preservation of bone and antler artifacts was good, including a remarkable sample of decorated objects. Figure 16a is a fragment of flat bone with a raised design suggesting the two limbs of an animal. The scoring of sharp blade marks is visible where several millimetres of the surface have been carved away to achieve the raised design. This technique called block engraving (Kew 1979) is the most common carving technique and principle of design in twodimensional Coast Salish art.  The Coast Salish spindle whorl in Figure 16b shows the  similarity between prehistoric and historic examples of this style. Figure 17a shows another small fragment of a block engraved composition in bone or antler. Three sets of raised, positive lines are similar in shape to feather motifs found on Coast Salish spindle-whorls and rattles. A Coast Salish example of this motif is shown in Figure 17b. Whereas art from the previous period exhibited a deeply engraved style that had an affinity to Northwest Coast art, the two fragments from Musqueam Northeast can be considered true examples of twodimensional block engraving. The antler spoons shown in Figure 18 and 19 are part of a cache of four spoons from Musqueam Northeast that are identical in shape to the scoop-shaped spoons from the Pender Canal site mentioned in the previous section. Although they are associated with the 2550 B.P. date for Zone A2, their position near the bottom of the zone led Borden (Borden and Archer 1975) to suggest that they might be closer to 2800 years old. Two of the spoons illustrated in Figures 18 and 19 showed representational decoration. The features in Figure  84  Figure 16. Musqueam Northeast block engraved design. a, fragment of antler block engraving; b, Coast Salish spindle whorl showing similar block engraving style.  b  85  Figure 17. Musqueam Northeast feather motif. a, Musqueam Northeast antler fragment with feather motif; b, Coast Salish spindle whorl showing feather motif.  Figure 18. Musqueam Northeast anthropomorphic antler spoon.  View A  View B  Figure 19. Musqueam Northeast zoomorphic antler spoon.  88  18 can be viewed in two ways. From the front (view A) the handle is carved into the form of a human face surmounted by an eared hat. Folded limbs are visible below the face. When viewed from the side (view B, the handle depicts an animal's wide mouth, eyes and ears. Because the spoon exhibited an unusual mixture of carving styles it was closely examined with the conclusion that the dorsal surface of the carving may have been partially obliterated (chewed by a mouse) and later crudely recarved. The most interesting feature of this carving is the negative T-shaped element defining the inside of each ear. This is the earliest prehistoric example of this common design device. In Northwest Coast art a T-shape is an elegant delineation of negative space formed at the juncture of three curving formlines. The use of this shape usually implies the use of the formline or form surface principle of design, but if we examine Figure 18 this principle of design is not in evidence outside of the ears at the tip of the spoon. This restricted use of a T-shape in an isolated part of an object~the rest of which exhibits another carving style-is found in other examples of Coast Salish art (Holm 1983:Figure 6:12). A similar mixture of styles can also be seen in the second zoomorphic spoon shown in Figure 19. It depicts a similar side profile of an animal with erect ears, but the design is conceived and executed with much greater sophistication than its mate. The lenticular eye, grinning mouth and wide eyebrow are defined with raised ridges that are close to the formline concept. The animal's profile has been delicately modelled and contrived so that the extension of the scoop of the spoon could represent the animal's lower jaw or tongue. The animal's ears are defined with U-forms: a negative U-shape delineates the interior of the ear while a raised positive U-shape forms the outside edge. On the back a slight constriction occurs at the neck and a curved margin marks the end of the zoomorphic design. An elongated wedge or T-shape is engraved near this margin but doesn't appear to relate to any other part of the design. When the composition is viewed as a whole, although formlines emerge around the ears and on the face the positive and negative elements of the two sections are not integrated. The erect ears, wide mouth and profile view of the zoomorphic  89  creature suggest the portrayal of a wolf, but the facial features also have the sinuous, serpentine expression of a snake. Perhaps, as in the previous spoon, this carving hints at two identities: the wolf and the eared or plumed lightning-snake. Figure 20 shows a spoon from Vancouver Island (with no site provenience) that is composed and executed in a similar style to the Musqueam example discussed above. Two other objects from the Locarno Beach component at Musqueam Northeast include a siltstone effigy of an animal head with the simple, incised features of a lenticular eye and mouth (not illustrated); and a fragment of a bone shaft ornamented at the top with a small head. Figure 21a illustrates the tiny carving. The top of the head is either deeply lobed or is meant to depict ears. The hollow eyes and triangular nose hole implies a human skull, but the triangle and notched head outline also suggests the beak and eared face of an owl. The dual identity of this carving is lent support by the Halkomelem word lexeme used for corpse, soul and owl.  Locarno Beach Another carving of a skull came from the Locarno Beach site, DhRt 6. The type site for this phase is represented by a small assemblage with late Locarno Beach phase radiocarbon dates of 2430 + 160 and 2270 ± 100 B.P (Borden 1970). (These dates were obtained when radiocarbon dating technology was first developed and may be inaccurate. The assemblage appears to be fully Locarno Beach in character [Matson pers. comm.].) Figure 21b shows how the natural shape of a deer metapodial bone has been used to represent the cranium of a human skull. The nasal hole is an engraved triangle, while the eyes and ears are drilled hollows. A mouth and teeth are incised, while the lower shape of the jaw has been carved. Even though the deer bone has much of the three-dimensional shape of a skull, a close examination of the finished carving shows how much effort was required to shape and sand the surface to achieve an even, smooth surface.  Figure 20. Undated zoomorphic spoon, Vancouver Island.  Figure 21. Locarno Beach carvings. a, antler owl/skull pin or scratcher; b, bone skull effigy with unmodified bone fragment from site.  92  Figure 22a shows a second anthropomorphic carving from the Locarno Beach site depicting a person wearing a hat. The antler carving is poorly preserved but a human face with deeply engraved eye margins, a small mouth and pointed chin are portrayed. This carvings shares several features with the antler spoon in Figure 18. In both carvings the eye margins are deeply engraved forming bulging eyeballs, and the nostrils are wide and flaring. Lastly, both wear small conical hats and are identical in size. Borden suggested that the figure was portrayed with a labret protruding from the lower lip, but I believe he incorrectly identified the carved out area under the nose as the mouth, and the small oval mouth as the labret. Borden uncovered two puzzling bone objects from the site which he has labelled "whale's tail" motifs, shown in Figure 22b-c. They do have a resemblance to the notched tails characteristic of minke, humpback and killer whales but the incurved edges of the flukes are anomalous. They are similar to the tail of a plumed serpent portrayed on a wooden atlatl (Figure 65), but whether they are zoomorphic depictions remains uncertain. The Locarno Beach site carvings do not show any of the "modern" Northwest Coast artistic traits found at Musqueam Northeast.  It is possible that the Locarno Beach site represents an older  assemblage predating the stylistic attributes found at the Musqueam N.E. site. The last object from a Locarno Beach phase assemblage in the Fraser Delta is a carving from the Crescent Beach site, DgRr 1. Figure 23 shows an antler image of a longbilled bird's head from the upper levels of a shell midden component dated from 3150 to 2350 B.P. (Percy 1975). The carving has several distinctive features. Firstly, the top of the head and the top edge of the bill have diagonally notched borders similar to the notched lines on the head of an antler animal figure from the Pender Canal site (Carlson 1986). The margin of the bird's mouth is also marked by an engraved line and notching but in this case the notching may refer to the saw-like bill of the merganser or "sawbill." A second similarity to the Pender Canal material is the hollowed-out pupil that may once have held an inlay. Lastly, the back of the head is hollowed in a similar way to several Pender Canal sculptures  93  Figure 22. Locarno Beach carvings. a, antler human head (drawing of artifact on right); b-c, antler fragments with possible whale tail motifs.  Figure 23. Crescent Beach antler bird head carving.  95  which Carlson (1983:203) hypothesizes may represent the portrayal of masks. Portions of the side and back of the bird's head are missing so it is not possible to fully assess whether this sculpture represents a bird mask, but the diagonal notched lines along the bill and crown are reminiscent of the cedar bark rope used to decorate several types of Kwakwaka'wakw masks. An example of this technique of decorating masks with twined rope is shown in Figure 24. Unlike the Pender Canal sculptures, the Crescent Beach carving has more complex features. The eye has a double incised line defining the eye margin, an angled T-shaped engraved space defining the eyeball and another raised ridge defining the perimeter of the pupil. It is the most complexly conceived eye form found so far in the prehistoric record, and marks the first full Northwest Coast style eye with a sharply pointed eye lid, and an eyeball plane which is angled and recessed from the pupil. It is tempting to view this carving as a more highly evolved example of the same style of carving that is seen at the Pender Canal site. Decorated objects from the Pender Canal and Helen Point sites that date to the Locarno Beach phase or likely slightly earlier were discussed in the previous section. A small fragment of carved antler from the Helen Point site is worth inclusion in this discussion. The fragment comes from the upper levels a the Mayne/Locarno Beach phase component dated circa 3000 B.P. (J. McMurdo 1974). engraved onto the flat antler.  A curved line has been deeply  This line has a V-shaped cross-section, with one almost  perpendicular side and another sloping side. It is a common convention of Northwest coast carving and its occurrence at Helen Point is the earliest record of this stylistic attribute.  Figure 24. Northwest Coast mask showing rope edging. U.B.C. Museum of Anthropology collection.  97  The Strait of Juan de Fuca  Hoko River At the Hoko River site, 45CA213, at the northwestern tip of the Olympic Peninsula in Washington State, a rare decorated wooden artifact was preserved. This wet site contains the mud-embedded refuse from a fishing camp dated from 2750 to 2210 B.P. (Croes and Blinman 1980). Figure 25 shows the profile of two crested birds which face each other and touch beaks to form the handle of a mat creaser. The mat creaser is associated with the lower layers of the site thought to date to 2750 B.P. Lenticular pupils and eye margins are deeply engraved, as are lines delineating feathers on the birds' crowns and wings. Although a simple composition, the carving style is block engraving. Joined double-bird and doubleanimal motifs are common to Coast Salish and West Coast mat creasers and combs although the bird heads face away from each other, instead of towards each other as in the Hoko River example. Usually the identity of the birds on mat creasers are ambiguous, but in this case the sharp beaks and crests suggest kingfishers or pileated woodpeckers. Since kingfishers are associated with fishing prowess and pileated woodpeckers nest in cedar trees, both birds would be appropriate symbols for the site activities of fishing and cedar-bark processing. A black staining from the use of lignite paint is visible on the crests and eyes of the birds (Croes pers. comm.).  This is the first prehistoric evidence of the use of paint on the  Northwest Coast. Figures 26a and b illustrate examples of decorated barbed, wooden projectile points found at the site. The unilaterally barbed point in Figure 26a has a deeply notched central ridge that suggests a backbone, and projecting barbs that are incised with lines similar to  99 Figure 26. Hoko River zoomorphic implements. a, barbed wooden point with backbone motif; b, wooden zoomorphic implement.  a  b  100  fins. Below the barbs are engraved chevrons that encircle the shaft. These decorative devices would appear to both enable line attachments to be secured to the points and suggest the ribs, backbone and fins of fish. Figure 26b is a wooden point of unknown use, or possibly a fish lure (Croes 1976:229). The tapering point features a head at one end with an incised mouth and lenticular eye. Behind the head are groups of incised crescents encircling the square shaft. The squared edges of the shaft are notched in the same area as the crescent designs occur. A hole in the shaft indicates that the wooden device was suspended. The lack of limbs or fins, and the banded markings suggest the depiction of a snake. These preserved wooden artifacts allow a glimpse of the variety of utilitarian items that were given decorative embellishment. One of the site's main researchers, Dale Croes, has compared the basketry from the Hoko site to other wet sites in the Pacific Northwest.  He finds that the Hoko site shares  more stylistic attributes with the Ozette site, which is geographically close but 2000 years younger, than it does with the Musqueam Northeast site with which it shares a Locarno Beach phase component. Croes (1985) proposes that distinct basketry styles in Wakashan and Salishan territories may have been in place in the Locarno Beach period indicating distinct ethnic groups. He asserts that the "cultural phase" or "culture type" such as Locarno Beach, should be viewed as a general economic strategy shared throughout a region rather than a cultural type. Unfortunately there is not an adequate sample of decorated artifacts from this site that would allow a detailed comparison with art styles from Locarno Beach phase sites in the Fraser Delta, but both the Musqueam Northeast and Hoko River sites share the block engraving style of carving.  101  North Coast sites  Three decorated objects from the Boardwalk site GbTo 31, complete the sample of material dating from 3300 to 2400 years old. The site is a former Tsimshian winter village with a component dating to the Middle Period. A small schist pendant depicting a bird, illustrated in Figure 27a, is thought to date to circa 3000 B.P. (MacDonald 1983:114). The small sculpture's form has been carved in low relief but since only the replica was examined the carving technique could not be discerned. Raised eye margins encircle channeled holes which may have been inlaid, and engraved crescents decorate the underside of the bird. Figure 27b and c illustrate two stone concretions which have been decorated with rib and backbone designs, and are thought to be approximately 3000 to 2500 years old. Both forms are long, slender and flat with slight indentations indicating heads.  One stone  features three incised lines at the "neck", two parallel lines running along the midline which likely depicts a backbone, and rough cross-hatching all over the surface which may represent fish scales. The second stone features a deeply notched dorsal edge, a tail section, and simple incised eyes and midline. MacDonald believes that both of these forms functioned as fish charms associated with salmon fishing (1983:114). Their forms are similar to fishshaped concretions found in later periods throughout the coast, but their forms also suggest phalluses which are found in northern stone art of the late prehistoric period.  Summary  The sample from the period dating from approximately 3300 to 2400 years old includes twenty-seven stone, antler and wooden objects. In comparing this sample with the older sample of eight objects dating from approximately 4500 to 3300 years old some continuities can be found. The decorated scoop-shaped spoon form found at the Pender  Figure 27. Prince Rupert Harbour stone effigies. a, schist bird pendant (MacDonald 1983:Figure 6:26); b-c, ribbed stone phallus effigies.  103  Island site is found in a cache of four spoons from the Musqueam Northeast site. Except for an unidentified spoon from Vancouver Island, these are the only known examples of this artifact form. More importantly, several of the stylistic traits found in the Pender Canal and Helen Point material can be found in artifacts from the later period. The raised notched ridges on the snout and crown of sculptures from Pender Canal described by Carlson (1986) appear to be similar to the notched ridges found on the bird's head from the Crescent Beach site. The deeply engraved circles and ovals surrounded by raised ridges found in sculptures from Helen Point and Pender Canal are found on the bird head from the Crescent Beach site. These similarities could be attributed to the inlaying technique, they may be characteristic of late Mayne/St. Mungo phase-early Locarno Beach phase components, or they may be associated with sites in Straits Salish territory. Future excavations from components within the 3500 to 3000 time period may answer this question. Where the record of art from the Mayne/St. Mungo phase and Locarno Beach phase sites differ is in their connections to historic art of the Northwest Coast. While the earlier sample had stylistic elements that were described as having a resemblance or affinity to some aspects of Coast Salish art, the later sample contains artifacts with direct iconographic, design and compositional resemblances to Northwest Coast art. A chart summarizing the stylistic attributes in art of this period is presented in Table II. Carving techniques that are new to this period include engraved lines that in cross-section have angled depressions, and two-dimensional block engraving where large sections of a design field are cut forming a raised, positive design. Design elements that are new to this period are the U-form and Tshapes.  Representational features such as eyes, feathers, noses, and mouths that were  depicted with simple, single elements in the previous period are now often compound features: a mouth has lips and teeth, an eye has an iris, eyelid and eyesocket. There is a single example from the Crescent Beach site of a full Northwest Coast style eye with an iris and pupil, angled eyeball, and pointed eyelid.  104  Table II. Prehistoric Style Attributes, 3300-2400 B.P.  ^ n=27: 1 DECORATED FORMS harpoon points human face pendants bird form pendants hafts and handles small pestle/pendants "slave killer" clubs whale bone clubs stone bowls stonefish/whaleforms atlatl weights browbands blanket pins comb stone effigy human figurine mat creaser spoon maul CARVING TECHNIQUES incising engraving block engraving inlaying modelling pecking + sawing cut-out antler DESIGN PRINCIPLES Formline Formsurface DESIGN ELEMENTS crescents negative C-shape T-shape U-form split U-form wedges Salish eye chevrons joint ovals/circles MOTLFS human animal + bird fish skull lightning-snake nib and backbone bear wolf/dog feather  9  6  4  cf  ^  <^  1  3  3  + . . . .  . . . .  . . . .  . . . .  . -  .  . . .  . . .  . + . . .  . . . .  . . . .  . . . .  . . . .  . . . .  . . . .  . . . .  . . .  . . + .  .  . + . .  . . . .  . . . + .  +  +  . +  +  . . . .  . .  . .  . .  . . -  . . .  . . .  . . +  . . . . .  . . . . .  +  -  +  .  . . . .  + + -• . . . . . . . . .  -  .  +  . . . . .  -  . . . . .  + . + -  -  . . .  ? . . .  -  . + -  .  _ . .  ? + . . .  . .  . .  . .  . . .  . . . . . . . .  . . . . . . . .  -  -  + ? ? . . . .  + ? . . . .  . . . . .  + + + + + . + . . . + + ? . . . . . . . . . . . . . . . . . . . .  -  105  The rib and backbone motif that occurred in the early period continues to be found in the 3300 to 2400 year old period. New to the later period is the depiction of the human skull in three different sites. A new concept being explored in this period is the image with a double-meaning or message: owl/skull, human/animal, and fish/phallus.  At least one  common Northwest Coast decorative form--the double-bird form mat creaser-dates to the Locarno Beach phase. Principles of design and composition that are directly related to historic Northwest Coast art do not occur at the Milliken or Locarno Beach sites. The block engraving style is found at Hoko River and Musqueam Northeast, while only one carving~the zoomorphic spoon from Musqueam Northeast-exhibits a style that is close to the formline concept. The sample from this period is small, therefore, any hypothesis based on this evidence must be tendered with caution. The evidence suggests that several common design elements and compositional concepts used throughout the Northwest Coast such as the U-form and Tshape, the interconnection of positive elements in a composition, and the formline concept could have evolved out of an engraving style without the stimulus of painting. These design elements and concepts are interrelated and difficult to separate since they are all based on balancing the negative space carved out of a design and the positive form surface that remains. The available evidence suggests that these Northwest Coast style concepts appear in the prehistoric record by 2500 B.P. The few examples of carving in three dimensions come from the Milliken site which occurs on the border of coastal and interior plateau cultures and from the Kamloops Indian reserve site. The two examples from the Milliken site have been roughly sawn to form rounded outline shapes but many of the features have been incised, only the Kamloops maul can be considered fully three-dimensional. The artifacts from these sites are clearly different from the decorated coastal artifacts from this period. Preservation of bone and antler from the Milliken site is poor so a comparison of this assemblage with the coast does not have much significance, but organic preservation at the Kamloops Reserve was very good and  106  reveals that only simple, geometric designs were found on bone and antler objects. Lastly, there is no firm evidence of the use of soapstone for sculpturing during the 3300 to 2400 B.P. period. Archaeologists note that many raw materials were being traded up and down the coast, and from the coast to the interior during the period relevant to this discussion. So it is likely that people in different regions were exposed to styles from other areas. There are clear relationships between the effigies with rib and backbone motifs from the Boardwalk and Milliken sites, and in the carving style seen in artifacts from the Gulf Islands, Fraser Delta, Olympic Peninsula and Boardwalk sites. These relationships suggest that when prehistoric decorated objects are found in the intervening regions between the north and south coast they will show similar stylistic attributes.  107  5.3 ART FROM 2400 TO 1500 B.P.  Marpole phase archaeological sites  The Marpole phase occurs at prehistoric sites in the Gulf of Georgia region dating from approximately 2400 to 1500 years. Sites with Marpole assemblages are concentrated in the geographic area accessible to Fraser River salmon runs (Mitchell 1971). Faunal remains and artifacts relating to the food quest indicate a pattern of resource exploitation of fish, shellfish and land mammals similar to historic Coast Salish patterns. Similarly, abundant heavy-woodworking tools and the remains of habitation features indicate the likelihood that people lived in permanent winter village houses similar to the historic Coast Salish (Mitchell 1971).  Although similarities between historic and prehistoric technology and cultural  patterns are evident there are also points of contrast. Marpole phase burials differ from historic burial practices with interment in middens, numerous grave inclusions including elaborate body ornaments, and cairn or stone-marked burials . Artifacts that are associated with Marpole phase assemblages include finely made ground stone artifacts such as slate knives, nipple-topped mauls and nephrite celts, unilaterally barbed harpoon points, and objects of personal adornment such as pendants, disc beads and labrets. A variety of stone sculptural forms including bowls depicting animal and human forms occur within this phase. As Burley has noted (1980:40) it is unclear over how large an area Marpole phase components can be found, or how this phase relates to prehistoric cultural sequences in other regions of the coast. Burley admits that although the Marpole culture is confined to the Strait of Georgia region the unilaterally barbed antler harpoon point that he has cited as  108  diagnostic of Marpole phase sites occurs both south and north of the core region. Contemporaneous assemblages from as far north as Alaska, in the central coast (Nuxalk and Heiltsuk region), and in particular the Prince Rupert Harbour sites, show affinities with the Marpole phase including status differentiation, large house features, and similar harpoon styles (Burley 1980:40). In the south on the coast of Washington State, the definition of prehistoric cultural sequences has proceeded slowly but Marpole traits have been cited for sites in the Skagit Delta region and as far south as the Columbia River (Mitchell 1971:74). Looking east, because Marpole sites have been found within the ethnographic region occupied by Central Coast Salish groups we might expect to find Marpole phase attributes in sites along the Fraser River up to the inland boundary of the Halkomelem Sto:lo. Looking west to Vancouver Island, McMillan (1980:100) finds affinity between Marpole materials and one assemblage at Shoemaker Bay at the head of Alberni Inlet, although sites on the outer islands of the west coast are considerably different. On the east coast of the island Capes (1977) and Monks (1977) have found Marpole similarities in sites at the northern end of the Straight of Georgia.  The Marpole site The Marpole site, DhRs 1, is situated near the north arm of the Fraser River at its outlet to the sea. The site falls within the territory used by the Musqueam and Kwantlen Salish (Burley 1979:505). The "Great Fraser Midden" once covered an area over 450 meters in length by 100 meters in width (Hill-Tout 1948). Although it is one of the best known sites in the Lower Mainland very little of it has been systematically excavated and it remains poorly documented. The deposits were disturbed late in the last century and sections of it have continued to be destroyed by urban development ever since.  The site was first  described by Hill-Tout (1902) when a road was cut through the midden.  Smith (1903)  uncovered many decorated artifacts along with 75 skeletons. During the years 1927 to 1932  109  Herman Leisk collected extensively at the site for the Vancouver Museum as the area was turned into a residential district (Hill-Tout 1948). The collections derived from all the early Marpole excavations are available to the researcher but there are few references to the spatial or temporal distribution of the artifacts. When Borden conducted several small excavations of the site from 1949 to 1957, only 60 meters of undisturbed midden deposit remained.  Baker (1974) was the last to  conduct excavations at Marpole in 1973 and Burley (1979) prepared the site's first excavation report which included a review of material from previous excavations. Five radiocarbon dates obtained from Borden's samples range from 2350 to 1780 B.P. (Wilmeth 1978) while Burley (1980) obtained the youngest date of 1510 + 90 B.P. Based on stemmed and leafshaped chipped stone points found by Borden, Burley (1979:538) hypothesized that some portions of the site may date to between 3500 and 4500 years old. But the bulk of the artifacts including most of Borden's collection and all of Hill-Tout's and Leisk's collections exhibit Marpole phase attributes. For this reason all of the collections mentioned above have been included in this sample even though the material collected before 1950 lacks radiocarbon dates and proper stratigraphic documentation. Forty-eight decorated artifacts from the Marpole site including 1 shell, 26 antler, 2 whale bone, and 19 stone objects are included in this analysis. As well, many local residents have reported private collections of decorated material from this site. The artistic richness of the Marpole midden is due in large part to the unusual number of burials uncovered at the site. In one month alone in 1932 Leisk excavated 103 adult and 13 juvenile burials (Leisk unpublished fieldnotes).  Examples of adult and juvenile skeletons exhibiting skull  deformation and numerous decorated objects in the graves were reported but they were not systematically recorded. No subsequent analysis has taken place to determine if burial patterns suggest evidence of status differences in the prehistoric Marpole community. Many of the decorated items from this period are abundant new forms not previously found in older sites including brow bands, zoomorphic stone bowls, bowls with  110  human images, and zoomorphic designs on unilaterally barbed harpoon points. The Marpole site collections contain several examples of each type of decorated artifact revealing that specific motifs are associated with particular types of objects. Due to this consistency and the size of the sample the artifacts below are discussed in their functional groups.  Barbed antler points: Six decorated barbed antler points from Marpole are shown in Figure 28. Figures 28a-c and 28e are the broken butt ends of harpoon points used for sea-mammal hunting, while Figure 28f is a antler point which would have been fixed to a leister or spear for salmon fishing. The five harpoon points depict long-beaked or long-mouthed creatures from whose mouths project the barbed shafts. In Figures 28b-d and 28f the eye margins are long and pointed, while in Figures 28a-d rows of horizontal lines extend along the back of the heads behind the cheeks. All the harpoon points, except Figure 28f, are well-made, deeply engraved compositions. Most likely these were functional hunting and fishing implements whose projections and notches served as line guards for the securing of the points to the harpoon shafts. The fact that all of the points have the end barbs broken off makes it plausible that the points were snapped off during harpooning. The similar design elements suggest that one particular creature is being portrayed. A squid, loon, fish or lightning-snake are several possible interpretations. Figure 28f likely represents the image of a lightning-snake with a horn or plume projection on the back of the head, a long tongue, and an elongated eye margin incised on the base of the shaft. Points in Figures 28e and 28f were found with burials (Leisk field notes).  Pendants: Seventeen decorated artifacts are classified as pendants because of their small size and perforation holes allowing suspension. Four of the pendants were found with burials. It is possible that several of these objects had other functions: two as small pestles, and one as  Ill Figure 28. Marpole zoomorphic antler harpoon point fragments, a-c, (Smith 1903:Fig. 52 a-c); d-f, zoomorphic creatures.  112  a blanket pin. Six of the pendants depict the frontal view of a human face (Figure 29a-f). Figures 29a-d are simple incised faces with stylized rather than realistic human features. Example 29e is a more sophisticated composition with modelling of the cheeks and forehead and deeply engraved negative features. The large human head sits on top of a small torso with folded arms and a rib cage. The worn edges of the pendant along with two suspension holes that have worn through indicate that this may once have been a larger carving depicting a complete human form. I find no stylistic evidence to support Borden's claim (1983:147) that this  sculpture bears resemblance to a human portrayed  on a  Kwakwaka'wakw soul-catcher. Figure 29f is made of polished lignite. The round sharply pointed eyelids of the face are identical to an eyelid shape found in Coast Salish wood sculpture (Suttles 1983:Figures 4:10 & 4:13). In this sculpture the sharp definition of the forecheek and forehead is done with incised lines rather than with deep engraving. Figure 30a is a fragment of an unknown object. It has the same eye shape and incised definition of the forehead and cheek as in the previous figure. In this profile view of a human head a form projects from the crown which was likely a bird shape the head of which has snapped off. Figure 30b also shows a pendant with the profile of a human head with a bird's head on top. The motif of a human head surmounted by a bird or bird's head has been reported at the Pender Canal site (Carlson 1986) in antler sculptures dating over three thousand years old and in late sites from Vancouver Island. These Marpole examples may be variations on the same theme. Bird forms are a second motif of Marpole site pendants as shown in Figures 31-32. Examples 31a-d have a similar simple profile of a long-beaked, long-necked bird with a cutout section forming a hole for suspension. Figure 31d might have functioned as a blanket pin or ritual scratcher as well as a pendant. Figure 32a shows a pendant with a the profile of two birds. Its whale bone composition and thickness suggests that it is the top fragment of a club which was reground to form a pendant. If this is the case, this is one of the earliest examples of this artifact form.  Figure 29. Marpole pendants with human face motifs. a, antler; b, shell; c-d, antler; f, lignite (Smith 1903:Figure 58).  Figure 30 Marpole human with bird motif carved fragments, a-b, antler.  Figure 31. Marpole bird form carvings. a-b, antler bird form pendants; c, antler bird form pestle/pendant; d, antler bird form pin/scratcher.  116  Figure 32. Marpole bird form carvings. a, whale bone club fragment (?) with bird heads; b, antler heron pestle/pendant.  b  117  The carving thought to represent a Great Blue Heron shown in Figure 32b, is one of the few examples of three -dimensional antler carving at the Marpole site.  The heron  pendant shows the same carving techniques found in older prehistoric sculptures: a combination of two-dimensional engraving and sculptural modelling.  The heron's beak,  crest, wings, legs and tail are sharply offset from the plane of the body by steep junctures. There is also a soft, naturalistic contouring of the head crest and body shape that gives this sculpture its visual appeal. The finer features of the carving are incised and engraved. The outlines of the feathers are delineated with engraved negative wedge-shapes (see back view), while the feather quills are incised. As in Figure 31c, the base of the sculpture has a rounded bulb which may have been used for grinding. A local resident living next door to Borden's 1957 excavations found this artifact in his yard approximately 30 cm below the surface of the midden (accession notes). The last three pendants shown in Figure 33 share a rib and backbone motif. Figure 33a is an antler pendant depicting an open-mouthed, snake-like animal with a sharply downturned eye. Incised crescents which run along the side of the body may represent ribs or scales, while two lines representing the spinal column and two deeply engraved chevrons are seen from the top view. Figure 33b shows a siltstone effigy with a toothed mouth, lines or shallow crescents running along the body and chevron shapes along the backbone. Figure 33c is one of only two decorated artifacts from the Marpole phase made of steatite. A small head has been defined by a sawn groove in this flat section of stone. An eye and suspension hole have been drilled, while on the back a longitudinal line with short intersecting lines depict the backbone. This little pendant is similar to undated steatite effigies found in the Yale area (Duff 1956a:Plate 20K). Figure 34 is a fragment of an unknown antler object which relates to the pendants discussed above. A creature with a long, thin body is portrayed with small, tightly folded limbs under its body. It has a wide mouth and teeth, crescents or gill operculum behind the mouth, and large lenticular eyes. Under the head are two rows of three dots while under the  118  119  120  hind end are three chevron shapes. From the top view a double row of engraved rectangles form an unusual checkered pattern for the backbone. Smith (1923:Plate II) shows the same motif on a barbed harpoon point from a Saanich area shell midden. This same design of rows of notched rectangles is occasionally seen in Coast Salish sxwayxwey masks identified as "snake-faces" (Suttles 1983:Figure 44c & d). In the masks the rectangles are painted white and red, while in Figure 34 the checkered effect is achieved by deeply carving out every other rectangle. The features depicted in Figures 33 and 34 most likely depiction a Northwest Coast creature known as the lightning-snake, and among the Coast Salish as sc'i'nkw, described as having the body of a snake with scales like a salmon (Jenness 1955:60).  