Ethnography, Analogy, and Arctic Archaeofaunas: Assessing the Limits of Zooarchaeological Interpretation
by
Lauren E.Y. Norman
A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy Department of Anthropology University of Toronto
© Copyright by Lauren E. Y. Norman 2015
Ethnography, Analogy, and Arctic Archaeofaunas: Assessing the Limits of Zooarchaeological Interpretation
Lauren E. Y. Norman
Doctor of Philosophy
Department of Anthropology University of Toronto
2015 Abstract
The use of analogy to infer past lifeways from archaeological material is integral to many types of archaeological investigation. There are many sources for analogy, but the ones that offer some of the richest interpretations use ethnographic and ethnohistoric records to understand archaeological materials and their patterns. The use of these documentary records has been particularly beneficial in places where contemporary people can be linked to their archaeological ancestors through material culture using the direct-historical approach. However, the use of documentary sources in constructing past lifeways has been critiqued, with questions raised about the use of a synchronic, subjective record of a rapidly changing historical present to infer the normally fragmentary and palimpsestic archaeological material. This study aims to clarify the use of the documentary record in interpreting dwelling activities from the archaeofaunal record.
By testing which ethnographically and ethnohistorically documented practices are visible archaeozoologically, archaeologists can identify activities, practices, and behaviours that can be accurately interpreted from the archaeological record using the documentary record.
The Arctic is an ideal location to study the use of the analogy in the archaeological record as it has both a detailed documentary record and a well-preserved archaeological record. This study
ii uses the direct-historical method to develop archaeofaunal expectations from the documentary record. Expectations for archaeozoological material were created to test for multiple stages of dwelling use: primary activities drawn directly from the documentary record, contemporary activities potentially invisible in the documentary record, and post-depositional activities.
Archaeofaunal materials from a fourteenth century Thule Inuit semi-subterranean dwelling at
Cape Espenberg, Alaska, were used to test these expectations. The strongest patterns that emerged were those relating to the primary activities directly reconstructed from the documentary record, specifically those relating to food storage, preparation, and consumption.
Although this is a single study, it indicates that archaeofaunal patterns can help differentiate between activity areas in houses, and that activities that relate to domestic subsistence practices are similar in the early Thule period and the historic period. More broadly, it also suggests that documentary records can be used to accurately interpret archaeofaunal patterns relating to food storage, preparation, and consumption.
iii
Acknowledgments
This work could have never been completed without the support of many people. First, I want to thank my supervisor, Max Friesen. Max introduced me to the Arctic and the academic world. He continually supports my goals and educates me in the ways of a successful academic. His sound advice, humor, and attention to detail have made my research, writing, and academic work infinitely better than I thought they could be. Max’s guidance through this process was invaluable, and the opportunities he afforded me to excavate, identify, analyze, and teach have given me experience that goes beyond a simple graduate degree. I cannot thank you enough.
I would also like to thank Gary Coupland, whose comments and questions about the Arctic ethnographic and zooarchaeological work on this dissertation and in class discussions have helped me clarify my ideas and connect the Arctic to the greater word. My conversations with Genevieve Dewar about zooarchaeological method and theory in class and in the Zooarchaeology Interest group continually challenge my interpretations of animal remains. David Smith and Liye XIe deserve many thanks for their thoughtful observations and suggestions. I want to thank Dr. David Morrison, the external examiner, for his comments, insights, and questions.
This work would have never been undertaken without the Cape Espenberg Project, whose co- principle investigators John Hoffecker and Own Mason invited Max and a few graduate students up to the amazing site of Cape Espenberg on the Seward Peninsula for the summer of 2011. I would particularly like to thank Owen Mason, John Hoffecker, Claire Alix, Christyann Darwent, Anne Jensen, and Shelby Anderson for their insights on the site and excavations. The field crew that excavated and mapped Feature 87 was invaluable: thank you Angélique Neffe, Arygris Fassoulas, and all the others who pitched in at the end. I particularly want to thank Michael O’Rourke, whose support in the field and in the lab was instrumental in this project. Mike digitized the Feature 87 maps, created the house map, and taught me the basics of GIS data analysis; without his support and help I would have not been able to start or complete this project.
I would like to express my gratitude to my fellow lab-mates, new and old. Specifically, I want to thank Lesley Howse and Susanne Needs-Horwarth for their insights into identification, analysis,
iv and all things animal bone related. My thanks go out to all the graduate students for their friendship and support: Alyson Holland, Danielle Macdonald, Jessica Taylor, Peter Bikoulis, Andrew Riddle, Lucille Harris, Adam Allentuck, Emily Hubbard, Danii Desmaris, Becky Goodwin, and many more!
The Cape Espenberg Project was funded by NSF grant (ARC-0755725) to John Hoffecker, Owen Mason, Nancy Bigelow, and Christyann Darwent. I have gratefully received substantial funding for my graduate work from the Social Sciences and Humanities Research Council of Canada, the Ontario Graduate Scholarship, and the Northern Science Training Program.
Grants from the Department of Anthropology, University of Toronto, allowed me to identify small seals at the Museum of the North, University of Alaska Fairbanks, where Aren Gunderson and Link Olson shared with me their knowledge on Alaskan mammalogy. Mark Peck also helped me identify small Charadriiformes at the Royal Ontario Museum, Department of Natural History.
I am thankful for my partner, Devin Clarke, who would take me out camping or recommend a glass of wine when I needed it the most. Finally, I want to thank my parents, Joan Young and Ron Norman and my sister Julia Norman for their unending love and encouragement in this and all things.
v
Table of Contents
Acknowledgments ...... iv
Table of Contents ...... vi
List of Tables ...... xvi
List of Figures ...... xviii
Chapter 1 Introduction ...... 1
1.1 The Use of Analogy in Archaeological Interpretation ...... 1
1.1.1 Limitations of Analogy ...... 4
1.2 Household Archaeology or Archaeology of a House? ...... 7
1.3 Brief History of Spatial Studies ...... 11
1.4 Spatial Studies of Archaeofaunal Assemblages ...... 12
1.5 Spatial Patterning in Houses ...... 15
1.6 Research Materials ...... 18
1.7 Research Goals ...... 18
1.8 Organization of Dissertation ...... 19
Chapter 2 The Cultural Setting ...... 20
2.1 Terminology ...... 21
2.2 The Inuit Tradition ...... 22
2.2.1 Okvik-Old Bering Sea ...... 23
2.2.2 Punuk ...... 24
2.2.3 Birnirk ...... 25
2.2.4 Thule ...... 27
2.3 Thule Inuit Origins ...... 28
2.3.1 Thule Origins in Context ...... 31
2.4 Thule Inuit Expansion ...... 35
vi
2.5 Thule Inuit Regionalization ...... 37
2.5.1 Modified Thule Inuit ...... 37
2.5.2 Historic Inuit ...... 40
2.6 Summary ...... 41
Chapter 3 Thule Lifeways ...... 42
3.1 Thule Way of Life ...... 42
3.1.1 Interaction and Social Organization ...... 43
3.1.2 Subsistence Strategies ...... 45
3.1.2.1 Marine Mammal Procurement Strategies ...... 46
3.1.2.2 Terrestrial Procurement Strategies ...... 48
3.1.3 Processing Strategies ...... 49
3.1.3.1 Butchering ...... 49
3.1.3.2 Storage ...... 50
3.1.3.3 Consumption ...... 51
3.1.3.4 Discard ...... 51
3.1.4 Dwellings ...... 51
3.1.4.1 Warm-Season Dwellings ...... 52
3.1.4.2 Cold-Season Dwellings ...... 52
3.1.5 Thule Way of Life Summary ...... 53
3.2 Documentary Records for the Seward Peninsula and Kotzebue Sound ...... 53
3.2.1 Documentary Sources ...... 55
3.2.2 Limitations of the Northwest Alaskan Documentary Record ...... 58
3.3 Northwest Alaskan House Histories ...... 60
3.3.1 House Construction ...... 60
3.3.2 House Activities ...... 62
3.3.2.1 Sleeping ...... 62 vii
3.3.2.2 Consumption ...... 63
3.3.2.3 Cleaning ...... 65
3.3.2.4 Working ...... 65
3.3.2.5 Socializing ...... 66
3.3.3 Moving Out ...... 66
3.3.4 Re-Use/Re-Building ...... 66
3.3.5 Abandonment and Burial ...... 67
3.3.6 Subsequent Disturbances ...... 68
3.4 Summary ...... 68
Chapter 4 Cape Espenberg Context ...... 69
4.1 Environmental Context ...... 69
4.1.1 Geology ...... 69
4.1.2 Climate ...... 72
4.1.3 Ecology ...... 73
4.1.3.1 Marine Ecology ...... 74
4.1.3.2 Terrestrial Ecology ...... 78
4.1.3.3 Freshwater Ecology ...... 80
4.2 Social Context ...... 81
4.3 Archaeological Context: History of Research ...... 86
4.4 Early Thule Inuit Semi-Subterranean Dwellings ...... 88
4.4.1 Cape Espenberg ...... 89
4.4.2 Cape Krusenstern ...... 89
4.4.2.1 Houses 7 and 8 ...... 89
4.4.2.2 Houses 4, 5, and 6 ...... 91
4.4.2.3 Houses 25a, 25b, 26, 27, and 29 ...... 92
4.4.3 Deering ...... 95 viii
4.4.4 Qimiarzuq ...... 96
4.4.5 Utqiagvik ...... 97
4.4.6 Eastern Arctic ...... 97
4.4.7 Summary of Early Thule Early Thule Inuit Semi-Subterranean Dwellings ...... 99
4.5 Description of Feature 87 ...... 100
4.5.1 Feature 87 History ...... 100
4.5.2 House Description ...... 102
4.5.2.1 Walls ...... 103
4.5.2.2 Floor ...... 103
4.5.2.3 Back Platform ...... 104
4.5.2.4 Side Benches ...... 104
4.5.2.5 Kitchen Area ...... 106
4.5.2.6 Katak ...... 106
4.5.2.7 Tunnel ...... 107
4.5.2.8 Midden ...... 107
4.5.3 Discussion ...... 107
4.6 Summary ...... 110
Chapter 5 Archaeofaunal Expectations ...... 112
5.1 Stages ...... 116
5.1.1 Primary Stage: Winter Village Faunal Assemblage Composition ...... 116
5.1.2 Secondary Stage: Household Activities ...... 117
5.1.3 Tertiary Stage: Contemporary Activities ...... 118
5.1.4 Quaternary Stage: Post-Abandonment Activities ...... 118
5.2 Expectations ...... 119
5.2.1 Primary Stage for All Areas ...... 119
5.2.1.1 Large Marine Mammals ...... 119 ix
5.2.1.2 Small Seals ...... 120
5.2.1.3 Large Terrestrial Mammals ...... 121
5.2.1.4 Small Terrestrial Mammals ...... 122
5.2.1.5 Birds ...... 122
5.2.1.6 Fish ...... 122
5.2.1.7 General ...... 123
5.2.2 Tunnel Expectations ...... 123
5.2.2.1 Primary Activities ...... 123
5.2.2.2 Secondary Activities ...... 123
5.2.2.3 Tertiary Activities ...... 125
5.2.2.4 Quaternary Activities ...... 129
5.2.3 Main Room Floor Expectations ...... 131
5.2.3.1 Primary Activities ...... 131
5.2.3.2 Secondary Activities ...... 131
5.2.3.3 Tertiary Activities ...... 134
5.2.3.4 Quaternary Activities ...... 137
5.2.4 Kitchen Expectations ...... 138
5.2.4.1 Primary Activities ...... 138
5.2.4.2 Secondary Activities ...... 138
5.2.4.3 Tertiary Activities ...... 141
5.2.4.4 Quaternary Activities ...... 143
5.2.5 Benches Expectations ...... 144
5.2.5.1 Primary Activities ...... 144
5.2.5.2 Secondary Activities ...... 144
5.2.5.3 Tertiary Activities ...... 149
5.2.5.4 Quaternary Activities ...... 152 x
5.2.6 Midden Expectations ...... 152
5.2.6.1 Primary Activities ...... 152
5.2.6.2 Secondary Activities ...... 152
5.2.6.3 Tertiary Activities ...... 155
5.2.6.4 Quaternary Activities ...... 157
5.3 Summary ...... 158
Chapter 6 Methodology ...... 160
6.1 Collection ...... 160
6.2 Identification ...... 162
6.3 Analysis ...... 163
6.3.1 General Subsistence Analysis ...... 164
6.3.2 Expectation Testing ...... 165
6.4 Some Definitions and Discussions ...... 165
6.4.1 Taxonomic Abundance Measures ...... 165
6.4.2 Skeletal Measures ...... 168
6.4.3 Bone Density and Utility Indices ...... 169
6.4.4 Taxonomic Composition ...... 170
6.4.5 Modification Analysis ...... 170
6.4.6 Fragmentation ...... 170
6.4.7 Expectation Indices ...... 171
6.4.8 Sample Size ...... 172
6.4.9 Numerical and Statistical Issues ...... 173
6.5 Analytic Framework ...... 173
6.6 Summary ...... 175
Chapter 7 Overall Subsistence and Economy ...... 176
7.1 Defining the Floor: Fill and Floor Comparison ...... 176 xi
7.1.1 Archaeological Context ...... 176
7.1.2 Density-Mediated Attrition ...... 178
7.1.3 Peri-Depositional Damage ...... 179
7.1.4 Breakage and Fragmentation ...... 181
7.1.5 Visible Human Modifications ...... 181
7.1.6 Element Representation ...... 182
7.1.7 Taxonomic Representation ...... 184
7.2 Fill and Floor Comparative Analysis Discussion ...... 189
7.3 General Subsistence for Feature 87 ...... 192
7.3.1 Density-Mediated Attrition ...... 192
7.3.2 Peri-Depositional Damage ...... 195
7.3.3 Breakage and Fragmentation ...... 195
7.3.4 Visible Human Modifications ...... 197
7.3.5 Element Representation ...... 199
7.3.5.1 Small Seals ...... 200
7.3.5.2 Caribou ...... 201
7.3.5.3 Canidae ...... 204
7.3.5.4 Bird ...... 206
7.3.5.5 Fish ...... 206
7.3.6 Taxonomic Representation ...... 206
7.3.6.1 Mollusca ...... 207
7.3.6.2 Actinopterygii ...... 207
7.3.6.3 Aves ...... 210
7.3.6.4 Mammalia ...... 211
7.4 General Subsistence Discussion ...... 219
7.5 Summary ...... 221 xii
Chapter 8 Spatial Patterning of Faunal Remains: Testing Expectations ...... 223
8.1 Testing Expectations ...... 225
8.2 General Results ...... 228
8.3 Tunnel Results ...... 230
8.3.1 Expectations Relating to Storage Activities ...... 231
8.3.2 Expectations Relating to Density-Mediated Attrition Activities ...... 234
8.3.3 Summary of Tunnel Expectations ...... 235
8.4 Main Room Floor Results ...... 236
8.4.1 Expectations Relating to Butchery Activities ...... 236
8.4.2 Expectations Relating to Cleaning Activities ...... 239
8.4.3 Expectations Relating to Density-Mediated Attrition Activities ...... 241
8.4.4 Summary of Main Room Floor Expectations ...... 241
8.5 Kitchen Results ...... 243
8.5.1 Expectations Relating to Preparation Activities ...... 243
8.5.2 Expectations Relating to Density-Mediated Attrition Activities ...... 246
8.5.3 Expectations Relating to Burning Activities ...... 247
8.5.4 Summary of Kitchen Expectations ...... 250
8.6 Benches Results ...... 250
8.6.1 Expectations Relating to Consumption Activities ...... 252
8.6.2 Expectations Relating to Manufacturing Activities ...... 254
8.6.3 Expectations Relating to Cleaning Activities ...... 254
8.6.4 Summary of Benches Expectations ...... 255
8.7 Midden Results ...... 257
8.7.1 Expectations Relating to Primary Butchery Activities ...... 257
8.7.2 Expectations Relating to Cleaning Activities ...... 259
8.7.3 Expectations Relating to Density-Mediated Attrition Activities ...... 259 xiii
8.7.4 Summary of Midden Expectations ...... 261
8.8 Summary ...... 261
Chapter 9 Conclusion ...... 263
9.1 Contributions to Method and Theory of Analogical Reasoning ...... 264
9.2 Contributions to Spatial Studies ...... 268
9.3 Contributions to Thule Subsistence ...... 270
9.4 Limitations and Future Work ...... 272
9.5 Concluding Remarks ...... 275
References ...... 276
Appendix A Overall Subsistence Analysis ...... 333
1 Squares Used in General Subsistence Analysis ...... 333
2 Calculations of MNE and MNI ...... 334
3 Calculations of Meat Weight ...... 337
4 Calculations of MAU for General Subsistence Analysis ...... 339
4.1 Fish MAU ...... 339
4.2 Bird MAU ...... 340
4.3 Dog/Wolf MAU ...... 340
4.4 Fox MAU ...... 341
4.5 Bearded Seal MAU ...... 341
4.6 Small Seal MAU ...... 342
4.7 Caribou MAU ...... 342
5 Bone Density and Utility Index Values ...... 343
Appendix B Expectations Testing and Spatial Patterning Analysis ...... 346
1 Squares Used in Spatial Patterning Analysis ...... 346
2 Quadrants per Archaeologically-Defined Area ...... 347
3 Expectation Calculations ...... 347 xiv
4 Elements Used in Calculations ...... 352
xv
List of Tables
Table 1: Radiocarbon dates for Feature87 ...... 101
Table 2: Activity frequencies ...... 116
Table 3: Summary of expectations from the tunnel. Note that all expectations are cumulative; the activities from each stage only affect the bones present in the previous stage...... 124
Table 4: Summary of expectations from the main room floor. Note that all expectations are cumulative; the activities from each stage only affect the bones present in the previous stage. 132
Table 5: Summary of expectations from the kitchen. Note that all expectations are cumulative; the activities from each stage only affect the bones present in the previous stage...... 139
Table 6: Summary of expectations from the benches. Note that all expectations are cumulative; the activities from each stage only affect the bones present in the previous stage...... 145
Table 7: Summary of expectations from the midden. Note that all expectations are cumulative; the activities from each stage only affect the bones present in the previous stage...... 154
Table 8: Squares screened through larger mesh (1/4-inch or 6mm) ...... 161
Table 9: Fill and floor taxonomic frequencies: whole values (NISP, MNI) and relative values (%NSP, %NISP, %MNI) (Table on previous page) ...... 186
Table 10: Diversity indices (NISP of the lowest taxonomic order) ...... 186
Table 11: Genereal subsistence taxonomic frequencies. Whole values (NISP, MNI) and relative values (%NSP, %NISP, %MNI) ...... 209
Table 12: Exepctations confirmation, total ...... 228
Table 13: Expectations confirmation, skeletal and comparative ...... 229
Table 14: Expectations confirmation, ratios and frequencies ...... 229
Table 15: Expectations confirmation, absolute and relative tests ...... 229 xvi
Table 16: Taxonomic proportions by area (%NISP) ...... 230
Table 17: Modification proportions by area (%NISP) ...... 230
Table 18: Indeterminate and size proportions by area (%NISP) ...... 230
Table 19: Tunnel expectation results. Expectations are placed in taxonomic categories regardless of their alphabetical label. See Chapter 5 for a detailed discussion of the construction of the expectations...... 232
Table 20: Main room floor expectation results. Expectations are placed in taxonomic categories, regardless of their alphabetical label. See Chapter 5 for a detailed discussion of the construction of the expectations...... 237
Table 21: Kitchen expectations results. Expectations are placed in taxonomic order regardless of their alphabletical label. See Chapter 5 for a detatiled discusion of the construction of the expectations...... 245
Table 22: Benches expectation results. Expectations are placed in taxonomic categories regardless of their alphabetical label. See Chapter 5 for a detailed discussion of the construction of the expectations...... 251
Table 23: Midden expectations results. Expectations are placed in taxamonic categories regardless of their alphabetical label. See Chapter 5 for a detailed discussion of the construction of the expectations...... 258
xvii
List of Figures
Figure 1: Harpoon head developmenet from Birnirk to Thule (from Stanford 1976:108) ...... 30
Figure 2: Location of Cape Espenberg and other sites from the text...... 70
Figure 3: Cape Krusenstern, House 7 (from Giddings and Anderson 1986:73) ...... 89
Figure 4: Cape Krusenstern, House 8 (from Giddings and Anderson 1986:79) ...... 90
Figure 5: Cape Krusenstern, House 4 (from Giddings and Anderson 1986:73) ...... 91
Figure 6: Cape Krusenstern, House 6 (from Giddings and Anderson 1986:73) ...... 91
Figure 7: Cape Krusenstern, House 5 (from Giddings and Anderson 1986:73) ...... 92
Figure 8: Cape Krusenstern, House 25a (right) and 25b (left) (from Giddings and Anderson 1986:61) ...... 93
Figure 9: Cape Krusenstern, Houses 29 (from Giddings and Anderson 1986:61) ...... 94
Figure 10: Cape Krusenstern, House 26 (from Giddings and Anderson 1986:65) ...... 94
Figure 11: Cape Krusenstern, House 27 (from Giddings and Anderson 1986:65) ...... 95
Figure 12: Qimiarzuq (Jabbertown), House 2 (from Larsen and Rainey 1948:171) ...... 96
Figure 13: Utqiagvik, House A (from Ford 1959:68) ...... 97
Figure 14: Cache Point, House 8 (from Friesen and Betts 2006:67) ...... 97
Figure 15: Tiktalik, House 5 (from Moody and Hodgetts 2013:8) ...... 98
Figure 16: Nelson River (from Friesen and Arnold 2008:530) ...... 98
Figure 17: Calibrated radiocarbon dates from Feature 87 ...... 102
Figure 18: Plan view of Feature 87 and its location on the Seward Peninsula...... 106
Figure 19: Archaeologically defined areas based on quadrants, Feature 87 ...... 166 xviii
Figure 20: Fill, small seal bone density; rs=0.45, p=.52; MAU=4; MNE=255 ...... 178
Figure 21: Floor, small seal bone density; rs=0.65, p=<.01; MAU=4; MNE=305 ...... 178
Figure 22: Fill and floor natural modification frequencies (%NISP) (Fill n=722; Floor n=498) 179
Figure 23: Fill and floor fragmentation (%NISP) (Fill n=722; Floor n=498) ...... 181
Figure 24: Fill and floor cultural modifications (%NISP) (Fill n=722; Floor n=498) ...... 181
Figure 25: Fill and floor small seal element distribution (%MAU) (Fill MNI=4, MNE=255; Floor MNI=4, MNE=305) ...... 182
Figure 26: Fill and floor small seal meat utility index (Fill MNI=4, MNE=255; Floor MNI=4, MNE=305) ...... 183
Figure 27: Fill bird element distribtuion (MAU) (MNI=5; MNE=50) ...... 184
Figure 28: Fill and floor vertebrae class frequencies (%NISP) (Fill n=772; Floor n=489) ...... 187
Figure 29: Fill and floor vertebrate taxonomic frequencies (%NISP) (Fill n=450; Floor n=413) ...... 188
Figure 30: Fill and floor vertebrate taxonomic frequencies (%MNI) (Fill MNI=37; Floor MNI=29) ...... 188
Figure 31: Small seal bone density ...... 194
Figure 32: Caribou bone density ...... 194
Figure 33: Bird bone density ...... 194
Figure 34: Fish bone mineral content ...... 194
Figure 35: General subsistence natural modification frequencies (%NISP) (NISP=23462) ...... 195
Figure 36: General subsistence fragmentation index (%NISP) (Total house NISP=23462; House, no kitchen NISP=10832, Kitchen NISP=12630) ...... 195
xix
Figure 37: Fresh fracture index (FFI) for identified land mammal cortical bone (n=157) ...... 196
Figure 38: General subsistence cultural modification frequencies (%NISP) (Total house NISP=23462; House, no kitchen NISP=10832, Kitchen NISP=12630) ...... 197
Figure 39: Burning frequencies by vertebrate class (%NISP) (Total n=23457) ...... 198
Figure 40: General subsistence burnt bone by size category and class (%NISP) (n=14285) ..... 198
Figure 41: General subsistence burnt bone by bone type (NISP) (n=2798) ...... 199
Figure 42: General subsistence small seal element distribution (%MAU) (MAU=47; MNE=2675) ...... 200
Figure 43: Meat utility index: rs=0.39, p=.15 (MAU=69) ...... 200
Figure 44: Iñupiat preference ranking: rs=0.36, p=.27 (MAU=34) ...... 200
Figure 45: General subsistence caribou element distribution (MAU) (MAU=3,MNE=110) ..... 201
Figure 46: General subsistence caribou food utility index: rs=0.41, p=.10 (MAU=3; MNE=110) ...... 202
Figure 47: General subsistence caribou meat drying index: rs=0.61; p<.01 (MAU=3; MNE=110) ...... 202
Figure 48: Genereal subsistence indeterminate mammal bone (NISP=9312) ...... 204
Figure 49: General subsistence caribou unsaturated marrow index: rs=0.70, p<.01 ...... 204
Figure 50: General subsistence dog/wolf (NISP=181, MAU=2, MNE=138) and fox (NISP=137, MAU=2, MNE=102) element distribution (%MAU) ...... 205
Figure 51: General subsistence bird skeletal distributions (MAU=4, MNE=79) ...... 206
Figure 52: General subsistence cranial and post-cranial distribution for fish (NISP=267) ...... 206
Figure 53: General subsistence meat weight (%) ...... 209
xx
Figure 54: General subsistence bird taxonomic distributions (n=74) ...... 210
Figure 55: General subsistence common vertebrate taxonomic groupings (NISP=6232, MNI=129) ...... 214
Figure 56: General subsistenc seal epiphyseal fusion by element (%MNE) (MNE=108) ...... 216
Figure 57: General subsistence seal epiphyseal fusion combined femur, humerus, and radius (MNE) (MNE=108) ...... 216
xxi 1
Chapter 1 Introduction
Analogical reasoning that uses the documentary record to interpret the material record is one of the key ways archaeologists understand the past. Despite this, there have been only a limited number of studies that test the ethnographic record, or more broadly the documentary record, against the archaeological record. This research starts to fill this gap by analyzing the similarities and differences between the activities recorded in the ethnographic record and those that are visible in the archaeofaunal record. Assessing similarities and differences will strengthen analogy and allow archaeologists to infer further similarities between the documentary and the archaeofaunal records. In order to test the strength of the analogical inference in interpreting activities within the house, this research focuses on the patterning of faunal remains within a dwelling feature in northwest Alaska during the Early Thule period. Although a few studies have used faunal materials to interpret spatial patterning within a feature, architecture and artefacts are most commonly used. Reconstructions of archaeofaunal spatial patterning will likely lead to different types of interpretations than either artefacts or architecture, which will lead to a better understanding of the past. In addition, the faunal material is analyzed to construct the use of subsistence resources during the occupation of an Early Thule semi-subterranean dwelling at Cape Espenberg, Alaska.
This introduction outlines some of the archaeological discussions around analogical interpretation, house and household studies, and spatial patterning studies. These three diverse theoretical discussions will situate this research within the broader archaeological framework before describing the specific context of the Arctic region in Chapters 2 and 3.
1.1 The Use of Analogy in Archaeological Interpretation
Analogical reasoning allows archaeologists to move beyond a list of material remains to interpret the behaviours, practices, and often beliefs of a past group of people. Analogy is used to interpret all scales and time periods of human society, from the earliest hunter-gatherers through the rise and collapse of state-level societies. Implicitly or explicitly, analogies are the basis of archaeological inference (Ascher 1961; Binford 1967; 1968; 1972; Chang 1967; Clark 1968; David and Kramer 2001a; Hegmon 2003; Hodder 1982; Hodder and Hutson 2003; Longacre
2
1978; 1991; Lyman and O’Brien 2001; Roux 2007; Shelley 1999; Stahl 1993; Watson 1986; Wobst 1978; Wylie 1985; 2002a; Yellen 1967). Broadly, an analogy is defined as a form of reasoning that uses established similarities between a source (e.g., ethnographic records, present day observations) and a subject (archaeological material) to propose more extensive similarities (David and Kramer 2001b; Gould and Watson 1982; Roux 2007; Shelley 1999; Wylie 1982; 1985). Analogies do not assume identical forms of behaviour, structures, or material, but identify key similarities and explore significant differences in order to establish a link between source and subject to propose further similarities (Ravn 2011; Wylie 1985). Therefore, because analogies are based on limited similarity but propose further similarities, they are always “‘ampliative’; they inevitably claim the existence of more extensive similarities in their conclusions than has been or could be established in their premises” (Wylie 1985:80).
There are two main overlapping approaches when using analogical reasoning to construct past human behaviour from material culture: the use of analogies to interpret the archaeological record, and the use of the archaeological record to test analogical similarities. The first approach works from the archaeological record and finds analogical correlates. This type of analogical reasoning is the most ubiquitous form in archaeological interpretation; it occurs whenever an archaeologist infers the functions, actions, or symbolism associated with past material based on ethnographic or ethnohistoric observations. An example of this is Morrison’s (1983b) interpretation of tool function in the Coronation Gulf area in order to identify hunting practices. One comparison Morrison (1983b:69) makes is between the use of the harpoon foreshafts recovered at the Clachan site and the foreshafts of the historic record; the morphological similarities between the two suggest that the archaeological specimen was used for similar purposes as the historical one. In this case, the archaeological foreshaft was “loose”. Most “loose” foreshafts in the historic period were used with a throwing harpoon in open-water hunting (Mathiassen 1927a: 27-28); therefore, Morrison (1983) concludes that much of the small seal hunting was done in open water. This simplified example shows how archaeologists use analogical reasoning to interpret archaeological materials based on ethnographic observations.
The second use of analogical reasoning establishes analogical correlates and tests them against the archaeological assemblages (Binford 1967). An example of this is when documentary correlates are tested against an archaeological dataset, such as the investigation by Mason (2012) into Iñupiat warfare based on the documentary records of Ernest S. Burch. Here, the expectations
3 for warfare are created from correlates described in the ethnohistoric record; these expectations are then tested against archaeological material to identify past societies that may have practiced warfare. These two ways of approaching interpretation of past human remains are not mutually exclusive (Shelley 1999; Stahl 1993; Wylie 1985), and both types of analogical reasoning are used together to reconstruct past activities, behaviours, and practices (e.g. Binford 1967; Friesen 2013a).
Analogical reasoning has undergone substantial discussion and critique. The most important discussions of analogy as a conscious and unconscious mode of interpretation started in the 1950s and continued through the 1980s; subsequently, there was little discussion of it during the 1990s and 2000s with the popularisation of postmodern theory (Ascher 1961; Binford 1967; 1972; 1978a; 1981b; Gould and Watson 1982; Hodder 1982; Hodder and Hutson 2003; Wobst 1978; Wylie 1982; 1985; 1988; 2002a). Most of the debate around analogy focuses on the use of the modern ethnographic observations to interpret early prehistoric remains, with limited analysis of the use of ethnohistoric and ethnographic sources to interpret direct cultural ancestors through the direct-historical method. The limited theoretical discussion about the direct-historical method is likely due to the fact that only a few areas in the world have both ethnographic and archaeological data from related cultural groups, which limits the use of the direct-historical approach globally. Where the direct-historical method is discussed, the critiques often focus on the changes in the culture brought about by Western contact. Beyond the debate on the direct- historical method, the broader use of analogical reasoning was challenged as non-empirical and invalid as a source of interpretation (Ascher 1961; Gould 1978; 1980; Gould and Watson 1982; Wobst 1978). However, further examination of archaeological interpretation has shown that analogical reasoning is a part of even the most empirically based interpretations and can be a strong interpretive tool (Wylie 1985; 1988; 2002b; 2002c).
One of the outcomes of this discussion has been the development of a critical method for using analogy to interpret the archaeological record. It is recognized that the analogies used in archaeological interpretation should be tested and strengthened. One way to test these analogies is by analyzing the similarities and differences in source and subject material in socially, biologically, and culturally constrained environments. Analysis of archaeological material using the direct-historical approach can provide such a context. It is important to establish which activities can or cannot be inferred in this constrained and culturally continuous environment in
4 order to move to other, less culturally continuous analogical inferences. Understanding the ways and circumstances in which the archaeological record relates to documentary accounts will provide a more accurate and precise interpretation of the prehistoric past. In addition, research to understand the similarities and differences among culturally continuous people may inform the documentary record itself, shaping these records rather than just imposing them on the archaeological material.
1.1.1 Limitations of Analogy
Discussions about the use of analogy for archaeological interpretation highlight some of the underlying problems and unconscious assumptions present in the use of analogical reasoning. One of the more salient critiques of analogical reasoning in general is that analogies are inductive ampliative arguments that can only claim probability of interpretation rather than validating or invalidating an interpretation (Wylie 1982). Because analogical reasoning will never supply either absolute confirmation or denial of an interpretation, the analogical relations are limited; known similarities between the source and subject do not indicate with certainty similarity in all other areas, but must be tested (David and Kramer 2001b; Gould and Watson 1982; Wylie 1982).
Another limitation when using the historical or ethnographic record to interpret archaeological material is the difference in temporal and spatial scales between the two data sources. Ethnographic analogy, ethnoarchaeology, and even historical records document a small time depth; archaeological evidence is often the result of repeated patterns over long periods of time. Despite the fact that past cultural processes and change have played a key role in the development of the present, ethnographers sometimes look at a single moment in time, without allowing for the cumulative effects of change (Peterson 1971: 242). Archaeology is interested in temporal aspects of the data, using history to develop site and artefact interpretations. These differences in temporal scales can impact the reconstruction of the past using present day observations. Not only are documentary sources temporally constrained, they are often spatially constrained as well. Historic, ethnographic, and ethnohistoric observations are most often limited to a few sites. Variation among groups and over landscapes may be absent in the ethnographic record, but present in the archaeological record. Both temporal and spatial differences indicate that analogical reasoning must be critically and carefully applied.
5
Documentary sources must be placed in their historical and cultural context in order to assess their limitations (Ascher 1961; Binford 1967; Dunnell 1978; 1992; Gould and Watson 1982; Hill 1998; Salmon 1982; Wobst 1978; Wylie 1982; 1985). Both informants and researchers bring different biases to the documentary record. Informants and researchers have different levels of knowledge, with different degrees of comprehension about their own knowledge. Although informants may tell ethnographers or ethnohistorians about practices and beliefs that the research is unable to observe due to spatial, temporal, or privacy reasons, these retellings are filtered through a myriad of factors including the informant’s and researcher’s identity, gender, age, or goals. Biases, oversights, goals, emphases, and purposes are all potential limitations on the final document produced by a researcher, and thus its reliability when used to interpret the past.
Some critiques of using the documentary record for understanding past human societies stem from the overarching assumption of uniformitarianism in using these records to interpret past behaviours (Gould 1978: 250). Behaviours not observed currently through the ethnographic records, historical documents, or actualistic research will be impossible to interpret in the archaeological record (Kramer 1979: 2). However, many documentary records were recorded at a time of great change or were specifically commissioned to observe change; thus, documentary records are often records of the process of change and of particular historical trajectories. As well, not all present behaviours may have analogs in the past. If adhered to strictly, the uniformitarian assumptions leave little room for unique historical events. These critiques assume the application of strict uniformitarian principles; with careful and critical application of some uniformitarian principles, the past can be constructed using some present day sources.
Another more subtle critique is the fact that when archaeologists use the documentary record, they often use anthropological terms such as tribe, band, or household and attempt to define them archaeologically (Freeman 1968: 262). This may both constrain the interpretations of the archaeological record to these anthropologically defined terms and limit the variability seen in the past.
A final critique about the use of the ethnographic record is the fact that many groups have changed drastically in the recent past. Many ethnographically described groups now occupy “marginal” regions, while in the past they would have also occupied the most productive areas (Guenther 2007; Clark 1968: 280). Clark (1968: 280) also argues that most hunter-gatherer
6 groups existing today have long been in contact with more complex societies and technologies and are therefore poor models for comparing with the prehistoric populations. The use of the direct-historical method for interpreting recent prehistory may moderate this critique if the documentary sources are recorded in the same region as the archaeological materials; however, these direct-historical methods are subject to a different set of critiques, described in Chapter 3.
Despite these critiques, analogical reasoning has been shown to be a valid form of interpretation if used carefully and critically (Wylie 1985, 2002a). Strong analogies are those that show extensive similarities between the source and subject, but also explore the differences (Wylie 1982, 1988). It is from these critiques and defenses of analogical reasoning that the present study was constructed.
Inuit documentary sources (i.e. ethnohistories, oral histories, and ethnographies) are extensive and intensive. Many of them give rich detail about many aspects of life in various Arctic regions (e.g. Balikci 1970; Birket-Smith 1929; Boas 1888; Burch 1998a; 2006; Damas 1969; 1972; Kroeber 1900; Mathiassen 1927a; 1927b; Murdoch 1892; Nelson 1899; Petitot 1887; Rasmussen 1931; 1932; Ray 1975a; 1983; Stefansson 1914a; Turner 1894). Similarities in house structure, material culture, and subsistence strategies between the historic Inuit and their Thule ancestors allow archaeologists to use the documentary record to interpret the behaviours and activities of Thule from ca. 1000 CE. What is often lacking is a test of the sources for interpreting ancestral archaeological material of the same region. When studies do critically evaluate the use of the ethnographic records, they do not always use the direct-historical method. Rather, they often use sources from a different region. For example, the research on early Thule in the Canadian Arctic uses the ethnographic records of the northern Alaskan Iñupiat to reconstruct social interactions, social strategies, and worldviews (e.g. Friesen 2012a; 2012b; Grier and Savelle 1994; Patton and Savelle 2006; Savelle and Wenzel 2003; Whitridge 1999a; 2002b; 2004).
The Arctic is an excellent location to test the accuracy and precision of the documentary record in interpreting archaeological materials due to both the extensive documentation of early historic Inuit groups across the Arctic and the excellent preservation of archaeological material at most sites. Additionally, the documentary record from the Arctic has been used extensively to interpret archaeological remains cross-culturally and as a basis for developing hunter-gatherer theory and global analogies. Thus, understanding which activities and practices are most similar to the
7 ethnographic record in the Arctic will help other archaeologists use this documentary record to interpret more spatially and temporally distant archaeological materials.
This study tests the documentary record at the spatial scale of the house, where human decisions about processing, butchering, consumption, and discard take place. Other studies have systematically used the documentary record at a broader scale. A few examples of these include the studies by Friesen on the interregional interaction of the Inuvialuit of the Mackenzie Delta (2012a) and the Inuit of the eastern Arctic (2012b) by comparison to Alaskan documentaries; the interpretations of Savelle and Wenzel (2003) of the social structure of the eastern Canadian Arctic Inuit using the documentary sources of northern Alaska; and an examination by Grier and Savelle (1994) about intra-site spatial patterning based on ethnographies from the Alaskan whaling communities. Although the documentary record has been used to interpret archaeological remains at the household level, only a few studies have tested the validity of using the documentary record against archaeological remains. Dawson’s work on changes and variability in the built environment (1995; 1997; 2001; 2006) and the behaviour of light within a dwelling (Dawson et al. 2007), and Harritt’s (2010) examination of the variations in house form at Wales are the few exceptions. Most of these studies examined houses in the later Thule period; testing the ethnographic record for use in the early Thule period has seen even less work (but see Hoffecker et al. 2012). Examining the similarities and differences in the faunal record compared with the documentary sources will allow archaeologists to strengthen their analogical reasoning. Once strong analogies have been established, further interpretation of behaviours, beliefs, and worldview will be possible in order to reconstruct the everyday lives of people in the Arctic. .
1.2 Household Archaeology or Archaeology of a House?
Houses have long been a focus of archaeology, specifically in spatial studies, with the sub- discipline of household archaeology emerging in the 1970s and 1980s from a symposium at the 1981 Annual Meeting of the Society for American Archaeology and subsequent publication (Wilk and Rathje 1982). At this time there was a gap between the overarching theories of culture change and the practical material objects of the archaeological record (Wilk and Rathje 1982). Wilk and Rathje (1982:618) argued that because households were often where social groups interacted directly with economic processes, it was at this scale that culture change should be studied. Although firmly established with a processual framework with much of the original
8 work focused on understanding culture change, economic, and ecological processes (Ashmore and Wilk 1988; Wilk and Rathje 1982), household archaeology has also been used in postprocessual interpretations of the past to contextualize agency, practice, and gender (Ashmore 2002; Brumfiel 1992; Hendon 1996; Robin 2003). Both processual and postprocessual frameworks continue to be applied to household reconstruction with recent studies looking at households as a historical construct situated within larger landscapes (Pluckhahn 2010).
As a fundamental social unit, analysis of the household can provide access to the everyday life of individuals (Allison 1999a; Ashmore and Wilk 1988:1; Franklin 2004:xiii; Hirth 1993:21; Pluckhahn 2010; Robin 2003:308). This perspective offers a contrast to the large-scale systems and processes that are often used to explain social and cultural change (Robin 2003; Wilson 2008). Additionally, households are good comparative units across space and through time (Blanton 1994; Hirth 1993:21; Pluckhahn 2010; Wilk and Netting 1984:1). Early research within the processual framework focused on patterns of domestic activities and behaviours that could be generalized to a certain site (i.e. season), region, or culture (Sabloff and Ashmore 2001). Households were often seen as the basic building blocks of larger social formations, as points of articulation between societies and economic and ecological processes, and as windows to evolutionary change (Pluckhahn 2010; Wilk and Rathje 1982). Since the 1980s, household studies have adapted to become more nuanced and interpretive in their approaches. Households are moving away from being analyzed as the basic building block of a society to being recognized as nodes of connection to other households, and as spatial dwellings for individuals with different affiliations through age, kinship, gender, and class; this recognition requires a more nuanced, historicized, and politicized understanding of archaeological households (Hendon 1996). Themes of household archaeology throughout different regions have looked at production and consumption, status differentiation, agency and power, gender, ritual and symbolism, and identity and ethnicity (Pluckhahn 2010; Robin 2003).
The term “household” has a number of definitions. Ethnographically described households consist of an activity group that engages in one or more of the following practices: production, consumption or distribution, reproduction, co-residence, and transmission (Ashmore and Wilk 1988; Wilk and Netting 1984). This functionalist definition does not preclude the social, ritual, or ideological necessities of a household, but does focus the definition on actions and interactions of people through household membership and cooperation in a set of practices (Coupland et al.
9
2009; Souvatzi 2008). Households are often defined as a fundamental social formation that is identifiable archaeologically as houses (Hirth 1993:21). However, households are more than just the physical house building; they are the also the remains of routine activities and habitual practices (Pluckhahn 2010:332). Archaeological households have been defined by Nash (2009) as a co-residential group that used the occupation surface, features, and the artefact assemblage of a dwelling where a dwelling is defined as one or more structures and both indoor and outdoor spaces. Despite this clear definition, there is often a tension between the broad activity-based definition of ethnographically defined households and the practical archaeological conceptualization of a household that focuses on the house structure and associated artefacts and features. This tension is difficult to resolve spatially and temporally. Spatially, households used many different spaces outside of the dwelling structure; temporally, many household activities occur at different places during the day, season, or year, and house structures can be used over multiple years or decades (Smith 1992). Thus, although an anthropological definition of household activities may more closely mimic the reality of how people interacted within their household unit, archaeological investigations are often more limited. Archaeologically, analyses focus on a section of interrelated house activities, often restricted to the dwellings and closely associated outdoor spaces. Although households, which are social units, are not excavated, their occupied dwellings and domestic artefacts are (Allison 1999a:2; Wilk and Rathje 1982:618).
The term household may be a misnomer for looking at a single house in a small village in the Arctic. Most ethnography in the Arctic does not employ the term “household”, but based on the above definition of it being a unit of production, consumption, transmission, and reproduction, a reading of Arctic ethnographies and ethnohistories indicates that the compound family described in Burch’s work (1998a; 2005; 2006) and the small village described in Ray’s work (1964) fall into the functional definition of a household. These larger family structures acted as an economic and social unit (Ray 1964). In certain seasons the household or compound family produces and distributes food through and to its members. There is common space in the form of a qargi in most villages in Alaska, and in all villages there are storage areas (Ray 1964). Even when these compound families disperse for (typically) the warm season, they are still part of the larger compound family/household (Ray 1964). In the Alaskan Iñupiat documentary records, people of different compound families did not live together in the same houses (Burch 2006). Thus, it can
10 be assumed that a domestic dwelling housed people from the same household, but likely does not represent everyone within the compound family or ethnographically defined household.
Due to the fact that an Arctic dwelling in the Kotzebue Sound area such as the one that forms the focus of this research is likely to represent only a part of the larger household or compound family, it is important to note that ethnographically people of the compound family used domestic dwellings differentially. Domestic dwellings in historic period of coastal Alaska were mainly used by female and younger members of the household, with men and older boys spending most of their time in the qargi or ceremonial house in most villages (Burch 2006; Ray 1964). Additionally, the individual occupants changed over the lifetime of a house. Thus, the necessity of looking at overall behaviours and patterns becomes a focus, with individual actions recognized and acknowledged as constituting these behaviours. Although it is beyond the scope of this dissertation, archaeology in the Arctic needs to expand to include the exterior areas and multiple contemporary houses at one village if it is to understand if there was a household structure in the past and if it functioned as detailed in the documentary record (but see Dekin and Kilmarx 1990; Hall 1984; Newell 1990; Polglase 1990; Sheehan 1990). Although household archaeology may not be the best term for the current research on a single house and midden area at Cape Espenberg, the value of looking at smaller units still exists. Houses, like households, are important for understanding the people who lived through broader societal changes and archaeological patterns.
For the present, archaeology of semi-subterranean dwellings gives a clear, definable border that allows a discrete unit of analysis to be compared and contrasted within a site and across the region. Further research into understanding the activities associated with all household members both within and among dwellings and in other types of structures needs to be conducted in the Arctic. This type of research will not be neatly bounded. The current research of a house is one part of the research needed to understand the village compound family and the integrated lifeways that were centered on this winter village. For this research, the dwelling and spaces people inhabited are places where attitudes and traditions were enacted to create and change the world in which people lived (Bourdieu 1977). This dwelling-based investigation will contribute to an understanding of domestic behaviour of the early Thule period in Alaska. The spatial organization of domestic activities has often been a focus of household, or dwelling, analysis, frequently with an emphasis on defining patterns of behaviour that can be generalized to whole
11 regions or time periods (e.g. King 2008; Robin 2002). Despite much of this research, detailed spatial analyses of faunal remains are lacking in published literature.
1.3 Brief History of Spatial Studies
Spatial analysis has been an important part of archaeology since the eighteenth century (Kroll and Price 1991; Wheatley and Gillings 2002). The most common type of spatial analysis originally was through the use of maps, plans, and visual inspection; early spatial studies did this analysis intuitively. In the late 1970s and 1980s, these subjective measures were reworked into more rigorous and robust sets of methods corresponding to the rise of New Archaeology. Various methods for testing and measuring archaeological patterns using technology, statistics, and computations were proposed and evaluated (Ashmore 2002; Clarke 1977; Flannery 1976; Hodder and Orton 1976; Wheatley and Gillings 2002). However, both the intuitive visual approaches and the spatial-scientific ones viewed space as a “neutral, abstract dimension in which human action took place” (Wheatley and Gillings 2002:8); space was the map against which human behaviour and cultural activity occurred. As a response to these characterizations, landscape studies emerged to highlight the socially constructed nature of settlements and households (Ashmore 2002; Bevan and Conolly 2006; Bradley and Fraser 2010; Carballo and Pluckhahn 2007; Cowgill et al. 1984; Flannery 1976; King 2008; Kroll and Price 1991; Nadal 2003; Smith 2010; Vaquero and Pasto 2001; Weiss et al. 2008).
Both archaeologists and anthropologists have long stated that space is not a neutral backdrop to human action, but is part of the medium that is shaped by and shapes social action. Ethnographers in the 1920s understood space, in the form of structures, as part of understanding socio-cultural surroundings. The built environment was recognized as a key component of society and culture (Lawrence and Low 1990). Space as a created medium was part of Bourdieu’s (1977) writings on the interaction of agency and structure, influenced Gidden’s (1984) writings on structuration, became an important component of Ingold’s (2000) notion of dwelling, and served as part of Hodder’s (1987) contextual archaeology. Archaeologists who study the built environment recognize space as a constitutive and constituting agent of society in order to understand behaviours, activities, social organization, and agency. Most studies use artefacts or architecture to understand space; zooarchaeological assemblages have been used only sporadically for this purpose. Faunal assemblages, often overlooked in discussions of space and
12 place, may offer analyses that can be compared and contrasted to the more typical architectural and artefactual analyses.
1.4 Spatial Studies of Archaeofaunal Assemblages
Spatial archaeology is able to look at all scales of analyses from individual buildings or burials through to settlements, landscapes, and regions (Ashmore 2002; Hodder and Orton 1976; Rossignol and Wandsnider 1992); however, most zooarchaeological analyses are at the settlement, regional, and landscape level, with a minority of spatial studies conducted at the intra-site level and very few at the intra-feature level (Ashmore 2002; Robertson et al. 2006; Rossignol and Wandsnider 1992; Wheatley and Gillings 2002; but see Darwent and Foin 2010; Friesen and Betts 2006; Hietala 1984; Kroll and Isaac 1984; Kroll and Price 1991; Rosell et al. 2012; Stahl and Zeidler 1990). Despite calls to use the complete archaeological assemblage to provide information on houses and households (Miller 1987; Shanks and Tilley 1987), the use of faunal assemblages remains limited.
The following section analyzes how studies have used faunal remains to interpret space at a small scale. Since there are a fairly limited number of intra-feature zooarchaeological studies, inter-feature faunal analyses are also included to understand how faunal material has been used in spatial studies. These analyses have been classified into four categories: studies dealing with site formation processes; studies of the function of different areas; studies that investigate the interaction of spatially defined areas; and studies that probe the cultural meaning of spatial patterning. Although the studies have been divided into four categories, they are not mutually exclusive, and most of the papers fall into more than one of these categories.
The first category includes studies of site formation processes using the zooarchaeological record (e.g. Asmussen 2009; Bunn and Kroll 1986; Lyman 1984; Otárola-Castillo 2010; Weissbrod et al. 2005). The sample has been limited to those studies that had a spatial component to their analysis. The goal of some of the studies was solely to understand site formation processes, such as actualistic studies by Kent (1981), Marean and Bertino (1994), and Wilson (1996) that look at how carnivores altered the spatial patterning of bones at sites. Other important ethnoarchaeological studies looked at understanding a wide variety of site formation processes such as discard patterning, trampling, and cleaning by observing present day peoples (Binford 1977; 1978a; Gould 1978; 1980; Kent 1984; 1987; 1990; Kramer 1979; 1982; Longacre and
13
Skibo 1994; O’Connell 1995; O’Connell et al. 1988; 1990; 1992; Watson 1979; Yellen 1977; 1991). Studies of faunal refitting have also been undertaken in order to determine discard patterning, sharing, and feature contemporaneity (O’Brien and Storlie 2011; Rosell et al. 2012). Although some studies focused just on site formation processes, many studies included site formation processes as part of their analysis of other social, cultural, or ideological aspects of lifeways (e.g. Gidney 2000; Hachem 2000; Hockett 1998; Leonova and Min’kov 1988; Logan and Hill 2000; Middleton et al. 2002; Olson et al. 2009; Stahl and Zeidler 1990; Tuma 2006; Zeder and Arter 1996). For example, Zeder and Arter (1996) first analyzed patterns of bone density and bone fragmentation in order to understand site formation processes before they examined the differences in bone spatial patterning. By understanding how the faunal assemblage was formed and what other post-depositional processes affected the assemblage prior to burial, they concluded that some of the differences in bone patterning resulted from differences in intensity of occupation, in social and economic status, and through different interactions between households.
In zooarchaeology, a majority of intra-site spatial patterning studies relate to questions of area function (Arroyo 2009; Bartram et al. 1991; Darwent and Foin 2010; Enloe and David 1992; Enloe et al. 1994; Friesen and Betts 2006; Gidney 2000; Howse 2008; Kroll and Isaac 1984; Kuckelman 2010; Leonova and Min’kov 1988; Logan and Hill 2000; Olson et al. 2009; Stahl and Zeidler 1990; Stalibrass 2000; Tuma 2006; Waguespack 2002; Wilson 1996; Zeder and Arter 1996). These types of analyses use detailed faunal data such as bone density, fragmentation patterns, refitting, and burning patterns, along with actualistic studies, the ethnographic record, and ethnoarchaeological record to understand the types of past behaviours based on differential faunal patterns across space. For example, Arroyo (2009) mapped the patterning of bone fragments in the Middle and Late Magdalenian and Azilian layers in El Mirón Cave, Spain. Arroyo (2009) identified spatially distinct events in different areas of the cave, including accumulated butchery waste near the back of the cave and a number of isolated consumption episodes throughout the cave.
A third category of intra-site faunal patterning studies the differences and similarities of spatial areas in order to understand social organization and relations. These studies used detailed faunal analyses such as taxonomic frequencies (Hachem 2000; Hockett 1998; Scott 1996; Teeter 2004), age/sex and survivorship profiles (Falconer 1995), bone element distributions (Junker et al. 1994;
14
Stokes 2000), and element/limb refitting (Waguespack 2002) to interpret the social relations and social organization (Teeter 2004; Waguespack 2002; Zeder and Arter 1996). Often, status differences and ethnic boundaries were investigated through these lines of faunal evidence (Gidney 2000; Hockett 1998; Junker et al. 1994; Schmitt and Zeier 2010; Scott 1996; Stokes 2000; Teeter 2004; Wilson 1996). This category of analysis relies heavily on the ethnographic, ethnohistoric, and ethnoarchaeological records to interpret behaviours from faunal remains. Often, zooarchaeological analyses are supplemented with other lines of archaeological evidence (e.g. Gidney 2000; Junker et al. 1994; Scott 1996). For example, Scott (1996) used faunal spatial patterning, along with historical records, to identify residents of an 18th-century colonial fur- trading settlement in northern Michigan. Based on historical records and differences in economic, ethnic, and religious groups identified by their food remains, Scott (1996) was able to identify the houses and areas in which they lived.
The final category of intra-site studies links spatial patterning to beliefs, ideology, and cognition (Falconer 1995; Kovacik 1998; Thomas 2005; Wilson 1996). Although studies in this category used the zooarchaeological record, they also relied on other lines of evidence such as strong links to ethnographically and ethnohistorically recorded groups (Kovacik 1998), contextualizing local patterns in relation to broader trends (Falconer 1995), and detailed, extensive understandings of the archaeological culture (Thomas 2005; Wilson 1996). For example, Kovacik (1998) used the direct-historical method to understand the spatial patterning of carnivores and birds of prey from a pueblo in Chaco Canyon, New Mexico. Here, Kovacik (1998) used ethnographic data from the American Southwest on the use of animal species and on the collective memory to inform his interpretations on memory at Chaco Canyon.
In all four categories, the most convincing spatial research was undertaken with strong direct historic or ethnoarchaeological connections (Clark 1998; Gidney 2000; Junker et al. 1994; Kent 1981; Kovacik 1998; Marean and Bertino 1994; Olson et al. 2009; Scott 1996; Schmitt and Zeier 1993; Stahl and Zeidler 1990; Tuma 2006; Wilson 1996). Although intra-site/inter-feature spatial analysis of faunal patterns without direct ties to a written or observed record is undertaken, these studies tend to be limited in the certainty of their findings.
Despite work being done on intra-site spatial patterning of the faunal record, very few of these analyses focus on intra-feature patterning of bones (Atalay and Hastorf 2006; Clark 1998;
15
Darwent and Foin 2010; Enloe et al. 1994; Friesen and Betts 2006; Howse 2008; Stahl and Zeidler 1990; Wilson 1996). Only a few of these intra-feature studies have used the documentary record to understand the behaviours and activities of a feature (Clark 1998; Darwent and Foin 2010; Friesen and Betts 2006; Stahl and Zeidler 1990; Wilson 1996). Although rare, there have been a few studies to look at the patterning of faunal remains within dwellings both within the Arctic (Darwent and Foin 2010; Friesen and Betts 2006; Howse 2008) and beyond (Atalay and Hastorf 2006; Enloe et al. 1994; Kroll and Price 1991; Stahl and Zeidler 1990; Wilson 1996). These studies investigated the spatial activities within households based on exploratory analyses of zooarchaeological data and comparisons to ethnoarchaeological studies. In none of these studies is a framework for understanding the differences in spatial patterning explicitly outlined. This research aims to fill in the gap both on the scalar level and methodological level.
1.5 Spatial Patterning in Houses
Spatial patterning in at the intra-feature level has often been studied based on the architecture (e.g. Blanton 1994; Kent 1990; Laurence 1994). An investigation of architectural remains may describe the behaviour of the people who built the house and the cultural patterning of space, but these studies do not necessarily lead to an understanding of the behaviours of those who inhabited the space (Allison 1999a). As argued by Rapoport (1990:15–20), the people who designed and built the houses are not the same and do not react to the environment in the same way as users. Although the people who built the house and set out the spatial divisions may have also inhabited the space, the majority of people living in and using the space were likely their relatives or descendants; alternatively, the people who occupied the house may be unrelated to the people who originally built the house (Allison 1999a; Blanton 1994; Rapoport 1990). This is often the case in the Arctic documentary record where houses may have been built mainly by men of the family, inhabited mostly by the women and children, and then subsequently reused over decades by descendants of the original builders (Burch 2006:64; Lee and Reinhardt 2003:123). Understanding the artefact and faunal assemblages should allow a more detailed construction of the way of life of domestic dwelling inhabitants than a study of architecture alone.
Some of the first investigations that used artefact assemblages to understand household spatial patterning were done in the 1970s in the American southwest (Hill 1968; LaMotta and Schiffer
16
1999; Longacre 1970). Subsequently, much research globally has been conducted on spatial patterning of artefacts within houses, with most of it focusing on the distribution of household activities (e.g. Allison 1999b; Cutting 2006; Daviau 1993; Friesen and Betts 2006). In most cases, an understanding of household activities through the artefact assemblage is mediated by an analysis of depositional and abandonment processes (e.g. Allison 1999a; LaMotta and Schiffer 1999; Lightfoot 1993; Savelle and McCartney 1994; Schiffer 1976; 1985; Stein 1992). It is from these studies on formation processes that the life history methodology for interpreting the assemblage of a household has been developed (LaMotta and Schiffer 1999; sensu Binford 1968:21–22; Schiffer 1987:13–15).