Browbands: Seven thin and flat decorated antler and bone pieces are shown in Figure 35 and 36. Three of the pieces were found with burials. Figure 35a - d are thought to be browbands with holes on the ends that would attach the decorated rectangles to fiber headbands or tumplines. The other carved fragments are of a similar size and thickness and may also be browband fragments.  In comparison to the pendants, these browband fragments show  consistently fine workmanship in the deeply engraved designs. Figure 35a shows a wedge or T-shape that most likely defines the margin of a Coast Salish style eye. Figure 35b also depicts an eye motif. In this example curved lines above and below the eye margin enhance the central motif. Although not depicting a human or animal image Figure 35c shows a zigzag design which is defined by negative engraved space on one side and a cut out margin on the other. It is one of many browband and bracelet fragments found at the Marpole site with a deeply engraved zig-zag design (Smith 1903:Figure 96). Figure 35d appears to be the incomplete design of a bird's face. Two parallel engraved lines form the forehead then curl to create an eye shape on one side. Chevrons within the "beak" complement the two lobes of the forehead. The whole composition suggests the depiction of an owl. Figure 35e features a  121  Figure 35. a-e Marpole antler browband fragments (example d,fromSmith 1903:Figure 51).  122 Figure 36. Marpole block engravings with formlines. a, antler fragment with curvilinear block engraving and formlines; b, antler fragment with formlines; c, Coast Salish spindle whorl.  a  c  123  deeply engraved dot and crescent design which may represent a backbone and ribs but too little of the original composition has been preserved. Figure 36a is a fragment of a complex curvilinear design similar to wing designs found on historic Coast Salish spindle whorls as shown in Figure 36c.  The fragment  suggests the inner section of a wing joining a bird's body with a face depicted inside. Figure 36b also shows a close relationship to Northwest Coast art. Two negative engraved T-shapes define a fragment of an elegant raised design. The interconnecting lines swell in the centre and diminish in the corners in the manner of formlines. The historic spindle whorl shown in Figure 36c shows a similar shaping of raised formlines as in the prehistoric fragments in Figures 36a and b.  Stone sculpture: Of 19 recorded decorated stone artifacts from the Marpole site 11 are bowls with human and animal representations.  The majority of the bowls are composed of easily-  worked sandstones which have been pecked to form outline shapes and facial features. There are single examples of the use of steatite, vesicular lava and granite. The granite and lava have been pecked in the same manner as the sandstone while the steatite seated human-figure bowl has been carved and deeply engraved. Following the typology developed by Duff (1956a) in his analysis of stone sculpture from southwestern British Columbia the Marpole site sculptures can be divided into four groups:  3 "Type B human figure bowls" depicting upright human figures with arms that enfold a circular bowl (Figure 37-38). . 2 bowls and 1 bowl fragment with a single human head facing outward (Figure 39). 2 "Type A zoomorphic bowls," where simple features have been added to the outside of the bowls (Figure 40). 3 "Type B zoomorphic bowls" with naturalistic animal outline shapes inside of which are bowl depressions (Figures 41-42).  124  The steatite human figure bowl in Figure 37 shows similar carving techniques to those used for antler sculptures of the prehistoric period. The sculpture has a cylindrical shape which gives it three dimensions but the carving techniques are primarily twodimensional. The features of the face and the fingers on the hands are engraved. From the front a topknot rises from the crown as in other prehistoric human sculptures. From the rear this projection appears to be a headdress with notches suggesting a backbone or a snake's body. Thin, weakly developed arms and legs project in low relief from the bowl form. Figure 38a, "the Marpole Image", was found over 1.5 meters below the surface of the midden on top of a stone cairn containing a burial (Hill-Tout 1948). The prismatic shape of the face gives a sense of three dimensions, but as in the previous example, the execution of the features shows a two-dimensional approach. The facial features are lightly pecked while the shape of rudimentary arms are pecked in low relief on the sides of the bowl. Although the figure holds a bowl shape in its arms it is not hollowed out. Figure 38b shares the features of an open mouth and topknot with Figure 37, but the facial features have been pecked to form raised positive features rather than the negative carved or pecked features in the previous two examples. All three seated human figure bowls show a crude execution of the design and rough workmanship compared to undated steatite human figure bowls from the Lytton-Lillooet area, although the theme of an upright human holding a bowl is common to both. The steatite human-figure bowl in Figure 37 was subjected to chemical and microscopic examinations which revealed that ochre~a substance associated with Northwest Coast ritual~was present in the bowl and on the figure (Rudy 1982) while traces of animal or vegetal compounds were absent. Its tiny size of 10 cm belies a practical function as a paint mortar and supports Duffs (1956a) conclusion that these bowls had a ritual function. Similarly, the sandstone human-figure bowl dubbed "the Marpole Image" in Figure 38a has a bowl shape but is without a functional bowl depression.  125 Figure 37. Marpole steatite seated human-figure bowl.  126 Figure 38. a-b, Marpole sandstone seated human-figure bowls.  a, s i d e view  a, f r o n t view  127  The sculptures of heads shown in Figure 39 are pecked so that the facial features stand out as raised ridges. In Figure 39a sawing and incising has also been used to define the features and rows or zigzag or double-chevron designs on the rim and side of the bowl. (When viewed from the top the bowl's rim has a chevron design, while a side-view of the rim shows another chevron. When combined or flattened out the design is a zig-zag.)  An  identical design is incised on the back of the stone pendant in Figure 33b identified as a possible snake. Figure 39b shows a good control of the stone medium, unlike the rough workmanship of most Marpole sculptures. The shape of arms has been pecked in low relief on the side of the bowl in a manner similar to seated human figure bowls.  In this  composition the bowl forms the back of the figure. Figure 39c is the fragment of a face that may have formed part of a sculptured bowl. An open mouth and lenticular eyes are pecked into the coarse-grained lava. Figures 39b and c show a prism shape to the face with the nose and forehead forming a sharp ridge. This face shape is common to human and animal stone sculptures from the Marpole phase period. Figure 40 shows two sandstone "Type A" zoomorphic bowls. Both bowls are oval shaped with pecked round eyes, wide mouths and crescent shapes behind the corners of the mouths. The features of Figure 40b are more deeply engraved and the eyelid margin extends along the rim of the bowl. Figure 41 and 42 show "Type B" zoomorphic bowls. The shape of the two bowls in Figure 41 suggest bird forms, while the bowls in Figure 42 are four-legged animals. Figure 41a was found by a private collector at the Marpole site. It's features are worn but a crested bird's head with a short, hooked beak can be discerned. The tail end of the bowl also has facial features but they are too worn to interpret. The bottom of the bowl has several irregular deep holes. Figure 41b shows the bird on its back with its head facing into the bowl in its belly. The eyes, mouth and feathers are incised onto the pecked wing and head forms. The base of the bowl ends in a short pedestal. The design of the bowl with a bird facing  128  Figure 39. Marpole stone bowls with human heads. a, human head bowl (Smith 1903:Figure 53a); b, sandstone bowl; c, fragment of human head bowl (?).  a  129  132  inward is similar to an undated steatite bowl found on the Skagit River (Duff 1956a:Plate 15). The squat bodies and short legs of the bowls in Figure 42 suggest turtle forms but the features are too worn and simple to allow a definite interpretation. In Figure 42a large bulging eyes are on top of the head while thin legs emerge in low relief on the sides of the bowl. A similar undated bowl was found at Hope (Duff 1956a:Plate 17F). The Marpole bowl has traces of ochre around the rim of the bowl and on the tail. In Figure 42b the animal's face points downward and the eyes are hidden underneath the bowl. The mouth is beaklike and suggests a bird although the four legs belie this interpretation.  Duff (1956a:61) also  noted that several bowls from the Yale area had bird-like tails and beaks and four legs. The distribution of the bowl forms discussed above has been analysed by Duff (1956a). With the exception of the "Type A" zoomorphic bowl the other bowl forms are common throughout the Strait of Georgia and Fraser-Thompson River regions.  Atlatl weights: Two sandstone objects thought to be decorated atlatl weights are illustrated in Figure 43. Figure 43a, found below a human skull (Smith 1913) depicts an anthropomorphic face. An unusual feature of the face is the round eye with an elongated, downturned eye margin. The face is defined by a prominent notched ring that may represent a headdress or hair style, or may as Smith suggests (1913:185) represent a plaited ring of cedar bark. Duff (1956:76) describes the following features for this object, "The head has two ear-like projections from the top...the humped back is decorated by six incised lines spaced evenly along its length." This may indicate a representation of the lightning-snake or thunderbird. Figure 43b is a broken piece of stone with an engraved motif on each side. One one side two curved sections stand out in low relief. Inside the left-hand section are curved lines which suggest a bird claw motif (see Suttles 1983:Figure 4:14h) while the right hand lines  133  Figure 43. Marpole stone atlatl weights. a, sandstone weight (Smith 1903:Figure 57); b, sandstone weight fragment with claw motif.  134  are difficult to interpret.  On the second side the drilled hole forms the interior of an  elongated eye while underneath engraved, curved lines are again difficult to interpret. Duff (1956) cites three other examples of undated prehistoric anthropomorphic and zoomorphic atlatl weights: two from the Yale locality and one from Chilliwack. Butler and Osborne (1959) illustrate two undated zoomorphic atlatl weights from the Dalles-Dechutes region of the Lower Columbia River.  Miscellaneous decorated items: A zoomorphic siltstone concretion is shown in Figure 44a. It has the shape of the proximal end of a "slave-killer" type of club with an animal head at the tip and a blunt projection below the head. A lenticular eye and simple mouth are incised on the face. The small size and soft stone composition of this object suggests that it was not a functional club. It was found with skeletal remains. Figure 44b is an antler haft with a bird's head carved on the end of the tine. An oval eye and mouth margin have been deeply engraved on the face. On the "body" rows of ticked lines encircle the neck while long parallel lines form an unusual geometric design. Traces of red ochre can be seen in the engraved lines.  The zoomorphic head with geometric  ornamentation of the antler shaft is most similar to examples of decorated digging sticks from the southern interior (Smith 1913:Plate III).  This artifact is not part of excavated  collections from the Marpole site. It was found at the site and donated by a private resident.  Summary The images presented on Marpole decorated artifacts show a range of human, bird, fish and animal motifs similar to those found in collections from previous periods. For the first time there is a large enough sample to demonstrate that human and animal motifs are highly correlated to specific artifact types.  At the same time, there is a general  135 Figure 44. Marpole miscellaneous decorated objects. a, miniature siltstone "slave-killer" club fragment (?); b, engraved antler zoomorphic handle (?).  b  136  heterogeneity of carving styles. A good example is provided by an examination of eye forms found on anthropomorphic and zoomorphic images. In an attempt to outline a typology for eye forms found at Marpole, 44 eyes were divided into categories based on outline shape, compound features (eyeball, iris, eyelid), and carving style. Twenty-five eye styles were identified! The exercise did prove useful in indicating that carving style and presence or absence of compound features showed no correlation to the identity of a creature or the material the object was made of. A significant correlation was shown between eye shape and animal identity.  Human eyes tend to have round or lenticular outline shapes whereas  animals and have both round and sharply pointed, downturned eyelids.  There is an  increasing complexity in the way features are delineated in comparison to previous prehistoric periods with 75 % of the eye shapes exhibiting secondary and tertiary elements. The Marpole site provides a large sample to compare with the following other Gulf of Georgia Marpole phase assemblages.  Beach Grove site The Beach Grove site, DgRs 1, is a winter village midden situated on Point Roberts, where historically both the Halkomelem speaking Tsawwassen and the Lushootseed speaking Nicomekl had villages, and close to where other Straits groups had reef net stations. Because the site was excavated by seven parties over the period 1956 and 1980 it presents a challenge to interpret. The site has a series of radiocarbon dates which are not necessarily correlated to the site assemblages obtained by the different excavations. The results of 14 radiocarbon dates suggest a Locarno Phase component at the north end of the site, and a larger late Marpole phase component reflecting more intensive use of the site from approximately 1550 to 1250 B.P. (Matson et al. 1980:99). Although the site is within the time span when many southern coast sites exhibit attributes of late prehistoric phases, the Beach Grove site is fully Marpole in character, therefore is included in the Marpole phase  137  analysis. The five decorated objects found in situ at the site support this interpretation by fitting well within the range of decorated forms and carving styles found in the Marpole period. Figure 45a shows what was likely a pendant with a broken suspension hole at the top. The bone piece shows the worn, simple features of a human face with drilled holes for eyes and barely discernable incised marking defining the nose and mouth. The forehead and facial margin is a separate raised ridge around the face. The artifact form, its motif of a human face, and the composition with a defined face margin is typically Marpole in style. Figures 45b and c are also similar to decorated objects from the Marpole site shown in Figures 31a-c. Both are cutout profiles of birds' heads carved in antler. Example b suggests the shape of an owl or hawk head while the long neck and down-curved beak in example c is unidentifiable.  Both objects have large carved out holes with smaller carved out slits for  suspension or use as buckles or toggles. Figure 45d is a small bone carving of unknown function although the profile shape of a bird's head with a long beak is a common Marpole motif. A round eyeball and angled eyelid shape with two crescents projecting down from the lower lid are unique to this carving, but both the eyelid shape and crescents below the eyes are found separately in Marpole phase art. Figure 45e is a carved fragment of bone or antler showing two curving shapes. The fragment shows the occurrence at Beach Grove of the two-dimensional block engraving style with raised curvilinear motifs also found at the Marpole and Esilao sites. Figure 46 shows a stone bowl in the shape of a human head. It was uncovered during bulldozer excavations, and therefore cannot be considered part of the sample from the Marpole component of Beach Grove. The rough-grained boulder was pecked and sawn in a manner similar to stone art from the Marpole period in particular Figure 39b. The pecked out negative space of the eye socket area defines the outline shape of the forehead, cheeks and bridge of the nose which are joined in a positive form surface. The face is prismatic in cross-section giving an added three-dimensional quality to the large sculpture.  The  138  Figure 45. Beach Grove bone and antler artifacts. a, bone human face motif pendant; b-c, antler bird form pendants, d, bone zoomorphic effigy; e, antler fragment with block engraving.  140  prismatic shape of the face, relationship of the facial features, general form and carving technique all conform to Marpole phase art style described in the previous section. The artifacts from the Beach Grove site all have their closest parallels in art the Marpole site.  Whalen Farm site The Whalen Farm site, which straddles the U.S.-Canada border, has two site designations: DgRs 14 and DfRs 3. It has been excavated by several archaeologists and illustrates the difficulty of trying to correlate artifacts from different excavations of one site. The site, on the eastern shore of the Point Roberts Peninsula facing Boundary Bay, is within the historic territory of Straits-speaking Semiahmoo Salish, and is thought to be strategically situated near reef-net fishing sites. The site was first mentioned by Smith (1925) who described a half-mile-long shell midden with house depressions. Smith also described the recovery of two antler tines with simple zoomorphic features incised on their distal ends (1925:317). Borden excavated part of the U.S. portion of the site in 1949 and obtained two radiocarbon dates (Wilmeth 1978) of 1580 ± 140 B.P. for the bulk of the material in the cultural deposits designated Whalen II, and 2450 + 160 B.P. for deposits designated Whalen I from the lowest horizon. Several years later Duff (1956b) reported on an isolated burial recovery 1.6 meters below the surface of the midden. It included an antler chisel handle with an anthropomorphic image and a chipped stone knife with an antler handle embellished with ticked lines. Seymour's 1972 excavations (1976) in the Canadian portion of the site proved clear Marpole phase associations and uncovered a stone bowl with a human face. No radiocarbon dates were obtained. Seymour found a thin historic deposit but asserted that below this were Marpole-age deposits. Lastly, excavations under the auspices of the Delta Museum in 1985 uncovered two barbed harpoon points with zoomorphic carvings and obtained three radiocarbon dates of 2360 ± 120, 2100 ± 70 and 2060 ± 110 B.P.  141  (Hammon 1986). A zoomorphic stone bowl and a zoomorphic antler carving not found in situ are also part of the collections from the site. Borden (1970) thought that the Whalen II assemblage was distinctive enough from Marpole material to indicate a separate cultural phase. He found the absence of pecked stone and ground slate objects, and the presence of microblades and certain chipped point styles significant enough to warrant the designation of a Whalen II phase. Archaeological work in the decades since Borden's work has shown that Marpole-age sites do exhibit some variation in chipped and ground stone artifacts.  As well, microblades and other items  Borden thought were distinctive of Whalen II have been found in other Marpole phase sites. Archaeologists now usually place the Whalen Farm site within the Marpole phase (Mitchell 1971, Burley 1980). The most recent radiocarbon dates indicate that the site was likely occupied for the full time-span of the Marpole phase. Unless otherwise noted, all of the objects discussed in the next section were found in situ within the Marpole-age deposits. Both of the two stone bowls found at the site are similar in form to bowls found at the Marpole site. Figure 47a shows a broken, small igneous stone bowl with a human-like face portrayed on the outside. It was found in situ (Seymour 1976) near the top of Marpoleage deposits. The features have been formed by sawing and pecking long indentations. This rough sawing technique created a broad face with elongated, raised ridges outlining lips, and eye margins. The eyes are lenticular with eyelids that are pointed on the outside and rounded on the inside. The margin of the face has been offset from the bowl by an encircling ridge. Figure 47b, a "Type B" zoomorphic bowl was not found in situ. The soft, shiny stone material has been roughly sawn and pecked. Incised nostrils at the end of a snout suggests that this is not a bird but further definition is impossible. An open mouth and tongue are portrayed and jawbone lines are incised on the underside of the head. Figure 48a shows one of two zoomorphic carved barbed harpoons found in the lowest levels of the 1985 excavation in basal sand deposits (Hammon 1986). The second example is  142  Figure 47. Whalen Farm stone bowls. a, fragment of human head bowl; b, "type B" zoomorphic bird bowl.  a  Figure 48. Whalen Farm antler zoomorphic artifacts. a, zoomorphic harpoon point; b, fragment with feather motif.  144  in a bad state of preservation. Although sections of the carving are worn, the features that can be discerned, and the general form of the artifact are identical to Figure 28c from the Marpole site. The elongated, pointed eye has been deeply engraved as have lines running behind the cheek. Dorsal and side surfaces of the object are worn suggesting that this may have been a functional harpoon rather than a purely decorative object. Figure 48b is an intricately carved bone fragment that is curved in cross-section. A raised motif of what appear to be four pairs of feathers forms the main design, while parallel engraved lines define another feather motif at one end. The composition resembles the back and tail feathers of a bird. Although Borden stated that there were no similar prehistoric examples of this motif (1983:160), a similar raised feather design from Musqueam Northeast is illustrated in Figure 17a. Although the base portions of the feather shafts are missing from the Whalen site fragment, it is unlikely that the negative spaces in the centres of the feathers are large enough to indicate wedges~as is seen in historic Coast Salish feather motifs. This motif and others like it from the Marpole period are clearly ancestral to the block-engraving style of Coast Salish feather motifs shown in Figure 17b. Figure 49 is an antler beaver-tooth chisel haft found with the flexed skeleton of an adult male. Although it was not uncovered during one of the major excavations, once the burial was exposed 1.6 meters below the surface soil, it was excavated and examined by an archaeologist (Duff 1956). . This finely executed sculpture combines incising, engraving and sculptural modelling techniques.  The oval eyeballs, and eyelids are defined by deeply  engraved lines while the eyebrows form a raised ridge and define the plane of the forehead. The head is offset from the body trunk by a sudden constriction similar to other prehistoric antler sculptures in Figures 6 and 9. As in other prehistoric sculptures the definition of limbs is rudimentary with the arms, hands, and what may be the sternum barely discernable.  Where this sculpture is unique is in the delicately modelled rendering of a  bulbous nose, grimacing mouth and furrowed cheeks. The general composition of the face with its raised forehead plane, raised eye features, large nose with defined nostrils and open  145  Figure 49. Whalen Farm anthropomorphic antler chisel haft.  146  mouth is similar to both stone and antler depictions of humans from the Marpole site. Duff (1956:69) postulated that this finely sculptured chisel marked the grave of a carver. Figures 50a and b show two antler artifacts found at the Whalen site in 1922 and described by Smith (1925). Example 50a may have functioned as an antler flaker. Figure 50b, although not visible in the drawing by Smith, is described as having short lines extending up perpendicular to the edge of the mouth. It has been carved out under the animal's jaw or beak, with the widest part of the hollow near the head. Figure 50c shows a fragment of an antler carving found in deposits excavated by a bulldozer. It is included as a comparison to Figure 23 from the Crescent Beach site and Figure 51b from Cattle Point. The three small antler carvings depict heads with carved out eye holes encircled by engraved lines. It is likely that the eyes once held inlays. The inlay technique, circular rim around the eye and composition which suggest small masks may relate to some of Carlson's (1987) discoveries at the Pender Island site which date one thousand years earlier. The Crescent Beach, Cattle Point and Whalen Farm site examples occur in assemblages whose time spans are different but do overlap to include the early Marpole period circa 2350 B.P.  A single, simple shell pendant (not illustrated) with  rudimentary human features comes from the Marpole-age component of the Crescent Beach site.  San Juan Island Sites The San Juan Islands are the southernmost gulf islands in the Strait of Georgia. These island must have been an important food resource area since the territories of the four Straits-speaking Salish groups-the Songhees, Saanich, Lummi, and Saamish-included these islands. San Juan Island, the largest in these southern group of islands has four archaeological sites with Marpole phase components where decorated objects have been uncovered.  147  Figure 50. Whalen Farm undated antler artifacts. a-b, antler zoomorphic objects (Smith 1925:Figure 1-2); c, zoomorphic head fragment.  a  148  The Cattle Point site, 45 SJ 1, was excavated by King (1950) before the Marpole phase and other prehistoric cultural phase designations for the Gulf of Georgia had been proposed. Decorated objects were found in components that King named the Maritime and Developmental phases.  Reanalysis of the Cattle Point assemblage has lead Mitchell  (1971:55) and Carlson (1983:199) to place both the Maritime and Developmental phases in the Marpole phase, although some Locarno Beach phase attributes are noted in the earlier Maritime component. Recent radiocarbon dates suggest that the site dates between 2860 ± 158 and 877 ± 159 B.P. (Burley 1980:33). Two antler decorated objects from the older Maritime component of the site include a carved antler harpoon or leister point, and the fragment of a carved antler zoomorphic head. Figure 51a shows the butt end of a broken antler harpoon or barbed point depicting the profile of an animal. The elongated eyelid, nostril, open mouth and long beak or jaw is similar in style to carved harpoons from Marpole phase material described previously. The carving differs from other examples in several details: a projection that likely functioned as a line guard is positioned so that it forms the creature's ear, there are no engraved horizontal lines that extend out from the back of the head, a curved line joins the lower eye margin to the mouth margin, and the top of the "snout" features a row of short parallel lines with two spurs projecting out from them. King thought that the carving resembled a wolfs head similar to Nuu-cha-nulth wolf masks, and indeed, projecting ears or horns are suggested. The curved line extending down from the eye and joining the mouth is similar to compositional elements on Figures 28 d and f from the Marpole site. The carving techniques include engraving and incising with a bevelled v-shaped incision shaping the negative space surrounding the pupil. Uneven lines and points of disjuncture attest that this example is not as well executed as other zoomorphic harpoons from Marpole sites. Figure 51b shows a fragment of a carved zoomorphic head. Similarities to Figure 23 from Crescent Beach and Figure 50c from Whalen Farm have been noted above. In this example the top of the "snout" features diagonal lines that join a a midline similar to a rib  Figure 51. Cattle Point decorated artifacts. a, antler zoomorphic harpoon point; b, antler zoomorphic head; c, siltstone fish effigy (drawn from King 1950).  top  view  150  and backbone motif, but the placement of this design on the head belies interpreting the motif as a rib design (or does it?). Although the lower half of the head is missing, side A shows an engraved line running from the lower corner of the eye to the end of the nose, while side B shows a nostril and smaller incised jowl line. Examples of Salish, Kwakwaka'wakw and Nuu-cha-nulth carvings feature carved or painted lines that run from an animal's eye to the nostril and appear to be compositional devices that unite the separate elements of the face. Figures 51a-b may be prehistoric examples of this same compositional principle. Figure 51c shows a small siltstone plaque with the a fish or whale lightly incised on one face. This artifact was found in Maritime phase deposits which likely represent the late Marpole period. The absence of a gill operculum suggests the depiction of a whale. Figure 52, an undated artifact from the northern Gulf of Georgia area, shows a carved antler zoomorphic head similar to Figure 51b. The circular eye sockets suggest the use of inlay, and the hollowed out underside of the head as well as the carved out eye is similar to carvings from the Pender Canal site (Carlson 1986), the Crescent Beach site (Figure 23), and the previous figure from Cattle Point. The Garrison site, 45 SJ 25, is a shell midden on Garrison Bay on the northwest side of San Juan Island. It has radiocarbon dates of 2100 + 100 and 1580 + 60 B.P. (Burley 1980). Figures 53a and b show two carved antler objects found with a burial. It is likely that both artifacts were browbands that broke and were recut and drilled to use as pendants. Figure 53a is clearly a fragment of what was once a larger composition. Four T-shapes create intriguing curved positive shapes that cannot be identified. The balance of negative shapes and positive formlines is similar to Figure 36a from the Marpole site and is reminiscent of some of the complex block-engraved compositions of Coast Salish spindlewhorls and rattles. Figure 53b depicts the upper portion of a lenticular eye with an eyeball and pupil. The broken suspension hole on the lower margin as well as the incomplete eye shape suggests that this artifact is also a modified section of a larger composition. The eye  Figure 52. Undated Gulf of Georgia carved antler zoomorphic head.  £H2J LIT  1,1  I  Figure 53. a-b, Garrison antler pendant/browband fragments.  153  motif, which is found on several brow-bands (Figure 35a-b) from the Marpole site, reinforces the hypothesis that this pendant once functioned as a browband. Figure 54a shows a roughly-carved representation of a face. The antler or bone object has a broken section missing from the top that was most likely a suspension hole. Although the workmanship is crude there has been some attempt to create two planes of the face. The eye socket area has been carved away to allow the forehead, nose and eyeballs to form a prominent plane. This artifact illustrates that even in a roughly conceived and executed object one can see that there is a conceptual model being followed-that of a human face presented in frontal view using engraving techniques to create several form surfaces. Figure 54b is a fragment of siltstone on which an eye, section of mouth, and facial margin have been deeply engraved. The eye form is similar to a Coast Salish eye style where the center of the eyelid closely follows the contours of the eyeball then angles sharply outward. The margin of the face is surrounded by a notched ridge which may relate to the notched facial margin on the stone face depicted on an atlatl weight at the Marpole site (Figure 43a), or the sectioned ridges representing a backbone in Figure 34.  Carlson  (1983:200) suggests that this image is "the earliest representation in the same style as a thunderbird" and may be a fragment of a spindle whorl. The block engraving style and the oval mouth formed by two deeply engraved parallel lines is similar to faces on Salish rattles and spindle whorls (Suttles 1983: Figure 4:3a,c and d) but the features that would identify this carving as a thunderbird representation are not specified. Although this artifact cannot be given a specific interpretation the engraving style, as well as shaping of the mouth, eye and facial margin can be compared to Coast Salish motifs and carving styles. Lastly, Carlson (1960:Figure 4, F:m) illustrates an antler graver from this site with a simple eye and mouth line incised on the distal end of the tine. Perhaps the most interesting artifact from the San Juan Islands is a zoomorphic barbed harpoon from the Whatmough Bight site, 45 SJ 280 (Burke Museum cat. notes), illustrated in Figure 55a. This artifacts exact provenience within the site, and the site's age  154  Figure 54. Garrison decorated artifacts. a, antler human face pendant; b, siltstone fragment with anthropomorphic (?) design.  a  155  Figure 55. Undated zoomorphic antler harpoon points. a, Whatmough Bight, San Juan Islands; b, Sooke locality, Vancouver Island.  156  is not known but it is nearly identical to a harpoon illustrated in Figure 28c from the Marpole site, and Figure 48a from the Whalen Farm site. A last example of this zoomorphic form is illustrated in Figure 55b. This surface find of a harpoon foreshaft from Sooke on Vancouver Island, features the long diagonally projecting mouth, elongated eye margin and lines running behind the head typical of this artifact type. In this example the outer eye corner is sharply downcurved, and two long wedges extend down the body in a manner similar to Figure 28d from the Marpole site. Although this artifact cannot be included in the research sample it should be noted that it comes from an area that was part of the territory of the Songhees Straits Salish which also included part of San Juan Island. The last decorated artifact from this locality comes from the Argyle Lagoon site, 45 SJ 2, on the southeast side of San Juan Island (not illustrated, see Carlson 1954:Plate 11). The antler haft has a crudely carved eye and mouth roughly with long, irregular lines extending from the corner of one eye (not illustrated).  False Narrows site The False Narrows site, DgRw 4, is on Gabriola Island one of the northern gulf islands in the Strait of Georgia. The False Narrows site is part of the historic territory of the Nanaimo Halkomelem Salish Indians who used the site in the summer months for fishing, sea-mammal hunting and shell-fish gathering.  Extensive midden deposits which stretch  over 1300 meters, house depressions, and numerous burials suggest that the site was once permanently occupied. The site was excavated by Mitchell (1966) and Abbott and Sendey (1967), and analysed by Burley (1979). Four cultural components are present at the site from the Marpole phase to a historic Salish component. Decorated artifacts come from the Marpole-age False Narrows I component, and the transitional late Marpole-Gulf of Georgia component with a date of 1710 ± 90 B.P. (Burley 1979a).  