Behavioural archaeologists such as Schiffer have extensively studied formation processes of houses through ethnographic, ethnoarchaeological, and archaeological research (LaMotta and Schiffer 1999; Schiffer 1976; 1985; 2010). Schiffer’s early work (1972; 1976; 1985) attempted to outline ways in which formation processes transform the archaeological assemblages in order to clarify question that could be asked for a particular archaeological assemblage. At the time of Schiffer’s (1976; 1977; 1985; 1987) early work, he argued that archaeologists developing New Archaeology, particularly in his region of the American southwest, were using artefacts to link activities and behaviours to spaces through erroneous implicit assumptions that artefacts recovered archaeologically were left in situ at the time of abandonment and that they represented a complete inventory of household objects. To start to understand the processes that affected assemblages, Schiffer identified formation processes that affected the assemblage post- abandonment. In an attempt to standardize archaeological research to look at formation processes, Schiffer (1972; 1976; 1977; 1985; 2010) originally identified a number of processes and refuse types at sites. The refuse types included primary refuse, abandonment refuse, de facto refuse, ritual deposits, post-abandonment uses, and secondary refuse (Schiffer 1987). These refuse patterns were thought to be a result of a number of different formation processes that could be classified into two types: cultural transforms (c-transforms) and non-cultural transforms (n-transforms) (Schiffer 1987). Each type of transform had multiple actualistic and ethnoarchaeological correlates; originally, behavioural and processual archaeologists believed that by identifying and outlining these processes, they would eventually be able to identify and account for all of the different processes that affected artefact assemblages. Over time, the understanding of these processes has been critiqued and the application has been altered. The
17 current work on formation processes is often classified as taphonomic studies. Most taphonomic issues are studied using zooarchaeological material and stratigraphy; more recent taphonomic work has used ceramic (e.g. Blanco-González and Chapman 2014) and other material culture assemblages (e.g. Bailey 2007; Chapman and Gaydarska 2007; Lucas 2012). Understanding the creation of sites is a necessary process of any archaeological interpretation.
The need for such studies is apparent when archaeologists consider the two themes that have emerged from the cross-cultural literature on formation processes (Schiffer 1976; 1987; 2010). First, there is not necessarily a direct relationship between artefacts and activities in a space; this applies to the fact that all objects used in a space are not necessarily deposited where they were used and the fact that not all objects are found where they were used (i.e. discard, ritual deposits, secondary refuse) (LaMotta and Schiffer 1999). Second, the archaeological assemblage may be a palimpsest of deposits relating to different phases of the house’s life history (LaMotta and Schiffer 1999). Schiffer and colleagues focus mostly on artefact assemblages, arguing that objects are rarely deposited in their primary activity location, especially in heavily used and maintained areas. Objects that are found on the floor are more likely to be a product of other contemporary, abandonment, and post-abandonment processes or deposition (LaMotta and Schiffer 1999).
A simplified version of Schiffer’s formation process model has three stages: use (habitation), abandonment, and post-abandonment (LaMotta and Schiffer 1999). A modified version of these three stages has been used to develop expectations for this research in order to align them with regionally specific documentary evidence. When comparing the expectations derived from the documentary record to the archaeological record, it is important to separate out the difference between the activities recorded by the observers (i.e. butchery, consumption, preparation) from the contemporary, but often invisible, activities often left out or poorly recorded by observers (i.e. cleaning, disposal, incidental burning). Western observers record very little detail about abandonment and post-abandonment activities, and so most expectations relating to these categories are derived from actualistic and cross-cultural ethnoarchaeological studies. All of these expectations are based on the bones that were typically brought into winter residential sites. As well, the effects of activities do not stop with the archaeological recovery, but continue to affect the assemblage through excavation, identification, and analysis. These processes are described in Chapter 5: Expectations and Chapter 6: Methodology. This analysis is not solely an
18 application of Schifferian assemblage formation processes to archaeozoological remains, but a more holistic view of the history of an assemblage.
1.6 Research Materials
In order to strengthen the use of analogy in interpreting spatial patterning, it is important to have good source and subject materials. The present study uses faunal material from an early Thule Inuit semi-subterranean house in the Kotzebue Sound area. This data set is ideal, because the documentary records from this region are very detailed, and the faunal assemblage is a high- resolution subject sample. The Arctic is an ideal location to test the use of analogy as the records of the early historic period are both extensive and intensive; as well, researchers are continuing detailed studies of the oral histories and recent past in the region. These excellent documentary records are used to create expectations from the source side of the analogy (Chapter 5). The documentary sources are reviewed in Chapter 3. Subject material was excavated from a house that was extremely well-preserved both taphonomically and structurally. The excavation was high-resolution, with faunal material collected in 50x50 centimetre squares and screened through 3mm mesh; the excavation and house are described in more detail in Chapters 4 and 6.
1.7 Research Goals
The main goal of this research is to test expectations based on the documentary record against the archaeological material in order to identify which parts of the documentary record can be used to establish meaningful interpretations of past lifeways. This will strengthen and clarify the analogical relationship between the source and subject material in order to calibrate the use of the documentary record in inferring further similarities between the archaeological record (subject) and the documentary record (source). To do this, expectations are generated from the documentary record and tested against the archaeological faunal assemblage from a semi- subterranean Thule Inuit house. These expectations also take into account the typical contemporary and post-depositional activities recorded in the documentary record, archaeological analyses, and taphonomic research. These expectations are tested against a high- resolution archaeozoological assemblage using absolute and relative measures. Testing the zooarchaeological record against the documentary expectations will begin assessing which behaviours and activities identified in the archaeological record can be more broadly interpreted using documentary records.
19
The expectations are generated for different areas within a typical Iñupiat dwelling. These expectations will be tested using new methods for understanding faunal remains within features. Establishing a methodology for intra-feature analyses is important for comparative purposes among regions and across cultures, to explore similarities and differences among faunal assemblages, and is a secondary goal for this research. In many societies, houses are the basic unit of analysis; in the Arctic, houses are one of the main sources of archaeological material and having a method to understand faunal spatial patterning within houses is important. Spatial studies of domestic dwellings have rarely focused on faunal material despite the fact that this material may be one of the best preserved and least disturbed features of the archaeological record, especially in the Arctic. The addition of another line of evidence in analyzing space will offer comparisons and contrasts to the typical studies of architecture, artefacts, and space.
Finally, this research aims to analyze the lifeways of an early Thule dwelling in the Seward Peninsula-Kotzebue Sound region. Although much research has been done in the area, very little has focused on the early Thule period or the early Thule subsistence practices of the coast. This high-resolution analysis of the faunal material at Cape Espenberg will provide an interpretation of a typical early Thule dwelling in the region, where little data has been published on any faunal materials.
1.8 Organization of Dissertation
This introduction has discussed the theoretical context for the thesis. It has contextualized analogical interpretations and intra-house spatial studies. It has also discussed the use of the house as a bounded unit of analysis. The following chapters will describe Arctic culture history (Chapter 2), the Thule way of life (Chapter 3), and excavations of Feature 87 (Chapter 4). Chapter 5 establishes archaeozoological expectations derived from documentary records and taphonomic processes. Following these expectations, Chapter 6 describes the collection, identification, and analytic techniques used to interpret the faunal assemblage. Before testing the expectations, the faunal assemblage is analyzed to interpret the general subsistence of the Thule occupants of Feature 87 (Chapter 7). Chapter 8 describes the results of the expectations tests. Chapter 9 concludes the thesis by outlining the contributions and limitations of this work.
20
Chapter 2 The Cultural Setting
The northern edge of the North American continent has a long history of interactions among different regions and cultural groups. Initial migrants into the Arctic spread from Alaska to Greenland following terminal Pleistocene deglaciation and were small-scale, mobile hunter- gatherers made up of nuclear families. A second group of people replaced these initial Palaeoeskimo groups starting around 200 CE in the Bering Strait and Chukchi Sea regions, and spread across the North American Arctic to Greenland around 1200 CE. These Inuit people are ancestors of the modern and historic groups encountered by European explorers. Even prior to European historical contact, Inuit groups were often interacting with regionally diverse groups within their own culture and beyond.
The following literature review will look at diachronic temporal and cultural similarities and differences. Most of the focus will be on the western Arctic, as the site used as the case study is located on Cape Espenberg, on the Seward Peninsula in northwest Alaska; however, an understanding of the regional history, including the migrations and interactions with the eastern region is important for interpreting the cultural divisions in Alaska. For this research, the western Arctic comprises the coastal and interior Mackenzie Delta west through Alaska to the Alaska Peninsula and the Siberian Bering Sea and Chukchi Sea coasts; the eastern Arctic is from Coronation Gulf east to eastern Greenland. The Thule Inuit, the focus of this dissertation, were defined and have been subsequently researched in greater depth in the eastern Arctic, with work in the western region largely focusing on both earlier and later groups despite the fact that the origins and development of Thule clearly occurred in the west.
Diachronically, there are two major cultural traditions referred in this review: the Palaeoeskimo1 and the Inuit2 traditions. In the western Arctic, coastal Palaeoeskimo cultural traditions were
1 Palaeoeskimo refers to the western Denbigh, Choris, Norton, and Ipiutak cultures and the eastern Pre-Dorset, Saqqaq, Independence I and II, and Dorset cultures. These groups are often divided into multiple different traditions, including the Arctic Small Tool tradition (Denbigh Flint Complex, Pre-Dorset, Independence I, and Saqqaq cultures), the Norton Tradition (Choris, Norton, and Ipiutak cultures), and the Dorset Tradition (Early/Middle and Late Dorset cultures); however, here they are treated as part of a larger cultural tradition with variable cultural iterations.
21 replaced, assimilated, or developed into Inuit cultural traditions around 200 CE3 (Gerlach and Mason 1992; Mason 1998; 2000; Mason and Barber 2003; Mason and Gerlach 1995a; Raghavan et al. 2014), although there is evidence that some groups existed until around 1200 CE (Anderson 1984; Gerlach and Mason 1992). Western Inuit culture diversified and eventually developed into ethnographic Alaskan Iñupiat and Yupik cultures. In the eastern Arctic, Palaeoeskimo people arrived later via a poorly understood migration around 3000 BCE (Friesen 2013a). They lasted until 1300 CE when Inuit groups from the west replaced them (Raghavan et al. 2014), although there may have been Palaeoeskimo groups that lasted in pockets of the Eastern Arctic until a later date (Hayes et al. 2005; Helgason et al. 2006; Hollinger et al. 2009). The Inuit who replaced the eastern Palaeoeskimo groups eventually developed into the ethnographic Inuit and Inuvialuit groups of Canada and Greenland.
After a brief note on terminology, this review outlines Arctic prehistory from the beginning of the Inuit tradition. The focus will be on the emergence and development of Thule Inuit culture within the Bering Strait regional context. Thule Inuit way of life will be discussed separately in Chapter 3.
2.1 Terminology
In the Arctic, cultural traditions are defined as all the archaeological cultures or cultural units that can be traced over generations via technological continuities (Giddings and Anderson 1986:XXVII; Willey and Phillips 1958); according to this definition, cultural traditions imply material culture continuity, but not necessarily genetic continuity. Cultural traditions are made up of cultural units that are defined as “the constellation of artifacts, methods of doing things, and patterns of arrangements that characterized a group of people at a particular time and in a
2 Inuit refers to the Okvik, Old Bering Sea, Punuk, Birnirk, Western Thule, Thule, historic Inuit, Inuit, Iñupiat, Yupik, Inuvialuit, and all other cultural descendants of Thule. ‘Inuit’ is used to describe the larger cultural tradition of the past 1000 years. Other terms that have described a similar cultural grouping are the Northern Maritime tradition (Collins 1964), the Thule tradition (Dumond 1977; Dumond 1987), the Neo-Eskimo tradition (Birket-Smith 1959; McCullough 1986; Steensby 1916), the Arctic Whale Hunting culture (Larsen and Rainey 1948), the Asian tradition (Bockstoce 1979), or the Inuk tradition (MacNeish 1959). 3 Unless stated otherwise, all dates are calibrated calendric dates.
22 particular place” (Giddings and Anderson 1986:XXVII). Cultural traditions and cultural units are important to define as over the years, different authors have used similar terminology and names to classify different suites of traits and different groups of people. All relevant cultural traditions (i.e. Palaeoeskimo and Inuit) and cultural units (i.e. Punuk, Birnirk, and Thule) are defined in this review and the characteristics that differentiate them from other cultural traditions or units are outlined.
2.2 The Inuit Tradition
Approximately 1500 years ago ancestors of modern Inuit, Iñupiat, Inuvialuit, and Yupik developed a novel way of life in the Bering Strait region based on the regular procurement of large marine mammals (walrus and baleen whales), a more sedentary settlement pattern, the use of cemeteries, marked social stratification, and inter-community forms of organization (i.e. trade, alliance, and conflict) (Mason 2009a; Mason and Barber 2003). They were the first of the Inuit tradition, whose descendants include modern Inuit, Inuvialuit, Iñupiat, and Yupik. The Inuit tradition encompasses people with a wide range of variability in technology and subsistence, increased settlement size, economic complexity, and social differentiation (Anderson 1984; Dumond 1977; Giddings and Anderson 1986; Mason 1998). The relationships among the archaeological groups in Alaska are difficult to determine because most sites lack modern, well- contextualized radiocarbon dates; most of them are dated via relative measures such as artefact types or stratigraphy (Gerlach and Mason 1992).
Based on evidence available now, Okvik-Old Bering Sea, Punuk, Birnirk and Thule seem to form a web of interaction throughout the Bering Strait region, with the Thule culture emerging around 1200 CE and rapidly expanding throughout the North American Arctic from the Alaskan Peninsula to eastern Greenland. Subsequent to this expansion, Thule groups started to adapt their lifeways to their new social and ecological environments. These developments were marked by variable changes throughout the Arctic, with some groups showing increased evidence of complexity and other groups showing decreased levels of complexity. In the western Arctic, particularly the northwestern regions of Alaska, this regionalization continued from the expansion of Thule through to the historic period. In many places across the Arctic, including northwest Alaska, there is a distinct change in settlement around 1450 CE as groups adapted to their regional ecological and cultural environments. Although the largest whaling villages like
23
Point Hope and Barrow maintained their population levels and settlement densities, overall, settlements became more dispersed with a movement into the interior regions around the mid- fifteenth century (Anderson 1984; Giddings 1952; Mason and Barber 2003). These groups continued to develop and adapt, which led to the distinct groups described by European and Russian explorers, traders, whalers, and ethnographers during the historic period. Despite this regionalization, at the time of contact, the Inuit, Inuvialuit, and Iñupiat from the Seward Peninsula to Greenland were genetically, linguistically, and culturally similar (Hayes et al. 2005; Maxwell 1985).
2.2.1 Okvik-Old Bering Sea
The first Inuit cultures were the Okvik and Old Bering Sea cultures that were contemporaneous with the Palaeoeskimo Ipiutak culture in Alaska. Okvik and Old Bering Sea are roughly contemporaneous localized cultures with different art and decorative styles that were originally thought to be temporally distinct (Ackerman 1984); Collins (1937) excavated and named the type-sites for Old Bering Sea, while Rainey (1941) named the Okvik culture. Now they are considered part of the same cultural group with different styles. Although still poorly dated, the Okvik-Old Bering Sea phase seems to have started around 550-650 CE and may have lasted until around 1050-1250 CE (Gerlach and Mason 1992:65; Mason 1998:254); in areas such as St. Lawrence Island, Cape Dezhneva, and sites on the North Chukotka coast, Old Bering Sea culture may have lasted into the fifteenth century CE (Mason 2009b).
Okvik-Old Bering Sea people of the Bering Strait region regularly hunted whales, walruses, seals, birds, and caribou (Ackerman 1984; Jensen 2009; Mason and Barber 2003). They used toggling harpoons along the sea ice edge, bow and arrow for interior land mammal hunting, multi-pronged bird spears, sealing darts with throwing boards, kayaks, umiaks, hand-drawn small sleds, snow goggles, fish leisters, fishhooks, chipped stone tools in greater numbers than ground stone tools, bone, ivory, and antler tools (Ackerman 1984). The most distinguishing feature of the Okvik-Old Bering Sea cultural phase was the production of intricate and beautiful pieces of art such as ivory carvings of animals and female figurines, elaborate art styles found on harpoon heads, ‘winged objects’, figurines, and drum handles (Ackerman 1984; Mason 2009a). Most of these pieces are found in burials. The designs are distinctly carved in the round with delicate curvilinear patterns of incised lines (de Laguna 1947). These decorations are similar in style to
24 the contemporary Palaeoeskimo Ipiutak styles. Although the Okvik-Old Bering Sea groups may have lasted until around 1400 CE and Ipiutak groups until at least 1100 CE, the height of interaction, population, and art production ended around 900 CE when drastic population reductions occur in both Ipiutak and Okvik-Old Bering Sea cultures on both sides of the Chukchi and Bering Seas. Most proxy climatic records imply that there was a drastic climatic event around 900 CE that may have impacted these cultures (Anderson et al. 2005; Hu et al. 2001; Mason 2009a; 2009b; Mason and Barber 2003; Mason and Gerlach 1995a). Prior to this shift, the Okvik-Old Bering Sea and Ipiutak cultural groups were living in warmer and stormier conditions with frequent open water; subsequent to this shift, the Punuk, Birnirk and Thule cultures emerged bringing with them innovations in sea mammal hunting technologies during cooler temperatures and heightened ocean productivity (Mason 2009b; Mason and Barber 2003).
2.2.2 Punuk
Although originally defined stylistically (Collins 1937), groups identified with Punuk culture also show an intensification of subsistence practices (bird and whale hunting), the adoption of various military technologies, and distinctive artistic styles and cosmological beliefs in comparison to the Okvik-Old Bering Sea cultural groups (Bandi 1995; Collins 1937; Mason 1998; 2009a; 2009b). Punuk style art and tools are dated to as early as 600 CE in the Bering Strait around St. Lawrence Island (de Laguna 1947; Mason 2009b). However, it is still uncertain where and when Punuk people developed. This may be due to the fact that Punuk likely had its greatest and earliest extent on the Siberian shores of the Bering Strait; although cemeteries have been relatively well excavated, settlements and dwellings have seen much less work (Bronshtein and Dneprovsky 2002; Dneprovsky 2002). As it stands now, the largest Punuk polities were likely on the eastern shores of St. Lawrence Island and on the south shore of Chukotka (Ackerman 1961; Chard 1955; Dneprovsky 2002; Mason and Barber 2003).
It is hypothesized that Punuk people spread outwards from a centre on St. Lawrence Island to other areas along the Bering Sea coasts of Alaska and Siberia, with early dates from the Alaskan shores around 800 CE (Harritt 2004). With the evidence available now, the height of Punuk influence was centred on St. Lawrence Island from 900 to 1000 CE, concurrent with a low storm episode and opening of a northern polynya along St. Lawrence Island (Blumer 2002; Mason 1998; Mason and Barber 2003). On the Siberian side, large population aggregates may have also
25 arisen at this time, potentially due to the expanding whaling economy or interaction networks (Harritt 2004). Along with these increases in population was the appearance of items related to offensive, defensive, and cohesive interaction (Ackerman 1984; Bandi 1995). Offensive and defensive items include Asian compound bows, bow braces, sinew twisters, bow guards, and slat armour (Collins 1937; Mason 2000). Drums potentially offer evidence for cohesive interaction (Collins 1937; Mason 2000). These indications of increased interaction may indicate that people of the Punuk culture were part of a complex network of trade, exchange, conflict, and alliance, with Okvik-Old Bering Sea and Birnirk people living contemporaneously in the north along the Chukchi Sea shores in Siberia and Alaska (Dneprovsky 2002; Harritt 2004). This period of interaction is described below, with evidence of Punuk influence ending around 1200 CE after variably influencing both Birnirk and Thule cultures (Bronshtein and Dneprovsky 2002; Dneprovsky 2002; Harritt 1994; Harritt 2004; Friesen 2013b; Mason 2009a; 2009b; McCullough 1989; Schledermann and McCullough 1980).
2.2.3 Birnirk
In contrast to the spread of Punuk outward from St. Lawrence Island along the coasts of the Bering Strait region, the Birnirk culture is first found along the shores of the Chukchi Sea, with a later movement towards the Bering Strait (Ackerman 1984; Mason 2009a). The most commonly held hypothesis for Birnirk origins is that it emerged from the Old Bering Sea culture in the Bering Strait region (Ackerman 1984; Anderson 1984; Bockstoce 1979; Collins 1937; 1964; Ford 1959; Stanford 1976). An alternative hypothesis is that Birnirk emerged on the Chukotka coast (de Laguna 1947) and then migrated to the Alaskan coast. Evidence for this colonizing model comes from the fact that only seven Birnirk sites are known in northwest Alaska, all of which are along the coast. Due to the lack of Birnirk sites in Alaska, it is hypothesized that Birnirk may have had a greater presence in Siberia (Gerlach and Mason 1992). Csonka (2003) suggests that Birnirk were likely once a strong presence at Ekven on the East Cape, which was likely an important centre for Birnirk culture. Based on sites known today, Birnirk culture seems to have two or three central regions: Barrow, the north Chukotka coast, and Cape Baranov (Mason and Bowers 2009). The southernmost extent of the Birnirk culture on the western Bering Strait shores is around Kotzebue Sound, with new evidence for Birnirk at Cape Espenberg, where Feature 87, the source of the faunal material for this thesis, is located. Giddings and
26
Anderson (1986) also classified one of the settlements at Cape Krusenstern, just north across Kotzebue Sound from Cape Espenberg, as Birnirk based on artefact categories and house styles.
Birnirk was one of the first cultures that was defined in the western Arctic. Stefansson (1941b) identified the distinctive harpoon head type at the Birnirk site in 1909, which was subsequently described by Wissler (1916) and Mathiassen (1929:50). Now, Birnirk sites or site components are identified based on a number of artefact types, including the two distinctive single-barbed harpoon heads termed ‘Birnirk’ and ‘Natchuk’ (Mason 2009a), along with other artefacts such as bola weights and wound pins which do not occur earlier in the archaeological record (Mason 2009b). The Birnirk subsistence economy is proposed to be different from their regional Ipiutak predecessors in a number of ways. A wide variety of harpoon heads show that Birnirk people were hunting seals in a number of new ways. Wound pins may indicate a new form of seal hunting along the ice edge, and are not found outside of Birnirk and Thule sites (Mason 2000). There is also intensification in bird hunting and fishing. However, there have been limited detailed faunal analyses to test the hypothesis of changes in subsistence between Birnirk and earlier cultures (but see Stanford 1976). This burst of innovation in subsistence pursuits is hypothesized to be linked to population growth and increased inter-regional interaction, potentially driven by the immigration of these new migrants into the Alaskan region (Mason 2009a). Birnirk is less well dated than Punuk but may have developed as early as 700 CE on the Siberian coasts, with major incursions into Alaska after 900 CE. Birnirk culture likely existed until 1300 CE (Mason 1998; 2000; 2009a; 2009b; Mason and Bowers 2009). The Birnirk decorative style is found in early Thule assemblages across the North American Arctic to Greenland, suggesting a Birnirk origin or at least heavy Birnirk influence on Thule cultural development (Anderson 1984). Prior to 1300 CE, the distinct styles of Birnirk, Ipiutak, and Punuk assemblages are maintained, even when found together within the same sites. This indicates that each group has a distinct identity and history, although there is evidence for interaction among all three cultures (Collins 1937; de Laguna 1947; Harritt 1995). By around 1300 CE, there is little evidence for these distinctions and there is growing evidence that many aspects of these cultures have been amalgamated into the new, dominant, Thule culture.
27
2.2.4 Thule
Thule was originally identified as an archaeological culture by Therkel Mathiassen on the basis of his research during the Fifth Thule Expedition from Greenland (Mathiassen 1927a). Thule culture, antecedent to the modern Inuit, Iñupiat, Yupik, and Inuvialuit of the Arctic, was more similar to cultures in Alaska and Greenland than to the Inuit of central Canada at the time of contact. As defined by Mathiassen, Thule people were maritime-adapted, lived in permanent winter settlements which consisted of substantial semi-subterranean sod-covered houses, and hunted marine mammals (whales, seals, and walrus) and caribou. As a general description, Mathiassen’s 1927 definition of Thule still holds true today. Although originally identified and defined in the eastern Arctic, Thule was quickly identified in the western Arctic by Larsen and Rainy (1948) at Point Hope. Since then, Thule has been identified on the coast and interior from Kodiak Island in southeast Alaska to the eastern shores of Greenland. Dumond (1987:118) defines Thule more specifically based on specific aspects of material culture in a specific geographic region: “all the midden-building, polished-slate-making, lamp-burning, kayak- and umiak-paddling Eskimos of later times, who extended from Kodiak Island to Greenland”. Thule is also defined based on artefact types; for example the Sicco and Natchuk harpoon heads are used as markers of early Thule sites in the eastern Arctic and linked to Thule origins in the western Arctic (Friesen and Arnold 2008; Morrison 1989; 2001).
These definitions are based on generalizations, and as Thule groups continue to be studied, differences and similarities are being teased apart. Thule, found from the Bering Strait to eastern Greenland over 300-400 years, exhibited vast variation in subsistence, technology, and material culture; some groups maintained a strong whaling component (McCartney and Savelle 1993; Savelle 1986; Savelle and McCartney 1999; Savelle and Wenzel 2003; Whitridge 1999a), others hunted seals for the majority of their diet (Morrison 1983a; 1983b), while others focused on terrestrial resources (Norman and Friesen 2010). Many of these differences are likely due to regionalization described below, but also likely a characteristic of the inherent flexibility Thule people inherited from their ancestors (Savelle 1987). Originally identified in the east and distinctly different from its Late Dorset predecessors across the Canadian and Greenlandic Arctic, Thule culture is easier to isolate and study in the eastern Arctic. In Alaska, the origins of Thule are unclear, with potential influences from Birnirk, Punuk, and even Ipiutak peoples.
28
2.3 Thule Inuit Origins
Since the discovery of Thule whaling village sites in the eastern Arctic and Mathiassen’s (1927a; 1927b) extrapolation of the isostatic uplift rates to estimate the age of Thule occupation at 1000 CE, archaeologists in Alaska have built assumptions of Thule origins based on a pre-1000 CE date. With the date of the early Thule migration into Canada pushed back to the late twelfth or early thirteenth century (Friesen and Arnold 2008), Thule origins in Alaska are now not restricted to those centuries prior to the first millennium CE. Rather, they could date anywhere up to the late twelfth century. With new radiocarbon dating of key sites, Thule origins are likely to shift to match the later expansion of Thule into the eastern Arctic, as is evident in the recent re-dating of Cape Krusenstern (Anderson and Freeburg 2013). Some groups of Thule in the eastern Arctic exhibit artefact traits that are almost identical to Punuk and Birnirk traits in Alaska, and so the origin of Thule out of these cultures must not date much before the expansion of Thule into the eastern Arctic (Friesen 2013b). Although the re-dating of Thule expansion between 1000 and 1200 CE is a first step in understanding Thule origins and development, this period of culture history is generally not well understood. One of the factors that limit our understanding of this period is the lack of extensive and systematic radiocarbon dating in northwest Alaska, a proposed area of Thule origins (Gerlach and Mason 1992; Mason and Bowers 2009). A second problem in understanding and interpreting the beginning of Thule is that material culture affiliation for artefacts, components, and sites is disputed (e.g. designs on Sicco harpoon heads). Despite these problems, Thule after 1200 CE represents a cultural unit that was important in both the western and eastern Arctic.
Thule culture could have originated from the Birnirk or Punuk peoples, although an uninterrupted descent solely from one of these cultures is unlikely. The most common theory for Thule origins is that they originated from the Birnirk culture, with significant influences from the Punuk culture of the Bering Strait region (Anderson 1984; Collins 1929; Collins 1937; Collins 1940; Dumond 2002; Ford 1959; Giddings and Anderson 1986; Hoffecker et al. 2012; Hollinger et al. 2009; Kankaanpaa 1996; Larsen and Rainey 1948; Mason 2009b; Maxwell 1985; McCullough 1989; McGhee 1969/70; McGhee 2009a; McGhee 2009b; Morrison 1989; 2001; Stanford 1976). Bandi and Blumer (2002) suggested that on St. Lawrence Island, Birnirk and Punuk peoples integrated to form a new culture, likely the Thule culture, but there is little evidence of this St. Lawrence Thule culture spreading or amalgamating in to the later prehistoric
29
Thule forms. Potentially, multiple groups could have made the shift to a broader Thule culture at a similar time if they existed in a system of closely linked interaction networks. However, sufficiently tight chronology is not available to test these hypotheses.
If Thule developed from Birnirk, there are two broad theories of where this occurred: in the Bering Strait centre where whaling hot spots are located or on the north Alaskan and Siberian periphery in response to cultural developments at the hot spots (Mason and Bowers 2009; Mason and Gerlach 1995a). There is limited evidence for a Birnirk to Thule transition in the Bering Strait region. Collins (1940; 1964) and Jenness (in Morrison 2002) originally hypothesized that Kurigitavik at Wales was a site of Birnirk to Thule transition with influences from Siberian Punuk culture. It does show a small percentage of Thule style harpoon heads, but with Sicco, the St. Lawrence Punuk-like harpoon head, as the most frequent (Mason and Bowers 2009).
Most evidence indicates that Thule developed on the margins of the Chukchi Sea, although an exact location or area is yet to be established as multiple sites show a Birnirk to Thule transition. Morrison (2001) has suggested that the collection of three harpoon head types (Sicco, Natchuk, and Tasik) were used by both Birnirk and Thule; based on these criteria, the Birnirk to Thule transition can be seen at multiple sites (Ford 1959; Hoffecker et al. 2012; Mason and Bowers 2009; Morrison 2001; Stanford 1976). Dates from Qimiarzuq (Jabbertown), Deering, Birnirk, and Uivvaq place the transition around the eleventh century (Hoffecker et al. 2012; Mason and Bowers 2009; Morrison 2001) with the Natchuk sealing harpoon head as a transitional artefactual cultural marker between Birnirk and Thule (Morrison 2001). A Birnirk to Thule transition on the Chukchi Sea margins lacks evidence due to erosion of potential areas of transition in the Beaufort Sea area and the lack of analysis of northern Siberia and northern Alaskan coastlines. However, there is some evidence for this transition in the northern Alaskan area. Collins (1929; 1937; 1940) originally hypothesized the relationship (de Laguna 1947; Ford 1959; Maxwell 1985; Morrison 1989; 2001) between Birnirk and Thule for St. Lawrence Island, Jenness (in Morrison 2002) for Wales, but Stanford (1976), following Ford (1959) was the first to supply evidence for a smooth transition based on both harpoon heads and faunal remains from Point Barrow (Figure 1). Along with harpoon head styles, an increase in regular, intensive whaling is also characteristic of the Thule transition, with less intensive, likely more opportunistic, whaling during the Birnirk cultural phase (Jensen 2012; Mathiassen 1927b; Savelle and McCartney 1994; Stanford 1976). However, the transition from Birnirk to Thule was likely much more
30
Figure 1: Harpoon head developmenet from Birnirk to Thule (from Stanford 1976:108)
31 complicated and nuanced than the presence or absence of three harpoon head types and intensification of whaling can indicate.
The use of the three harpoon heads types as a marker of the Birnirk to Thule transition and early Thule phase brings with it problems. As Mason and Bowers (2009) state, the inclusion of Sicco harpoon heads complicates the straightforward assumption of a Birnirk to Thule transition since the Sicco harpoon head is closely tied to the Punuk people. One hypothesis is that Punuk people may have moved from St. Lawrence Island to the North Slope where they interacted with Birnirk-Thule people. Birnirk-Thule people then moved away from the Punuk immigrants east to the Canadian Arctic and west to the Siberian Arctic (Mason and Bowers 2009). However, this does not account for the fact that Thule groups are found shortly after this eastward expansion along the entire coast of Alaska. If Punuk groups displaced Thule from the North Slope, there should be evidence of a distinct Punuk culture from St. Lawrence to the North Slope. This is not the case, as most Punuk artefacts are found within Thule assemblages. Skeletal evidence has also brought into question the smooth transition of Birnirk people to Thule people as Hollinger et al. (2009) argue that Birnirk groups were not biologically related to the later Thule groups that occupied the areas (Hollinger et al. 2009; contra Jensen 2009). However, this contradicts the archaeological data (Jensen 2009) and new genetic analysis (Raff et al. 2013; Raff et al. 2015). Even if the genetic and archaeological evidence indicates a close relationship between Birnirk and Thule, Punuk culture certainly influenced the early Thule through Thule adoption of the decoration on Sicco harpoon heads (Friesen and Arnold 2008; Morrison 1999; 2001). In addition, the accumulating evidence for a cultural continuum from various sites, not just around Point Barrow, is inherently problematic as these sites have variable evidence for the relationships among Birnirk, Punuk, and Thule cultures across northern Alaska (Harritt 2004; Hoffecker et al. 2012). Understanding how Thule emerged is likely not going to be as easy as identifying a direct descent and subsequent expansion from either Punuk or Birnirk, but may be conceptualised as process of becoming that was generated through interactions with multiple different groups at different sites where, eventually, Thule culture emerged.
2.3.1 Thule Origins in Context
Thule culture traits emerged at a time of complex, multi-cultural communities with shifting networks of trade, alliance, and conflict in northwest Alaska, the Bering Strait region, and the
32 eastern coasts of Siberia. These complex interactions occurred at a time when people were increasing their focus on whaling. It is during this process that Thule culture developed and expanded.
Mason (1998; Gerlach and Mason 1992; Mason and Gerlach 1995a) has constructed a hypothesis of multiple contemporary groups that parallels the “nation-society” model for the nineteenth century for the centuries before the emergence of Thule. Gerlach and Mason (1992) suggest that the cultural differences identified at this time—Okvik-Old Being Sea, Ipiutak, Punuk, and Birnirk—were used to create distinct social boundaries based on competitive exclusion for access to resources. These groups were marking their social boundaries via artistic, artefact, and harpoon head styles in the Bering Strait region between 50 BCE and 1000 CE4 (Bronshtein and Dneprovsky 2002; Gerlach and Mason 1992). By 1000 CE, Punuk, Birnirk, and some Thule traits are documented across northern Alaska (Mason 2009b). Accompanied by these traits are other shifts: a greater reliance on polished-slate implements, the introduction of oil-burning lamps made of clay, distinctive modes of pottery production, an increased reliance on marine mammals and marine mammal hunting techniques, and in some cases an increased emphasis on whaling. Over the next two hundred years or less, these changes, in some areas, gave rise to the Thule Inuit people who migrated into the eastern Arctic.
Thule people did not carry a strictly uniform culture in the form of artefact styles or subsistence traits, but they represent an adherence to a broad underlying Thule way of life. This way of life is maritime-focused, with semi-permanent winter settlements composed of families living in semi- subterranean houses and subsisting on stored food (Mathiassen 1927b). Although the material culture is not completely uniform, there are no material indicators of distinct social boundaries unlike those found in the preceding period. One explanation for this is that Thule people out- competed and outlasted the contemporary Birnirk, Punuk, and Ipiutak groups due to innovation, adaptation, or warfare. An alternative hypothesis is that Thule culture developed in multiple places from multi-cultural communities and amalgamated certain material traits from the
4 It is important to note that re-dating of sites indicates that many of the original dates for sites, cultures, and transitions were a few hundred years too old; with new methods, important events such as the Thule migration and end of Late Dorset in the eastern Arctic are being re-dated to a few hundred years later. Except where noted, material and cultures have not been re-dated and thus may indicate an age that is too old.
33 previous cultures to form the new Thule culture. In this scenario, different sites will have different traits depending on the interaction of individual groups at the time, as is evident is areas such as St. Lawrence Island (Bandi and Blumer 2002), Wales (Harritt 2004; 2010), or Cape Krusenstern (Anderson and Freeburg 2013). Due to social or environmental factors such as the need for larger communities to hunt whales, the Thule culture may have negated the need to maintain social boundaries; instead of competitive exclusion, people were incorporating under a new social and environmental reality where cooperating in larger groups worked to maintain key resource areas. In reality, the group interaction was likely based on both competitive exclusion and social cohesion depending on the region, people, and time period as is described in documentary sources (Burch 2005; 2006). However, if Punuk, Birnirk, Early Thule, and Ipiutak traits existed in the Bering Strait region at the same sites starting around 1000 CE, and the emergence of Thule is re-dated to around 1200 CE as suggested above, there are potentially a few hundred years where social boundaries within multi-cultural villages may have been altered through a series of small changes including marriage, trade, and artefact sharing. These alterations may have introduced cultural commonalities that were adopted by the village. As suggested by Maxwell (1985) for the Birnirk to Thule transition and as stated by Mason (2000:245) for the emergence of Birnirk: “The mechanism(s) producing hybrids are straightforward: conflict or alliance producing the amalgamation (Burch 1998a:320–327), with the emulation or imposition of higher status cultures on “lower” or poorer groups (McGhee 1988:371)”. This type of interaction would allow for the creation of a broad way of life with local variations depending on the individual, the family, their situation, the village, and their interactions at broader scales. This could account for the variation seen in the eastern Arctic with the initial Thule migrants; groups from different communities were moving, bringing different amalgam of traits with them. Although the mechanics of this scenario are at this time solely a hypothesis, re-dating and recent excavations at larger sites hint at this possibility (Anderson and Freeburg 2013; Harritt 2004; 2010; Gerlach and Mason 1992; Mason 2000).
At the same time that groups were developing strong ties to whaling localities and increasing the population size of these large sites, people at smaller sites such as Cape Espenberg and Onion Portage were using resources such as seals, caribou, fish, or birds (Harritt 2004). The settlement and resource patterning of large whaling sites and small secondary resource sites was potentially maintained within an inter-regional interaction sphere where resource and social trade between
34 these two types of settlements was necessary for mutual survival (Harritt 2004). If the larger settlements were the site of Thule cultural formation based on cross-cultural interactions, these smaller sites may have also come under the influence of the emerging Thule culture due to tight inter-regional interaction spheres.
In this social cohesion hypothesis (c.f. Gerlach and Mason 1992), people from different groups may have lived in villages together during a period when an increase in storms may have led to a strain on the trade and interaction networks of the previous 1000 years (Gerlach and Mason 1992). Gerlach and Mason (1992) suggest that this ecological change may have affected groups differently; the Okvik-Old Bering Sea and Ipiutak polities seem to have declined significantly around 1250 CE, while the Punuk, Birnirk, and Thule groups may have readjusted their relationships. Concrete evidence for this hypothesis is scarce, although both re-dated and older un-dated sites indicate that it may yet be supported with more research. Mason (2000) identifies multiple sites between 500 and 1000 CE where contemporaneous spatially distinct cultural groups (Birnirk, Punuk, Ipiutak, and Thule) lived together in close proximity at the time of Thule development, such as Deering, Cape Krusenstern, Kukulik, and S’keliyuk. Mason suggests that this time was a period of “ethnic enclaves, each with its own history of matial [sic] and marital interchanges, sibling rivalries, and long-standing feuds, as well as trade relationships” (Mason 2000:230). At Wales, Harritt (2004; 2010) is re-dating and conducting new excavations that show distinct and contemporaneous Birnirk-Thule and Punuk settlements from 800 CE until around 1100 CE. At Cape Krusenstern, Anderson (Giddings and Anderson 1986:93) gives evidence of “Ipiutak-like lithics and Birnirk and early Thule organic artefacts”. New dating at Cape Krusenstern shows a longer, more sustained, and more complicated occupation history than suggested before (Anderson and Freeburg 2013); however, more detailed work on occupations and dwelling structures and more dating of artefacts is needed before any claims can be made regarding multi-cultural communities. At the site of Kitnepaluk, Mason and Barber (2003) suggest this originally designated ‘pure Thule’ site may in fact have more stylistic elements in common with the Punuk culture than Birnirk or Thule. Other sites such as those on St. Lawrence Island and at Ekven show mixtures of Birnirk, Punuk, and Thule traits but have not been reported or dated in enough detail to assess contemporaneity (Bronshtein and Dneprovsky 2002; Csonka 2003; Dneprovsky 2002; Gulløv and McGhee 2006; Mason 2009b). Another site that has yet to be dated but that has multiple sets of diverse harpoon heads that are potentially a result of
35 multiple communities is the Nunagiak locality on the North Alaskan coast (Ford 1959; Mason 2009b). As Mason (2009b:96) states: “The diverse number of harpoon head types within the Nunagiak mounds…implies either repeated interactions with discrete ethnic intrusions or a lengthy sequence of internal stylistic development”. Like most other sites in Alaska, determining the history of this site must wait further dating, but Nunagiak is a strong candidate for testing the hypothesis that Thule culture emerged when people in multi-cultural communities coalesced into a new cultural unit. Although the dating of Punuk counterweights or “winged object” is limited, the presence of these distinctive Punuk artefacts at sites such as Uivvaq and Nunagiak in north Alaska in the eleventh century indicates that through trade or migration, people using Punuk implements and decorations were living in close proximity to Birnirk and Thule peoples (Ackerman 1984; Collins 1937; Ford 1959; Hoffecker et al. 2012; Mason 2009b). Ackerman (1961) also identifies Punuk and Birnirk in close stratigraphic context at S’keliyuk, with distinct Punuk styles appearing on Birnirk harpoon heads. Evidence is accumulating to support the hypothesis that Thule culture, emerging from a “patchwork of distinctive societies” (Gerlach and Mason 1992:67), changed the social and political fabric of the Arctic. Thule culture may have emerged in areas where people needed to cooperate in large groups to acquire large marine mammals. In the previous centuries, Birnirk, Punuk, and sometimes Ipiutak groups were living in close proximity to each other; as these social boundaries were permeated, a new Thule culture developed that adopted and incorporated the previous cultural diversity.
2.4 Thule Inuit Expansion
The origin and expansion of Thule Inuit are likely closely intertwined. As described above, although Thule culture traits appear in multiple areas starting around 1000 CE, the close links to Birnirk and Punuk in the eastern Arctic indicate that the transition from these earlier groups to Thule did not occur too long before the migration itself (Friesen 2013b). After the initial expansion, Inuit development is largely seen as an in situ process of development and specialization (Jensen 2009), although small-scale migration, interaction, and movement were likely continuous processes. The migration of Thule to the eastern Arctic has been examined in much greater detail than the Thule development and expansion in the western Arctic (Friesen and Arnold 2008; Friesen 2013b; Gulløv and McGhee 2006; Helgason et al. 2006; Hollinger et al. 2009; Krupnik and Chlenov 2009; Marchani et al. 2007; McGhee 1984a; McGhee 2000; Morrison 1999; 2000; Sheppard 1998; but see Dumond 2009; Jensen 2009; Marchani et al. 2007;
36
Stanford 1973). During the expansion east, Thule likely met with variable ecological and social environments; although there is little direct evidence to support interaction with the Late Dorset Palaeoeskimos, the Thule appear to have had no problem establishing themselves in the areas with access to large sea mammals (Friesen 2000; 2013b; Maxwell 1985; Whitridge 1999b).
The southward expansion of Thule traits into the western Arctic was variable and likely involved multiple movements that occurred over hundreds of years. Each expansion would have been different, with interactions through violence or social integration occurring when Thule Inuit occupied previously inhabited areas, and of adapting to the ecological environment when entering unoccupied areas (Anderson 1984). As in the eastern Arctic, initial Thule expansion likely occurred at productive sea mammal sites, with later expansion into areas with less access to whales and walrus. The earliest southern-most expansion puts Thule cultural traits at the Naknek River drainage around the twelfth century CE (Dumond 2009). This expansion likely involved some movement of people carrying the Thule traits from the north, but also an incorporation of local people into the Thule culture due to the strong retention of local traits. In the thirteenth century, Thule people are documented moving into, or at least occupying, different parts of Kotzebue Sound (Mason and Bowers 2009). As stated above, this was often cited as a potential source for Thule cultural origins. This has been tested by Mason and Bowers (2009) and does not seem to be the source for a Birnirk-Thule transition.
Debate continues about the motivations for this massive Thule expansion. Recent re-dating of the Thule migration across the Arctic to Greenland, and the movement of traits, and likely people, to the south and interior of Alaska, has been put at the late 12th or 13th century CE, when climate and environmental conditions were deteriorating in relation to open-water whale hunting (Maschner and McGhee 2009). This re-dating suggests new hypotheses about the causes of the Thule expansion, both in the western and to the eastern Arctic. For many years, multiple factors have been suggested to explain the large and diverse expansion of Thule people and their cultural traits. These motivations for migration are classified by Friesen (2013b) into three broad categories: the pursuit of bowhead whales driven by changes due to climate change (Bockstoce 1976; McGhee 1969/1970; Stanford 1976); search for raw materials such as meteoritic iron and Norse iron and bronze (Gulløv and McGhee 2006; McGhee 2009a; McGhee 2009b); and high population pressure with increased frequency of warfare where social conditions were favourable for emigration (Mason 1998; Mason and Barber 2003). Current interpretations of the Thule
37 migration into the Canadian Arctic blend several of these push and pull factors (Krupnik and Chlenov 2009), and a combination of these factors is also likely the cause of the western expansion. More work is needed in Alaska to understand the origins of this widespread Thule expansion into both previously occupied and unoccupied territories.
2.5 Thule Inuit Regionalization 2.5.1 Modified Thule Inuit
After the initial explosive expansion of the Thule culture around 1200 CE, groups started to settle into their new territories, developing regionalized acquisition strategies to procure subsistence resources. This settling in process likely took a few generations during which Thule Inuit learned about the new ecological and social environment, and re-adjusted their settlement pattern, economy, and social practices (Friesen and Arnold 2008). Regionalization, as a settling in process to a new environment, is influenced by both natural and social aspects of the new context, and the particular historical trajectory of the group (Jacobs and Stenton 1985; McGhee 2009b). Regionalization as a whole likely occurred much differently in the Alaskan region compared to the eastern Arctic due to differences in climate, ecology, population density, interaction spheres, ethnic groups, and social pressures. In the western Arctic, regionalization has been suggested to start at a number of different times: Bockstoce (1979) and Giddings and Anderson (1986) see regionalization beginning around 1500 CE; Mason (2009a) argues for a date around 1700 CE; for the Kotzebue Sound area specifically, Giddings and Anderson (1986) suggest that the Western Thule phase ends around 1400 CE when the assemblage at Cape Krusenstern shows a decline in whaling and settlement complexity. In the Mackenzie Delta region, Betts and Friesen (2004) suggest a qualitative shift occurring within a few generations during the early fifteenth century. Although many of the dates are not well contextualized, are run on poor material, or are quite old, it is likely that in different areas of the Arctic, regionalization occurred at different times for different reasons.
In the western Arctic, Thule groups settled into local whaling hot spots and moved into previously unoccupied areas, adapting their subsistence patterns to their local biological and social environments. At Cape Krusenstern, after 1400 CE, settlements changed from large multi- room dwellings to small, single-room dwellings in sparsely populated settlements (Anderson 1984). At the same time, whales declined as part of the subsistence economy, and people focused
38 more on fishing. In places such as at Point Hope and Barrow where whaling continued, settlements increased in size and density during the regionalization period (Anderson 1984). This regionalization continued through to the ethnohistoric period, where the regional variations of the Inuit culture were numerous and individual cultural groups were known by their specific names. Despite regionalization and diversification of groups, art and technical styles co-evolve through the region, leading to a hypothesis that interregional interaction via communication, trade, alliance, and conflict was both extensive and intensive (Anderson 1984).
Alternatively, economic and social specialization may have influenced the regionalization of Thule groups. As Yesner (1985) hypothesizes and Sheehan (1995) elaborates, trade goods may have allowed groups to concentrate on specialized seasonal hunting activities. This would have alleviated some the pressure for conflicted scheduling of certain groups involved in the trading networks (Morrison 1994). Social gatherings, such as trade fairs (Betts 2005; Palsson and Stefansson 2001) would have facilitated the gathering of groups. Evidence of this specialization is found in numerous areas, including the Mackenzie Delta, and north and northwest Alaska. In these areas, coastal villages and interior settlements were occupied at the same time of year; they likely were part of a trading network of maritime and terrestrial resources where essential marine oil and caribou skins were traded (Morrison 1994; Sheehan 1995). These were goods were part of an integrated regional economy and culture. In Alaska, Sheehan (1995) proposes that this occurred from around 1200 CE to the late nineteenth century. Morrison (1994) agrees that this complementary trade-based subsistence strategy dates to the early Thule time from Alaska and the western Arctic, citing the Kobuk Valley and Kotzebue Sound as an example of this from northwest Alaska (Giddings 1952).
Tied to regionalization is evidence of major population increases in many areas, which would have led to intensification, and potentially diversification, of resources (Betts and Friesen 2004; Sheehan 1995). The development of socio-territories and territoriality has been tied to the development of locally based resource subsistence economies (Betts 2005). These socio- territories have archaeological and archaeofaunal signatures (Betts 2005; Eldridge 2012) and are well defined in the western Arctic documentary records (Burch 1998a; 2006; Ray 1964; 1975a; 1983).
39
In the eastern Arctic, regionalization also varies regionally and temporally. Overall, eastern Arctic regionalization was likely influenced by climate and social change as whaling collapsed in many areas, forcing people to establish a broader base of subsistence (Whitridge 1999b). Unlike regionalization through increased trade relations as suggested by Yesner (1985) and Sheehan (1995) for Alaska, Whitridge (1999a) has hypothesized that the regionalization in the central Canadian Arctic was affected by a collapse of communication and trade networks, not a strengthening of them. This decline of Thule culture and transition to regional Inuit groups is seen in changes to the style of winter dwellings, settlement patterns, and social interaction, as well as changes in subsistence patterns with later Inuit groups focusing more on sealing and less on whaling (Jacobs and Stenton 1985; McCartney 1980; Schledermann 1976a; Whitridge 2002b). A decrease in interaction may have led to a diversification of culture and subsistence strategies.
Both increased and decreased trade and communication relations may be mechanisms of regionalization that operated in the Arctic; the central Canadian Arctic environment, social fabric, and cultural history is different than the Mackenzie Delta and northwest Alaska at the time of contact. In the Mackenzie Delta and northwest Alaska, the extensive trade networks seem to lead to increasingly ‘complex’ social organization (Sheehan 1995). Morrison (1994) postulates that this economic model was brought into the eastern Arctic with the Thule migration to the area, and that early Thule sites in the Central and Eastern Arctic show a diversity of same-season subsistence pursuits that indicate such a complementary strategy. He suggests that western central Canadian Arctic sites like Clachan and Beulah in western Coronation Gulf were reliant on open-water summer/fall ringed seal hunting, while other sites like Lady Franklin Point on western Victoria Island were based on caribou (Morrison 1994). This could indicate that Thule people were trying to maintain subsistence based on a cultural precedent from the homeland with stored food being the key to winter survival; eventually, this model was not sustainable in the central Arctic and after regionalization people started hunting ringed seals on the ice in the winter and hunting caribou and fishing in the summer. Savelle (1987) and McCullough (1986) have suggested that early Thule subsistence strategies were ‘collector’ based with a strong reliance on stored food. They also postulate that early Thule did not rely on hunting seals at breathing holes, but that they developed this subsistence practice as they adapted to the local region. As Inuit groups in the eastern Arctic became more regionalized, their material culture
40 became diversified. By the historic period, groups in the eastern Arctic were part of distinct, named societies that had different subsistence-settlement patterns than the Inuit in the west.
Although the interregional interaction networks were likely quite different in the east and west, the function was the same: groups maintained cultural ties to vast areas. Even with the collapse of highly integrated groups in the central Arctic, interaction on a small scale would have occurred. In both areas, the interregional networks included interaction with non-Inuit people: in the western Arctic these included groups in Siberia, while in the eastern Arctic they interacted with the Norse. Sub-Arctic peoples interacted with the Inuit in most areas of the North American Arctic to varying degrees.
2.5.2 Historic Inuit
Historic Inuit culture is demarcated at different times in the Arctic depending on the time and type of contact groups had with the early European explorers. In Alaska, the earliest explorers were Russians, commissioned to acquire furs from the Siberian coast, the Aleutian Islands, and eventually southern Alaska in the late eighteenth century (Ray 1975a; VanStone 1984). In the Canadian Arctic, the first European explorers were the Norse, who first interacted with the Thule Inuit in the late thirteenth century (Maxwell 1985). Later European contact in the Arctic came from explorers and scientists, most of whom searching for the Northwest Passage. Differences in motivation—fur trade for the Russians and exploration for the Americans and British—resulted in different interactions with the native inhabitants. The Russian fur trade was quite disruptive, as local groups adapted their subsistence and settlement patterns, incorporated trade goods, and altered their social interactions. Explorers were generally less disruptive, as they had little interest in the inhabitants except for the supply of information, food, and water (VanStone 1984).
The contact between the Europeans and the local Inuit was variable. This variable interaction resulted in a wide variety of historical and ethnographic documents. Some of the expeditions into the Arctic brought ethnographers and archaeologists with them; other expeditions recorded very little, like the Russian-sponsored expeditions around the Chukchi Sea (Ray 1975a). Additionally, the ways in which local Inuit groups incorporated European goods into their cultural tradition varied greatly (Friesen 2013a; Walls 2009); in northern Alaska trade goods were rare until the middle of the nineteenth century, despite the fact that the first Russian trade goods appeared in
41 the eighteenth or perhaps seventeenth centuries (Anderson 1984). Ethnographic observations of the Inuit, Iñupiat, and Inuvialuit will be discussed in Chapter 3.
2.6 Summary
The house that is the focus of this dissertation was occupied in the fourteenth century, after the major Thule Inuit expansion, but prior to the local regionalization. The centuries prior to the occupation of the houses at Cape Espenberg saw the Thule people actively participating in and reacting to the inter-regional interactions, changing environmental regimes, and global historical forces that characterized this period(Maschner and McGhee 2009). Adaptations to the different interactions, environments, and historical forces continued through the Thule period into the historic period; the intensity and extent to which different settlements participated in this regional and pan-regional interactions likely fluctuated. Feature 87, occupied during the fourteenth century, would have housed people who knew the local and regional area well and who had been and were continually adapting to large and small-scale environmental, climatic, social, and cultural changes. They were also likely a part of regional and inter-regional interaction networks connecting them to the interior, major whaling villages, and other communities.
42
Chapter 3 Thule Lifeways
This chapter outlines the lifeways of Thule Inuit as constructed by archaeological data using documentary records as sources for interpretation. The first section of this chapter outlines the early Thule way of life. Here, early Thule is defined as the cultural group who extended from Alaska to Greenland from ca. 1000 CE to local regionalization, which varies across the Arctic, but can roughly be estimated at between 1400-1500 CE. The second section of the chapter describes the documentary sources used to describe the space in houses of northwest Alaska and to construct the expectations in Chapter 5. The third section of this chapter outlines the construction and use of space in houses of the northwest Seward Peninsula and southern Kotzebue Sound area based on the documentary sources. These descriptions will help contextualize the expectations developed in Chapter 5.
3.1 Thule Way of Life
This general construction of the Thule way of life will focus on subsistence, faunal processing, and dwellings, with brief descriptions of technology and interaction. It is important to note that for many studies of Thule lifeways, the Iñupiat documentary record from the Alaska North Slope region, from Point Hope to Point Barrow, is used as the source of analogy for early Thule archaeological sites across the Arctic, often preferentially over the local Inuit documentary records. These northwest Alaskan groups often show greater similarities in subsistence patterns, dwelling structures, and artefact technologies to early Thule archaeological assemblages than the local ethnographically recorded Inuit. Studies that have specifically used the northwest Alaskan documentary record as source material have looked at subsistence (e.g. McCartney 1980; McCartney and Savelle 1985; Savelle 2000; 2005; Savelle and McCartney 1999), interregional interaction (e.g. Friesen 2012a; 2012b), spatial patterning of sites (e.g. Savelle and Wenzel 2003; Whitridge 1999a), social roles (e.g. Grier and Savelle 1994; Whitridge 1999a; 2002a), and ideology (e.g. Patton and Savelle 2006; Whitridge 2004). However, the local direct-historical method, hunter-gatherer theory, and other cross-cultural comparisons have also been used to interpret the archaeology. It is from an amalgamation of these accounts, tested in the archaeological record, that a general Thule way of life has been constructed.
43
3.1.1 Interaction and Social Organization
Although the focus of this research is on subsistence strategies and use of space in a dwelling, it is important to place Thule Inuit groups within the larger regional and inter-regional context. Alliances, conflicts, and trading relations can drastically impact the resources available to groups, as well as the types and locations of their settlements.
Thule people likely maintained wide-ranging networks of trade and exchange based on kinships, formal sharing partners, and informal friendships. These networks existed among Thule Inuit groups, and between Thule Inuit and subarctic Aboriginal groups, Norse, and other Siberian and Alaskan Inuit groups. Within these networks, both friendly and aggressive interactions occurred. Archaeological evidence of these networks is seen in long-distance trade items such as soapstone, copper, and iron (Maxwell 1985; McCartney and Mack 1973; McCullough 1989; McGhee 1972; 2000; Morrison 1987; 1991; Schledermann and McCullough 1980). The creation and retention of extremely similar artefact types and styles over a large area provides additional evidence for the interconnectedness of early Thule people (Friesen 2012a; Mason and Bowers 2009; Savelle 2000). Western Arctic early Thule Inuit interaction and networks have been less studied than their eastern Arctic counterparts. Although a great deal of work has been done on the previous centuries and cultures in the western Arctic (i.e. Mason 1998), much of the interaction and network construction for early western Thule Inuit is taken directly from documentary records, especially those of Burch (1980; 1998a; 2005; 2006), Bockstoce (1979; 2009), and Ray (1964; 1975a; 1983).
Social organization of early Thule people is described at different scales from the regional group through to the family. Based on the documentary records of Burch (1998a; 2006) and Ray (1975a) and the archaeologically recorded similarities between early Thule and ethnohistoric groups’ subsistence-settlement patterns, dwellings, and artefacts, researchers have suggested that at least some Thule Inuit were arranged in regional groups or societies (Burch 2006; Friesen 2012b; Mason 2009b; 2012; Mason and Bowers 2009; McGhee 1997; Ray 1975a; 1983). Ethnographically, members of these groups shared a common dialect, group name, and territory (Burch 1980; 1998a; 2005; 2006; Ray 1964; 1975a; 1983), and in Burch’s (1998a; 2006) definition, a distinctive subsistence pattern. These regional groups were usually composed of one large village with several small linguistically related villages and the territory in between them
44
(Ray 1964; 1983). People in the smaller villages joined the large village in intergroup activities (Ray 1964; 1975a; 1983), although individuals and families sometimes moved between regions (Spencer 1959). Within these regional groups most families followed a general pattern; however, individual families often dispersed to different locations within the region for the warm season. More archaeological work needs to be done to establish the antiquity of the regional boundaries and subsistence-settlement patterns identified in the documentary record. General Thule settlement patterns were highly variable from Alaska to Greenland depending on the season, resource locations, and scheduling (Friesen 2012a; Savelle 1987). Both winter and summer settlement patterns are tied to preferred resource access, usually marine resources (Helmer 1992; Jacobs and Sabo 1978). However, site location decisions were also affected by local environment, neighbouring camps, protection from adverse weather (i.e. storms, insects), aesthetic considerations, ideological interpretations, and kin relations (Grier and Savelle 1994; Jacobs and Sabo 1978; Savelle 1987).
Below the scale of the regional group, interpretations of Thule social organization at the village level are slightly less reliant on the documentary record. Villages often consisted of semi- subterranean houses occupied over a number of generations (Ackerman 1984; Arnold and Hart 1992; Friesen 2012a; Giddings and Anderson 1986; Levy and Dawson 2009; Maxwell 1985; McCartney 1977; Savelle 1987). In most areas, these houses were accompanied by storage pits or caches (Giddings and Anderson 1986; McCartney 1980). These dwellings were likely occupied in the cold season. Most small hunting and fishing camps were not inhabited in the winter (Giddings 1952; Giddings and Anderson 1986). Although not many large villages have been excavated and thus contemporaneity of houses cannot be accurately assessed, the number of houses in a village could range from four to over 50 (Mason 2009b; Mason and Bowers 2009; Maxwell 1985; McCartney 1977; Savelle 1987). Qargich (pl.; qargi sing.)5or communal houses are sometimes found in villages (Habu and Savelle 1994; McCullough 1989; Norman and Friesen 2010; Patton and Savelle 2006; Savelle and Wenzel 2003; Spencer 1959). In the documentary record, umialgich (pl.; umialik sing.), family heads or “chiefs” (Burch 2006:66-67) are sometimes associated with a qargi. Although qargich are found in the archaeological record,
5 Spelling of Inuit words follows Alaskan Iñupiat orthography from Burch (2006)
45 it is uncertain if the same social structures of the umialik were also present. In the documentary record of many coastal whaling groups, umialgich were heads of large households that also could rally a whaling crew and the necessary technology for hunting whales (Bockstoce 1976; Burch 2006; Harritt 1995). In the documentary record, many groups called upon the umialgich and other leaders for decision-making and guidance; however, the autonomy of the nuclear family and the necessity for mobility to procure a variety of resources often negated the formalization of leadership in many villages (Burch 2006; Harritt 1995; Spencer 1959). Transitory leadership was most often found in the smaller villages, while more permanent leadership was found in settlements focused on whaling (Burch 2006). This style of informal, impermanent leadership may also reflect the early western Thule social structure in some small villages not focused on whaling, although this would need to be tested further.
Early Thule Inuit family structure and household composition have yet to be systematically studied in the archaeological record. In the documentary record, the typical cold-season dwelling held two or more nuclear families, with approximately eight people on average (Burch 2006; Jenness 1922). The early Thule archaeological record has similar house plans and sizes as those in the documentary record, which may indicate that a similar number of people inhabited the early Thule dwellings (Giddings and Anderson 1986; Lucier and VanStone 1995). Based on the warm-season dwellings locations, early Thule people likely dispersed seasonally into smaller family groups to hunt and fish. These families, likely consisting of a married man and women and any dependent children, probably functioned as the basic economic and social unit as recorded ethnographically (Burch 2006).
3.1.2 Subsistence Strategies
Early Thule Inuit generally procured and stored large amounts of meat and blubber for consumption over the winter. In order to store large amounts of food, some groups often employed a focused subsistence strategy for part of the year. This was often conducted in the form of communal hunts, an integral part of the Thule cultural tradition (Friesen 2012a; Grønnow et al. 2011; McCartney 1995; Savelle 1987; Savelle and McCartney 1988; Woodburn 1980). Although subsistence strategies likely do not completely determine the Thule way of life, they influence it to a great degree. The regional environment limited options available to Thule groups, although trade was a large part of Thule lifeways. Different groups procured different
46 animals for storage ranging from bowhead whales in areas in western and northern Alaska (e.g. Stanford 1976) and the central Canadian Arctic (e.g. Grier and Savelle 1994) to birds and seals in Kotzebue Sound (e.g. Moss and Bowers 2007). Despite their focus on a few specific resources for the majority of their diet, Thule Inuit also procured a broad range of secondary resources.
3.1.2.1 Marine Mammal Procurement Strategies
Although Thule Inuit are commonly characterized as “native whalers, par excellence” (McCartney 1995:840), they were excellent hunters of all mammals. Although marine mammals supplied food and fuel resources, they also influenced social structure, belief systems, and worldview (Borré 1994; Hill 2011). Across the early Thule world, sea mammals were hunted at open water in the spring, summer, and fall, and their resources were stored for winter use. Bowhead and gray whales were the largest of these, but belugas, narwhals, walrus, bearded seals, and small seals were procured depending on the ecology of the territory (Friesen and Arnold 1995a; 1995b; Lucier and VanStone 1995; McCartney and Savelle 1993; McGhee 1974; Morrison 1983b; Savelle 1995). Both whaling and sealing are described briefly in order to contextualize two common aspects of Thule subsistence.
Intensive hunting of large baleen whales is often seen as the hallmark of Thule Inuit culture. It was first highlighted by Mathiassen (1927b) and since then has often been a marker of distinction between Thule and earlier cultures such as Birnirk (Bockstoce 1976; Mathiassen 1927b; Savelle and McCartney 1994; Stanford 1976). Most often, the largest Thule settlements are found in areas with access to these large baleen whales, such as at Point Barrow, Point Hope, and Cape Prince of Wales in Alaska and around Somerset Island in the central Canadian Arctic. The practice of hunting whales impacts Thule settlement patterns and social organization (Friesen 1999; Grier and Savelle 1994; McCartney and Savelle 1993; Whitridge 1999a). Based on the documentary and archaeological evidence, bowhead whale hunting was most often done in umiaks (large skin boats) with six- to nine-man crews using harpoons, seal skin floats, and lances (Bockstoce 1976; Maxwell 1985; McCartney 1980; 1995; McCartney and Savelle 1985; 1993; Savelle 2005; Savelle and McCartney 1990; 1999; 2002; 2003; Whitridge 2002b). Whales were towed ashore, flensed, and divided among the whaling crew according to established social hierarchy (Bockstoce 1976; Savelle and Wenzel 2003; Whitridge 2002b). Hunters selected smaller whales, specifically yearlings, likely to avoid the danger in hunting and the difficulty in
47 processing larger whales or calves still by their mother’s side (McCartney 1980; 1995; Savelle and McCartney 1994; 2003).
Although whaling is often associated with the hunting of large baleen whales, most notably bowhead and gray whales, smaller whales such as the beluga and narwhal were also hunted (McCartney 1995; Ray 1975a; Savelle 1995). Beluga whales were hunted in different locations along the coasts of Siberia, Alaska, and northwest Canada (Arnold 1994a; Betts and Friesen 2004; 2006; Friesen 1995; 1999; Friesen and Arnold 1995b; 1995a; Friesen and Morrison 2002; McGhee 1974). Ethnographically, belugas were caught in nets, harpooned at the ice edge, from kayaks, from wooden platforms, or at openings in the sea ice (Boas 1888; Kleivan 1984; Nelson 1899); the antiquity of these methods has not been systematically examined but a variety of these techniques were probably used in the early Thule period. Like their larger counterparts, the hunting of smaller whales likely influenced subsistence-settlement systems, social organization, interaction, exchange, interpersonal relationships, taboos, rituals, singing, and other intangible aspects of Inuit society and social characteristics of some sites in the prehistoric western Arctic (Betts 2005; 2007; Betts and Friesen 2004; 2006; Friesen 2013a; Lantis 1938; McCartney 1980; McGhee 1974; Savelle 1995).
Despite the visibility of whale hunting in the documentary record and much of the archaeological record, for most Thule coastal winter sites, small seal remains are the most abundant (Giddings and Anderson 1986; Jacobs and Stenton 1985; McCullough 1989; Mohl 1979; Morrison 1983a; Park 1989; Rick 1980; Schledermann 1975a; Staab 1979; Stenton 1983). The ubiquitous presence of seals indicates that seal hunting was an integral part of Thule lives, especially at sites where whaling was an occasional or unavailable economic strategy. Morrison (1983b) offers a summary of Thule seal hunting based on faunal analyses from the Clachan site, where small seal (ringed seal) remains constitute an overwhelming percentage of the assemblage (89.7% by number of identified specimens). Such a high percentage is not unusual in Thule assemblages (Darwent and Foin 2010; Mathiassen 1927b; McCullough 1989; Morrison 1983a; Park 1989; Rick 1980; Schledermann 1975b; Stanford 1976), and therefore sealing warrants more research than it has historically been given. Where studied, Thule sealing techniques are often different from those of their historic descendants. Breathing-hole sealing, done in the winter with toggling harpoon heads through the sea ice, seems to be limited or absent in the early Thule record (Morrison 1983a; 1983b). Most faunal analyses indicate a mainly open-water hunting strategy
48 during the spring, summer, and fall (Gotfredsen 2010; Henshaw 2003; Park 1989; 1999; McCullough 1989; Mohl 1979; Morrison 1983a; 1983b; Moody and Hodgetts 2013; Stenton 1983; 1987; Whitridge 1992; Woollett 2010; Woollett et al. 2000). Open-water sealing would have used similar techniques and technologies to that of open-water large sea mammal hunting (Steensby 1916). The breathing-hole complex may have developed as a response to stable ice conditions in the central Canadian Arctic (Morrison 1983a; 1983b). If this is the case, breathing- hole hunting in the western Arctic may have only entered the Thule Inuit toolkit in late Thule periods; however, little work has been done on coastal subsistence in the early Thule period in Alaska to compare it to the later periods. Historic groups did use breathing-hole hunting techniques and technology in some areas at certain times, but not to the extent that was found in the central Canadian Arctic. Without the ability to pursue seals at breathing holes in fast ice, early Thule groups would have needed to rely on stored foods to a greater extent than their ethnographic descendants.
3.1.2.2 Terrestrial Procurement Strategies
For the majority of archaeologically identified Thule groups, subsistence strategies were focused on the coast. However, like many of their ethnographic descendants, Thule Inuit also procured terrestrial resources in the form of large and small land mammals, birds, and freshwater fish. These resources were either obtained seasonally or traded, with evidence of some Thule groups relying heavily on terrestrial resources for most of the year (Friesen 2013c; Morrison 1997c; Moss and Bowers 2007; Sheehan 1995; Spencer 1959; Stewart 2005)
Caribou, an important primary resource for Thule in some areas and an integral secondary resource in other areas (Burch 2012; Coltrain et al. 2004; Sheehan 1995), were likely hunted in a variety of ways, from individual snares to communal drives (Friesen 2013c). Evidence of caribou drive systems is found throughout the Arctic (Binford 1978a; Brink 2005; Friesen 2013c; Grønnow 1986), although only a few have been mapped or described. Thule Inuit likely used bows and arrows for hunting caribou from the ground and lances from kayaks (Maxwell 1985). Caribou were an integral part of the ability of Thule Inuit to hunt in the winter, as caribou skins provided lightweight, durable, and warm protection from the elements, while bone, antler, and sinew were excellent materials for tool and clothing manufacture (Burch 2012; Morrison 1997b; Sabo and Jacobs 1980). Where available, muskoxen were also hunted.
49
Documentary accounts indicate that birds, foxes, hares, rodents, freshwater fish, and other small game were hunted in a multitude of ways. Artefacts from Thule assemblages indicate that technologies such as bolas for bird hunting, snares and traps for small terrestrial mammals or birds, and hooks, leisters, and lures for fish were all used to varying degrees in the early Thule period (Maxwell 1985; Moss and Bowers 2007; Stewart 2005; Whitridge 2001) When subsistence techniques are discussed, most interpretations of Thule assemblages use documentary sources to a varying degree to reconstruct terrestrial subsistence strategies (Moss and Bowers 2007).
The discussion above gives a brief overview of the different types of resources acquired by early Thule and touches on some of the diverse techniques used for procuring these resources. Most early Thule Inuit used most if not all of these techniques at some time in the year to acquire a variety of resources; the varying degrees to which the resources made up part of the diet, the time spent acquiring different resources, and the scheduling of procurement events means that there are a myriad of possibilities for yearly subsistence strategies.
3.1.3 Processing Strategies
Butchery, preparation, consumption, and discard have been extensively studied in the Arctic through faunal remains and the material record. Animal processing was likely an integral part of Thule lifeways, since much of the procured food would have to be stored for long periods of time. Processing and sharing strategies create and maintain the worldview of many modern Inuit groups (Omura 2002, 2013). This worldview may be an underlying aspect of Inuit culture that dates back to the early Thule period; however, this interpretation has yet to be tested. Studies of Thule Inuit processing (butchery, storage, etc.) are outlined in this section.
3.1.3.1 Butchering
Food and architectural utility indices have been developed and used to varying degrees of success in the Arctic. These indices produce ideal expectations against which archaeological assemblages are tested. Binford’s (1978a; 1984) ethnoarchaeological work on caribou butchery in the Arctic informs most caribou analyses. These indices have been adjusted and altered to produce simpler food utility (Metcalfe and Jones 1988), meat drying (Friesen 2001), and marrow and grease indices (Morin 2007). Indices have also been created for studying seal butchery (Diab
50
1998; Lyman et al. 1992). Archaeological evidence of whale processing is rare (but see Betts and Friesen 2013), although a utility index for whales has been produced for meat (Savelle and Friesen 1996) and architecture (Savelle 1997).
One of the most important ethnographic, and it is assumed archaeological, aspects of butchery is the sharing process (Boas 1888; Balikci 1970; Binford 1984; Borré 1994; Burch 1980; 1988a; 2006; Damas 1972; Diab 1998; Jenness 1922; Omura 2002; 2013; Rasmussen 1931; 1976; Searles 2002; Stefansson 1913; Spencer 1959); despite attempts to identify sharing, it remains a near-invisible pattern in the archaeological record. Butchering and sharing large sea mammals such as whales, walrus, and bearded seals, was performed according to specific practices. In northwest Alaska, the general sharing practice was to give everyone involved in helping the crew obtain the whale a portion of the whale (Burch 1988a; 2006; Spencer 1959; Whitridge 2002b). Caribou and other medium and small sized animals were often kept by the family of whoever killed them; however, the products of communal hunts were often divided among participants or the entire community (Birket-Smith 1959; Burch 1988a; Burch 2006; Burch 2013). In many cases, seals were divided and exchanged with particular people, such as kin or specific sharing partners (Damas 1972).
3.1.3.2 Storage
Thule Inuit likely stored animal remains in a variety of ways including freezing, drying, and fermenting. Research into storage practices in the Thule archaeological record is limited, but has been done for seals (Park 1999), caribou (Friesen 2001; Friesen and Stewart 2013), and fish (Knudson and Frink 2011; Knudson and Frink 2010a). These studies look at processing and drying through faunal remains and soil chemistry, although more work has been completed on late Thule and early historic sites than at early Thule sites (Frink 2007; Knudson et al. 2004; Knudson and Frink 2010a; 2010b; 2011). The presence of hide, meat, and blubber in some Thule sites (e.g. Dekin and Kilmarx 1990) indicates that meat was often frozen in excavated caches. Surface stone caches and excavated pits have been identified archaeologically as ways Thule Inuit stored their food (Frink 2007; Giddings and Anderson 1986; Maxwell 1985; McCartney 1980). Food, materials, and tools were also likely stored on raised racks around the site in order to prevent dogs and other carnivores from eating them; however, these are difficult to identify and have undergone little archaeological analysis (Newell 1990; Polglase 1990). Fermentation
51 was also an important storage technique recorded in the ethnographic and ethnohistoric accounts. This type of storage is almost invisible archaeologically as food was stored in perishable containers such as seal skin pokes or excavated soil caches; most interpretation of Thule oil and meat storage relies heavily on documentary records.
3.1.3.3 Consumption
Arctic people carefully prepared and cooked foods for some meals (Harry and Frink 2009). In the documentary record, and it is presumed also archaeologically in many cases, cooking was done using direct fire-boiling (a pot set directly on the fire), indirect/stone-boiling, and, rarely, grilling (Burch 2006; Harry and Frink 2009). Soapstone vessels were the preferred cooking utensil in the east, but where soapstone was not abundant pottery vessels were used (Ackerman 1984; Harry and Frink 2009). In typical ethnographic contexts, Inuit prepared one hot meal a day, snacking on raw frozen or dried food throughout the day (Burch 2006). Despite evidence for burnt bone in nearly all sites, cooking and food consumption is another understudied aspect of early Thule lifeways.
3.1.3.4 Discard
Although middens are well excavated in the Arctic, discard is rarely the subject of investigation. Binford (1978a) used ethnoarchaeological data to try to understand discard patterns, which could then possibly be traced back into the archaeological record. From Binford’s analyses and other ethnographic and ethnohistoric accounts (Jenness 1922; Stefansson 1914a; Turner 1894), it can be generalized that Inuit often discarded bones and other debris in specific locations, although weather and dogs potentially affected these deposits. Areas for specific discard included piles of bones for marrow processing, separate storage of bones for grease production, undifferentiated house debris, and final discard (e.g. Jenness 1922). These areas may be visible archaeologically, but excavations will need to be conducted in the areas outside of houses in order to test Thule Inuit discard patterns against the documentary record.
3.1.4 Dwellings
Houses are the nexus of social, economic, and domestic life; for ethnographically recorded Inuit and Iñupiat, they were the physical centres of daily life (Burch 2006; Whitridge 2004). In the Thule Inuit archaeological record, houses are extremely well studied due to their visibility on the
52 landscape, their excellent preservation, and their abundance of artefacts and faunal specimens. These houses show a wide variability geographically, temporally, and regionally; they varied with season, construction materials, degree of mobility, cultural traditions, and specific individual requirements (Burch 2006). The similarities in Thule Inuit dwellings are summarized below; however, it should be noted that much more variability exists than is presented in this overview.
3.1.4.1 Warm-Season Dwellings
Early Thule Inuit warm-season dwellings were likely skin tents with wooden poles. Often, the only archaeological evidence is the stone used to hold down the skins (McCartney 1977; Savelle 1987). These offer very little preserved archaeological remains due to the limited occupation times and high degree of exposure to weathering processes on the surface.
Ethnographically, intermediate-season dwellings called qarmat (qarmaq sing.) were used in the fall (Mathiassen 1927b:133). The structure of these dwellings varied widely, but they were often slightly excavated, with a heavier ring of stones or sod, covered by a skin tent roof (Park 1988). Archaeologically, these dwellings have been dated mostly from late Thule settlements (Park 1988; LeMoine and Darwent 2010; Morrison 1983b; Sabo and Jacobs 1980).
3.1.4.2 Cold-Season Dwellings
Cold-season dwellings are typically deep semi-subterranean features with entrance tunnels; based on documentary records, they were usually occupied in the fall and winter when people gathered on the coasts (Burch 2006). Early Thule houses generally consisted of one main room, although the floor plan could vary greatly. One of the most elaborate architectural arrangements recorded is the eight-room house at Qimiarzuq (Jabbertown; Mason 2009b; Mason and Bowers 2009). Based on documentary records, winter houses were built to reduce the volume of air inside the dwelling so roofs at the edges were often lower, floor area was small, and houses were likely only built to house a certain, restricted number of people. The floor was likely kept cool for storing meat, with lamps kept above floor level (Arnold and Hart 1992). People slept, rested, and worked on raised platforms in the warmer air (Arnold and Hart 1992). Houses were covered in sod and snow, and benches were lined with willow or heather (Arnold and Hart 1992). These semi-subterranean dwellings were likely not just physical structures, but places and spaces where
53 memory, society, and tradition were created and maintained (Whitridge 2004). This can be seen archeologically when whalebone was used in house construction. Levy and Dawson (2009) suggest that the use of whalebone was symbolic, as it actually reduced the structural soundness of the house. Houses have been the source for much Thule material, and as such, they have been used to evaluate cultural and historical relationships with other arctic societies, to examine temporal and spatial diversity, to explore change in subsistence-settlement systems, and to look at differences and similarities in social relations, organization, and ideology (Dawson 2001; Maxwell 1981; 1985; McCartney 1977; McCullough 1989; McGhee 1972; 1984b; Park 1989; Patton and Savelle 2006; Savelle 1987; Savelle and Habu 2004; Schledermann 1976b; Whitridge 1999a; 2002a; 2004). Spatial patterning has also been examined for early Thule houses in both the Mackenzie Delta (Friesen and Betts 2006) and northeast Greenland (Darwent and Foin 2010). These studies of semi-subterranean dwellings across the North American Arctic contextualize this current research on faunal spatial patterning.
3.1.5 Thule Way of Life Summary
This overview of the Thule way of life was constructed based on analogical interpretations of the archaeological record based on information from a variety of sources. Both the analogical sources and archaeological record have been generalized and summarized to give an overview of early Thule life. The final section of this chapter will work from the documentary record to outline probable broad-scale activities in the house. Instead of using a generalized analogical interpretation from the broader documentary record, the activities for creating expectations are based on the direct-historical approach with details taken from the ethnographies and ethnohistories from northwest Alaska. This final section outlines these activities, which will be used in Chapter 5 to detail the specific expectations of faunal material in each area of the house. First there is a brief description of the documentary records used in this research.
3.2 Documentary Records for the Seward Peninsula and Kotzebue Sound
One of the advantages of working in the Arctic is that detailed ethnographic and ethnohistoric accounts from the contact period and the present day were, and continue to be, recorded for much of the area. By using analogy in its many forms to help interpret the archaeological record, archaeologists can construct a more detailed understanding of the past. However, these analogies
54 must be applied critically and tested for perceived similarities and differences. One way to approach this is to use the direct-historical record for a particular region and create expectations to test against the archaeological record of that region. Both confirmation and rejection of an expectation must be explored further and ‘boot-strapped’ (Wylie 2002b) using multiple lines of evidence.
The direct-historical approach is a form of analogical reasoning built around time and the notion of heritable continuity between ancestor and descendant (Lyman and O’Brien 2001). This approach allows particular historical interpretations to be made in order to identify ethnic affiliation of archaeological material (Sapir 1916; Wedel 1938; Willey 1953), to build relative chronologies (Kroeber 1916; Nelson 1916; Wissler 1917), and to understand human behaviour, which is the most frequent application (Lyman and O’Brien 2001; Steward 1942; Trigger 2006). The current study uses the direct-historical approach to construct expectations in order to test the archaeological record to strengthen the use of the documentary record in interpreting the use of space within a house in the early Thule record.
In order to apply the direct-historical approach in outlining expectations, cultural relationships and similarities between the present group and past groups must be highlighted so that further similarities can be proposed. As was summarized in Chapter 2, Inuit, Iñupiat, Inuvialuit, and Yupik are genetic, linguistic, and cultural descendants of Thule Inuit (Bockstoce 1979; Burch 1980; Ford 1959; Giddings 1952; Giddings and Anderson 1986; Stanford 1976). In the case of the Seward Peninsula, there is continuous occupation of the area by culturally similar groups since the eleventh century CE (Bockstoce 1979; Giddings 1952; Giddings and Anderson 1986; Harritt 1994; Schaaf 1988; 1995). Burch (1980) describes the northern Seward Peninsula and Kotzebue Sound region as exhibiting regional long-term cultural continuity with extreme short- term social and demographic fluctuations. Modern inhabitants recall their ancestors visiting and living at Cape Espenberg in the historic period, with oral history accounts of this site extending slightly further back into prehistory (Burch 1980; Burch 1984; Burch 1998a; Burch 2006; Burch 2012; Burch 2013; Ray 1975a; 1983; Sobelman 1985). Although the historic period brought about great changes in the way of life of northwest Alaskan Iñupiat, the general lifeways can be traced back for hundreds of years; understanding the specifics of these similarities and exploring the differences is a goal of this part of this research.
55
Since cultural descent has been established, similarities in material culture between the early Thule archaeology and the historic period documentary must be examined to determine similarities in behaviours and practices (Nicholson 1976; Stahl 1994; Trigger 1995; 2006). This must be done carefully because particular behaviours can be manifest in the material culture record in different ways over time; meaning and form can and do change independently (Trigger 2006). For this research, the documentary records of northwest Alaska, specifically the northern Seward Peninsula and southern Kotzebue Sound are used to establish expectations to explore the differences and similarities in the archaeological record. These expectations are the starting point for a establishing a contextual synthesis that uses multiple lines of evidence to strengthen the use of analogical reasoning in constructing the past (Wylie 2002b). The next section of this chapter outlines the methods and highlights some critiques of the documentary records used for constructing the activities in semi-subterranean dwellings from northwest Alaska described below.
3.2.1 Documentary Sources
Ernest S. Burch, Jr., Dorothy Jean Ray, and a number of other historic sources (e.g. Beechey 1831; VanStone 1960; 1977; von Kotzebue 1821) documented the region around Cape Espenberg. Dorothy Jean Ray’s research on the area mostly focused on art and interregional interaction in the Bering Strait region to the south. Some historic documents, such as those of Otto von Kotzebue (1821) recorded information about the groups in the area, but rarely discussed the dwellings of the inhabitants of the northern Seward Peninsula. Therefore, most of the following description of the creation of the documentary record will focus on the work of Burch. Additionally, the following section’s descriptions of semi-subterranean dwellings and their associated activities, as well as the expectations constructed in Chapter 5, are mostly derived from the work of Burch. However, some sources offer or corroborate the descriptions of Burch, so all published material is described in this review.
Otto von Kotzebue (1821) was the first known Westerner to visit the region. Von Kotzebue’s observations of the Iñupiat were brief, with only a few visits ashore in 1816. Some of the first descriptions of houses come from von Kotzebue, although these descriptions only refer to the architecture and orientation. Ludovik Choris, the artist of the expedition, also made notes
56
(VanStone 1960) about the inhabitants, their materials, and their physique. However, he does not draw or describe in detail the dwellings of the area.
Both F. W. Beechey (1831) in 1826 and 1827 and Aleksandr Kashevarov (VanStone 1977) in 1838 observed the people of the region and recorded much more information than the previous explorations. Kashevarov was the first to bring an interpreter during his exploration. Neither explorer commented on the dwellings of the people, with most of their descriptions focused on the Iñupiaq appearance, mannerisms, and materials. John Simpson (1875) observed and recorded ethnographic data during the Franklin search expeditions of 1848-1855, and learned to speak the Iñupiaq language. Much of Burch’s work references Simpson’s original descriptions, including descriptions of houses come from Hotham Inlet in Kotzebue Sound (Simpson 1875:255–257).
Much of the work done by Dorothy Jean Ray focuses on historical sources from the Russians and Americans and accounts written by explorers, travelers, teachers, and missionaries (Ray 1975a; 1975b; 1983). She brings these historical sources together with her own observations made during extensive fieldwork. Although there is some description of houses in Ray’s observations of the Iñupiat, these observations are limited to shape, orientation, and architecture.
Ernest S. Burch, Jr. worked in northwest Alaska from 1960 to 1999, interviewing people and investigating the history of the region. Although his region of study overlapped slightly with the ethnographic work of Nicholas Gubser (1965), John Murdoch (1892), and Robert Spencer (1959) to the north and the work of E. W. Nelson (1899) and Dorothy Jean Ray (1964; 1967; 1975a; 1975b; 1984) to the south, most of his work investigated ethnographically unknown territory in between. Burch (2006:2) defines the early contact period as 1800-1848 CE. This is the earliest time for which both oral and documentary sources provide information, but is also a time where Western influence was minimal, but present. Western people were in the region, but were interested mainly in exploration. Some trade vessels did visit, but Burch argues that neither traders nor the explorers greatly impacted the Iñupiat way of life. There is a definite end to this period at 1848 CE when the first American whaling ship arrived in north Alaskan waters and when the first of many British navy ships arrived as part of the massive search for the missing Sir John Franklin expedition. Whalers increased in numbers dramatically over the next few years and by the early 1880s had nearly exterminated bowhead whales and walruses. They also brought with them independent trading vessels that established the fur trade in the area. At least
57 one of the ships and crews searching for the Franklin expedition stayed in the study area over the winter, with many vessels present in the summer for seven years. The first epidemic recorded was in the late fall of 1851. Therefore, this early time from 1800 to 1848 was minimally influenced by Westerners.
Burch uses both documentary and oral sources to describe and analyze the people of northwest Alaska (2006:11–15). The documentary sources consist of some of the sources described above, as well as other published and unpublished reports of explorers in the 19th century. The oral sources use the stories and accounts about the 19th century, recorded and observed by Iñupiat oral historians in the 20th century. Burch relies both on his own observations and interviews and the interviews and information collected by others. Although Vilhajalmur Stefansson (1914a; 1944) did not study the region, a number of the informants he lived with between 1908 and 1918 were from the Kotzebue Sound area. Knud Rasmussen (1925; 1933), a native of Greenland, was able to communicate with the Iñupiat in northern Alaska at the end of his sled journey across the northern part of the continent in 1924. Finally, Edward Curtis (1930) visited western Alaska in the summer of 1927 to photograph traditional Native American life. He collected myths, legends, customs, and commented on the conditions in communities. Burch (2006:14) also brought together the more recent unpublished ethnographic works of Froelich Rainey, Don Charles Foote, Albert Heinrich, and Charles Lucier. Each of these researchers worked in the study region during the mid-twentieth century.
Burch’s (1998a; 2005; 2006) information also comes from Iñupiat historians, interviewed between 1960 and 1999, who learned about traditional ways from their parents and grandparents. He conducted participant-observation fieldwork and formal interviews in the community of Kivalina, and then conducted further interviews in all the communities in the study area. Burch (2006:15) employs an ethnohistoric method called ‘upstreaming’ to bring these disparate documentary and oral sources together. Similar to the archaeological method of the direct- historical approach, this method interprets the information about the past by comparing it to the patterns of the present (Burch 2006:15). This method does not just use patterns of the present to describe the past, but compares the earlier patterns to the later ones via a direct sequence of events; it is “going from the known present to the unknown past” (Fenton 1962:12). Using this method, Burch curated information about a wide variety of topics on the early historic period, including the ecological adaptations, subsistence patterns, and economic practices of the Iñupiat
58 nations in the area (Burch 1998a), their boundaries and borders (Burch 1998b), the mechanisms through which these boundaries and interregional relationships were maintained and transgressed (Burch 2005), and finally a detailed description of how these groups were organized and functioned internally (Burch 2006). It is this last volume that most fully describes house form, function, and everyday activities.
3.2.2 Limitations of the Northwest Alaskan Documentary Record
General limitations and critiques of the use of ethnohistoric and ethnographic sources were discussed in the introduction; however, there are specific limitations on the use of these particular sources in interpreting the early Thule site at Cape Espenberg. Most significantly, there are few reliable sources that specifically describe the area of Cape Espenberg. Only von Kotzebue (1821) met with people of the Cape Espenberg region, and none of his notes describe the house features for these people. Later investigations did not acquire information from the Cape Espenberg area, but rather the broader region of Shishmaref or Deering. During the time of Burch’s investigations, the village at Cape Espenberg had been abandoned (Burch 1998a). Currently, people from Shishmaref use the region of Cape Espenberg for warm-season subsistence pursuits, but no one resides there in the winter (Sobelman 1985; Wisniewski 2005). Cape Espenberg represents a blank spot in the documentary record. Therefore, the house descriptions and expectations are derived from a broader region of northern Seward Peninsula and southern Kotzebue Sound.
A second critique of using the documentary record to interpret archaeological remains is that historians, explorers, ethnographers, and other researchers were not recording objects or behaviours that directly generate the archaeological record. They rarely discussed discarded or broken artefacts, and even more rarely did they make observations on the use, reuse, or discard of animal remains. Burch (2006) presents one of the few sources of information about use and discard behaviours for animal remains outside of ritual or ideological purposes; even here, much of the information about bones must be inferred. For example, only animal parts not bone themselves are recorded as being brought back to the site; the specific bones must then be extrapolated from these parts. Site formation processes, such as trampling, dog gnawing, or burning must also be inferred from descriptions of other activities.
59
A third critique for the Alaskan documentary record is that the oral and written evidence that forms the documentary record is fragmentary at best. In Mason’s (2012:73) analysis comparing Burch’s reconstruction of conflict to the archaeological evidence in northwest Alaska, he highlights that “opportunism and serendipity play a substantial role in the conduct and transmission of oral history and anthropology”. Thus, the reconstruction of early contact life in northwest Alaska will never be complete. Additionally, this documentary record may not only reflect the early contact period, but may be a palimpsest of many generations. Mason’s (2012) archaeological analysis of conflict in Alaska shows that the most intense period of warfare lasted from 1400 to 1700 CE. However, the reconstruction from the oral and written histories of the records compiled by Burch (2005) indicates that conflict was endemic between 1800 and 1820 CE. The discrepancy between the oral accounts and the archaeological evidence for the pervasiveness of conflict in northwest Alaska in the early contact period is an important reminder that the oral histories are sometimes a blending of many stories, activities, and practices over time (Mason 2012:86).
Finally, as with all ethnographic observations, historical recordings, and documentary or oral sources, there are more general critiques. These include the major critiques described in Chapter 1. Specifically for the direct-historical method, ethnographies and ethnohistories must be understood both as a biased recording of the cultures they are describing and as historical artefacts of the time in which they were written, which includes the circumstance of the researcher (Binford 1967; Dunnell 1978; Hill 1998; Ravn 2011; Salmon 1982). For the people around Kotzebue Sound, this includes acknowledging the ways in which the research was conducted, what period it represents, and who was doing the data collection. Other factors not explicitly outlined in these reconstructions are gendered knowledge, restricted access, and unconscious mirroring of previous recordings. All of these factors may bias the documentary records. Where applicable, the limitations of the documentary record are highlighted.
Although analogies are intended to identify similarities, they cannot be applied without detailing the changes that occurred between the deposition of the archaeological record and the recording of the documentary record (Dunnell 1978; 1992; Hill 1998; Stahl 1993). For the Arctic, this change includes sustained contact with whalers, traders, explorers, scientists, and missionaries (Bockstoce 2009; Burch 2013; Fienup-Riordan 1990; Ray 1975a; 1983). As well, historical change due to environment, climate, interaction, and epidemic disease all affect the documentary
60 record, some of which have been analyzed by Ray (1975a; 1983) and Burch (2005; 2006; 2010; 2012). Since both source context and historical change must be accounted for when applying the direct-historical approach to archaeological interpretation, a potentially more useful application of the direct-historical approach is to create expectations that can then be tested using multiple lines of evidence (Binford 1967; Binford 1968; Shelley 1999; Stahl 1993; Wylie 1985). This allows archaeologists to explore similarities and differences between the material of the archaeological and documentary records. Chapter 5 constructs expectations against which the archaeological record is tested in order to explore if and how the use of space has changed by determining how closely the archaeological record is reflected in the documentary record. Before the spatial patterning of activities in a dwelling can be tested, these activities must be outlined.
3.3 Northwest Alaskan House Histories
For the final section of this chapter, the known documentary history of a semi-subterranean dwelling is reconstructed based on the writings of Burch (1988b; 1980; 1998a; 2005; 2006; 2012), Ray (1975a; 1983), and others (e.g. VanStone 1960; 1977; von Kotzebue 1821) who recorded or reconstructed the “ethnographic present” of the nineteenth century. These reconstructions focus on the groups of the northern Seward Peninsula. However, some of the activities and processes were not recorded in the specific sources that describe the northern Seward Peninsula, and so some of the descriptions are based on the wider northwest Alaskan documentary record; these will be identified where applicable. Activities are divided into tasks, with specific reference to their location; however, most of these tasks overlap both temporally and spatially and are likely not performed in isolation from other activities. Considering the tasks both as a set and as individual actions is important in understanding the recovered faunal assemblage.
3.3.1 House Construction
The basic pattern of house plans around the Cape Espenberg region described in the documentary record is remarkably similar to the excavated early Thule dwellings. The earliest description of a semi-subterranean dwelling by early Russian explorers in the Kotzebue Sound and northern Seward Peninsula region was by von Kotzebue in 1816; his description of a house on Shishmaref Island on the northwest edge of the Seward Peninsula, southwest of Cape Espenberg, describes a floor plan that is extremely similar to the floor plans for typical early
61
Thule houses. The house had two rooms and an entrance passage. The first room was a storage room, ten feet (3 metres) long by seven feet (2 metres) wide (von Kotzebue 1821:200). It had a wood roof and wood walls and stored blubber and tools in a pit along one long side (von Kotzebue 1821:200). The entrance tunnel was dug into the earth and ended in a small antechamber (von Kotzebue 1821:200). The entrance to the main room was through a 1.5 foot (0.5 metre) hole in a wooden plank (von Kotzebue 1821:200). The main room was ten feet (3 metres) square and had a raised back platform and small side shelves; it was made of smoothly finished driftwood on the floors, walls, and roof with a small bladder-covered window at the apex of the roof (von Kotzebue 1821:200–201). Kotzebue (1821:201) records that all the houses in Shismaref were built on this floor plan of an entrance/storage room, tunnel, and main room, except one that had two extra side rooms to house a larger family. Choris also records the same house style in his notes on the subject (VanStone 1960:147). Burch (2006:216) describes the houses in the Cape Espenberg region as semi-subterranean houses that consisted of one main room, built on a rectangular plan, with sitting and sleeping platforms opposite the entrance; the entrance tunnel often had a number of small rooms or alcoves for storage or cooking. Lee and Reinhardt (2003:87) describe houses on the Seward Peninsula originally described in Fair et al. (1996) and Gerlach (1996) from the 1920s that follow the floor plan of a rectangular main room, entrance passage, and large entrance/storage alcove.
Archaeologically, a similar floor plan based on one main room, an entrance tunnel, and a storage room is found in the early Thule period in the Cape Espenberg and northern Seward Peninsula region. The longevity of the floor plan in the region from early Thule to the historic period is likely due to the cultural continuity through time in this region. Although dwellings were constructed based on other floor plans by the historic period, none were as frequently used in the northern Seward Peninsula as the one described above (Lee and Reinhardt 2003:87). With a similar floor plan between the source (documentary records) and subject (archaeological record), the use and activities performed in the internal space can be tested to assess similarities and differences.
In both the archaeological and documentary records, spaces within the house were physically delimitated during the initial building phase. Ethnohistorically, semi-subterranean houses were built in the late summer and early fall when the permafrost had melted (Burch 2006:216). At this time, permafrost may still have been encountered, and so fires were built on the areas to loosen
62 the soil (Burch 2006:216). One of the main concerns when building a house was the ability to heat it; thus, houses were normally built as small as possible based on the number of people who would be living in them and were subsequently quite crowded (Burch 2006:98). Houses, once built, could be altered. There are no descriptions of houses being refurbished in the Cape Espenberg or larger northern Seward Peninsula region, but semi-subterranean houses were often refurbished with sod, wood, or whalebone at the start of the yearly occupation in many other regions (Lee and Reinhardt 2003:10, 14, 16, and 52). However, these reconstructions mostly added additional rooms or alcoves to the houses rather than altering the main room or previously built rooms. The original floor plan, and any subsequent changes, restricted the activities performed in the living space.
Animal remains were sometimes a part of the built domestic environment. Most notable is the variable use of whale bones as structural or symbolic elements (Lee and Reinhardt 2003; Patton and Savelle 2006). Other large mammal bones may have also been used as supports or to repair houses (Burch 2006). Other than these structural elements, the construction of space is likely to have been invisible in the faunal record.
3.3.2 House Activities
As people inhabit space, they incorporate memories, cultural and social ideals, and the physical and environmental context in which they live. On a more mundane note, the limited amount of space in a Thule cold-season dwelling likely mimicked the limitations of the Iñupiat dwellings recorded in documentary sources from the northern Seward Peninsula region; here, utilization of space was separated both spatially and temporally (Burch 2006:98). In most cases, only one, potentially two, activities could occur within the winter dwelling at the same time. Such an overarching constraint on activities necessitated storage of items not in use and removal of used or discarded items. Middens, discard piles, and storage places are archaeological evidence of these practices. The following sections describe in more detail the activities and tasks that occurred in winter dwellings based on the documentary sources.
3.3.2.1 Sleeping
In the northern Seward Peninsula region, members of a house slept on the raised back platform (Burch 2006:216; Lee and Reinhardt 2003:87). These platforms were mainly built of wood in
63 this region and were almost always raised in winter dwellings to keep people off the cold floor and in the warm air. Bedding consisted of grasses and furs, which were often shaken out prior to sleeping (Burch 2006:224). In large households, lower ranking family members would sleep on the floor or in the tunnel. The tunnel was also used as a sleeping location when couples desired privacy.
3.3.2.2 Consumption
Processing animals for consumption was different throughout the Iñupiat region as recorded in the documentary record, with very few descriptions of processing specifically for the northern Seward Peninsula region. The following descriptions of consumption processes are based on the ethnographies from the broader northwest Alaska region. Animals were first hunted, then butchered, often in multiple stages, then processed, shared, cooked, consumed, and discarded, although not always in this order. This larger process of consumption has been suggested to be one of the ways the Inuit in the central Arctic create and maintain their worldview (Omura 2002; 2013). Despite the importance of hunting and butchery to zooarchaeological remains, this review is most concerned with the activities that occur within the house. These included some primary butchery, secondary butchery, processing, distribution, cooking, and consumption.
Primary butchery refers to the practice of breaking carcasses down into smaller, transportable packages. Large animals such as whales, walrus, and bearded seals were almost universally butchered first at a kill or primary butchery site (Burch 2006:159–160, 164, 167–168). These large animals were typically butchered by a number of people, not all of whom were part of the hunt (Burch 2006:160,164). Smaller animals such as small seals, fox, or hare, were often brought back whole to the site as they were more easily transported (Burch 2006:172–176). Women took over management of most resources once they were killed; thus, retrieval of hunted animals, butchery, and processing was mostly done by women (Burch 2006:64–65, 81). Butchery of other resources, such as fish or birds, was dependent on the season and method of procurement. If large amounts of these small game were caught, such as in bird drives or fish runs, most of the processing would occur at the site of procurement; if these animals were caught individually or in few numbers, they were brought back to the village site and processed there (Burch 2006:146, 179).
64
Secondary butchery, where animals were processed for storage, exchange, or immediate consumption, mostly occurred at the village site and was completely controlled by women (Burch 2006:64–65, 81, 264, 310). Although much of the secondary butchery was often completed outside in the warm seasons where drying, freezing, or other types of storage occurred, winter processing necessitated that the meat be brought into the house to be thawed prior to secondary processing (Burch 2006:138–139).
Processing for storage, sharing, and consumption included drying, fermenting, and freezing. Once people brought meat into the household, either from the hunt or from distribution, it was stored or prepared for cooking and consumption (Burch 2006:255–256). Most of the preparation for cooking occurred in the hearth areas; Burch (2006:97, 98, 216, 225–226) states that cooking areas for the northern Seward Peninsula were often found off the tunnel, although other documentary records for the region make no mention of cooking areas. More broadly, the houses with one main room, entrance tunnel, and side rooms did have a separate kitchen area in the northwest Alaska documentary records (Lee and Reinhardt 2003:79, 80, 87). If meat was not cooked, it could be stored frozen or placed in containers, often seal skin pokes, and stored in seal oil. Many types of food, such as caribou meat, fox meat, or berries were stored in this manner which allowed food to be stored for long periods of time (Burch 2006:147, 214, 222).
Ethnographically in the northwest Alaskan region, one hot meal a day was prepared in the cold- season. Cooking was typically done by placing hot stones in water in order to boil the contents of the container that included meat, oil, blood, and vegetable content (Burch 2006:212, 226). Two or three cycles of heating stones and placing them in the containers were required for most cooking (Burch 2006:226). Men and teenage boys typically ate the hot meal in the qargi, while women and children ate it in the domestic dwelling (Burch 2006:54, 255–256). If men ate in the house, they were served first, while women and children ate second; this sequential patterning reduced some of the crowding in the dwellings (Burch 2006:98, 264). During the winter ceremony and visiting season, or at other time of celebration, people often gathered in the qargi for a meal, songs, ceremonies, and rituals (Burch 2006:264). At these times, visitors would often be welcomed into the domestic dwellings for meals. People sat where they could while eating, which was typically on the back platform (Burch 2006:97). Although consumption of frozen meat did occur, most meat was heated prior to eating. Raw, frozen, or dried meat was put out in the mornings by the woman of the household to be available throughout the day for people to eat
65
(Burch 2006:264). Meat from caribou, fish, or birds was also dried and consumed as quick meals, rehydrated into stews, or included as part of the main meal (Burch 2006:139, 146, 179, 181, 182, 264).
3.3.2.3 Cleaning
Discard of food remains or other types of trash is not fully explored in the documentary or archaeological record for either the specific area or the broader northwest Arctic region (Burch 2006:268). However, most ethnographers record neat and tidy dwelling spaces (Bockstoce 1977; Burch 2006; Murdoch 1892; Thornton 1931; von Kotzebue 1821). For the broader northwest Alaska region, household cleaning was regularly conducted using the wing of a large bird as a broom (Burch 2006:184, 268). Certain animal remains were discarded in specific ways in specific regions; bowhead whale skulls were often dumped in the ocean, sometimes followed by other whale bones that were not used for dwelling construction, caches or graves (Burch 2006:269). Beluga whale bones were sometimes burned, dumped in the sea, or taken to a site behind the beach depending on the group (Burch 2006:269). Burch (2006:269) mentions that garbage was disposed in a convenient nearby location, but that the volume of it must have been significant enough to warrant it being systematically removed. Dogs were often responsible for removing much of the waste, but there would be more animal remains than even the large dog teams of the documentary record could consume (Burch 2006:268–269). Most waste was tossed outside of the house and on top of the house (Burch 2006:269; Larsen and Rainey 1948).
3.3.2.4 Working
In the broader northwest Alaska region, most work was completed outside in the light; however, in the winter when there was little to no daylight outside, work such as tool making or clothing manufacture was completed in the house or qargi (Burch 2006:53–57). In either building, light was extremely important, with people working around the light source. Material for both tool making and clothing manufacture was often kept in alcoves or storage spaces in the main room and tunnel (Burch 2006:98, 216). Women also dried clothing, melted water, and thawed meat in the house over the stone lamps (Burch 2006:224–225).
66
3.3.2.5 Socializing
Winter villages were aggregations of people that often lived in small family units throughout the rest of the year; when people first gathered at their semi-permanent dwellings for the cold season, it was considered a time for celebration and visiting (Burch 1980; Burch 2006; Burch 2013). In communities with a qargi, these buildings were often the centre for socializing and ceremonies for all members of the community (Burch 2006:54, 105–106, 220). However, domestic dwellings were also an important location for socializing as people would drop into houses to eat, interact, and chat (Burch 1980; Burch 2006; Burch 2013). Domestic dwellings were the central location for women to work, socialize, and eat in the winter months (Burch 1980; Burch 2006:64). Socializing is likely to be mostly invisible in the faunal record.
3.3.3 Moving Out
Semi-subterranean house habitation ended in the spring when temperatures increased and frozen walls and floors melted; at this time, people started to disperse to warm season locations to acquire resources. In the summer, houses were uninhabitable for both humans and animals as they often filled with water. Throughout northwest Alaska, these winter dwellings were almost always re-occupied over the course of a decade or generations (Lee and Reinhardt 2003:123); in the Seward Peninsula region, houses were often abandoned (Ray 1975a:173). Moving out activities were therefore affected by many factors, such as plans to return, decisions to build a new house, and season of return. People took with them things that were needed for the next season, but often left stored in caches the tools and objects they would not need until returning the next cold season; these villages of semi-subterranean houses were a family’s base of operations (Burch 2006; Ray 1964; 1983). If the house was to be truly abandoned, much of the house structure and possessions were likely removed and stored or transported to the new house site (Lee and Reinhardt 2003:113). This also occurred in the re-use or re-building phase when people return to the site.
3.3.4 Re-Use/Re-Building
Houses were reused multiple times (Lee and Reinhardt 2003:123). When people returned to a house, they cleaned out the summer debris and often chipped away the ice lens that filled the house. Families would replace any structural elements in need of fixing, and would often acquire new sods for the exterior (Burch 2006). Tools, skins, and other objects were removed from their
67 summer storage and racks and winter storage facilities were constructed. It is at this time that re- building of additions or closing off of rooms occurred; this would re-order activity areas. One of the only times they were not refurbished for winter use was when a death occurred in the house (Ray 1964; 1983); after this, people were not allowed to enter the house (Burch 2006; Lee and Reinhardt 2003:113–114, 154–155). However, archaeological excavations on the Alaskan North Slope have shown that people sometimes did re-enter the house after a death in order to take useful materials and objects (Hall and Fullerton 1990a). At Utqiagvik in Barrow, excavation revealed a house had collapsed and killed a family; here, the remaining sets of clothing and utensils were fewer than should be expected for a family, indicating that people had taken some of these materials from the entrance tunnel (Hall and Fullerton 1990a). At other times, houses were abandoned because of dampness or inundations in the spring (Ray 1964; 1983). These houses could be used as sources of wood for fires and construction (Lee and Reinhardt 2003:123).
3.3.5 Abandonment and Burial
Sometimes when houses were abandoned for a number of years, other families could re-occupy the abandoned house. Abandoned houses were sometimes used as midden locations or animal dens, although the water or ice lens in houses often made them unsuitable for either use. People often scavenged abandoned houses for structural materials (Lee and Reinhardt 2003:123). When these structural elements had been removed or had deteriortated, houses underwent quick collapse, burial and incorporation into the permafrost. These collapsed houses left slight depressions in the ground that were sometimes used as midden areas or as locations for racks or caches. Animals were often attracted to buried houses as the vegetation is richer and the soil is often less compact. Foxes, voles, lemmings, ground squirrels, snow buntings, and even dogs or wolves often made house depressions their homes in later years.
As houses collapsed there may have been movement of archaeological material. The roofs of semi-subterranean houses were often used as activity areas, lookouts, or storage facilities (Burch 2006). Material deposited on the roof may become part of the house depression. However, most houses have a distinct floor layer that is covered by thick sod used to insulate houses.
68
3.3.6 Subsequent Disturbances
Human, non-human, and environmental agents affected houses after they collapsed. Erosion is often a major factor on coastlines. Freeze-thaw events in the active layer of permafrost move, split, and deteriorate objects and houses. Rodent or carnivore burrowing could also disturb the site. Humans cause some of the biggest disturbances to houses; they often scavenge structural elements or surface objects, both immediately after a house is abandoned and at later time periods. Acquiring artefacts through surface collection or digging has a long history in the Arctic, but more recently has increased exponentially. Finally, in terms of current context, archaeological excavation is one of the most disruptive activities that affect the house in the course of its life history, resulting in a complete removal of all artefacts and discarded remains and displacement of structural elements.
3.4 Summary
This chapter outlined the general history of semi-subterranean houses within the context of the Alaska Thule way of life using specific documentary references to the northern Seward Peninsula and broader references to the northwest Alaskan region when needed. Documentary descriptions of a general house history inform the specific faunal expectations formulated for Feature 87 in Chapter 5. This chapter also outlined the sources used to construct both the house histories and the expectations; specific critiques of these northwest Alaskan documentary sources were also discussed. The following chapter will contextualize the archaeozoological material by describing the environmental, social, and archaeological context in order to extrapolate expectations against which to test the faunal record.
69
Chapter 4 Cape Espenberg Context
This research focuses on the spatial patterning within one house, Feature 87, at Cape Espenberg (site KTZ-087) on the Seward Peninsula in northwest Alaska (Figure 2). This chapter situates the Cape Espenberg site by describing its physical and social environment. The specific archaeological context of the northern Seward Peninsula and Kotzebue Sound region is described prior to a history of excavation at Feature 87. The architecture of Feature 87 is described and then placed in context with other archaeological Thule Inuit houses within the region, and then more broadly.
4.1 Environmental Context
Cape Espenberg is a spit of land that is home to numerous archaeological sites.6 The spit of Cape Espenberg sits on the border between two ecological regions, Seward Peninsula and Kotzebue Sound, which help shape its past and present physical environment. It is situated on the southeast shores of the Chukchi Sea, just north of the Bering Strait (Figure 2). The Seward Peninsula is the western-most point of the mainland in North America with Cape Prince of Wales, 200 kilometres southwest of Cape Espenberg, and only 88 kilometres from Asia (Ray 1975a). The physical environment will be divided into geology, climate, and ecology for descriptive purposes; in reality, all three are closely interrelated and much more complicated than presented.
4.1.1 Geology
The geology of the Cape Espenberg area is similar to much of northwest Alaska; it is composed of coastal lowlands, offshore bars, barrier islands, sets of beach ridges, and where the mountains meet the shore, steep headlands (Ackerman 1998). Thaw lakes dot the coastal lowlands, while streams drain the mountainous interior along the steep headlands. The spit is a part of
6 For this discussion, the spit will be referred to as Cape Espenberg while the archaeological sites of the same name will be distinguished as Cape Espenberg sites.
70
Figure 2: Location of the Cape Espenberg site and other sites discussed in the text.
71
the Seward Peninsula in the Bering Strait, and abuts the Chukchi Sea and Kotzebue Sound (Kessel 1989). The Chukchi Sea offers continental shelf and deep-water marine environments.
To the north and east, Kotzebue Sound hosts major estuarine (Selawik, Kobuk, and Noatak rivers) and shallow-water habitats not found in other Bering Strait areas (Ackerman 1998). The spit straddles the Arctic Circle (66°30’N, 163°30’W) and extends for about thirty kilometres west to east, varying in width from one to two kilometres (Mason 1990) (Figure 2). The spit is more than 100 kilometres from the tree line within the region of continuous permafrost (Mason 1990). The archaeological sites lie at the southern tidal entrance of Kotzebue Sound, due south of Cape Krusenstern (Mason and Jordan 1993). Cape Espenberg spit and archaeological sites are part of the Bering Land Bridge National Preserve administered by the National Park Service, Alaska, U.S.A.
Over the last 4000 years, the coastline of Cape Espenberg has been built up by the formation of beach ridges and dunes, and changed by erosional agents such as wind and storm surges. The low tides, slow rates of sea level change, and surplus of sediment led to the formation of a series of beach ridges (Mason 1990; Mason and Jordan 1993). After beach ridges formed, high winds delivered sand onto the ridges, which formed dunes. Grass and other plants colonized and stabilized the dunes; soon after the dunes stabilized, there is evidence for human occupation (Mason 1990). However, dunes and ridges are affected by storm surges and wind erosion that modify the landscape. These beach ridges and dunes have been the focus of Dr. Owen Mason’s research at Cape Espenberg and around Kotzebue Sound for decades (Mason 1990; Mason and Bigelow 2008; Mason and Jordan 1993; 2002; Mason and Ludwig 1990). The ridges are numbered starting at the shoreline and working inland; these parallel ridges are separated by inter-ridge swales of variable width. Within the swales, there are numerous circular or oval thaw lakes and marshy basins interspersed with polygonal tundra. Due to low evaporation, there is long-term persistence of water on the landscape. Four major drainages (rivers or tidal channels) crosscut the Espenberg beach ridge complex, with the Espenberg River being the largest (Mason 1990). Cape Espenberg sits on the eastern shore of the Chukchi Sea; presently, ice covers the entire Chukchi Sea during the winter months, with the open water season varying from four to five months between June and October (Mason 1990). Sea level stabilized in Kotzebue Sound after 2000 BCE and has been nearly constant since ca. 550 BCE (Mason 1990:34–35). Although
72 tides are quite small with ranges of less than one metre, storm surges, winds, and changes in atmospheric pressure can cause rapid changes in water elevations of from one to three metres (Kessel 1989; Mason 1990).
Feature 87 is part of the site KTZ-087. Site KTZ-087 is found on ridge E; Feature 87 is found on sub-ridge 5. Ridge E-5 formed around 1300 CE, and perhaps earlier than 1200 CE as a result of large storms that deposited sand on the spit (Bigelow et al. 2013; Mason 1990). This ridge’s eastern edge was eroded sporadically by storm surges between ca. 1400 and 1800 CE (Mason 1990; Mason and Jordan 1993). Site KTZ-087 is located inland from the eastern edge of the spit, and was probably not affected by the erosional periods that continue to truncate the coastal edge of the ridge. These stormy periods are an important part of the climatic history of the area and are described below. Although sea levels were steady, data from the Cape Espenberg Project has shown that the shorelines were changing at this time likely due to seawater inundation around 1000-1100 CE; by 1300 CE the swale was dry enough for peat formation and is likely when people started to occupy ridge E-5 at KTZ-087 (Bigelow et al. 2013).
4.1.2 Climate
Currently, the climate of Cape Espenberg is comparable to the broader Kotzebue Sound region. The Seward Peninsula coast has a moderate maritime summer climate with cold and windy continental winters. In Kotzebue, 55 kilometres northeast of Cape Espenberg, February, the coldest month, has daily average temperatures from -23.3 to -16.1°C, while July, the warmest month, has daily averages of 9.7 to 15.6°C; the average yearly temperature is -5.7°C. Precipitation is most frequent and greatest during the maritime summer climate, averaging 257 millimetres per year. Snowfall averages about 99 millimetres per season. Shishmaref, on the outer Seward Peninsula coast, about 100 kilometres southwest of Cape Espenberg, has a mean temperature range in the summer from 1 to 12°C and in the winter of -24 to -14°C (Kessel 1989). The climate and precipitation at Cape Espenberg is likely similar to what is described for Shishmaref and Kotzebue.
Past climate is often difficult to determine and its effect on human populations is even harder to extrapolate. Although temperature proxies are often used by archaeologists to determine climate in the past, Jordan (2009) argues that the extent and cycle of sea ice was likely more critical to human populations than surface air temperature in the north. Large weather systems are
73 influenced by the position of the ice margin as the formation of clouds is associated with ice surface melting. In periods when the ice margin retracted, cloud cover extent and duration may have been reduced (Clark et al. 1996). The sea ice cover is based on a number of different factors, including sea surface temperature, storminess, cold water upwelling, and wind direction. Although temperature may not be as important to people in the north as other aspects of climate, it is often used as a general indicator of climate change. In the Seward Peninsula area, there were times when temperature was colder and warmer than today. Surprisingly, colder temperatures might not always be worse; Mason (2009b; Mason and Barber 2003) hypothesized that cooler temperatures co-occurred with colder ocean conditions, which increase marine productivity. In addition, certain areas will experience more stable ice conditions during cold winters, which may be beneficial for sea ice hunting (Mason and Jordan 1993). Thus, cooler climates may actually be more beneficial to certain past human groups (Mason 2009b; Mason and Barber 2003). Tree ring data from the Seward Peninsula indicates that from 1000 to 1200 CE the climate was cooler than in the 20th century, with warming after 1200 CE (D’Arrigo et al. 2005). Dendroclimatological analyses from the Seward Peninsula also show narrow rings every 30-45 years between 1150 and 1425 CE; narrow rings may translate into drought conditions, which indicate this period was warming in northwest Alaska (Mason and Jordan 1993). Subsequently, cooling occurred as the northern hemisphere moved into the Little Ice Age. Dendroclimatological evidence shows wider tree rings, which are indicators of greater rainfall and cooler climates (Mason and Jordan 1993). Feature 87 was occupied during the warming period in the 14th century. This warming climate may have had adverse effects in some areas, while opening other areas up to human occupation. Mason (2009b; Mason and Bowers 2009) has postulated that Cape Espenberg may have been an ideal location to spend these warmer years due to the reliable presence of small seals, birds, and other resources when large sea mammals, such as whales and walrus, may have been fluctuating in other areas. Exploring this hypothesis is one of the goals of the broader Cape Espenberg Project.