157  One of the most interesting conclusions drawn from the analysis of site materials is "that the False Narrows mortuary data support the tenet that a ranked society has been in existence within the Gulf of Georgia from at least Marpole times onward...(Burley 1979a:336)." By assessing differences in the quality and quantity of burial inclusions Burley cites evidence of differential treatment accorded to burials. In the top "elite" group are three burials of adolescent or young adult males with burial goods ranking high above all others. (It should be noted however that this ranking is based on arbitrary values given to different types of decorative objects. If dentalium beads were to be given a value of 2 instead of 1, with clam shell beads valued at .5 then female burials would be equally represented in the elite group.) Figure 56a is a lignite (coal) pendant which was found along with over 2500 shell beads in the top ranking burial of an adolescent male.  The intrusive burial has been  assigned to the Marpole phase False Narrows I component. Although Burley suggests that this pendant might represent a beetle, particularly when viewed from the side, I feel it depicts the front view of a simplified human face. This conclusion is partially based on the number of other pendants from Marpole period sites that depict human faces. If this is the case, round eyes and a wide curved mouthline are represented on this highly polished oval of black lignite. The mouth line has been deeply scored at an angle to create a swelling upper face plane. Five roughly incised lines are inscribed laterally across the pendant but do not appear to relate to the compositional elements or standard of workmanship of the rest of the piece. A second pendant from the False Narrows I component in Figure 56b is a roughly incised disc of siltstone depicting a face. The triangular nose and grinning aspect of the face as well as the loose style of execution is reminiscent of Figure 29a from the Marpole site. Figure 57a-c shows three stone zoomorphic effigies. Figure 57a, from False Narrows I, has rough incising of an eye and mouth on both sides of a siltstone concretion. While the identity of the latter two examples cannot be identified Figure 57b from False Narrows I, and 57c from the late Marpole dated False Narrows II, appear to depict fish.  158  Figure 56. False Narrows human face motif pendants, a, lignite pendant; b, siltstone pendant.  a  159  Figure 57. False Narrows stone effigies. a, zoomorphic image on siltstone concretion; b-c, siltstone fish (?) effigy fragments; d, head fragment from seated human-figure bowl (?). a  b  d  160  The division of the Marpole deposits at this site into two temporal units gives the first refined dating of an example of stone human figure sculpture. Figure 57d depicts the features of a human face engraved and sawn onto a wedge-shaped fragment of sandstone from the late Marpole dated deposits. One can only speculate whether this fragment was originally part of a seated human figure sculpture, but it does bear resemblance to a sandstone human figure bowl thought to be from the Washington-British Columbia border area (Holm 1983:Figure 205). The carving style of the False Narrows sculpture is unrefined but it shares two features in common with other seated human figure bowls-a wedge-shaped face and the depiction of an open mouth. Figure 58 presents an interesting puzzle. The bottom fragment in Figure 58a is associated with a burial from the False Narrows II component, while the top fragment is from the False Narrows I component. The two pieces might not have been associated if not for two fragments of one sculpture found at Orcas Island in the San Juan Islands shown in Figure 58b. The Orcas Island zoomorphic sculpture is from an undated private collection. It shows a four-legged animal with a long tail and body which is identical in form and carving style to the False Narrows fragments. In the Orcas Island example two parallel lines along the tail and back likely represent the backbone and tailbone, while in the False Narrows example a single line with short diagonal lines suggests a rib and backbone motif. The sharply defined forelimbs on both sculptures and the general form of the animal brings to mind the zoomorphic sculpture from Pender Canal in Figure 7, and squat-bodied animals depicted in Coast Salish artifacts as shown in Figures 58c and d. They share rib and backbone motifs, long squat bodies, and sharply angled leg joints. The proportions of the sculptures in Figure 58a-b suggest the shapes of otters or other members of the weasel family.  161  Figure 58. Zoomorphic effigies. a, False Narrows zoomorphic fragments; b, Orcas Island zoomorphic fragment; c-d, Coast Salish zoomorphic motifs.  a.  b  c  162  Musqueam North Musqueam North, DhRt 3, near the mouth of the north arm of the Fraser River is situated within the modern community of the Halkomelem Musqueam Salish. Musqueam North is an early Marpole phase village midden with a basal date of 2340 ± 60 B.P. (Wilmeth 1978).  Figure 59 is an antler pendant from the site which may also have functioned as a  small pestle, as in Figure 31c and 32b from the Marpole site. Portions of the little human figure such as the head and chin have been delicately modelled, while other features such as the nose, eyebrows, and hair have been engraved or incised using two-dimensional carving techniques. The nose, forehead and lips form a flat facial plane while the eyes and cheeks are contoured. A greenish stain in the eye sockets suggests that a copper inlay once formed the figure's eyes. All of the features of this artifact work together to create a pleasing composition: the projection on top of the head with a suspension hole suggests a hair topknot, and the tapered pestle is of a size to represent the figure's torso. Figure 60 is one of the most sophisticated compositions from the Marpole period. The small antler zoomorphic face holds in its mount the stump of what was likely the stem of a blanket pin. The dorsal presentation of a flat face with eyes on top of the head and a wide mouth suggests the portrayal of a snake. The eyes have bulging eyeballs and long tapered eye margins which conform to the shape of the eyeball, slant down at the outer edges and join together at the nose. The mouth is a raised ridge which swells slightly in height at the corners and at the nose creating a grinning realism. What is unique in this piece is the modelling of the negative space around the mouth and behind the eyes creating slanted planes that rise gradually to the ridges forming the eye and mouth margins. The eyes, nose and mouth are joined by these raised margins in a manner analogous to formlines. All of the features of the composition combine to create a gracefully composed face with a definite serpentine character. What is also remarkable is that this piece is only 3 cm long. If the mouth did hold the stem of a pin it would have represented the reptile's tongue, which is an interesting choice considering that if the stem came out of the back of the head it could have  Figure 59. Musqueam North anthropomorphic antler pestle/pendant.  164  Figure 60. Musqueam North zoomorphic antler carving.  165  represented the snake's body. As with many human sculptural images the top of this animal's head has a projecting topknot.  Musqueam Northeast Musqueam Northeast, DhRt 4, the second of three sites on the Musqueam Indian reserve contains two distinct temporal units. The Locarno Beach phase assemblage dated from 2550 to 2970 years old has been previously discussed. The second unit has not been radiocarbon dated but its assemblage has been assigned to the late Marpole period circa 1500 years B.P. (Archer 1972:26). Figure 61a was found in a transition zone between the Locarno Beach and Marpole deposits and might belong to either period. It is a small antler fragment depicting the profile of a zoomorphic face. The long, curved mouth has been formed with an engraved v-shaped line which widens to a cutout open mouth. The other features of the face are defined with double incised lines. The long upturned mouth and lines running horizontally at the back of the head are features similar to the creatures on barbed antler harpoons from Marpole-age sites. The width and thickness of this fragment is within the range of harpoon points found in Marpole assemblages but the cutout mouth appears too fragile to function as a harpoon. Figure 61b also from the late Marpole deposits at Musqueam Northeast is a small fragment of bone depicting a face in profile. Although it is only a tiny fragment of a larger composition, features familiar in art of the Marpole period can be discerned. A block engraving style has been used to define a long, pointed eye with a downturned outer margin. The eyeball is a drilled hole while the eyesocket area is defined by the raised cheek and forehead plane. A v-shaped cut for the mouth is visible behind the broken margin of the snout. Figure 62 was named xwanaimus, "smiling face", by Musqueam elder James Point when it was first discovered. Like several other Marpole phase examples of pendants  167  168  depicting human images a topknot projects from the figure's crown.  In this case the  projection is as long as the face and may have functioned as a pestle. This antler image of a human ranks as one of the finest artistic compositions of the Marpole period. The carving style is primarily two-dimensional in that there are abrupt junctures in the facial planes and deeply engraved areas defining the main features of the face. At the same time, the mouth and chin are delicately modelled with the upper lip swelling upwards from the base of the cheeks. The handling of negative and positive space in the face can be compared with modern examples of Northwest Coast art such as the Kwakwaka'wakw human face mask illustrated in Holm (1987:Figure 41). The manner in which the forehead, cheeks and lips are a joined area of high relief while the rest of the face becomes a continuous lower plane of negative space swelling up to points of high relief are the same. This style is also found elsewhere in the prehistoric record, in particular in several examples of seated human-figure bowls as will be discussed in a later section. The treatment of the facial features in Figure 59 from Musqueam North is also somewhat similar.  The Point Grey Site The Point Grey site, DhRt 5, on Burrard Inlet is also within the territory of the Musqueam Halkomelem Salish. The site is thought to be a Marpole phase summer fishing camp with radiocarbon dates of 1970 ± 100 (Burley 1980), 1690 ± 120, and 2210 ± 90 B.P. (Coupland and Unfreed n.d.). A single decorated object was recovered from the site. Figure 63 shows a Type A oval stone bowl of vesicular lava with facial features pecked onto one end. Two round eyes, an oval mouth and two nostrils complete the face. Because the eyes are on top of the head a top view of the bowl presents another face with the bowl's depression forming the mouth.  169  170  Port Hammond site The Port Hammond site is a large shell midden on the north bank of the Fraser River near the Pitt River. The site has been identified as cxwi't (Jenness 1955:20), an old village site of the Katzie Halkomelem-speaking Coast Salish.  Although Smith's (1903)  excavations at the turn of the century were not systematic the site has been evaluated by Duff as being at least 500 to 1000 years old (unpublished notes) and by Burley (1980:43) as being Marpole-age. The presence of unilaterally barbed antler points and composite toggling harpoon valves suggests a middle to late Marpole occupancy of the site. Smith (1903:187) reported that the artifacts from the Port Hammond site were most commonly found in a portion of the site used for burials. Decorated artifacts shown in Figure 64 include a steatite human figure bowl, two bowls with human heads, a pipe fragment with a zoomorphic shape, and an antler carving of human figures. The human figure bowl is one of three classified by Duff (1956a) as a "Type A" bowl, where the figure reclines on its back with the bowl describing a hollow in the belly. All three examples of "Type A" bowls were found in the Fraser Valley. Like Figure 37 from the Marpole site, this example of a human figure bowl in Figure 64a is made of steatite and has been incised and engraved. In the Port Hammond figure the workmanship is of slightly higher quality. Notable features are the Salish-style eyes identical to the eyes on the stone pendant in Figure 29f from the Marpole site, a notched headdress or hairstyle surrounding the face, and the depiction of ribs, a backbone and buttocks on the back of the figure. As in several other undated seated humanfigure bowls there is a small but deep hole in the top of the head. The bowl with a human head shown in Figure 64b has the pecked, raised features of . an oval mouth and lenticular eyes. The eyebrows and forehead are joined in a raised ridge which describes two heart-shaped lobes for the upper face (similar to Figure 64a). The third bowl from the site and the zoomorphic pipe fragment are too poorly illustrated to describe here (Smith 1903:Figure 55a & d).  171  Figure 64. Port Hammond decorated artifacts. a, steatite "type A" human figure bowl; b, human head bowl (Smith 1903:Figure 53b); c, antler carving (Smith 1903:Figure 59).  a  172  The elaborate antler carving in Figure 64c whose present provenience is unknown, is poorly illustrated but well-described by Smith (1903:186-7): A human figure will be recognized at the left-hand end of the piece of antler. It is squatting, the hands raised to the chin, the elbows touching the knees...A similar figure occupies the opposite end...but in place of the human head we find a bird's head, the beak of which is cut out of a branch of the antler. This may represent a mask worn by the man. The lower convex side of the object, near its middle, has been shaved off over the whole undecorated extent between the two terminal figures, and the centre seems to have been hollowed out. There is a hollow in the ends of each of the two prongs on the right-hand side of the illustration. It looks as if some object had been inserted in these two hollows, the one rising from the head of the bird, the other protruding from its beak.  Several features of this sculpture resemble seated human-figure bowl sculptures: the pose of the figures, the notched headdress or hair, and the hollowed area on top of one of the figure's head. The tightly-flexed pose is also found on much older sculptures described by Carlson (1987) from the Gulf Islands and Figure 9 from the Helen Point site. If the sculpture does represent a human wearing a mask then it is the first prehistoric depiction of this ceremonial item and gives credence to Carlson's (1987:10) interpretations of carvings representing miniature masks at the Pender Island site.  173  Puget Sound Sites  Fishtown site The well known "Skagit atlatl" illustrated in Figure 65 was found in the Skagit River delta in front of the Fishtown site, 45 SK 99, within the territory of the Swinomish Lushootseed-speaking Coast Salish. The site itself has been attributed to the Gulf of Georgia cultural phase with a date of 1220 ± 70 B.P. (Onat 1976:134). The yew wood atlatl has been given a date of 1700 ± 100 B.P. by AMS (accelerator mass spectrometry) dating (Fladmark et al. 1987). Although it is not known whether the site extends back to the same age as the atlatl I believe that it is reasonable to associate the atlatl with the site. When used as a base from which to hurl a spear the decorated portion of the atlatl would have hung upside down and acted as a weight. The Skagit atlatl stands apart from all other Northwest Coast prehistoric artifacts in its elaborate design and detailed realism. At first the researcher might despair at this indication of the rich artistic record in wood that has been lost. The stone and bone artifacts that form the bulk of this sample seem crude in comparison. Upon closer scrutiny, the atlatl fits comfortably within the range of attributes so far described for art of the southern coast, and shows that identical techniques were used in carving wood and antler.  Previously  described snake-like creatures have been only cautiously interpreted as representing lightning-snakes, but this artifact clearly depicts the feather-crested lightning-snake or to use the Halkomelem term, sc.i'nk.we (Jenness 1955). The pose of the snake suggests coiled up agility and power, while its face seems to express a fierce eagerness. This unique expressive figure gives a glimpse of the sculptural potential that wood provided to the prehistoric artist. Six feather-crests rise from the head and back of the snake. Short, folded fore and hind limbs are indicated as is a tail or tail feathers. Ribs are depicted on the creature's sides,  Figure 65. Fishtown wooden atlatl (Skagit atlatl).  175  while skin folds define the neck. The skin folds add to the fierce, grinning aspect of the face with its two rows of bared teeth set in a wide mouth. The creature rests its head and forelegs on a human head. The human face is framed by chin-length hair. Three faint lines run vertically through the coiffure ending in a circular depression at the base. Three long pairs of streamers or feathers extend from the base of the head to the belly of the snake. Although the outline shape of the sculpture is three-dimensional two-dimensional engraving techniques define the details.  Negative spaces between the third and forth  plumes on top of the snake and under the snake's neck have been cut out, while other similar areas-between the fifth and six plumes and behind the human's head-have not been carved away. Engraved wedges are found throughout the sculpture defining the eyemargin, ribs, and streamers (these are not split U-forms as suggested by Fladmark et al.). Inside each feather an engraved line marks the quill in a manner similar to the heron image from the Marpole site in Figure 32b.  The feather crest behind the snake's eye is anomalous in  featuring an elongated wedge which might suggest that an ear rather than a feather is represented here. When viewed from the top the snake's tail and hind legs are defined by two-dimensional block engraving. The tail has a shape similar to two fragments from the Locarno Beach site (Figure 22b-c) identified as a possible whale's tail motif. On the human face the area around the lips and the nostrils are deeply engraved. Small wedges and an engraved crescent define eyebrows while the eye is an engraved hole. Luckily, the inlay material used for the eyes remains in the snake's pupil but its composition has not been identified. All of the carving techniques, design elements and motifs used in creating this sculpture are found in Marpole phase art.  The configuration of the human head is  remarkably similar to the stone atlatl weight from the Marpole site illustrated in Figure 43a and also bears a resemblance to the face on a undated seated human figure bowl from the San Juan Islands (Holm 1983:Figure 206). No other prehistoric examples of a snake with numerous feather crests has been found. The closest image is Figure 34 from the Marpole  176  site which shows a sinuous creature with folded limbs and a toothy grin.  Examples of  plumed snakes are found in Coast Salish art as in the houseposts from Musqueam illustrated in Suttles (1983:Figure 4:11 & 12).  Biederbost site A second example of wood sculpture was from Washington State was found at the Biederbost site, 45 SN 100, on the Snoqualmie River. A bowl with two faces was found embedded in clay by a private resident in 1959. It was not until many years later that the site was excavated and described as a "Marpole-type" assemblage from a fishing site or possible winter village with radiocarbon dates of 1940 ± 80 and 2000 ± 80 B.P. (Nordquist 1976). The site is within the territory of the Snoqualmie Lushootseed-speaking Coast Salish. The oval bowl in Figure 66 features two pointed prow-like ends, while in the centre two projections rise sharply to form handles. The bowl also has four short legs. Two similar prow-shaped wooden bowls were found at the Ozette site. The shape is also common in Northwest Coast oil dishes. Although only drawings and descriptions of the bowl could be examined (Nordquist 1960)  it is evident that deep engraving constituted the carving technique.  Two  anthropomorphic faces are carved into either ends of the bowl. On each face a feather-like motif radiates from the side of the head while elongated wedges define long vertical ridges running over the crown.  The prominent nostrils and wide open mouth with engraved  crescents at the outer margins relate to other human images from this period of prehistory. The only unusual features on the faces are deeply furrowed foreheads. The design described as a feather motif is identical to the plumes on the lightning-snake depicted on the Skagit River atlatl. By a unique coincidence, this same motif has been found on a fragment of a wooden artifact, described in the following section, found on the Fraser River 150 kilometers from the coast.  177  Figure 66. Biederbost wooden bowl. (Nordquist 1960:Plate 1).  178  Lower Fraser Region  The Esilao site The Esilao site, DjRi 5, is a former pit house village on the banks of the Fraser River on the border between the Coast Salish and the Interior Salish territories.  Inside a  prehistoric pithouse that had been burnt and buried was discovered the charred fragments of a carved box. Charcoal from immediately below or inside the pithouse was dated 2080 ± 130 and 2000 ± 120 B.P. Figure 67a and b show two fragments from the carved box. Figure 67a shows a deeply engraved design that is likely the remains of a feather motif, while 67b show deeply engraved long, parallel lines with engraved diamond shapes within a joined line. This motif is unique to the prehistoric period but it is similar to the depiction of long rows of teeth in Nuu-chah-nulth art. Figure 67c is the largest fragment of decorated wood. It shows an elaborate block engraving design that is similar to block engraving on spindle whorls depicting the wings of thunderbirds (Suttles 1983: Figure 4:14e,g,h,i). The piece is too small and too complex to interpret but it adds significantly to our knowledge of the use of wood as an artistic medium. The composition and carving style are identical examples of Coast Salish block engraving. The preserved wooden remains from The Fishtown, Biederbost and Esilao sites also indicate that certain aspects of artistic vocabulary and techniques common in Marpole sites were either common to other prehistoric cultures outside of the Marpole phase area, were traded out of the Marpole area, or certain attributes of the Marpole phase including artistic techniques extended south and east beyond the Strait of Georgia. Moving further into the British Columbia interior along the Fraser River there is only one more example of prehistoric art that pre-dates 1500 years from the Lehman site, EdRk 8, between Lytton and Lillooet. Figure 68 shows a stone maul with a zoomorphic head at the top that was thrust up by a farmer's plow. It was included in a description of the artifacts from Zone I at the Lehman site with a date of 2185 ± 150 B.P. (Sanger 1970) although it was not found in situ. Although it cannot be considered securely dated for the  Figure 67. a-c Esilao wooden block engraved fragments.  181  purposes of this sample it is likely approximately 2000 years old since the site was abandoned after this time. The features of a bird's head are pecked onto the handle of the maul in a manner similar to the 3000 year old zoomorphic maul from a Kamloops area site shown in Figure 12. Round eyes protrude from an eye socket with a downturned, pointed outer margin. The beak is slanted up and a pecked depression defines the mouth. On top of the head is an uneven hole.  Even though this artifact is very worn it shows a greater  mastery of stone sculpturing technique than any examples from this period on the coast, although finely made undecorated mauls are found in Marpole sites.  West Coast of Vancouver Island  Only one example of a decorated object has been found in dated deposits in the entire Wakashan area. Excavations at Yuquot, DjSp 1, in Friendly Cove on Nootka Sound reveal 4200 years of occupation at this site. A fragment of a decorated whale bone club was found in Zone Ha deposits dated 2101 and 1916 + 100 B.P. (Dewhirst 1980). The club handle in Figure 69a shows a deeply engraved mouth line which slants down at the outer edge, a short line for an eye, and a deeply notched margin on the upper head and neck. The notched margin once extended further up the crown of the head but was broken off. The notching which extends down to a groove outlining the head suggests feathers, a backbone or perhaps scales. Whale bone clubs are widely distributed throughout the Northwest Coast but are most common in the West Coast area where whales were hunted. This prehistoric example is a crudely executed image but it does show a clear relationship to the bird head images found on many of the historic clubs (Smith 1907:Figure 171).  182  Figure 69. Whalebone clubs. a, Yuquot club fragment; b, Boardwalk club handle (cast of original); c-d, Kamloops locality clubs (Smith 1923:Plate 37).  183  North Coast Sites  Prince Rupert Harbour sites Completing the survey from this period are eight decorated artifacts from Prince Rupert Harbour sites. A unique burial cache from the Boardwalk site, GbTo 31, a winter village of the Tsimshian, provides the second prehistoric whale bone club for this sample. The cache which is thought to date from 2500 to 2000 years old (MacDonald 1983:111) includes four clubs, a stone dagger and copper tubes. The decorated club in Figure 69b features a human head surmounted by an animal. The theme of the club is similar to Northwest Coast clubs of the early historic and late prehistoric period featuring the profile of two heads on the handle of a narrow-blade club. In the Boardwalk site example, as in several undated prehistoric clubs from the Kamloops region, shown in Figure 69c, the upper head is fashioned as a headdress worn on top of the human's head. The animal is long and narrow with a crooked tail and engraved dashed lines along its flank that suggest a snake or a weasel skin. A circular hump on the animal's back has a deeply engraved hole which may have held an inlay. This hump is a puzzling feature, although MacDonald interprets it as the human's ear I believe it belongs to the animal's body. The carving techniques used in defining the details of this club are identical to techniques found in Marpole period sites. The features are deeply engraved and large areas of negative space are cut away to create two form surfaces. The forehead, fore cheek, chin and nose are joined in a positive, primary plane; while the eye socket and crescent behind the cheek form the negative, secondary plane. The raised ridge around the eye suggests that an inlay once formed the eyeball. Although MacDonald describes the design elements in the face as ovoids and U-forms they do not follow the use of the terms as defined by Holm (1965:37-41). Although the eye form is similar to an ovoid it does not have a tapered formline with convex curves. The cheek shape is defined by a negative emergent wide  184  crescent rather than a positive U-form.  I do agree with the intent of MacDonald's  description which is to suggest that this club is surprisingly close in style to northern Northwest Coast art. The composition shows an elegant balancing of negative and positive space, with the positive design elements linked in a continuous flowing manner-tapering at the cheek to a thin line. Two other clubs dating from approximately 1850 to 1750 B.P. were found at the Boardwalk site. MacDonald suggests that their small size (approximately 17 cm) indicates a ceremonial rather than a functional use as clubs. Both depict a zoomorphic head at the proximal end of an antler tine. A protruding section of a smaller tine forms the striking portion of the club. One club, shown in Figure 70a, has the suggestion of a second eye and a beak at this striking edge, as well as two rows of rectangles forming a backbone motif that runs the length of the club (a similar double backbone motif is illustrated in Figure 34 from the Marpole site). A zigzag motif encircles the handle. The second club has a curved line behind the face which suggests the profile of a bird's head and wing. The carving style of both clubs is simple and rough although their design closely relates to finely made undated "slave killer" clubs found throughout the Northwest Coast (Wardell 1964: Fig. 155). A stone effigy in Figure 70b from the Baldwin site, GbTo 36, has several attributes that relate to Marpole sculptures. The deeply engraved concretion has two large eyes, curved horns, and thin fore and hind legs tightly flexed under the body. Prominent ribs run the length of the body as does a double ridge representing the backbone. The thin, folded limbs and the double backbone are similar to Figure 34 from the Marpole site while the double backbone is also found on the club in Figure 70a. It is likely that this creature represents the lightning-snake. Although there have been problems in dating this site this sculpture is thought to be approximately 2000 years old (MacDonald 1983:113). A second stone sculpture was found at the Dodge Island site, GbTo 18, in a component that is thought to be approximately 2000 years old (MacDonald 1983:115). The schist bark shredder (MacDonald 1983:Figure 6:29) has the general outline shape of a bird  185  Figure 70. Prince Rupert zoomorphic effigies. a, Boardwalk antler miniature "slave-killer" club (cast); b, Baldwin zoomorphic effigy on stone concretion (cast).  a  186  with the shape of a head and beak on one end and a tail on the other. A mouth line has been sawn onto the beak, otherwise there are no other distinguishable features. This object bears closest resemblance to mat creasers from the Coast Salish area although MacDonald states that other zoomorphic examples of stone shredders occur in undated collections from the Prince Rupert area. The Lachane site, GbTo 33, is a waterlogged section of a winter village midden. The site adds much to our understanding of fiber and wood craft of this period, notably the recovery of basketry, bent-wood boxes, bowls and bowl preforms that show similarity to Coast Tsimshian construction techniques. Whenever wooden artifacts are preserved they stand out from all other decorated objects from the prehistoric period. A red cedar handle from the Lachane site, shown in Figure 71, is no exception to this rule. Organic material surrounding the decorated wooden handle was dated 1630 ± 100 B.P. (Inglis 1976). The handle has two raised designs joined by a lower section of wood. Two pegs projecting from the underside would have affixed the handle to a base. Since the carving is curved it is thought that the base was the lid of a wooden bowl. On one section of the carving is an eye, while the other features a curved motif of what may be a feather or crest. The eye has an oval eyeball and pointed eye margin. Negative engraved wedges with angled depressions define the pointed margins, while a curved, engraved line separates the eye brow from the margin of the eye. This is the first prehistoric example of an eye motif where the eyeball, margin and eyebrow are joined in a continuous raised line. Although a similar-shaped eye with engraved, angled wedges is found on an antler bird's head from the Crescent Beach site dated 1000 years older (Figure 23). Can this carving be considered an example of a composition with formlines? In a manner like formlines the positive elements of the carving are joined in a continuous, smoothly curving line. Although there is a tapering of the eyebrow and of the curving lefthand motif, most of the line is of a uniform thickness with no attempt to avoid thickly joined points of juncture. The engraved wedges do not extend out into T-shapes that could shape  187  Figure 71. Lachane wooden handle with eye motif (MacDonald 1983:Figure 6:33).  0  2cm  188  the comer areas into elegantly balanced junctions of negative and positive space. This composition should be considered a formline although it is not quite its modern equivalent. (We should remember that the concept of formline is an arbitrary value defined by presentday researchers.) A stone club in Figure 72a from the Kitandach site, GbTo 34, is thought to date to between 2000 to 1500 years old. It's phallus-shape and engraved longitudinal lines are similar to a club from the Hagwilget Canyon cache on the upper Skeena River described by Duff (1975:Figure 101). A last decorated example from the same site in Figure 72b, thought to be 2000 years old, consists of an antler haft with a crude zoomorphic head at one end.  Summary  There is a wide array of art forms, carving techniques, and quality of workmanship in Marpole phase art.  At first glance these factors mask the underlying coherence of  Marpole art, but a closer inspection reveals a high degree of stylistic homogeneity both within and between sites in the south coast. There is a much smaller sample of art from areas peripheral to the Strait of Georgia, but significant parallels in artistic development are found here as well. Of equal significance is the fact that the entire carving repertoire of Central Coast Salish art can be seen in art of the Marpole phase of 2000 years ago. Table III shows the attributes that are shared among the sites discussed in the previous section, and the following summary interprets some of these findings. The sample of artifacts from 2400 to 1500 B.P. is the earliest period in which satisfactory assessment can be made of the art record; previous periods yielded too few artifacts to expect to find points of similarity between sites. The quantity of Marpole art is due not only to the many excavated shell midden sites of this period but also to prehistoric burial practices of interring the dead in shell middens with frequent inclusions of decorated  189  Table III. Prehistoric Style Attributes, 2400-1500 B.P.  n=102: DECORATED FORMS harpoon points human face pendants bird form pendants hafts and handles small pestle/pendants "slave killer" clubs whale bone clubs stone bowls stonefish/whaleforms atlatl weights browbands Bcratchers/blanket pins comb stone effigy human figurine mat creaser spoon  6 3 4 1 1 8 1 8 48 - + + + -+ - - + + + + - - + - - - - +  + ? + - - - - - + - + + + + + + ? + + - - - - +- + + - - + + - + - - - + - - ? + - - - - + - - + + +?  maul  CARVING TECHNIQUES  incising  engraving block engraving inlaying modelling pecking sawing cut-out antler DESIGN PRINCIPLES Formline Formsurface DESIGN ELEMENTS crescents negative C-shape T-shape U-form split U-form wedges Salish eye chevrons joint ovals/circles MOTIFS human animal  bird fish skull lightning-snake rib and Backbone bear wolf/dog feather  3 2 1 5 1 1 8  .  -+  .  .  .  .  .  .  .  - - - + - +  .  .  .  .  .  .  .  + + + -  .  + + . . + . .  .  + - - - - + - + + - - - - + - +  + - +  -  - - +  + - + - - - -  +  +  +  - + + + - - - -+ + - + + + - - - + + - - + + - + . + . + . . . + . . + . . . . +  + -  -  +  +  - - - + - - +  + + -  -  + + - -  + - -  - + +  + + . . . + . . .  + + + + + + -  +  + -  + -  -  -  + +  -  ? -  + - - - ? + ? ? - - - . + + - + - - + - . - + . . 4  -  -  + + - -  -  + + - -  - -  +-  -  191  objects and other personal items. Stratigraphic records show that some decorated items were directly associated with skeletons, while others were not, but in many cases this information was not recorded.  