4.1.3 Ecology
As described above, the divisions between climate and geology and their effect on people are difficult to describe in isolation from one another; to complicate matters, the interactions of animals and people within this physical environment are also intimately linked. The following sections discuss the basics of three overlapping ecosystems as they occur today: marine,
74 terrestrial, and freshwater. Although many organisms occupy all three ecosystems at one time or another, such as sea birds, this rough division of the environment will help organize the descriptions. Additionally, ecological systems will have changed over time; the archaeofaunal results will be compared against this present-day baseline to explore similarities and differences in species composition.
4.1.3.1 Marine Ecology
Past and present marine ecology variability is the result of differences in nutrients, water temperature, and ice cover around continental shelves and coasts (Ackerman 1988; Ackerman 1998; Mason and Gerlach 1995a). In the broader Bering Strait and Chukchi Sea regions, cold, nutrient-rich waters flow along the Siberian coast, while warm, less saline, and nutrient-poor waters flow along the Alaskan coast (Aydin and Mueter 2007). Because of the difference in nutrient upwelling throughout the region, certain areas have differential concentration of marine mammals. Areas with the most concentrated nutrients correlate to the three most culturally important areas in western Arctic prehistory: Point Hope, the greater East Cape region, and northern St. Lawrence Island (Mason and Gerlach 1995a). Other areas, such as Kotzebue Sound and the Seward Peninsula, are areas of low nutrient productivity with lower concentrations of marine mammals.
In addition to the sea nutrient levels and water temperature, sea ice is an important part of the environment and ecology of the north. During the winter months (October to March), sea ice covers the Chukchi Sea and extends southward to the edges of the Bering Shelf (Ackerman 1988). During the summer (May/June to September/October) the Bering Sea is fully ice free as are the central and eastern portions of the Chukchi Sea; the polar drift ice is present for most of the year in the western portion of the Chukchi Sea (Ackerman 1988). Sea ice begins to melt in mid- to late-March and melting is fairly rapid, with ice leads (open, narrow stretches of water) occurring to the west and east of St. Lawrence Island and opening up into the Bering Strait and the Chukchi Sea (Ackerman 1988; Ingraham 1981; McNutt 1981). By late May to June, the majority of ice has melted in the Bering Sea and the rest is dispersed as drift ice (Ackerman 1988). Ice in the Chukchi Sea may last until mid-June, with land fast ice persisting several weeks longer (Harritt 1994). At Shishmaref, average freeze-up is in early November, with break-up dates in late June (Kessel 1989). Kotzebue is ice-free for less time than Shishmaref; freeze-up at
75
Kotzebue is in early October, while break-up is in early July. Ice cover freeze-up and thaw times at Cape Espenberg is likely somewhere between Shismaref and the greater Kotzebue Sound.
In the winter, most animals are found in open water areas, which become extremely important resource areas. Open water leads are long linear cracks in the ice, often found at the margins of fast/shore ice (ice that forms on and stays attached to the shore or the bottom) and drift/pack ice (ice that forms in the water and that is not attached to the shore). Leads close and open depending upon tidal and wind conditions (Ackerman 1988). Polynyas are open water areas; a short-term anomalous polynya sits off of Shishmaref, west of Cape Espenberg, when the winds blow from the north (Mason and Gerlach 1995a). Dune formation can record predominant wind direction, which can be used to approximate times when the polynya may have formed in prehistory (Mason and Gerlach 1995a; Mason 1990). The polynya off the coast of Shishmaref may have been open during the occupation of Feature 87, which is evidenced by the wind direction based on dune formation at this time. Most of the access to sea mammals depends on open water in the form of leads or polynyas; these spots of open water are habitats for plankton and fish under the fast ice and thus attract sea mammals and seabirds. Outside of this short-term anomalous polynya, open water is rare on the northern Seward Peninsula and Kotzebue Sound areas in the winter; most of this area is enclosed by fast ice at this time (Mason 1998). In the winter, the polynya at Shishmaref would have been the only access to open water, while in the spring and fall leads would have allowed access to animals in open water.
Attraction for organisms to the open water is both for access to the air surface and light. In both polynyas and leads, Arctic cod (Arctogadus glacialis) and other small fish feed on the zooplankton under the sea ice (Mecklenburg et al. 2002) and are then in turn food for numerous sea birds such as northern fulmars (Fulmarus glacialis), murres (Uria spp.), gulls, and kittiwakes (Rissa spp.) (Kessel 1989). When ice is present in fall freeze-up and spring break-up, leads attract alcids, particularly murres, glaucous gulls (Larus hyperboreus), slaty-backed gulls (Larus schistisagus), ivory gulls (Pagophila eburnean), Ross’ gulls (Hodostethia rosea), black guillemots (Cepphus grille), migrant eiders (Somateria spp.) and long-tailed ducks (Clangula hyemalis) (Kessel 1989). Other marine species congregate at open water; these include small seals (ribbon seal, Histriophoca fasciata; ringed seal, Pusa hispida; and harbour/spotted seals, Phoca spp.), walruses (Odobenus rosmarus), and polar bears (Ursus maritimus) (MacDonald and Cook 2010; Wynne 2008). Very few mammals inhabit the sea ice in the winter, but of these, the
76 ringed seal (Pusa hispida) is of great importance to past and present people and can be hunted in the open water or through breathing holes. Polar bears can also be found hunting ringed seals on the sea ice (MacDonald and Cook 2010). Bearded seals (Erignathus barbatus) also can be found on and under the sea ice, but are not as frequent as ringed seals (Wynne 2008)7. Although seals are year-round residents, harbour/common (Phoca vitulina), spotted (Phoca largha), and ribbon (Histriophoca fasciata) seals are found only in open water in the winter (MacDonald and Cook 2010; Wynne 2008). After break-up and the onset of open water conditions, pack ice may remain grounded on the shoals east of Cape Espenberg, allowing people to stalk basking ringed seals (Mason and Gerlach 1995a). Whales are found in the area during their spring and fall migrations when they travel to and from the southern, warm waters to occupy the ice-free waters of the Bering Strait, Bering, and Chukchi Seas. These include the beluga (Delphinapterus leucas), gray (Eschrichtius robustus), fin (Balaenoptera physalus), minke (Balaenoptera acutorostrata), and bowhead whale (Balaena mysticetus) (Wynne 2008).
Kotzebue Sound is a slightly different habitat than the west coast of Seward Peninsula, especially in the winter. The Sound is quite shallow, and thus freezes over early and rapidly. In the winter, ringed seals, bearded seals, and polar bears are the only marine mammal resource in the bay (MacDonald and Cook 2010). However, in the spring, summer, and fall, the shallow, warm sound is home to numerous faunal resources. The estuaries of southern Kotzebue Sound provide nurseries for fish and larvae and are thus areas of beluga concentrations (Mason and Bowers 2009). Other species of fish (described below in ‘Freshwater Ecology’) migrate into the rivers and throughout the sound in the warm months. The warm, shallow waters of the inner Kotzebue Sound region are used by birds feeding on the fish and benthic invertebrates (i.e. loons, eiders (Somateria spp.), scoters (Melanitta spp.), harlequin ducks (Histrionicus histrionicus), red- breasted mergansers (Mergus serrator), pelagic cormorants (Phalacrocorax pelagicus), Larus
7 Taxonomic classifications are changing rapidly with new information from genetics altering the hierarchy and relationships of species. The following sources were used in for taxonomic purposes for fish, birds, and mammals; all were accessed via online taxonomy databases at the indicated dates. Fish taxonomy, ecology, and geography range follows individual entries from FishBase (Froese and Pauly 2014) from May, 2014. Bird taxonomy follows the AOU Checklist of North & Middle American birds, including the 54th supplement (AOU 2014) from May 2014; bird ecology follows Kessel (1989). Mammalian taxonomy follows Mammal Species of the World, 3rd edition (Wilson and Reeder 2005) from May 2014. Mammal ecology and geographic range follows Recent Mammals of Alaska (MacDonald and Cook 2010) and Guide to Marine Mammals of Alaska (Wynne 2008).
77 gulls, black-legged kittiwakes (Rissa tridactyla), arctic terns (Sterna paradisaea), common murres (Uria aalge), and horned puffins (Fratercula corniculata)) (Kessel 1989).
Marine waters are habitats for various species of fish throughout the year and birds in open water seasons. Pacific herring (Clupea pallasii) spawn inshore and feed offshore (Mecklenburg et al. 2002) and can be caught most easily when fast ice has formed. Several species of lampreys are also found in the marine and in some cases freshwater ecosystems of the Seward Peninsula. Numerous cod species are found in the marine and brackish waters: Alaska pollock (Theragra chalcogramma), a schooling fish; saffron cod (Eleginus gracilis); and other small cods (Gadidae) (Mecklenburg et al. 2002). People of the Seward Peninsula and Kotzebue Sound areas used all of these fish ethnographically at different times of the year. Birds are found at open water in the spring and fall, as described above, and at other marine areas in the summer. Loons (Gavia spp.), waterfowl, and gulls use protected marine waters; these types of waters are found around rivers mouths relatively close to Cape Espenberg in both Kotzebue Sound and on the Seward Peninsula (Kessel 1989). Colder, deeper waters meet warmer waters along the coasts of western Seward Peninsula; at this boundary, auklets (Aethiini) and thick-billed murres (Uria lomvia) feed. Coastal cliffs, found between Cape Espenberg and Deering are home to pelagic cormorants, glaucous gulls, horned puffins, black-legged kittiwakes, and murres (Kessel 1989). Arctic terns and common eiders occasionally use coastal beaches such as Cape Espenberg; however, loafing and feeding shorebirds, gulls, and waterfowl are more commonly found on these beaches. Areas with salt grass meadows are used by nesting and feeding brant (Branta bernicla) and emperor geese (Chen canagica), black turnstones (Arenaria melanocephala), glaucous gulls, and Sabine’s gull (Xerna sabini), and for foraging and resting by other waterfowl and shorebirds (Kessel 1989).
Although nutrients, water temperature, and geography all contribute to marine resource distributions, ice cover is probably one of the most important factors affecting human interaction with the environment. Open water is essential in the spring and fall for accessing marine birds and mammals, and when available, open water can become a powerful draw for all faunal resources in the winter. Summer waters are often less productive for Inuit than those in other seasons as winter ice cover allows people to travel out on the ice to hunt marine mammals and birds and to fish through the ice. Summer is often a time when people focus on other ecosystems to supplement their resources on the northern Seward Peninsula.
78
4.1.3.2 Terrestrial Ecology
The terrestrial ecosystem at Cape Espenberg is moderately productive in the warm season; one of the attractions is the abundance of flora at the Cape. Although Cape Espenberg sits more than 100 kilometres beyond the tree line, within the region of continuous permafrost, the spit hosts a number of plants that are distinctive in the area due to the well-drained soil. Modern Iñupiat use these plants for fuel, food, artefact construction, and other material and ideological purposes (Sobelman 1985). The most abundant shrubs include three species of willow (Salix spp.), dwarf birch (Betula nana), and crowberry (Empetrum nigrum) (Crawford 2012). Dry heath-type plants are found on the dunes, while wet tundra-type plants are found in the swales and low-lying areas. Lyme grass (Leymus arenarius) is found on the newly formed dunes with older dunes covered in crowberry, bearberry (Arctostaphylos spp.), lowbush cranberry (Vaccintum spp.), and other shrubs. In the swales, there are sedges (Carex spp.), Labrador tea (Rohdodendron tomentosum), bog bilberry (Vaccinium uliginosum), sphagnum moss (Sphagnum spp.), and other shrubs (Jones 2010; Sobelman 1985). On both dunes and drier swales, berries dominate the vegetation (Jones 2010). Driftwood, another plant resource, today is relatively sparse at Cape Espenberg and consists primarily of spruce (Picea spp.), poplar (Populus spp.), and willow, with a low frequency of alder (Alnus spp.), paper birch (Betula papyrifera), and larch (Larix spp.) (Alix 2009).
Many birds nest in in the shrubs, dunes, and swales around Cape Espenberg, but the majority of their time is spent feeding on and around water and are therefore described as part of the freshwater ecology. There are a few birds, such as the Lapland longspur (Calcarius lapponicus) and some sandpipers (Scolopacidae) that are found in the drier dwarf shrub areas (Kessel 1989). These types of small birds were not normally attractive to people in the Arctic and are often considered intrusive in archaeological deposits. Both willow (Lagopus lagopus) and rock ptarmigan (Lagopus muta) are found on the Seward Peninsula and are one of the only terrestrial resources available in the winter (Kessel 1989; Moss and Bowers 2007).
The herbivores of the Seward Peninsula feed on some of the plant species described above, while the carnivores are often attracted to the bird eggs and other small mammals. In the past few hundred years, the terrestrial mammal populations have fluctuated dramatically, with extirpation and re-introduction of multiple species. Up until the mid-eighteenth century, caribou (Rangifer
79 tarandus) were numerous in the area (Harritt 1994; MacDonald and Cook 2010). Burch (2012) argues that there was a distinct caribou population on the Peninsula that calved in the spring around the Espenberg Peninsula, spent summers in the elevated areas near modern-day Shishmaref and to the south, and was sparsely distributed in the fall and winter. Additional data suggests that calving may have also occurred on the northwest coast of the Seward Peninsula near Cape Espenberg (Harritt 1994). In 1890, Sheldon Jackson introduced Siberian reindeer as a subsistence resource for modern populations and today there are multiple reindeer herds managed by different families (Wisniewski 2010). Additionally, there are some wild caribou on the Seward Peninsula (Wisniewski 2010). Other large terrestrial mammals include muskox (Ovibos moschatus) and moose (Alces alces) (MacDonald and Cook 2010); both of these species have uncertain prehistoric distributions on the Seward Peninsula. Muskoxen seem to have existed on the peninsula since the Pleistocene, but were extirpated by the early 1900s, then re-introduced recently (MacDonald and Cook 2010). Although moose are represented from archaeological sites in the interior (Larsen 1968), they are not know in the Seward Peninsula area prior to 1950, when an expansion of their territory brought them into parts of the Seward Peninsula (Harritt 1994; Wisniewski 2010). Other terrestrial mammals found on the Seward Peninsula include carnivores, lagomorphs, and rodents (MacDonald and Cook 2010). Some of the most important for people are foxes (Vulpes spp.), wolves and dogs (Canis lupus), hares (Lepus spp.), Arctic ground squirrel (Spermophilus parryii), and lemmings and voles (Arvicolinae).
For the most part, the resources from the terrestrial ecosystem are procured in the spring, summer, and fall months, with most species migrating or hibernating when there is snow cover. Arctic fox (Vulpes lagopus) and Alaskan (Lepus othus) and snowshoe hares (Lepus americanus) often overwinter on the Seward Peninsula (MacDonald and Cook 2010). Ptarmigan are the most reliable terrestrial resource in the winter, as they congregate in flocks in the underbrush, making them easier to snare than when they are dispersed in the summer (Kessel 1989; Moss and Bowers 2007). Although foxes, hares, and ptarmigans are found in the area in the winter, most of these animals move into the interior for the majority of the winter season (Kessel 1989; MacDonald and Cook 2010).
80
4.1.3.3 Freshwater Ecology
The freshwater and marine ecosystems are difficult to differentiate as most species move through both ecological zones seasonally. Only the northern pike (Esox lucius), landlocked lake trout (Salvelinus namaycush) and landlocked Arctic char (Salvelinus alpinus) stay within the freshwater ecosystem. Most of the salmonids of the Seward Peninsula and Kotzebue Sound spawn in fresh water but spend much of their life in the marine ecosystem. Anadromous salmon found on the Peninsula and around Kotzebue Sound include coho/silver (Oncrhynchus kisutch), chum/dog (Oncorhynchus keta), pink/humpback (Oncorhynchus gorbuscha), sockeye/red (Oncorhynchus nerka), and chinook/king salmon (Oncorhynchus tshawytscha) (Mecklenburg et al. 2002); none of these salmon species run in large numbers at Cape Espenberg, with the closest large salmon run in the Yukon River to the south (Groot and Margolis 1991). However, there are smaller salmonid populations in the Kobuk River and other small streams including several species of whitefish (Coregoninae), sheefish/inconnu (Stenodus nelma), and Arctic grayling (Thymallus arcticus) (Groot and Margolis 1991). A number of species of char (Salvelinus spp.) are found in and around Cape Espenberg, including Dolly Varden char/trout (Salvelinus malma), Arctic char, and lake trout.
As opposed to the salmonid population, other fish are found mostly in freshwater with only brief forays into salt or brackish water. Burbot (Lota lota), mostly found in fresh water, spawn in brackish environments (Mecklenburg et al. 2002). Alaska blackfish (Dallia pectoralis) are small, bottom feeding fish mostly found in swamps, ponds, lakes, and streams with vegetation for cover (Mecklenburg et al. 2002). These fish congregate in the vicinity of open water during the winter, making them easy prey for both other animals and humans. Northern pike are also found in freshwater and brackish water in the area. Various species of smelt (Osmeridae) are found in the lakes, rivers, and near-shore areas around the Seward Peninsula (Mecklenburg et al. 2002). The longnose sucker (Catostomus catostomus) inhabits freshwater lakes and rivers and is often prey for larger fish species. Capelin (Mallotus villosus) spawn near shore but depending on the populations, spend most of their time in the ocean or rivers and streams. Various species of sculpins (Cottoidae) are found in both freshwater and marine habitats (Mecklenburg et al. 2002).
Again, the division between the ecosystems blurs when talking about birds; often birds feed on fresh and marine water and nest in the terrestrial areas. Areas with flowing waters are used
81 extensively by breeding harlequin ducks and red-breasted mergansers, glaucous and mew gulls (Larus canus), and American dippers (Cinclus mexicanus) (Kessel 1989). The long, linear lakes at Cape Espenberg are primary breeding areas for loons, waterfowl, and phalaropes (Phalaropus spp.) (Kessel 1989). Wet meadows are home to pectoral sandpipers (Calidris melanotos), long- billed dowitchers (Limnodromus scolopaceus), and red-necked phalaropes (Phalaropus lobatus). Normally in July, flocks of post-breeding shorebirds and waterfowl gather at productive feeding sites at Cape Espenberg prior to their southward migration (Kessel 1989).
As with the terrestrial ecosystem, the freshwater ecosystem is mostly used during the warm season, with salmonids and migratory birds being caught during the spring and fall migrations. In particular, nesting and moulting birds are important subsistence resources in the area throughout the warm months. Like their marine counterparts, many fish species are caught through the ice during freeze-up and break-up, with a few species caught in the winter. Although there is much work to be done on the prehistory of Cape Espenberg, this diversity of resources has likely made Cape Espenberg an attractive location for people to live on a broad range of resources year-round at some times, or as a seasonal camp for procurement of specific resources at certain times of the year.
4.2 Social Context
Chapter 2 outlined the general Arctic culture history; this section will briefly outline the local culture history of the Seward Peninsula and Kotzebue Sound, with specific focus on Cape Espenberg. The earliest people of the Seward Peninsula date to just after rising sea levels covered the Beringia continent, about 10,000 years ago (Harritt 1994; Larsen 1968; Mason and Gerlach 1995a); the spit of Cape Espenberg has been occupied intermittently since its formation 4000 years ago.
The earliest sites in the region, those of the Palaeoarctic tradition, are found in the interior of the Seward Peninsula and are centred on river valleys; subsequent to this earliest tradition are sites from the Palaeoeskimo8 tradition. Early sites include the Trail Creek Cave site (Larsen 1968) and
8 In defining the chronology for the region, Giddings and Anderson (1986) use Arctic Small Tool Tradition to classify the Denbigh Flint complex, Choris, Norton, and Ipiutak cultures. This dissertation uses “Palaeoeskimo” and “Inuit” to name the traditions in order to coincide with a broader North American Arctic classification
82 the Serpentine Hot Springs Fluted Point site (Goebel et al. 2013). On the Seward Peninsula, Palaeoeskimo sites date from 2000 to 700 BCE and are found both inland and on the coast. Interior sites have been identified at Trail Creek Caves (Larsen 1968) and Kuzitrin Lake (Schaaf 1988). Onion Portage in the Kotzebue Region has Denbigh winter houses, features not found yet on the Seward Peninsula (Dumond 1987; Giddings and Anderson 1986). The earliest Palaeoeskimo people in the area were those of the Denbigh culture who are thought to have moved seasonally between the coast and interior (Harritt 1994; Larsen 1968). The earliest occupation at Cape Espenberg consists of Denbigh seal hunting sites, used as seasonal spring and summer coastal camps (Giddings and Anderson 1986; Larsen 1968; Schaaf 1988).
After a hiatus, the spit was re-occupied by people of the Choris complex around 900 BCE (Mason and Gerlach 1995b). Choris is also found at numerous sites on the Seward Peninsula such as the Trail Creek Caves, Onion Portage, and Agulaak Island, and around Kotzebue Sound at sites on the Choris Peninsula and at Cape Krusenstern (Giddings and Anderson 1986; Harritt 1994). The Choris occupation at Cape Espenberg is similar to the occupation by Denbigh people; both show evidence for briefly occupied, ephemeral camps at hundred-year intervals. Choris likely maintained a similar subsistence strategy to Denbigh, with coastal occupations in the spring and summer for seal hunting, and interior occupations in the cold months for caribou hunting (Giddings and Anderson 1986).
After Choris, people of the Norton culture inhabited the spit. Their use of the site prior to 200 CE was likely only as a sealing camp, although knowledge of this site is limited (Harritt 1994; Mason 1998; Schaaf 1988). Norton sites are found throughout the Seward Peninsula and along the coast including Kugzruk Island, Trail Creek Caves, Ikpek, Agulaak, Cape Nome, Cape Krusenstern, Iyatayet, and Gungnuk (Bockstoce 1979; Dumond 1987; Giddings and Anderson 1986; Harritt 1994; Larsen 1968; Mason 2009b). Coastal Norton sites in this region may represent a variety of subsistence strategies ranging from an exclusive focus on seals, to a combination of whaling, sealing, and caribou hunting (Giddings and Anderson 1986; Harritt 1994).
Based on current evidence, there is a gap in occupation after the Norton culture followed by a brief re-occupation of the area by the Ipiutak culture around 650 CE (Harritt 1994). These are the first permanent houses at Cape Espenberg (Harritt 1994). Harritt (1994) suggests that the Ipiutak
83 occupation was primarily during the winter. In addition to the occupation at Cape Espenberg, Ipiutak are found at Trail Creek Caves, Point Spencer, and Deering (Harritt 1994; Larsen 1954; Moss and Bowers 2007). The coastal sites show evidence of winter occupation, probably for seal hunting. The Ipiutak culture around Kotzebue Sound declines rapidly during a cool, stormy period around 900 CE and is replaced by the Birnirk, Thule, and Punuk cultures (Mason and Barber 2003).
Despite extensive archaeological survey, there are no sites between Ipiutak and early Inuit9 occupations, likely representing a real occupational hiatus at the spit. On the Seward Peninsula and in the Kotzebue Sound region, Inuit occupations of Punuk, Birnirk, and Thule cultures overlap temporally and often geographically. Thule occupations are the most numerous, with a number of Birnirk sites in the area. The only known Punuk site components on the Seward Peninsula are at the Kurigitavik Mound in Wales (Harritt 1994; 2004). Birnirk is found at both Cape Espenberg and Cape Krusenstern (Giddings and Anderson 1986). Other sites with Birnirk components are at Cape Nome and Cape Prince of Wales (Dumond 1984; Giddings and Anderson 1986). Kotzebue Sound and the Seward Peninsula are the most southern extent of the Birnirk population. In the area, Birnirk subsistence seems to focus primarily on seal hunting, supplemented by caribou and other mammals (Giddings and Anderson 1986), although detailed faunal analyses have yet to be completed on any of the Birnirk materials in the area. Recent dating puts the Birnirk occupation of Cape Espenberg at ca. 1000 CE, and marks it as one of only a few dated Birnirk occupations (Mason 1998; 2009b)
Thule people occupied the Kotzebue Sound and Seward Peninsula region starting about 900- 1000 CE, with dense occupation of Cape Espenberg indicated by the number of houses and sites (Darwent et al. 2013; Mason 1990). Giddings and Anderson (1986) separate the Thule culture into two periods in the Kotzebue Sound region: Western Thule and Kotzebue. Although there are differences between these two Thule occupations, there is no significant break in the cultural continuum (Dumond 1987; Harritt 1994). Harritt (1994) argues that there is not enough evidence to separate the Thule into different periods, and uses the term “Western Thule” to describe the
9 In the chronology of the region developed by Giddings and Anderson (1986), they use Northern Maritime Tradition to classify the Birnirk, Thule, and Kotzebue Period cultures. Here, “Inuit” will be used to designate the larger tradition as previously stated in Chapter 2.
84 entire occupation. Although the differences identified by Giddings and Anderson (1986) for Western Thule and Kotzebue Period cultures are discussed below, the terms “early Thule” and “late Thule” will be used to designate the two periods. Despite the debates on terminology, early Thule occupations seem to have large villages with multi-roomed houses; later Thule settlements show a reduction in village sizes and house complexity (Anderson 1984). Giddings and Anderson (1986) suggest that early/Western Thule settlement and social patterns were based on successful whale hunting. After 1400 CE, only groups in the best locations for whale hunting focused on this subsistence pursuit; most other late/Kotzebue Thule people broadened their subsistence base (Giddings and Anderson 1986). This regionalization mirrors other Alaskan Thule groups’ shift to local resources at this time. The evidence for whale hunting at Cape Krusenstern is equivocal and whales currently do not migrate in great concentrations in the area.
Early and late Thule sites are found throughout the study region at Point Hope, Kivalina, Cape Krusenstern, Kotzebue, Sisualik, Cape Espenberg, the Ikpek area, the Kitluk River area, the Nugnugaluktuk River drainage, the Sullivan Bluffs area, Deering, and Shismaref, although they are still best known from Cape Krusenstern, both within Kotzebue Sound and throughout Alaska (Giddings and Anderson 1986; Harritt 1994; Schaaf 1988). Cape Krusenstern was one of the only places where single component early Thule materials were identified (Giddings and Anderson 1986). At Cape Krusenstern, Giddings and Anderson (1986) excavated five houses in the largest of the early Thule sites. The early Thule settlement seems to have been occupied by a relatively large number of people. Although there is no conclusive evidence for whaling, large storage pits and composite houses suggests to Giddings and Anderson that whaling occurred in the area. Harritt (1994) questions this interpretation, and suggests that there is evidence for whaling near Ikpek, although not at other early Thule sites in the region. Feature 87 at Cape Espenberg dating to the mid-14th century, also represents a discrete early Thule occupation and is very similar to the early Thule Cape Krusenstern house structures as described below.
Late Thule (Kotzebue period) is described as a local variant of broader late prehistoric Thule culture (Giddings and Anderson 1986). At Cape Krusenstern and the Choris peninsula, house form drastically changes from a typical Thule Inuit form (square with a rear platform) to a local house form (trapezoidal with a central hearth) (Collins 1937; Giddings and Anderson 1986). The trapezoidal form is found in southern and interior Alaska, with a few examples from northern Alaska (Young 2002). Giddings and Anderson (1986) recorded late Thule sites at Battle Rock,
85
Cape Krusenstern, Sisualik, Kotzebue, Choris Peninsula, Trail Creek, Deering, Cape Espenberg, and Espenberg River in the Kotzebue Sound and Seward Peninsula region.
Cape Espenberg has been sparsely populated from the late prehistory to the modern day. After 1700 CE, people largely abandoned the site, with no permanent occupation after 1900 CE. Currently, the local name for Cape Espenberg is Inuigniq (“no more people”) (Schaaf 1995). The first explorers in the region record people at the site of Cape Espenberg (Ray 1975a; von Kotzebue 1821), although almost no people occupied the site by the time of later, more extensive, ethnographic recordings. Both Ernest S. Burch, Jr. and Dorothy J. Ray have recorded limited information about the groups of Cape Espenberg from ethnohistoric interviews.10 Burch (1998a; 2006) identifies the inhabitants as members of the Pittagmiut nation; Ray (1975a) calls this historic Iñupiat nation the Buckland tribe. In the early 19th century, a population count based on early explorer accounts estimates that only 300 people lived along the coast between Cape Prince of Wales and Barrow in winter settlements, with nearly two-thirds of the population living in five communities: Point Hope, Nuwuk, Utqiagvik (at Barrow), Wales, and the four Bering Strait islands (Foote 1965). Fewer people lived in the secondary sites: Kivalina, Cape Espenberg, and other sites between Cape Lisburn and Icy Cape (Foote 1965). The largest village in the Cape Espenberg area in the 1880 census had a population of forty and was located at the mouth of the Espenberg River (Ray 1964). People living here focused on Goodhope Bay (southern Kotzebue Sound) and visited the Cape Espenberg spit primarily for sealing during late winter/early break- up in May to early July (Mason 1990). After the 1880s, this nation seems to have disappeared in Northwest Alaska, with only one person remaining in the Goodhope district (Burch 1998a:285– 287,296). Burch (1984) mapped Cape Espenberg as a border between two societies: a no-man’s land or a defensive outpost. Presently, Cape Espenberg is a peripheral settlement with 20 to 30 modern Iñupiat/Yupik using the land seasonally to procure seals, walrus, and waterfowl (Mason and Gerlach 1995a; Ray 1964; Sobelman 1985; Wisniewski 2010). This seems to have been the case for the last 200 years.
10 See Chapter 1 and 3 for further discussion of ethnographic and ethnohistoric records.
86
4.3 Archaeological Context: History of Research
This following section puts the recent archaeological excavations in historical context. However, it should be noted that collecting, curating, and reusing artefacts from the past by modern and prehistoric Inuit and Iñupiat groups likely impacted most of the sites in Alaska and that this section only outlines the history of Western archaeological research.
Archaeological work on the Seward Peninsula was done by some of the earliest Arctic researchers. Diamond Jenness conducted the first archaeological survey and excavation in 1926; he identified the village sites around the modern communities of Wales and Teller (Jenness 1928). This same year, Aleš Hrdlička surveyed the coast around Norton Sound on the south coast of the Seward Peninsula (Hrdlička 1930). Henry Collins in 1928 and 1929 explored the coast around Norton Sound and up the Bering Sea coast to Shishmaref (Collins 1929); Collins returned in 1936 to excavate sites at Cape Prince of Wales (Collins 1940). In the 1940s, Louis Giddings, Wendell Oswalt, Froehlich Rainey, and David Hopkins conducted surveys and excavations on the Seward Peninsula, including the excavation of the Ipiutak ceremonial house at Deering (Giddings 1964; Larsen 2001). In 1950, the Bering Strait Expedition was the first large scale, collaborative archaeological investigation of the Seward Peninsula based on the complementary investigations of Giddings, Larsen, and Rainey in different regions (Giddings 1964).
J. Louis Giddings conducted archaeological research that resulted in a culture sequence still in use today. He continued his work in 1956 and 1958 and focused on the Choris Peninsula and Eschsholtz Bay of eastern Kotzebue Sound and the Buckland River Valley of the northern Seward Peninsula (Giddings and Anderson 1986). During the 1958 season, Giddings’ survey included Cape Espenberg, previously ignored by archaeologists (Giddings 1967; Giddings and Anderson 1986). Giddings (1967: 22-23) was not hopeful that the sequence at Cape Espenberg would prove fruitful, as he suspected that the dunes constantly moved, and that later ridges had covered earlier ridges, thus obscuring the actual house depressions. However, his survey located materials from the Denbigh and Choris-related cultures—a successful outcome as he was looking for the earliest occupations (Giddings 1967; Giddings and Anderson 1986). Subsequent to his initial survey of Kotzebue Sound, Giddings returned to Cape Krusenstern, across the sound from Cape Espenberg, from 1958 to 1962 to conduct extensive survey and excavation (Giddings 1967; Giddings and Anderson 1986). Again, these surveys and excavations focused on early
87 occupations (Giddings and Anderson 1986). Subsequent to his excavations at Cape Krusenstern, Giddings moved to the inland site on the Kobuk River of Onion Portage (Giddings 1967; Giddings and Anderson 1986). Here, the cultural chronology paralleled that of Cape Krusenstern.
Archaeological work continued steadily in the northern Seward Peninsula and Kotzebue Sound region from then on, with surges in the 1970s and 1980s prior to and during the creation of the Bering Land Bridge National Preserve. The first extensive survey was done by Powers et al. (1982) in 1974 on the southern and interior portions of the Seward Peninsula. In the 1980s, Jean Schaaf conducted extensive coastal and inland surveys that included Cape Espenberg (Schaaf 1988; 1995). Following Jean Schaaf’s original work, Roger Harritt oversaw work at Cape Espenberg in 1988, 1989, and 1990 (Harritt 1994). As described above, Owen Mason was also conducting geomorphological and climatic work in the area (Mason 1990). Harritt’s (1994) work was the first to look at faunal remains at Cape Espenberg. The amalgamated analysis of faunal material from the site showed a reliance on marine resources that were supplemented by terrestrial resources during the winter months. Comparatively, sites in the interior of the Seward Peninsula show use of caribou during all seasons, with other resources gathered seasonally (Harritt 1994).
Recently, excavations have been conducted at Wales (Harritt 2004; 2010), Deering (Moss and Bowers 2007), Serpentine Hot Springs (Goebel et al. 2013), and Cape Espenberg (Darwent et al. 2013). The recent work at Cape Espenberg started in 2007 and then continued from 2009 to 2011 as part of the large, interdisciplinary Cape Espenberg Project (CEP). Part of the project mapped a number of sites (Darwent et al. 2013), including KTZ-087 and KTZ-088 previously mapped by Schaaf (1988) and Harritt (1994). Results of these excavations are in preparation for publication and have been presented at regional and international conferences.
Archaeological work in the past has established a general culture history for the region; research continues to refine these broad narratives while exploring the particulars of people living at specific sites. The previous discussion of the physical, ecological, and social contexts of the Seward Peninsula and the Kotzebue Sound region have placed the site of Cape Espenberg within this broader culture and environmental history in order to elucidate the lifeways of the people who inhabited one dwelling on the spit. The following discussions now focus on the specific
88 context of Feature 87, an early Thule Inuit semi-subterranean dwelling, in order to situate the zooarchaeologically-based analyses in the following chapters.
4.4 Early Thule Inuit Semi-Subterranean Dwellings
Thule Inuit semi-subterranean dwellings are often both the most visible structures in the prehistoric Arctic landscape and the most archaeologically rich. Therkel Mathiassen (1927a) originally described these houses as part of the Thule culture based on excavations north and west of Hudson Bay. His synthesis compared the expedition’s findings to excavated remains from Siberia to East Greenland and to contemporary cultures in Alaska and Greenland (Mathiassen 1927b); he recognized the surprising similarity in house form of the ancient Thule people across 6000 kilometres from Siberia to Greenland. This type of semi-subterranean house was quickly realized to have a long history in Alaska antecedent to the Thule culture, likely dating back to the origins of the Inuit culture (Collins 1937; Larsen and Rainey 1948).
Collins (1937:256–286) placed ethnographically and archaeologically identified houses into a number of house types based on structure and construction material. The most commonly found house type along northwest Alaskan coasts between approximately 1000 and 1900 CE was called the Point Barrow type. This highly recognizable structure was a semi-subterranean dwelling comprised of a relatively long, sloping entrance tunnel that led to the main living room. People entered the living area through an opening in the floor near the front wall (katak). Collins (1937) also recorded a number of small annexes or storage alcoves along the tunnel, in addition to a kitchen. The kitchen itself was a separate space connected to the tunnel where much of the cooking and processing of food was done. These houses were constructed of driftwood and whale bones and were covered in sod blocks for insulation. Many houses of the Thule period have been identified as Point Barrow, or early Thule, style houses. In the late Thule period (starting around 1450 CE), house shapes change and diversify; prior to ca. 1450 CE most of the coastal houses in northwest Alaska across to East Greenland are of the Point Barrow/early Thule style.
Despite the extensive research conducted in Alaska, there is paucity of fully excavated early Thule Inuit houses. Much of our knowledge of house structure and form comes from Cape Krusenstern houses and eastern Thule houses, described below. Most fully excavated houses are dated to the late Thule/early historic period (e.g. Ford 1959; Hall and Fullerton 1990a).
89
Subsistence information and artefact descriptions mostly come from test pits within houses or middens, often with their locations uncertain (i.e. Stanford 1976). Feature 87 at Cape Espenberg, therefore, offers a unique perspective on early Alaskan Thule Inuit behaviours, use of domestic space, and house life history.
4.4.1 Cape Espenberg
Thule Inuit houses on the same ridge at Cape Espenberg as Feature 87 are similar to and different from Feature 87. Feature 68 was originally thought to be a multi-roomed house with rooms connected by entrance tunnels that joined at an entrance shed; during excavation, it became clear that it was really two separate features with no structural connection (Darwent et al. 2013). Feature 68A is a single room winter house with an external hearth/pottery production area and an entrance tunnel (Darwent et al. 2013). Further analyses of houses on this ridge will contribute to the preliminary interpretations offered in this dissertation, but house form seems to be fairly consistent within the two excavated dwellings.
4.4.2 Cape Krusenstern
Giddings and Anderson (1986) describe ten early Thule houses at Cape Krusenstern, directly north of Cape Espenberg across Kotzebue Sound; all houses followed the early Thule pattern with a main room, kitchen, tunnel, and in some cases a side room (Giddings 1967; Giddings and Anderson 1986).
4.4.2.1 Houses 7 and 8
The following description comes from the text and extrapolation from the image published by Giddings and Anderson (1986:78). Based on their artefact types and location, the earliest Thule houses are House 7 and 8. House 7, the larger of the two, has been dated between 700 and 1400 CE (Giddings and Anderson 1986). House 7 (Figure 3) has a main room and a side room connected by a tunnel Figure 3: Cape Krusenstern, House 7 (from Giddings and Anderson to the main tunnel. Both rooms were built in the four- 1986:73)
90 corner-post construction style. Both rooms have fully planked floors that run perpendicular to the tunnels. Whale bones may be found at the entrance to the tunnel. The main room of House 7 has both rear and side platform areas marked off by posts. The rear platform is covered with planks paralleling the rear wall and extending to about the middle of the room. The side platform, situated on the opposite side of the kitchen, is covered with planks parallel to the side wall. Giddings and Anderson suggest that this side platform was an extension built onto the house, although there is no evidence that it was built after the rest of the house; without it, the house would have been identical in size to the other room of House 7 and most other early Thule houses. House 7 has a smaller room that extends from the front left corner of the main room. This room is entirely constructed of split upright planks without any supporting frame, although Giddings and Anderson suggest it may have been A-shaped. The room had charcoal over the floor, and therefore was likely a kitchen.
House 8 has a slightly different floor plan than House 7; again, the description is derived from the text and image (Giddings and Anderson 1986:78, 79). House 8 has one single main room and large kitchen extending off of the front left corner of the main room (Figure 4). The main room was built on the four-corner-post construction. On the image, evidence for a back platform is in the upright posts that mark the front of the platform. The kitchen is relatively large and its floor was covered in broken ceramics, bones, and a thick greasy charcoal midden. The tunnel is constructed of paired upright posts and horizontal wall logs; the floor is partially planked. Figure 4: Cape Krusenstern, House 8 (from Giddings and Anderson 1986:79)
91
4.4.2.2 Houses 4, 5, and 6
The house plan descriptions for House 4, 5, and 6 are taken from the written descriptions and images of Giddings and Anderson (1986:72, 73). Houses 4 and 6 are of the same type as House 8 (Figure 5, Figure 6). Both houses have a main living room with a back platform, a kitchen off the front corner of the house, and an entrance tunnel. Of the House 4-6 cluster, House 4 is the smallest with only a single living room, a kitchen, and tunnel. Giddings and Anderson state that the tunnel runs diagonally towards the adjacent tunnel entrance to House 5, but based on the site map, House 4 and 5 are at Figure 6: Cape Krusenstern, House 4 least 20 metres apart and both face the same direction. (from Giddings and Anderson Based on the floor plan image, House 4 does not have 1986:73) any planked floors and the back platform is demarcated with a single log running perpendicular to the tunnel just slightly to the rear of the middle of the room. The house was constructed with four corner posts. The kitchen is attached to the house via a small tunnel that runs off the front left corner of the house. A hearth area takes up almost the entire kitchen area. What may be whale bones are also indicated on the house plan: one in the kitchen, one on the main room side floor, and one on the main room platform.
House 6 is slightly larger than House 4, with a single main room constructed with four corner posts (Figure 6). The kitchen attaches to the house in the same way as House 4: off the front left corner. The kitchen is attached to the house with a small, planked tunnel. The floor is partially planked. There is no indication of a hearth area in the House 6 kitchen. The main tunnel is partially planked, with the planking running parallel to the tunnel. Figure 5: Cape Krusenstern, House 6 (from Giddings and Anderson 1986:73)
92
House 5 is slightly more complicated than Houses 4 and 6 with a main room, a kitchen, and a side room attached to the main tunnel via a side tunnel (Figure 7). The side room has a planked floor. Both the main room and side room have a four-corner post construction. The main room also has a platform across the back half, demarcated by a single horizontal log just to the rear of the middle of the room. The side room is almost completely planked. The kitchen attaches to the main room off the front right corner by a tunnel and is outlined by upright posts. There Figure 7: Cape Krusenstern, House 5 (from is a small hearth area indicated on the image Giddings and Anderson 1986:73) plan at the back side of the kitchen. A small rectangular alcove with a planked floor and log walls is off the main tunnel near the entrance to the side room. Corner posts and horizontal wall logs delimitate the alcove.
4.4.2.3 Houses 25a, 25b, 26, 27, and 29
Giddings and Anderson (1986) identified a number of later Thule settlements at Cape Krusenstern (ca. 1300-1400 CE). They excavated five houses from the largest of these settlements in a cluster of six house depressions: Houses 25a, 25b, 26, 27, and 29. The house depressions were large and deep, with well-preserved logs and planks as well as posts. Most of the rooms had either fully or partially planked floors, which were often covered in a midden. This suggested to them that the floors were layered in dirt during later occupations or that the midden was part of the roof debris. None of the houses had more than one layer of floor planks. The artefacts were mostly found in the floor planks and roof midden, with no artefacts found under the floor in the sterile gravel. Description of all the houses comes from Giddings and Anderson’s (1986:59-60) in text descriptions and floor plan images. Based on artefact styles, Giddings and Anderson (1986:70) suggest that House 26 and 27 were about fifty years earlier than Houses 25a, 25b, and 29. However, the association of both Houses 26 with House 27 and House 29 with Houses 25a and 25b are based on similarities in construction details, proximity,
93 and details of artefact styles in the field notes as the artefacts from both House 26 and 29 were lost when a barge sank in a storm from Nome to Seattle after the 1959 field season.
Houses 25a and 25b are a pair of adjacent houses. Giddings and Anderson suggest that they are slightly more recent than Houses 26 and 27 based on their artefact assemblages. Houses 25a and 25b were both constructed in the four-corner-post style (Figure 8). House 25a has a rear platform separated from the front by a plank parallel to the rear wall with two posts slightly to the rear of the plank. The kitchen area extends diagonally from the front Figure 8: Cape Krusenstern, House 25a (left) left corner of the main room. It is connected to and 25b (right) (from Giddings and Anderson the main room by a narrow tunnel, which 1986:61) widens out to form the kitchen room. The kitchen floor is composed of a thick, charred, greasy material, probably sea mammal fat. There were piles of bones in the corners. The long tunnel extends from the middle of the front wall diagonally to the right, where it comes closest to the tunnel of House 25b. There is a small rectangular storage compartment off the tunnel near the main room on the kitchen side that is made of horizontal wall logs.
The structure of House 25b was not as well preserved as House 25a, although there is evidence of a main room and entrance tunnel. The main room has a four-corner-post construction (Figure 8). There is no clear indication of a platform at the back. The main room floor seems to have three to four floor planks running perpendicular to the side walls. The tunnel extends from near the middle of the front wall diagonally to the right, where it comes close to the entrance of House 25b. Although neither shows evidence of an outer storm shed, there were planks at ground level around the tunnel entrances that may indicate a surface construction that may have connected the two tunnels (Giddings and Anderson 1986:59). In between the two houses there is a whalebone- framed cache pit, nearly touching the front corner of House 25b. There is no entrance to it from either house or any trace of midden on the floor.
94
House 29 is similar in construction to Houses 25b and 26, with a single main room, entrance tunnel, and kitchen (Figure 9). The main room was built as the four-corner-post type. The only evidence for a rear platform is a single post about halfway between the rear and front of the house near a side wall. The kitchen extends diagonally off the front left corner. The kitchen is connected to the main room by a tunnel. The kitchen floor is covered with charcoal and broken stones, potentially used as boiling stones. The main tunnel extends from about the middle of the front wall diagonally to the right. About halfway down the tunnel, a small room extends off the side opposite that of the kitchen. Figure 9: Cape Krusenstern, This room was constructed out of whalebones. Houses 29 (from Giddings and Anderson 1986:61)
House 26 consists of a main room, tunnel, and kitchen area. The main room was again constructed around four corner posts (Figure 10). Giddings and Anderson (1986:65) comment that the poor preservation of the house may have removed interior main room features such as a dividing element between the front living and rear sleeping areas. Unlike any of the other early Thule houses excavated at Cape Krusenstern, the kitchen of House 26 is off the entrance tunnel and not the main room. The kitchen is a long, narrow room with walls of closely spaced vertical poles. The floor was covered in a dense charcoal midden. The entrance Figure 10: Cape Krusenstern, House 26 tunnel adjoins the main room in the center of the front (from Giddings and Anderson 1986:65) wall and ends in a storm shed. The storm shed at the end of the tunnel is a small rectangular room, framed by four corner posts.
95
House 27 is the largest and most complex house excavated at Cape Krusenstern with five individual rooms (Figure 11). There is a main room with a kitchen area attached to the front right corner and two side rooms adjoin the tunnel on the opposite side from the kitchen. Both side rooms and the kitchen have tunnels attaching them to the house. The main room and the two side rooms follow the basic four-corner post construction. These three rooms are planked with wood. A line of three posts marks off a rear platform that is parallel to Figure 11: Cape Krusenstern, House 27 (from the rear wall. The side rooms do not have a Giddings and Anderson 1986:65) rear platform demarcation. The larger side room, closer to the main room, has a smaller room extending off the rear corner. The kitchen, off of the main room, is triangular in shape and Giddings and Anderson suggest that the rear wall may have sloped inward toward two mid-wall posts in the fashion of a Nunivak Island house illustrated by Collins (1937: 258-260). The middle of the kitchen floor was covered in a thick, oil-soaked charcoal mass over an unplanked floor; animal bones were piled along the outer end of the room. The entrance tunnel joins the main room squarely in the middle of the front wall.
4.4.3 Deering
Excavations at Deering, on northern Seward Peninsula, 68 kilometres southeast of Cape Espenberg on Kotzebue Sound, have uncovered two Thule semi-subterranean houses. These houses were excavated in 1998 and 1999 and date to approximately 1000 to 1300 CE (Moss and Bowers 2007). Both houses were made of driftwood, whale bone, and sod. House 1 was mostly excavated and is composed of a single main room, a long entrance tunnel, and a side room kitchen.11 Floorboards covered much of the interior of the main room; there were several bowl-
11 There has been no figure published for this house.
96 shaped subfloor caches and several hearths (Moss and Bowers 2007). House 1 had a 6.4 metre long entrance tunnel and a side room that functioned as a kitchen (Moss and Bowers 2007). House 2 was partially excavated; only the main room and part of the tunnel were uncovered so the presence/absence of a side room is unknown. The main room in both houses was roughly square (Moss and Bowers 2007).
4.4.4 Qimiarzuq
Although most of the material excavated by Larsen and Rainey (1948) from Point Hope was Ipiutak, the house excavated at the site of Qimiarzuq (previously Jabbertown) was associated with the early Thule period. House 2 is one of the largest and most complicated houses excavated in northwest Alaska for the Thule period (Figure 12). The house has two main rooms that each has an attached kitchen. Both main rooms are attached to a single entrance tunnel. The first main room was paved with logs on top of gravel cemented by sea mammal oil. Its kitchen has a partial wood floor
Figure 12: Qimiarzuq (Jabbertown), House 2 and includes a hearth area; much of the floor (from Larsen and Rainey 1948:171) was covered in a thick layer of sea mammal oil mixed with animal bones, baleen, wood, hair, and feathers. The second main room was partially paved with wood. The kitchen attached to the second main room also contained a hearth. The final room in the house structure was attached to the main tunnel and was classified as a storeroom
.
97
4.4.5 Utqiagvik
Along the North Slope, Ford (1959) excavated House A, an early Thule house at Utqiagvik (Figure 13). House A is composed of a main room with an entrance tunnel; much of the wood was removed from the house. The main room floor had planks in the middle of the floor, although other planks may have been removed. The entrance tunnel was disturbed, but Ford estimates that it extended at least 9 feet (2.7 metres) to the south and was constructed using a whale jaw, scapula, skull, and wooden beams. Entry to the main room Figure 13: Utqiagvik, House A (from Ford was through a hole in the floor; evidence for this 1959:68) comes from the well-worn segment of an oval hole in one of the front floor planks.
4.4.6 Eastern Arctic
Moving east, houses of a similar style are found in early Thule sites across the Canadian and Greenlandic Arctic. In the Mackenzie Delta, House 8 at Cache Point was excavated and described in detail by Friesen and Betts (2006) (Figure 14). This house contains a main room, a long entrance tunnel, and a kitchen. The main room floor is partially paved with logs. The entrance tunnel was five metres long and met the house floor at a trap door. The kitchen did not have any wall structures but was defined by a hearth and heavily burnt soil. This house was occupied Figure 14: Cache Point, House 8 (from Friesen between 1200 and 1400 CE and represents an and Betts 2006:67)
98 early Thule Inuit occupation in the Mackenzie Delta.
Further east, in the central Canadian Arctic, Tiktalik and Nelson River are two other sites with well-preserved semi-subterranean house structures. Tiktalik House 5 has a main room constructed of log posts and adzed boards (Morrison 2000) (Figure 15). The tunnel was at least four metres long with a kitchen alcove to the east, identified by the burned sea-mammal oil-impregnated gravel surrounded by a series of post-holes. Located on Banks Island, Nelson River is often considered the best candidate for the earliest Thule occupations east of Alaska (Morrison 1989, 1999). The house is composed of two main rooms joined at a common tunnel entrance
(Friesen and Arnold 2008) (Figure 16). The larger room Figure 16: Tiktalik, House 5 (from floor is covered in planks and poles and has a raised Moody and Hodgetts 2013:8) earthen platform at the back. Adjoining this room is a separate external kitchen alcove. The second room also was paved in wood planks, although there is no raised area (Arnold 1994b).
Early Thule Inuit house structures in the Canadian Arctic have been the focus of much research, unlike in Alaska. Although the following houses are similar in plan to the Point Barrow/early western Thule style houses, they all have significant differences in construction material. Naujan and Malerualik were some of the first recorded by Mathiassen (1927a; 1927b) and follow the same early western Thule style. Pearce Point, Co-op (Morrison 2000), Vaughan, Jackson (Taylor 1972), Figure 15: Nelson River (from Friesen and Memorana (McGhee 1972), Hazard Inlet (Habu Arnold 2008:530)
99 and Savelle 1994; Savelle 1997; Savelle and Habu 2004), Talaguak, and Okivilialuk (Sabo and Jacobs 1980), to name a few, are all early Thule sites with early western Thule style semi- subterranean houses. Evidence indicates that the earliest of these were constructed out of driftwood and whalebone (McCartney 1977; 1979; Savelle 1986). Very quickly, construction materials were adapted to the local environment; with driftwood exceedingly rare in the central Arctic, most of the construction material was whale bones and stones, which altered the shape of the houses (Dawson 2001; Patton and Savelle 2006; Savelle 1986; Schledermann 1976b). However, space within the houses was still divided into a sleeping platform and a working floor area (Patton and Savelle 2006; Savelle 1997). Kitchen and cooking areas also changed, with later houses losing the external kitchen and incorporating it into the main room of the house itself (Maxwell 1981; Savelle and Wenzel 2003; Schledermann 1976b).
Some of the houses furthest from the Bering Strait are found at Skraeling Island and adjacent northwest Greenland. Despite their distance, they are also very similar to Alaskan house styles of early western Thule in the Bering Strait region (Holtved 1944; Schledermann 1978; Schledermann and McCullough 1980). Domestic dwellings have round to angular outlines and most have a separate kitchen entered from the main room and located parallel to the entrance passage. There are neither traces of a flagstone floor nor any clear signs of sleeping platforms, although some houses have a sandy floor at the back that is slightly raised above the main floor. The small kitchen space was accessed via a short tunnel; the hearths in the kitchen were raised, which is likely due to successive reconstruction and use (Schledermann and McCullough 1980). In addition to similarities in house style, similarities in artefacts, specifically ceramics, indicate that people may have transported these materials directly from Alaska (Schledermann and McCullough 1980). These houses may represent the direct movement of a group of Alaskan Thule Inuit into the eastern Arctic.
4.4.7 Summary of Early Thule Early Thule Inuit Semi-Subterranean Dwellings
Most early Thule houses use a similar floor plan: a main room with a raised back platform, a detached kitchen, and a long entrance tunnel. Although there are variations, the basic form remains the same from the Bering Strait to northwest Greenland. Similarity in form does not necessarily indicate that the activities in these areas were the same. Despite this, few studies have examined spatial patterns within Thule houses (but see Darwent and Foin 2010; Dawson 2001;
100
Friesen and Betts 2006). The current research aims to further investigate the activity patterns within a western early Thule dwelling. Understanding the use of space in the western Arctic will allow this work to be compared with past and future work on spatial patterning to see if use of area stays the same or changes across the Arctic and through time. The following section describes the house that is used for the spatial patterning analysis; the history of excavation at Feature 87 is outlined and then the architecture of the house itself is described.
4.5 Description of Feature 87 4.5.1 Feature 87 History
Feature 87, as it is called now, has a long history, starting with the original construction and occupation by Thule Inuit up to the full excavation and re-burial by the Cape Espenberg Project in 2011. Jean Schaaf conducted the first recorded excavations at Feature 87 during her work on the Bering Land Bridge National Preserve survey. Although the house was situated outside of the original survey area, the objective in 1988 was to record a part of the house that was undisturbed by previous digging and to establish the age and cultural affiliations (Schaaf 1988). It was unnumbered and placed within the incorrect site, KTZ-088, rather than the correct designation of KTZ-087. Schaaf’s original work identified a number of “looter’s”12 pits on the surface, along with scattered cultural artefacts. Examination of the surface materials led Schaaf (1988) to suspect the remains were from Birnirk groups.
Harritt (1994), reinvestigating Cape Espenberg, gave the house its first number (Feature 30) and excavated a two by one metre long trench. Although he located the house in the original site designation, he noted that it was provisional, pending further research in the area. Harritt’s (1994) excavations of Feature 87 (labelled Feature 30 in site KTZ-088) indicated that whale bones were present on the modern surface, as were small piles and scatters of bone and pottery. Excavations identified two activity areas: the upper-level and lower-level (Harritt 1994). The upper activity area was a sparse deposit of bones and artefacts. The lower level was identified as
12 Note that “looter’s” is a term from Harritt (1994). Use of the term “looter” has negative connotations; terms such as ‘subsistence digging’ have been used to identify people who dig for profit to supplement their way of life (see Hollowell (2006) for discussion). In the Arctic, digging for and collecting of material culture has a long history (i.e. Staley 1993). ‘Digger’ will be used to denote the previous excavations that have occurred in Feature 87 prior to recorded rchaeological excavations.
101 wooden roof deposits that overlay at least two surfaces with a large number of objects, including bone and pottery. Of the artefacts recovered, the most interesting was a section of an antler armour plate. Some of the pottery recovered also had curvilinear paddled surface decoration, which Harritt identified as a variant of Oswalt’s Barrow Curvilinear Paddled and Ahteut Curvilinear Paddled types based on temper and decoration. A calibrated date of 1272 CE (730±90 BETA 28011) was calculated based on a charcoal sample about 20 centimetres above the lowest deposit, but around 30 centimetres below the roof fall (Harritt 1994). Harritt concluded that the assemblage is consistent with the late prehistoric Western Thule, as defined by Giddings and Anderson (1986), dating to around 1200-1300 CE. Based on the whale bone visible on the surface, Harritt suggested that the occupants were whaling, as per the description of early (Western) Thule in Giddings and Anderson (1986). A new round of excavations of Feature 87 was initiated at the beginning of the Cape Espenberg Project in 2009 but did not uncover the floor. In 2011, Feature 87 was again excavated under the direction of Dr. T. Max Friesen, which uncovered an extremely well preserved house feature. In addition to excavation of the main room, the tunnel was also uncovered, although lack of time only allowed a portion of the tunnel roof fall to be removed. A burnt area, likely a kitchen, was excavated to the east of the tunnel, which dates to the same time as the main floor. Finally, a portion of the midden was excavated to the southwest of the tunnel. Radiocarbon dating was performed on six samples from different areas of the house including the kitchen area, midden, and main room; based on a 2- sigma calibration, five of the dates fall between 1275 and 1380 CE; a sample from the house fill gave a date in the early 15th century which likely indicates later use of the site after the house was abandoned (Table 1, Figure 17)
Sample ID Unit S Unit E Level Quad Species Element 14C Age 1 SD F87_1 14 2 2B - Rangifer tarandus radius-ulna 680 15 F87_2 14 2 2E SW Rangifer tarandus metatarsal 625 15 F87_3 6 2 2e - Rangifer tarandus metatarsal 535 15 F87_4 2 4 x3C SE Rangifer tarandus tibia 570 20 F87_5 4 2 x3C NW Rangifer tarandus metatarsal 650 20 F87_6 Kitchen charcoal 640 25
Table 1: Radiocarbon dates for Feature87
102
Figure 17: Calibrated radiocarbon dates from Feature 87
Feature 87 has a typical early Thule house floor plan; however, the preservation and completeness of the house is unique for the early Thule record in Alaska. The wall planks, floor boards, and structural supports were all preserved, as were the large back platform planks. In many houses, many or all of the well-made wood pieces are removed for reuse in other houses or as raw material for other tools. This suggests that later inhabitants of the site did not scavenge architectural elements in the house. This excellent preservation also permits a detailed description of the house, without resorting to the documentary record to fill in the blanks. A brief description is provided below.
4.5.2 House Description
Feature 87 (Figure 18) is a semi-subterranean wood structure. The sub-rectangular main room was 3.5 metres wide by 4 metres long and was composed of a wide platform at the rear that originally would have been raised, and two narrow benches on each side of the floor. The floor was covered with adzed wood planks that ran parallel to the tunnel between the three benches; under the two side benches the floor was cemented, likely with sea mammal oil. The eight-metre long sunken entrance tunnel joins the main room in the middle of the front wall. There is a kitchen area to the southeast of the main room that was likely used contemporaneously, although
103 the articulation with the house or tunnel is equivocal. Originally, a number of large baleen whale mandibles were thought to be part of the house structure, but excavation showed them to be later additions, likely part of a later rack structure.
4.5.2.1 Walls
Large upright posts were uncovered in each of the four corners of the main room. In between these complete posts, the walls were composed of split logs with their flat, adzed sides facing inwards. Between 10 and 11 split logs made up the back and side walls; the southwest side wall was cemented into sterile sand, likely through the use of sea mammal oil. The front wall was made of whole upright logs, with larger upright posts situated where the tunnel and front wall met. A buttressing log was found outside the southwest walls and at the north end of the back wall; excavations did not proceed deep enough to uncover any buttressing logs on the southeast side wall.
Although the tunnel was not fully excavated, large upright posts outside the tunnel buttressed the horizontal wall logs at the north end. The kitchen area did not have any discernible wall features, although there is evidence of a post near the centre.
4.5.2.2 Floor
As mentioned above, the floor of the house had a number of different features. The house was built on sterile dune sand, into which the house depression was dug and the floor and walls constructed. The floor under the back platform (3.5 metres wide by 1.5 metres long) was sterile, loose sand, covered in vegetation such as crowberry bushes, grasses, and mosses. There were also furs and hides between the sterile sand and the back platform planks.
The floor under the two side benches (just under 1 by 2 metres for the west bench and 0.5 by 2 metres for the east bench) was remarkably different; although the matrix was also sterile sand, it had been cemented in place, likely with sea mammal oil, and was rock solid. There was no evidence of vegetation or hides on the floor under either of the side benches. There were small pieces of wood under these benches right above the cemented sterile soil; the wood on the eastern edge was covered in a white and pink mould. There were two areas with evidence of burning under each of the side benches; under the west bench, the compact soil was burnt, while the east bench had burning on the floor planks (Figure 18).
104
The main part of the floor was covered in finely adzed, thin planks that ran parallel to the long axis of the house laid over sterile, loose sand (Figure 18). There were seven flat plank boards and one split log. All of these were placed parallel to the tunnel and were very closely fitted. At the front, there was a sill log interior to the front wall planks and posts.
4.5.2.3 Back Platform
The back platform was constructed out of five 3.5 metre long, thick adzed planks placed perpendicular to the tunnel and floorboards. Two thick logs were placed perpendicular to the planks (parallel to the floorboards) on either end of the platform. The west support may also have been supported with a sub-adult baleen whale humerus. The perpendicular supports were placed on thick, round buttressing logs at the front of the platform and the back wall. At the back wall, the supports were embedded in the wall between the wall logs. The buttressing logs were 3.5 to 4 metres long. The back buttressing log supported the back wall and the platform supports; the two buttressing logs at the front of the platform supported the platform supports and the side bench supports. A smaller log, also 3.5 to 4 metres long was placed in front of the two front buttressing logs and under the floor planks. Between the smaller log and the larger buttressing logs were upright posts, likely supports for the front of the platform.
4.5.2.4 Side Benches
The side benches were both encountered at a higher elevation than the back platform and floor and were distinct from the back platform construction. Both side benches are similar in width (just under 0.5 metres), with the west bench composed of one plank about 2 metres long and the east bench composed of one plank that has split about 1.5 metre long. The east bench width matches to the cemented sand width and has two post supports, one at the north and one at the south. The west bench is narrower than the cemented sand area; as well, the bench supports are at the edge of the floor/cement sand boundary, making the original bench potentially a metre wide. If this was the case, there is potentially one bench plank missing on the west side.
105
Figure 18: Plan view of Feature 87 and its location on the Seward Peninsula.
106
4.5.2.5 Kitchen Area
Once the fill was removed from Harritt’s previous excavations, there were a number of distinct burnt layers to the east of the tunnel. Excavation revealed an amorphous burnt area consisting of burnt sea mammal oil, crushed burnt bone, and charcoal that is approximately 2 by 2 metres. It sits about 2 metres to the southeast of the main room and about 1.5 metres from the tunnel. There are multiple layers of burnt material below the excavated surface, and so the kitchen area likely built up quickly, raising the floor above the other house floor levels. There is no concrete evidence of an articulation with either the main room or the tunnel. The pit Harritt originally dug probably obscured the visibility of the articulation to the tunnel or the house, although it is likely that it did in fact articulate with the house. The dates from the kitchen area are contemporaneous with the dates for the occupation of the house and not the fill layers. Although it may have been used at a different season in the open air, it is more likely that it was used as a kitchen area for the house during the cold season with a skin roof, such as those kitchen areas described from documentary sources in the Mackenzie Delta and the Bering Strait regions. Because of all of these considerations, this area is considered as part of the house itself.
4.5.2.6 Katak
The articulation of the tunnel and the main room was not fully uncovered. Excavation did uncover the entrance to the tunnel in the main room floor; however, the bottom of this entrance (at the tunnel level) was embedded in permafrost. In the main room, the floor planks around the entrance (katak) were rounded on the southern end, either intentionally to allow a smooth entry into the house or from use as people pulled themselves up from the sunken tunnel cold-trap. Once the floor planks were lifted, there was dark soil with artefacts, bones, and wood all sloped down to the tunnel at approximately a 45-degree angle, indicating an original hole through which people entered the main room. This hole was never fully excavated, but the organic soil matrix and artefacts went more than 0.5 metres below the other sterile sand of the rest of the floor and under the sill log at the front of the house.
The katak area in the tunnel was not fully excavated, but a few features were observed. The sill log, mentioned above, was on the inside of the front wall uprights and the entrance depression angled below this log. It is unlikely that people stepped over this log to exit, as knots and branch stubs were visible and sticking out on the superior surface. These protrusions would have made
107 moving across the log from the superior side inconvenient at best and harmful at worst; the smooth, rounded inferior surface would have allowed uninhibited access through the hole. At the base of the tunnel along the front house wall, there was a short log that potentially formed a step into the house itself; the log had been worn smooth through use.
4.5.2.7 Tunnel
The roof/sides were uncovered for the entire length of the 8-metre long tunnel, but only the 2.5 metres of the tunnel closest to the main room were excavated below the roof fall. It is not clear if the floor level was reached, as permafrost prevented further excavation. The tunnel would have acted as an excellent cold trap as it was built on the low side of the dune and excavated into the dune, which placed it well below the level of the main room floor.
4.5.2.8 Midden
The midden squares were excavated starting approximately one metre from the end of the tunnel. The full extent of the midden was not uncovered, but in two squares, the midden was excavated to sterile sand. Two samples, from an upper and lower layer, were radiocarbon dated and fall within the calibration curve for the main room occupation in the 14th century. Thus, the midden material is analyzed as a part of the house occupation.
4.5.3 Discussion
Overall, Feature 87 is in most ways a typical early Thule Inuit dwelling of the Point Barrow/early western Thule type. It consisted of a rectangular main room, long sunken entrance tunnel, and a raised back platform. The kitchen area was likely used at the same time as the household occupation based on radiocarbon dating. However, Feature 87 exhibits differences not seen in other Thule houses.
The most unusual feature is the occurrence of the side benches. One possible reason for the lack of side benches recorded in other archaeologically excavated houses is that the excavators may not have recognized these structures. Early archaeology often moved quite quickly and may have completely missed upper side benches during excavation; today, side benches may be mistaken as roof fall. Side benches have been identified in ceremonial (qargich) structures in the archaeological record; however, Feature 87 also has a wide back platform, which qargich do not
108 have. Preservation may also be a reason for lack of evidence of side benches; this feature is one of the better-preserved house features in Thule Inuit archaeology. The side benches were found in the matrix above the floor material, and likely were part of the active layer of permafrost which thaws and freezes each year; the other features likely were buried and integrated into the permafrost quickly, which would have led to better preservation. As stated above, not only was the house taphonomically well preserved, but also that there is no evidence of material scavenging, unlike most other early Thule houses. In other houses, there may have been side benches, but they may have been removed when houses were abandoned and/or rebuilt. Although excavation techniques, preservation, and scavenging cannot be completely eliminated as factors for the lack of side benches in other dwelling features, the associated change in floor under the two side benches in Feature 87 indicate that these areas were distinct from the main floor or back platform. Changes in flooring have not been recorded in other early Thule dwelling features, and so may corroborate the uniqueness of the main room layout of Feature 87.
The side benches are called ‘benches’ rather than ‘platforms’ as they more closely resemble the benches normally found in qargich than side platforms recorded ethnographically and in prehistory. Side platforms have been recorded at sites across northwest Alaska and into the Canadian Arctic. However, these are more similar to the wide sleeping platforms than narrow benches found at Feature 87. The earliest evidence of side platforms within semi-subterranean houses comes from the Norton settlements at Point Hope (Larsen and Rainey 1948). This platform arrangement became widespread in the Ipiutak period and continued to a lesser extent in the Birnirk period (Ford 1959). At Cape Krusenstern, Birnirk houses and the early Thule House 7 had single side platforms; these were made of multiple planks or logs and took up about half of the room (Giddings and Anderson 1986). Once house form diversifies in the late Thule period, side platforms are present in the archaeological and documentary record in different house types throughout the North American Arctic (Collins 1937; Young 2002). All of the published side platforms are thought to function like back platforms, as they are wide, made of multiple planks, and take up a significant amount of room space. They are not like the narrow, single plank side benches in Feature 87, which, like the platforms, may have provided warm areas for working, but would not have been practical for sleeping (Patton and Savelle 2006).
Narrow side benches like those excavated in Feature 87 are observed in the archaeological and documentary records from qargich. Qargich are often quite similar to cold-season domestic
109 dwellings; both are semi-subterranean with the main room constructed of wood, whale bone, or stone, and accessed by a long entrance passage (Dumond 1987; Murdoch 1892; Patton and Savelle 2006; Rainey 1947; Spencer 1959). They are identified archaeologically by their lack of sleeping platforms and the presence of side seating benches along the inner wall (Irving 1962; Patton and Savelle 2006; Savelle 1997; VanStone 1970). Most of our knowledge of qargich comes from the documentary record as only a few excavated qargich have been published; of the excavated qargich, only a few are from the early Thule period. The earliest excavated features that are considered analogous to ethnographically described qargich are found in western Alaska; Lutz (1973) excavated a Norton period ceremonial house at Ungalaqliq on Norton Sound, while Larsen (2001) excavated an Ipiutak ceremonial house from the at Deering on Kotzebue Sound. Excavations at the site of Utqiagvik uncovered a late prehistoric ceremonial house (Hall and Fullerton 1990b). VanStone (1970) also excavated a late prehistoric qargi at Akulivik, on the Nushagak River in Alaska. One early Thule qargi was excavated by Friesen at the Pembroke Site in the central Canadian Arctic (Norman and Friesen 2010); it conformed to ethnographic expectations as it was larger than the contemporary winter dwellings and had stones arranged as benches around the entire interior walls. One, and potentially two, Thule qargich were excavated at Hazard Inlet (Habu and Savelle 1994; Savelle 1987; 1997). The eastern Arctic has recorded a number of other Thule period qargich, such as those excavated at Qariaraqyuk (Whitridge 1999a; 1999b) and Ruin Island (McCullough 1989). Others have been reported from Barter Island on the North Slope (Jenness 1990) and the late prehistoric period at Crow Village in southwest Alaska (Oswalt and VanStone 1967).
Narrow side benches have been used archaeologically to identify qargich and their appearance in Feature 87 may indicate that this structure may have functioned as both a domestic and ceremonial structure. Alternatively, this house may just be a domestic dwelling with extra seating. Regardless of whether this house can be categorized as a domestic or hybrid domestic- ceremonial dwelling, the side benches mark the house as unique and may be tied to different activity areas than are normally found in Thule Inuit houses. Interpretations of the use of space within this house are affected by these unique structures within the main room.
110
4.6 Summary
Cape Espenberg is composed of beach ridges that have formed successively over the past 4000 years. On these beach ridges is one of the last remaining sequences of human occupation in northwest Alaska that has not been heavily impacted by climate change, sea level changes, or undocumented digging. Not only is the site preservation excellent, but also the continuous building of beach ridges has led to the use of different areas of the site during different time periods. This horizontal stratigraphy limits the reuse of house features as people returning to the site most often built closest to the coastlines, thus moving their dwelling places as ridges accumulated. Although Feature 87 was likely a palimpsest of a number of years of occupation, it still offers a somewhat synchronic view of the early Thule period in the area.
The geological, climatic, and ecological context of the area today have been described in order to establish a base line from which to compare the excavated materials. Feature 87 at Cape Espenberg dates to the fourteenth century, at the end of the tumultuous period in Arctic history where migrations occurred within the context of increasing warfare, increasing social complexity and concomitant regional social collapse, significant demographic shifts, and the rise and expansion of the Thule culture (Maschner and McGhee 2009). Although there is limited research on this period in this specific area, the local social context has been outlined as well.
The subsistence strategies for the broader Seward Peninsula and Kotzebue Sound region focus on small sea mammal and caribou hunting. These subsistence strategies date back to the Palaeoeskimo Tradition. At some periods, in ideal locations, whale and walrus hunting became the focus of some groups. However, the reliable small sea mammal resources of the Chukchi Sea and the caribou and other terrestrial resources of the interior Seward Peninsula allowed groups to establish long-term occupations in the region. Although each regional group likely had slightly different subsistence strategies, only people who inhabited Wales in the south after 1000 CE were able to procure large whales on a regular basis (Harritt 1994). Based on limited and coarse- grained faunal analyses, the subsistence strategies for the Thule at Cape Espenberg focused on small seals. The small drainages along the spit may have been important as they carried break-up flow, which accelerates the decay of land-fast ice and assists in seal hunting along the open water leads (fractures in the sea ice) (Mason and Gerlach 1995a). Currently, Cape Espenberg is used as a seasonal camp to hunt seals and birds, fish, and collect berries (Mason and Gerlach 1995a;
111
Wisniewski 2010). In addition to its economic draws during certain prehistoric periods, Cape Espenberg would have been of strategic importance as the northern extremity of the Seward Peninsula and its close proximity to Kotzebue and Cape Krusenstern, both which are visible from Cape Espenberg (Mason and Gerlach 1995a).
As previously described, Feature 87 was occupied during the 14th century and sits within the early Thule period when houses across the Arctic were fairly homogenous. Feature 87 is one of the best-preserved semi-subterranean dwellings in the western Arctic. Feature 87 conforms to many of the expectations about what an early Thule Inuit house at this time should look like, but also has some unusual features not found in many, if any, other published excavations of dwelling features. Like the other houses of this period, Feature 87 has a long, sunken entrance tunnel, a square main room that is divided between a back platform and front floor, and a separate kitchen area, likely attached to the main room. However, Feature 87 has side benches not recorded for domestic dwelling features elsewhere in the archaeological region. These side benches make Feature 87 unique in the region, while still conforming to the overall template of an early Thule cold-season dwelling. This review of early Thule house structures and the description of Feature 87 have established both the need for complete excavations at early Thule dwellings, and the need for fine-grained analyses of spatial use within early Thule dwellings. The unique form of Feature 87 makes the use of space within this dwelling important for understanding how people lived and engaged with their built environment in the early Thule period.
112
Chapter 5 Archaeofaunal Expectations
The previous chapters have given context for the expectations detailed below. Chapter 1 outlined the ways in which the expectations and final interpretations fit into previous analyses of spatial patterning for faunal remains in houses and sites, and how spatial studies have been conducted in houses. It also introduced the use of analogy in the archaeological record. Chapter 2 introduced the cultural context of the early Thule period. Chapter 3 drew upon general ethnohistoric, ethnographic, cross-cultural, and experimental records to sketch the Thule way of life, and then used the direct-historical approach to detail the construction, use, abandonment, and disturbance of houses in the northern Seward Peninsula and southern Kotzebue Sound region. Chapter 3 also gave a description of how the documentary records were constructed and analyzed the critiques for the northwest Alaskan documentary sources. As described in Chapter 3, although there are a number of sources for the Iñupiat in the Seward Peninsula region, the work of Ernest S. Burch, Jr. offers the most complete and in-depth analysis of the area. Most of the expectations are constructed from the detailed descriptions of Burch. Chapter 4 gave the general environmental, social, and archaeological context of the region, with specific attention paid to the history and architecture of Feature 87.
In the present chapter, the specific activities performed within the house will be mapped onto the floor plan of Feature 87, and expectations will be outlined. After a brief introduction on formulating expectations and the inherent assumptions associated with these expectations, the first section of this chapter describes the four stages that form the final expectations. Each stage draws on evidence from the ethnographic, ethnohistoric, ethnoarchaeologic, and taphonomic records in order to establish faunal expectations. Here, activities are defined as any cultural or natural process that interacts with the faunal record; these are similar to Schiffer’s formation processes (Schiffer 1983; 1987; 2010; LaMotta and Schiffer 1999) as described in Chapter 1, but are expanded to include all activities from initial deposition through excavation and analysis. The second section of the chapter details the activities for each stage that lead to the final expectations in each area. The first subsection describes the primary stage activities that bring bones to the site, which is the same for all areas. The subsequent subsections are divided by area: tunnel, main room floor, kitchen, benches, and midden. Secondary, tertiary, and quaternary
113 activities are described and summarized in associated tables; each table has a final expectations column that specifies the expected bone ratios and frequencies.
The purpose of formulating and testing simplified expectations is to unpack assumptions inherent, and often unidentified, in using documentary analogy and ethnoarchaeological studies to interpret the archaeological record. Because expectations are inherently simplified and model versions of the lives people enacted in the past, they are unable to completely reflect past activities. However, as stated in Chapter 1, analogical reasoning can be used to effectively test for gross similarities and highlight specific differences between the documentary and the archaeological records. In order to simplify the documentary assumptions and get at large-scale practices, the activities that occur in the house, the animal remains affected by these activities, and the taphonomic effects have all been reduced to the level of general patterns. By testing the assumptions made by using the documentary, ethnoarchaeological, and taphonomic records in interpreting archaeological materials, correlates that may be useful, obscured, or in need of further investigation can be identified
An important note is that the expectations assume that the house was left in a similar state to when it was originally abandoned; the quaternary expectations deal with some aspects of post- abandonment activities, but the history of the house shows little impact of intensive post- abandonment activities for the floor material. Even the later addition of whale mandibles into the fill did not affect the material on the floor, as the whale bones did not reach floor level. Another important note is that the floor material may not be related to the original construction; when houses are re-occupied, they were cleaned and re-furbished, thus potentially eliminating much of the previous years’ occupations (Ray 1983:180). Therefore, the floor material is likely only a snapshot of the final months, and at most year, of occupation in Feature 87. Conversely, middens in the documentary record were indicated as being an amalgam of multiple years of use, and thus will be a palimpsest of these years’ activities (Burch 2006; von Kotzebue 1821). This will be further investigated when comparing the expectations relating to the midden and house to the spatial patterning results.
One of the simplifications performed for this research is to construct stages from activities in order to detail faunal expectations. These stages describe simplified versions of the archaeological, documentary, and taphonomic records; in reality, the activities that occurred
114 within and between stages would have overlapped, interrelated, and comingled in complex ways. Despite this, the activities and their correlated faunal expectations have been divided into primary, secondary, tertiary, and quaternary stages based on the information available in the documentary and taphonomic records. The primary stage establishes expectations based on activities that initially brought faunal remains to the winter sites such as hunting, fishing, gathering, processing, and storage. These activities are directly derived from the documentary sources. The secondary stage defines expectations based on the visible activities recorded by the documentary sources that occurred within the domestic dwellings at winter sites such as butchery, storage, cooking, skinning, tool and garment manufacture, and discard. These expectations are derived most extensively from a close and narrow reading of the documentary sources, with the use of ethnoarchaeological and actualistic records when the documentary records are not explicit. The tertiary stage defines expectations based on contemporary activities to the secondary stage activities; these are separated from the secondary activities because they are not explicitly recorded by documentary sources but may obscure the patterns of secondary activities. These activities include carnivore activities, secondary burning, trampling, and cleaning. These may be briefly or obliquely mentioned in the documentary sources, but have been expanded and extrapolated using actualistic, ethnoarchaeological, and taphonomic studies. The quaternary stage develops expectations based on activities that occurred within the domestic dwellings and middens subsequent to occupation such as animal disturbances, weathering, and additional human actions. Some of these quaternary activities come from the documentary record, but most are derived from taphonomic studies.
Just as the activities are simplified into stages, the full range of available species is simplified by aggregating them into a smaller number of taxa identified in the documentary and archaeological records. The animal remains are divided into six categories: large marine mammal, small seal, large terrestrial mammal, small terrestrial mammal, bird, and fish. This division reflects both general taxonomic divisions and general transport and processing strategies for these taxa recorded in the Northwest Alaskan Iñupiat documentary sources (Burch 2006; Ray 1975a; 1975b; 1983). Large marine mammals such as whales, walrus, and bearded seal, although hunted differently, all underwent primary butchery away from the winter sites, typically with only the soft tissues such as meat, blubber, and skins transported to dwelling sties. Certain bones were transported back, as outlined below. Small seals (ringed, spotted/harbour, and ribbon) underwent
115 primary processing both away from the site and at the winter site (Burch 2006:153). Soft tissues were always brought back, as were most of the bones. Large terrestrial mammals such as caribou, moose, muskoxen, and bears underwent primary butchery away from the site. As with large marine mammals, meat, fat, skins, and other soft tissues were transported back to the camp; additionally, a high number of bones were transported back to the site. Small terrestrial mammals were habitually brought back whole to the winter village site and butchered there (Burch 2006:170). This category includes wolves, foxes, hares, ground squirrels, and minks. Like small mammals, birds were typically brought back whole to the village. Fish processing and transport generally removed most of the bones from the fish in the summer at the fishing and processing locations. In the winter, fish were not processed away from the site but were brought back whole (Burch 2006:146).
This chapter deals with many overlapping and complex ideas, assumptions, and expectations. Since my research goal is to test documentary assumptions and develop a methodology for assessing spatial patterning of faunal remains within domestic dwellings, the summary of expectations for all four stages will be separated and outlined by archaeologically identified areas. This summary is in the form of tables for each of the areas (Table 3, Table 4, Table 5, Table 6, and Table 7). The activities for each stage will be detailed in each section along with descriptions of the expected impact of these activities on the bones in that area; these expectations will be referenced in the table summaries. The primary stage activities will all be the same because these activities establish the baseline for the bones that enter the site. Secondary, tertiary, and quaternary stage activities in the tables will be described based on the archaeologically defined area and will be cumulative; for example, the activities described in the tertiary stage for the kitchen will be applied to the expected bone ratios and frequencies from the secondary stage in the kitchen to formulate the tertiary expectations. Final expectations are shown in the last column on each table; these will be the expectations against which the archaeological spatial data is tested. Outlining the primary, secondary, tertiary, and quaternary stages in steps will show the logical progression and assumptions inherent in formulating the final expectations. Similarities and differences between theses final expected frequencies and the archaeological record will be summarized and interpreted in Chapter 8.
116
5.1 Stages
This section describes the four stages of activities that lead to the four sets of expectations in each area of the house. Table 2 shows the frequency of the activities for each stage in each area of the house. Secondary Activities Tunnel Main Room Floor Kitchen Benches Midden Storage +++ + + Butchery +++ + ++ Preparation + +++ Consumption + ++ +++ Manufacturing + +++ Discard + + + +++ Butchery Marks + ++ ++ +++ ++ Burning +++ ++
Tertiary Activities Carnivore Activities ++ +++ Secondary Burning +++ + Trampling ++ +++ ++ + + Cleaning1 - / ++ - - / + - - - - - +++
Quaternary Activities Animal Disturbances + ++ +++ Weathering + +++ Human Actions + ++
1 - signs indicate removal of materials; + signs indicate addition of material Table 2: Activity frequencies
5.1.1 Primary Stage: Winter Village Faunal Assemblage Composition
In order to accurately define expectations, bones that typically entered the winter sites must be identified; from this baseline, expectations relating to the subsequent three stages of activities can be outlined. Although ethnographers and historians did not identify most of the bones that entered the site, with a few exceptions, the following account extrapolates from the meat or skin packages that are mentioned in the documentary sources and from the accounts of how these animal remains were used as tools, garments, or other artefacts. For example, there is no mention of the bacula from walruses being specifically removed from the carcasses; however, the fact that they were used as clubs and for raw material indicates that these must have been transported to the village (Burch 2006:169). In this section only, the categories are not divided based on areas, but on the taxa themselves. All areas of the house have the same primary expectations. After a detailed description of processing of each taxa category, the expectations are outlined. These correspond to the Stage 1 expectations in Table 3, Table 4, Table 5, Table 6, and Table 7. There
117 is also a “General” row where overall expectations (i.e. modification frequencies, fragment size expectations, general comparative frequencies) are outlined.
5.1.2 Secondary Stage: Household Activities
The expectations outlined in the secondary stage are based only on the activities that occurred in the house, taken from the documentary record. Not all of the bones brought to the site will be found in all the areas; bones not a part of the secondary activities in these areas are indicated in the Stage 2 columns of the summary tables. In most cases, the documentary record does not explicitly state which bones are affected, but rather gives an overview of the process the carcasses undergo or the result of the activities. Ethnoarchaeological and cross-cultural sources are used to help identify the specific bones that were a part of these recorded activities. Although the documentary record mentions many activities that occur within the house, this research focuses on those that are most likely to affect bones: storage, butchery, cooking, consumption, use in manufacture, and discard. Even these six sets do not encompass the complexity of activities within and around the house. Bones do not just enter the house and undergo one activity; most bones will move from one activity to another (e.g., storage to butchery). Comparative expectations based on these six activities are deliberately naïve and are based on the assumption that the bones are discarded at the final secondary activity location within the house. Tertiary stage activities, such as cleaning to a midden area, are described separately. The secondary stage expectations are cumulative; they are based only on the expected bones that are brought back to the site (Primary Stage expectations). These will show where bones should be if activities from the documentary record accurately predict the distribution of bones, and if no subsequent factors impact their frequencies. Although the activities overlap spatially, the documentary record and physical construction of the house suggest that activities can be generally associated with particular locations in Feature 87 (Table 2: Activity frequencies). For simplification, the archaeologically identified domestic space is associated with its most frequently occurring activity. The activities and their impact on the anatomical representation of the six taxa are described for each activity; modifications are outlined following the anatomical discussions. The only modifications for the secondary stage should be butchery marks and evidence of burning. Secondary stage expectations are found in the Stage 2 column for Table 3, Table 4, Table 5, Table 6, and Table 7.
118
5.1.3 Tertiary Stage: Contemporary Activities
The activities in the tertiary stage are contemporary to the secondary activities. They are separated from secondary activities because they are separate actions that can drastically affect the household activities. Additionally these activities are ‘invisible’ in the documentary sources; they consist of activities occurring at the time of occupation and that may be mentioned in passing in the records, but are not explicitly described. As outlined above, these activities are extrapolated from documentary references based on actualistic studies. The expectations outlined at the end are cumulative; thus, the tertiary activities are applied to the expected secondary stage remains to give the tertiary stage expectations. Changes or additions to the secondary expectations are indicated on the tertiary stage expectations both within the expectation descriptions and summaries at the end of the section. Only four activities are discussed here: carnivore activity, burning, trampling, and cleaning. The documentary and ethnoarchaeological evidence for each activity will be outlined followed by the effects these tertiary activities are expected to have on the secondary expectations. In many cases, the taxon specific anatomical representation will not change from the secondary to tertiary expectations; specific taxonomic expectations are mostly ratios and comparative frequencies, and so destruction or breakage of all bones will not necessarily alter the ratio. In some cases, the destruction will cause a different ratio (i.e. high ratio of high-density to low-density bones) to be expected or will remove some bones preferentially which will cause the ratio to change.
5.1.4 Quaternary Stage: Post-Abandonment Activities
Activities in the quaternary stage affect the faunal assemblage after the house has been abandoned. The three outlined here are animal disturbances, weathering, and human actions; they are derived from taphonomic studies and documentary sources. The general effect of these activities on faunal assemblages is described, followed by specific effects these activities would have on the tertiary expectations that have been outlined above. Again, often the ratios or frequencies from the tertiary expectations to the quaternary expectations do not change but are enhanced.
119
5.2 Expectations
This section describes how the activities affect the taxa for each area. The primary stage activities are organized slightly differently than the other three stages. For the primary stage, the bones of the six taxa are described in detail. These are then outlined in each of the tables (Stage 1 column in Table 3, Table 4, Table 5, Table 6, and Table 7). The secondary, tertiary, and quaternary stages are described separately for each of the areas so that the expected changes from stage to stage are easy to follow. The overall effect of the tertiary and quaternary stage activities is described for the first area (the tunnel), and then the effects of these activities on the other areas are referenced back to the tunnel area. Since different activities will often produce similar results, the expectations are summarized at the end of each area-stage section. There is table for each section summarizing the expectations. For this chapter, all activities and changes in bones are described in detail in order to highlight the logical steps and assumptions to create the final expectations; archaeological results of these expectations are discussed in Chapter 8.
All expectations are summarized in Table 3, Table 4, Table 5, Table 6, and Table 7. All primary expectations are given in the Stage 1 column from Expectation A to J. As described below, some expectations are overwritten by actions that affect the entire taxonomic category, while other faunal remains are not expected to be present at all in certain areas. These expectations are outlined and summarized in the tables. Additionally, secondary, tertiary, and quaternary activities can add expectations to an area; all expectations added beyond the primary expectations are given letters after J. If the expectation is related to specific taxa (e.g. low frequencies of fish in the tunnel), they are placed out of alphabetical order in the correct taxonomic category.
5.2.1 Primary Stage for All Areas
5.2.1.1 Large Marine Mammals
As described above, large marine mammals underwent initial processing prior to entering the village (Burch 2006:170). This category includes baleen whales, beluga whales, walruses, and bearded seals. Generally, the flippers were the only common part to enter the site, as most butchery was done on the ice or shore where the animal was killed (Burch 2006:149, 153, 159, 164, 167). Large baleen whales were rare at Cape Espenberg, but when they were acquired in the Chukchi Sea, they were skinned, fleshed, and butchered close to the shore (Burch 2006:159).
120
The baleen, ribs, and jaw bones would have been transported back to the village for construction and household material (Burch 2006:159). Additionally, front flippers would have been brought back to the village (Burch 2006:162–163). The smaller beluga whales were hunted in Kotzebue Sound and processing took place close to the edge of open water (Burch 2006:164–165). Beluga skin with the blubber attached was carefully prepared and stored (Burch 2006:159), which introduced forelimb flippers to the village as they were kept with the skin. Other than the forelimbs, there is no mention of beluga bones used for other purposes or transported to the site in the regional documentary records. Walrus were often butchered near their kill sites; when they were on ice or in the open water, it was done quickly, but when they were killed on the beach, processing occurred more thoroughly and slowly (Burch 2006:167). In most cases, the head, forelimb, and hind flippers were removed immediately when the carcass was skinned (Burch 2006:168). Flippers were frequently brought back to the village sites to ferment for a few days, while the other elements were discarded (Burch 2006:168). Other than the flippers, bacula and ivory were coveted items for tools and raw materials that would have almost always been transported to village site (Burch 2006:169). For bearded seals, the flippers were removed and transported with the skin which was then processed at the residential site for boot soles, boat covers, ropes, or as containers for meat (Burch 2006:154). Meat was removed and dried, with only the hind flippers sometimes kept intact to ferment and be consumed a few days later (Burch 2006:154–155). In summary, for whales and large marine mammals, the entire animal was never brought back to the winter village; processing occurred where it was easiest, with the soft tissues carefully prepared and transported back to the site. In general, the only bones that would make it back to the site were hind- and fore-flipper bones (Expectation A) and potentially some specific bones such as walrus or bearded seal bacula, or whale ribs and mandibles (Expectation B).
5.2.1.2 Small Seals
The most economically important mammals in the documentary accounts for the area were the small seals. Although often heavy, they were relatively easy to transport over the snow and ice due to their torpedo-like shape; thus, many small seals were brought back whole to the villages (Burch 2006:153). When small seals were not brought back to the village, they were processed by groups of women along the ice edge or coast near the village (Burch 2006:153). Here, primary processing of the animal included removing and discarding the head and hind flippers at the butchering locale. Skins, with the fore flippers attached, were transported back to the winter
121 site, along with the meat and blubber (Burch 2006:154). Most of the small seal skeletal remains should therefore be present on the site (Expectation C), with a slightly lower frequency of heads and hind flippers (Expectation D).
5.2.1.3 Large Terrestrial Mammals
The two sets of large terrestrial mammals in the region are ungulates (caribou, moose, and muskox) and bears (polar and grizzly). Although the purpose for acquiring terrestrial mammals affected the butchery practices, generally only the high-utility meat packages were brought back to the village. Neither moose nor muskox processing is described, so the data for ungulates comes from the vast amounts of information available for processing caribou (Binford 1978a; Burch 2006; 2012). Caribou were mostly processed away from winter village sites, although this varied greatly, with caribou killed close to the winter village brought back and processed there. The first stage in processing late fall or winter caribou at any location was to cut off the head around the shoulders and remove the legs (Binford 1978a:48; Burch 2006:138). In the late fall, the caribou carcasses were deboned and frozen where they were killed; if the temperature was not low enough to freeze, the carcass was deboned at the kill site and meat was cut into thin strips and dried (Burch 2006:138). When caribou were killed away from the site, the carcasses were cached and retrieved at a later date (Binford 1978a:48, 50). In the winter, frozen large terrestrial mammal packages would have been brought into the house to thaw prior to secondary butchery or preparation (Burch 2006:138). Most of these packages were likely composed of overall high-utility bone elements. Antler, although not specifically recorded as returning to the village site, must have been regularly brought to the site since it is a prime raw material for tools (Burch 2006:140, 202–203). Caribou lower leg bones were also removed from the carcass and preferentially brought back to the village site to be smashed and boiled for consumption (Binford 1978a:157; Burch 2006:138). In most cases, rib cages were left for the dogs (Burch 2006:138). Bears were generally butchered and transported in the same way as other large mammals (Burch 2006:170). Burch (2006:170) specifically mentions that the bear heads were never brought back to the village, but were either buried or impaled on a small tree or shrub. Teeth and claws were used as tools and therefore transported to the village (Burch 2006:170). Generally, villages will have high frequencies of large terrestrial mammal high-utility elements compared to low-utility elements (Expectation E) (Burch 2006; cf. Binford 1978a); however, despite the low-utility of lower leg and extremity bones, these elements were valuable to people for consumption or tool
122 manufacture. There would be high frequency of lower limbs relative to whole skeletal elements (Expectation F).
5.2.1.4 Small Terrestrial Mammals
In contrast to large terrestrial mammals, small terrestrial mammals13 are habitually transported whole to the site (Burch 2006:170). This category includes marmots, ground squirrels, muskrats, hares, beavers, foxes, mustelids (i.e. wolverines, otters, martens, minks, and weasels), wolves, lynx, and porcupines. Despite these species being recorded in the southern Kotzebue Sound and northern Seward Peninsula area, not all of these were available around Cape Espenberg. Of the small terrestrial mammals, marmots, ground squirrels, and hares were used mainly for their pelts, although they were also eaten regularly (Burch 2006:172–174). Foxes and wolves were rarely eaten, and were mainly caught for their pelts (Burch 2006:175, 176). Beavers were mainly caught for food; their teeth and claws were also used as tools (Burch 2006:175). Otters, minks, martens, and lynx were caught for both food and fur, while wolverine and weasels were hunted mainly for their fur (Burch 2006:175). Porcupines were mainly eaten, but they were also used for other things; in particular, porcupine bones were cleaned and saved to help cure aching joints (Burch 2006:176). Different processing and consumption of these animals may affect their skeletal distribution, but in general, the entire skeleton should be represented equally at the site (Expectation G).
5.2.1.5 Birds
Ethnohistorically, around 80 species of birds were used in the northwest Alaska region (Burch 2006:177–178). In the majority of cases, whole bird carcasses and collected eggs were transported to the village site and into the house for butchering and processing (Expectation H).
5.2.1.6 Fish
Although not always visible archaeologically, as Burch (2006:140) states “[f]ish were an important source of food for both people and dogs in all parts of the study region [Northwest Alaska]”. Fish caught in the winter were frozen and cached immediately without any processing;
13 This category includes both medium- and small-bodied terrestrial mammals, but is shortened to ‘Small Terrestrial Mammals’ for brevity.
123 these would then be eaten by dogs and humans frozen, thawed, or boiled whole (Burch 2006:146). Due to the lack of processing and quick freezing, fish caught in the winter likely entered the house whole. Sometimes fish heads were collected and boiled to produce fat, which was stored and consumed in the winter (Burch 2006:146). Summer-caught fish were mostly dried, with their heads and spines removed (Burch 2006:146). Fish caught in the summer were also putrefied in various ways; in almost all cases, heads and spines were removed after a period of fermentation (Burch 2006:146). Thus, most summer-caught fish would have only rib bones brought back from the fishing location. Although fish remains may not make it back to the site in the same frequency that fish were consumed, when they were caught in the winter, all parts of their skeleton should be represented. For a winter house, there may be slightly elevated frequencies of ribs, but overall the expectation would be a fairly equal ratio of cranial and post- cranial remains (Expectation I).
5.2.1.7 General
The only modifications present on the bones at this stage would be a low frequency of butchery marks from primary butchery (Expectation J).
5.2.2 Tunnel Expectations
5.2.2.1 Primary Activities
See 5.2.1 Primary Stage for All Areas (Table 3, Stage 1 column).
5.2.2.2 Secondary Activities
(Table 3, Stage 2 column)
5.2.2.2.1 Storage
In-house storage would be mostly in specific small rooms, sub-floor caches, or the tunnel (Burch 2006:97, 216); for Feature 87, the tunnel is the only potential area for medium or large amounts of storage in the house (Table 3). Bones would only remain in the storage area if they were left during abandonment or forgotten during the house occupation; on most other occasions, meat packages and associated bones would be moved through the house and be discarded in other areas. Most of the bones from animals that entered the site may be originally placed in storage in
124
Stage 1: Stage 2: Stage 3: Stage 4: Taxonomic Village Faunal Household Activities Contemporary Post-Abandonment Final Expectations Category Composition Activities Activities Large marine A. High ratio of A. Stored as whole A. Increase in ratio; A. Slight increase in A. High ratio of mammal appendicular : axial package; no changes still high ratio; still high appendicular : elements axial elements B. High frequencies of B. These bones not B. N/A B. N/A B. Bacula, whale ribs, bacula, whale ribs stored in tunnel and mandibles not and mandibles expected to be present in the tunnel Small seal C. Even skeletal C. Stored as whole C. High ratio of dense C. Slight increase in C. High ratio of distribution package; no changes : less dense ratio; still high dense : less dense elements elements D. Low frequencies of D. Stored as whole D. Low frequencies of D. No change D. Low frequencies of head and hind package; no changes hind flippers hind flippers flippers Large E. High ratio of high- E. Stored as whole E. Even ratio of high- E. Slight increase in E. Even ratio of high- terrestrial utility : low-utility package; no changes utility : low-utility high-utility : low- utility : low-utility mammal elements elements utility bones elements F. High frequency of F. Stored as whole F. Increase in ratio, F. Slight increase in F. High frequency of lower limbs package; no changes still high ratio; still high lower limb Small G. Even distribution of G. Stored as whole G. High ratio of dense G. Slight increase in G. High ratio of terrestrial skeletal elements package; no changes : less dense ratio; still high dense : less dense mammal elements elements Bird H. Even distribution of H. Stored as whole H. High ratio of dense H. Slight increase in H. High ratio of skeletal elements package; no changes : less dense ratio; still high dense : less dense elements elements Fish I. Even distribution of I. Fish not stored in I. Intermediate I. Slight increase in I. Intermediate cranial : post-cranial house frequency of frequency; no frequency of elements vertebrae comparative change vertebrae N. Low frequency of N. Low frequency of N. Low frequency of fish remains fish remains fish remains General J. Low frequency of J. No further butchery; J. No further J. No further butchery; J. Low frequency of butchery marks no change butchery; no change no change butchery marks K. Low total number of K. Intermediate K. Slight increase; no K. Intermediate total specimens (NSP) total number of comparative change number of specimens (NSP) specimens (NSP) L. Intermediate L. Slight increase; no L. Intermediate frequency of comparative change frequency of indeterminate indeterminate bones bones M. Intermediate M. Slight increase; no M. Intermediate frequency of comparative change frequency of carnivore marks carnivore marks Table 3: Summary of expectations from the tunnel. Note that all expectations are cumulative; the activities from each stage only affect the bones present in the previous stage.
125 the tunnel; however, since the tunnel is more accessible than other storage areas (i.e. behind the house racks, etc.), it may be where meat packages and bones were stored for convenient access during the winter months (Burch 2006:97, 216). The packages specifically recorded as being stored for short periods of time are large marine mammal flippers (Burch 2006:154–155, 159, 162–163, 168); the tunnel is the most logical place for them to be stored in Feature 87 due to its proximity to the main room. The only taxa unlikely to be stored here are fish; these are specifically described as being stored outside of the house (Burch 2006:146, 147) (Expectations I and J). Additionally, the large bones of large marine mammals (whale ribs and mandibles) are unlikely be stored in the narrow confines of the tunnel. Although these were used in tunnel construction in the documentary sources (Burch 2006:159), they are not part of the structure of Feature 87 (Expectation B).
Otherwise, the tunnel should represent all the bones brought to the site, as it may be a storage area for most of these remains (Expectations A, C, D, E, F, G, and H). The remains are not expected to undergo any other butchery or processing prior to being stored in the tunnel. Butchery marks should be low since most bones will have only undergone one stage of processing (Expectation J); there should be no evidence of burning.
5.2.2.3 Tertiary Activities
(Table 3, Stage 3 column)
5.2.2.3.1 Carnivore Activities
Carnivores, specifically dogs, had a potentially immense impact on faunal assemblages. Ethnographically, keeping dogs from eating artefacts, stored food, and even the dog ties themselves was a perpetual problem (Burch 2006:147, 283, 284). For the most part, dogs were kept outside, tied away from storage areas, houses, and each other (Burch 2006:284). In some cases, dogs were allowed to run free (Kotzebue 1821:199-200). Dogs were fed a variety of foods and were given bones to chew on; additionally, dogs would often scavenge the garbage of households and any other faunal substance they came across (Burch 2006:26, 138,140, 156, 159, 165, 169,174, 269, 283, 284); foxes, in the absence of dogs, would have the same type of effect on assemblages but with a much reduced intensity. In some circumstances, such as very cold weather or when puppies were born, dogs were brought into the house (Burch 2006:284). Most
126 of the documentary data suggests dogs and puppies were kept in the tunnels when they were allowed in the house (Burch 2006:284). In these areas, meat for human consumption was likely stored away from the dogs, but discarded bones were likely free for the chewing.
Carnivore activity will have two effects on the faunal assemblage: the physical destruction of bones and direct indicators of carnivore activity such as punctures, gnaw marks, and acid etching. Overall, the effects of carnivore-mediated destruction will differentially affect bones. Axial bones and cancellous bones that are less dense and full of grease (Binford 1978a; Lyman 1994; Lam et al. 2003) will be more significantly affected for all mammals than cortical bone (i.e. mid-shaft fragments) and dense bones (i.e. femurs). In general, large mammal bones are denser than small mammals, with birds and fish bones often even less dense and hence more susceptible to destruction and elimination through carnivore gnawing (Faith et al. 2007; Lam et al. 1999; Lam et al. 2003; Lam and Pearson 2004; 2005; Lyman 1984; Morlan 1994). This type of attrition that is correlated with bone density is called density-mediated destruction and will be used to refer to all the destructive effects of carnivore action on bone.
Tunnels are expected to have intermediate frequencies of secondary carnivore activity: high in comparison to other areas of the house, but lower than the exposed midden. Dogs may have been brought into the tunnel in the winter, but were unlikely to have consistent access to the faunal assemblage in this area.
5.2.2.3.2 Secondary Burning
Secondary burning, different from the primary burning described for the kitchen (Section 5.2.4.2.1) is burning of bone that is not being prepared for consumption but rather burned during the process of cooking, heating, or lighting. This secondary burning can be either accidental by being too close to the fire or intentional as a source of fuel. Burning will alter the bone assemblage in two ways: through colour, and potentially size, changes and through the breakage of bones into smaller pieces (Clark and Ligouis 2010; Conard et al. 2008; Costamagno et al. 2010; Morin 2010; Shipman et al. 1984; Théry-Parisot et al. 2005).
Ethnographically, tunnels were rarely the location for either cooking or lighting, thus they are expected to have no to low amounts of burning and thus no effect on the expectations (Table 2: Activity frequencies).
127
5.2.2.3.3 Trampling
Everyday movement of people through the house and around the site will have physically broken bones at the time of occupation. This is likely enhanced by the small space in these houses that restricted movement to certain areas. Although trampling of remains is not indicated in the documentary record, it can be expected to be highest in areas that were most well-used in the winter (the main room floor), medium in areas of less frequent use (kitchen and tunnel), and low in areas of limited movement (benches) or with fewer spatial constraints (midden) (Table 2: Activity frequencies). Although both the benches and midden would experience physical pressure, the passive act of sitting and sleeping on the benches leads to a low frequency of physical destruction, while the unrestricted nature of the midden, and the fact that it was a garbage heap, likely mitigated the flow of people. For all of the bones, the physical destruction would have broken large pieces into smaller pieces.
As stated above, the tunnel likely had medium amounts of trampling activity, thus reducing some large bones into smaller ones.
5.2.2.3.4 Cleaning Activities
Although Arctic dwellings were often deemed by Europeans to be smelly and small, their neatness was clear in the documentary records of the region (von Kotzebue 1821:200; Ray 1975a). How this cleaning occurred is not the subject of documentary records, although some aspects of disposal and the artefacts used for cleaning are recorded, as seen below. Alternatively, ritual or ideological reasons for cleaning, which are often a part of worldview, are not normally mentioned in the northwest Alaskan documentary record. Cleaning likely affected all areas of the house; in some areas there was the removal of remains, while in others there was the addition of cleaned remains (Table 2: Activity frequencies). Cleaning is last in the discussion of contemporary activities, since it likely affected the frequencies of all tertiary activities. Cleaning would have removed both large and small sized bones, but likely through different actions in different spaces.
The tunnel was likely periodically cleaned, especially of large remains, although potentially to a lesser extent than any of the other interior spaces since it was not habitually lived in. The tunnel may have been the repository of some of the cleared bones, especially for small remains swept
128 from the main room floor. Additionally, small remains, most often cleaned via sweeping (Burch 2006:268), were probably not frequently removed from the tunnel.
5.2.2.3.5 Tertiary Expectations Summary
(Table 3, Stage 3 column)
The tunnel should have intermediate frequencies of carnivore, trampling, and cleaning activity, with little to no burning (Table 2: Activity frequencies). This means there should be a comparatively intermediate frequency of direct indicators of carnivore activity on bones (Expectation M) and medium destruction of bones into indeterminate pieces (by carnivore gnawing and trampling) in comparison with the other areas (Expectation L). There should be an intermediate frequency of bone fragments (NSP) in the tunnel compared to the rest of the house (Expectation K).
Intermediate frequencies of carnivore gnawing and trampling will result in the destruction of less dense bones in comparison to denser remains. In general, axial bones (vertebrae and ribs) are less dense than limb and extremity bones. For large marine mammals, the high ratio of extremity to axial bones should continue to be strong due to the preferential destruction of axial (less dense) bones (Expectation A). For small seals, the destruction of axial elements from Stage 2, Expectation C will remove the axial and cranial elements, leaving the dense bones (Stage 3, Expectation C). The only exception to this should be the low frequency of hind flippers, which are high-density elements, but already in low frequencies (Expectation D). For large terrestrial mammals, the high ratio of high-utility to low-utility should be reduced and should become an even ratio, as many high-utility bones (i.e. vertebrae and ribs) are also low-density (Expectation E). The second expectation for large terrestrial mammals should increase the frequency of lower limbs (dense); however, this frequency is relative in comparison with the other areas, so the expectation stays the same (Expectation F).
Small terrestrial mammals will have most of their skeletal elements removed due to medium amounts of carnivore activity and further reduced with medium amounts of trampling; there should be a high ratio of dense to less dense bones (Expectation G). If birds are not totally destroyed by carnivores, they will also show a positive correlation with bone density (Expectation H).
129
Most large marine mammal bones, as describe in the secondary activities, are not stored in the house and so are not affected by tertiary activities (Expectations B). Although fish are not stored in the tunnel, they may enter the tunnel during cleaning activities: both carnivore gnawing and trampling will likely destroy most of the fragile fish cranial and rib remains, leaving the denser vertebrae in intermediate frequencies compared to other areas of the house (Expectations I). Overall, fish remains are expected to be present in only very low frequencies (Expectation N).
5.2.2.4 Quaternary Activities
(Table 3, Stage 4 column)
5.2.2.4.1 Animal Disturbances
Rodents and carnivores cause most animal disturbances in the area. Rodents offer less disruption than carnivores. Animals such as arctic ground squirrels and lemmings can burrow into the house and fill, and can often move bones around (from floor to surface, throughout the fill, etc.) in small quantities. This movement likely affects medium and small bones. They can also gnaw on bones, leaving distinct bone modifications on the remains. Additionally, they may die in the burrows, thus adding remains to the faunal assemblage not acquired by the humans who lived there. The effect of rodents, either through modifications, bone movement, or addition of bones, is likely to be limited in the case of Feature 87, since taphonomic analysis indicates it was buried quickly and deeply (Chapter 4).
Larger animals, such as canids, may also enter or dig into houses, displacing material from its pre-abandonment context. Canids such as foxes and dogs may modify bones by gnawing, similar to the effects described for the tertiary set of activities. For this stage of canid activities, the amount of gnawing within the house depends on the amount and type of collapse. Destruction and gnawing may be extremely variable depending on the entrance point of canids; however, the current assumption is that in an un-collapsed house, although the wet tunnel was the easiest point of entry (low frequency), the warm main room likely attracted canids (medium frequency; see Table 2: Activity frequencies). Based on the previous discussion of the house, very little disturbance, either human or canid, likely occurred after abandonment, so there was likely no subsequent canid disturbance to the house after collapse, although any exposed midden would be subject to destruction and modification. Even prior to collapse, the main room and tunnels often
130 fill with water in the warm months and freeze in the cold months, making the abandoned houses inhospitable to canid scavengers.
The tunnel is expected to have low amounts of carnivore activity in the quaternary stage.
5.2.2.4.2 Weathering
Weathering has more predictable effects on bones than either animal or human action; it typically affects the most exposed, least dense bones, although there are numerous exceptions. Weathering results in flaking and bleaching of bones exposed to the elements, and eventually can result in the physical destruction of bones (Behrensmeyer 1978; Andrews 1995). The remains in the midden will be most affected by weathering, with those remains in the main room and tunnel being least affected because they are buried quickly (Table 2: Activity frequencies). Kitchens and benches, such as at Feature 87, are sometimes at a higher level than the main room floor or tunnels and have less overburden, which may make their assemblages more susceptible to weathering. High-density bones, such as mammal cortical bones, are much more likely to survive than low-density bones, such as mammal cancellous bones, fish, and bird remains.
The tunnel is expected to have experienced no weathering activities.
5.2.2.4.3 Human Actions
Humans are probably the greatest disturbers of abandoned houses as they scavenge construction material for building other houses, dig into the house for artefacts and raw materials, and finally fully excavate the house. Although initial scavenging may remove architectural and artefactual materials, they are less likely to intentionally remove objects from the faunal assemblage. When architectural features such as bench or floor planks are the objective rather than artefactual objects, the movement of faunal material may be much greater and less predictable. Architectural scavenging is unlikely for Feature 87 as described in Chapter 4, since most of the important and normally curated wood pieces were left intact. Due to lack of evidence for architectural scavenging in Feature 87, human impacts are limited to moving through the house to scavenge other material; these activities may cause some trampling and movement of remains, although both are likely to be negligible. Human trampling on the midden is likely the highest, with low amounts in the tunnel (Table 2: Activity frequencies). Finally, subsequent looting and sanctioned excavations described in Chapter 4 had the highest impact on the faunal assemblages. Luckily,
131 only the 2011 excavations reached the floor material, leaving most of the occupation surface intact. However, after the 2011 excavations, the entire faunal assemblage has been removed, likely resulting in the greatest change to the history of the assemblage since its deposition. Identification and analysis of this material are described in Chapter 6: Methodology.
The tunnel is assumed to have low human trampling activity.
5.2.2.4.4 Quaternary Expectations Summary
(Table 3, Stage 4 column)
The tunnel was not greatly affected by post-depositional activities. Low amounts of carnivore and human activity may have resulted in some density-mediated attrition. This would enhance many of the ratios and frequencies already recorded for the tertiary stage (Expectations A, C, E, F, G, H, I, K, L, M, and N). Because the quaternary activities are so low for the tunnel, there is no change in the expected bone ratios, comparative frequencies, or modifications from the tertiary expectations. Any change that does occur is by degree, not kind, and is likely unrecognizable in relational analyses. Expectation B is not expected to be present in the tunnel.
5.2.3 Main Room Floor Expectations
5.2.3.1 Primary Activities
See Section 5.2.1. Primary Stage for All Areas (Table 4, Stage 1 column).
5.2.3.2 Secondary Activities
(Table 4, Stage 2 column)
5.2.3.2.1 Butchery
Although some butchery of animals transported to the site would have occurred outside the domestic dwellings (see Section 5.2.6.2.1. Midden Discard), in the winter when carcasses were frozen, much processing would have occurred within the warmth of the dwelling (Burch 2006:138). The only space warm and large enough to accommodate regular butchery in Feature 87 is the main room floor.
132
Stage 1: Stage 2: Stage 3: Stage 4: Taxonomic Village Faunal Household Activities Contemporary Post-Abandonment Final Expectations Category Composition Activities Activities Large A. High ratio of A. Large marine A. Low frequency of A. No change; still low A. Low frequency of marine appendicular : axial mammals not extremities frequency extremities mammal elements butchered on main room floor B. High frequencies of B. Not expected to be B. N/A B. N/A B. Bacula, whale ribs bacula, whale ribs present on main and mandibles not and mandibles room floor expected to be present on main room floor Small seal C. Even skeletal C. High frequency of C. High frequency of C. Increase in C. High frequency of distribution heads and hind mandibles and hind frequency, still high mandibles and flippers flippers hind flippers D. Low frequencies of D. Overwritten by D. N/A D. N/A D. Overwritten by head and hind Expectation C Expectation C flippers Large E. High ratio of high- E. Low ratio of high- E. Low ratio of high- E. Decrease in ratio; E. Low ratio of high- terrestrial utility : low-utility utility : low-utility utility : small, low- still low utility : small, low- mammal elements elements utility elements utility elements
F. High frequency of F. High ratio of F. Increase in ratio; F. Increase in ratio; F. High ratio of lower limbs phalanges : whole still high still high phalanges : whole limbs limbs Small G. Even distribution of G. Slightly high G. Slightly high G. Increase in G. Slightly high terrestrial skeletal elements frequency of heads frequency of heads frequency; still high frequency of mammal mandibles Bird H. Even distribution of H. High frequency of H. Increase in H. Increase in H. High frequency of skeletal elements lower limbs frequency; still high frequency; still high lower limbs Fish I. Even distribution of I. Fish not butchered I. Intermediate I. Slight increase in I. Intermediate cranial : post-cranial on the main room frequency of frequency; still frequency of elements floor vertebrae intermediate vertebrae L. Low frequency of L. Reduced; still low L. Low frequency of Fish fish General J. Low frequency of J. Intermediate J. Reduced, but still J. Reduced, but still J. Intermediate butchery marks frequency of intermediate intermediate frequency of butchery marks butchery marks K. Low total number K. Intermediate total K. Increased, but still K. Intermediate total of specimens (NSP) number of specimens intermediate number of (NSP) comparatively specimens (NSP) M. High frequency of M. Increase in M. High frequency of indeterminate frequency, but still indeterminate bones high bones N. Intermediate N. Intermediate frequency of frequency of carnivore marks carnivore marks Table 4: Summary of expectations from the main room floor. Note that all expectations are cumulative; the activities from each stage only affect the bones present in the previous stage.
133
Large marine mammals were likely not butchered within the house due to their size (Burch 2006:170). The parts of the large marine mammals that most frequently occur at the site, the extremities, are known to be directly consumed after storage (Burch 2006:154–155, 162–163, 167, 168). Since the main floor is not used primarily for either storage or consumption, large marine mammal bones, including appendicular elements, are likely to be in low frequencies, if present at all (Expectation A). Other large marine mammal bones may have been used in the construction of the main room; for example, at Feature 87, a subadult baleen whale humerus is used for buttressing one of the back bench supports. Bones used as raw materials for tools (i.e. bacula) were likely moved from storage to the areas of tool manufacture and did not accumulate on the main room floor (Expectation B). As a result, there are not expected to be many large marine mammal remains on the main room floor.
Small seals, if processed within the house, were likely butchered on the main room floor. Heads and hind flippers, if still attached and not removed outside, were likely removed and left on the floor. Other low-utility elements, such as front flippers likely stayed attached to either skins or high-utility meat packages and were moved onto either the benches for manufacturing garments or into the kitchen for further preparation, respectively. Thus, heads (cranium and mandibles) and hind flippers (metatarsals and phalanges) are expected to be discarded on the main room floor in higher frequencies than other small seal elements (Expectation C, overwrites Expectation D).
Large terrestrial mammal packages may have been brought directly into the main room from the storage areas in order to be butchered (Burch 2006:138, 145–155, 162–163, 168). These packages would have had specific low-utility bones attached to high-utility bones (i.e. on caribou limb packages, high-utility humeri and radio-ulnas are often attached to low-utility metacarpals, carpals, and phalanges) (Binford 1978a:40). During butchery on the main room floor, these packages were separated into elements, with the high-utility bones being moved to the kitchen (preparation) or bench (consumption) areas; low-utility bones would be discarded on the main room floor, resulting in high ratios of low-utility to high-utility elements (Expectation E). Lower limbs (carpals/tarsals, metapodials, and phalanges) will have had low-utility elements (phalanges) removed and discarded on the main room floor, with the metapodials being moved to the kitchen; this will result in a high ratio of phalanges to whole limbs (Expectation F).
134
Small terrestrial mammals were likely skinned and butchered in the main room. Bodies of small terrestrial mammals would be moved to the kitchen for preparation and skins with attached extremities would be moved to the benches for garment manufacture. If the heads were not previously removed, they would have been removed and discarded on the main room floor resulting in higher frequencies of heads (Expectation G).
Bird wings and legs packages were likely removed from the body in the main room, with the long bones of these packages being moved on to the benches for use in tool making and the bodies being moved to the kitchen for preparation. The remaining bones (lower limbs) are likely to be in highest frequencies on the main room floor after they were discarded (Expectation H).