It is not known if personal items of adornment were  intentionally broken, burnt and scattered in Marpole phase burials, as is indicated for older burials at the Pender Canal site (Carlson 1987). References to the quantity and quality of burial items, the occasional finds of burial cairns and cists, and deformed and undeformed skulls, show that differential treatment was accorded to both the living and the dead, suggesting inherited status differences in prehistoric communities of this period. The carving skill and the overall design of Marpole decorated objects show a great range of variation. There are finely-crafted and skillfully composed sculptures in wood and antler that suggest the handiwork of part-time specialists. There are also many roughly executed decorated items with uneven workmanship and a poor sense of design. What is interesting is that even these crudely executed pieces show conformity to a standard repertoire of art forms, design principles and motifs. For example, Marpole phase pendants show a great range in quality of execution, but no matter what their quality, pendants depict either the front view of an anthropomorphic face or the profile of a bird. This intrasite and intersite uniformity of pendant motifs recommends this item as a diagnostic artifact of the Marpole phase. The variability in the quality of carving in pendants contrasts with that of browbands, which are uniformly well-made. Even examples of browbands with geometric decorations which where not included in this sample show even, deeply engraved zigzag motifs. The investment of carving skill in these items suggests that they were more highly valued than pendants or other modestly decorated items such as antler tine handles. This variability of quality suggests that pendants might have been carved and worn by anyone, while browbands might have been made by skilled carvers for higher status individuals. Perhaps browbands drew attention to the high, sloping foreheads of upper class individuals and were associated with the display of social position. Speculations such as this can be  192  addressed in future excavations with more thorough analysis of burial remains to assess the relationship between high status individuals and certain types of decorated objects. Compared to carving in antler and wood, the relatively poor quality of workmanship in Marpole phase stone art is puzzling. Except for two pendants and several incised stone plaques, Marpole stone art takes the form of sculptured bowls. The standardization of the bowl forms into four main types with widespread distributions (outlined by Duff:1956a) suggests that the bowls were well-established forms in the Marpole period. Bowls found in situ are simple, pecked compositions or roughly engraved and designed steatite objects. Both compare poorly with undated, finely carved steatite bowls found most frequently in the MidFraser region. The motifs and the general forms of the bowls are the same in both regions but the workmanship varies greatly. Duff interpreted this difference in quality as indicating a temporal evolution of bowl forms from older simple zoomorphic bowls and bowls with heads to the later elaborate seated human-figure bowls. Since the bowls from the Marpole site were the only ones to come from a dated assemblage, Duff used distributional information to make inferences about the temporal development of this sculptural form, working with the assumption that those bowl forms with the widest distribution~the simple Type A (Figure 40) zoomorphic bowls-were the oldest. Duff postulated that the less-refined renditions of the seated human-figure motif from the Marpole site were early coastal expressions of a motif which later became more elaborate as the style spread up the Fraser River where soapstone was available. What new evidence is there to add to Duff s study of prehistoric stone sculpture and its temporal and geographic origin? Since all four bowl forms occur for the first time at several Marpole phase sites-with at least one human-figure bowl found deep within the Marpole site midden deposits-there is at present no evidence to support Duff s notion that one of these bowl forms is ancestral to the others or that there was an evolution of bowl forms culminating in the seated human-figure bowl. Type B zoomorphic bowls and Type A and B human figure bowls have similar compositional styles. Both feature figures with well-  193  defined heads and facial features that project from the bowl, while poorly-defined limbs and other body features emerge in low relief from the outside bowl shape. A 3000 year old sculptured maul from the Thompson area and several small effigies from the Milliken site dating from 2400 to 2800 years old are evidence of stone sculpturing in the Mid-Fraser and Thompson River regions pre-dating stone sculptures from the coast. As this locality is also the only source of the easily worked mineral steatite, it would seem plausible that sculptured bowl forms could have evolved in this area. One hypothesis is that the seated-human figure form evolved in the Mid-Fraser and Thompson River region by the Marpole period and was poorly imitated by coastal groups who hitherto had only a stone pecking and grinding industry and poor quality stone for carving. Alternatively, undated finely composed and carved seated human figures from the south coast region such as the "Sechelt Image" shown in Figure 73, may date from the Marpole period rather than later periods, suggesting the possibility of well-made human figure bowls on the coast at this time. The principles of composition and the design motifs in "classic" steatite stone bowls from the Lillooet area fit within the range of compositional and motif styles found at Marpole phase sites. To give specific examples: the carving style used to define facial features in the undated seated human-figure bowls in Figures 74a and b, can be compared to the pendant/pestle from Musqueam North in Figure 62, a lignite pendant from Marpole in Figure 29f, and a sandstone atlatl weight from Marpole in Figure 43a. These shared traits suggest that the stylistic "vocabulary" used in creating the seated human-figure bowls of the Mid-Fraser region was also used by carvers at the Marpole site. The problem remains that the dated seated figure bowls are cruder examples of the seated human motif while the finely-made examples from the Mid-Fraser are undated. Are the coastal examples copies or prototypes? I believe there is insufficient evidence to support either an inland or coastal origin hypothesis, nor is it useful to view the Marpole sculptures as either copies or prototypes. We know that the most complex aspects of Marpole art-seen in the block-engraved boards from the Esilao site-reached to the gateway of the interior by  Figure 73. "Sechelt Image" undated human figure sculpture.  195 Figure 74. a-b, Lytton area undated steatite human figure bowls.  a  b  196  2000 B.P. and at some point, motifs and carving styles of the Marpole period were shared by prehistoric carvers of the Mid-Fraser River as demonstrated by undated sculptures from this region. Both regions had an indigenous sculptural tradition, one in antler and the other in stone.  The occurrence of the bowl form in both regions indicates shared artistic style  elements and presumably a similar ritual context during the Marpole period and perhaps later periods. The long standing sculptural tradition and access to easily worked steatite allowed interior carvers to elaborate on the seated human-figure form to include multiple animal motifs while coastal sculptors with their rough-grained boulders were forced to convey only the basic form of this motif. We must not forget that coastal carvers may have used wood to express equally complex and powerful expressions of the seated human figure bowl motif.  The evolution of carving techniques in the Marpole period: Looking over the period between 3500 and 1500 years we can see a gradual evolution of carving styles toward a Northwest Coast sense of composition. In the early period, dating from 3500 to 2600 years ago, antler carvings are deeply engraved and cut away creating sharply raised ridges defining such features as eyebrows and spines, and carved out areas defining eyes and mouths. If the carved out features are viewed as negative elements and the raised features as positive elements we can see changes over time in the treatment of these elements. Early period antler carvings show a sense of balance between the negative and positive elements, and some features are joined in a continuous raised surface such as the sternum and arms of the human figure from Helen Point in Figure 9. Beginning at the Musqueam N.E. site 2550 years ago and continuing in Marpole phase sites, there occur more sophisticated examples of the positive elements of a composition being joined to create raised positive form surfaces or lines (Figures 19 and 36a). Some compositions exhibit a balance and unity between the negative and positive elements in a design that mirrors Northwest Coast art. This "fully-developed" Northwest Coast sense  197  of design takes many forms in both two- and three-dimensional carvings: as formlines and emergent T-shapes in Figure 36b and 53a, as raised positive elements in the face in Figure 59, and as block engraved motifs in Figure 36a where the negative and positive elements are expressed as continuous lines. These types of compositions are rare but widespread among Marpole phase sites and sites of similar age in regions outside of the Gulf of Georgia. This style of joining negative and positive elements in a composition was not only expressed in different ways during the Marpole phase as shown above, but the compositional techniques were transferred from one medium to another. The sharply defined positive and negative features and steep junctures marking shifts in form surfaces in engraved antler sculptures can also be seen in pecked stone compositions. Although the stone medium must have presented a challenge, prehistoric sculptors attempted to peck out the same raised ridges and sharp junctures between the main parts of human and animal forms as seen in the "Marpole Image" in Figure 38a.  In a similar manner there are examples of incised  designs that utilize principles developed for deep engraving. In Figures 30a and 29f the lines around the eyebrows and behind the cheeks only make sense as flattened references to the two planes of the face. In these incised examples, what would be an abrupt juncture between the positive and negative elements of the face in a three-dimensional sculpture is defined as a line. The few wood carvings that have been preserved illustrate that the same techniques used in carving antler were used in carving wood. The wooden artifacts found at the Fishtown and Biederbost sites make use of the plastic nature of wood to create threedimensional forms but the artistic embellishment of these forms—the definition of wings, eyes and other features-is two-dimensional. Northwest Coast art in general follows this tendency. The four wooden artifacts from this period are all elaborate compositions which suggest a thorough mastery of the larger format and ease of expression that wood provides. The style of deeply engraved curving parallel lines found in Figure 36a from the Marpole site, Figure 66 from the Biederbost site, and Figure 67 from the Esilao site is  198  similar to the designs on several Coast Salish spindle whorls (Kew 1979:Figures 1, 4, 8, 16) and is one illustration of the many links of Marpole phase art to Coast Salish art mentioned in the previous section's analysis.  Design elements and motifs: There is rare but widespread occurrence of engraved T-shapes, wedges, crescents and formlines or form surfaces by 1500 to 2000 years ago. These design elements appear to have been shared by all regions represented in this sample, suggesting that these elements occurred at least as far north as the Skeena River and south to the Skagit River. Although all of these design features are first evident at the Musqueam Northeast site at approximately 2550 B.P., assemblage dates are too imprecise and regional samples outside of the Lower Mainland too small to allow any conjecture on the regional origins of these design elements. I believe it is possible to see the gradual emergence of the formline and form surface concept in carvings of this period although not enough carvings from the central coast have survived to tell us whether "modern" Northwest Coast formlines (as defined by Holm 1965) were in use 1500 B.P. In the south coast where there is a large sample of artifacts we can see that there were consistent choices of forms that were decorated and consistent motifs chosen to decorate these forms: pendants with human faces or bird shapes, browbands with eyes, and bowls with seated humans or crouching animals.  Engraved geometric designs-usually  parallel lines with ticks or triangles-commonly occur on Marpole phase antler handles or hafts and are similar to Interior Salish designs on digging stick handles. All of these traits allow a more refined understanding of Marpole phase art and should be considered diagnostic characteristics for this cultural phase. This period also sees an increasing trend toward compound facial features: eyes with lids, sockets and eyebrows; mouths with lips, teeth and tongues. In particular, eyeforms approach the fully-developed Northwest Coast  199  style. The Salish-style eye can be seen on antler and stone carvings from the Marpole and Garrison sites. A casual examination of petroglyphs from the Gulf of Georgia area reveals motifs, design elements, and compositional principles common in Marpole phase art. For example, a petroglyph located near the Crescent Beach site (Ham 1982:119c) utilizes an obtuse angle on a natural rock face to create a wedge-shaped anthropomorphic face as found in sculpture from the Marpole phase. Although it is difficult to comment on the presence and absence of motifs in regions with a small sample, some motifs such as a human head surmounted by a bird, appear to be restricted to Marpole phase sites while others are broadly shared. A creature that likely represents a lightning-snake, with a serpentine body, curved horns or head plumes, thin limbs and prominent teeth and eyes is portrayed in a stone effigy from the Baldwin Site in Prince Rupert Harbour, in antler and wood carvings from the Marpole site, and in a unique wooden artifact from the Skagit River. One of these objects is an atlatl and the other a possible atlatl weight. Several undated stone artifacts in private collections also appear to be shaped like atlatl weights and are embellished with snake and lightning-snake motifs (Keddie 1988) suggesting that the depiction of this creature was an attempt to aid hunting proficiency. Zoomorphic decorations on toggling harpoon points (in one case identified as a lightning-snake) from several Marpole phase sites also suggest that the images where used to affect the outcome of hunting and fishing activities. Similar decorated club forms occur at Yuquot and Prince Rupert, and another type at Prince Rupert and Marpole. In conclusion, many elements of the Northwest Coast art tradition are in place by 2000 B.P., along with irrefutable evidence that wood carving was well-established by this period. Design elements, principles of composition and mythical creatures were likely as familiar to prehistoric societies on the coast as to modern inhabitants. The artistic forms such as masks, decorated houseposts and rattles and their associated ritual and social  200  complexes may also have been in place or may have still been evolving but the prehistoric art record does not reveal this.  201  5.4 ART FROM 1500 TO 150 B.P.  The period from 1500 to 1000 years ago saw the full achievement of ethnographic economic and settlement patterns possessed by the aboriginal peoples of the coast. The number of local culture types following the Marpole phase is a reflection of regional groups pursuing the resources offered by local habitats. Mitchell (1971) places the Stselax phase on the southern mainland (Borden 1970), the Esilao phase at the mouth of the Fraser Canyon (Borden 1970), and the San Juan phase for the Gulf Islands (Carlson 1970) within the Gulf of Georgia Culture type representing the archaeology of the present Coast Salish people. The transition from the Marpole phase to the later phases was gradual, occurring at different rates in different locations, therefore from 1500 to 1000 years ago is considered a transitional period (Burley 1980, Matson et al. 1980). On the whole, there is an increase in assemblage homogeneity in the different regions of the coast and an unexplainably smaller inventory of stone, antler and bone artifacts. The use of wood and its poor preservation in prehistoric sites may partially explain the decreased inventory. Bone and wood most likely supplanted the use of stone and antler, particularly for fishing. While the use of chipped stone tools decreased during the Marpole phase (Matson et al. 1980), by the late period it was reduced to a minor industry. Decorated objects are generally rare compared to the previous phase. This may be a reflection of the increased use of wood and changed mortuary patterns (mortuary houses, isolated grave sites and tree burials) discouraging the archaeological recovery of decorated objects. The near-perfect preservation of several houses at the village of Ozette, dating to approximately 1719 A.D. (Samuels 1983:24), reveals a large inventory of mainly wooden decorated artifacts.  Throughout the coast there is evidence of weaving and spinning  industries in this period. Population increases and likely also warfare since fortified villages and defensive sites are characteristic of the late period.  202  In the Interior Plateau archaeologists have drawn similar temporal divisions between the last two prehistoric horizons (Plateau archaeologists prefer to use the term horizon instead of cultural phase). The Plateau horizon, which begins approximately 2400 years ago at the same time as the Marpole phase, is replaced by 1200 B.P. by the Kamloops horizon (Richards and Rousseau 1987). As in the relationship between the Marpole phase and the ensuing late phases on the coast, the Kamloops horizon is seen as a continuation and intensification of food acquisition and storage techniques of the previous horizon.  The  atlatl's replacement by the bow and arrow is seen in the ubiquitous side-notched point of the Kamloops horizon. Several new types of barbed harpoons and leisters were used for catching migrating salmon.  There was extensive use of ground stone, especially steatite and  soapstone, and an increase in bone, antler and copper artifacts.  Changes in Kamloops  horizon housepit remains suggest the use of mats and light poles in some localities in addition to the heavily built earth and timber pithouses. There is evidence in both coastal and interior sites of increased trade between the regions. Whale bone, dentalia and slate from the coast were traded inland while steatite, nephrite, and vitreous basalt were traded to the coast. Archaeologists working in both regions have hypothesized that social organization may have been more complex in earlier prehistoric periods. Based on the decrease in decorative objects in the late period, Borden (1983:159) saw the Marpole phase and artistic climax unmatched by the late prehistoric Stselax phase or historic period. In the Mid-Fraser region, both the size of housepits and pithouse villages appear to have been larger between approximately 2000 and 1000 years ago, leading some archaeologists (Stryd 1973, Hayden et al. 1985) to postulate that social organization and ceremonial complexes may have been more complex. Decorative  objects continue to be common in burials in the Mid-Thompson and  Fraser region in the Kamloops horizon leading Sanger (1970) to view only the historic period or protohistoric period as being markedly different in its level of social organization. Burial customs may be the key to understanding the apparent decrease in coastal art of the late  203  period, and the continuation of a lively artistic tradition in the western plateau.  The  following examination of the art evidence will assess the differences between the last prehistoric period and the preceding ones. The art evidence from the last prehistoric period is a small, poorly dated sample. Whereas many older prehistoric sites occur slightly inland along former shorelines, most late period sites are in choice locations along rivers and coastlines where recent settlement continues. Late period prehistoric villages such as Stselax have been continuously occupied and have evolved into Indian reserves and present-day communities.  The uppermost  prehistoric deposits have been disturbed by the activities and settlement patterns of the past 200 years. The following sample includes a large number of artifacts found by private residents and "pot hunters" which were  later donated to museums or shown to  archaeologists working at site excavations.  Supporting archaeological information that  might provide a context for the art objects in the following sample is scanty. Two artifact forms, whale bone clubs and antler human figurines, occur almost exclusively as isolated finds and are therefore discussed together rather than in the context of individual sites.  Gulf of Georgia Region  Maple Bank site Although there are numerous shell middens on southern Vancouver Island the early settlement of the area by Europeans and piecemeal salvage and survey work has resulted in low recovery of datable decorated artifacts, yet stone and antler carvings in private collections attest to an active prehistoric art tradition. The Maple Bank site, DcRu 12, is in Esquimalt Harbour which was occupied by the Songhees, a Straits Salish group. The shell midden site has a Developed Coast Salish component within which was found two decorated  204  artifacts (Keddie 1976). Figure 75a shows a small fragment of carved antler. The artifact is from Zone A which lies above the Zone B deposit dated from 1460-1140 B.P. (Keddie pers. comm.) Although only four engraved lines are visible, they form the configuration of part of an eye and mouth. A suggested reconstruction of the design is included in the illustration. Of note is the way the pupil is joined to the upper eyelid using a negative C design element. Other prehistoric examples of this eye shape are illustrated in Figures 86a, 103a, 108b, and 109c and are discussed later in this chapter. Figure 75b illustrates one of the more important finds in the prehistoric art record. The comb was found at the bottom of Zone B dating from 1140 to 1460 B.P. (McMurdo 1975) and is associated with a date of 1310 ± 70 (Keddie 1987). The handle of the comb depicts the profile of an animal that likely represents a dog or a wolf. Distinct to this piece is the lively posture of the animal: the back curves, the creature's paw rests under the chin, and its bent foreleg rests on its folded hind limb. Although the depiction of animals with cutout shapes defining the limbs is found in antler sculpture at the Pender Island site dating over three thousand years old, four-legged animals are not a motif found in Marpole phase sites. At variance to sculptures from previous periods is the attention given to the definition of the body and limbs which are given perfunctory treatment in most prehistoric sculpture. The wrist joint is marked by an engraved crescent and long tapering claws are portrayed on the paws. A double line defines the backbone while engraved crescents define the ribs and short straight spurs off the backbone may either refer to the opposing set of ribs or vertebrae. The folded hind limb is defined by an engraved wedge element. The head has a square slightly upturned snout in the style of West Coast and Salish art. A nostril and round eye and barely visible upper mouth margin define the facial features. A rounded ear is exaggerated in size to allow for a suspension hole. This carving has a number of features that give it a modern aspect. The engraved crescent at the wrist and the long, tapering claws is a typical forepaw motif in Northwest Coast art. Equally interesting is the tightly flexed lower limb which forms a U-shape within  205  Figure 75. a-b Maple Bank antler carvings.  206  which the angled engraved line almost forms a split-U. The natural curving quality of the animal's pose is found in Coast Salish wooden grave monuments from Saanich and Musqueam depicting fishers (Suttles 1983:Figure 4:13b & c). Keddie has suggested that the elongated limbs and posture of the creature represents "a transformation being, that is, a supernatural creature in transition between animal and human forms" (Keddie 1987). Although the forelimb is unusually elongated, none of the features of this carving show human characteristics. But Keddie proposes an interesting interpretation: the figure may represent a dog and the important transformation of wool into blankets and wealth.  A  similar comb from the Prince Rupert site is shown in Figure 102.  Stselax Village site The Stselax village site, DhRt 2, excavated in 1959, is the third site in this sample on the Musqueam Indian Reserve and lends its name to Borden's last prehistoric phase designation for the Lower Mainland. A radiocarbon date of 660 ± 130 B.P. (Borden 1968) was obtained for deposits near the bottom of the excavation. Borden's field notes reveal some of the problems inherent in excavating late period sites. At Stselax the shell matrix of the site formed the foundations of historic houses. In mixed deposits containing historic and prehistoric material Borden assigned artifacts to the historic or prehistoric period based on his judgement of the quality of decorative carving.  I believe that his assumptions and  assignments to the historic period should be re-examined--in particular, the deeply carved items that were assumed to have been made with steel tools. Figure 76a-c from the Stselax site, shows the decorated portions of three antler objects presumed to be blanket pins. Example 76a was found  in situ and portrays the  outline shape of a head with a short incised line on the lower beak or jaw. Example b, found during a surface search of the site, depicts a bird's head with a slightly upturned, open beak. A round pupil and eye margin defines the eye and an engraved crescent with an  Figure 76. Stselax antler carvings. a-c, blanket pin/scratcher fragments; d, antler fragment.  208  elongated spur defines the lower margin of the beak. Figure 76c, found in screened earth south of the site, is elaborately engraved and depicts an animal with short, pointed ears behind the head. It is finely carved and deeply engraved with an oval eyeball and a raised, positive elongated eye margin. The margin of the mouth is also defined with a continuous deeply engraved line. The engraved lines are V-shaped in cross-section with the angle of the V widening in places to shape the positive elements of the design. The small ears and slanted, elongated eye extending along the length of the head and turning down at the outer edge, suggest the depiction of a lightning-snake. Figure 76d, a site surface find, is a small fragment of a carved design showing a mouth and corner of an eye. Figures 76c and d are similar in the optically stimulating interplay between the negative engraved lines and the positive raised formlines. The carved fragments in Figure 77a-b further illustrate the sophisticated balancing of positive and negative line. Figure 77a is a curved antler object of unknown use found in situ in the house floor stratum.  Short projections on one end show no signs of wear  indicating their function, while the other end has broken off. The object in cross-section has convex and concave faces, with the convex side featuring deeply engraved crescents and wedges in an intriguing non-representational design. The combination of two wedges joined at their apex with a crescent element below is repeated twice and bounded by long curving lines. Another engraved wedge with a raised positive rim like an ear motif can be seen in the lower left corner. Three negative C elements run in a band design near the top of the carving.  The graceful composition looks remarkably freshly carved.  Examples of what  appear to be non-representational compositions are found in Coast Salish art such as on decorated spindle whorls. As well, the design element of two joined wedges is found on some examples of Coast Salish sxwayxwey masks (Suttles 1983:Figure 4:5d). Figure 77b, made from the epiphysis of a whale vertebra and found in screened material from the excavation, has a curved margin that may indicate a spindle whorl fragment. Deeply engraved lines in a block engraving style define the corner of an eyelid and  209 Figure 77. Stselax antler carvings. a-b, block engraved carvings; c, anthropomorphic comb.  a  210  a slightly down-turned beak. The lower margin of the beak is further defined with an engraved line. Figure 77c shows one of two combs found in  situ at the Stselax site. The handle of  the comb has the roughly carved features of two eyes and a nose. Notches in the sides and on top suggest a neck and lobed forehead. Five anomalous horizontal lines run across the face (in an identical manner to the lignite pendant from the False Narrows site in Figure 56a). The other comb is an incomplete lower fragment with a non-representational design of raised horizontal and vertical lines. Borden used the zoomorphic comb as evidence of an artistic decline in the late prehistoric period: On pondering this artistic expression of the late period one is struck with the realization that it would be difficult to depart further from what one would normally recognize as Northwest Coast art style than is exemplified by this anthropomorphic comb...(1983:163). Borden attributed the well-made decorated objects from the Stselax site to the historic period because he thought that their deep engraving style suggested the use of steel tools. Some of his assignments may be correct since the site was occupied in the historic period, but other undisputed examples of fine carving from the late prehistoric period, such as the comb from Maple Bank in Figure 75b, prove that although the art record is smaller, the quality of execution did not decline. Figure 78 is a carved antler fragment from uppermost historic-prehistoric deposits at Stselax. An animal's head and forepaw is carved on both sides of the fragment. As Borden (1983:165) illustrates, the carving is identical in composition to animals portrayed on a nineteenth century Coast Salish spindle whorl. Similar details include an engraved line that widens into a wedge shape from the nostril to the eye, and becomes a crescent defining the cheek.  In the prehistoric or early historic fragment these elements are part of a  continuous line, whereas in the spindle whorl there is a small break dividing the two elements. The eyes are similarly composed with oval eyeballs and lenticular eye margins denned by deep engraving. In the prehistoric example the engraved space inside the eye is  Figure 78. Stselax zoomorphic block engraved design.  212  V-shaped in cross-section with a T-shape at each corner. Although the section showing the paw is worn, it appears that the claws were defined with short incised lines. Figure 79 shows a whale bone club from a private collection recovered while excavating a house basement on the Stselax site.  The well-preserved carving is deeply  engraved in a block engraving style on both side of the club. The lower head depicts a bird with a short hooked beak like a hawk, a large, round eye and a forehead that extends to the top of the beak. A curving engraved line forms both the top of the forehead of the lower head and the margin of the lower beak of the upper head. The upper head has similar features with the addition of an anomalous circular element or possible eye motif at the back of the two heads. The motif is wrapped around the end of the club and appears on both sides. This club conforms to the shape and motifs of "West Coast" style whale bone clubs the handle of which shows the profile of two bird's heads or a human head surmounted by a bird's head. Although over a dozen of these clubs have been found in or near late period sites and historic Coast Salish settlements, nothing is known about their manufacture or use among the Salish. Other prehistoric occurrences of this artifact are discussed later in this section. Although there are seven other sites in the Lower Mainland region with assemblages dating from the last 1000 years, none of these sites have yielded discoveries of decorated artifacts.  in situ  Fortuitously, a wooden club shown in Figure 80 was  dredged up from the north arm of the Fraser River and dated by radiocarbon AMS method to 1000 ± 130 B.P. (Fladmark et al. 1987). Although the object appears to have eroded during its time in the river obscuring the finer details of the carving, the overall composition is preserved. The club has a resemblance to Northwest Coast "slave killer" type stone and antler clubs with the form of a projecting blade approximately one quarter of the way down the shaft (earlier prehistoric examples from the Prince Rupert Harbour site are illustrated in Figure 70a). The Fraser River club features a rounded, projecting ball for the striking head of the club. Grasping this ball in its forearms is a creature with a prominent hooked bill and limbs in a crouching posture. Incised lines representing claws are visible on the end of the  Figure 79. Stselax whale bone club fragment.  214  Figure 80. Wooden club from the Fraser River.  215  limbs. On the handle of the club is carved another head, perhaps representing a snake, from who's open mouth projects a forked tongue. Behind this head is a fringe or collar decorated with short lines radiating around the face. Both figures show three-dimensional modelling of the facial features with prominent, oval eyebrows and well-defined mouthparts. The twoheads and horizontal crouching posture suggests the depiction of a lightning-snake or twoheaded snake.  Helen Point site One object found in situ and two beach surface finds come from the Helen Point site on Mayne Island, DfRu 8, on the Saanich Indian Tsartlip reserve. The late prehistoric component at the site is designated the San Juan phase and represents the prehistoric culture of the Straits Salish (Carlson 1960). Figure 81a was found in the San Juan phase component of the site dating from approximately 1150 to 750 years old (Carlson 1985, pers. comm.). The carved antler fragment is deeply engraved and presents either a frontal or profile view of a human face. All sides of the composition show breakage so it is difficult to recreate the original design but the top margin may have included a suspension hole for the carving.  The back of the head and forehead has been ground to a smooth edge which  suggests a profile view, although the full frontal view of the human face is more common in prehistoric art. A lenticular eye points upward at its outer edge and is surrounded by an engraved line which describes both the shape of the lenticular eye socket and the surrounding lower cheek and forehead area. A wide nostril, open mouth and mouth margin are defined. A barely discernable line on the cheek indicates that the chin area may also have been defined. Although the eye shape is somewhat unusual, the configuration of the facial features is within the range of styles seen previously. The carved antler fragments in Figure 81b-c are undated surface finds. Example b is a well-made carving of a human face surmounted by a bird's head. The features of the face  217  rise in low relief and are reminiscent of antler sculptures from earlier periods with steep surface junctures defining the features or like prehistoric stone sculpture rather than the deep engraving style more common in antler carving of the Marpole and later phases. The eyebrows and nose form a continuous raised ridge. The area underneath the eyebrows is carved away, emphasizing the prominence of the round eyes. The division between the forehead and hair is also deeply carved. The head and neck of what appears to be bird is portrayed on top of the human head. What may have been the rest of the bird's body has broken away, as has the right side of the human's face and the lower body. The motif of a human head surmounted by a bird is found on two examples of antler sculpture from the Marpole site, shown in Figure 30, as well as several 3000 year old sculptures from the Pender Canal site (Carlson 1986, pers. comm.). This carving may address the same theme. The carving style suggests that this piece might belong to one of the earlier components of the site. Carlson (1983:203) suggests that this carving represents a mask and may relate to the Coast Salish sxwayxwey complex mask which portrays birds' heads above the face mask. Figure 81c shows a carved antler fragment of an unknown object with an eye motif. An uneven line describes either a lenticular eyelid or eye socket within which sits a prominent engraved eye with a carved out centre.  