Fish were unlikely to be butchered in the main room as they were stored outside and moved directly into the kitchen for preparation in soups or onto the benches to be consumed whole; therefore, there are not expected to be any fish remains on the main room floor (Expectations I).
Due to more intense butchery, there should be an intermediate frequency of butchery marks visible on bones; these will be more frequent than in the tunnel storage area, but lower than the kitchen or benches (Expectation J). There should be no primary burning on the main room floor. The main room floor should have a low frequency of total bones (NSP) in comparison to the rest of the house as most of the bones will have been moved into the kitchen or onto the benches (Expectation K).
5.2.3.3 Tertiary Activities
(Table 4, Stage 3 column)
5.2.3.3.1 Carnivore Activities
As described above (Section 5.2.2.3.1. Carnivore Activities), carnivores were likely absent from the main room floor during house occupation; therefore, there are no changes in the faunal remains from the secondary to tertiary stages on the main room floor due to carnivore activity.
135
5.2.3.3.2 Secondary Burning
Secondary burning is likely to be absent on the main room floor since there are no indications of either cooking or lighting in this area (see Section 5.2.2.3.2. Secondary Burning and Table 2: Activity frequencies).
5.2.3.3.3 Trampling
As stated above (Section 5.2.2.3.3. Trampling), trampling was likely highest on the main room floor due to frequency of use and types of activities. Trampling on the main room floor would have broken the large bones into smaller pieces, making some of them unidentifiable below class. In general, fragile craniums were likely affected the most; small, dense extremity bones and dense mandibles were likely affected the least. Additionally, trampling and kicking may have moved some bones from the main room floor to under then benches.
5.2.3.3.4 Cleaning Activities
The main room would have had material removed and deposited during cleaning events. Likely, large remains, such as heads, were removed and small remains, such as indeterminate bones, were added during the cleaning process. Main room floors were also likely swept periodically, removing most of the medium and small remains (Burch 2006:268). Cleaning activities likely moved bones to the edges of the main room floor, into the tunnel, and under the benches. Cleaning may have occurred less on the main room floor than in either the kitchen or on top of the benches as those areas were used more intensively for sitting and working; remains left on a floor may be less of a nuisance than those left in high-use areas.
Small remains removed from the benches (see Section 5.2.5.3.4. Cleaning) would have been transferred to the main room floor where they and small remains discarded from secondary and tertiary activities in the main room floor were likely caught in the floorboards and along the edges of the room. Large remains were likely removed completely.
5.2.3.3.5 Tertiary Expectations Summary
The main room floor likely has no carnivore activity or secondary burning, but a high frequency of trampling and cleaning activities (Table 2: Activity frequencies). This affects the secondary expectations in a few ways, most importantly by removing (via cleaning or trampling) the fragile
136 and/or large cranial remains and raising the frequency and ratio of small, dense extremities and mandibles.
There are no large marine mammal bones expected on the main room floor from secondary stage activities. Some large marine mammal extremities may have entered the assemblage due to cleaning of the benches. These are dense, so trampling likely did not affect them. However, in comparison to the tunnel and benches, the frequency of extremities is expected to be low (Expectation A). Bacula, ribs and mandibles are not expected to be present during tertiary activities (Expectation B).
Small seal cranium may be removed due to cleaning of large bones and destruction of fragile bones through trampling, leaving high frequencies of dense mandibles and hind flipper bones (Expectation C).
Large bones of large terrestrial mammals are not expected to be on the main room floor from the secondary stage activities since most of the low-utility bones that remain are small; however, some low-utility bones may be larger, and so would be removed. Thus, the only bones expected would be small, low-utility bones, leading to an expectation of a low ratio of high-utility bones to small, low-utility bones (Expectation E). Large terrestrial mammal phalanges are both small and dense and will survive both trampling and cleaning in high frequencies; tertiary activities will increase, but not change, the secondary ratios of Expectation F.
Small terrestrial mammals would have their fragile craniums destroyed, leaving the more dense mandibles on the main room floor (Expectation G).
Dense bird bone, which includes the lower limbs and extremities, will be in high frequencies and ratios after tertiary activities. Therefore, birds would still have high frequencies of lower limbs (Expectation H).
Fish, like large marine mammals, are not expected to be on the floor during secondary activities; however, they may have been added in low frequencies during cleaning activities. Trampling is expected to have destroyed most of the fragile fish remains, leaving only the dense vertebrae. These are expected in intermediate frequencies compared to other areas (Expectation I). Overall, it is expected that fish will be in relatively low frequencies in the main room (Expectation L).
137
Trampling likely raises the total frequency of bones (Expectation K) and the total frequency of indeterminate bones (Expectation M). There is not expected to be any carnivore or burning modifications. Butchery marks will be intermediate in frequency: although their visibility may be reduced due to the destruction of identifiable bones, they are still likely higher than in other areas (Expectation J).
5.2.3.4 Quaternary Activities
(Table 4, Stage 4 column)
5.2.3.4.1 Animal Disturbances
As described above (Section 5.2.2.4.1. Animal Disturbances), the main room floor would likely have had no rodent activity and a medium amount of canid activity prior to collapse. This would result in both intermediate frequencies of carnivore modifications on bones and destruction of bones.
5.2.3.4.2 Weathering
Weathering processes are not expected to have affected the main room floor due to quick collapse and deep burial.
5.2.3.4.3 Human Actions
As described above (Section 5.2.2.4.3. Human Actions), human actions were not likely a major activity in Feature 87, with negligible trampling effects on the main room floor after abandonment (Table 2: Activity frequencies).
5.2.3.4.4 Quaternary Expectations Summary
Carnivore activity is likely the only quaternary activity that will have affected the faunal assemblage in the main room. This activity will affect the assemblages by adding carnivore marks to the bones (Expectation N) and by destroying less dense remains and creating more indeterminate bone pieces (Expectation M). All of the bones will have been affected by intermediate frequencies of density-mediated destruction; however, carnivore-mediated destruction will enhance the tertiary expectations rather than change them. Therefore, the ratios and frequencies of large marine mammals, small seals, large terrestrial mammals, small
138 terrestrial mammals, birds, and fish are unlikely to be changed comparatively from the tertiary expectations (Expectations A, C, E, G, H, I, and L). All of these ratios and frequencies have been already reduced to dense bones from tertiary activities. Bones from Expectations B and D are not expected to be present on the main room floor.
5.2.4 Kitchen Expectations
5.2.4.1 Primary Activities
See Section 2.1. Primary Stage for All Areas (Table 5, Stage 1 column).
5.2.4.2 Secondary Activities
(Table 5, Stage 2 column)
5.2.4.2.1 Preparation
Meat was most often prepared for consumption in the kitchen. Preparation of meat packages would have included deboning, further butchery, and cooking. Once preparation was complete, bones still attached to the prepared meat would have been moved out of the kitchen onto the benches for consumption, leaving the removed and discarded elements from preparation in the kitchen.
Large marine mammal meat was likely prepared in the kitchen but, as described above, few bones likely even entered the site. The bones that did enter the site, such as the ribs, mandibles, and appendicular elements, were not recorded in the documentary records as being prepared in the kitchen; the ribs and mandibles were used for construction, while the extremities were left in accessible storage areas for consumption within a few days. Thus, no large marine mammal bones are expected in the kitchen (Expectations A; Expectation B overwritten by Expectation A).
Small seal high-utility packages were brought into the kitchen, where the attached low-utility parts (such as metacarpals and front phalanges) were removed from the high-utility elements and discarded (Expectation C). Heads and hind limbs were not consumed often and so were not assumed to have entered the kitchen area (Expectation D).
139
Stage 1: Stage 2: Stage 3: Stage 4: Taxonomic Village Faunal Household Activities Contemporary Post-Abandonment Final Expectations Category Composition Activities Activities Large A. High ratio of A. Large marine A. N/A A. N/A A. LMM not expected marine appendicular : axial mammals not to be present in the mammal elements prepared in the kitchen kitchen B. High frequencies of B. Overwritten by B. N/A B. N/A B. Overwritten by bacula, whale ribs, Expectation A Expectation A and mandibles Small seal C. Even skeletal C. High ratio of low- C. Slightly high ratio C. Slight increase in C. Slightly high ratio of distribution utility : high-utility of dense : less ratio; still slightly dense : less dense elements dense elements high elements D. Low frequencies of D. Heads and hind D. N/A D. N/A D. Heads and hind head and hind flippers not flippers not expected flippers prepared in kitchen to be present in the kitchen Large E. High ratio of high- E. Intermediate ratio of E. High ratio of dense E. Slight increase in E. High ratio of dense : terrestrial utility : low-utility high-utility : low- : less dense ratio; still high less dense elements mammal elements utility elements elements F. High frequency of F. Lower limbs will be F. N/A F. N/A F. Lower limbs will be lower limbs smashed and smashed and rendered rendered unidentifiable unidentifiable Small G. Even distribution of G. Slightly high G. Increase in G. Slight increase in G. High frequency of terrestrial skeletal elements frequency of distal frequency; still frequency; still high distal limbs mammal limbs slightly high Bird H. Even distribution of H. Whole bird bodies H. N/A H. N/A H. Whole bird bodies skeletal elements cooked in bags with cooked in bags with no discarded no discarded remains in kitchen remains in kitchen Fish I. Even distribution of I. Even distribution of I. Slightly high I. Slight increase in I. High frequency of cranial : post-cranial cranial : post-cranial frequency of frequency; still high vertebrae elements elements vertebrae L. High frequency of L. Intermediate L. Slight decrease in L. Intermediate fish frequency of fish frequency; still frequency of fish intermediate General J. Low frequency of J. Intermediate J. Low frequency of J. Slight decrease in J. Low frequency of butchery marks frequency of butchery marks frequency; still low butchery marks butchery marks K. High frequency of K. High frequency of K. Slight increase in K. High frequency of indeterminate indeterminate bone frequency; still high indeterminate bone mammal bone M. Highest frequency M. Increase in M. Slight increase in M. Highest frequency of of burn marks frequency; still frequency; still burn marks highest highest N. High ratio of burnt N. Even ratio of burnt N. No change N. Even ratio of burnt small mammal : all bones across taxa bones across taxa taxa Table 5: Summary of expectations from the kitchen. Note that all expectations are cumulative; the activities from each stage only affect the bones present in the previous stage.
140
Large terrestrial mammal packages, already stripped of their low-utility bones in the main room butchering area, were likely deboned in the kitchen for preparation of meat for soups or grilling (Burch 2006:212). The discarded bones would therefore be high-meat utility elements. In comparison to the preparation areas, the ratio of high- to low-utility bones would be higher, but in comparison to the consumption areas (see Section 5.2.5.2.1. Consumption), the ratio would be lower (Expectation E). Lower legs, the other frequently occurring large terrestrial mammal elements, were likely processed in the kitchen; processing involved smashing and boiling the lower leg bones to extract grease and fat to be used in preparation and to be consumed (Binford 1978a:157–159; Burch 2006:138). This preparation would leave few, if any, bits of identifiable large terrestrial mammal bone (Expectation F) but would increase the frequency of indeterminate mammal bone in the kitchen (Expectation K).
For small terrestrial mammals, the heads and skins with attached extremities were removed in the butchery area; the rest of the body would have been transported to the kitchen. Here, if further butchery occurred, it likely removed the distal limb bones (radii, ulnae, and tibias) from the body, which were discarded in the kitchen (Expectation G). The rest of the body would be moved out onto the benches for consumption, leaving no trace in the kitchen.
After the legs and wings were removed in the butchery areas, the bird bodies may have been further butchered in the kitchen to prepare for consumption; however, there is likely to be little to no discard of bird remains as most were cooked whole in bags (Burch 2006:179, 180, 181, 182, 183, and 184) (Expectation H).
Frozen fish were most often prepared through deboning and/or cooking in the kitchen. If fish were deboned, the bones would have been discarded in the kitchen, resulting in a higher percentage of fish remains in the kitchen than in either the storage or butchery areas, but likely lower than in the consumption areas (see Section 5.2.5.2.1. Consumption) (Expectation L). The most common preparation of fish in the winter was boiling the fish heads in soups (Burch 2006:146). Fish heads used in soup may have been prepared in the kitchen, with the bodies (and associated ribs) being moving out to the consumption areas. When heads were used in soup, they were not likely discarded from the soup in the kitchen, but were discarded on the benches after
141 consumption. Thus, the skeletal distribution of fish was likely fairly even in the kitchen due to deboning and preparation (Expectation I).
Cooking in the documentary record mainly consisted of boiling, although grilling is also mentioned (Burch 2006:212). Burn modifications should be highest here compared to the other areas (Expectation M). Small terrestrial mammals were more frequently grilled than other animals, and so should have a higher frequency of burning (Burch 2006:212) (Expectation N). Butchery marks are expected in greater frequencies here than in the tunnel or main room floor as animal remains have undergone another stage of preparation, but lower than on the benches where they are further reduced for consumption (Expectation J).
5.2.4.3 Tertiary Activities
(Table 5, Stage 3 column)
5.2.4.3.1 Carnivore Activities
As described above (Section 5.2.2.3.1. Carnivore Activities and Table 2: Activity frequencies), carnivores were likely absent in the kitchen during house occupation; therefore, there are no changes in the faunal remains from the secondary to tertiary stages in the kitchen due to carnivore activities.
5.2.4.3.2 Secondary Burning
The kitchen would likely have the highest frequency and ratio of burnt bone due to secondary burning. Kitchens were used to prepare and cook food, with most preparation done over an open wood or oil fire. All of the taxa present in the kitchen are likely to have some degree of secondary burning, especially in comparison with the other areas. Burning, like dog gnawing, breaks down less dense bones (Clark and Ligouis 2010; Conard et al. 2008; Costamagno et al. 2010; Morin 2010; Shipman et al. 1984; Théry-Parisot et al. 2005), both destroying them and making them analytically unidentifiable to one of the six taxa. Because of this, there should be a higher frequency of dense, burnt large mammal bones in the kitchen than in the other areas. There should also be a high frequency and ratio of small, indeterminate, burnt fragments in the kitchen area in comparison to the other areas.
142
5.2.4.3.3 Trampling
The kitchen, a well-used location in the house, is assumed to have intermediate amounts of trampling activity. As described above, trampling affects bones by breaking them into smaller pieces, reducing the number of large, identifiable remains and increasing the number of small, indeterminate remains. In the kitchen, this will result in many of the bones becoming indeterminate, especially the fragile, burnt remains.
5.2.4.3.4 Cleaning
Large bones in the kitchen would have been removed to the midden by hand, with small remains likely taken out periodically. This intermediate amount of removal via cleaning would likely not change many of the taxa ratios, but would result in burnt remains entering the midden (see Section 5.2.6.3.4. Cleaning).
5.2.4.3.5 Tertiary Expectations Summary
The medium amount of trampling and cleaning likely does not affect the taxa ratios to any great extent. However, burning may have destroyed some of the less dense bones by breaking them into indeterminate pieces.
Bones of large marine mammals are not assumed to have entered the kitchen for preparation (Expectation A).
Secondary stage expectations for small seals (high ratio of low-utility to high-utility) will be overwritten by an expectation based on density-mediated attrition from burn destruction. There is likely to be a slightly higher ratio of dense to less dense remains (Expectation C).
The large terrestrial mammal intermediate ratio of high-utility to low-utility elements may change as less dense bones are removed through burning and trampling. This will likely result in a signature related to density with a high ratio of dense to less dense elements (Expectation E).
The high frequency of small terrestrial mammal distal limb bones is likely to increase since the distal limbs are relatively dense in comparison to the rest of the skeleton (Lyman, Houghton, et al. 1992; Novecosky and Popkin 2005; Pavao and Stahl 1999) (Expectation G).
Bird bones are not discarded in the kitchen during preparation (Expectation H).
143
Fish will likely have a fairly even distribution of skeletal remains, although there may be a slight increase in the frequency of vertebrae due to the higher density of vertebrae to other skeletal elements (Expectation I). Tertiary activities may reduce the frequency of fish remains in the kitchen in general (Expectation L)
In general, both trampling and burning likely reduce the overall number of identifiable bones and increase the number of small indeterminate remains (Expectation K). Burning should increase absolutely from secondary to tertiary stages, but it will not change in comparison to other areas (Expectation M). Also, secondary burning does not distinguish between taxa, and so there should be an even ratio of burning among taxa (Expectation N). The reduction of the identifiable remains and the reduction in size of remains coupled with the increased presence of burn modifications would likely affect the visibility of butchery marks. This would reduce the overall comparative butchery frequency to make it low in comparison to other areas of the house (Expectation J).
5.2.4.4 Quaternary Activities
(Table 5, Stage 4 column)
5.2.4.4.1 Animal Disturbances
As stated above, there is likely no post-abandonment rodent or carnivore activity associated with the kitchen (Table 2: Activity frequencies).
5.2.4.4.2 Weathering
Weathering may have affected the kitchen assemblage to a small degree since it had less overburden than the tunnel or main room floor. However, the weathering is likely to be minimal, with some cracking due to freeze-thaw effects. This would likely slightly reduce the size of bone fragments and make them harder to identify.
5.2.4.4.3 Human Actions
Human actions are not likely to have affected the kitchen since people were not likely moving through the kitchen to scavenge other household or artefactual materials.
144
5.2.4.4.4 Quaternary Expectations Summary
The only quaternary activity to affect the kitchen would be weathering, and this likely had a low impact. This would enhance the taxonomic anatomical expectations due to low density-mediated destruction, although this would not alter the ratios (Expectations C, E, G, and I). Modification frequencies are not likely to change in type, but the low density-mediated destruction would enhance these frequencies (Expectations J, K, L, M, and N). Bones from expectations A, B, D, F, and H are not expected to be present in kitchen.
5.2.5 Benches Expectations
5.2.5.1 Primary Activities
See Section 5.2.1. Primary Stage for All Areas (Table 6, Stage 1 column).
5.2.5.2 Secondary Activities
(Table 6, Stage 2 column)
Benches were the location for many activities including sleeping, eating, tool manufacturing, and socializing (Burch 2006:216, 264). Additionally, storage may have been important under the benches rather than on top of them. Consumption and manufacturing are described here because they are the activities most often associated with faunal remains and activities on the benches. Storage of animal remains under the benches is not expected to have been of primary importance due to the warmth of the house and limited accessibility of the under-bench areas. Summary expectations relating to the consumption and manufacturing activities are defined after the individual activity descriptions (see Table 6).
5.2.5.2.1 Consumption
The primary activity related to faunal remains that occurred on the benches was consumption. Here, people ate the food directly from storage, after it had been butchered, or after it had undergone preparation in the kitchen. Using the naïve assumption that bones were discarded at their place of activity, the bones that were attached to meat during consumption are expected to be discarded on the benches for the secondary set of expectations.
145
Stage 1: Stage 2: Stage 3: Stage 4: Taxonomic Village Faunal Household Activities Contemporary Post- Final Expectations Category Composition Activities Abandonment Activities Large A. High ratio of A. Discard of bones on A. Removal of bones A. No activities; A. High ratio of marine appendicular : axial benches after equally; no change in ratio no change appendicular : axial mammal elements consumption; no change elements B. High frequencies of B. High frequency of bacula B. Curation of raw material B. No activities; B. High frequency of bacula, whale ribs and unidentifiable sources for too no change bacula and and mandibles manufacture; no change unidentifiable Small seal C. Even skeletal C. High frequency of high C. Intermediate frequency of C. No activities; C. Intermediate distribution utility elements high-utility bones no change frequency of high- utility elements D. Low frequencies of D. High frequency of front D. Increase in frequency; still D. No activities; D. High frequency of head and hind flippers high no change front flippers flippers Q. High ratio of small : large Q. No activities; Q. High ratio of small : elements no change large elements Large E. High ratio of high- E. High ratio of high-utility E. Removal of these large E. No activities; E. Low frequency of terrestrial utility : low-utility : low-utility elements bones during cleaning; no change high-utility elements mammal elements low frequency of high- utility elements F. High frequency of F. High frequency of lower F. Increase in frequency due F. No activities; F. High frequency of lower limbs limbs to curation; no change no change lower limbs L. High frequency of antler L. Increase in frequency; L. No activities; L. High frequency of no change no change antler Small G. Even distribution of G. Low frequency of heads G. High ratio of small : large G. No activities; G. High ratio of small : terrestrial skeletal elements elements no change large elements mammal Bird H. Even distribution of H. Low frequency of heads H. High frequency of wings H. No activities; H. High frequency of skeletal elements and legs no change wings and legs P. High ratio of small : large P. No activities; P. High ratio of small : bird elements no change large elements Fish I. Even distribution of I. Discard of bones from I. All bones left; no change I. No activities; I. Even distribution of cranial : post-cranial soups and frozen fish no change cranial : post-cranial elements after consumption; no elements change K. Highest frequency of K. Increase in frequency; K. No activities; K. Highest frequency of fish still highest no change fish General J. Low frequency of J. High frequency of J. Intermediate frequency of J. No activities; J. Intermediate butchery marks butchery marks butchery marks no change frequency of butchery marks M. Intermediate frequency M. Decrease in frequency; M. No activities; M. Intermediate of burn marks still medium no change frequency of burn marks N. Low total number of N. No activities; N. Low total number of specimens (NSP) no change specimens (NSP) O. High ratio of small : O. No activities; O. High ratio of small : large bones no change large bones Table 6: Summary of expectations from the benches. Note that all expectations are cumulative; the activities from each stage only affect the bones present in the previous stage.
146
For large marine mammals, the only bones associated with consumption are the appendicular bones. As described before, they were likely placed in storage for a few days in order to ferment (Burch 2006:154–155, 168), then consumed and discarded on the bench. Therefore, they will have a high ratio of appendicular to axial bones (Expectation A).
Small seal remains will have undergone primary butchery on the main floor, preparation in the kitchen, and finally consumption on the benches. Bones associated with consumption are the high-utility bones. This will leave a high frequency of high-utility bones on the benches (Expectation C). Heads and flippers are not expected to be on the benches, as they were not typically consumed.
Large terrestrial mammal high-utility bones are expected in high ratio compared to low-utility bones on the benches due to their association with meat consumption (Expectation E). Lower limb bones are not expected to be present on the benches because they were processed and discarded in the kitchen.
Small terrestrial mammal heads and lower limbs are expected in lower ratios than the other skeletal elements (e.g. ribs, vertebrae, upper limbs); heads and lower limbs are not attached to much meat, while the bodies and upper limbs are attached to more meat. The heads and lower limbs are assumed to have been removed during the butchery and preparation activities associated with the main room floor and kitchen. Expectations relating to small terrestrial mammals will also be affected by manufacturing activities; these expectations are described in the summary.
Bird wings, lower legs, and heads are expected in low frequencies compared to other skeletal elements due to their lack of meat and removal at other stages in processing. Again, the expected low-ratio elements would have been removed during the butchery and preparation activities associated with the main room floor and kitchen. Expectations relating to birds will also be affected by manufacturing activities and so the expectations are described in the summary.
Fish are found in the highest frequency on the benches in comparison with other areas (Expectation K). As discussed above, fish may have been prepared as part of a soup or eaten raw. In both cases, all the fish remains were transported and discarded on the bench; fish heads from
147 soups and post-cranial material from frozen bodies. Fish remains are therefore likely to be fairly equally represented by both cranial and post-cranial remains (Expectation I).
During consumption activities, knives may have been used to cut meat off bones and into smaller pieces; these butchery marks would be additional to any previous butchery marks made during primary or secondary butchery, or during preparation. Burning is expected in intermediate frequencies as some of the bones associated with consumed meat were likely burned during preparation (i.e. grilling over open flame), especially for small terrestrial mammals (Burch 2006:212). Manufacturing activities alter both of these observations and so expectations are given in the summary section.
5.2.5.2.2 Use in Manufacture
As mentioned above, benches were the location within domestic dwellings where tool and garment manufacture occurred (Burch 2006:216, 264). In the dark of winter, these domestic dwellings would have been one of the only places with sufficient light and warmth for working skins, bones, and stones (Burch 2006:216). Bones used as raw material to manufacture tools or attached to skins for garment manufacture will be in highest frequency with a high ratio in comparison to other bones due to curation of bones for future manufacture and discard of bones after they have been used for tool manufacture.
Large marine mammal bacula will be in high frequencies on the benches compared to any other area as they are often used as raw material for tools. Other large marine mammal bones, often unidentifiable, may also be in highest frequencies on the benches as raw material (Expectation B).
Small seal bones are rarely used in tool manufacture, so they are expected in very low frequencies on the benches for tool manufacture activities. Small seal skins were sometimes used to manufacture garments or other items such as tent covers or bags. The front flippers were often left attached to skins (Burch 2006:155); these front flipper bones are expected to occur in higher frequencies than all other skeletal elements (Expectation D).
Large terrestrial mammal lower limb bones are associated with tool manufacture if they were not smashed and boiled for grease extraction in the kitchen. These bones would be used and stored on the benches for tool manufacture (Expectation F). Antler, an extremely important raw
148 material source for tool manufacture, will also likely be found in high comparative frequencies on the benches (Expectation L).
There is expected to be a high frequency and ratio of small terrestrial mammal extremity bones because they were attached to skins used for garment manufacture. Again, expectations for small terrestrial mammals are based on manufacturing and consumption activities and so are given in the summary section.
Bird wing and leg bones are also used as raw material for tool manufacture and so are likely found in the highest frequencies on the benches in areas of tool manufacture. Again, expectations for birds are based on manufacturing and consumption activities and so are given in the summary section.
Fish are not typically used in tool or garment manufacture and are not expected on the benches due to manufacturing activities.
Butchery marks will be high in the areas for tool and garment manufacture because bones have undergone primary and secondary butchery, and potentially will have been processed heavily during tool and garment manufacture. Consumption and manufacturing contribute to modification frequencies and so these expectations will be outlined below.
5.2.5.2.3 Benches Secondary Expectations Summary
The benches in this analysis are a combination of consumption and manufacturing activities. Because of this, many of the expectations are broader than expected for only one activity (Table 6).
For large marine mammals, Expectation A is based on the consumption of flippers, while Expectation B is based on the use of bones in manufacture.
For small seals, Expectation C is based on the consumption discussion, while Expectation D is based on garment manufacture.
For large terrestrial mammals, Expectation E is based on consumption activities, while Expectations F and L are based on manufacturing activities.
149
For small terrestrial mammals, consumption activities would lead to a low expectation of heads and extremities; however, manufacturing expectations assume a high ratio of extremities to other bones. Therefore, the small terrestrial mammal skeleton is likely to show a fairly even distribution (bodies consumed, feet attached to skins used in manufacture); the only missing element will be heads (Expectation G).
Birds are similar to small terrestrial mammals with consumption activities assuming a low frequency of heads, wings, and legs; however, wing and leg bones are used in tool manufacture. This creates an expectation of even skeletal distribution (bodies consumed, wings and legs used in manufacture), with low frequencies of heads (Expectation H).
Fish are a bit simpler because they are not commonly used in tool or garment manufacturing; therefore, the only activity affecting these expectations is consumption activities (Expectations I and K).
For both consumption and use in manufacture, the general expectations for all bones should show the highest frequencies of butchery marks of all the activity areas (Table 2: Activity frequencies) (Expectation J).
Additionally, the benches should have an intermediate frequency of burn modifications due to bones being prepared for consumption (burning during preparation) (Expectation M).
5.2.5.3 Tertiary Activities
(Table 6, Stage 3 column)
5.2.5.3.1 Carnivore Activities
As described above (Section 5.2.2.3.1.), carnivores were likely absent on the benches during house occupation; therefore, there are no changes in the faunal remains from the secondary to tertiary stages on the main room floor due to carnivore activity
5.2.5.3.2 Secondary Burning
Secondary burning likely occurred on the benches, but in low amounts (Table 2: Activity frequencies). Secondary burning was likely not due to cooking, but caused by light sources. Evidence of light sources is found on both the back bench (presence of a large ceramic lamp) and
150 side benches (burnt areas). As described above (Section 5.2.2.3.2.), burning alters the bone by changing the colour and breaking it into smaller pieces. An intermediate frequency of burning is already present in the secondary expectations, and so this secondary burning does not change the expectations. There will also be a low amount of fragmentation due to low amounts of burning.
5.2.5.3.3 Trampling
As stated above (Section 5.2.2.3.3.), trampling was low on the benches; sitting and sleeping may have broken some of the large bones into smaller pieces, making some of them unidentifiable below class.
5.2.5.3.4 Cleaning
The removal of both large and small bones through cleaning activities likely affected the remains most drastically in the tertiary stages as this area was used more intensively for sitting and working and was likely kept clean.
Like other areas, large bones were removed by hand to the midden or main floor area. Small remains were removed from the benches and transferred to the main room floor when bedding was picked up and shaken. Because both large and small remains are removed, the taxonomic ratios and frequencies are not likely to alter drastically from the secondary stage expectations. Overall, there are fewer remains on the benches than on the floor or kitchen. If remains were left on the benches, they would likely be small, unobtrusive remains, such as fish.
5.2.5.3.5 Tertiary Expectations Summary
Cleaning activities are assumed to have affected the benches the most as remains were removed to clear the area for sitting and sleeping. Some cleaning activities on the main room floor may have moved bones under then benches, which would increase the number of small remains. Carnivore activity is expected to be absent, while both secondary burning and trampling are likely low. These latter two activities will have reduced some of the bone to small, indeterminate pieces, but to a lesser extent than in the other areas. However, cleaning will have altered the expected frequencies by removing most large and small bones, leading to low overall specimen count (Expectation N). There is expected to be a higher ratio of small to large remains due to the
151 fact that large remains are preferentially removed and small remains are likely to get stuck down the sides of benches or walls or moved under the benches from the main floor (Expectation R).
Large marine mammal extremity bones are similar in size and would be treated similarly in cleaning activities; therefore, the ratio is not expected to change (Expectation A). The frequencies of bones used for tool manufacture is also expected to be the same as these materials were likely saved during cleaning activities as raw material sources for manufacturing (Expectation B).
For small seals, some of the larger high-utility bones will be removed, resulting in intermediate frequencies of high-utility bones (Expectation C). Small flipper bones are likely to fall through the bench platform planks and be missed in cleaning activities; since hind flipper bones are not expected to be on the bench from secondary activities, the front flippers are expected in high frequencies (Expectation D). There will also be a high ratio of small to large elements, as small elements may be left behind or disregarded during cleaning (Expectation Q).
There are only expected to be large bones of large terrestrial mammals from secondary activities on the bench, and most of these will be removed during cleaning (Expectation E). Bones used for manufacturing tools (lower limbs and antler) will be curated and will be in high frequencies (Expectations F and L).
Large remains of small terrestrial mammals will be removed, leaving just small elements; this will result in a high ratio of small to large elements relative to other areas (Expectation G).
Birds will also have most of their large remains removed, but likely will have a high ratio of small to large remains relative to other areas (Expectation P). Also, wings and legs may be curated for tool manufacture (Expectation H).
Fish elements are small already and so cleaning activities are less likely to differentially remove some of their elements; therefore, their anatomical ratio will not change (Expectation I). As elements from other taxa are removed during cleaning, fish remains will have even higher comparative taxonomic frequencies (Expectation K).
152
Although the evidence of butchery marks will be lower as bones are removed, it is still expected in intermediate frequencies compared to other areas (Expectation J). Burn frequencies may be reduced, but will be intermediate in comparison to other areas (Expectation M).
5.2.5.4 Quaternary Activities
(Table 6, Stage 4 column)
5.2.5.4.1 Animal Disturbances
There are expected to be little, if any, subsequent animal disturbances on the benches due to quick collapse and deep burial (see Section 5.2.2.4.1. Animal Disturbances).
5.2.5.4.2 Weathering
Again, there is expected to no weathering activity on the faunal material from the benches (see Section 5.2.2.4.2. Weathering).
5.2.5.4.3 Human Actions
With limited evidence for human activities at the house subsequent to abandonment and prior to burial, the bench is assumed to have no changes to the expectations (see Section 5.2.2.4.3. Human Actions).
5.2.5.4.4 Quaternary Expectations Summary
With little to no animal disturbances, weathering, or human actions, there is no expected change in the expectations from the tertiary to quaternary stages for the benches (Expectations A, B, C, D, E, F, G, H, I J, K, L, M, N, O, P, and Q).
5.2.6 Midden Expectations
5.2.6.1 Primary Activities
See Section 5.2.1. Primary Stage for All Areas (Table 7, Stage 1 column)
5.2.6.2 Secondary Activities
(Table 7, Stage 2 column)
153
5.2.6.2.1 Discard
Primary butchery may have taken place outside of the house, although this model expects most of the butchery to be conducted on the main room floor due to the need to thaw carcasses prior to butchery (see Section 5.2.3.2). It is expected that the discard of some remains from primary butchery practices likely ended up in the midden. However, the majority of bones to enter the midden are expected to result from tertiary cleaning activities (see Section 5.2.2.3.4. Cleaning and Table 7, Stage 3 column).
Large marine mammals were butchered mostly at the locations they were harvested and are not expected in high frequencies in the midden. Specifically, appendicular elements and bones used in manufacture or construction are not expected in the midden at this stage (Expectations A and B).
Heads and hind flippers of small seals may have been removed prior to entering the house and may be in intermediate frequencies in the midden compared to other areas (Expectation D). The rest of the carcass is expected to have been moved into the house from this stage and not be represented in the midden (Expectation C).
Low-utility elements of large terrestrial mammals may have been removed during butchery and discarded in the midden, like on the main room floor. This would leave a high ratio of low-utility to high-utility bones in the midden (Expectation E), with no evidence of lower limbs since these would have been moved into the kitchen for processing (Expectation F).
The small terrestrial mammal and bird heads may have been removed outside and discarded in the midden, leaving a higher frequency of heads for both these taxa (Expectations G and H).
Fish are not expected to be butchered near the house in the winter or summer (see Section 5.2.1.6) and therefore will be absent in the midden (Expectations I).
In general, the frequency of bones discarded in the midden for secondary activities should be low (Expectation K). Additionally, there should be higher ratios of large bones (e.g. heads) to small
154 Stage 1: Stage 2: Stage 3: Stage 4: Taxonomic Village Faunal Household Activities Contemporary Post-Abandonment Final Expectations Category Composition Activities Activities Large A. High ratio of A. Large marine A. Intermediate A. Decrease in A. Intermediate marine appendicular : axial mammals not frequency of frequency; still frequency of mammal elements butchered near appendicular intermediate appendicular midden elements elements B. High frequencies of B. Large marine B. Curated on benches; B. N/A B. Bacula, whale ribs bacula, whale ribs mammals not not expected and mandibles not and mandibles butchered near expected to be midden present in the midden Small seal C. Even skeletal C. Enter house directly; C. High ratio of high- C. Increase in ratio; C. High ratio of high- distribution not in midden density : low- no relative density : low-density density elements change elements D. Low frequencies of D. Intermediate D. Intermediate D. Increase in D. Intermediate head and hind frequency of heads frequency of hind frequency; no frequency of hind flippers and hind flippers flippers relative change flippers Large E. High ratio of high- E. High ratio of low- E. High ratio of high- E. Increase in ratio; E. High ratio of high- terrestrial utility : low-utility utility : high-utility density : low- no relative density : low-density mammal elements elements density elements change elements F. High frequency of F. Enter house directly; F. N/A F. N/A F. Lower limbs not lower limbs no lower limbs expected to be present in the midden Small G. Even distribution of G. Slightly high G. Slightly high G. Increase in G. Slightly high ratio of terrestrial skeletal elements frequency of heads relative ratio of relative ratio; no cortical : cancellous mammal cortical : cancellous relative change bone bone Bird H. Even distribution of H. Slightly high H. Slightly high H. Increase in H. Slightly high ratio of skeletal elements frequency of heads relative ratio of relative ratio; no cortical : cancellous cortical : cancellous relative change bone bone Fish I. Even distribution of I. Fish not butchered I. Low frequency of I. N/A I. Fish not expected to cranial : post-cranial near midden addition and high be present in the elements frequency of midden destruction; not expected General J. Low frequency of J. Intermediate J. Addition and J. Low frequency J. Low frequency of butchery marks frequency of destruction of bone; of butchery butchery marks butchery marks no change marks K. Low total number of K. High total number K. Increase in K. High total number specimens (NSP) of specimens (NSP) frequency; no of specimens (NSP) relative change L. High ratio of large : L. Even ratio of large : L. Low ratio of L. Low ratio of large : small bones small bones large : small small bones bones M. Low frequency of M. No change M. Low frequency of burnt bone burnt bone N. High frequency of N. Increase in N. High frequency of carnivore marks frequency; no carnivore marks relative change O. High frequency O. High frequency of of weathering weathering Table 7: Summary of expectations from the midden. Note that all expectations are cumulative; the activities from each stage only affect the bones present in the previous stage.
155
bones (e.g. phalanges) due to the removal of large, low-utility bones for most taxa prior to these animals being moved into the house (Expectation L). Finally, butchery marks should be in intermediate frequencies in comparison to the rest of the areas as animal remains are only undergoing primary butchery (Expectation J).
5.2.6.3 Tertiary Activities
(Table 7, Stage 3 column)
Most tertiary and quaternary activities are expected in highest frequencies in the midden (Table 2: Activity frequencies). However, bone may survive better in the midden at all stages if the midden accumulates quickly; in other words, middens can often be excellent places of preservation because they get covered quickly with other material, which inhibits many types of tertiary and quaternary activities. Despite this often quick preservation, the comparative nature of this analysis still means that most of the tertiary and quaternary activities will have the highest frequency in the midden even if absolute frequencies are low.
5.2.6.3.1 Carnivore Activities
Carnivore activities are expected to be highest in the midden due to the fact that dogs and other canids have access to the middens more than any interior house space (see Section 5.2.2.3.1. Carnivore Activities). Carnivore activities are expected to leave high frequencies of carnivore marks on the bones and destroy cancellous and low-density bone to a greater degree than high- density and cortical bone.
5.2.6.3.2 Secondary Burning
Secondary burning is absent in the midden, as people would not set up fires or lamps where they discard material. However, there may be evidence of burnt bones due to cleaning activities (see Section 5.2.6.3.4. Secondary Burning).
5.2.6.3.3 Trampling
The midden likely had some trampling, but was likely avoided by humans during their everyday activities (see Section 5.2.2.3.3. Trampling). Trampling may have reduced some bones to small pieces.
156
5.2.6.3.4 Cleaning
The addition of material to due to cleaning activities is assumed to have been highest in the midden (see Section 5.2.2.3.4. Cleaning). Large bones, removed by hand from all of the interior house areas, are expected to be in high frequencies in the midden.
5.2.6.3.5 Tertiary Expectations Summary
The midden is likely to be most affected by the addition of remains from cleaning activities. These activities would add all types and sizes of bones; however, large remains are removed more frequently as they are more obtrusive in the house than small remains. Although cleaning activities will add larger rather than smaller bones to the midden, high levels of carnivore destruction will lead to an even ratio of large to small remains relative to other areas (Expectation L) and changes in the taxonomic anatomical representation described below. Carnivore activity will also leave a high frequency of carnivore marks (Expectation N). With the addition of bones, the overall bone frequency is expected to be high in comparison to other areas (Expectation K). The frequency of butchery marks is assumed to stay the same because carnivore destruction will destroy some marks, while the addition of bones with butchery marks from other areas will increase the frequency (Expectation J).
Small extremity bones of large marine mammals are expected in the midden, as they were likely cleaned from the consumption areas (Expectation A). However, the other bones used as raw material for tool manufacture and construction are not likely to be found in the midden in high quantities due to curation of these bones in the manufacturing area (Expectation B).
Cleaning will add mostly large skeletal elements to the midden from butchery and consumption areas. However, destruction of low-density bones due to carnivore activities will leave a high ratio of high-density to low-density bones (Expectation C). Small seal heads are likely destroyed due to high frequencies of carnivore gnawing and low frequencies of trampling, leaving an intermediate frequency of hind flippers (Expectation D).
The addition of large terrestrial mammal elements from consumption areas will add high-utility elements to the midden, thus potentially negating the high ratio of low-utility to high-utility elements and changing the ratio to an even representation of high- and low-utility elements. However, density-mediated attrition due to carnivore gnawing is expected to result in a high ratio
157 of high-density to low-density bones for large terrestrial mammals rather than a signature based on utility (Expectation E).
Like small seals, the heads of small terrestrial mammals are likely destroyed in higher frequencies than more dense elements, thus negating the high ratio of heads. However, carnivore gnawing likely destroyed many of the cancellous bones, leaving a high ratio of cortical to cancellous small terrestrial mammal bones relative to other areas (Expectation G).
Bird heads and any added cancellous bones were likely destroyed by carnivore activity, thus leaving a high ratio of cortical to cancellous bird bone relative to other areas (Expectation H).
Fish, not present originally in the midden, may have been added in low frequencies. They would then have been destroyed by carnivore action in this secondary stage. Thus, there is expected to be no or very low frequencies of fish bones in the midden (Expectations I).
5.2.6.4 Quaternary Activities
(Table 7, Stage 4 column)
As described above, the midden was likely affected most frequently and intensely by quaternary activities. Although a quick burial may reduce the impact of all of these quaternary activities, they will still be in highest frequency and intensity in the midden compared to the interior of the house.
5.2.6.4.1 Animal Disturbances
Both carnivore and rodent activity may have affected the midden. Animal marks on bones will be in highest frequency, and destruction of bones will lead to a higher ratio of small, indeterminate bones.
5.2.6.4.2 Weathering
Weathering, although mediated by the amount of time the midden was exposed, is likely to be highest for midden remains (Table 2: Activity frequencies). Weathering results in flaking, bleaching, and destruction of bone described in Section 5.2.2.4.2. Weathering.
158
5.2.6.4.3 Human Actions
Human actions are likely to have a low impact on midden formation for quaternary activities. Although some trampling after house abandonment when people re-occupied the site is likely, it is expected to be low. This again will result in destruction of bone through mechanical processes (see Section 5.2.2.4.3. Human Actions).
5.2.6.4.4 Quaternary Expectations Summary
Although there are comparatively high levels of quaternary activities that may destroy elements based on bone density, the anatomical ratios and frequencies for taxa will not change from the tertiary to quaternary expectations because the midden materials have already undergone high levels of density-mediated attrition by tertiary activities (Expectations A, C, D, E, G, and H). Large marine mammals, small seals, large terrestrial mammals, small terrestrial mammals, and bird ratios and frequencies will stay the same, as they are relational measures.
In general, animal destruction, weathering, and human trampling will all reduce the number of large, identifiable bones and increase the number of small, indeterminate bones. This will increase the absolute number of bones; the comparative frequency will still be high (Expectation K). The ratio of large to small bones will potentially change from equal to low, with more small fragments represented due to density-mediated attrition of carnivore gnawing, weathering, and human actions (Expectation L). Density-mediated attrition will destroy some of the butchery marks, leading to a low frequency of butchery marks in the quaternary stage (Expectation J). Low frequencies of burnt bones are still likely (Expectation M), while carnivore marks will still be high, especially with the addition of quaternary carnivore action (Expectation N). Finally, weathering will have the highest frequency in comparison with other areas (Expectation O), although the overall weathering frequency for the midden will depend on the rate of burial. Expectations B, F, and I are not expected to be present in the midden.
5.3 Summary
This chapter outlines the expectations to be tested with the archaeological record. The context of these expectations has been described in great detail in the preceding chapters; the following chapters will contextualize the zooarchaeological assemblage, starting first with a discussion of
159 methods (Chapter 6), a description of the general subsistence (Chapter 7), and finally an analysis of the spatial patterning (Chapter 8).
160
Chapter 6 Methodology
In order to understand the history of the archaeofaunal assemblage, the collection, identification, and analyses must be described and the methods outlined. The following methods are described for both the general subsistence analysis and the expectation testing. Detailed calculations are found in the appendices. Chapter 7 describes the general subsistence analysis, with figures and tables, and Chapter 8 interprets the spatial patterning by testing the expectations and discussing interpretations.
6.1 Collection
As described in Chapter 4, Feature 87 has undergone a number of known archaeological investigations. Jean Schaaf’s (1988) excavation detailed no collection of faunal remains. The faunal material collected from Harritt’s (1994) excavations was amalgamated with other site fauna from KTZ-088 (a different site designation than the current designation of KTZ-087) to give a broad understanding of subsistence practices at the site; this material is unavailable for the current study. The excavations in 2009 of the Cape Espenberg Project did recover faunal material from the feature in arbitrary layers; however, the floor was never identified in these excavations and so could not be correlated to the excavation in 2011 Cape Espenberg Project excavations of Feature 87.
The excavations of Feature 87 in 2011 under the direction of Dr. T. Max Friesen completely excavated the main room, partially excavated the tunnel, and sampled the kitchen and midden areas. The excavation first removed the sod layer (Level 1), and then excavated 1x1 metre squares of the ‘fill’ in arbitrary ten centimetre levels (Levels 2, 2b, 2c, etc.). Test excavations uncovered a dark, organic layer just above the floor and back platform. Once identified, this layer was labelled Level 3. Level 3 was excavated in cultural layers; for example, above floor material was labelled Level 3, directly on floor material was Level 3b, and below floor material was Level 3c. For the platform, these layers were slightly different, with above wall fall, under wall fall, below bench, and on floor all different levels. Only material under the side benches was labelled Level 3, with a similar system as the floor layers; above bench material was not collected as Level 3. All Level 3 was excavated in 50x50 centimetre squares with all faunal
161 material bagged by quadrants within the 1x1 metre square. This allowed for an efficient, yet high-resolution excavation of the floor material. The entire main room was excavated down to sterile sand. The first two metres of the tunnel were also excavated to the floor level, although the permafrost prevented excavation to sterile soil.
Most material excavated from Level 3 was screened using 1/8-inch (3mm) mesh. This screen size is recommended for capturing a wide variety of small bones, including some small fish remains (Partlow 2006; Vale and Gargett 2002; Zohar and Belmaker 2005). The larger 1/4-inch (6mm) mesh was used in the following cases. On the last day of excavation, time constraints necessitated the use of 1/4-inch mesh; in these cases, mesh size switched midway during screening, and both materials screened from the smaller (1/8-inch) and larger (1/4-inch) were put into the same level/quadrant bag. In other cases, 1/4-inch mesh was used by mistake or when individual quadrants were not excavated. The squares where larger screen size was used are shown in Table 8. This may slightly bias the sample in favour of larger bones (from larger animals or more complete bones) and this was taken into account during the analysis; however, because the analysis combines the different levels in each quadrant, the impact of the different screen sizes is largely mitigated. South East Level Quad Screen Size 3 1 3b NE 1/8-inch (3mm) & 1/4-inch (6mm) NE NW 3 3 x3d 1/8-inch (3mm) & 1/4-inch (6mm) SE SW No quad 5 2 3b NE 1/8-inch (3mm) & 1/4-inch (6mm) NW NE 5 2 3d 1/8-inch (3mm) & 1/4-inch (6mm) SW 5 2 3e SE 1/4-inch (6mm) SE 5 3 x3d 1/4-inch (6mm) NE SE 6 2 3 1/4-inch (6mm) SW 7 2 3 SE 1/4-inch (6mm) 8 2 3 NE 1/4-inch (6mm) Table 8: Squares screened through larger mesh (1/4-inch or 6mm)
Both the kitchen area and the midden area were excavated differently. For the kitchen area, 100 percent samples were collected from specific areas since the matrix itself was burnt bone, charcoal, hardened seal oil, and bits of ceramics. These bulk samples were removed for the
162 southwest corners of the squares at the occupation level and then sorted in the lab. The midden area was excavated in 1x1 metre squares with the southwest corner excavated, screened, and collected separately using the 1/8-inch mesh. The other three quadrants were collected together and put through the 1/4-inch mesh. There were no distinct cultural layers within the midden, so excavation proceeded by arbitrary 10cm levels labelled as Level 2 until sterile soil or permafrost was reached.
Processing at the field laboratory included a rough count of bones and check for artefacts; no bones were removed or identified during the process. Bones that were artefacts or were considered tool debitage (i.e. worked antler) were categorized and identified with the artefact assemblage. All other faunal material was shipped to the University of Toronto for identification.
6.2 Identification
The majority of faunal remains were identified at the Zooarchaeology Lab in the Department of Anthropology at the University of Toronto. All material was identified to the lowest taxonomic level possible, with uncertain identifications listed at a higher taxon with a recorded indication of the probable lower taxon identification. The taxonomic classification of the specimen (Lyman 2008:4–7), skeletal element, side, size (compared to lab comparative specimens), proportion, orientation, and size class (i.e. <1cm, 1-2cm, 2-5cm, or >5cm) were recorded. Secondary information was recorded for type of mammal (terrestrial or marine), bone type, fusion, presence of juvenile cortex, age, and sex. Modifications and their locations on the bones were also recorded. Cut marks were identified using a hand-held magnifying glass (10x); the number of cuts per specimen was recorded. Burning colour and percentage was recorded. Weathering was recorded by type (i.e. flaking, bleaching, etc.) and using Behrensmeyer’s (1978) weathering stages. The type of gnaw marks (puncture, gnawing, acid etching) was recorded, along with size and orientation. Impact scars, identified by the presence of one or more bone flake scars, were recorded. Any root etching or soil staining was also recorded, along with the percentage of the specimen it affected. Large terrestrial mammal cortical bones were analyzed for fracture freshness using Outram’s (2001; 2002) fracture freshness index (FFI). Here, three criteria (fracture angle, fracture surface texture, and fracture outline) were given a score from zero to two and then added together to give a score from zero to six.
163
Identification using different comparative collections allowed a fairly large portion of the material to be identified to genus or species. Since the Bering Sea fish at the zooarchaeological lab at the University of Toronto and the Ichthyology Department at the Royal Ontario Museum are limited, fish specimens were not identified as narrowly as might be possible with better collections. However, the limited number of fish in the assemblage dissuaded trips to other collections to further identify these fragments. Some bird and mammal remains were taken to other collections for identification. Small charadriiformes (gulls, terns, and jaegers) were identified at the Royal Ontario Museum, Ornithology Department. Unlike fish, small seals, an integral part of most Iñupiat and Inuit diets, were important to identify as narrowly as possible. Elements that were used by Hodgetts (1999) in distinguishing between seal species of the North Atlantic were taken to the Museum of the North, University of Alaska Fairbanks. There, multiple skeletons of ringed seal, spotted seal, harbour seal, and ribbon seal were examined in order to identify the remains on hand.
6.3 Analysis
Different sets of specimens were used in general subsistence and spatial patterning analyses. For general subsistence, most material labelled Level 3 was used. For the spatial patterning analysis, material was more narrowly selected in order to standardize the spatial location and collection methods of the material. Appendix A outlines the squares used in general subsistence analysis; Appendix B outlines the squares used in the spatial patterning analysis. In both analyses, material from the completely excavated main room, from the partially excavated tunnel, and samples from the kitchen and midden were used.
Neither analysis used material excavated from the 2009 field season. Two squares and one quadrant on the floor were potentially affected by the previous excavations: S4E3, S5E3, and the southwest quadrant in S3E3. However, excavation in 2011 revealed that none of the 2009 squares were excavated down to floor planks, with a 7-10 centimeter matrix overlying the planked floor in these squares. All other previously excavated squares ended well within the fill layers.
Material above the side benches was not included in the analyses because it was not originally identified as Level 3 and thus was only screened through 1/4-inch mesh and collected as a whole
164 square. It was considerably higher up than the other material, which also affected the preservation of the faunal remains.
6.3.1 General Subsistence Analysis
In order to provide a robust interpretation of the general subsistence of the people living in Feature 87, as much of the material as could confidently be attributed to the earliest occupation of the semi-subterranean house was included (Appendix A). For the house occupation, material considered above wall fall was not included; only material found below the wall fall was used in analysis. Two quadrants from the burnt area were used; these were collected as bulk samples and then screened in the lab. These two samples came from the floor level of the kitchen; this level t was dated to the same period as the house floor (Figure 17). Since these were collected in the lab, which is different from the rest of the house samples, they will likely have a higher overall recovery than can be expected from field-recovered samples. This will likely bias the kitchen samples through collection of higher numbers of smaller bone pieces and greater percentage of collection in general. Despite the bias toward greater collection frequencies, the kitchen is included in the general subsistence analysis; however, the inclusion of this area may alter many of the analyses for the floor. Additionally, the inclusion of the kitchen area material may bias the samples if compared to other Thule houses where no kitchen was excavated. Therefore, the kitchen material is sometimes removed from analyses such as fragmentation and burning in order to assess the house without the effects of the kitchen material. The total house analyses are always included; explanations are given for analyses where the kitchen has been removed.
Midden material is also included in the general subsistence analysis. Although many different factors may have affected the midden, such as greater carnivore gnawing or weathering, it represents a part of the subsistence that is important to account for, since much of the original material from the house may have been left or disposed here. Material used in the analysis for general subsistence comes from Level 2e, which is dated to the same time as the house floor occupation, cal. 1294-1394 CE and has a stylistic indicator of an early Thule occupation (Sicco harpoon head).
In order to investigate the difference between house floor and house fill, a square was identified from top to bottom. S3E4 was chosen as it was one of the only squares fully within the house walls that were previously unexcavated. It sits over the east side of the rear bench. As stated
165 above, all levels above Level 3 were screened through 1/4-inch mesh; this change in screen size may affect the analyses, but should still give a good indication of the overall differences between fill and floor material. This was the only fill material identified and analysed in this study.
6.3.2 Expectation Testing
Selection of animal remains for the expectation testing was more conservative. Only bags with full provenience were used (square, level, and quadrant), and again, only those found below wall fall were used. Material from the burnt area was used from the bulk sample of the southwest quadrant of two squares. Midden material only comes from the southwest quadrant of Level 2e (described above), as it is the only part of the square with both a specific quadrant and screened through 1/8-inch mesh. In order to test the expectations, the house was divided by quadrants into archaeological recognized areas (Figure 19). These followed as closely as possible the limits of identified features; however, there is some overlap as some quadrants have a small part within another area. A list of quadrants per area is found in Appendix B.
6.4 Some Definitions and Discussions
Before proceeding with the analysis, a number of terms and definitions need to be outlined. The most basic term needed is for analyzed bones. A specimen in this dissertation means an individual skeletal remain, either complete or not (Lyman 2008:4–8). Identifiable or identified specimens mean identified to a biological taxon and represented by a complete or partial bone, tooth, or shell (Driver 2011; Lyman 2005; 2008). The specifics, such as the taxonomic level that the specimen was identified to, will be outlined in each analysis.
6.4.1 Taxonomic Abundance Measures
Since taxonomic abundance measures are the basis for all subsequent analysis, these are defined and discussed here. The two most basic and frequently used taxonomic abundance measures are number of identified specimens (NISP) and minimum number of individuals (MNI). NISP is the most fundamental unit by which faunal remains are calculated (Lyman 2008:27). Normally, NISP only includes specimens identified to taxonomic order and skeletal element (Lyman 2008:27). Another measure is the number of specimens (NSP); it includes all the specimens identified to a specific taxon. Some analyses do include the unidentified fragments (i.e. bones
166
Figure 19: Archaeologically defined areas based on quadrants, Feature 87
167 identified to taxon but not skeletal element); where these are used, it will be indicated as NSP. MNI is a derived calculation that indicates the minimum number of individuals in that taxon as represented by their skeletal remains. Neither of these taxonomic abundance measures are accurate representations of the animals used in the past, but can be compared and contrasted, and also used to derive other calculations and results. At best, these taxonomic abundance measures are an ordinal-scale measure; although they can be presented in ratio-scale terms, they can only be interpreted at an ordinal scale (Lyman 2008:72). Although some of the problems associated with both NISP and MNI are outlined below, NISP is the preferred taxonomic abundance measure as it is a fundamental taxonomic measurement.
NISP is an observed measure, and so is not subject to the pitfalls of other, derived measures of taxonomic abundance. However, it does have its own problems that have been discussed widely in zooarchaeological and paleontological literature (e.g. Bokonyi 1980; Chaplin 1971; Gautier 1993; Grayson 1973; 1979; Lyman 2008; Ringrose 1993). Some of the more important issues are the variability in number of bones for different taxa, the effects of differential recovery, the effects of butchering patterns, and the interdependence of skeletal remains (Lyman 2008:29– 30,36–38). Most of these problems can be mitigated analytically. NISP itself is a poor indicator of diet as one whale bone may be more important that 100 arctic ground squirrel bones. Thus, it is important to note that NISP is not an absolute indicator of the animals or animal proportions people were consuming.
MNI, a derived measure, is often used to counteract the problem of specimen interdependency found in NISP. MNI is most often calculated as the most commonly occurring kind of skeletal element of a taxon, thus it helps avoid counting the same individual twice (Lyman 2008; Ringrose 1993). Like NISP, MNI also has problems (Fieller and Turner 1982; Grayson 1973; 1979; 1984; Klein and Cruz-Uribe 1984; Lyman 2008; Plug and Plug 1990; Ringrose 1993; Turner and Fieller 1985) including the difficulty and difference used in calculating it, and the fact that is often exaggerates the importance of rare taxa. One of the most significant problems is the fact that different aggregates of specimens in a collection will result in different MNI values (Lyman 2008:45–47). MNI was calculated for the fill layer and floor layer of S3E4 (Appendix A; Table A.1). For the general subsistence analysis, MNI was calculated based on the entire occupation, including the midden and is used only in the discussion of overall subsistence at the site. By aggregating all the semi-subterranean occupation data into one collection for MNI, the
168
MNI counts the absolute minimum number of individuals that could have been brought to the site and used in Feature 87 (Appendix A; Table A.2). This is likely a gross under-representation of the actual number of animals procured and used during the occupation.
Another measure of taxonomic abundance is the amount of meat weight represented in the faunal collection (White 1953). Meat weight estimates the amount of meat or consumable tissue represented by a zooarchaeological collection (Lyman 2008:89). For each taxon, the average biomass of the animal is calculated than multiplied by the number of individuals of that taxon, and then by the proportion of the animal that is considered edible. Meat weight estimates are most often based on MNI calculations (Betts and Friesen 2013; Friesen and Arnold 1995a; Grayson 1984; Jackson 1989; Lyman 2008; Reitz and Wing 2008; White 1953), although bone weight has also been used (Barrett 1993; Reed 1963; Uerpmann 1973). Bone weight is a controversial taxonomic abundance that is not used in this analysis (Grayson 1984; Lyman 2008). There are numerous problems inherent in using meat weight estimates, including using MNI as the basis for calculations, estimating the proportion of edible tissue regardless of cultural preferences, averaging taxa weight throughout the year, across ages and sexes, and assuming that the entire animal was actually eaten (Lyman 2008; Reitz and Wing 2008). Meat weight is at best an ordinal measure and only provides broad estimates of the actual meat contributions to the diet. This measure, analyzed in conjunction with MNI and NISP, helps interpret the taxonomic make- up of the assemblage (Appendix A; Table A.3).