Pender Canal site The Pender Canal site, DeRt 1 and 2, from which came several prehistoric antler sculptures discussed in an earlier section, also contained two sandstone sculptures thought to be approximately 1000 years old (Carlson 1986, pers. comm.). The sculptures shown in Figure 82a-b are oval in cross section and present the profiles of whales or fish with prominent dorsal fins. The size and shape of the fins, particularly the bottom example, suggest the depiction of killer whales. Short incised lines indicating mouths appear to be the only embellishments of the sculptures.  Similar sculptures found at Pedder Bay on  Vancouver Island are shown in Figure 86b (see also Smith 1907:Figure 188 for Saanich area  218 Figure 82. Gulf Island decorated objects. a-b, Pender Canal sandstone whale effigies; c, Montague Harbour antler zoomorphic haft.  219  examples). Petroglyphs depicting whales and porpoises are found on Gabriola and Hornby islands, north of Pender Island within Coast Salish territory (Lundy 1974:108). Jenness (unpub. notes) states that whales were hunted by Straits Saanich groups.  Montague Harbour site Figure 82c is from the Montague Harbour site, DfRu 13, on Galiano Island, in the territory of the Cowichan Halkomelem (Mitchell 1971:20) or the Straits Saanich (Suttles 1985).  It is from a Developed Coast Salish component dating from 790 ± 130 B.P. to  approximately 200 B.P. (Mitchell 1971). The underside of the curved antler section has a V shaped groove which suggests this object's use as a knife haft. The curving handle has several incised lines outlining a simple mouth or beak, two crossed lines for an eye and a line forming a right angle which may indicate a wing or flipper. The general form of the object as well as its simple manner of decoration is unique.  English Camp site Five decorated objects were discovered at the English Camp site, SJ 24, situated on Garrison Bay next to the Marpole phase Garrison site. Radiocarbon dates range from 1500500 B.P. (Kornbacker 1989:85) The most interesting decorated object from this site is an antler digging-stick handle with a zoomorphic face, shown in Figure 83.  Digging-stick  handles with this shape are common interior plateau artifacts and are often incised with geometric designs. This coastal example features the face of a long beaked animal, likely a bird, on the distal end of the antler tine. A round eye and long mouth have been deeply engraved while two rows of diagonal lines along the mouth or beak margin have been incised. The margin between the head and neck is defined by a abrupt change in the surface plane, as well incised straight lines mark each side of the head and the end of the beak. The simple zoomorphic image resembles a carved antler tine from the Whalen site. Figure 84a  Figure 83. English Camp antler digging stick handle.  221 Figure 84. a-b, English Camp antler carvings.  222  shows a fragment of antler or whale bone upon which the features of a face are deeply engraved. A diamond-shaped eye has a sharply downtumed spur which runs to the bottom of the face. This line is paralleled by engraved lines running from the cheek and forehead. The corner of a mouth occurs along the broken edge of the carving. Although the fragment is too small for either a functional or representational identification the carving style and eye shape is similar to other carvings of this period. Figure 84b shows another incomplete face in profile although this object appears to be unfinished rather than a broken fragment.  The lenticular eye with a carved out eye  socket area creating a second form surface is part of the carving repertoire seen in Marpole sites. Two fragments of siltstone with incisions depicting lenticular eyes, a mouth and other unidentified lines also come from this site.  Pedder Bay Site Although prehistoric art discoveries from the Saanich Peninsula are plentiful, few of them come from datable contexts. Keddie (1984) states that most sites with late period components occur in different locations than those with Marpole and Locarno Beach phase components. This allows the surface finds from these sites to be placed with some certainty within the late period, although their relationship to the historic period remains unknown. The Pedder Bay site, DcRv 1, has one of the earliest dates for a Developed Coast Salish phase component, dating from 1580 B.P. (Mitchell 1971:62) to the historic period (Keddie 1984). The site was formerly a large fortified village of the Straits Songhees, with burial cairns behind the village. A carved antler fragment of what may be a browband or bracelet was found in  situ but the design is too fragmented to analyse. Seven artifacts were  found at the site by private collectors and although they cannot be used for this study's dated sample, they provide useful information to add to the small number of dated objects from the late period.  223  Figure 85a is a highly-polished and carved antler figure of a four-legged creature. A tunnelled hole in the creature's belly may have allowed the sculpture to be suspended in a pendant. The creature's wide mouth gaps open and its tongue curves out and under the chin. Lenticular eyes and the lips are engraved, while a nose with nostrils projects from the tip of the snout. The creature's body is short with a pointed upturned tail end. A raised ridge like a collar separates the head from the torso. Thin, flexed legs project in low relief from the body. Between the legs is a raised surface with an incised lines that may either represent a rib and backbone motif or wings. What is unusual is the way the design field is raised slightly from the side of the creature. The identical motif on a raised surface is found on a steatite sculpture of a bird from the Chase burial (Figure 124b). The features of this sculpture from Pedder Bay suggest the portrayal of a lightning-snake. The following objects from Pedder Bay are part of private collections photographed by Royal British Columbia Museum archaeologists.  Figure 85b is an unusual antler  composition depicting two human figures. It appears that the head of one figure and the limbs of the second figure have broken off, but a small head at the bottom left of the composition may be the head of the figure on the left.  At first glance, the general  configuration of the two bodies and the facial features of the larger head are reminiscent of small Tlingit sculptures. The motif of two humans in profile is unfamiliar to Salish area prehistoric or historic art. On the other hand the style of the eye shape is found in several examples of prehistoric stone art from the Salish area. The raised rim of the eyelid closely conforms to the large oval eyeball-a common Salish stylistic feature~and the corners of the lids are in the centre of the eye, without the slight downward droop of many northern Northwest Coast eye forms. Both figures have deeply engraved ribs. Sections of antler have been cut away and the spaces smoothly finished giving a rounded three-dimensional effect to the figures rather than the "cookie-cutter" appearance of many antler prehistoric sculptures. Figure 86a is a rare unbroken antler blanket pin or ritual scratcher with the same motif of an animal's head with a wide, open mouth as is found on pins from the Stselax site.  224  225  Figure 86. Pedder Bay decorated artifacts. a, antler pin/scratcher; b, sandstone fish/whale effigy; c, stone zoomorphic head.  a  226  As in Figure 76c, the eye is elongated to fill the entire head. In the Pedder Bay carving, the eye form is sophisticated in conception in that much of the design is hinted at rather than explicitly depicted. Only the lower edge of the pupil and the inner corners of the eyeball are defined by negative C and wedge elements. The entire eyelid is only suggested by the shape of the remaining positive space in the design field.  The definition of "inner" negative  engraved space without the definition of the outer lines of features is a common design device found in Coast Salish spindle whorl and rattle designs (Suttles 1983:Figure 4:14e, j & 1). Figures 86b-c are stone effigies.  Figure 86b appears to be made of siltstone or  sandstone similar to the two whale forms from the Pender Canal site shown in Figure 82a-b. The Pedder Bay effigy may refer to a fish since the dorsal fin is less pronounced. Like the examples from Pender Island, the stone is simply decorated with incised lines indicating a mouth and eye.  Figure 86c, made of siltstone is a zoomorphic head with a lenticular  engraved eye socket and eye. A lip margin and open mouth have also been engraved. Two lines run from the corner of the mouth to the back of the head. Another line runs from along the top of the head. Faint lines behind the eye are difficult to discern and likely connected with a section that has broken away. Two other small stone effigies from uncertain contexts that may relate to the Pedder Bay object are shown in Figure 87a-b. Figure 87a was placed at the shoulder of a bark wrapped burial found at Boundary Bay (Leisk fieldnotes 1930, Vancouver Museum).  The shape of the siltstone object suggests a penis, although the  elongated downcurved eye margin, protruding tongue, and three incised lines representing the backbone also suggests the depiction of a lightning-snake.  A hole has been drilled  through the tip of the tail section. Figure 87b, made of steatite has two spurs running down from the outer eye margin and incised nostrils like the previous example. A small hole has been drilled through the nostrils. Its features are also similar to the stone image from Pedder Bay. It is an undated find from the Qualicum Beach locality on the east coast of Vancouver Island.  227  Figure 87. Undated stone effigies. a, Boundary Bay locality siltstone carving; b, Qualicum Beach steatite carving.  228  An antler mat creaser with two undecorated zoomorphic heads, similar to Coast Salish examples, was found at the site but is not illustrated. Lastly, Figure 89b an antler fragment of a human face is similar in carving style and motif to antler figurines found in the Skagit Delta, Gulf Islands and Bella Coola area. It is a burnt, broken fragment carved in a block engraving style. A stylistic analysis of this and other related figurines is detailed later in this chapter.  Cadboro Bay Site The Cadboro Bay site, DcRt 15, is a historic village of the Straits Songhees as well as a prehistoric village midden with associated burial cairns. Although radiocarbon dates for the site are problematic, there are two distinct components thought to represent both a late Marpole and an early Developed Coast Salish phase (Wilmeth 1978:57). Decorated artifacts from the site include two whale bone clubs described by Smith (1907:Figures 167 & 168) which are included in a later discussion, a carved antler fragment found in situ in surface deposits, a simple zoomorphic stone club, and a siltstone fragment with a zoomorphic head found by private collectors. The antler fragment found in situ is illustrated in Figure 85c. It depicts the profile of a human head on top of which is an animal's head and fore limb. The broken margins of the carving show that the composition once included the lower portion of the human figure and more of the animal's body. The motif recalls a similar theme depicted in the whale bone club from Prince Rupert Harbour (Figure 69b), and more closely resembles antler carvings from the Marpole site shown in Figure 30. The definition of a raised ridge curving around the forehead and cheek is similar to the incised line on Figure 30a, and to an undated antler carving from English Camp, San Juan Island in Figure 84b.  Since the motif relates to  material from Marpole-age sites this fragment may be from the late Marpole component at Pedder Bay. In the Cadboro Bay carving the identity of the animal on top of the head is  229  uncertain. Its blunt snout suggests an animal rather than a bird or supernatural snake and its foreleg resembles a seal's flipper. The two engraved circles on its back may be part of a backbone motif. The convention of depicting the backbone as a line of deeply engraved circles is found on stone sculpture from the Columbia River region on the border between Washington and Oregon states (Wingert 1952:Figures 5 & 8).  Antler Figurines and Whale Bone Clubs  In late period Gulf of Georgia sites there are two unique forms exhibiting an unusual degree of stylistic conformity: whale bone clubs and antler human figures. Examples of both forms occur mainly as isolated finds in poorly dated sites on Vancouver Island, the Gulf Islands, and the San Juan Islands. But sufficient examples exist to allow a fruitful analysis of these two widespread prehistoric artifact types. For this reason these two decorated forms are discussed as a group rather than in the context of individual sites.  Antler human figurines Carved antler figures occur in at least 14 separate archaeological contexts. Nine of the carvings are from sites in the southern Strait of Georgia and northern Puget Sound. One carving is from the Bella Coola area and one was found inland in the Yakima Valley in Washington. Numerous similar carvings come from late prehistoric sites along the Lower Columbia. Table IV lists the attributes of the carvings. Although there is some variation in the Columbia River sculptures general characteristics of figurines that occur as the same site have been lumped together in the chart. General characteristics of the carvings shown in Figures 88-94 include: the depiction of an anthropomorphic figure usually wearing a short skirt; well-defined arms, legs, hands  Table IV. Late Period Human Figurines.  *W*\, / //// ///////////// 1. Head shape (half offrontalview) 2. Headdress  + - - + - +  + +  or hair ornament 3. Eye shape  O  4. Earrings/ear holes 5. Armbands/tattoos  4 - 3 - - - 1 - 3 - - 2 1 2 +-+ +--+  Legbands/tattoos  O S - ®© ®  ®  •  ~' ® ^ ^  °  + + + +  +--+  +  6. Ribs  - - - +  +-+-  +  7. Clavicle/necklace  + .+ + + - + - - - + - + + -  8. Skirt  + + + + ?-  - - + + - - + +  9. Female genitals genitals 10.Male Navel  + - - - + ?  11. Arms: straight folded 12. Circle at wrist joint  ' - + - + + + + + + + + + + + - - - + +  13. Line down forehead and nose  - - - + + + + -  +--  231  and feet; a head that slopes in towards the forehead; a coiffure often decorated with ear and hair ornaments; tattooing or ornamentation of the neck, wrists, arms, legs and chest; and depiction of the ribs, sternum, and clavicle. There are regional variations in the figurines. In the carvings from Bella Coola, Georgia Strait, and Puget Sound the arms are folded over the chest while in the Columbia River examples and one carving from the San Juan Islands, the arms are held straight at the sides of the body. Elaborate headdresses are a feature of Columbia River and Yakima examples. These southern carvings and one of the northern examples depict the sternum, clavicle, and rib cage. The carvings from the Puget Sound and Gulf of Georgia have exaggerated pointed heads, and in some, a line that runs from the top of the forehead and continues down forming the bridge of the nose. The head shape likely calls attention to forehead deformation which in early historic times was a mark of upper class birth. The neck is defined by two deep notches, giving the jaw a squared-off look. Carvings from this region exhibit deeply engraved carved lines with distinct engraved, rectangular-shaped eye margins.  Male or female  genitalia are depicted in four carvings from these Central Coast Salish area sites. Although none of the carvings are closely associated with radiocarbon dates, over half of the carvings were found in situ in sites with broadly defined ages. All of the available temporal evidence suggests that the figures are approximately 1000 years old. Figure 88a from the Anutcix site, FaSu 10, in the Kwatna locality, is associated with the Anutcix phase assemblage dated to between 1750 and 750 B.P. (Carlson 1983:124). Carlson posits that the elk antler figurine may have been obtained in trade since elk is not a local resource and is not found in prehistoric contexts. Figure 88b was found by a private collector at Montague Harbour, DfRu 7, on Galiano Island.  Archaeological sites in the locality date from  approximately 200 to 3000 years old (Mitchell 1971:64). This figure appears to display ribs, a deeply engraved line that may represent the clavicle, and male genitals.  Figure 88c was  found by Harlan Smith in an eroded section of shell midden at the Maple Bank site, DcRu 12 (Keddie 1987). The site dates from approximately 1460 B.P. to the historic period. The  Figure 88. Antler human figurines. a, Anutcix (Carlson 1983:Figure 7:2); b, Montague Harbour; c, Maple Bank.  233  Maple Bank figure has female genitals and what appears to be facial tattoo markings. Both Figures 88b and c have a line extending down the forehead and bridge of the nose. Figure 89a was found at DeRu 1, at the northern tip of the Saanich Peninsula, by a private collector. The lower half of figure has broken away, but the remaining portion exhibits typical features. Extreme head deformation is suggested by the head shape. Facial tattoos and a line running down the forehead are similar to the previous example. Deeply engraved markings on and below what may be a skirt are difficult to interpret.  Two  triangular notches at the bottom broken margin may indicate genitals. Figure 89b from private collections found at DcRv 1, Pedder Bay, likely date with the site to between 1580 B.P. to the present. Even though the figure is not complete, the head shape, facial features and hand held under the chin conform to other examples. Figure 90a, one of the best preserved figures, was found in a cave on Sucia Island, the northernmost island in the San Juan group. Nearby sites at Fox Cove and Fossil Bay have dates of 1710 and 1514 B.P. (Burley 1979, Kidd 1969). Notable features include: traces of red ochre on the carving, engraved ovals at the wrist joints, and two raised chevrons at the neck that may represent a necklace or beard. The eyes and possibly the wrists once held inlays. Two almost identical figures shown in Figures 90b and 91a were found in Skagit River Delta sites at Conway, 45 SK 59, and Fishtown, 45 SK 33. The Skagit examples are not as well carved and lack engraved ovals in the wrists. Figure 90b from the Conway site was found in situ during test excavations. The site is dated between 1270 and 800 B.P. (Thompson 1978). Figure 91a from the Fishtown site was also found in situ and is thought to date to a approximately 1000 B.P. (Carlson 1983:124). Other undated carvings executed in a similar style may related to the material presented here.  Figure 91b is a whale bone carving from DcRu 32, at the entrance to  Victoria Harbour. Uncovered in 1914 during excavation of the Ogden Point breakwater, the carving shows a figure with facial features and upraised arms identical to the carvings  234  235  Figure 90. Antler human figurines, a, Sucia Island; b, Conway.  236 Figure 91. Antler human figurines, a, Fishtown; b, Victoria Harbour.  237  previously described. The carving does not depict the lower limbs, instead two bird forms decorate the lower portion of the composition. Carlson reports the discovery at the Pender Canal site of an antler human form with a fringed skirt and cut out legs similar to the examples discussed above (1986). Earspools and the figure of a bird decorate the human's head. This carving was found in a burial area dated over 3000 years old. A number of similar antler human figurines have been found along a stretch of the Columbia River, on the border between Washington and Oregon, from Sauvies Island to the Deschutes River. This area spans the boundary between the forests west of the Cascades and the semi-arid Columbia Plateau to the east, and is within the territory of the Upper Chinook Indians. It should be noted that this section of the river also marks the cultural boundary between coastal groups and the Plateau Sahaptin. As a physiographical and cultural borderland, and in its distance from the coast, the Dalles-Deschutes area corresponds to the Fraser Canyon of British Columbia. Along this stretch of river are several prehistoric pithouse village sites, and associated burial areas and fishing stations. The area was long recognized among relic hunters as a rich source of decorated stone and bone artifacts, but it was not until the 1920's that Strong el al. (1930) attempted systematic excavations. Caldwell (1956) and Butler (1965) carried out further salvage studies during construction of The Dalles dam and reservoir.  Wakemap Mound, 45 KL 26, situated next to a former winter village of the  Wishram, is the only site satisfactorily reported and radiocarbon dated. Two temporal strata have been identified for the village site with a separate, later age for the cremation burial area (Butler 1965:9). Wakemap I dates from 1100 to 550 B.P., Wakemap II dates from approximately 550 to 250 B.P., and Late Cremation I dates from approximately 350 to 150 B.P. Figures 92a and b, thought to be head-scratchers or hair ornaments, are two examples of decorated material from Wakemap Mound. Figure 92a was found near the  238  Figure 92. Columbia River antler human figurines. a-b, Wakemap Mound; c, Bridge Camp; d, historic Chinook carving.  239  bottom of Wakemap I and is thought to be over 1000 years old. As in examples from Gulf Island and Skagit Delta sites, the antler figures have short skirts, hair ornaments, deep notches at the neck forming a squared-off shoulders, a necklace or clavicle, arm and leg decorations or tattoos, and ribs. Unlike the northern examples, the arms are held at the sides and the forehead is not sloped. Figure 92c is a unique antler sculpture of a figure carrying a child on its back. It is from waterlogged deposits at the Bridge Camp site on Sauvies Island and is estimated to date between 350 and 1450 years old. The terminal date is based on absence of European trade goods and 4 m of silt deposits capping the cultural deposits (Butler 1965:7). No other figurines portray skeletal features so explicitly with clavicle, sternum, ribs, elbows, kneecaps, and neck vertebrae depicted. Although portions of the carving are similar to the deep engraving style of art from southern British Columbia, the carving style, skeletal chest features and proportions of the body closely resembles a Chinook figure carving shown in Figure 92d. Butler postulates that the human figurines found in Lower Columbia sites are of two types that represent different time periods. Figurines similar to the example in Figure 92a are from older site components, while later carvings (shown in Figure 93) are found in cremation burial sites that date from 350 years ago to the early historic period. Some the figures have high, deformed foreheads and hair styles similar to figurines from the Gulf Islands and Puget Sound sites, while others appear to be wearing headdresses. The unusual almond-shaped eyes and grinning mouths with protruding tongues are also portrayed in local pictographs and stone sculptures (Wingert 1952:no.8; Strong 1959:Fig. 32). Figure 93e was identified by a Wishram native informant as representing Tsi'La, the water snake guardian spirit. The grinning pose of the face represented spirit possession and sickness brought on by spirit possession (Butler 1965:8). Three undated discoveries related to this discussion are shown in Figure 94. Figure 94a is from an undated Lower Columbia River site near the John Day river. It is made of  240  Figure 93. Columbia River antler human carvings. a-c, The Dalles (Strong 1959:Figure 42); d, Stewart; e, Over; f-g, Dalles-Deschute region (Strong et al. 1930:Plate 9).  Figure 94. Undated human figurines. a, Lower Columbia river (Strong 1959:Figure 40); b, Yakima Valley (Smith 1910); c, Waldron Island.  242  steatite and is closest in form to the figurines from southern British Columbia and northern Washington.  Figure 94b was found with a burial in the Yakima Valley area (Sahaptin  territory) by Smith (1910). The elaborate headdress of this figure is unusual, but the fringed skirt, skeletal and decorative details, and general shape are familiar. Lastly, Figure 94c was found among the roots of a tree on Waldron Island in the Gulf Islands of Puget Sound. It is included here because of its closer similarity to material from the Lower Columbia River region.  Summary The most striking resemblance among all of the figures is the bold, linear nature of the surface decoration. A strong visual impact is created by the deeply incised parallel lines of the hair, headdresses, shirts and ribcage. The emphasis on the central part of the hair and the diagonal lines of the coiffure in some of the figures would seem to echo the ribcage and sternum. On the northern examples where ribs are not often depicted the lines of the hair might be a symbolic reference to a rib motif. The flattened heads and the headdresses likely refer to high ranking individuals. Several authors (Holm 1987, Carlson 1983) refer to the figures as women, but at least two of the carvings have male genitals.  A second  similarity between all of the figures is the flat, squared off nature of the compositions. The squared shoulders and absence of modelled low relief elements integrated into the composition seem closer to the artistic sensibility of Duwamish and Swinomish spirit canoe figures and Chinook bowl figures than to Central Coast Salish art. Wooden figures with similar squared shoulders, short skirts, and engraved body ornaments were collected by Capt. Vancouver at Bainbridge Island near Seattle in Lushootseed territory (Burke Museum exhibit).  The navel surrounded by engraved rings shown in Figures 92b and 93a is a  common motif in Central Coast Salish spindle whorls (Suttles 1983:84).  243  It would not be fruitful to hypothesize where this figurine originated, since radiocarbon dates discussed above suggest that carvings from both the Columbia River and Georgia Strait-Puget Sound region are over 1000 years old. It is clear that the carving style of the figurines has closest parallels to the historic art of Lushootseed, Twana and Chinook speaking groups. In the north, the prehistoric occurrence of this figurine is restricted to sites within the territory of Straits and Lushootseed speaking Salish. The figurine from Kwatna in the Bella Coola region may represent an isolated item of trade. Lack of ethnographic information prevents any secure interpretation of the cultural context of this figurine. Within the territories where the figurines have been recovered, head deformation and female facial tattooing were associated with higher status families. Among the Chinook head flattening was noted as "a badge of aristocracy" (Ruby and Brown 1976:47), and tattooing among the Coast Salish as well as head deformation was "the mark of a well-to-do person" (Barnett 1955:74).  These figurines would appear to flaunt the  outward signs of status and wealth-flattened heads, earrings, headdresses, necklaces and tattoos~and may once have been associated with the rights and privileges of higher status individuals. Carlson (1983:201) suggests that the size and side perforations of the figurines indicate their use as pendants, and that they are analogous to shamanic pendants from Siberia. Keddie (1987:1) also suggests that these figurines may have shamanic associations, being similar to small antler carvings found in shamans' graves in the northern Northwest Coast. He notes that the hand posture is similar to aspects of a curing ritual described by an early Hudson Bay Company official. There remains no concrete information from the Coast Salish area on the significance of these figurines or the particular characteristics portrayed on the human figures in the carvings.  244  Whale bone Clubs Clubs carved from the dense, lower jaw bones of whales have been collected along the length of the Northwest Coast and inland along the Fraser and Columbia River systems. The carved clubs are often called West Coast whale bone clubs for several reasons. Firstly, it is among the West Coast groups-the Nuu-chah-nulth, Nitinat and Makah that whales were extensively encountered and hunted, making whale bone easily available.  Secondly, the  motifs and design elements portrayed on the clubs relate most closely to West Coast art. Lastly, many whale bone clubs were collected by early explorers from the villages of the west coast of Vancouver Island. This study has discovered that whale bone clubs are common artifacts in prehistoric sites in the Salish regions of southern Vancouver Island. Boas discussed and illustrated a collection of fifty whale bone clubs. He described the carved design as: the head of the eagle or thunderbird represented in the type commonly found in the art of the Nootka, with a bird head-dress of the type of the eagle-head masks so commonly used by the Nootka Indians (1907:403). Drucker stated that the clubs were called tci'tu and served as a war chiefs badge of office (1951:335). The lower faces on the clubs have distinct, downturned mouths and large slanted eyes that take up much of the face. Over half of the examples presented by Boas (some of which are shown in Figure 95) have zigzag lines, double lines, or lines with internal vertebrae designs running down the centres of the blades. Many of the shaft designs end in a face with radiating feathers.  Most of the clubs likely represent the thunderbird  surmounted by the lightning-snake or wearing a lightning-snake headdress. As West Coast artist Ron Hamilton has stated, "The thunderbird wears the lightning-snake as his robe of power (pers. comm. 1986)." The design on the blade represents the lightning-snake's body either with a zigzag lightning motif, or a double backbone motif. West Coast informants indicated to Boas (1907) that notches, dots and engraved wedge-like designs on the heads represented feathers.  245  246  —  Whale bone clubs have been found in three early archaeological contexts as  previously discussed in the section dealing with art from 2400 to 1500 years old. Figure 32a from the Marpole site, dating from 2400 to 1500 years old, is likely the handle of a whale bone club that was reground to form a pendant. It depicts the profile of two bird's heads facing away from each other. The Boardwalk site in Prince Rupert harbour yielded a 2500 to 2000 year old burial with an elaborately engraved club shown in Figure 69b. It depicts the profile of an anthropomorphic head surmounted by an animal headdress. Figure 69a shows a handle of a zoomorphic club from the Yuquot site on the west coast of Vancouver Island. The handle as well as several broken club blades were found in a component dating from 1900 to 1150 B.P. The carving suggests the head of a bird or lightning-snake with feathers along the back of the head and a downturned mouth. These early discoveries give proof that the main motifs seen on clubs from the historic period were established at least 2000 years ago although the early prehistoric examples do not exhibit the same carving style and motifs.  The prehistoric finds also  indicate that whale bone clubs were as widespread 2000 years ago as they were in early historic times. A number of whale bone clubs have also been found in late prehistoric sites. Most are surface finds, but several were found in situ. Boas notes finding clubs in shell middens from Plumper's Pass, Cadboro Bay, and the Songish reserve. This study has also found clubs in collections from the Ozette site; Esquimalt Lagoon, DcRu area; Georgeson Bay on Galiano Island, DfRu 24; the Victoria Yacht Club site, DcRt 8; Stselax Village site, DhRt 2; a slate version from Nicomen Island, DhRm 3, in the Fraser Valley; and the Chase burial site, EeQw 1, near Kamloops. Three stages of manufacture have been documented at the Ozette site including a modified section of whale bone, a roughed out club shape, and a completed club (Kirk 1978:103). The completed club exhibits motifs and design elements identical to clubs collected in the historic period. None of the examples found in Salish area sites of the late period differ significantly from the majority of clubs from the West Coast. This would support the notion that the clubs  247  were traded as finished products rather than in raw material form. Of the 70 clubs reviewed for this study only a few have a single face. Four diverge slightly from the standard motif in presenting an anthropomorphic face surmounted by the body of a snake or animal: one is the 2000 year old example from the Boardwalk site, two shown in Figure 69c were collected by Smith in 1897 from the Kamloops district (although no information on their collection is recorded they are included in a discussion on prehistoric artifacts), and the last example was found in the Columbia River region. These prehistoric clubs, as well as the Marpole site club fragment with two opposite profiles shown in Figure 32a, indicate either that a greater diversity of club forms existed earlier on in prehistory, or that artists outside of the West Coast region varied the motif to suit their tastes. The consistency of the thunderbird and lightning-snake motif on clubs found throughout a wide geographic area indicates a potent relationship between the club and its carved representation. The key to this relationship is the material of the club itself-whale bone. Whales are hunted by the thunderbird and lightning-snake and are a proof of their hunting prowess.  It seems fitting that lightning-snakes and thunderbird images were  applied to harpoon points, atlatl weights and whale bone clubs-tools where accuracy and speed were needed for success. The images helped to ensure that the weapons found their mark and struck with force.  Olympic Peninsula  Ozette site The Ozette site is near the northwestern tip of the Olympic Peninsula in the state of Washington. It was a winter village of the Makah that was partially buried by a mud slide. The presence of iron but the absence of other trade goods, as well as recent dendrochronology studies (Samuels 1983:24) suggest that the site is approximately 200 years old. Although the  248  site is not very old, its importance lies in the unprecedented recovery of a complete inventory of household contents rather  than the usual midden refuse and burial remains.  Unfortunately very little of the archaeological work has been published. Information used in this study was gleaned from short articles and a visit to the Makah Museum where many but not all items from the site are on display. Of over 42,000 items recovered from the site, 66 decorated artifacts have been illustrated, while there are references to 300 to 600 decorated items in the total sample (Croes & Blinman 1980, Daugherty & Friedman 1983, Kirk 1978). For the purposes of this study the art record from Ozette has been used to evaluate the decorated forms, materials, motifs, design elements and carving styles present in a single village at a single point in time. These variables are compared with the sample from Gulf of Georgia sites. Materials and decorated forms: Wood was by far the most popular artistic medium at Ozette used in 54 of the 66 decorated artifacts, followed by 9 antler, 3 whale bone, 2 bone, and 1 stone item. Aside from the dense and tough whale bone used for handles and clubs, the choice of materials such as wood or antler does not appear to be correlated to particular types or functions of objects. Particular types of implements were chosen over others for artistic embellishment, and these choices may indicate the status given to various occupations. Of the 66 decorated artifacts available for this study, 50 are clubs, combs, pendants, weaving implements, tool handles or bowls.  It is interesting to note that tools used by both men and women are equally  decorated. Daugherty and Friedman (1983) state that in the total sample of art from Ozette, clubs, tool handles, weaving and spinning tools and combs are almost always decorated, while boxes, bowls, harpoon shafts, harpoon valves, and loom posts are often decorated. As might be expected, functional objects that are worn down during use such as stone blades, wedges, awls, fish hooks, awls and canoe bailers are unembellished. But cultural preference rather than functional requirements is indicated in the decoration of sea mammal hunting equipment to the neglect of land mammal hunting equipment (bows and arrows).  249  Significantly, the only large sculptural forms found in the site were two carved wooden effigies of whale "saddles" (the dorsal fin and surrounding section of the whale's back) which are thought to be the most prized section of whale meat. The two effigies are elaborately ornamented with sea otter teeth and turban snail opercula. The decorated harpoon points from Marpole phase sites in the Gulf of Georgia region appear to match, on a smaller scale, the attention given to sea-mammal hunting equipment at the Ozette site. The decoration of implements associated with blanket weaving at Ozette stands out in its similarity to ethnographic Salish art. Suttles (1983) suggests that among the Coast Salish, tools that transform raw material into wealth-notably weaving implements-are frequently decorated. His thesis should be examined for its relevance to Makah culture and the weaving material at Ozette. The decoration of all tools handles such as chisels and adzes, while other seemingly good candidates for embellishment such as boxes and bowls receiving less frequent attention, deserves consideration. The most obvious explanation is that the decoration of tools, like hunting equipment, was thought to help ensure the successful completion of the task at hand. A pouch containing whaling implements found in the same house at Ozette, also contained small carvings identified as amulets (Daugherty pers. comm.). Boxes and bowls performing passive functions as containers for storage and serving may not have needed emblems of strength and prowess. Alternatively, Suttles thesis of the decoration of tools transforming raw materials into wealth could be applied to woodworking implements and sea mammal hunting equipment.  These are speculative  explanations that should be examined in the light of archaeological and ethnological information contained in the eventual completed study of the Ozette site. Decorative inlays were found on four items. As mentioned above two whale fin effigies show extensive inlay decoration, one with a non-representational design of opercula, and a second larger effigy shown in Figure 96a, with inlaid sea otter teeth forming a thunderbird figure surrounded by rows of teeth. John Webber's drawing in 1778 of a house in Nootka Sound depicts an identical whale fin effigy with rows of inlays (Drucker 1965:147).  250  251  A cedar box or drum depicting a thunderbird's face, shown in Figure 96b is also inlaid with sea otter teeth. Since sea otter teeth are found on only the latter two artifacts, we can assume that inlaying was reserved for prestigious items and/or representations of the thunderbird. A small carving of what appears to be a bird form is inlaid with operculum. Daugherty and Friedman (1983:185) state than opercula inlays were also found along the front of sleeping platforms and in whaling harpoon shafts.  Of the 66 artifacts under  investigation only the two whale fin effigies, the box drum, a large bowl in the shape of a man, and two wall planks decorated with whales, thunderbirds and wolves might be objects associated with high status individuals. All are made of wood and would not have been preserved under normal processes of site decomposition. The single stone carving from Ozette, shown in Figure 97c, is a steatite effigy of a crouching human with a distended abdomen and female genitals. The pose and possible reference to pregnancy is similar to the antler carvings in Figure 114 from the Bell site.  Carving styles: The Ozette site, as in all other prehistoric sites in this sample, exhibits a wide variety of carving styles. Representative examples of the different carving styles are shown in Figure 97 to 99. Figure 97a shows a realistic carving in the round of a reclining male figure bowl. Figure 96b shows a box/drum executed in a block engraving style, while the weaver's sword in Figure 97b shows a combination of three-dimensional modelling and deep engraving. A comb decorated with cut-out shapes forming the profile of two zoomorphic forms is shown in Figure 98a. Several objects are carved in a style suggesting trade or influence from other areas. Figure 98b is one of several tiny abstract carving reminiscent of Yurok carvings (Drucker 1955:Plate 21). A small wooden human head that may have functioned as a pendant, shown in Figure 98c, is similar to Quinault (a Salish group on the Olympic Peninsula) art. An unknown bone object shown in figure 98d is similar in style to Chinook art. The diversity of  252  Figure 97. Ozette carvings (Daugherty and Friedman 1983:Figures 10.4,10.9, 10.14). a, wooden human form bowl; b, zoomorphic weaver's staff; c, steatite female effigy.  Figure 98. Ozette carving styles. a, antler comb detail; b, antler human figure; c, wooden human face pendant; d, antler carving (b-d, Kirk and Daugherty 1978).  254  carving styles at the Ozette site proves that the general stylistic diversity seen in most prehistoric sites is not a result of a lengthy occupation of a site. Whether acquired by trade or made in the village, Ozette carvers were exposed to a wide array of carving styles. All of these styles except the abstract form shown in Figure 98b and the Quinault and Chinook style carvings in Figures 98c and d can be found in prehistoric and historic art of the Gulf of Georgia and Lower Fraser River regions. Daugherty and Friedman have suggested that there is a relationship between an object's style of carving and its function: The unusually complete inventory of art work present at Ozette indicates a correlation between realistic representational art and secular functions on the one hand, and conventionalized art and ceremonial use on the other. The correlation is not perfect, however, for stylized themes also appear on objects of apparently secular function. Quite possibly items that seem intended for secular use may have performed a ceremonial function as well (1983:195). I interpret the relationship between carving style and artifact somewhat differently.  In  examining the sample of 66 illustrated items from Ozette certain motifs such as the thunderbird, rather than artifact types are represented in a conventionalized style, while other motifs such the human form, are usually portrayed more realistically. In examining 9 combs, which likely were used for the same function, a diversity of carving styles is evident, weakening the evidence for function determining the choice of carving style. Daugherty is correct in so far as certain images like the thunderbird are likely used for sacred or high prestige items, while images of human and seals decorate more mundane implements. As well, two-dimensional engravings tend to be conventionalized while three-dimensional forms are realistically modelled. In contrast to the diversity of carvings styles, there is a limited range of motifs and design elements found in Ozette art.  Human images are ubiquitous on tool handles.  Sometimes a human head with or without a hat is portrayed, while a human head and torso is also common. This association between tools such as chisels and the human form has persisted for over 3000 years as Figure 6, from the Glenrose site attests. Human forms are  255  also found on loom posts, mat creasers, clubs, bowls, and pendants. Most of the human images are modelled in three dimensions, but a few are more abstract, incised images. The facial plane is usually flat with features shown in low relief. This contrasts with prehistoric human images from Salish sites which often have prismatic-shaped face planes and deeply engraved features. Visual punning is seen on one club which when viewed from the side shows a seal's head, and when viewed from the front depicts a human head. Dogs or wolves are portrayed on mat creasers, weaving implements, and combs. Since dogs were used as a source of wool the use of their images on weaving implements would not be surprising, but this identification is uncertain. An antler comb on display at the Makah Museum shows two canines in profile on the handle. The cutout profiles are similar in style to an antler comb shown in Figure 75b, dating 1000 years older. Many other zoomorphic and anthropomorphic images are identical to tools such as mat creasers and spindle whorls of the Coast Salish. Double-headed forms, snake-like creatures with horns, and a creature with fins or plumes on its back likely represent lightning-snakes, two-headed snakes, or some other supernatural sea-creature.  These forms decorate combs, weaver's  swords and mat creasers. One antler effigy with a double-backbone, rows of engraved ribs, and legs portrayed in low relief is similar to a carving from the Pedder Bay site, shown in Figure 85a, and an undated antler carving from Orcas Island (Smith 1907:Fig. 194). Considering the proximity of the Olympic Peninsula to Vancouver Island, it is not surprising that both regions share many motifs, carving styles and artifact forms. An equal or greater number of implements have motifs that are distinctive of West Coast art, including whales, thunderbirds, seals, and wolves. A wall plank and several small heads that may represent mask effigies likely depict wolves.  In addition to the three-  dimensional whale fins mentioned previously, an engraved whale with a rib and backbone motif is found on a wall plank. Three-dimensional seal forms are found on a large number of wooden clubs, while flat whale bone clubs have the standard two-headed motif characteristic . of this artifact.  256  A drawing of a wall screen found at Ozette is shown in Figure 99. Two thunderbirds with outspread wings are flanked by two wolves. The Ozette site has the only prehistoric depictions of the winged thunderbird figure which is so ubiquitous in Coast Salish and West Coast two-dimensional art. All of the thunderbird images have a distinct slanted eye shape. In some cases, as in Figure 99, only the outer half of the eyelid shape is defined. The wolves on the same section of cedar board, and a whale depicted on another wall plank also have this unique wedge-form behind the eye like a half eyelid. The slanted eyeform and its association with thunderbird and lightning-snake images has been previously noted in other sections of this thesis. The eyes of humans and birds are depicted as single incised dots or engraved lens shapes. Canines, as shown in Figure 97b, have elongated, pointed eyelids. The negative C element occurs at Ozette on a wooden comb depicting a sea creature (Daugherty & Friedman 1983 :Fig. 10:10). Only a few different design elements are used in Ozette art.  What may be  considered engraved negative T-shapes or elongated wedges are used to define the inner margins of eyelids such as the canine heads in Figure 97b. In several cases this design element has a flat instead of concave base as in Figure 96b.  Daugherty and Friedman  (1983:189) state this design element is used as well in three other flat carvings to define feathers.  Figure 96b also shows thin lines that outline the facial features and design  elements. These additional lines give a double outline to all of the features.  It seems  plausible that these double lines were painted in to form solid lines. Zigzag designs are found on three of the 66 decorated objects in this sample, and incised or inlaid rows of dots are on four objects. Zigzag motifs are common in prehistoric art of the Salish and Chinook, while dotted lines are a common West Coast design feature. In summary, Ozette art has many points of similarity with prehistoric art from the Salish area. There are many shared carving styles and decorated artifact forms, while at the same time certain motifs and design elements are closer to West Coast rather than Salish  257  Figure 99. Ozette wooden wall screen (Kirk and Daugherty 1978:frontispiece).  258  art. Daugherty (pers. comm. 1985) surmises that a whaler lived in one corner of one of the excavated house at Ozette, explaining the decorated whaling harpoons, whale fin effigies and decorated objects found in that corner of the house. As might be expected, the thunderbird-famed as a whale hunter~was associated with men of status who claimed the right to hunt whales. The thunderbirds on the wooden screen in Figure 118 may have been displayed during potlatches to refer to the whaling prerogatives of the prehistoric inhabitants of this house, as has been documented in historic potlatch accounts (Hyde 1984). Red and yellow ochre as well as several paint brushes were found, but interestingly, no masks were uncovered.  Central and North Coast sites  Kwatna Inlet sites Approximately ten decorated objects have come out of three late period components in Kwatna Inlet, excavated by Carlson and Hobler between 1969 and 1972.  Unfortunately  these items are either too worn or non-diagnostic to provide a picture of the prehistoric art record of the Bella Coola region. Four decorated artifacts were found in the Anutcix phase component of the Nutlitliquotlank village site, FaSu 2. All four objects are thought to be approximately 700 years old (Carlson 1983:122). Figure 100a is a whale bone implement thought to be a shuttle or spindle. Although the object is very worn, engraved wedge-shapes and what may be a face can be discerned. Figure 100b is an antler valve from a sealing harpoon. The inside has an elegant engraved T-shape which shapes the surrounding surface into intersecting formlines. Figure 100c shows what may be a bone pin with a crude eye form on the knob. Not illustrated is a flat stone with a crudely fashioned head and four limbs (Carlson 1983:Fig.7:4d).  Figure 100. Kwatna antler carvings. a, whale bone shuttle; b, antler toggling harpoon valve; c, bone object; d-e,antler pendants (Carlson 1983: Figures 7:3a-b, 7:4a,b,e).  260  At the Anutcix site, FaSu 10, three decorated objects date between 1750 and 750 B.P. (Carlson 1983:122).  An antler human figurine similar in style to figurines from  Vancouver Island and the Gulf Islands, has been discussed in detail in another section (Figure 88a). Figure lOOd is an antler pendant approximately 750 years old. The tapering body and cutout section under the head is similar to a snake/female figurine from the Bell Site shown in Figure 111. Carlson has interpreted the top of the Anutcix figure as a bird's head or skull with deeply engraved eyes and a mouth. It could also be interpreted as a face with a pair of arms held under the chin. The tapering section of the carving would then represent a torso or snake's body. A hole in the top of the head brings to mind the many Fraser River human stone sculptures that also have small holes in the top of the head. Might this be another representation of a snake/human? Figure lOOe is another pendant from the same component with an outline shape suggesting a bird. Lastly, Figure 101a is a bone spoon handle with a cutout T-shape at the tip. It is thought to be approximately 150 years old. The Axeti site, FaSu 1, is a midden dating to the precontact Kwatna phase from 450 to 150 years old. Figure 101b comes from a water-saturated section of the site where organic material was covered in silt at the edge of a fish trap. It is a delightful antler sculpture of two identical human figures joined at the hands. The core of an antler tine has been carved out, then the outline shapes of two figures have been defined by cutting out large sections of the outer surface. The simple figures have broad, oval heads with ticked lines suggesting hair. The eye sockets and eyeballs are oval or ovoid. The squared-off shoulders, bent arms, and two triangular cuts in the genital area might indicate a stylistic connection with the antler figurines illustrated in Figures 88-91. Two objects not found in- situ interest to this study. A hammerstone grinder found at the Axeti site has pecked lenticular eyes and other features too worn to discern. Figure 101c shows a soapstone hammer ploughed up in a field in the Bella Coola Valley. The low relief modelling of the arms and legs is an identical technique found in seated human-figure  261  Figure 101. Kwatna carvings. a, antler handle fragment; b, antler human figurine, c, Bella Coola area undated stone maul (Carlson 1983:Figures 7:5, 7:6b,d).  262  bowl sculptures in the Lower Mainland and Lower Fraser River region. The figure's posture, raised notched backbone and lenticular shaped head is also stylistically similar to seated figure bowl sculptures.  Prince Rupert Harbour Sites The general archaeology of sites in the Prince Rupert Harbour locality has been discussed in earlier chapters. The late period in this region designated Period I, dates from 1450 B.P. to the time of contact with Europeans (MacDonald & Inglis 1981:52). MacDonald and Inglis state that during this period large zoomorphic stone mauls, bowls, and clubs are common along with bone, stone and tooth pendants, beads, bracelets, and labrets.  The  record of art for the late period is surprisingly poor considering the rich record of Tsimshian stone and bone art. A decorated fragment of a comb from the Garden Island site, GbTo 23, in Prince Rupert harbour is shown in Figure 102b. It is estimated to be approximately 1150 years old (MacDonald 1983). Its stylistic relationship to the comb from Maple Bank shown in Figure 75b--with a dog or wolf portrayed with a human-like seated posture-is obvious and indicates that this motif may be associated with late prehistoric combs (see also Figure 98a from Ozette). Even the number of teeth in the combs from the Maple Bank and Prince Rupert site are the same. The Garden Island example exhibits a more formal stylized posture and proportioning of the head, trunk and limbs which could be seen a prehistoric indication of the regional trend towards the formal, symmetrical quality of northern Northwest Coast art. The folded fore and hind limbs form a pleasing "biventral" symmetry common in northern Northwest Coast art. Interestingly, this same motif can be found on a historic Tlingit trap stick made of whale bone (Vaughan and Holm 1982:153). Three slightly curved engraved lines form a rib motif along the animal's back. The prominent tongue curving down, the slanted eye, and drooping pointed ear are all featured  Figure 102. Garden Island antler comb.  264  on petroglyphs from Kitselas Canyon on the Skeena River (MacDonald 1983:Figure 6:38) and are further evidence of the expression of the Tsimshian region art style in prehistoric art. There are no formlines in this composition but the careful balancing of the engraved lines within the positive ground shaping the negative and positive space shows a similar design sensibility. Three other decorated artifacts from the Garden Island site are thought to be approximately 950 to 750 years old. Two of the items are carved pins: one of a bird's head shown in Figure 104a, and another with a crude head and eyes (MacDonald 1983:Fig. 6:14a). The third item, illustrated in Figure 103a, is a flat piece of bone with two holes drilled for suspension. It appears to be a fragment of a larger eye design. The top and bottom margins of the eye socket are complete but the sides are presumably broken off. Inside is an another ovoid forming the eyeball. An engraved negative C shape inside the eyeball defines the pupil. Drawings by Holm (1965:Fig. 51 & 52) shown in Figure 103b-d clarify this design concept. Figure 103b uses painted formlines to illustrate the negative C shape defining the pupil while Figure 103c illustrates the same concept in an engraving.  The prehistoric  example is eroded and cracked but the bevelled angles of the engraved lines are still visible. As in finely carved contemporary northern carvings the eye is not a flat surface: the inner edge of the eye socket is deeply recessed and the plane of the eye gently swells towards the eyeball.  Figure 103d illustrates this carving technique.  Another way of describing the  angled slopes within the eye is that the negative portions of the composition are recessed. MacDonald is correct in calling this artifact "a classic North Coast eye form" (1983:109). Figures 104b-d are small pins or pendants from undated late prehistoric or early historic components at the Kitandach and Lachane sites in Prince Rupert Harbour. Examples b and c may relate to pin/scratcher forms from the south coast with open mouths and tongues. Figure 104d is a canine tooth that has been carved into a bird-shape. Copper wire has been wrapped around the neck. The many Tlingit examples of carved canine teeth thought to be shamans' amulets (for an example see Holm 1987:226), has led MacDonald to  265  Figure 103. Negative C carving. a, Garden Island antler eye form; b, example of painted negative C eye form (Holm 1965:Figure 51a); c, example of engraved negative elements (Holm 1965:Figure 21e).  266 Figure 104. Prince Rupert Harbour carvings. a, Garden Island bone pin/scratcher; b, Lachane antler pendant; c, Kitandach bone pendant; d, Kitandach tooth pendant.  267  propose that some of these small effigies from Prince Rupert sites once performed the same function (1983:109). Figure 105 from the Kitandach site, GbTo 34, is thought to be approximately 950 to 750 years old. The bone comb reveals an intricate engraved design that includes split-U forms, eye forms and T-shapes. Since several design elements are incomplete on the outside edge of both faces, it would appear that a portion of the original composition has broken away and the edge has been refinished. Although eye shapes, feather designs and ear motifs are intimated by the design elements, the piece as a whole defies interpretation. Despite the puzzling configuration of the design, the engraved lines shape the surrounding space into formlines in characteristic north coast style. MacDonald has noted that the engraved wedgeshaped elements in the eye (that he calls cuniforms) are Salish in style, rather than Tsimshian. The long rectangle shape of the eye socket combined with the long eyelid and round eyeball also has a West Coast appearance. One might speculate that at some point in the late period, painting gained primacy over engraving allowing greater freedom in shaping the lines of a composition and creating the possibility for tertiary design elements such as split-U forms. Figure 106 is a drawing of two fragments of a zoomorphic bone comb from a Period I component at the Lachane site. The design elements are difficult to examine in the drawing but recessed ovals on the back and shoulders may have held inlays and indicate joints and vertebrae.  268  269  Figure 106. Prince Rupert bone comb (MacDonald 1983:Figure 6:13b).  270  The Mid-Fraser and Thompson River Region  Milliken and Esilao sites The Milliken site DjRi 3, and the nearby Esilao site DjRi 5, are on the Fraser River near Yale within the border of the Coast Salish Stalo Indians.  The upper cultural  components of the Milliken site are poorly understood due to disturbance of older layers during pit house excavations and a lack of radiocarbon dates from the upper layers. Borden's Emery phase dating from approximately 1600 to 750 years old is a tentative cultural designation. Borden himself admitted that artifacts ascribed to this phase were those left over from a mixed assemblage after items thought to belong to preceding and ensuing phases were subtracted (1983:156). The Esilao site is a late prehistoric pit house village with a burial area containing a radiocarbon date of 570 ± 100 B.P. as well as artifacts from the historic period predating A.D. 1827 (Mitchell 1963:83). This site lends its name to the Esilao phase, thought to date from 750 to 150 years old. Despite the uncertainty over the exact age of material from the late prehistoric period at the Milliken and Esilao sites, it is worthwhile evaluating the decorated objects from these assemblages. Figures 107a-b are decorated pipe bowls from the Emery phase at the Milliken site. Figure 107a is made of steatite and depicts a seated human figure with legs and arms drawn close to the body. The figure's posture, clearly defined ribs, modelling of the legs and arms, and protruding shoulder blades immediately brings to mind seated humanfigure bowl sculptures from the Fraser River region. One half of the figure's head and chest is missing, but a pierced ear is visible. The top of the head has been widened to form a bowl for the tobacco which gives the head a broad, flat profile. The second pipe in Figure 107b is made of siltstone and is carved into the shape of a bird. A hole in the side suggests that the carving was discarded due to this breakage. Both pipe bowls would have been mounted on hollow wooden stems.  271  Figure 107. a-c, Milliken-Esilao stone carvings.  a, t h r e e  views  272  Figure 107c from the early historic deposits at the Esilao site has a tubular pipe shape more typical of interior pipes from the late period. A bird's head forms the bowl of the pipe. The eye has a downturned spur at its outer edge similar to bird and snake images from the Marpole site. The large downturned beak suggests the portrayal of a hawk or raven.  Duff (1956:84) states that decorated tubular soapstone pipes are found in large  numbers in the Lytton region and less commonly in the Fraser Valley and Kamloops areas. Pipes are found in prehistoric Plateau and Kamloops horizons on the Canadian Plateau (Richards & Rousseau 1987:89). Illustrations of decorated pipes in Duff (1956:Plate 19) and Smith (1923:Plate XIV & XV) show that a variety of human and animal forms were portrayed on the pipes. With human forms the pipe bowl is at the top of the head, while in animal forms the bowl is usually the creature's mouth. Figure 108 is a unique steatite spindle whorl from the Milliken site. The piece was recovered from late Emery phase material sloughed off a trench wall.  Borden (1983)  assigned this piece to the historic period based on its sophisticated composition which he felt was only found in historic Salish spindle whorls, although he qualified his opinion with the statement, "future data and deeper insight may eventually justify placing this remarkable artifact into an earlier period." The motifs and design elements on the whorl are consistent with prehistoric art from the late period described in this section, therefore I will take Borden at his word and put the steatite carving back into the prehistoric record. On the rounded convex face of the whorl (side A) single eye shapes appear to swirl around the central hole. The eyes are elongated with pointed eyelids. The majority of the eyelids point down at their outer corners. The hole in the centre of the composition is also surrounded by an eyelid making the hole into the pupil. This same motif also represents the lips of humans in Coast Salish spindle whorls. The suggestion of a human face is aided by an engraved oval line which runs from under the eye/mouth to the edge of the whorl, and two nostrils which sit above the eye/mouth. No attempt has been made to place the eyes inside this face in their  Figure 108. Milliken steatite spindle whorl.  side  a  side  b  274  correct position. Deeply engraved dots are sprinkled throughout the composition and also circle the outer margin. On the reverse face (side B), the bodies of four serpents curl around the central hole. They too have rows of engraved dots along their bodies and elongated eyes with downturned eyelids. The pupils of their eyes are slightly different in being joined to the upper eyelid. This particular negative C eye configuration has been noted in several examples of late period art. The eye shape and open mouth of the Milliken carving suggests the depiction of a lightning-snake. A seated human figure bowl found in North Saanich also has a serpent with rows of dots along its body (Duff 1975:63).  Lochnore-Nesikep sites The Lochnore-Nesikep locality sites, on the Fraser River between Lytton and Lillooet, are in the territory of the Upper Thompson Indians.  Sanger has described two  zoomorphic artifacts from Kamloops phase late period assemblages (1968a, 1971). Figure 109a is an antler haft from the McPhee site, EdRk 6, a housepit village thought to be less than 500 years old. The haft has an engraved rib and backbone design. Every other vertebra is deeply incised forming a checkered line similar to Figure 34 from the Marpole site. Figure 109b is from a burial at Texas Creek, EdRk 1, a privately excavated site thought to date between 550 and 350 years old. This object is from a private collection containing over 168 artifacts examined by Sanger (1968a).  The collection shows a wide range of  functional and decorative artifacts embellished with non-representational ornamentation typical of the Plateau area. The steatite carving in Figure 109b is the single zoomorphic representation. Diagonal lines on the creature's side may indicate ribs although not all the lines are parallel. The spine is notched, and when viewed from the top shows a row of engraved chevrons. A short line defines a mouth, otherwise no other features on the head are portrayed. The low, long profile and short legs suggests the depiction of a lizard.  275 Figure 109. Mid-Fraser River carvings. a, McPhee antler haft; b, Texas Creek steatite zoomorphic carving; c-d, Lytton area antler carvings (Smith 1899:Figure 114).  276  Figure 109c is an undated antler carving from a burial in the Lytton locality (Smith 1899). It is included to show the use of the negative C element in the Mid-Fraser region.  Bell Site The Bell site, EeRk 4, is situated on the east side of the Fraser River a few miles north of Lillooet within the territory of the Lillooet Interior Salish.  The Bell site is a  prehistoric pithouse village with nine decorated artifacts dating from 1560 to 1250 years old from late Lillooet or early Kamloops horizon components (Stryd 1983). Six of the sculptures from the Bell site were found in housepit #19, with four of the items found in a single burial. The housepit has both Lillooet and Kamloops horizon projectile point styles, so although radiocarbon dates of 1430 ± 60, 1515 ± 90 and 1250 ± 200 B.P. suggest a late Lillooet occupation, Stryd (1983) assigns the burial and house floor assemblage to the early Kamloops horizon circa 1000 B.P. More recently, Stryd (Lawhead et al. 1986) has suggested that this burial be reassigned to the much earlier Shuswap horizon dating from 2400 to 3500 years old, based on stylistic similarity of a human figurine (Figure 111) to sculptures from the Pender Canal site.  In their review of Canadian Plateau prehistory, Richards and  Rousseau (1987) suggest that since incised antler is absent in the Shuswap horizon the burial and its artwork likely belong to the late Lillooet horizon as the radiocarbon dates suggest. I concur that stylistic similarity is not enough evidence to reassign the burial, since a human figurine similar to Bell site examples shown in Figure 114 was found at the Ozette site (Figure 97c). The burial in housepit #19 was of an infant placed together with four decorated objects and 249 dentalium beads, in a pit dug into the house floor. Figure 110 shows an antler comb from the burial depicting what may be two birds in profile joined at their beaks. This striking composition is unique in the prehistoric period although elements of the design are similar to artifacts from coastal sites. First, the eyelid is similar to the Coast Salish eyelid form found in prehistoric antler carvings from the Marpole (Figure 29f & 30a) and  277 Figure 110. Bell antler bird form comb.  278  Garrison (Figure 54b) sites. The Bell site version of this eyelid is unique with distinctly down-turned edges giving an angled shape to the eye margin. But the execution of the eyelid form with the concave wedge shapes and angled engraved cuts is identical to Marpole phase examples. In the same manner as the combs from the Maple Bank and Garden Island site, cutout spaces are used to shape the outline of the figures, and in this case also the eyes and mouths. The motif of two birds with touching beaks was also found at the Hoko River site (Figure 25). In the Bell site example the upright postures and pointed bills of the creatures suggest owls, geese or cranes although the lack of wings makes this uncertain. The design may also be intended to contain a visual pun being read as the head of one creature with two eyes and a downturned mouth (perhaps a frog) as well as two birds. The incised horizontal bands that decorate the animals' bodies and shaft of the comb is a subtle interplay of negative triangular ticks that form a positive zigzag pattern. This motif is also found on three antler digging stick handles the same site (Stryd 1983), in several other undated decorated objects in the Mid-Fraser and Thompson region (Sanger 1971) and on unilaterally barbed harpoon points from the Marpole site (Smith 1903:182). A second antler carving from the burial in housepit #19 is shown in Figure 111. It portrays the striking visage of a creature with a human-like upper body and the lower body of a rattlesnake.  Female genitals are portrayed above the tail rattles.  Stryd accurately  perceived stylistic similarities with sculptures from the Pender Canal site dating approximately 3000 years old. The distinctive elongated slits of the eyelids in the Bell site figurine are similar to the eyes of a sculpture from Pender Canal (Carlson 1986 pers comm). The notched ridge of the crown and the intricate cutout shapes that define the arms and back are also similar to the Pender Canal sculptures. But there also similarities to later sculptures from interior and coastal sites. Similar elongated eyelids are found on a steatite figure from the Chase burial site (Figure 124a). The suggestion of notched headdresses  279  Figure 111. Bell antler snake-female effigy.  280  occurs on carvings from Locarno Beach and Marpole phase as well as on undated seated human-figure bowl sculptures. The bulging ridges of the eyelids and forehead contrast with the flat surfaces of the lower face and arms. A sharply defined forehead also shapes a wide nose. The wide mouth is open and slightly downturned. The shape of the eyelids and the downturned mouth gives the face a grimacing expression. The head is joined to the body by the slender stalk of the backbone. The arms are flexed at the sides of the body and the forearm joins the sternum: no attempt has been made to make this section realistic. The lower snake-half of the body has six encircling engraved lines which portray the rattles, and a carved out area depicting a female vulva. The body of the creature is covered in vertical bands of short parallel lines likely suggesting the scales or markings of a snake. These incised markings are found on prehistoric utilitarian items from the Canadian Plateau. A mythological association for this figure has not been found although Teit (1912:366) mentions a Thompson Indian story about a cannibalistic woman whose genitals have a poisonous bite like a rattlesnake. Stryd (1983) suggests that the head of the figure may be the portrayal of a mask. The back of the head has been hollowed out, the top is notched as in certain styles of headdress, and there is a small hole above the end of the eye as a mask might have for attachment to the head (the other side of the face is broken off). Carved out head areas have been noted in antler sculptures from coastal sites in this sample. Whatever its meaning, the figure's countenance and human-snake features suggests the portrayal of a supernatural creature. Figure 112a shows a small stone animal figure from the burial in housepit #19. The small sculpture which can be seated upright, has been carved, engraved, and finely polished. An open mouth, nostrils, simple lenticular eye, and a tiny projecting ear are portrayed on the large head. Thin, weakly developed legs stand out in bas relief against a large abdomen. The section of legs which once projected from the body have been broken off although they may never have been fully depicted. A short tail is visible between the two hind legs. Taken  281  Figure 112. Mid-Fraser/Thompson zoomorphic carvings. a, Bell site siltstone bear (?); b, Shuswap area steatite bear (?).  