6.4.2 Skeletal Measures
The basis of MNI is a skeletal measure known as the minimum number of elements (MNE); it is also the basis of another important skeletal quantitative unit described below and known variously as MAU or %survivorship. A simple definition of MNE is that it is the minimum number of skeletal elements or portions necessary to account for the specimens in the analysis. The calculation of MNE has undergone decades of debate in the literature (Andrews 1990; Binford 1981a; Binford 1984; Klein and Cruz-Uribe 1984; Klein et al. 1999; Lyman 1984; 1994; Marean and Frey 1997; Marean and Spencer 1991). All of this debate has led to calls to be explicit about how MNE measures are calculated (Grayson 1984; Lyman 1994; 2008; Reitz and Wing 2008); a full description of MNE calculations for this study are given in Appendix A, Section 1.2, Tables A.1 and A.2). Generally, side, fusion, portion (both epiphyseal and
169 diaphyseal), and size were used to calculate MNE. Both MNE and MNI suffer from the same problems; they are only a minimum count and thus are hard to statistically compare, and they are influenced by sample size, aggregation, and definition. Again, as with most other measures, MNE is at best an ordinal scale (Lyman 2008).
MNE is used to measure skeletal part frequencies and is the basis for another important measure, minimum animal units (MAU), originally defined by Binford (1984:50). MAU values are calculated by adding the MNE values together for a single element and dividing these by the number of times it occurs in a complete skeleton (Lyman 2008); this standardizes the MNE counts to individual skeletons. MAU values are often normed by dividing all MAU values by the greatest observed MAU value and multiplying them by 100. This results in %MAU values, or %survivorship. In these analyses, the normed MAU values are labelled %MAU. Norming elements make graphs easier to interpret and compare across areas or species (Lyman 2008). Calculations of MNE, MAU, and MNI for this assemblage are found in Appendix A, Section 1.4 (Tables A.4-A.10). It is important to note that for calculations of MAU, the MNEs of all small seals (Histriophoca fasciata, Phoca largha, Pusa hispida, Histriophoca/Phoca spp., Phoca/Pusa spp., and small seal identifications) were added together in order to get general small seal MAU values. This general small seal MAU value is used in all the analyses. Only certain small seal elements were identifiable to species, and so using only these elements would drastically skew the results of skeletal frequencies in favour of species identifiable elements.
6.4.3 Bone Density and Utility Indices
There are a number of different quantitative measures for interpreting element distributions, as calculated by MNE, MAU, or %MAU. The most common are those that compare bone density and food utility to elements; these comparisons are most often presented as indices and are typically compared by scatter plots and statistical correlation. When compared as scatterplots, the index values are plotted along the x-axis, while the element frequencies are along the y-axis; these are compared visually to hypothetical curves and interpreted. Statistical correlations are normally calculated by Spearman’s rank order correlation coefficients (rs). Both food utility indices and bone density indices are used as a frame of reference for assessing skeletal part frequencies (Lyman 1994). Details of which indices were used are described when they are used. Bone density analyses have been an important part of zooarchaeological interpretations as
170 experimental research and ethnoarchaeological data have shown that a number of taphonomic processes are mediated by bone density (Binford and Bertram 1977; Haynes 1983; Lyman 1994; Marean and Spencer 1991). Density-mediated attrition is a term that covers the myriad of factors that can destroy bones based on density; analysis beyond element correlation to bone density is needed to unpack which processes or actions were likely responsible for this attrition. Maximum values for both bone density and utility indices are plotted against the maximum taxonomic MAU values. A full list of index and MAU values is outlined in Appendix A, Section 1.5 (Tables A.11-A.14).
6.4.4 Taxonomic Composition
For the fill and floor comparative analysis, the material is analyzed using indices of heterogeneity and evenness, which measure diversity of an assemblage. Heterogeneity measures the number of different taxa within the samples; evenness measures the distribution of specimens within each taxa. Together these measures can be used to assess taxonomic diversity comparatively. There are several measures of taxonomic heterogeneity, with the most popular in zooarchaeological analyses being the Shannon—Wiener index, often referred to as the Shannon index (Lyman 2008). Taxonomic evenness analyzes how individuals are distributed across taxa; the one most commonly used in zooarchaeological literature is the Shannon index of evenness (Lyman 2008). Both of these indices are calculated using NISP. In this analysis, both were calculated with PAST software (Hammer et al. 2001).
6.4.5 Modification Analysis
Modification analyses were performed at the specimen level. For example, cut mark frequencies are based on the number of specimens with identified cut marks, regardless of the number of cut marks on the specimen itself.
6.4.6 Fragmentation
One way to calculate fragmentation is by analyzing the fracture freshness index (FFI), as described above; unfortunately, this only works for cortical bone of land mammals, normally large ones. In order to examine the overall fragmentation, Lyman (2008) uses methods that calculate the extent and intensity of fragmentation. The extent of fragmentation looks at the proportion of specimens in a collection that are anatomically incomplete, while the intensity of
171 fragmentation looks at how small fragments are and if they are identifiable (Lyman 2008). The fragmentation index is calculated by tallying the number of specimens that are complete, subtracting them from the total NISP, and dividing the result by the total NISP to get the frequency of fragmentation (Lyman 2008). For example, if 10 out of 100 bones were complete, the fragmentation index would be (100-10)/100=0.9. Lyman (2008) defines completeness as 100 percent; however, completeness in this analysis is anything with 50 percent or above completeness. Because of this change, the degree of fragmentation is likely a low figure; it would be higher in all cases if only complete (100%) bones were used. In any case, fragmentation is best used as a comparative measure to look at differences between the fill and floor material and between areas on the floor. Intensity of fragmentation is tested by dividing the total number of identified specimens (NISP) by the total number of specimens (NSP). The resulting frequency indicates how much of the assemblage was identifiable, which suggests how intensely the assemblage was fragmented. Used together, these fragmentation measures can be used along with other factors to interpret taphonomic patterns, burning patterns, and human actions.
6.4.7 Expectation Indices
Since documentary expectations have rarely been explicitly tested in the faunal record, the following methods have been created to test the expectations. All expectations are all set out as indices. Indices can either be ratios or frequencies, depending on how the expectation is tested.
Ratios refer to measures in which certain elements or sets of elements are compared to other elements. Specific taxonomic skeletal ratios (i.e. large terrestrial mammal high- to low-utility elements) are compared to the expected ratio of the skeleton for that animal. For example, for high- to low-utility ratio of large terrestrial mammal elements, the expected normal skeletal ratio is 74:45, or 1.6. If the ratio for that area is above the normal ratio (1.6), then there are more high- utility elements than low-utility elements compared to normal skeletal representation; if the number is less, there are more low-utility elements present. Some ratios do not have a normal skeletal value against which to compare them; fish vertebrae and ribs are so variable that there are no average vertebrae or rib counts. These ratios cannot be compared to a normal skeletal ratio, but can only be compared between areas; these are called comparative ratios.
Frequencies are calculated by dividing the number of bones by the total number in the assemblage for that area. For example, to calculate skeletal frequencies of bird wings and legs,
172 the normal numbers of wing and leg bones are added together than divided by the total number of bird bones in an average skeleton. For bird wings and legs, the frequency is 40/110=0.4. Skeletal frequencies results from the faunal data are compared with expected normal skeletal frequencies. Like the ratios, some frequencies cannot be compared to a normal skeletal frequency, and so can only result in relative comparisons. These frequencies are called comparative frequencies.
Both skeletal ratios and skeletal frequencies can be confirmed absolutely in comparison to the normal skeletal ratios/frequencies; these can also both be confirmed relatively by ranking the results based on comparisons to other areas. Neither comparative ratios nor comparative frequencies can be confirmed absolutely by comparing them to a normal skeletal measure; the comparative expectations can only be confirmed in comparison to other areas. Both calculations and absolute and/or relative confirmation of the ratios and frequencies are tested and discussed in Chapter 8. Both specific calculations (Tables A2.1-A2.7) and the bones used for each category (i.e. high utility, wings and legs) are listed for each taxon in Appendix B, Section 2.3 and 2.4.
6.4.8 Sample Size
Sample size affects archaeological and ecological measurements (Cannon 2001; Drennan 2009; Grayson 1984; Grayson and Delpech 1998; McCartney and Glass 1990; Orton 2000; Robertson 1999; Thompson 1987; Van der Veen and Fieller 1982). Studies of sample size for faunal assemblages has mainly focused on how sample size relates to taxonomic diversity (Amorosi et al. 1996; Banning 2002; Drennan 2009; Lyman 2008; Orton 2000), and the effect of sample size on derived measures such as taxonomic proportions (Grayson 1978; 1981; Jackson 1989; Lyman 2008) and utility indices (Faith and Gordon 2007). For establishing taxonomic diversity, Amorosi et al. (1996:133–134) suggest a sample size of 300-400 specimens. For the general analysis, all assemblages are well above this threshold. Faith and Gordon (2007) suggest that for food utility indices, the minimum number of elements (MNE) needed to asses transport strategies is 250; however, lower sample sizes only lowers the ability to determine the significance of correlation of the bulk transport strategy. Other than the bulk transport strategy, an MNE of 50 is considered adequate to establish other transport strategies. Therefore, it is only when a bulk transport strategy (the majority of animal elements brought back to the site) is identified that a
173 sample size below 250 MNE will be insufficient; all other sample sizes above a MNE of 50 will be considered representative.
The ratios and frequencies used in testing the expectations relating to spatial patterning are also affected by sample size. For all areas, a general sample size of above 350 will be considered adequate to compare the areas to each other, following Amorosi et al. (1996). However, there has been nothing written that I have found during my research about a representative or adequate sample size within taxa when comparing faunal remains across space. This type of analysis is different than acquiring adequate taxonomic representation or the ability to interpret elements in transport strategies. Using comparative ratios lessens the impact of small sample sizes; even with small sample sizes, ratios can often be compared among areas. However, in order to establish meaningful analyzes of the differences in ratios and frequencies across areas, the threshold of 50 specimens is used for this analysis. This is an arbitrary threshold that is expected to show variability and patterns of bones within taxonomic categories. When analyzes involve taxa with fewer than 50 specimens, the expectation is still analyzed, but is not considered as meaningful.
6.4.9 Numerical and Statistical Issues
Statistical significance is tested where appropriate. In this analysis, the statistical significance threshold is when the p value is <.05. For display and writing purposes, most numbers are rounded to one decimal place in text discussion and two decimal places in tables; however, when some values are extremely small, numbers are rounded to the second or third decimal place in text and tables, respectively. In all cases, calculations are based on the entire number, rather than the rounded number.
6.5 Analytic Framework
A number of frameworks have been proposed recently for understanding not only the consumption practices associated with animal remains, but also the reconstruction of contexts of consumption and treatment of animals (Bar-Oz and Munro 2004; Marciniak 2005; Orton 2012). These frameworks look at ways to understand the human-animal relations of the past, which is often described as social zooarchaeology. They also advocate for using multiple lines of evidence in order to resolve problems of equifinality. The multivariate taphonomic approach suggested by Bar-Oz and Munro (2004) proposes a three-part framework that focuses on the
174 impact of taphonomic agents on the assemblage; this framework is geared towards untangling the equifinality of density-mediated attrition. They propose three phases to frame the analysis: descriptive, analytical, and comparative. Orton’s (2012) framework also uses multiple lines of evidence to understand the taphonomic processes, but incorporates the descriptive, analytical, and comparative phases within each of his five proposed stages. This type of framework offers a more holistic understanding of the assemblage and provides more direct information on human activities. Although the framework is broken down into five stages where each stage informs the subsequent stages, in reality, the stages overlap and provide feedback to the others (Orton 2012). This type of analysis is similar to the bootstrapping methods highlighted by Wylie (2002b) where multiple lines of intertwined evidence can be used to arrive at interpretations of the material.
The five stages of Orton’s (2012) framework are: Stage 1—Density-mediated attrition; Stage 2— Peri-depositional damage; Stage 3—Breakage and fragmentation; Stage 4—Visible human modification; and, Stage 5—Assessment of element representation. Within these stages, most of the typical zooarchaeological analyses can be completed and assessed. The benefit of using this framework is to not only a way to organize the overlapping data, but also to integrate a more holistic understanding of the assemblage. This framework generally moves from interpreting the taphonomic history of the assemblage to understanding human effects on the assemblage. By the time skeletal frequencies are analyzed, human actions often can be distinguished from taphonomic agents. In the following analyses, this framework is used as an organizational and mnemonic tool to interpret the assemblages. However, a holistic understanding of an assemblage must go beyond these five stages, and so two other stages are included to bracket the five proposed by Orton: archaeological context and taxonomic representation. Although Orton (2012:322) does suggest that excavation and recovery biases should be taken into account for assemblages with limited documentation, this information is important to contextualize the data. Information about the site and excavations, the observations of the excavators, matrix of the assemblage, and other associated data is important when interpreting the remains. Even within the same site, differential recovery techniques or identification procedures may produce very different results. A final analytical stage has also been added that looks at the taxonomic representation of the assemblage. Orton (2012:232) includes assessment of taxonomic representation in his fifth stage, assessment of element representation. However, taxonomic frequencies can help interpret overall subsistence patterns, seasonality of occupation, and other
175 broader aspects of the archaeological site and therefore are useful to discuss separately from anatomical representation.
6.6 Summary
In order to understand the human actions that lead to the current faunal assemblage, the historical context of the assemblage has to be reconstructed. Identification procedures, taphonomic factors and faunal methodology are important when establishing the archaeological and historical context of an archaeofaunal assemblage. A holistic understanding of the assemblage history will allow a fuller understanding of human actions associated with the deposition and creation of the faunal assemblage.
This chapter has outlined the methods used to analyze general subsistence and spatial patterning. The following two chapters will present the results of the faunal analyses through the framework described above and based on these basic definitions and methods.
176
Chapter 7 Overall Subsistence and Economy
This chapter will outline and interpret the general zooarchaeological assemblage, specifically as it relates to site formation processes, subsistence and economy, and general human-animal relations. First, the fill and floor material will be compared to see if there is a difference in the two assemblages. Second, the overall subsistence of the house will be interpreted by combining all floor, kitchen, tunnel, and midden squares. Both analyses loosely follow Orton’s (2012) analytical framework for social zooarchaeology as described in Chapter 6. Spatial differences will be discussed in Chapter 8.
7.1 Defining the Floor: Fill and Floor Comparison
The first portion of the chapter will explore the differences in life history of the fill and floor material within one square in order to outline the clear boundaries of the floor. Establishing a clear floor-level faunal assemblage is critical in understanding the nature of subsistence strategies and human-animal relations at the time of house occupation. Fill material often has a plethora of bone and artefact material, which is sometimes used to establish subsistence patterns for the people who originally occupied the house. If the fill material is related to the floor-level occupation, then this is a valid dataset for the current analyses; however, differences and similarities between fill and floor material must be established rather than assumed. This analysis will pinpoint the processes and actions that may have led to the current fill and floor zooarchaeological assemblages.
7.1.1 Archaeological Context
In order to compare the floor and fill material, one metre square was chosen within the house that did not have evidence or records of excavations prior to 2011 (S3E4). The only known disturbance prior to excavation was in the southwest quadrant where one of the upright whale bone mandibles was dug into the house depression. This may have moved some material vertically and horizontally. However, the disturbances were not significantly noticeable during excavation and would have only affected a small portion of the extreme southwest corner.
177
In this square, the floor material was easily distinguished due to the excellent preservation of the fallen roof and walls. Material that is below wall fall and was identified as Level 3 (floor) is considered floor material, while any material above the wall fall and identified as Level 2 (fill) is considered fill material. This floor level is dated to ca. cal 1300-1400 CE based on a date from an adjacent square at floor level (Figure 17). In the fill, the material right above the wall fall is different than the uppermost fill levels based on differences in preservation, artefacts, and features, although there is no distinct separation. In the uppermost levels, the excavator noticed that the bones were softer and more fragile than those in the lower fill levels. These uppermost levels (Levels 2-2e) also had a number of distinct concentrations of artefacts, faunal remains, and features, including an area of burnt bones, dark soil, and charcoal and a distinct area of wood chips. In the lower fill levels, there were fewer distinct features and the bones were not noticeably softer or fragile. Additionally, these lower fill levels (Levels 2f-3) had a different soil composition; in these levels, there were banded, vertically-oriented strata that contained charcoal, burnt bone, and lithic material. The orientation of these banded strata dips towards the centre of the house, potentially indicating that they were collapsed roof material and post- abandonment activity areas. The purpose of this analysis is to see if there is a difference between the fill and floor rather than distinguish differences in all occupations; therefore, all the levels that are above the wall fall have been combined into a fill layer (Level 2, 2b, 2c, 2d, 2e, 2f, 2g, and 3) and all the levels below the wall fall into a floor layer (Level 3b, 3c, 3d). Although there are no direct dates on any of the fill levels in S3E4, an upper level was dated from a nearby square outside of the house (S6E2) to 535±15 (Figure 17). This date, at level 2e (70-80cm below datum 1), corresponds roughly to Level 2b (50-60 cm below datum 2 +20cm). This upper level is slightly younger than the floor.
All levels were excavated by trowel and screened. The floor levels were screened through 1/8- inch (3mm) mesh, while the fill levels were screened through 1/4-inch (6 mm) mesh for the Level 2’s and 1/8-inch (3mm) mesh for Level 3.
178
7.1.2 Density-Mediated Attrition
Bone density is used as a first step to measure the amount of density-mediated attrition that may have affected the site, both from natural and cultural processes. Density-mediated attrition, as described in Chapter 6, is analyzed in two ways: by scatterplot and by statistical correlation. Both are important in this analysis. Bone density was compared to small seal bone Figure 20: Fill, small seal bone density; rs=0.45, frequencies (Histriophoca fasciata, Pusa p=.52; MAU=4; MNE=255 hispida, Phoca largha, and any other small seals) based on minimum animal units (MAU)14. MAU for the fill and floor material was plotted against bone mineral density for phocids (Chambers 1992 in Lyman 1994:248) and Spearman’s rank order correlation coefficient was calculated to test for correlation and significance for the fill (Figure 20) and floor material
(Figure 21). For the fill material there is a Figure 21: Floor, small seal bone density; rs=0.65, positive, but insignificant correlation. For p=<.01; MAU=4; MNE=305 the floor material, there is a positive, significant correlation, which indicates that there is an association between seal bone survivorship and the densest seal bones. Many human and non- human factors can lead to skeletal frequencies that are positively correlated with bone density. Non-human factors include carnivore attrition (Binford and Bertram 1977; Brain 1969; Brain 1976; Brain 1981), mechanical destruction (trampling, sediment overburden, freeze-thaw action) (e.g. Guadelli 2008; Lyman 1994), and fluvial transport (Behrensmeyer 1975); human factors
14 See Chapter 6 for methodology and Appendix A for full calculations of MAU and minimum number of elements (MNE) for small seals
179 include extraction of bone grease and bone marrow (Binford 1978a; Brain 1976; Church and Lyman 2003; Lyman 1993; Outram 2001; Outram and Umberto Albarella 2005), differential transport (e.g. Binford 1978a:323; Lyman 1994:257), or use of bones as tools (e.g. Lyman 1984). One of the interesting features of the floor is the presence of fish cranial remains that are much more susceptible to density-mediated attrition than mammal remains (Butler 1993; Butler and Chatters 1994). The presence of fragile fish remains suggests that factors other than uniform natural processes affected the floor assemblage. The insignificant correlation of bone density for small seals in the fill material does not negate the possibility that this layer was affected by density-mediated attrition, just that the small seal skeletal frequency does not reflect this. Factors that may have impacted bone density correlations are discussed in the fill and floor summary in conjunction with other lines of evidence. For now, it is important to identify the difference in density-mediated attrition between fill and floor material. Along with the evidence below, it indicates that the taphonomic factors were quite different for both the fill and floor material.
7.1.3 Peri-Depositional Damage
Modification frequencies can help discern the different processes that may have affected the fill and floor material. Natural taphonomic indicators, such as weathering, root etching, and gnaw marks show slight differences between the fill and floor material (Figure 22). The weathering frequencies, although low in both layers, affect a higher number of bones in the fill than the floor layer, which would be expected if bones in the fill were exposed to the elements for longer periods of time than those from the floor. Slightly surprising is the lack of root etching on the fill material but the low percentage in the floor Figure 22: Fill and floor natural modification material. There is very little actualistic frequencies (%NISP) (Fill n=722; Floor n=498) or taphonomic research on the effects of root etching on faunal material (Denys 2002; Lyman 2008), although it is most often seen on bones from shallow buried assemblages (O’Connor 2000) where bones spend some time in the active root zone (Todisco and Monchot 2006). In the Arctic, root etching is often associated with the active layer (e.g. Betts 2007), that layer of
180 sediment that freezes during the winter but thaws during the summer. The lack of root etching on bones from the fill layers may indicate that burial for most of the assemblage was sufficiently rapid to remove them from the active root zone (Todisco and Monchot 2006). It has also been suggested that bones with root etching may indicate a period without much disturbance (Cook 1986); conversely, a lack of root etching does not prove that bones were subject to greater disturbances. Based on the evidence that some of the bones were weathered (<6.5%), it is likely that the fill material was exposed to the elements for a longer period of time than the floor material; however, the lack of root etching on the fill material may suggest that this material was disturbed, which would reduce the formation of root etching. For the floor material, weathering and root etching occur in equally low frequencies (<1%). A limited amount of exposure could produce limited weathering and root etching; however, this exposure likely did not occur within the house, as opposed to outside of the house. The deep burial of the floor in square S3E4 and lack of disturbance under the wall fall makes it highly unlikely that there was vertical mixing, and the limited numbers of bones that were weathered or root etched supports this interpretation. One interpretation is that defleshed bones were stored outside where they underwent some weathering and root etching and then only a few were brought into the house for further processing and butchery.
The low frequencies of gnaw marks for both fill and floor material makes it difficult to assess differences or similarities. The higher frequency of gnaw marks on the floor indicates that carnivores or dogs may have affected the assemblage at some point in the taphonomic history, while the lack of gnaw marks in the fill suggest that carnivores had little, if any, impact on the fill assemblage. Carnivore destruction is at least partially density-mediated (Faith et al. 2007; Marean and Spencer 1991; Marean et al. 1992), and so higher frequencies of carnivore gnawing should lead to higher correlation between bone mineral density and bone survivorship (Faith et al. 2007), as seen for small seal remains on the floor (Figure 21). However, the low gnaw mark frequency and presence of low density fish cranial remains suggests that gnawing did not affect the whole assemblage. Other factors, such as taxonomic abundance, anatomical frequencies, butchery patterns, seasonality, or site use, must be affecting the differences in bone assemblage in addition to limited carnivore activity.
181
7.1.4 Breakage and Fragmentation
Difference in fragmentation for the fill and floor assemblage is tested by examining the extent and intensity of fragmentation, as described in the methodology (Lyman 1994). The fragmentation index shows the amount of bone that is fragmented compared to the amount of whole bone (Figure 23). Mammals and birds were used to calculate bone fragmentation, with fragmented bones counted as anything less than 50 percent complete. The fill material is more than twice as fragmented as the floor material. However, this Figure 23: Fill and floor index does not use the unidentifiable bones, and so does not fragmentation (%NISP) (Fill n=722; Floor n=498) indicate the intensity of fragmentation. One way to calculate the intensity of fragmentation is to divide the total number of identified specimens (NISP) by the total number of specimens (NSP); the higher the number, the less intensely fragmented the assemblage. The ratio for the fill is 0.51 (51% identified) and for the floor it is 0.76 (76% identified) (Figure 23). Both measures of fragmentation show that the fill is both more intensely and extensively fragmented, while the floor has less fragmentation in general and more complete bones specifically. The greater fragmentation in the fill likely was produced by a number of factors, including natural processes such as weathering and trampling described above, and burning and other human factors described below.
7.1.5 Visible Human Modifications
Figure 24 shows the difference between the human modifications of the floor and fill material. The large disparity in burning modifications (colour, shrinkage, and cracking) likely has to do with the differences in the uses of each layer (fill and floor). The upper layers have evidence of a hearth, with charcoal, burnt bones, and stained soil. In the lower layers, there is limited evidence of burning; this was expected as this floor material comes from the bench while most of the Figure 24: Fill and floor cultural modifications (%NISP) (Fill cooking and bone burning is found in the kitchen (see Chapter n=722; Floor n=498)
182
8). Burning and fragmentation may be correlated here as burning makes bone more brittle. This can lead to breakage during or after the burning episode (Lyman 1994; Shipman et al. 1984; Clark and Ligouis 2010; Costamagno et al. 1999; 2005; Medina et al. 2012; Stiner et al. 1995).
Butchery marks are low in both assemblages, with a higher frequency in the floor material. This difference in distribution may be a result of many factors, including spatial, taxonomic, and fragmentation differences. The floor material comes from a bench area, where expectations state there should be higher frequencies of butchery marks due to tool and garment manufacture and consumption practices (see Chapter 5). Although there is not enough data to suggest the activities that occurred in the upper layers, outside areas were often used for primary butchery and storage, which may leave fewer cut marks than consumption and manufacturing practices (Binford 1978a; Lyman 1994; but see Lyman 2005). The butchery marks indicate that a greater degree of consumption and manufacture was occurring on the floor, potentially with some primary butchery and storage activities occurring in the fill; however, no concrete interpretation of fill material can be made with such a limited spatial exposure.
7.1.6 Element Representation
Element representation is presented for small seals as all other species had low NISP values. Skeletal frequencies can be compared between the two layers based on bar graphs showing the MAU values (Figure 25); statistical tests are also performed to examine correlations. The
Figure 25: Fill and floor small seal element distribution (%MAU) (Fill MNI=4, MNE=255; Floor MNI=4, MNE=305)
183 anatomical representation of small seals is significantly correlated between the fill and floor material (rs=0.5; p=.01). In both assemblages, the appendicular elements (i.e. radius, ulna, tibia- fibula, metatarsals, phalanges) are better represented than the axial elements (vertebrae, ribs, innominates). However, there are some differences, including the relative frequency of tibia- fibulas and front phalanges.
Element representation can be compared to a meat utility index for ringed seals (Pusa hispida) (Figure 26) (Diab 1998). Meat utility indices compare the bone survivorship to the expected frequency of bones if they were only selected for meat and other consumable products. Both fill and floor small seal assemblages show slight negative, insignificant correlations to meat utility
(Fill: rs=-0.3, p=.34; Floor: rs=-0.5, p=.07). This lack of correlation indicates that the bones represented in the assemblage are not the highest meat utility elements. Likely other human and non-human factors were affecting the distribution of small seal elements.
Figure 26: Fill and floor small seal meat utility index (Fill MNI=4, MNE=255; Floor MNI=4, MNE=305)
184
Comparisons between fill and floor material for other animals were not valid, as either the fill or floor had fewer than ten elements. Although birds are well represented in the fill (NISP=75, MNI=14), they are poorly represented in the floor assemblage Figure 27: Fill bird element distribtuion (MAU) (MNI=5; MNE=50) (NISP=7, MNI=4). This precludes a statistically valid comparison between the two layers, but emphasizes the difference in the avian class between the fill and floor material. In the fill material, birds are represented by high frequencies of wing and leg bones (Figure 27). Other than birds, only arctic ground squirrel (Spermophilus parryii) remains in the fill have more than 10 specimens. Arctic ground squirrel element frequencies are uneven, but indicate that likely only one individual was present. This is supported by the fact that, when the elements could be analysed for age, they all had unfused epiphyses, some with juvenile cortex around the fusion line. These remains are found throughout the fill (Level 2, 2c, and 3), which would indicate that different taphonomic and human actions may have mixed or vertically moved some of the faunal assemblage of the fill.
7.1.7 Taxonomic Representation
Gross differences and similarities between the fill and floor taxa were tested using a test for difference (two-sample t-test=t-test) and a test for correlation (Spearman’s rank-order correlation coefficient=rs). The two-sample t-test used the NISP values of common groups to test for difference; this statistical test compares the mean of each sample to test for significant difference. Based on the NISP values of common groups, there is no statistical difference between the two layers (t-test=0.04, p=.97). The Spearman’s rank-order correlation coefficient used the NISP values of the common groups to test for correlation based on rank order; this statistical test measures statistical dependence between the fill and floor NISP values. Based on NISP values of common groups, there is no statistical correlation between the two layers (rs=0.2, p=.58). Despite the lack of evidence for distinct differences, there is also little evidence for correlation. The lack
185 Fill Floor Fill Floor Taxon NISP %NSP %NISP NISP %NSP %NISP MNI %MNI MNI %MNI Gastropoda 5 0.56 3 0.55 1 2.63 2 6.06
Bivalvia 2 0.23 17 3.12 1 2.63 2 6.06
Gastropoda/Bivalvia 20 2.26 18 3.30
Mollusca Total 27 3.05 38 6.97 2 5.26 4 12.12
Clupea pallasii 0 0.00 0.00 2 0.37 0.48 0 0.00 1 3.03 Salvelinus alpinus 0 0.00 0.00 1 0.18 0.24 0 0.00 1 3.03 Coregoninae 0 0.00 0.00 1 0.18 0.24 0 0.00 1 3.03 Salmonidae 0 0.00 0.00 2 0.37 0.48 0 0.00 0 0.00 Gadidae 0 0.00 0.00 1 0.18 0.24 0 0.00 1 3.03 Actinopterygii 0 0.00 4 0.73
Actinopterygii Total 0 0.00 0.00 11 2.02 1.69 0 0.00 4 12.12
Anser spp. 0 0.00 0.00 1 0.18 0.24 0 0.00 1 3.03 Goose 0 0.00 0.00 1 0.18 0.24 0 0.00 0 0.00 Somateria spp. 2 0.23 0.44 0 0.00 0.00 2 5.26 0 0.00 Melanitta spp. 1 0.11 0.22 0 0.00 0.00 1 2.63 0 0.00 Clangula hyemalis 8 0.90 1.77 0 0.00 0.00 2 5.26 0 0.00 Duck 12 1.35 2.66 0 0.00 0.00 1 2.63 0 0.00 Anatidae 1 0.11 0.22 1 0.18 0.24 0 0.00 1 3.03 Lagopus spp. 7 0.79 1.55 0 0.00 0.00 2 5.26 0 0.00 Calidris alpina 1 0.11 0.22 1 0.18 0.24 1 2.63 1 3.03 Calidris spp. 1 0.11 0.22 0 0.00 0.00 1 2.63 0 0.00 Gull 0 0.00 0.00 1 0.18 0.24 0 0.00 1 3.03 Sterna hirundo 1 0.11 0.22 0 0.00 0.00 1 2.63 0 0.00 Sterna paradisaea 7 0.79 1.55 0 0.00 0.00 3 7.89 0 0.00 Sterna spp. 6 0.68 1.33 0 0.00 0.00 0 0.00 0 0.00 Tern 10 1.13 2.22 0 0.00 0.00 0 0.00 0 0.00 Charadriiformes 5 0.56 0 0.00
Aves 13 1.47 2 0.37
Aves Total 75 8.47 12.64 7 1.28 1.21 14 36.84 4 12.12
Spermophilus parryii 17 1.92 3.77 0 0.00 0.00 3 7.89 0 0.00 Castor canadensis 0 0.00 0.00 1 0.18 0.24 0 0.00 1 3.03 Ondatra zibethicus 5 0.56 1.11 0 0.00 0.00 2 5.26 0 0.00 Lemming* 27 0 0.00 6 15.79 0 0.00
Arvicolinae* 8 0 0.00 7 18.42 0 0.00
Lepus spp. 2 0.23 0.44 0 0.00 0.00 2 5.26 0 0.00 Rodentia 1 0.11 0.22 0 0.00 0.00
Canis lupus 2 0.23 0.44 13 2.39 3.15 2 5.26 3 9.09 Vulpes spp. 3 0.34 0.67 11 2.02 2.66 2 5.26 3 9.09 Erignathus barbatus 6 0.68 1.33 5 0.92 1.21 2 5.26 2 6.06 Histriophoca fasciata 2 0.23 0.44 0 0.00 0.00 1 2.63 0 0.00 Phoca largha 0 0.00 0.00 3 0.55 0.73 0 0.00 2 6.06 Pusa hispida 9 1.02 2.00 10 1.83 2.42 2 5.26 3 9.09 Pusa/Phoca spp. 9 1.02 2.00 9 1.65 2.18 1 2.63 0 0.00 Pusa/Phoca/Histriophoca spp. 334 37.70 74.06 338 62.02 81.84 3 7.89 4 12.12 Neovison vison 0 0.00 0.00 1 0.18 0.24 0 0.00 1 3.03 Rangifer tarandus 3 0.34 0.67 10 1.83 2.42 1 2.63 2 6.06 Ovibos moschatus 1 0.11 0.22 0 0.00 0.00 1 2.63 0 0.00 Cetacea** 70 0 0.00
Marine mammal 3 0.34 6 1.10
Terrestrial mammal 10 1.13 24 4.40
Mammalia 290 32.73 45 8.26
Mammalia Total 697 78.67 87.36 476 87.34 97.09 22 57.89 21 63.64
Mammalia/Aves 87 9.82 12 2.20
Vertebrates 0 0.00 1 0.18
Indeterminate Vertebrate Total 87 9.82 13 2.39
Total 886 545 38 33 *Probable intrusive taxa; excluded from all calculations **Probable intrusive specimens from upright whale mandible; excluded from all calculations
186
Table 9: Fill and floor taxonomic frequencies: whole values (NISP, MNI) and relative values (%NSP, %NISP, %MNI) (Table on previous page) of difference in taxonomic means is likely a result of the limited number of taxa available in the Arctic region, as well as the fact that seals that dominate both assemblages (Table 9). The lack of correlation is interesting, and is explored in the taxonomic representation analysis below. Despite the lack of difference in taxonomic representation, the assemblages are noticeably different in other taphonomic modifications, human modifications, and skeletal frequencies as described above.
Another way to measure similarity and difference is through diversity indices such as heterogeneity and evenness. The heterogeneity index used here is the Shannon—Wiener index, which normally varies between 1.5 and 3.5; larger values equal greater heterogeneity and more individual taxa (Lyman 2008:192). Evenness measures how individuals are distributed by taxa. Here, evenness is calculated as the Shannon index of evenness, which can vary between zero and one; a value closer to one indicates that the individuals in taxa are equally abundant (Lyman 2008:195). When all of the identified taxa are used, diversity indices show that the fill material is slightly more heterogeneous and even (Table 10a). When only mammal diversity is compared, the differences in heterogeneity almost disappear; both fill and floor have low heterogeneity and low evenness (Table 10b). Birds show a greater heterogeneity and lower evenness at the fill level than any of the other taxa (Table 10c). However, the bird NISP for the floor is only seven, which limits the statistical validity for the bird material. The differences in these diversity measures are minimal, but are interesting when combined with other taphonomic, modification, and skeletal evidence as discussed below. a All Taxa Fill Floor Heterogeneity 1.31 0.92 Evenness 0.15 0.12 b Mammal Taxa Heterogeneity 0.75 0.74 Evenness 0.16 0.21 c Bird Taxa Heterogeneity 2.24 1.61 Evenness 0.72 1.00 Table 10: Diversity indices (NISP of the lowest taxonomic order)
187
The mammal frequencies are similar between the fill and floor layers. Mammal remains contribute to most of the identified remains in both the fill and floor material at 90% and 96% respectively (Figure 28). Unidentified shell remains contribute small, but noticeable amounts to the fill and floor layers (3% and 7% respectively).
Differences in the bird and fish remains are interesting; for the fill, there are no fish remains, but a substantial Figure 28: Fill and floor vertebrae class percentage of bird remains (10%), while for the floor, frequencies (%NISP) (Fill n=772; Floor n=489) there are small, equal percentages of fish and bird remains (2%). The lack of fish in the fill layers may be a result of mechanical, chemical, and carnivore destruction. The limited degree of carnivore gnawing on all the remains indicates that this may have impacted the fish assemblage to a very limited degree; alternatively, fish may have been abundant but may have been destroyed by dogs, which would leave very limited carnivore marks and very few fish remains. More likely, the factors that produced the high degree and intensity of fragmentation for other vertebrate remains may have destroyed the less dense fish remains (Butler and Chatters 1994; Lyman 1984; 1994; Wheeler and Jones 1989; Willis et al. 2008). Fish may never have been discarded in the fill material due to seasonal and occupational differences. Ethnographically, fish are processed differently in each season, with all bones except ribs being discarded at the procurement site in the summer; in the winter, fish freeze quickly and are brought back to the site whole (Burch 2006:146). Based on this small sample, the absence of fish in fill material may indicate that it was a warm season occupation, while the presence of cranial and post-cranial fish remains on the floor may support the interpretation that it was a cold-season occupation.
Birds make up more of the fill assemblage than the floor (10% in the fill, 2% in the floor). In the fill, ducks, ptarmigan, and terns dominate the assemblage, while geese make up most of the floor assemblage. In both assemblages, most of the birds are migratory, except ptarmigan. Both assemblages also contain dunlin (Calidris alpina), a small shorebird.
188
Based on NISP, small seals make up most of the assemblage in both fill and floor assemblages (Figure 29). When MNIs are calculated, small seals still make up most of the floor assemblage, but are second to furbearers in the fill material (Figure 30). Small seals were the focus of
Figure 29: Fill and floor vertebrate taxonomic frequencies both fill and floor subsistence (%NISP) (Fill n=450; Floor n=413) practices.
Artiodactyla remains are not frequent in either the fill or floor. The fill has fewer caribou (Rangifer tarandus) remains, but has one instance of a muskox (Ovibos moschatus). A number of factors may be affecting the frequency of artiodactyla remains, such as transport strategies from kill sites in the interior, or that fact that neither Figure 30: Fill and floor vertebrate taxonomic frequencies subsistence strategy actively (%MNI) (Fill MNI=37; Floor MNI=29) pursued these animals.
Furbearers are ranked second after seal by NISP in both fill and floor material, while they rank first by MNI in the fill assemblage. The number of different furbearers in both fill and floor raise their MNI values; when even just one bone is present, one individual is counted. Other than foxes, none of the furbearers overlap between the two assemblages. Furbearers in the fill include arctic ground squirrels, muskrat (Ondatra zibethicus), hare (Lepus spp.), and fox (Vulpes spp.). Furbearers from the floor assemblage are beaver (Castor canadensis), fox, and mink (Neovison vison). Mink, muskrat, and beaver most often live in the forests, meaning that people either travelled into the forested regions during their yearly round or traded for these animals. Animals linked to forest habitats are found in both the fill and floor material, which indicates that there
189 were close ties through trade or travel with other parts of the region in different seasons and at different times of occupation.
Lemmings and specimens identified to the subfamily Arvicolinae are not included in the calculations or figures but are interesting in terms of taphonomy. They were excluded from the analysis because they are burrowing rodents, with little economic or social value to Iñupiat people. They likely represent intrusions into the site. Arctic ground squirrels are also potentially intrusive; however, they were important sources of fur for garments and have been included in these calculations. In any case, lemmings, ground squirrels, and Arvicolinae are only found in the fill layer; none of the specimens were found on the floor. Excavation notes detail a number of rodent burrows in fill levels that corroborates the intrusive nature of the lemming bones. These burrows, along with the bones, indicate that rodents may have been moving material around in the fill layer. These burrows could only be dug in unfrozen soil, which suggests that the fill levels were at some point in the active layer.
Cetacea (whale) remains were also not included in the analysis. All of the remains were in the fill and were collected near to the upright whalebone mandible in the southwest corner. All were marked on the level forms as having flaked off the mandible upright. Again, these remains are interesting to note because they help construct a holistic view of the formation of the assemblage.
7.2 Fill and Floor Comparative Analysis Discussion
Taken together, these analyses show that there were different occupation histories, taphonomic processes, and taxonomic compositions for the fill and floor assemblages. Based on the contextual analysis, the fill represents a palimpsest of occupation, while the floor represents a discrete occupation. The earliest deposition of fill material was likely from the use of the house roof during the house occupation. After the house was abandoned and collapsed, there were a number of other occupations, although with limited spatial analysis, it is difficult to tell what type of activities these occupations represent. In the middle of the fill material is a concentrated area of burnt bone, charcoal, and seal oil, indicating that this was used at one time for cooking or heating. Additionally, there was likely a work area, as evidenced by a layer of wood chips. The most noticeable feature during excavation was the placement of two large baleen whale mandibles in the fill. Although initially thought to be part of the structural support of the house, excavation revealed them to be situated above the floor material, on top of the roof fall. These
190 pieces likely represent the remains of a raised rack, which was common in the documentary record in the area (Burch 2006:119, 120; Lee and Reinhardt 2003:77, 78; Ray 1975a:173). At some point, pieces of the whale bone broke off into the fill material; these were collected next to the whalebone mandible and were identified as part of the mandible itself. The material on the floor is much shallower, and was likely deposited over a shorter period of time. The floor material was found below wall and roof fall, which indicates that it was likely deposited during the house occupation, potentially over as little time as a few months. The lack of carnivore gnawing, weathering, and structural scavenging suggests that the house was left mostly undisturbed after abandonment. This occupation history, along with the lack of taphonomic processes described below, indicate that human decisions were likely the most important factor in the creation of the floor assemblage.
The analysis shows that taphonomic processes affected the fill material to a greater degree than the floor material. Although not all of the fill material has indications of weathering, some of the bones have flaking, bleaching, and cracking that indicate that they were exposed to the elements. However, the lack of root etching on fill material indicates that these weathered bones were subject to disturbances during the weathering process and prior to vegetation formation. This is not surprising in the Arctic, given the slow process of plant growth, vegetation cover, and soil formation (Ping et al. 2008). A secondary line of evidence that indicates that the fill material was affected by taphonomic processes is the fact that in some of the fill layers, the bones were so fragile that they disintegrated during the excavation process; this was not the case for the floor material. Although all microenvironments are different (Sutcliffe 1990), the more frequent freeze-thaw cycles in the active layers of the fill likely led to higher frequency of fragmentation of mammal and bird bones (Figure 23) and decreased frequency of identifiable bones (Guadelli 2008; Ping et al. 2008). The effect of freeze-thaw cycles on archaeological assemblages has had limited study. Studies that do look at the freeze-thaw effect focus on movement and cryoturbation, especially of artefacts (Carr 1984; Denys 2002; Hilton 2003; Goebel et al. 2013; Masson 2010; Todisco and Monchot 2006; but see Matthiesen et al. 2014; Milne 2003). A recent study by Guadelli (2008) showed that over the course of hundreds of freeze-thaw cycles, bones and teeth can be rendered unidentifiable. This type of weathering action leaves very few marks on the bones, but results in dramatic changes to the faunal record (Guadelli 2008). The mechanical destruction of bones through freeze-thaw cycles, along with trampling and burning,
191 may account for the high degree of fragmentation in the fill as compared to the floor. Additionally, the thawing of the fill layers would have allowed other chemical weathering processes to affect the bones, as well as the mechanical process of disturbances of rodent burrowing.
The floor material has quite a different depositional history than the fill, based on archaeological context, taphonomy, and human factors. The very low frequencies of weathering and root etching on the floor, the presence of fragile fish remains, and the low rates of fragmentation indicate that the assemblage was not exposed to the same types of processes or rates of attrition as the fill. There is no evidence of intensive fragmentation, intrusive rodents, or vertical mixing. Additionally, there is a clear layer of collapsed wall and roof material above the floor material. During the occupation and re-occupation of the floor, material may have been subject to processes such as freeze-thaw and trampling found in the active layer, but to a much lesser degree than material in the fill. After abandonment, the thick wall and roof overburden would have collapsed and covered the material quickly and deeply, likely leading to rapid incorporation into the permafrost. After abandonment and collapse, the house floor material was not likely disturbed to any great degree; permafrost, encountered at the floor level during excavations, would have limited the mechanical and chemical destruction of bones, while low frequencies of carnivore gnawing and structural scavenging indicate that human and animal actions were negligible.
Overall, the floor and fill material are very similar in their taxonomic composition; the people who occupied both areas were hunting small seals for the majority of their subsistence. Although taxonomic frequencies are not significantly different between the fill and floor material, they do show some interesting features. Lemming and/or vole remains are completely absent from the floor but are second most abundant in the fill material (Figure 21 and Figure 20). These remains, along with rodent burrows identified during excavation, indicate that the ground was not frozen and was accessible for part of the year, while the floor material was likely covered and frozen quickly which prevented rodent action. This corroborates the interpretation that taphonomic processes affected the fill to a greater extent than the floor assemblage. Taxonomic frequencies for the fill levels show that, as with the floor, people were hunting small seals for a majority of their subsistence needs. However, the difference in bird frequencies and small furbearers suggests that the fill levels may have been used more frequently in the warm season, when there
192 were a greater number of bird species present on the landscape. One of the primary resources at Cape Espenberg currently are birds, both migratory and nesting (Kessel 1989; Schaaf 1988). People who deposited the fill layer material were likely taking advantage of these resources. However, this general subsistence strategy for the fill is hindered by the palimpsestic nature of the fill levels; analysis of the rest of the fill material both horizontally and vertically would be needed to identify fine-grained differences and changes through time and space. In comparison, the floor level, which will be discussed below in general subsistence analysis, shows a heavy reliance on marine mammals, specifically small seals. There is limited evidence for bird remains, and those that are present have likely been stored from the warm season. Similarly, fish remains are present at the floor level likely due to the acquisition and differential processing of fish in the winter and use of stored summer fish, but also the excellent preservation of the floor material.
7.3 General Subsistence for Feature 87
The floor and fill comparison has shown that there are many differences and some similarities in archaeological context, architecture, taphonomy, and faunal composition. This analysis uses the entire floor assemblage (main room, kitchen, tunnel, and midden) to analyze general subsistence patterns. Based on archaeological context, floor material can be identified over most of the main room and tunnel. As described in Chapter 6, kitchen and midden material is also included in the general subsistence analysis in order to get the most complete interpretation of the faunal assemblage. Here, material was collected from the levels that date to the house occupation. Differences across space will be the focus of Chapter 8; however, in some cases, the kitchen material may dominate the assemblage as it was filled with a large number of unidentifiable burnt remains. In these cases, the kitchen is removed and analyses are calculated for the house with and without the kitchen. This analysis will follow the social zooarchaeological approach proposed by Orton (2012), while including information about the excavations, taxonomic frequencies, and identification to create a framework to interpret the assemblage holistically (see Chapter 6).
7.3.1 Density-Mediated Attrition
As described in the fill and floor analysis of density-mediated attrition, bone density is analysed by scatterplot and statistical correlation. In all of the analyses, the minimum animal unit (MAU) was used to compare against bone density measures. Phocid seal bone density was used to
193 compare bone density in small seals; as described in Chapter 6, the inability to identify to species whole classes of bones (i.e. vertebrae, phalanges, metacarpals, etc.) would drastically skew the results if only specimens identified to species were used. Caribou bone density was used from density measures based on the mean shape-adjusted density values (BMD2) from caribou (Lam et al. 1999:357; Lam et al. 2003). Bird bone was plotted against an average duck element volume density (bone mineral content/volume; Broughton et al. 2007). Although combining the species is not ideal, the use of only the duck specimens resulted in very few data points. The duck average was chosen for two reasons: first, ducks were the most abundant taxa identified in the assemblage; and second, the average duck density values include both high density (diving ducks such as Mergus merganser) and lower density ducks (dabbling ducks such as Anas acuta; Broughton et al. 2007:383). Since none of the species are directly comparable to the ducks in the assemblage, this analysis cannot be assumed to be as accurate as the small seal or caribou analyses. Fish bone density was also plotted to see if there were differences among the classes. Fish bone density (VD) and bone mineral content (BMC) were taken from studies on Chinook salmon (Oncorhynchus tshawytscha) (Butler and Chatters 1994:417)15.
Bone density indices show variable correlation to bone survivorship depending on the taxon. Figure 31 shows a positive, significant correlation between small seal bone density and survivorship for the whole house floor level assemblage. Caribou remains also show a positive, statistically significant correlation to bone density, though the scatter plot is wedge-shaped (Figure 32). Bird bones also have a statistically significant correlation (Figure 33). Both caribou and bird wedge-shaped scatterplots indicate that although the less dense remains are rare, the denser elements are not necessarily common (Orton 2012). This implies that although density- mediated attrition is an important factor, other factors also likely affected the assemblage. As described in the fill and floor bone density section, these other factors may be taphonomic (e.g. carnivore attrition, trampling) or human actions (e.g. extraction of bone marrow, making bone tools). Interpretations of these bone density indices are discussed along with other lines of evidence in the general subsistence summary. Fish remains show no significant correlation to bone density (VD) or bone mineral content (BMC) (Figure 34), probably due to the limited
15 All of the values for the bone density and utility indices can be found in Appendix A, Tables A.11-A.14.
194
Figure 31: Small seal bone density Figure 32: Caribou bone density rs=0.77, p<.01; MAU=30; MNE=2675 rs=0.61, p<.01; MAU=3; MNE=110
Figure 33: Bird bone density Figure 34: Fish bone mineral content rs=0.64, p=.13; MAU=4; NISP=79 rs=0.21, p=.48; MAU=3.5; NISP=192 Bone density (VD): rs=0.12, p=.69
number of remains (n=192). However, the most frequent fish remains were vertebrae (n=172), the highest density bones besides otoliths; otoliths were the second most frequent (n=7). Seals, caribou, and birds all show a positive correlation to bone density; however, there are differences in the scatterplots between seals and caribou and birds. These differences, along with the presence of fragile fish remains and excellent wood preservation indicates that less dense bones were not likely being destroyed by universal causes, such as weathering or freeze-thaw action. Other actions, likely human, created these different bone density profiles. There are many factors that could affect bone density including processing and sharing, taphonomic factors, and disposal factors (Lyman 1994). In order to investigate this further, modification frequencies and fragmentation indices need to be examined.
195
7.3.2 Peri-Depositional Damage
Figure 35 shows the natural modification frequencies recorded for this assemblage. All three processes can impact bone density in assemblages by destroying less- dense remains; the low frequencies of all three in this assemblage (<1% of NISP) indicate that they had little impact on the assemblage as a whole. This supports the interpretation from the comparative analysis of fill and Figure 35: General subsistence natural modification frequencies floor material that most bones on the floor were not (%NISP) (NISP=23462) exposed to the elements for a significant length of time prior to burial and was not within the active root layer. Multiple lines of evidence, as described in the comparative analysis, indicate that most of the house material was buried quickly and entered the permafrost. The only exception to this may be the kitchen material and the midden material. As described in Chapter 6, the kitchen material was closer to the surface, which may have put it in the active layer. The midden may have been exposed for longer periods of time at some points, or it may have also been buried quickly. These differences will be explored in Chapter 8.
7.3.3 Breakage and Fragmentation
Fragmentation in the house assemblage is tested for extent and intensity of fragmentation. The fragmentation index measures the extent of fragmentation in an assemblage and is most useful as a comparative tool. However, it can be used, along with other measures, for a single assemblage to get a general sense of the preservation and state of destruction of the material. When analysed by class for the whole house, fish and mammals are the Figure 36: General subsistence fragmentation index (%NISP) (Total least fragmented, with birds showing fragmentation house NISP=23462; House, no kitchen for almost half of their remains (Figure 36). NISP=10832, Kitchen NISP=12630) Fragmentation analyses of the whole house are important; however, the high number of fragmented
196 remains from the kitchen may skew the fragmentation results. Therefore, the fragmentation index was also used to look at the extent of fragmentation for the house with no kitchen material and for the kitchen alone. Figure 36 shows that the mammal frequencies show a much greater percentage of fragmented remains in the kitchen (36%) than in the rest of the house (16%); this percentage of mammal remains for the rest of the house is fairly similar to the total house remains fragmentation (18%). Birds are fairly similar across the three different groupings, but fish show a lower fragmentation percentage for the kitchen than the rest of the house or the total house remains. Fragmentation extent is not greatly impacted by the kitchen material.
The index for extent of fragmentation does not account for the unidentifiable bones that were not given a completion percentage. Another measure of fragmentation is percentage identified, which measures the intensity of fragmentation. A simple way to calculate this is by dividing the NISP by the NSP. For the assemblage as a whole, only 26% of the bones were identified to family taxon or lower. For the kitchen alone, only 4% of the bones were identified, while for the rest of the house, just over 50% of the remains were identified. This indicates that the kitchen material is much more intensely fragmented than the rest of the house, and that this material does affect the total house fragmentation intensity. With an identification rate of only 26% for the whole house, a large number of small, unidentified remains make up most of the assemblage, which are mostly found in the kitchen. The fragmentation index suggests that fragmentation does not affect all bones equally; those that were identified were mostly whole. These two indicators suggest that when bone was broken, it was comminuted into small bits, especially in the kitchen.
Another measure of fragmentation is the fracture freshness index (FFI) used for land mammal bones as described in Chapter 6. The FFI uses three indicators to measure the freshness of a bone: fracture angle, fracture surface, and fracture outline (Outram 2001; 2002). Each indicator is given a score from zero to two, which are then added together to Figure 37: Fresh fracture index (FFI) for get a score from zero to six for each broken identified land mammal cortical bone (n=157) land mammal diaphysis. The lower the score
197 on the index, the closer the bone was to being fresh, with higher numbers indicating less fresh bone (Outram 2001; 2002). Freshness is impacted by a number of factors including time from death, attached meat, or heating. This index has been used as a line of evidence to identify marrow and bone grease production (Outram 2001; 2002) and bone used as fuel (Théry-Parisot et al. 2005). In this assemblage, an FFI was recorded for identified land mammal cortical bones. As seen in Figure 37, FFI scores of three dominate, but scores of one and two are well represented, with scores above three less frequent. This indicates that most fractures occurred after the bones were partially dried or heated (Binford 1978a; Outram 2001). However, the average FFI score is just under three (2.98; n=157), which means that most of the fragmentation was done on near-fresh bones. This may indicate that the fragmentation pattern was due to marrow extraction and bone-grease rendering (Outram 2001; 2002) when bones were partially dried or heated prior to breakage.
7.3.4 Visible Human Modifications
Figure 38 shows three types of cultural modifications to bones: cut marks, impact scars, and burning. Although there is not much difference in the cut marks or impact scars between the whole house and the house with the kitchen removed, the burning modifications are quite different. Thus, Figure 38 shows the cultural modifications in three categories: total house; house, no kitchen; and kitchen. Like weathering, root etching, and gnawing modifications, the butchering marks are extremely low, with less than one percent of the bones Figure 38: General subsistence cultural modification frequencies affected in the total assemblage. Birds had the highest (%NISP) (Total house NISP=23462; relative frequency of cut marks (1.6%), although mammal House, no kitchen NISP=10832, Kitchen NISP=12630) remains had the highest number of elements with cut marks (n=57). Cut marks were found on a variety of animals, elements, and element locations. Most of the cut marks were found on the appendicular skeleton of birds and mammals. One interesting feature is that three of the five elements with cut marks are identified to large baleen whale (suborder Mysticeti). Cut marks, in association with other factors, have been used to support the hypothesis for active whaling (Betts and Friesen 2013). Birds and walrus also have
198 high frequencies of cut marks, which may be a result of differences in processing type and intensity, skill levels of people processing the animals (i.e. children butchering birds), or physiology. Dog/wolf (Canis lupus) elements had indications of skinning marks on the mandibles and distal limbs (Val and Mallye 2011).
Burning is the most ubiquitous modification in the assemblage, affecting over sixty percent of the total specimens for the entire house (Figure 38). If the kitchen material is removed from the analysis, burning only affects 23% of the rest of the assemblage. When looking at only the kitchen material, burning affects 93% of the remains. This drastic difference will be explored more in Chapter 8, but does show that burning was an important factor in the Figure 39: Burning frequencies formation of the faunal assemblage in the kitchen, and less so by vertebrate class (%NISP) in the rest of the house. However, despite this drastic (Total n=23457) decrease, it is interesting to note that burning affects a quarter of the counted bone in the rest of the house. Using the total house burning frequencies, it can be seen that burning is found on all vertebrate categories except birds (Figure 39). However, the most frequently burnt material is from the bird/mammal (n=10853) and indeterminate category (n=189). Although it is not possible to securely identify whether the burning fragmented the bones or the bones were fragmented prior to burning, most of the indeterminate bone was burnt: over 60% of the indeterminate mammal, over 91% of the bird/mammal, and over 50% of the indeterminate. Only around 7% of the mammal bone identified to family or lower was burnt, with even lower percentages for fish and no burnt bird. All of the indeterminate burnt bird/mammal bone and over 97% of the burnt mammal bone was 100% burned; this indicates either high intensity burning broke bones into pieces in the fire, or that bones were broken into pieces prior to entering the fire (Costamagno et al. 2010; Costamagno et al. 2005; Medina et al. 2012; Figure 40: General subsistence burnt bone by size category and class (%NISP) (n=14285) Shipman et al. 1984; Stiner et al. 1995). Size
199
also corresponds to the degree of burning; the smallest bones (<1cm) are most frequently burned in all classes (Figure 40). In fact, almost all (92.6%) of the indeterminate bone less than one centimetre was burnt; over two-thirds of the mammal bone less than one centimetre was also burnt. As the bone fragments get larger, the burning decreases (Figure 40). This trend, in addition to the percentage of total burnt
Figure 41: General subsistence bone for these categories, indicates that mostly small burnt bone by bone type (NISP) specimens that cannot be identified to a particular taxa or (n=2798) skeletal element are burned. Most of the burnt remains were either cancellous (51%) or indeterminate (42%) (Figure 41); within the bone type categories, the majority of both cancellous (63%) and indeterminate (71%) bones were burnt. These numbers suggest that cancellous bone was preferentially burnt. Experiments and archaeological research have shown that spongy bone is a better source of fuel than compact bone (Morin 2010; Théry- Pasrisot and Costamagno 2005; Théry-Pasrisot et al. 2005), with bone often broken prior to being used as fuel (Costamagno et al. 2010).
7.3.5 Element Representation
Anatomical frequency was assessed for small seals, bearded seals, caribou, dog/wolf, fox, and birds using minimum animal units (MAU). If the whole animal was killed and brought back to the site, these profiles should be fairly equal. In reality, the skeletal distributions are rarely even. Differences in MAU frequencies can be caused by both natural and cultural factors, such as differential destruction, transport, butchery practices, and sharing. MAU frequencies can also be compared to known utility indices to see if the bone survivorship of that particular species matches the expected frequencies for certain human behaviours. The individual utility indices are described below.