282  as a whole, the large head with its small ears, the distended belly and the thin legs give the figure the appearance of a foetus or newborn bear. Two similar steatite figures with thin forelegs and undeveloped hindlegs were found at the Chase burial site described in the following section. The last item from the Bell site burial is a small steatite pendant shown in Figure 113a. The top half is an anthropomorphic head while the bottom portion is an undecorated cylinder with a suspension hole. The forehead consists of four deeply notched lobes. Directly underneath are two engraved lenticular eyes. The cheek and chin are defined by a deeply engraved circle which is broken at the top by a broad wedge-shaped nose. The lips project from the encircling groove of the cheek and chin. The configuration of the facial features and the notched forehead is in a similar style to art from coastal and interior sites mentioned previously. Figure 113b is from the floor of housepit #5 at the Bell site and is associated with a date of 1380 ± 65 B.P. (Stryd 1983). Although executed in a different carving style than Figure 113a the two carvings resemble each other in form.  The antler carving has a  crescent-shaped carved out mouth and lenticular eyes directly underneath a prominent, notched headdress or hairstyle.  A hole in the lower section of the carving may be an  unfinished suspension hole or may represent genitals (or both). Chisel marks are visible all over the carving giving it a rough, unfinished appearance. The figure was found with four fish vertebrae, a fir twig, and a large rock. Figure 114a-b shows front, back and side views of two antler figures found together on the floor of housepit #19 but not with the burial. The complete figure shows a human with arms and legs tightly flexed against the body. The arms and legs stand out in low relief while a spinal column is indicated by a deeply notched backbone. In contrast to the thin limbs, the figure's abdomen is large and swollen.  A long carved depression below the  abdomen likely represents female genitals. The features of the head are carved with wide cuts that crudely portray lenticular eyes, an open mouth, and a broad triangular nose with  Figure 113. Bell anthropomorphic carvings, a, steatite pendant; b, antler carving.  284  Figure 114. a-b Bell antler anthrpomorphic carvings.  a:  front  b:  front  285  wide nostrils. The small figure stands upright on a flat base. The secondfigureis damaged but shows similar features and carving style although the wide nose does not have nostrils and the mouth is an engraved oval. The depiction of female genitals and a prominent belly suggests the depiction of pregnancy although this is an unfamiliar motif in the art of any region of British Columbia. The features and surface of both carvings are worn and the edges of the engraved lines are rounded rather than sharply defined. A broken stone object that is likely the top portion of a maul, shown in Figure 115, was also recovered from Housepit #5 and is associated with the date of 1380 ± 65 B.P. The worn gabbro stone has widely pecked depressions defining the oval eyes and slightly projecting forehead and what appears to be a downturned beak. Lastly, a badly worn and fragmented section of antler (not illustrated) from housepit #23 is associated with a date of 1560 ± 90 B.P. It has the rudimentary features of two circular eyes and an open circular mouth. It is hollowed out inside and may have functioned as a haft. Finding four decorated objects together in a burial provides an unusual opportunity to assess the art styles present in one location at one point in time. Each of the four items from the burial in Housepit #19 are executed in a totally different style, yet each has strong stylistic associations with prehistoric objects from both coastal and plateau archaeological sites. The heterogeneity of carving styles found in Marpole phase sites appear to hold true for objects found at the Bell site. On the other hand, Bell site objects also show stylistic differences from coastal sites. The configuration of facial features in the antler and steatite figurines in Figures 111, 113 and 114 with wide, flat triangular noses, flat faces with simple oval eyes and no eyebrows is distinct. The little figurines that appear to depict female genitals are also a unique motif. The only coastal example of this motif is from the Ozette site dating approximately 400 years old (Figure 97c). The well-made sculptures and quantity of dentalia found with the burial of an infant at the Bell site suggests that some degree of status may have been accorded to this young individual.  Differences in the quality and quantity of decorative artifacts interred in  286  Figure 115. Bell stone zoomorphic maul fragment (Stryd 1983:Figure 9:6).  287  prehistoric burials, especially plentiful inclusions with child burials have also been noted at sites in Kamloops (Richards and Rousseau 1987:38), Cache Creek (Pokotylo et al. 1987), and Chase (Sanger 1968).  Seton Lake site The Seton Lake site, EeRl 21, is a prehistoric pithouse village in the Lillooet vicinity. A small steatite bowl illustrated in Figure 116 was found in a housepit with a date of 1220 ± 85 B.P (Stryd 1983:174). The oval bowl features two identical faces which project from either end. Both the carving technique and the configuration of each face is unusual. The bowl is finely carved and polished. The face is a broad oval within which the mouth, eye socket and forehead area are deeply carved out. It is somewhat confusing to discern the relationship of the carved out negative crescents at the top of the face. Are they eyebrows, or are the crescent-shaped lines below them the eyebrows? I believe that this composition follows the convention seen in Marpole phase art where in two-dimensional compositions the forehead area is marked by an incised or engraved line that is not representational (as in Figures 29f and 30a). The engraved crescents define the upper margin of the eyebrows with the lower margins of the brows are defined by the engraved eye sockets. The Seton Lake faces, like their Marpole phase counterparts, have heart-shaped forehead margins with sharply down turned peaks at the bridge of the nose (compare to Figure 37). Although unusually simple, abstract shapes define the features of the face, the composition as a whole is very sophisticated. The positive plane that forms the nose, eye brows and cheeks swells at the centre of forms and diminishes at the junctures with other features in the manner of formlines. Several suggestions have been put forward as to the identity and meaning of the faces and general bowl form. Diagonal lines in groups of three form a zigzag pattern around the rim of the bowl. An engraved line encircling the bowl just below the top edge forms a  288  Figure 116. Seton Lake steatite bowl.  289  raised rim. A bowl with a human face from the Marpole site, shown in Figure 39a, also features a zigzag motif around the rim of the bowl. The bottom of the Seton Lake bowl has a deeply engraved line which runs lengthwise along the bowl and notches the chin of each face. Duff (1975:49, 172) interpreted the bottom of the bowl as representing the "mounded cleft" of the female genitals, while the upright bowl represents the open, lipped vagina. Duff thought the faces could represent humans or owls. When viewed in isolation, I would not interpret the shape of the bowl as representing female genitals, but with the evidence of the artifacts from the Bell site, one of which has the unmistakable depiction of the "mounded cleft" of the vulva while three others have markings which suggest the depiction of the vaginal opening, I concur with Duff s interpretation.  I also believe certain elements of the bowl~the two heads,  the zigzag pattern on the rim, and the possible backbone on the bottom-could represent a lightning-snake or two-headed snake. The bowl form of two heads facing outward is found on only one other stone bowl. Found nearby at Lillooet by a private collector, the rough-grained bowl in (Duff 1975:27) was uncovered by a plow. The deeply engraved eye socket and mouth area, the small oval eyes, and the configuration of the eyebrow and nose are similar to the bowl from Seton Lake, but the faces on the Lillooet bowl have a prismatic shape with the sharp angle at the bridge of the nose or beak. The Lillooet bowl suggests the face of an owl with the notched ridge that encircles the face representing the feathered eye ring.  Undated Sculpture from the Mid-Fraser and Thompson River Region Several other undated finds from the Mid-Fraser and Thompson rivers have motifs that relate to the material uncovered in archaeological contexts discussed above. Figure 112b shows a steatite carving of an animal's head which forms the bowl of a pipe. It was found by a private collector at a grave site in the Thompson region and is similar in form to the "bear" sculpture from the Bell site burial.  290  One of several similarly styled steatite carving from the Yale area is shown in Figure 117a. This carving suggests both a phallus and a snake. The end of the body may represent the rattlesnake's tail rattle as is also depicted in the female-snake figure from the Bell site. The underside of the effigy features a longitudinal line intersected by short lines suggesting a backbone motif. Figure 117b, found near Lytton, and the human head bowl in Figure 117c, found between Lytton and Yale, both depict a human face with a notched headdress similar to the artifacts in Figures 111 and 113a-b from the Bell site, and to previously mentioned two and three-dimensional carvings from Marpole phase sites. Figure 117d is a headless human figure in steatite from the Yale area which may have functioned as a pendant. The neck and ankles are decorated with incised lines which may represent decorative bangles or tattoos. An engraved notch at the base of the torso likely represents female genitals. Figure 118 from an undated grave site at Kamloops, may provide a link between some of the female figures found at the Bell site and the seated human figure bowl complex. It portrays a human figure with a long torso, slightly protuberant belly, and thin legs and arms folded close to the body in a similar style to the figurines shown in Figure 114 from the Bell site. The body of a snake or lizard extends along the figure's backbone. The coarsegrained soapstone stone sculpture is distinct in portraying the figure resting on a large human head. A large hole in the forehead of the head is in a position to suggest the depiction of the vaginal opening of the upper figure. The overall shape of the figure and head has been modelled in three dimensions whereas the features such as eyes, mouths and fingers are simply portrayed with incised lines. Hill-Tout (1899:18), the object's collector, was told by Indian informants that the figure represented a woman giving birth and that the object was used by a shaman during girls' puberty ceremonies. The hole held water which the shaman sprinkled on the girl upon her return from seclusion in the woods. This account of the bowl's association with shamans or ritualists agrees with other information on the use of decorated stone bowls discussed in  291 Figure 117. Mid-Fraser River undated steatite carvings (c, Smith 1907:Figure 192b).  b  c  292 Figure 118. Kamloops area seated human figure bowl (Smith 1923:Plate 36).  293  detail by Duff (1956a:56), although not all accounts deal with girls' puberty rituals. A Yale informant (Duff 1956a:57) stated that "Bowls with the 'basin' in the head were said to be female; those with the basin held in the arms were male." These undated objects from the Mid-Fraser and Thompson region confirm the prevalence of unique regional motifs including female figures, rattlesnakes and bear effigies that relate to the distinct ceremonial complexes of the Plateau region.  Chase Burial site The Chase Burial site, EeQw 1, is situated on the South Thompson River near Shuswap Lake on territory belonging to the Interior Salish Neskainlith Shuswap. The site once contained over 50 burials dating from 400 to 200 years old (Sanger 1968b). The age of the site was determined by the presence of trade copper and small side-notched points, and by the absence of trade beads and iron items. Unfortunately, most of the site was vandalized by local residents who discovered the richly endowed graves.  Sanger was only able to  examine six intact graves and inspect pot-hunted artifacts shown to him by local residents. Despite the destruction of much of the site, Sanger made the following assessment: ...at Chase there was found one of the largest collections of aboriginal art work ever excavated from a single Plateau site, where archaeological assemblages often contain only the stark stone (and sometimes a few bone and antler) implements associated with the subsistence quest. The Chase collection demonstrates the extremely limited range of material culture that has been recovered in the past from Plateau sites (1968:131). The six scientifically excavated burials were of three children, and three adults. Only one of the child burials lacked grave inclusions; the other burials had extensive inclusions of ornamental and functional items. Thirty-eight artifacts from both the private and archaeological collections were decorated with non-representational motifs.  These  decorated forms included digging-stick handles, harpoon points, clubs, pendants, and miniature bows. Fourteen artifacts were embellished with representational motifs. The  294  majority of the latter items were in the possession of private individuals who allowed them to be sketched or photographed. Figure 119 shows the single wooden artifact from the site. It is believed to be a mask even though its size is smaller than a human face. To date, it is the only mask to survive from the prehistoric period. Protruding peg-like eyes are reminiscent of Coast Salish sxwayxwey masks. The nose and central portion of the mask have disintegrated but the remaining sections show sharply pointed lenticular eyelids and engraved lips.  Ochre  staining was visible on the mask which was found with a section of twined sedge matting. The discovery of punctured pecten shells at the Chase Burial site (unfortunately the provenience is not known) increases the possibility that the wooden mask may be part of the sxwayxwey ritual complex since these shell rattles are associated with the Coast Salish sxwayxwey ritual (although it should be noted that they are not exclusive to this ritual [Suttles 1983]). Figure 120a and b are identified as antler handles portraying human heads. Figure 120c and d are steatite carvings of human heads. The face in Figure 120c has two defined facial planes, raised triangular forecheeks, and broad nostrils characteristic of Lower Fraser and Gulf of Georgia stone sculpture from the Marpole phase onward. The pipe in Figure 120d conforms to examples from the Lytton area depicting human heads on bowl section pipes. Figures 121a-c illustrate three antler tine clubs with the "slave killer" shape common on the Northwest Coast. Example 121a has a notched backbone motif, as well as a row of dot and circle designs interspersed with short parallel lines (two typical Plateau design elements). An elongated eyelid runs along one quarter of the antler shaft. Figures 121b and c have zoomorphic heads with simple eye and mouth details. The striking points on both clubs are in the position of the animals' forelegs. Example b has two zigzag lines encircling both sides of the deeply engraved neck section. Similar clubs have been reported in late Marpole phase sites on the coast and in the Middle period Boardwalk site in Prince  295 Figure 119. Chase Burial wooden mask (Sanger 1968a:Figure 10).  296  297 Figure 121. a-c, Chase Burial antler clubs (c, Sanger 1968a:Figure 7a).  298  Rupert Harbour. Figure 122a shows the decorated tip of a digging stick handle. It is simply decorated with two eyes and a snout. Groups of parallel lines decorate the "neck" of the creature. This artifact was found along with an antler sap scraper and dentalium necklace with the remains of a young child. Figure 122b shows the handle of a whale bone club decorated with the typical two-headed motif of West Coast whale bone clubs. The next four carvings in steatite reveal close stylistic connections with seated human figure bowls and carvings from the Bell Site near Lillooet.  Figures 123a and b  suggest the depiction of bears or canines. They have thin, weakly developed forelimbs and crude definition of the lower body typical of sculpture from the Lower Fraser region. Figure 123a is similar to Figure 112a from the Bell site burial.  Figure 112b, discussed in the  previous section and identified as "from a graveyard near Shuswap", likely belong to the Chase burial collection. Figure 123b has a hole in the stomach or genital region and another hole in the base similar to several seated human figure bowl sculptures. Figure 124a is a steatite seated human-figure bowl sculpture, although the motif is crudely rendered. Many of the characteristic features of the sculptural complex are present including the seated posture, the bowl (in this case a shallow depression) in the belly of the figure, depiction of the clavicle, a snake running along the backbone and shoulder blades, and a hole drilled into the top of the head. Interestingly, the eyelids are long slits identical to the eyes of the rattlesnake/female from the Bell Site burial (Figure 111). The crudeness of the front section of the sculpture suggests that the chest and bowl may have cracked off and been recarved by unskilled hands. The steatite carving in Figure 124b shows a curious mixture of styles. The head is finely modelled in three dimensions to represent an eagle or hawk, while the legs are modelled in low relief along the side of the bowl, and the wings are two-dimensional leafshaped discs. The wings have been engraved with rows of lines indicating feathers. This same style of portraying disc-like wings projecting from the side of the carving is seen on an antler image presumed to be a lightning-snake from the Pedder Bay site (Figure 85a).  299 Figure 122. Chase Burial carvings. a, zoomorphic digging-stick handle (detail); b, whale bone club (Sanger 1968a:Figure 9a).  300 Figure 123. Chase Burial zoomorphic steatite carvings (Sanger 1968a:Plate VI).  a  301 Figure 124. Chase Burial steatite carvings. a, seated human-figure bowl; b, bird form bowKSanger 1968a:Plate VII).  302  Although bird-form bowls are common in the Lower Fraser and Lower Mainland region, they are usually reclining figures. The posture, upturned face, and hole in the top of the head of the Chase Burial bird bowl is identical to seated human figure bowls. Could the eagle head and odd wings point to the representation of a thunderbird or lightning-snake? The steatite sculptures from the Chase Burial site are clearly related to stone sculpture from the Lower Fraser and Gulf of Georgia region. The connections with coastal art styles, the discovery of a wooden mask and possible pecten shells associated with the sxwayxwey dance, the club forms, and the different treatment accorded the burials investigated by Sanger brings up the question posed for the Bell site burial material. Was the coastal system of inherited ranking more widespread among the Plateau Salish than previously thought? These questions will be addressed in the concluding chapter of this study.  303  Table V. Prehistoric Style Attributes, 1500-100 B.P.  # - f / / f / / * * n=77+: DECORATED FORMS harpoon points human face pendants bird form pendants hafts and handles small pestle/pendants "slave Jailer" clubs whale bone clubs stone bowls stonefish/whaleforms atlatl weights browbands Bcratchers/blanket pins comb stone effigy human figurine mat creaser spoon maul CARVING TECHNIQUES incising engraving block engraving inlaying modelling pecking sawing cut-out antler DESIGN PRINCIPLES Formline Forms urface DESIGN ELEMENTS crescents negative C-shape T-shape U-form split U-form wedges Salish eye chevrons joint ovals/circles MOTIFS human animal bird fish skull lightning-snake rib and backbone bear wolf/dog feather  62  # / / / / /  1 2 1 6 6 * 7 7 9 4 2 10 10 14 • ? - ? - + - ? - - - -  - - +  + + + - + - - - . + . . . .  ? - - - - - + - +  -  + + + + + . . . + . + . . + + + + . . < . +- - - - - - + - + + -  - + + + - + -  + + -  + + . .  . .  . .  -  -  + + . .  . .  . .  . + . . . . -  + -  -  + + . . .  . .  . -  + + + -  . .  . -  +  . . .  +  .  .  . .  . .  . . .  .  +  -  + + . .  . . . .  .  +  + -  + + + - - - + - + + - + - + - +  - -  + - +  + - +- + + - -  + +-  + -  -  -  -  + + + + + -  -  -  + + .  .  .  ? - ? + + - ? + - - - + - - - + + + + +. + + . . .  + - + ?  + +  . .  ? . +  + + . .  CHAPTER 6. CONCLUSION  The primary objectives of this study, as outlined in the introduction and methodology sections of this thesis are to describe the temporal and spatial characteristics of prehistoric Northwest Coast art, to examine the findings in the light of current theories and research questions in Northwest Coast archaeology, and to comment on the possible factors behind stylistic variation in the prehistoric art record. Each of these concerns is addressed in separate sections of this concluding chapter. The "summary" section gives an overview of stylistic developments in each of the four temporal periods representing the prehistoric cultural sequence for the Gulf of Georgia region.  Where possible the record for southwestern British Columbia is compared with  evidence from other regions of the province. The following "theoretical implications" section presents a model of the evolution of prehistoric art in British Columbia and addresses theories on the same subject presented by other researchers as discussed in Chapter 3. The "interpretations" section comments on the possible significance behind stylistic variation in decorated objects, evaluating ideas presented in Chapter 2 and proposing new hypotheses worthy of investigation in future archaeological work.  6.1 SUMMARY  Beginning at least 3500 years ago and continuing to the present day is a record of surprising continuity of motifs, carving styles, design elements, decorated forms, and principles of composition that links the prehistoric art record to the historic Native Indian art tradition known as Northwest Coast art. Only the Gulf of Georgia area has yielded an adequate regional sample of prehistoric art, and here we find evidence of the early  coalescence of the Coast Salish regional style variant of Northwest Coast art.  Table VI  summarizes stylistic developments in the prehistoric period.  The earliest art  Only a handful of decorated objects have survived from St. Mungo and Mayne phase sites dating from 4500 to 3300 years old. Despite the uncertain dating of these earliest decorated forms, they unequivocally show carving conventions and motifs common toalthough not exclusive to-Northwest Coast art.  These attributes include the rib and  backbone motif, circular joint marks, the use of inlays, and the deep engraving style of carving with abruptly stepped surface planes.  Although not illustrated in this thesis,  Carlson's (1986) new discoveries at the Pender Canal site provide additional examples of these attributes and confirm the use of circular and oval shell and stone inlays to decorate antler carvings. The Pender Canal finds also confirm that the inhabitants of Pender and Mayne Island had developed a distinctive antler carving style and set of motifs. Single, undiagnostic carvings from sites in the central and north coast do not allow stylistic comparison with the south, but do indicate that a decorative antler and bone carving industry existed at least 3500 years ago in these regions as well.  Locarno Beach phase art  Imprecise dating of Locarno Beach phase components and a small inventory of decorated objects make it difficult to hypothesize when or how artistic developments took place during this time period. But it is clear that near the end of the Locarno Beach phase approximately 2500 years ago an artistic vocabulary had evolved that represents the  306  Table VI. Stylistic Developments in Prehistoric Northwest Coast Art.  YEARS B.P.  DECORATED FORMS  CO  DESIGN PRINCIPLES  3500  z o  j j  spoons hafts  |  mauls?  j j  mat creasers harpoon points  deep engraving modelling inlaying cut-out antler  C3  z => Z  3000  DESIGN ELEMENTS  Biasing  UJ  i  CARVING TECHNIQUES  joint ovals  "V" engraving  UJ CO  o o o  complex eye-form formsurface formline  block engraving  U-form T-shape  2500  pecking chevrons wedges crescents  j browbands j pendants j wooden boxes, bowls j stone bowls | whale bone dubs | aflatls j pins/scraichers | "slave killer" clubs  2000  1500  |  combs negative C-shape  1000  a  o S 500  j wooden dubs | antler figurines j whale/fish effigies Northern formline  Split-U  coalescing of the Northwest Coast carving style. At the same time evidence for the Coast Salish regional style of carving is manifested in the block engraving technique seen on artifacts from the Musqueam N.E. site. A very small sample from the Yale, Prince Rupert, and Olympic Peninsula localities suggests, but in no way proves, that other regional preferences for materials, decorative forms and carving styles had developed. Four artifacts from the Musqueam N.E. site (shown in Figures 12a, 13, 16 and 17) provide most of the evidence for the developments taking place during this cultural phase including: the fragment of a raised feather motif identical to a motif seen in Coast Salish art, two fragments of block engraving identical to modern Coast Salish carvings, and two antler spoons exhibiting U-form and the negative T-shape elements. The spoon in Figure 19 also hints at the evolution of the formline, exhibited in this carving as a continuous raised surface joining and defining facial features within the composition. The formline concept appears to have evolved during the Locarno Beach phase out of the raised, positive lines created by the block engraving style, as shown in the raised lines of the feather motifs in Figure 17. The artist who carved the antler spoon in Figure 19 achieved the same effect of raised, positive lines without using a deep engraving or block engraving style. The antler spoons from the Musqueam Northeast site are identical in shape but not in carving style to spoons thought to date almost 1000 years earlier from the Pender Canal site. A carved bird's head from the Crescent Beach site exhibits the oldest Northwest Coast style complex eye form with an elongated eyelid, an angled V-shaped eyeball depression, and a round eyeball. The circular hollow of the eyeball (that may have once held an inlay), the hollowed-out head and the notched margins to the mouth and crown are identical to carving techniques seen in Pender Canal site carvings (Carlson 1986). These attributes are not seen in other Mainland Gulf of Georgia sites and hint at separate stylistic attributes in southeastern sites in the Gulf of Georgia region corresponding to Straits Salish territory.  308  A wooden mat creaser from a 2700 year old Locamo Beach phase site at Hoko River on the Olympic Peninsula shows the use of the block engraving technique, the use of decorative paint, and the double bird motif common in historic mat creasers. In other regions of the coast, there are only a few decorated items dating from 3300 to 2400 years old, but they reveal both significant similarities and what could be interpreted as already evolving regional preferences in decorated forms.  Similarities include deeply  engraved oval holes suggesting the use of inlays in a small stone pendant from the Prince Rupert area. More interesting are two Prince Rupert site concretions dating from 3000 to 2500 years old with backbone and vertebrae motifs similar to south coast carvings. Their phallus shapes are similar to a number of undated stone clubs from the Skeena and Bulkley River localities (Duff 1975:114-119). The Prince Rupert examples may represent the first manifestations of this unique northern sculptural form and a general interest in stone carving that may have been absent, or for which there is no present evidence, in Locarno Beach phase sites. Stone decorated forms are also found at the eastern border of the Gulf of Georgia region at the Milliken site. Several small, soft-stone carvings exhibit the familiar backbone, vertebrae and rib motifs. The forms achieve a three-dimensional sculptured quality by using rounded stones and adding shallowly pecked, sawn and incised body details. The stone medium, the zoomorphic forms, and the carving techniques are more akin to the small, stone sculptures of the late prehistoric record of the Yale and Lillooet localities than to the carvings from Mayne and Locarno Beach phase assemblages on the coast. A stone maul with pecked zoomorphic features from the Kamloops locality associated with a date of 3000 B.P. predates by one thousand years the earliest pecked mauls, but if accurate, would confirm the existence of an early stone carving industry in the Mid-Fraser and Thompson River region that blossoms later in time. The sample for this prehistoric time period is extremely small, therefore, more precise information on any of the sites mentioned above could affect the conclusions  309  presented here.  Nevertheless, the motifs, carving styles and design elements with their  single examples from sites with Locarno Beach phase components are well-represented in the following Marpole phase. Although more refined dating or new discoveries may move the dates for the first examples of particular stylistic attributes 500 years forward or backward, present evidence suggests that the fundamental Northwest Coast design concepts of form surfaces and formlines with raised positive design defined by negative carved-out spaces and standardized design elements had evolved by 2500 B.P. Motifs such as the rib and backbone and carving techniques such as inlaying also link the prehistoric and historic art tradition although it must be acknowledged that these attributes are not exclusive to Northwest Coast art. A question arises whether the stylistic attributes discussed above should be described as part of a coalescing Northwest Coast or Coast Salish art tradition, since the evidence for these attributes comes from sites in the south. I have used the term Northwest Coast in the previous paragraph because, with the exception of block engraving, the design concepts described above are common to the historic art traditions of all regions of the Northwest Coast of British Columbia.  Marpole phase art  Most of our knowledge of prehistoric art comes from Marpole phase components of Gulf of Georgia sites dating between 2400 and 1400 years old. The motifs, design elements and carving styles first seen in the Locarno Beach phase are well represented in Marpole phase art,  furthermore the repertoire is expanded. The large sample from this cultural  period shows an established set of style conventions. Decorated forms are standardized with specific motifs and carving styles used to decorate particular forms. Although there are only  310  a small number of similarly-aged carvings from other regions of British Columbia identical decorated forms, motifs, and design elements are found in these regions. New decorated forms such as stone bowls, harpoon points with zoomorphic images, pendants with human head and bird-form motifs, and pendants with pestle-like bases are found in three or more Gulf of Georgia sites. About half of the decorated forms seen in Marpole phase art are found in historic Northwest Coast art, while browbands and pendants fell out of use. Twenty stone bowls are associated with Marpole phase assemblages in the Gulf of Georgia region. Their forms include all five types known to occur in the region (Duff 1956a): seated human figure bowls, reclining human figure bowls, bowls with human faces, simple zoomorphic bowls and sculptured zoomorphic bowls. None of the bowls found in situ exhibit as refined a style of carving or design as exhibited in undated examples in museum collections. But the general forms of the undated bowls with wedge-shaped faces, large heads, headdresses or top-knots, open mouths, and weak limbs modelled in bas-relief are identical to the bowls found in situ in Marpole phase deposits.  As well, the facial  configurations and eye shapes seen in undated bowls are similar to anthropomorphic antler carvings from Marpole-age assemblages. On the basis of stylistic attributes any of the undated bowls could date to the Marpole phase, although this does not preclude them from dating to a later period of prehistory. Attributes not found on Marpole phase seated human-figure bowls that may have evolved later in time include secondary zoomorphic images on the front of the bowls and snakes portrayed in high relief along the backs of the human figures. There are single examples of a deeply engraved wooden box from the Esilao site and a bowl from the Biederbost site in the Skagit Delta. These finds indicate that carved boxes and bowls so ubiquitous in Northwest Coast art are at least 2000 years old. Although few prehistoric wooden decorated objects have been found, those preserved show that wood was carved in a manner similar to antler. (Although this may only hold true for the small-sized  wooden carvings found to date.) All four examples of decorated wood from Marpole-age sites are sophisticated in design and execution and compare well to historic examples of Coast Salish or Northwest Coast wood carvings. At Yuquot on the West coast of Vancouver Island and at Prince Rupert Harbour are examples of whale bone clubs with shapes and motifs similar to those of the historic period. The Boardwalk site assemblage in Prince Rupert Harbour also includes two clubs similar to historic "slave-killer" clubs (Wardell 1964:Plate 155). Fragments of what are thought to be both types of club forms are also found at the Marpole site. A highlight of the relatively large sample of Marpole phase art is the association of certain motifs with specific decorated forms: seated human-figures with stone bowls, human faces with pendants, elongated zoomorphic creatures with harpoon points, and bird shapes with what are thought to be toggles or pendants. For some of these decorated forms the association of the image with the object retained its potency over several millennia continuing to historic times. Examples are the lightning-snake figures depicted on harpoon points (Carlson 1983: Figure 11:1a; Figure 28g) and the persistence of the two headed profile on whale bone clubs. It is sometimes difficult to identify the anthropomorphic and zoomorphic images depicted in Marpole art. Recognizable motifs of the Marpole phase are limited to human, bird and lightning-snake images. There is great interest in the human figure, and in particular the human face. (It must be noted that these carvings have been termed "human" or "anthropomorphic" although they may represent non-human entities.) Facial features are usually rendered in detail while the rest of the body often receives perfunctory treatment. Several carvings of humans at the Marpole and Port Hammond sites show an eyelid form with short spurs projecting from the eye common in Coast Salish art. The bird, which is the second most common motif, has a variety of shapes: with long legs or long beaks suggesting herons or diving birds, with hooked beaks suggesting hawks or owls, or with nonspecific bird shapes. Both human and bird-forms are carved in a variety of  312  styles from simple outline shapes of birds to intricate three dimensional carvings. One can speculate that the emphasis on the open-mouthed human face portrayed on pendants and stone bowls of the Marpole phase finds its modern expression in the open-mouthed faces depicted on Coast Salish spindle-whorls, and that the bird motif continues in popularity in the ubiquitous bird-form mat-creasers. The third most common motif is thought to represent the lightning-snake and is usually given full descriptive rendering as a sinuous creature with horns or feather tufts, an open mouth, protruding tongue and slender crouched limbs. This creature is found in Marpole-age assemblages in sites as far north as Prince Rupert Harbour and south to the Skagit River. The lightning-snake or plumed serpent is depicted in prehistoric art of Mexico and southwestern United States, so although the motif has great antiquity on the Northwest Coast, it cannot be considered distinctive of this art tradition. Less certain among Marpole phase motifs are identifications of the thunderbird. The human face on the Skagit atlatl in Figure 65, the face on the wooden bowl in Figure 66 and the face on the atlatl in Figure 43a all with feather motifs behind the head are good candidates for the thunderbird motif as it sometimes appears in Northwest Coast art with human facial characteristics. The wooden box fragment from the Esilao site in Figure 67c, showing what appears to be a straight-edged wing similar to thunderbirds on Coast Salish spindle whorls, may be another representation of this figure. All of the carvings that have been interpreted as lightning-snakes and thunderbirds stand out in the sophistication of their design and execution.  The attention to detail and carving quality corresponds to  findings at the Ozette site where thunderbird motifs are neatly carved in a conventionalized deep engraving style and often adorn high status items. This relationship will be explored more fully in the interpretive section of this chapter. Design elements show a strong correlation to certain motifs and to the block engraving style of carving.  T-shapes are found in unidentified examples of engraved  formlines from the Marpole and Garrison sites.  Engraved wedge-shapes occur in two  313  contexts: to delineate feathers and to define the inner corners of eyelids.  Wedges are  associated with lightning-snake and thunderbird images in the prehistoric record. In Coast Salish art short wedges are used in a variety of animal and bird motifs but elongated wedges delineate the feathers of thunderbird figures. A design technique or motif used to define feathers is a raised formline doubled in on itself. First seen in a carved antler fragment from the Musqueam N.E. site dated to the late Locarno Beach phase, it is seen in the Marpole phase on carvings depicting lightningsnakes and birds. It is interesting to note that the wedge shape, T-shape, and feather motif are all found in unusually sophisticated and well-made compositions. Chevron and crescent elements found at the Marpole site are also correlated with lightning-snake figures as are checkered backbone and double backbone motifs from Gulf of Georgia and Prince Rupert sites. The double backbone motif is identical to double lines down the shafts of West Coast whalebone clubs (Boas 1907:Fig. 167b &c). There are examples where chevrons and checkered backbones appear in carvings of lightning-snakes in Coast Salish art, but crescents are also widely used in other designs. The U-form which first appeared at Musqueam Northeast on a spoon dated at least 2500 years old is not found in any carvings from Marpole age sites. Lastly, the engraved diamond shape seen in the wooden box from Esilao in Figure 67b-c is common to Nuu-chah-nulth rather than Coast Salish art. The conformity observed above would seem to be contradicted by the differences in carvings styles and quality of execution seen in Marpole phase art. Almost every site has antler or wood examples of simple incising, three-dimensional sculpting, two-dimensional block engraving or combinations of these techniques. Stone carvings are pecked, engraved, drilled, sawn and incised. This heterogeneity of carving techniques is a feature of Coast Salish art as well. Underlying this appearance of prehistoric heterogeneity in carving styles are some revealing consistencies. Roughly conceived and executed designs are often incised in antler and incised and sawn in stone. But deeply engraved antler and wood carvings and  pecked stone sculptures are generally well-made in terms of workmanship and design conception. A second correlation between good craftsmanship and particular motifs has been noted above. I believe there is evidence to suggest either that a limited number of carvers used the more sophisticated carving techniques, design elements, and principles of form, or that these related techniques and design conventions were reserved for carving higher status items and motifs. Although this study has not focused on rock art, a casual examination of petroglyphs from the Gulf of Georgia area reveals parallels with the motifs, design elements and compositional principles of Marpole phase art. Most examples of rock art can not be dated but on the basis of style fit comfortably within the stylistic parameters of Marpole phase art. This study has pointed out many hitherto unrecognized similarities between antler, wood and stone carvings in Marpole phase art, and between Marpole phase art and historic Northwest Coast art.  The formalized design principles, consistent associations between  specific motifs and particular objects, and rules pertaining to the use of design elements within engraved compositions attest to a fully-developed Northwest Coast art tradition during the Marpole phase. Many attributes of this prehistoric art tradition are found over a larger geographic area extending from the Skeena River to the Skagit River and from the coast to the gates of the Fraser Canyon.  Late period art  The sample of art from the late period dating from approximately 1400 to 150 years old is smaller than that of the previous period but better represents each region. Only seven poorly dated decorated items from the Lower Mainland area are attributed to the Late phase compared with 74 from Marpole phase sites. Despite the limited sample, general continuity with the preceding prehistoric period can be seen as well as several new traits distinguishing  315  the art of this time period.  Oddly lacking are examples of the regional variation that  characterizes historic Northwest Coast art. This may be partially attributed to the lack of preservation of wooden artifacts and large decorated forms such as masks, houseposts and mortuary sculpture. Decorated whale bone and antler "slave killer" clubs which first appeared in the archaeological record 2000 years ago are common in late period sites. Whereas whale bone clubs exhibited a range of motifs in the Marpole phase, in the Late period they closely conform to the double-headed thunderbird motif of historic clubs.  Their widespread  distribution seen in the previous period continues in this period in sites as far south as the Columbia River and east to the Chase Burial site in the Mid-Thompson River region. Several pendants are in the sample but only two steatite pendants from the Bell site conform to the standardized human face motif seen in the Marpole phase.  Small  anthropomorphic and zoomorphic effigies with no known function are common in this period. A number of antler figurines exhibiting a unique human form, carving style and range of decorative elements date to approximately 1000 B.P. The figurines have been found in 14 prehistoric sites in the present day territory of the Columbia River Chinook and further north in the territory of the Straits and Lushootseed Salish, with one example from Bella Coola.  Prominent features of the carvings are angular outline forms with elongated  foreheads, skirts, and numerous body ornaments or tattoos. Within the territories where the figurines have been recovered, head deformation and female facial tattooing were associated with higher status families (Ruby and Brown 1976:47; Barnett 1955:75). Stylistically similar figurines have been collected in the early historic period in the territory of the Lushootseed and Chinook. Outside of the Ozette site the recovery of wooden decorated objects is poorer for the late period than for any other. This might explain the puzzling absence of certain decorated forms such as bowls and tool hafts that are found in both the former prehistoric and latter historic periods. A wooden mask from the Chase Burial site is the only prehistoric example  found to date of this important historic decorated form. Importantly, the mask has peg-like eyes similar to Coast Salish sxwayxwey masks, although the lower half of the mask is dissimilar. A widely distributed new decorated form of late period components is the comb. It was used as a hair ornament and scratcher associated with male and female rites of passage amongst both coastal and interior Salish groups (Barnett 1955:151, Teit 1909:588). Combs from the Maplebank site near Victoria, a Prince Rupert Harbour site, the Bell site near Lillooet, and the Ozette site in Washington exhibit similar motifs and carving techniques with sections of antler cut out to create zoomorphic outlines. Combs are generally well-made with careful attention to carving technique and motif details. Decorated pins or ritual scratchers found at the Marpole site in the previous period are found in late period components at the Stselax, Pedder Bay, Kwatna, and Garden Island sites. The Stselax and Pedder Bay examples have a similar zoomorphic form while the central and north coast examples do not. Zoomorphic images inscribed on sedimentary rock have been found- in previous prehistoric periods but in the Late period a distinct form appears in two sites. The outlines shapes of sandstone whales or salmon forms have been found at the Pedder Bay and Pender Island sites, both in Straits Salish territory. Sandstone fish forms were found in Marpole and late Marpole phase deposits at the False Narrows site. Similarities between decorated forms in Interior Plateau sites and coastal sites have been integrated into the above discussion but notable differences are also evident.  While  there is an absence of stone sculpture in coastal sites this art form is well represented in the form of bowls, pipes, mauls and sculptured effigies at the Bell, Seton Lake and Chase Burial sites in the Mid-Fraser and Thompson River region. It is important to note that the welldated components containing decorated stone artifacts are from 1400 to 1200 years old and might be better associated with Marpole-age art, while stone art thought to be less than 500 years old occurs in poorly dated burials from the Chase, Milliken and Esilao sites.  A stone pipe bowl depicting a seated human from the Milliken site and a seated human-figure bowl from the Chase Burial site, although poorly dated as noted above, reveal that this popular motif of the Marpole phase survived into the Late prehistoric period. Carving conventions for stone art of the Marpole phase, with the limbs presented in basrelief while the head is strongly modelled in three dimensions, is found on stone bowls and stone zoomorphic effigies from the Bell and Chase Burial sites as well as in an undated maul from the Kwatna locality. Zoomorphic decorated digging-stick handles are unique to the Chase burial site although there are other undated examples from burials in the same region (Smith 1903). There are several motifs seen in the Marpole phase that continue to be found in assemblages from Late period sites. Representations of a feathered serpent thought to represent a lightning-snake are found at the Pedder Bay and Stselax sites, as well as on undated whalebone clubs from several Gulf of Georgia sites. The checkered backbone motif, associated with the lightning-snake in Marpole phase and Coast Salish art, is seen in an antler fragment from the Lochnore-Nesikep locality. The lightning-snake is much less common in Coast Salish art than the two-headed snake form. Two-headed images which may refer to this creature are found for the first time in late period components. A stone bowl from the Seton Lake site thought to be 1200 years old features an identical head at either end of a steatite bowl, and a club from the Fraser River dated to 1000 B.P. portrays a hook-beaked creature with folded limbs at one end and a creature with a feathered collar and forked tongue at the other. The rib and notched backbone motif so prevalent in the Marpole phase is rare in the Late period, while the raised, notched backbone does not occur. This may be due to the limited number of three-dimensional human-figure carvings found in situ in late period components. Bird images appear on a stone bowl from the Chase Burial site and in an antler fragment from the Stselax site but they are short, sharp beaked birds (perhaps ravens or owls) rather than the long-legged and long-beaked birds of the Marpole phase.  318  There are many more new motifs in this last prehistoric period than there are continuing similar themes. Four-legged animals such as dogs, wolves, and bears are found in interior and coastal sites. Whale images date to 1000 B.P. at the Pender Island site and are common at the Ozette site (both the Saanich and Makah were known to take whales). Seals are common on clubs at the Ozette site, and may be depicted on one carved fragment from the Pedder Bay site. Appearing for the first time are definite thunderbird images (although they were suspected in the Marpole phase) at the Ozette site and on whale bone clubs. Interestingly, the ubiquitous Coast Salish thunderbird image with the straight wing and crested head is not found in the prehistoric record of the Gulf of Georgia. A rattlesnake is portrayed on a half-human half-snake creature at the Bell site and a snake is portrayed along the backbone of the seated human-figure bowl from the Chase Burial site.  Several undated seated human-figure bowls portray rattlesnakes.  The  rattlesnake-human effigy along with other effigies from the Bell site portray female genitalia, not unusual in Northwest Coast art but not previously seen in prehistoric art. Antler human figurines discussed previously from Gulf of Georgia sites and human figure sculptures from the Ozette site depict male and female genitals. The above review of decorated forms and motifs from late period site components reveals relationships between certain decorated forms and particular motifs and carving styles as was seen in art from the Marpole phase.  Certain design elements are also  associated with particular motifs and the deep engraving style of carving. The negative C shape, a new design element appearing in five Late period assemblages, is associated with the eyes of snake-like creatures. It is seen on a carving thought to represent a "sea-monster" at Ozette, a bird or snake effigy from Lytton, a snake-like pin/scratcher from the Pedder Bay site, and snake figures on a spindle whorl from the Milliken site. The negative C is also seen in a carved eye design from the Garden Island site in Prince Rupert.  In the northern  example the C shape defines ovoid forms within a complex engraved eye composition.  319  All of the carvings with the negative C element are well-made deeply engraved designs which include engraved chevron, crescent and wedge motifs. As was seen in the Marpole phase, formal engraved design elements are associated with finely executed block engraved compositions, and may also be associated with the depiction of particular supernatural creatures.  Engraved wedge elements, T-shapes, and crescents are used to  decorate other well-made deeply engraved objects from Gulf of Georgia and Ozette site assemblages in such objects as combs, whale bone clubs and weaving staffs. The Prince Rupert eye design described above, thought to date from 950 to 750 years old, is the only occurrence of the ovoid design element in the prehistoric record. A similarlyaged bone comb fragment from the Kitandach site at Prince Rupert shows the first use of split U forms within a complex formline composition. It is in the large assemblage of decorated objects from the Ozette site where the relationship between the deep engraving style, specific design elements, and particular motifs is most clearly evident. Elongated eye shapes with wedges are used exclusively in the designs of whales, thunderbirds, and canines (likely wolves), while wedge shapes are also used as feather motifs on thunderbird designs. These motifs appear on what have been interpreted as high status items such as boxes, wall planks, and whale bone clubs. The only other items that are carved in the deep engraving style are combs and weavers' swords with canine, thunderbird, and snake designs. The significant relationship appears to be between the deep engraving style and the representation of particular creatures rather than between this style and particular types of objects since not all combs and weaving staffs have engraved designs. The use of sea otter tooth and snail opercula inlays is also reserved for objects depicting thunderbirds, whales and lightning-snakes. The sample from late period sites throughout the Northwest Coast and western Plateau has shown continuity with the previous periods with some new decorated forms, motifs, and design elements added to the artistic repertoire. The principles of design, range of design elements, and carving techniques seen in Marpole phase art continue as the  underlying stylistic vocabulary of late period art. Although the sample of art from this last prehistoric period is smaller than the previous period there is no evidence of a stylistic decline in the quality of carving, range of design elements, or adherence to Northwest Coast principles of design.  6.2 THEORETICAL IMPLICATIONS  The major finding of this study is that by the end of the Locarno Beach phase or beginning of the Marpole phase the essential character of the Northwest Coast art style had developed. Characteristics include an art style based on engraving where negative space is carved out to create a raised positive design. The positive design field is often connected in formlines or form surfaces, and the negative and positive shapes created in a composition utilize a small number of standardized design elements. There are new developments in the late period, but essential stylistic attributes described for Marpole phase art remains the same. How can developments in the prehistoric period be generalized to answer some of the long-standing questions and assumptions about the evolution of Northwest Coast art? This next section addresses the theoretical implications of the summary presented above.  Continuity A most important point regarding Northwest Coast art is that we have no prototype. There is no sense of development of an art style unless it can be proven that the stone carvings of the Fraser River area and Columbia River basin are the forerunners of the Northwest Coast style (Inverarity 1968:784). It [Marpole stone sculpture] does not seem to be a logical development out of the earlier Locarno Beach phase culture type...On a general level, stylistic elements of the stone sculpture assemblage have their most comparable analogues further to the north (Burley 1980:24).  321  Many researchers have not perceived the stylistic continuity within the prehistoric record and between prehistoric art of the Gulf of Georgia region and historic Central Coast Salish art. The absence in the prehistoric art record of masks, houseposts, spindle whorls and other common decorated forms of historic art can create an impression of dissociation between the historic and prehistoric record of art. Art historians and archaeologist are also unfamiliar with Coast Salish art and its relationship to Northwest Coast art, so in many cases they have been unable to perceive the expression of typical Northwest Coast design principles in both Coast Salish art and Gulf of Georgia region prehistoric art.  In the  instances where stylistic similarities have been noted between prehistoric art and Northwest Coast art-as in seated human-figure bowls~the resemblance has been attributed to north coast art rather than Salish art. The evidence presented in the summary suggests stylistic continuity from the late Locarno Beach phase to historic Coast Salish art. There is, as well, convincing evidence of the stylistic similarity between seated human-figure bowls and Marpole phase antler and stone art.  The relationship between north and south coast art The perception of profound differences between Coast Salish art and the art of central and north coast groups has resulted in an assumption that northern and southern art evolved independently, with the north developing the concepts of formlines and standardized design elements, and Salish art only later imitating these design principles. ...During the nineteenth century Coast Salish artists of the Lower Fraser and southeastern Vancouver Island were subjected to strong outside stylistic influences of radically different character from the simple geometric style that had been in vogue during the Late Period (Borden: 1983:165). Researchers (Boas 1950:295; Borden 1983:165; Drucker 1955:178; Wingert 1949:119) who have characterized the Salish style as realistic, simple, representational, three-dimensional,  322  and geometric; and north and central coast art as abstract, conventionalized, twodimensional, and symbolic have then searched for evidence of the antiquity, diffusion and interaction of the two styles. The evidence presented in this study suggests that, as far back as the record extends, three-dimensional, naturalistic forms and two-dimensional incising and engraving techniques have equal antiquity. The earliest prehistoric carvings in this study show the interplay between three-dimensional human and animal forms and overlying twodimensional engraved elements. Northwest Coast concepts of design have great antiquity in Gulf of Georgia prehistoric sites and may possibly have originated in this region. Is Coast Salish art representative of an older prehistoric art style? In its retention of the interest in engraved design over painted design, yes, but in the historic style of simple naturalistic sculptures, no. In northern Northwest Coast art the elegant northern formline and two-dimensional painted designs dominate the art, while the Coast Salish have retained an interest in deep engraving and generally use paint as a wash.  Although stylistic  continuity between the historic and prehistoric record has been stressed, Coast Salish art reveals some interesting stylistic changes. Instead of overlying three-dimensional human and animal forms with deeply engraved design elements, as seen in prehistoric art, the historic style generally keeps the two styles distinct, even within the same composition. Three-dimensional human and animal forms tend to be unembellished while flat surfaces are elaborately engraved. • The spindle-whorl in Figure 125 demonstrates how these two design principles are often integrated into the same composition: the human carved in low relief with its features defined by sculptural modelling and the background figures defined with engraved design elements. This study has been unable to shed light on the evolution of separate regional styles outside of the Gulf of Georgia. Unique regional motifs are evident in north coast and interior sites, and single examples of the ovoid and split-U design elements associated with the northern formline system appear in Prince Rupert sites at approximately 800 B.P.  323  Salish spindle whorl.  Holm has suggested two scenarios for stylistic developments in the prehistoric period: ...a direct, representational, two dimensional incised ancestor led, on the one hand, through the influence of painting to the sophisticated formline system, and on the other hand to a simple naturalistic sculpture, and finally to convergence of the two lines with formlines overlaying and modifying the sculptural forms. Another way of seeing it would be the two lines, sculptural and flat, developing together rather than coming together at a later time. It could work either way, but I think they did both develop from the flat system (1983:44). I propose that from the Locarno Beach phase onward the flat, engraved style and the threedimensional sculpture style developed together giving Northwest Coast art its distinctive character. The formline concept developed very early on out of the raised, positive lines created by deep engraving in antler. I concur with Holm that the northern formline design concept increased in sophistication with the stimulus of painting until it dominated two and three-dimensional north coast art. It is perhaps due to the unique qualities of antler that Northwest Coast art owes its evolutionary direction, since a section of antler provides both a two-dimensional surface that is easy to deeply engrave and a rounded three-dimensional form giving volume to the figures represented.  The relationship between the interior and the south coast  Several questions arise regarding the relationship between prehistoric art of the Interior Plateau and the Gulf of Georgia. Is there evidence, as Borden (1983:143) suggests, of a stone sculpture complex dating to 2800 B.P. at the Milliken site that was the stimulus for stone carving in the Marpole phase? The sample of stone sculpture at the Milliken site is small and poorly dated, and all but one are flat, crudely incised effigies. Other evidence of possible pre-Marpole stone sculpture from the Mid-Fraser and Thompson region consists of two zoomorphic mauls from the Kamloops Indian Reserve and the Lehman site associated  with components dated approximately 3000 and 2100 years old respectively. Neither of these finds are dated with certainty. It is not until approximately 1500 B.P., when stone sculpture is well-established on the coast, that small zoomorphic and anthropomorphic steatite and other soft-stone carvings appear in the Mid-Fraser and Thompson River region (Stryd 1983:175). There is not enough evidence of early stone sculpture in the Mid-Fraser and Thompson River region to suggest that art of this region would influence Marpole phase art. It may be the case that the use of steatite for sculpture in the Marpole phase originated in the Fraser Canyon where the mineral occurs naturally. Since steatite can be carved in a manner similar to antler it may have enhanced an already well-developed antler carving industry. Duff has suggested that steatite carving intensified and reached its most sophisticated development in the Yale-Lytton locality (1956a:92). As discussed in Chapter 5, the undated, sophisticated human figure carvings from this area fit within the range of Marpole stylistic attributes. Since this decorative style is seen only on the western border of the plateau it seems reasonable to assume that the impetus for this art style came from the coast. There is evidence that stone sculpture of the Mid-Fraser and Thompson River region exhibits unique motifs including rattlesnake, bear and female images that likely address quite different concerns from coastal art.  Furthermore, the interest in elaborate stone  carvings continued up until the contact period in the mid-Thompson region (as seen in the Chase Burial material) while it languished on the south coast. Another explanation would be that stone sculpture continued to be recovered in late period burial contexts in this region while burial customs changed on the coast. It is interesting to speculate on the possible relationship between burial customs in the Marpole phase on the south coast and late Lillooet and Kamloops horizon components in the Mid-Fraser and Thompson River region. All have examples of burials with elaborate grave inclusions including sculptures, utilitarian objects, and numerous personal ornaments.  326  Some of the individuals with extensive grave inclusions are infants and children who in an egalitarian society would not normally achieve the status that the rich burial goods imply. Might this burial pattern be characteristic of an incipient ranked society?  A Marpole phase artistic climax  It is one of the salient conclusions of this study that...the [prehistoric] stone sculpture is more highly developed than Salish wood sculpture. We must assume either that Salish wood sculpture has declined or that the former wood-carvers were not the present Salish tribes of the area (Duff 1956a: 108).  Duffs conclusions in his early study of prehistoric stone sculpture, together with Borden's (1983:158) proposal of a Marpole phase artistic climax and a late period post-climax "decline in artistic creativity" remain the prevailing model of artistic evolution for the Gulf of Georgia region. One of the primary conclusions of this study is that this model has not been supported and must finally be laid to rest. This model is refuted by evidence presented in Chapter 5 and earlier in this chapter. The evidence may be summarized as follows:  1. There is a prehistoric antecedent to Marpole phase stone sculpture in Locarno Beach and Marpole phase antler carving. 2. Sophisticated stylistic attributes seen in Marpole phase components continue undiminished in art of the late period and in Coast Salish art. 3. Marpole phase stone sculpture is stylistically closer to Coast Salish two-dimensional art rather than three-dimensional wood sculpture.  It is possible, as Borden and Duff have suggested, that the stone sculpture industry climaxed during the Marpole phase. As stone industries declined in the late period and were abandoned early in the historic period with the introduction of metal one would expect a parallel decline and abandonment of stone sculpture. But even this assumption is uncertain  due to the poor sample from late phase components and changes in burial patterns affecting the archaeological recovery of stone sculpture.  6.3  INTERPRETATIONS  In the conclusion to Chapter 2 there were two approaches identified that had potential for interpreting the cultural dynamics behind stylistic variation in prehistoric art of the Northwest Coast. The first approach uses ethnographic information relating to specific ritual complexes to assign meaning to similar material culture patterns in the prehistoric record. The second approach views stylistic variation as a strategy of information exchange where messages relating to group membership are relayed through objects.  The two  perspectives overlap in their interest in the expression of material culture patterns associated with particular ethnic groups.  Ethnographic Analogy  There is an assumption by some researchers that the Northwest Coast art style developed as a result of the unique social forms that evolved among coastal groups, therefore prehistoric examples of this style indicate the prehistoric achievement of Northwest Coast cultural patterns. Boas was the first anthropologist to suggest that the Northwest Coast art style was developmentally related to the northern crest system and the validation of rank and privileges. He postulated that the use of crests had stimulated the motifs and style of art that characterizes Northwest Coast art (1950:281). MacDonald (1983) and Ham (1985) have also implied that decorated objects in the prehistoric record denote a prehistoric society with complex social forms similar to historic  328  Northwest Coast groups. Ham (1985:121) cites the discovery of an incised zoomorphic stone image in a St. Mungo phase component as evidence of a complex society. While MacDonald (1983:104) regards the absence of decorated objects in early prehistoric components of Prince Rupert sites as an indication that "graphic expressions of corporate identity" had not yet become meaningful. This study has commented on the fact that the development of essential characteristics of the Northwest Coast art style coincides in time with the coalescence of Northwest Coast economic and demographic patterns.  Can we then use evidence of  Northwest Coast stylistic attributes as a sign of prehistoric achievements in social and economic spheres as some researchers have done? I do not believe so. First of all, although this study has asserted that key stylistic attributes of Northwest Coast art are seen by approximately 2500 B.P. some of these same attributes can be found in older St. Mungo/Mayne phase components before the Northwest Coast cultural pattern developed. It would be inaccurate to use art as a sign of the relationship between decorated objects and corporate groups or complex societies in prehistoric cultural phases when other evidence of social complexity are unproven. Although comparisons on a general level between art and society in historic and prehistoric contexts may not be valid, can specific decorated forms associated with historic ritual complexes be used as evidence of prehistoric ritual behavior? This study's findings suggest that single attributes on their own are not sufficiently unique either temporally, geographically or culturally to serve as indicators of particular ritual practices or ethnic preferences. On the other hand, when a cluster of two or three associated stylistic attributes matches an arrangement seen in historic Northwest Coast art~as in the deeply engraved thunderbird images carved on historic and prehistoric whale bone clubs-then a case can be made for assigning a similar meaning or function to the prehistoric decorated object. Carlson's (1983) research has focussed on four concerns that he postulates are the stimulus for Northwest Coast art: secret societies, shamanic spirits, crest spirits, and  329  guardian spirits. His assumptions and methods were discussed in Chapter 2. What can now be evaluated is whether his assertions are supported by this study's findings. Small hollowed-out antler anthropomorphic and zoomorphic heads from the Crescent Beach, Helen Point, and Milliken sites are cited as evidence (Carlson 1983:203) of miniature masks associated with secret societies. Unfortunately only one of these effigies is securely dated. Two additional undated zoomorphic effigies of this type were examined in this study. Although it is possible that these images represent masks this study has been unable to find any other evidence of masks in the prehistoric record. More problematic for Carlson's hypothesis is the lack of ethnographic evidence for the use of zoomorphic masks or miniature mask effigies in the Central Coast Salish area. Carlson, Burley and Sanger have cited evidence of the possible prehistoric occurrence of sxwayxwey masks including a human face effigy with wide eyes surmounted by a bird's head from the Helen Point site (Figure 81b), punctured pecten shells from the False Narrows site similar to those used by sxwayxwey dancers, and a miniature wooden mask (Figure 119) with protruding eyes and punctured pecten shells at the Chase Burial site. Since pecten shell rattles are also used among the Nuu-chah-nulth and Coast Salish (Suttles 1982:64) in more general rituals they alone cannot be used as evidence of the sxwayxwey complex. Although one or two characteristics of the Chase Burial and Helen Point carvings may be similar to sxwayxwey masks attributes such as the distinctive facial configuration and carving style of this mask are missing. This study has found more examples of deeply engraved zoomorphic images on Marpole phase barbed harpoon points that Carlson (1983:199) has postulated represent guardian spirit helpers. I believe there is some correspondence between the prehistoric and historic contexts of this decorated form supporting Carlson's hypothesis. But we should be not assume that this correspondence has significance beyond its particular context. This does not constitute prehistoric proof of other aspects of the guardian spirit complex (such as  the vision quest) or that the motif represents a guardian spirit in other prehistoric contexts as Carlson (1983:200) has suggested. Stylistic similarities between carved antler figures from Gulf of Georgia and Prince Rupert area sites and Siberian shamanic emblems have been posited but not documented (Carlson 1983:201, Keddie 1987:2, MacDonald 1983:104). No other evidence of shamanism or ritualism has been found in this study except already documented decorated combs, scratchers and stone bowls which are associated with, although they are not exclusive to, Coast Salish ritual practices. Is there evidence of combinations of associated stylistic attributes that can indicate historic cultural practices or the stylistic preferences of a regional group? Although I believe that these correspondences will eventually be found, the present sample of decorated objects is too small to point to any ritual-specific or ethno-specific combination of stylistic variables. For example, there is presently no evidence to suggest that the distinctive motifs and forms of whale bone clubs have greater antiquity among West Coast groups. Although motifs, carving styles, principles of composition, and design elements similar to Coast Salish art are well-established in Marpole phase art only the checkered-backbone motif may be exclusive to Coast Salish art. The block engraving style is seen at the Ozette site in Makah territory so cannot be considered distinctly Salish, but a case may be made for close connections between deeply engraved curvilinear designs as seen in Figure 36a and those seen in Salish spindle whorls. As more evidence from other regions of the coast is uncovered we will be able to comment further on whether the styl