200
7.3.5.1 Small Seals
The MAU profile for small seals is fairly even, although it does show that axial elements (cranium to innominate) are, as a group, underrepresented (Figure 42). Hind limbs are the most
Figure 42: General subsistence small seal element distribution (%MAU) (MAU=47; MNE=2675)
commonly occurring elements, with the distal elements such as metatarsals and hind phalanges the most frequent. This pattern has no correlation with the meat utility index (MUI) for ringed seals (Pusa hispida) (Figure 43) (Diab 1998). When compared to Iñupiat preference ranking (Diab 1998), there is no correlation Figure 43: Meat utility index: rs=0.39, p=.15 (MAU=69) (Figure 44). The ringed seal MUI and preference ranking was used since the majority of small seals identified to species were ringed seals. Meat utility indices have been generated for the eastern arctic harp (Phoca groenlandica) and hooded (Cystophora cristata) seals (Lyman, Savelle, et al. 1992), which may be a better representation of spotted (Phoca largha) or ribbon (Histriophoca fasciata) seals based on Figure 44: Iñupiat preference ranking: rs=0.36, p=.27 (MAU=34) body shape and meat distribution; however,
201 neither of these indices has been tested on the Alaskan seals. No other indices have been produced for Alaskan seals.
These insignificant, uncorrelated relationships for small seals may be a function of multiple factors. The first is bone density; since MAU is correlated to bone density, destruction of less dense bones likely affects the meat utility index (Diab 1998). Although neither weathering nor gnawing likely affected the assemblage to a great degree, selection of these more porous, less dense bones for burning may have impacted the anatomical representation. Sharing among extended families was common in the documentary record (Burch 2006; Omura 2013), and this may also be a cause of differential skeletal representation. Although the lack of heads may be a result of transport decisions, the presence of low-utility flippers counters this claim; heads and hind flippers were both removed for transport in the documentary records (Burch 2006:154). The high number of hind flippers may indicate a preference for them in Feature 87; alternatively, it may indicate that this was the portion of the seal that was shared with the people in the house. The lack of vertebrae is also suggestive of other actions, such as use of vertebrae for dog feed or fuel. These lines of evidence will be brought into conversation with other pieces of evidence to assess the strongest interpretations in the summary. Alone, these frequencies and indices can only suggest interpretations.
7.3.5.2 Caribou
Caribou element skeletal representation shows a fairly even appendicular distribution, with a general low frequency of axial elements (Figure 45). One exception to this is the high frequency of cervical vertebrae, specifically the axis (second cervical vertebrae). This element
Figure 45: General subsistence caribou element distribution (MAU) (MAU=3,MNE=110)
202 representation represents a positive significant correlation with bone density as described above; however, the bone density scatterplot shows that although the least dense bones are rare, the densest remains are not always abundant (Figure 32). Multiple utility indices have been developed for caribou that may be able to interpret the skeletal frequencies better than bone density. The caribou food utility index (FUI) is a simplification of Binford’s (1978a) modified general utility index (MGUI), which is intended to predict which body parts of large land mammals would remain at, kill sites and which would be transported back to the residential sites (Metcalfe and Jones 1988). A positive correlation of MAU with FUI will indicate that people were transporting the bones associated with the highest food utility (meat, fat, marrow, and bone grease) back to the site, while a negative correlation would indicate that a majority of bones were low utility, and thus the assemblage may represent a kill site. When compared to the FUI, there is no significant correlation (Figure 46). There is an insignificant negative correlation, which indicates that there is greater frequency of low utility bones than high utility bones. The lack of correlation to the FUI means that Figure 46: General subsistence caribou food something other than choices based solely on utility index: r =0.41, p=.10 (MAU=3; s food utility led to the survivorship of caribou MNE=110) elements in the assemblage.
Another utility index that is often used is the meat drying index (MDI) (Friesen 2001). This index is similar to the original drying utility index (Binford 1978a) that looked at which bones are selected in which order during meat drying activities. Friesen’s (2001) MDI looks at a number of issues that affect which bones
Figure 47: General subsistence caribou meat accompany meat when it is dried including drying index: rs=0.61; p<.01 (MAU=3; total amount of meat attached to the bone, the MNE=110) ratio of bone weight to meat weight, and the amount of marrow and brain that accompanies
203 the bone package. As with the food utility indices, the elements with high MDI values are expected to be stored for drying more frequently than those elements with low MDI values. Element distributions are negatively, and significantly, correlated to the caribou MDI (Figure 47). This means that the most frequently occurring bones are the ones that are least likely to be dried for storage. The caribou bones that have survived in the assemblage were not those with the highest food or drying utility.
Neither the food utility nor meat drying index is positively correlated to the density index; the meat drying index is negatively correlated with the density index (Friesen and Stewart 2013). This means that the bones most likely to be selected for drying are also low in density; therefore, assemblages such as this with a negative correlation to the MDI and a positive density correlation could either have been affected by density-mediated processes or have a negative drying profile. As discussed above, other taphonomic processes such as weathering or gnawing did not likely affect the assemblage, and so there is no other evidence corroborating the fact that the assemblage was exposed to destructive agents. Therefore, there must be other factors contributing to the anatomical frequency.
The presence of lower limb bones may indicate that these bones were being saved from early fall to late spring caribou kills for marrow processing and bone grease production. In this season, caribou were hunted for food and were therefore processed carefully and thoroughly (Burch 2006:138–139). In the fall through spring, if the kill site was a long distance from the settlement, caribou were cached; eventually, most of the caribou, including the lower limbs, were transported back to the winter village site (Burch 2006:138). Depending on the temperature, the meat was dried, fermented, or frozen. The lower leg bones were saved in order to render bone grease or soup stock (Burch 2006:138). Alternatively, the goal of summer hunts was usually to acquire hides; most often, the majority of the caribou meat and all of the bones were left at the kill site (Burch 2006:139).
The bones that are the lowest in frequency are bones that are porous and full of grease, such as vertebrae, ribs, and sternum (Burch 2006:138). These porous bones may have been used as fuel in fires. The low density, porous axial elements are more useful for fuelling fire than high- density elements (Théry-Pasrisot and Costamagno 2005), with land mammals more useful in fuelling fires than marine mammal bones (Vaneeckhout et al. 2013). These axial elements were
204 also used to produce bone grease (Binford 1978a:158). Processing of axial bones for bone grease involves smashing them into small pieces and then boiling them in water; the fat floats to the top where it solidifies and is skimmed off as it cools (Binford 1978a:158). This will leave small pieces of unidentifiable mammal cancellous bones. When analyzed by bone type, cancellous bone makes up the majority of the Figure 48: Genereal subsistence indeterminate mammal bone indeterminate mammal bone (Figure 48). (NISP=9312)
A test of the interpretation that caribou bones were being saved from cold season kills for their bone grease and marrow properties is to compare the element frequencies to the unsaturated marrow index (UMI) (Morin 2007). The UMI is based loosely on the principles of Binford’s (1978a) marrow index. The UMI is used to predict the elements preferred for marrow extraction. When compared to caribou element representation at Feature 87, the index is strongly positively and significantly correlated to the UMI (Figure 49). This indicates that the most frequently occurring bones are also those with the highest preference for use in marrow extraction. In the documentary record, it is noted that bones were brought back to be used for bone grease and marrow extraction, which may be occurring here (Burch 2006:138). As Burch mentions, bones were saved until the end of the season for bone grease extraction, which matches the FFI Figure 49: General subsistence caribou histogram (Figure 37) with some bones broken unsaturated marrow index: rs=0.70, p<.01 fresh, while most were less-fresh when broken.
7.3.5.3 Canidae
The skeletal representations of dog/wolf and fox elements are different, even though the NISP and MNI counts are very similar (Figure 50). The only similarity is the high frequency of heads for both taxa. Axial and appendicular elements are more equally represented for dog/wolf than for fox. The hind limb is the most frequent, with lower parts of both hind and fore limbs well
205 represented for dog/wolf. In contrast, fox remains are more uneven, with high and low representation of both axial and appendicular elements. Statistically, the assemblages are not significantly different (t-test=-1.2, p=.24) or correlated (rs=0.2, p=.45) when dog/wolf and fox MAU values are compared. Both the element frequency graphs and the statistical tests indicate that these taxa were treated in both similar and different ways.
Figure 50: General subsistence dog/wolf (NISP=181, MAU=2, MNE=138) and fox (NISP=137, MAU=2, MNE=102) element distribution (%MAU)
In the documentary record, foxes, wolves, and dogs were only eaten in times of famine (Burch 2006:175, 176, 283). Foxes and wolves were mainly hunted for their pelts for clothing, trim, and as amulets (Burch 2006:175, 176). Dogs that died were skinned and their pelts were used as parka trim, or pants (Burch 2006:283). The high frequency of cranial remains for both taxa indicates that the canids probably entered the house assemblage whole, and were then processed for skins, and potentially for food. All of the dog/wolf elements are fused with no juvenile cortex, indicating that they are adult. The position of the cut marks on the mandibles and distal ends of long bones indicates skinning and processing (Val and Mallye 2011). The skeletal frequency supports this with a fairly equal representation of all remains. Additionally, this distribution may indicate that the dog/wolf individuals were kept more intact during disposal than the fox remains. Both the cut mark patterns and the age of the dog/wolf remains in the assemblage support the hypothesis that adult dogs died and were then skinned.
206
7.3.5.4 Bird
When combined, the bird skeletal frequencies show a distinct preference for long bones, which are often the densest (Broughton et al. 2007). However, not all of the densest bones survived, as seen in the wedge-shape of the density scatterplot (Figure 33 and Figure 51). Long bones may have been curated for tools, toys, or household goods (Burch 2006:184, 239, 244), while natural or cultural agents destroyed other remains.
Figure 51: General subsistence bird skeletal distributions (MAU=4, MNE=79)
7.3.5.5 Fish
When combined, fish anatomical frequencies are very uneven, likely due to the small sample size of identified elements (n=192). The only useful comparison is between cranial and post-cranial remains (Figure 52). Although the postcranial remains are more frequent, the presence Figure 52: General of cranial remains is important for understanding the season in which subsistence cranial and the fish were caught and processed, as discussed below. post-cranial distribution for fish (NISP=267) 7.3.6 Taxonomic Representation
As discussed above, there are a variety of taxa that make up the house assemblage. The majority of remains are unidentified, but those that are identified suggest that a variety of procurement strategies and techniques were being used during the house occupation. There are a number of different measures used for taxonomic abundance in this analysis: NISP, MNI, and meat weight
207
(Table 11; Figure 53). The theory, methods, and problems of each of these measures have been discussed in Chapter 6: Methodology.
7.3.6.1 Mollusca
There are a large number of shells, both gastropods and bivalves, in the sample. These shells were not identified below class, as the comparative collection at the University of Toronto does not have invertebrates from the Chukchi Sea. Because shells were not identified below class, they were not included in meat weight calculations. Mollusca (shells) contribute 2.7% to the overall NSP, nothing to the NISP (since none are identified to species), but significantly more to the MNI at 28.6%. These shells are removed from other taxonomic calculations, as invertebrate and vertebrate remains are notoriously difficult to reconcile in zooarchaeological assemblages (Reitz and Wing 2008). For now, it is interesting to note that these shells must have been brought to the house by people who collected them off the beach or in the intertidal zone during the course of their daily activities (Burch 2006:184–185). Molluscs were also eaten from the stomachs of walrus by walrus hunters at the kill site (Burch 2006:185), though this is unlikely to be the source of the molluscs in Feature 87.
7.3.6.2 Actinopterygii
Fish remains are infrequent, making up only 2.3% of the vertebrate assemblage by NISP, only 9.2% by MNI, and about 0.2% by meat weight. There are a variety of fish including herring, salmon, and tomcod (small cod-like species) (Table 11). Vertebrae make up about 75% of the fish remains, while cranial remains make up just over 25%. The presence of fragile cranial remains indicates that density-mediated attrition may not have affected the assemblage extensively; there is no significant correlation between bone density (Butler and Chatters 1994) and MAU (rs=0.12, p=.70). There is also no significant correlation between bone mineral content
MAU (rs=-0.19, p=.53).
The presence of both cranial and postcranial remains suggests that the fish were processed and eaten within the dwelling, which may be related to winter fishing. There were a multitude of ways to catch fish including using weirs, seines, dip nets, hook and lines, or bows and arrows (Burch 2006:140–145). Although fish were often a common dog food (Burch 2006:140, 146), their passage through the dog’s digestive system would have destroyed most of the fish remains;
208 Taxon NISP %NSP %NISP MNI %MNI %MNI* 78 0.32 33 18.13 Gastropoda 181 0.75 19 10.44 Bivalvia 385 1.60 Gastropoda/Bivalvia 644 2.67 52 28.57 Mollusca Total
Clupea pallasii 4 0.02 0.06 1 0.55 0.77 Oncorhynchus spp. 1 0.00 0.02 1 0.55 0.77 Salvelinus alpinus 1 0.00 0.02 1 0.55 0.77 1 0.00 0.02 0 Salvelinus spp. Coregoninae 1 0.00 0.02 1 0.55 0.77 Salmonidae 45 0.19 0.72 3 1.65 2.31 3 0.01 Salminoformes Gadus spp. 3 0.01 0.05 2 1.10 1.54 Gadidae 88 0.37 1.41 3 1.65 2.31 208 0.86 Actinopterygii Actinopterygii Total 355 1.47 2.31 12 6.59 9.23
Anser spp. 1 0.00 0.02 1 0.55 0.77 Goose 5 0.02 0.08 1 0.55 0.77 Cygnus columbianus 1 0.00 0.02 1 0.55 0.77 Somateria spp. 6 0.02 0.10 1 0.55 0.77 Melanitta spp. 1 0.00 0.02 1 0.55 0.77 Clangula hyemalis 3 0.01 0.05 2 1.10 1.54 16 0.07 0.26 0 Duck 9 0.04 0.14 0 Anatidae Lagopus muta 1 0.00 0.02 1 0.55 0.77 7 0.03 0.11 0 Lagopus spp. Gavia stellate 2 0.01 0.03 1 0.55 0.77 Gavia adamsii 2 0.01 0.03 1 0.55 0.77 Gavia spp. 5 0.02 0.08 2 1.10 1.54 Podiceps spp. 1 0.00 0.02 1 0.55 0.77 Buteo lagopus 1 0.00 0.02 1 0.55 0.77 Grus Canadensis 1 0.00 0.02 1 0.55 0.77 Calidris alpina 1 0.00 0.02 1 0.55 0.77 Calidris spp. 1 0.00 0.02 1 0.55 0.77 Larus hyperboreus 2 0.01 0.03 1 0.55 0.77 Laridae 2 0.01 0.03 1 0.55 0.77 Sterna paradisaea 4 0.02 0.06 1 0.55 0.77 Sterna spp. 2 0.01 0.03 2 1.10 1.54 53 0.22 Aves Aves Total 127 0.53 1.19 22 12.09 16.92
Spermophilus parryii 13 0.05 0.21 4 2.20 3.08 Castor Canadensis 3 0.01 0.05 1 0.55 0.77 1 1 Cricetidae** Lepus othus 2 0.01 0.03 1 0.55 0.77 5 0.02 0.08 0 Lepus spp. Canis lupus 181 0.75 2.90 6 3.30 4.62 Vulpes vulpes 2 0.01 0.03 2 1.10 1.54 Vulpes lagopus 21 0.09 0.34 4 2.20 3.08 Vulpes spp. 114 0.47 1.83 1 0.55 0.77 9 0.04 0.14 0 Canidae
209
Ursus spp. 1 0.00 0.02 1 0.55 0.77 Odobenus rosmarus 8 0.03 0.13 1 0.55 0.77 Erignathus barbatus 73 0.30 1.17 4 2.20 3.08 Histriophoca fasciata 5 0.02 0.08 2 1.10 1.54 Phoca largha 49 0.20 0.79 9 4.95 6.92 5 0.02 0.08 0 Phoca/Histriophoca spp. Pusa hispida 201 0.83 3.22 20 10.99 15.38 Pusa/Phoca spp. 181 0.75 2.90 3 1.65 2.31 Pusa/Phoca/Histriophoca spp. 4928 20.45 79.06 26 14.29 20.00 2 0.01 0.03 0 Phocidae Neovison vison 1 0.00 0.02 1 0.55 0.77 3 0.01 Carnivore, marine 1 0.00 Carnivore, terrestrial 22 0.09 Carnivora Alces alces 2 0.01 0.03 2 1.10 1.54 Rangifer tarandus 199 0.83 3.19 5 2.75 3.85 Ovibos moschatus 4 0.02 0.06 1 0.55 0.77 4 0.02 Artiodactyla Balaena mysticetus 2 0.01 0.03 1 0.55 0.77 3 0.01 Mysticeti Delphinapterus leucas 5 0.02 0.08 1 0.55 0.77 19 0.08 Cetacea 91 0.38 Marine mammal 152 0.63 Terrestrial mammal 4414 18.31 Mammalia Mammalia Total 10725 44.50 96.50 96 52.75 73.85
11877 49.28 Mammalia/Aves 373 1.55 Vertebrates 12250 50.83 Indeterminate Vertebrate Total
Total 24101 182
*%MNI without shell (MNI=130) **Probable intrusive taxa; excluded from all calculations
Table 11: Genereal subsistence taxonomic frequencies. Whole values (NISP, MNI) and relative values (%NSP, %NISP, %MNI)
Figure 53: General subsistence meat weight (%)
210 thus, the remains of the house assemblage were likely discarded as a result of human consumption. In documentary records, fish caught in the winter quickly froze and were then eaten partly frozen or put whole into soups. In either case, the entire fish would be brought back to the house (Burch 2006:146). This is in contrast to summer fish processing that was done away from the residential site, at the fishing site. There, most of the bones (except ribs, fins, and tail bones) were removed so fish could be filleted and dried (Burch 2006:146). These dried fish were stored in oil or brought back to the residential site; they had very few bones and are likely invisible archaeologically. Fish were often fermented during both cold- and warm-season processing (Burch 2006:146–147) which would remove the heads and vertebrae entirely. The presence of both fish crania and vertebrae within the house indicate that they were caught and consumed in the winter.
7.3.6.3 Aves
Bird remains make up a smaller percentage of the vertebrate NSPs than fish (1.1%), but a larger portion of the MNI (16.9%) and meat weight (0.3%) measures (Table 11; Figure 53). There are a wide array of birds, from geese, ducks and terns, to loons, raptors, and dunlins (Table 11). Waterfowl make up the majority of the remains, based Figure 54: General subsistence bird taxonomic both on NISP and MNI (Figure 54); gulls distributions (n=74) and terns are next, followed by ptarmigans and shorebirds. Additionally, there is one specimen each of a rough-legged hawk (Buteo lagopus) and a sandhill crane (Grus canadensis). If taxonomic abundance can be correlated to subsistence economy, these rank orders are very similar to those subsistence resources identified by Burch (2006:178–184) for Iñupiat in northwest Alaska. Waterfowl were economically important to 18th century Iñupiat due to the fact that this category encompasses a large number of individual species, but also because the Seward Peninsula is a breeding and migrating zone, specifically at Cape Espenberg (Kessel 1989). Waterfowl were hunted in the warm season at any time, but most frequently during migration, while they were nesting, and when they were moulting (Burch 2006:179). A number of waterfowl remains contained medullary bone (Gavia spp., and Clangula hyemalis), indicating
211 that they were caught during the egg-laying season in the spring, between late-May and early- June (Kessel 1989:63, 67, 112).
Ptarmigan (Lagopus spp.) are the only fowl species identified in the assemblage. They are an important species, especially in the late winter, as they are one of the only animals to be present year-round in this area. One specimen with medullary bone also indicates that at least one was hunted during the spring from late-May to mid-June (Kessel 1989:133, 134).
Gulls and terns together are the second most frequent bird group in the assemblage, although they only represent five individuals. Gulls and terns are seasonal migrants in the area, most likely caught in the early spring (Burch 2006:182).
Only two specimens are identified as shorebirds, but these represent two individuals. They are a type of sandpiper, found along the wave line along the coast. They again are migrants, found in the warm season along the shores; dunlins (Calidris alpina) are found on the Seward Peninsula from when the snow has started to melt from the tundra in mid-May through early-September (Kessel 1989:176–177). They could be snared or hunted with bow and arrow; Burch (2006:182) suggests that young children, specifically boys, mainly caught them with slingshots while learning to hunt. Other times they were taken opportunistically by people out hunting or in need of a change in diet.
Raptors and cranes were also seasonal migrants and were hunted in various ways. Both were eaten and used in a variety of ways; for example, raptor feathers were used as decorations, while crane wings and bones were used as tools and toys (Burch 2006:183–184).
7.3.6.4 Mammalia
The previous discussion of skeletal frequencies for mammal taxa has covered a number of important issues. This section will focus on those mammals not previously discussed and will interpret the subsistence strategies used to acquire these resources based on the presence and frequency of the mammal remains. Mammals are the most abundant class by far in all measures of abundance including both invertebrates (NSP=50.5%) and vertebrates (NSP=95.7%; MNI=73.8%; meat weight=99.5%).
212
7.3.6.4.1 Whales
Whales make up only a minor portion of the assemblage based on NISP or MNI, but are the highest ranked taxa in the assemblage once converted to meat weight (Table 11; Figure 53). There are two species of whale identified, the large baleen bowhead whale (Balaena mysticetus) and the smaller toothed beluga whale (Delphinapterus leucas). One of the major problems in Arctic zooarchaeology is how to account for bowhead or large baleen whales in Thule subsistence strategies (Betts and Friesen 2013; Coltrain et al. 2004; McCartney 1995; McCartney and Savelle 1985; Savelle and McCartney 1988; 1990; 1994; 1999; Savelle et al. 2000). Most of the research on bowhead whaling has focused on the central Canadian Arctic, where bowhead bones and whalebone house remains are a highly visible part of the archaeological record, but whaling was almost completely absent in the documentary record. In Alaska, where bowhead whaling was still practiced at the time of contact, there has been less intense investigation into the nature of large baleen whale hunting outside of the major bowhead whaling sites of Wales, Point Hope, and Barrow (but see Mason and Gerlach 1995b; Savelle and McCartney 2002; 2003). Although there is some whalebone on the surface at Cape Espenberg, none of it has been directly associated with Thule occupations. Bowhead whales were regularly pursued in the spring at only four locations in the immediate regions around Cape Espenberg in northwest Alaska: Tikiġaq (Point Hope), Uivvaq (near Cape Lisburne), Nuvua (south of Point Hope) and Kiŋigin (Cape Prince of Wales) (Burch 2006:156). Only at Tikiġaq and Kiŋigin was whaling of primary importance, although landing a whale at any site would have an enormous impact on the local economy (Burch 2006:156).
Active and regular hunting of large baleen whales has been assessed based on size, age, element representation, and cut marks. In the house assemblage, there is one phalanx identified to bowhead whale; however, there are a number of other elements identified to the suborder Mysticeti, which includes the grey and bowhead whales. These specimens are all forelimb elements (one humerus, three metacarpals, and one phalanx). This type of element representation has been suggested by Whitridge (1999b:110) and used by Betts and Friesen (2013) as one indicator of whaling at Thule sites. Although the highest ranked elements for meat utility are the hyoids, ribs, thoracic, lumbar, and caudal vertebra (Savelle 1997), these are not expected to be found at occupation sites. Most whale bones are too large and heavy to be removed from the
213 processing site, and most of meat can be removed without any attached bones (Savelle and Friesen 1996).
Although the limited number of remains in Feature 87 does not indicate regular hunting, the element representation and the presence of cut marks on two of five of the elements identified to large baleen whale suggest that these whales were not just scavenged for construction material or dog food, the only use for scavenged whales in the documentary record outside of famines (Burch 2006:156), but were potentially hunted. Betts and Friesen (2013) use both element representation and cut mark frequency as indicators of bowhead whaling specialization at sites in the Canadian Arctic. If large baleen whales were occasionally hunted, the procurement of a single bowhead whale would far surpass the meat provided by any of the other taxa in the assemblage. Like most marine mammals, baleen whales were shared among the family and village (Burch 2006:159–160, 162–163). Large bowhead bones were often re-used over many years, which may mean that these bones represent the remains of whales caught years, or decades, previous to the occupation represented by most of the faunal remains.
The other whale identified in the assemblage is the beluga whale (Delphinapterus leucas). Both vertebral and limb elements were identified, although none of these elements exhibited cut marks. As with the bowhead whales, the beluga whale hunt was not likely a focus of subsistence at Cape Espenberg during the occupation of Feature 87. They would also have been taken when they were encountered. Belugas precede bowhead whales up the ice leads north in the spring (MacDonald and Cook 2010). Spring ice leads were likely where belugas were hunted from Cape Espenberg, as the ice leads would open right along the coast depending on wind and current directions (Burch 2006:161).
If whales had been hunted during the occupation of Feature 87, they were likely caught occasionally during the early spring, when ice leads were opening up and when semi- subterranean houses were still occupied. Despite the low MNI and NISP, the meat weight measure shows that catching even just one whale would have had a substantial effect on the subsistence resources of the people in the site.
214
Figure 55: General subsistence common vertebrate taxonomic groupings (NISP=6232, MNI=129)
7.3.6.4.2 Small Seals
Small seals make up the majority of the assemblage based on NISP and MNI (Table 11; Figure 55); they rank second to whales based on meat weight (Figure 53). The high NISP, MNI, and meat weight measures suggest that small seals were the subsistence focus of the people at Feature 87. When broken down by species, the majority of the seals were ringed seals (Pusa hispida), followed by spotted seal (Phoca largha), and ribbon seal (Histriophoca fasciata), although the majority of specimens were not identified below the genus level (91.8%).
Ringed seals are ice-dependant seals that rest on ice floes and follow the ice north in the summer. Ringed seals migrate into the area when the ice begins to form in the fall and remain in the area until the summer (MacDonald and Cook 2010; Wynne 2008). Ringed seals maintain breathing holes in the land fast ice, which allows them to stay in the area throughout the winter. The other seals must follow the open water, and so ringed seals are often the only seals found in the area from freeze-up until mid-March (Burch 2006:149). In the spring and fall, seals could be hunted from the ice-edge; when freeze-up occurred, ringed seals could be hunted from leads and pools, but were more commonly hunted via breathing holes. The most productive method was to hunt seals with nets in the early fall before freeze up along the shore or after freeze up under the ice. Ringed seals were also stalked in the early spring on the sea ice or by boat in the fall (Burch 2006:148–152).
215
Spotted and ribbon seals inhabit ice floes and open water. They both summer in the open water and live among the ice floes in the winter and spring. This means that they are most accessible to humans in this area in the summer, since they follow the ice edge south during the winter months (Burch 2006:148; MacDonald and Cook 2010; Wynne 2008). Ribbon seals are more ice bound than the spotted seals, with the ribbon seal almost never coming to land. Both seals could be hunted via boats in open water or by stalking, harpooning, or netting on the ice-edge.
Small seal catches were often divided among families, although the exact division of the harvest depended on the hunt and species of seal (Burch 2006:153). Sharing may have had a large impact on anatomical frequency. Age at death can be used to pinpoint the season in which seals were caught. The age at death of small seals can be estimated based on epiphyseal fusion (Storå 2001; 2002). Storå’s (2001) work on small seal fusion identified life stages for certain small seal species in the Baltic Sea. Although the small seal species are different for the Bering Strait and Chukchi Sea region, the four broad life stages outlined by Storå (2001; 2002) should be applicable to this assemblage. The ringed seal, which is most abundant in the current assemblage, was one of the species used to develop these stages. The four stages are based on life history stages rather than calendar ages: yearling, juvenile, young adult, and old adult (Storå 2001; 2002). Only the first stage can be correlated to a calendar age (less than a year), although the juvenile stage likely refers to seals that are less than four or five years old. In this analysis, the last two stages are combined to form a general adult stage. Combining these stages allows the use of a more than one element to estimate age; this provides a more robust estimate of age at death for the seal assemblages. Although most of the bones can be used to determine age broadly, the femur has the most useful epiphyseal fusion sequence: at the yearling stage, the femur has no fused epiphyses; at the juvenile stage, the proximal epiphysis is fused; and, at the young adult (in this analysis the adult) stage, the distal epiphysis is fused (Storå 2001:222). Both the humerus and radius also have fusion patterns that can distinguish between yearling, juvenile and young adult/old adult (Storå 2001:222). In calculating the age, the fusion stages were either fused or unfused; a visible line of fusion was grouped with the fused elements (Storå 2002:52). Figure 56 shows the relative frequency of each element in each age group. The femur and humerus show similar patterns, while the radius is slightly different. When combined, they show a bimodal distribution with almost an equal portion of seals less than one year of age and those who have reached full maturity (Figure 57). It is also interesting to note that there is evidence
216
Figure 56: General subsistenc seal epiphyseal fusion Figure 57: General subsistence by element (%MNE) (MNE=108) seal epiphyseal fusion combined femur, humerus, and radius (MNE) (MNE=108) for a number of extremely young seals; there are a number of epiphyses with 100% juvenile cortex and indistinct morphological features as well as a number of vertebrae with unfused bodies and vertebral arches. This indicates that these seals were killed within three to six months of birth (Crockford and Frederick 2011; Storå 2001). This age data indicates that seals were being caught throughout the cold season, from late fall to late spring. Yearlings are present in the late spring of their first year, and may still be present in the fall of that year. Juveniles are typically found in open water, as they cannot compete for territory on the land fast ice near shore. Here, adults and very young seals will make up most of the spring population (Wynne 2008).
7.3.6.4.3 Other Marine Mammals
The other marine mammals are bearded seals and walrus. Although both are present in low frequencies based on NISP and MNI (Figure 55), they contribute around the same meat weight as artiodactyla species (Figure 53). Although meat weight is only a rough ordinal scale measure, it suggests that when these large marine mammals were caught, they were important sources of food, as were caribou, moose, and muskox. Both bearded seals and walrus were likely hunted opportunistically.
Bearded seals were hunted like spotted seals; they were hunted in open water or when hauled out on ice floes or land (Burch 2006:148–151). Bearded seal skins were an important source of raw material for boot soles, boat covers, rope, and storage covers (Burch 2006:153, 154). The bearded seal remains show that flippers were the most frequent element in this assemblage.
217
Walrus were hunted when they were swimming, hauled out on an ice floe, or hauled out on land (Burch 2006:165). All of the elements present in Feature 87 are axial elements (vertebrae and ribs) with a few pieces of unworked ivory. Both walrus and bearded seals were shared among the hunters, which may be the cause of the low and uneven distribution of skeletal remains (Burch 2006:153, 168–169).
7.3.6.4.4 Artiodactyla
Artiodactyla are the most numerous land mammal grouping in terms of NISP and meat weight, but second to furbearers for MNI (Figure 53 and Figure 55). Caribou (Rangifer tarandus), muskox (Ovibos moschatus), and moose (Alces alces) have all been identified in the artiodactyla group.
Caribou are an important species, even to coastal Thule groups, since they provide much of the raw material needed for hunting on the ice such as furs, sinew bone, and antler. Caribou could be taken communally at a drive system or individually through stalking (Burch 2006; 2012). In this region, individuals or small groups would travel into the interior during the winter to hunt caribou and bring the kill back to the winter settlement (Ray 1964; 1975a; 1983). As discussed above, the element distributions indicate that the remains found in the house may have been curated after removal of meat and sinew for marrow and bone grease processing; the positive correlation to the unsaturated marrow index and fracture freshness index support this interpretation. When present, the long bone ends are fused with no juvenile cortex, which suggests that most of the caribou hunted were adult. This may indicate a fall hunt, as juvenile caribou would be likely killed in the spring and summer (Burch 1972; 2006; 2012; Ray 1975a; 1983). Alternatively, this age distribution may indicate selective hunting all year round for adults.
Both musk ox and moose contribute very little to the assemblage in terms of NISP, MNI, or meat weight. Previous research has suggested that there were no musk ox on the Seward Peninsula, although the presence of horn and appendicular bones in this assemblage suggests otherwise. Moose are unlikely to have been encountered in the immediate area, preferring the trees along rivers and below the tree line. Like the presence of beaver and mink it looks like people were travelling south to the tree liner or northeast to the trees along one of the rivers that drains into Kotzebue Sound. Alternatively, they were trading for these forest-dwelling animals.
218
7.3.6.4.5 Dog and Wolf
Canis lupus specimens make up fewer than 3% of the NISP assemblage and fewer than 5% of assemblage by MNI; they make up around 1% of the meat weight, although they were likely only eaten in times of famine. Dog/wolf (Canis lupus) specimens have not been identified below species level, as this is difficult even with good comparative collections. Often the identification is done based on size, but wolves are a similar size to the dogs in the area and so no definitive identifications were made to the subspecies level. Most of the remains may be dog (Canis lupus familiaris) since dogs were kept in the Thule period, and wolves were not frequently found on the coastal tundra in this area (Burch 2006:176). Based on the age and even skeletal distribution, these remains likely represent a number of dogs that died, were skinned, and then were discarded as a unit. Both dogs and wolves were eaten only rarely; both were used mainly for their fur to trim clothing (Burch 2006:176, 283). Dogs were also used to help travel, hunt, and protect the settlement.
7.3.6.4.6 Other Terrestrial Mammals
7.3.6.4.6.1 Furbearers
A large number of different land mammals classified as furbearers have been identified at Feature 87; they make up 3% of the assemblage by NISP, 11% by MNI, but only 0.3% by meat weight (Table 11; Figure 53 and Figure 55). The assemblage includes elements from arctic ground squirrels (Spermophilus parryii), beavers (Castor canadensis), Alaskan hares (Lepus othus), hares (Lepus spp.), red foxes (Vulpes vulpes), arctic foxes (Vulpes lagopus), and minks (Neovison vison). The ground squirrel, hares, and foxes are all typically found on the coastal tundra; the mink and beaver are more commonly found in forested regains. Ethnographically most of the furbearers were caught in the warm season, although Alaskan hares and foxes could be caught year round. Other than the foxes described above, the assemblage only consisted of a few bones for each animal. For animals found locally, it is assumed that they were brought back whole to the site where they were skinned for pelts and some were eaten (Burch 2006:172–176).
Two of the furbearers are most often found in the forests of the interior or rivers; the presence of beaver and mink in the assemblage is somewhat surprising as the closest forests would have been at the tree line to the south or up one of the rivers that drain into Kotzebue Sound to the east. Like the moose discussed above, the beaver and mink could have been caught during seasonal
219 rounds, during winter excursions to the interior, or traded with other groups. However, the elements identified for the mink and beaver (and moose) are not only associated with pelts being traded (i.e. lower limb bones) or useful tools (i.e. beaver teeth for wood working). Mink mandibles and beaver clavicles would not have been part of a traded pelt; these elements would only be found in animals transported whole. This indicates that at least some of the people from Feature 87 were travelling into the forested regions of the Kotzebue Sound region to acquire resources at some point in the year.
7.3.6.4.6.2 Bears
Bears make up less than 1% of the assemblage by NISP, around 1% by MNI, and just fewer than 2.5% by meat weight (Table 11; Figure 53 and Figure 55). Only one bear rib was identified, indicating that either a polar or grizzly bear was hunted by people of the village and then potentially shared. There may also be a limited number of bear remains at the occupation site because, based on the documentary record, bears were processed at the kill site due to their large size (Burch 2006:170, 171). People of the interior most often hunted grizzly bears, while seal hunters encountered polar bears (Burch 2006:170, 171); however, the generic nature of the rib does not allow a species identification. In any case, killing a bear was never easy and almost always involved working in groups (Burch 2006:170, 171).
7.4 General Subsistence Discussion
Multiple lines of zooarchaeological evidence reveal four general trends. First, there is a distinct floor assemblage that underwent different accumulation, taphonomic, and excavation processes than the fill layer above it. Second, the excellent preservation of the floor material eliminates most natural formation processes as causes for the formation of the floor record. Third, some of the anatomical frequencies in species are the result of using bone as fuel. Finally, the overall subsistence of the site focused on small seals, but also relied on a multitude of other resources caught throughout the year, with a focus on winter resources.
The chapter started with a discussion of the material from one square, S3E4, and the similarities and differences in the history of these assemblages. Analyses of density-mediated attrition indicate that there is no significant correlation between survivorship and bone density for the fill, but that there is a positive, significant correlation for the floor. Based on modification
220 frequencies and archaeological context, taphonomic factors likely did not produce the small seal element profiles or bone density plots on the floor. Evidence for weathering, root etching, and gnawing is present, but in very low frequencies. This, along with the archaeological context that has preserved wood, fur, and other fragile faunal remains, indicates that the floor material was buried quickly when the roof collapsed. Excavation revealed that most of the floor material was within the permafrost, even in the summer months, and so mechanical and chemical processes were limited. The fill material had higher weathering frequencies, indicating that this assemblage was exposed to the elements for a greater length of time. Additional contexts such as disintegrating bones in upper levels, rodent burrows, and burnt layers, indicate that the fill was subject to more destructive processes than the floor in square S3E4. The high bird and furbearer frequencies point towards warm-season occupation for the fill assemblage, despite the fact that small seal remains dominate the fill assemblage. Small seals also dominate the floor assemblage, which affected the statistical tests of difference and led to no statistically significant differences between the fill and floor. However, other factors such as archaeological context, taphonomic indicators, fragmentation indices, and skeletal representation all highlight the differences between the fill and floor material.
The first three sections of general subsistence (archaeological context, density-mediated attrition, and peri-depositional factors) confirm the hypothesis that the floor was well preserved based on the one square fill/floor comparison. As with the single metre square, bone density correlates positively and significantly with small seal MAU, bird MAU, and caribou MAU. However, the modification frequencies and archaeological context again indicate that the floor material was quickly buried and frozen, which likely mitigated the mechanical and chemical taphonomic processes. The following sections (breakage and fragmentation, human modifications, and skeletal representation) examined if the destruction of less dense bones was caused by factors other than natural processes. Fragmentation is an important factor in the assemblage overall, and even more important when looking at individual species skeletal representation. Fragmentation is likely affected by burning, which is the most frequent human modification to the remains overall, even with the kitchen material removed. Cut marks also are important, especially for understanding the use of whales and canids.
Although small seal and caribou skeletal frequencies are quite different, they are both missing porous bones that are used to render bone grease and can be used as fuel for fires. The use of
221 these bones as fuel is supported by the high percentage of small cancellous mammal bones in the assemblage and high number of small burnt remains. Likely, people were bringing whole seals back to the site. There, seals underwent butchery to store and eat the meat, while axial bones may have been either fed to dogs or comminuted to feed the fire. Most caribou elements seem to also have been brought back to the site, with high marrow utility bones most frequent. Based on the fracture freshness index, bones were stored and then broken, likely to get marrow and bone grease or to be used in the fire.
Finally, an overview of the assemblage shows that even though whales may have significantly contributed to the diet in terms of meat weight, the evidence does not suggest that they were hunted regularly. However, the successful hunt of a large whale would drastically impact the community’s subsistence resources for the season. Cut marks and element frequencies make a strong case that whales were hunted, not scavenged. The bowhead whales, along with the other large marine mammals, were likely hunted when people encountered them rather than being the focus of a subsistence strategy. Additionally, whales may have been hunted at a distance from the site, and only select elements were brought back. Alternatively, small seals were hunted regularly; they contributed second most to meat weight and first to %NISP and %MNI, which indicates that they were the focus of the subsistence strategy at Feature 87 during the fourteenth century. Caribou contributed the most to diet of all the land mammals, with furbearers a close second. Birds, fish, and molluscs rounded out the remainder of the faunal assemblage. Age profiles for seals, caribou skeletal frequencies, and low frequencies of warm-season taxa all indicate that this house assemblage represents a cold season site, likely occupied from early fall to late spring. During this time, hunting parties seem to have travelled into the interior, potentially as far as the tree line, to hunt beavers, moose, and mink in the forests. However, the main focus of subsistence was on small seals, specifically ringed seals at breathing holes when the oceans froze. During the spring and fall, people caught migrating birds, whales, and other small seals with a variety of methods.
7.5 Summary
The division between the fill and floor material was analyzed in the first part of the chapter. Fill material is likely a palimpsest of multiple occupations in different seasons over hundreds of years. The floor material is also a palimpsest, although it only represents occupations in the cold
222 season, and likely only represents the last few months of occupation. The floor material was examined together, including material from the kitchen and midden. As described above, small seals were likely the focus of the resource strategy at Feature 87 in the winter, and were supplemented with other resources such as caribou, bearded seal, birds, and fish. There is evidence that whales were hunted at Cape Espenberg, but their contribution to the diet is difficult to assess. In some analyses, the kitchen material was removed as it distorted the burn modifications and fragmentation indices due to the high percentage of small, indeterminate, burnt remains. These types of spatial differences are analyzed in comparison to the expectations in the following chapter.
223
Chapter 8 Spatial Patterning of Faunal Remains: Testing Expectations
The Arctic documentary record has been used to interpret many different aspects of Thule lifeways. However, it has rarely been used to interpret behaviours or activities within houses, despite the fact that houses are often considered a basic unit for understanding human actions and behaviour. In order to use the documentary record as an analogy for Thule lifeways in the past, it needs to be tested against the archaeological record. Many documentary accounts in the Arctic were recorded based on the activities, behaviours, and beliefs that were structured by dwelling features. Understanding dwelling space, as described in Chapter 1, explores how people in the past were creating their lives on a small scale. The first goal of this chapter is to model a method to study space in dwellings. The second goal of this chapter is to strengthen the use of analogy in interpreting activities within dwellings by exploring differences and similarities between the expectations of the documentary record and the faunal assemblage.
This chapter draws on the context and discussion of the previous chapters to describe and interpret the archaeological results of documentary expectations. As mentioned above, Chapter 1 briefly described the history and use of spatial patterning in the archaeological study of houses, and specifically focused on how faunal material has been used to understand space. Chapter 1 also explored the theory behind using the direct-historical method to establish and test expectations. In order to construct spatial expectations against which to test a method for understanding the use of space within a house, Chapter 2 outlined the cultural history of the Thule, focusing on groups in the western Arctic (Alaska and Bering Strait region). The culture history and lifeways of the Thule Inuit has been generally established through interpretation of the archaeological record based on documentary sources, general hunter-gatherer theory, and cross-cultural analysis. These lifeways were described in Chapter 3, along with a typical history of an early Thule semi-subterranean house in northwest Alaska. Chapter 4 described the specific environmental, social, and archaeological context of the faunal material used to test the expectations, and a specific description of the house on which the expectations were based, Feature 87 at Cape Espenberg. In Chapter 5, the typical history of an Iñupiat house, the Feature 87 house itself, and the analogical and spatial theory were used to establish expectations based on a close reading of the documentary records of the northern Seward Peninsula and southern
224
Kotzebue Sound area. Before these expectations were tested, the methods used and the general subsistence patterns for the house were examined (Chapters 6 and 7). The present chapter provides the results of the expectations tests and interpretations. These results suggest ways in which documentary sources and taphonomic assumptions can be used to interpret archaeological behaviours, specifically those relating to spatial patterning within a house structure.
Expectations were outlined based on four stages of activities. The primary stage activities were based on the procurement, butchery, and transportation practices that introduced animal remains into a cold-season site. The faunal expectations derived from the primary activities were the same for all areas of the house and are found in the Stage 1 column for all expectations tables (Table 3, Table 4, Table 5, Table 6, and Table 7). These expectations were based solely on the historic, ethnohistoric, and ethnographic documentary sources from the area. The secondary activities were the main activities that occurred in each area of the house, identified in the documentary record (Table 2). In reality, most activities likely overlapped and occurred in multiple areas; however for this model, activities were identified in the areas where they most frequently occurred. The tunnel was associated with storage, the main floor with secondary butchery, the kitchen with preparation, the benches with consumption and manufacture, and the midden with discard. In the expectations tables, the faunal signatures resulting from these secondary activities are recorded in the Stage 2 columns. The tertiary activities are activities that occurred during the occupation of the house, but that are not well described in the documentary sources, and that may have had drastic effects on the faunal assemblage. These include processes like carnivore gnawing, trampling, cleaning, and secondary burning; expected faunal remains were recorded in the Stage 3 columns. Quaternary activities occurred after the house was abandoned; these include carnivore activity, weathering, and human disturbances. The results of these expectations are recorded in the Stage 4 columns. The final expectation columns record the faunal signatures that are predicted to result from the cumulative impacts of all four stages of activities; these final expectations are tested in this chapter. A total of 72 expectations were outlined in Chapter 5, Table 3, Table 4, Table 5, Table 6, and Table 7. The specific results of each expectation are detailed below. This chapter discusses the most important results for each area and concludes with a brief summary.
225
8.1 Testing Expectations
Expectations were tested using the specimens from the floor that were collected in quadrants (Appendix B, Table B.1). The material was analyzed and visualized in ArcGIS. ArcGIS provided an easy, flexible database that generated all of the results by area based on the expected frequencies and ratios for each area; this type of spatially focused analysis was much more easily managed in a spatial analysis program than in a traditional database. Furthermore, these data can be used in future for a more exploratory analysis rather than one based on pre-defined expectations.
Expectations were tested through the use of indices. Indices are either ratios or frequencies. Ratios compare specific elements to other specific elements; frequencies compare specific elements to the total number of bones in that assemblage for the area. For example, a ratio compares large terrestrial mammal high-utility to low-utility elements; a frequency calculates the proportion of lower limbs in an area (see Chapter 6, section 6.4.7).
Those expectations that can be compared to a normal skeletal measure are called skeletal indices; ratios and frequencies that cannot be compared to a normal skeletal measure are called comparative indices. Normal skeletal measures are used for expectations dealing with known numbers of elements based on standard anatomy (e.g. bird wings and legs, small seal high-utility elements, etc.). These skeletal indices can be compared to normal skeletal indices for taxa, since both sets of elements can be estimated in a typical skeleton for that taxon. For example, there are 74 high-food-utility caribou elements (the most numerous large terrestrial mammal species present in the taxon category) and 45 low-food-utility elements; the normal skeletal ratio is 74:45, or 1.64. The ratios from the archaeological results for high- to low-utility elements are then compared to this normal skeletal measure.
Some indices cannot be compared to a skeletal norm and are only compared among areas; these are called comparative indices (e.g. ratio of small to large small seal specimens or the ratio of fish cranial to post-cranial bones). Normal skeletal measures could not be calculated for taxa with highly variable elements or when there is no expected normal measure (e.g. fish, modification frequencies, etc.). The final columns in Table 19, Table 20, Table 21, Table 22, and Table 23 indicate into which category expectations fall.
226
The expectations have been divided into these two categories to clarify how expectations were tested and confirmed. For skeletal indices there are normal skeletal measures against which to confirm or contradict an expectation. These normal skeletal measures are calculated based on the normal average number of bones in an element set or entire skeleton. Therefore, these skeletal expectations can be confirmed absolutely in comparison with a normal measure. This is referred to as “Absolute” under the confirmation heading in Tables 19-23. However, skeletal expectations can also be tested by comparing the results among the five areas. This is referred to as “Relative” under the confirmation heading in Tables 19-23. Comparative expectations cannot be tested against a normal measure, and so cannot be tested absolutely. The results of these comparative expectations can only be tested against the other house areas by comparing them to the house average index and ranking them with other areas.
Since all the expectations can be relatively tested, these results are more useful for developing area signatures, as discussed in Chapter 9. This methodology of comparing results across the house may allow this research to be broadened to other areas of the Arctic during Thule periods and potentially to other regions and time periods as well.
The results outlined below are based on the expectations from Chapter 5, which are summarized by area in Table 3, Table 4, Table 5, Table 6, and Table 7. These tables are not repeated in full in the results tables (Tables 19-23). The results table for each house area has nine columns. The first two columns are ‘Taxonomic Category’ and ‘Final Expectations’. The ‘Taxonomic Category’ column divides the animal into simplified taxa as described in Chapter 5. The ‘Final Expectations’ columns give the final expectations from the last columns in Chapter 5, Table 3, Table 4, Table 5, Table 6, and Table 7. These expectations are the expected faunal indices for each area based on the cumulative impacts of primary, secondary, tertiary, and quaternary activities. Even if a taxonomic expectation is not expected to be present in that area according to the final expectations, it is stated and tested. The ‘Results’ column gives the result of the index calculation based on the archaeological material. The calculations for each result are found in Appendix B. The next column, ‘Normal Skeletal Measure’, shows the normal skeletal measure for expectations classified as skeletal. This measure is given when the expectation can be tested against a known normal value based on the known number of skeletal parts for the taxa. The elements used in these calculations are found in Appendix B. The ‘Average House Results’ column gives the average result for the house. The raw numbers and calculations for each
227 expectation are given in Appendix B. ‘Relative Ranking’ details the rank for that expectation in relation to the other areas of the house; there are five areas in total and so ranks fall between one and five. There are two columns that state if the final expectations were met or not: ‘Absolute’ and ‘Relative’. The ‘Absolute Confirmation’ column compares the results to the normal skeletal measure, while the ‘Relative Confirmation’ compares the results to the average house results and the relative ranking. The final column, ‘Expectation Type’, identifies the expectation type as either skeletal or comparative as described above.
Primary expectations were outlined based on the faunal remains that were expected to enter the site. In the tables in Chapter 5, these are listed in the first column and represent expectations A to J in all of the areas. As described in detail in Chapter 5, some of these original faunal expectations were not expected to contribute to the assemblage at later stages, however they are tested anyway. If whole taxa are not expected to be present in an area (e.g. large marine mammals are not expected in the kitchen), the overall taxonomic frequency for that area is tested. If certain skeletal elements are not expected (e.g. large terrestrial mammal lower limbs in the kitchen), those indices are tested. This allows the analysis to move beyond testing just the presence or absence of faunal material, but to explore to what degree the expectations were confirmed or denied. Thus, even faunal indices that are not expected to be present in an area are tested against the normal skeletal measure (if applicable) and the average house results; the result is given a relative ranking and the type of expectation is noted. Further discussion of the results and comparisons are given in the text if necessary.
Just as some primary expectations are not expected to be present in some areas, there are secondary, tertiary, and quaternary expectations that are added to the list of expectations for an area. These are all given letters that fall after J; however, if these expectations relate to specific taxa (e.g. low frequencies of fish in the tunnel, Expectation N), they are placed out of alphabetical order in the correct taxonomic category (e.g. Expectation N in the tunnel is placed with the fish, under Expectation I).
The following sections of the chapter describe the overall and area based results of the expectations for the house. In both the overall and area-specific descriptions and interpretations, expectations are calculated in two different ways in order to measure the degree to which expectations are confirmed. Two different calculations are needed since skeletal expectations are
228 tested twice; these expectations may only be confirmed based on one test (either absolute or relative) and not the other. The first calculation, “General Confirmation”, is based on confirmation of either the absolute or relative test for skeletal expectations and confirmation of the relative test for comparative expectations. Therefore, if a skeletal expectation has only the absolute test but not the relative test confirmed, the entire expectation is considered confirmed; the skeletal expectation would also be confirmed if the relative test was confirmed but the absolute test was not confirmed. The General Confirmation test corresponds to the number of expectations outlined in Chapter 5 (n=72). The second calculation, “Discrete Confirmation”, treats each absolute and relative test for skeletal expectations as a separate test, which increases the number of tests in each area (n=101). Table 12 gives the numbers and percentage confirmed for both calculations.
8.2 General Results
Overall, the expectations were confirmed for 56% of the total expectations (General Confirmation); however, when skeletal expectations were separated, the expectations were only 51% confirmed (Discrete Confirmation) (Table 12). The general confirmation is fairly low; however, as described below, some of the areas, such as the kitchen and midden, have a greater percentage of confirmed expectations than others. General Confirmation (either Discrete Confirmation of Absolute and absolutely or relatively) Relative Tests Area Total Confirmed % Confirmed Total Confirmed % Confirmed Tunnel 14 9 64.29 21 12 57.14 Main Room Floor 13 6 46.15 19 7 36.84 Kitchen 13 10 76.92 18 13 72.22 Benches 17 7 41.18 23 10 43.48 Midden 15 8 53.33 20 10 50.00 Total 72 40 55.56 101 52 51.49 Table 12: Exepctations confirmation, total
Different types of test indices and different types of expectations have different proportions of confirmation for Feature 87. Skeletal expectations were confirmed 53% of the time, with comparative expectations confirmed slightly less, at 49% (Table 13). When the indices are broken down into ratios and frequencies, frequency expectations were confirmed 56% of the time, while ratios were only confirmed 42% of the time (Table 14). This suggests that skeletal measures may be more useful for interpreting faunal specimens than comparative measures.
229
Skeletal Expectations (Absolute & Comparative Expectations Relative Tests) (Relative Tests) Area Total Confirmed % Confirmed Total Confirmed % Confirmed Tunnel 14 7 50.00 7 5 71.43 Main Room Floor 12 5 41.67 7 2 28.57 Kitchen 10 7 70.00 8 6 75.00 Benches 12 6 50.00 11 4 36.36 Midden 10 6 60.00 11 6 54.55 Total 58 31 53.45 43 21 48.84 Table 13: Expectations confirmation, skeletal and comparative
Ratios Frequencies Area Total Confirmed % Confirmed Total Confirmed % Confirmed Tunnel 10 5 50.00 11 7 63.64 Main Room Floor 4 1 25.00 15 6 40.00 Kitchen 4 2 50.00 14 11 78.57 Benches 5 2 40.00 18 8 44.44 Midden 8 3 37.50 12 7 58.33 Total 58 13 41.94 70 39 55.71 Table 14: Expectations confirmation, ratios and frequencies
Absolute Relative Area Total Confirmed % Confirmed Total Confirmed % Confirmed Tunnel 7 3 42.86 14 9 64.29 Main Room Floor 6 4 66.67 13 4 30.77 Benches 6 3 50.00 17 7 41.18 Kitchen 5 3 60.00 13 11 84.61 Midden 5 3 60.00 15 7 46.67 Total 29 16 55.17 72 38 52.78 Table 15: Expectations confirmation, absolute and relative tests
However, each area is different, with skeletal and comparative indices variably confirmed in different areas.
The following sections describe and interpret the results for each area. In addition to the results tables for each area, Table 16, Table 17, and Table 18 show the overall results that will be used to interpret the expectations in the areas. Table 16 shows the proportions of each taxon by area and their relative rank. Table 17 shows the modification proportions by area, while Table 18 shows the indeterminate and size proportions by area; both also indicate the relative rank by area. Further tables with the raw numbers and calculations of ratios and frequencies are found in Appendix B (Tables B.1-B.7).
230
Large Large Small Marine Terrestrial Terrestrial Area Mammals Rank Small Seals Rank Mammals Rank Mammals Rank Birds Rank Fish Rank Tunnel 0.55 4 89.57 2 3.00 4 2.00 3 3.77 1 1.11 4 Main Room 1.61 3 79.91 3 3.67 1 11.00 1 1.83 3 1.98 3 Kitchen 0 5 78.50 5 0.48 5 0.32 5 0.32 5 20.38 1 Benches 2.63 1 78.83 4 3.65 2 5.64 2 2.14 2 7.11 2 Midden 2.14 2 90.91 1 3.74 3 1.60 4 1.60 4 0 5 Total 1.76 81.13 3.21 5.63 2.11 6.17 Table 16: Taxonomic proportions by area (%NISP)
Butchery Gnaw Area Marks Rank Marks Rank Burning Rank Weathering Rank Tunnel 0.32 3 1.51 4 24.75 2 2.83 1 Main Room 0.98 1 1.79 2 12.72 4 1.32 2 Kitchen 0.01 5 0.01 5 92.40 1 0.02 5 Benches 0.77 2 1.85 1 22.06 3 0.87 3 Midden 0.28 4 1.68 3 11.76 5 0.84 4 Total 0.30 0.76 61.26 0.63 Table 17: Modification proportions by area (%NISP)
Mammal All Area Indeterminate Rank Indeterminate Rank Indeterminate Rank Small Rank Large Rank Tunnel 21.69 2 27.81 2 49.50 2 89.36 3 10.59 3 Main Room 10.98 4 26.51 3 37.49 4 81.45 4 18.51 2 Kitchen 81.03 1 13.36 5 94.39 1 99.82 1 0.18 5 Benches 8.36 5 17.85 4 26.21 5 81.23 5 18.62 1 Midden 11.76 3 35.01 1 46.78 3 89.64 2 10.36 4 Total 53.86 17.46 71.31 92.78 7.18 Table 18: Indeterminate and size proportions by area (%NISP) 8.3 Tunnel Results
The tunnel had 14 expectations tested, with nine of them generally confirmed either absolutely or relatively (Table 12). When divided into discrete tests, there were 21 total expectations, with 12 of them confirmed (Table 12). This suggests that the documentary and taphonomic models for the tunnels somewhat reflect the archaeological reality, but that other factors are also influencing the distribution of bones. The description of the expectations below do not discuss some of the expectations proposed to be intermediate or average, such as the intermediate expected frequency of indeterminate bones (Expectation L), the average expected overall number of bones (Expectation K), and the average frequency of gnaw marks (Expectation M). These intermediate indices have a variety of ranking relative to other areas, from second to fourth, with little distinguishing them from the other areas with similar results. Although all other expectation
231 results are discussed, as stated in Chapter 7: Methodology, an arbitrary threshold of 50 specimens is used to establish meaningful results; taxa with specimens below this threshold give weaker results than taxa with large sample sizes. For the tunnel, all the taxa except small seals and the general expectations are below 50 specimens and need further investigation in other houses before explicit interpretations can be established.
Overall, faunal material was expected to show storage practices and an intermediate effect of density-mediated attrition activities in the tunnel.
8.3.1 Expectations Relating to Storage Activities
The tunnel was expected to have a number of faunal signatures that are related to storage practices. Large marine mammal appendicular elements (Expectation A), large terrestrial mammal high-utility and lower limb elements (Expectations E and F), and dense bird elements (Expectation H) are the only final expectations that show evidence of these storage behaviours. Three of the four expectations were confirmed through both absolute and relative testing. The first is the high ratio of appendicular to axial elements for large marine mammals (Expectation A) (Table 19). It was expected that storage of large marine mammal limbs would occur in the tunnel where they were placed after the butchery of these mammals away from the site. Despite the absolute and relative confirmation for the high ratio of large marine mammal appendicular to axial elements, there were only five large marine mammal specimens in the tunnel, the second lowest of any area. Even though the expectations were met, they are likely not meaningful due to sample size.
Lower limbs of large terrestrial mammals such as caribou were preferentially transported back to the site and stored for further processing and consumption. The high frequency both absolutely and relatively of lower limbs in the tunnel suggests that this may have been the case in Feature 87 (Expectation F). This interpretation is hampered by the small sample size (n=27).
The final confirmed expectation relating to storage was the high ratio of dense to less dense bird elements (Expectation G). Whole bird bones were expected to be stored in the tunnel prior to preparation for consumption and use in manufacture; however, they were expected to be
232
Normal Average Confirmation Taxonomic Skeletal House Relative Type of Category Final Expectations Results Measure Results Ranking Absolute Relative Expectation Large A High ratio of appendicular : 3.00 1.62 1.89 1 Yes Yes Skeletal marine axial elements mammal B Bacula, whale ribs, and 0.20 N/A 0.18 3 N/A No Comparative (n=5) mandibles not expected to be present in the tunnel Small seal C High ratio of dense : less dense 1.06 1.30 1.18 3 No No Skeletal (n=807) elements D Low frequencies of hind 0.29 0.27 0.30 2 No No Skeletal flippers Large E Even ratio of high-utility : low- 0.75 1.64 1.10 5 No No Skeletal terrestrial utility elements mammal F High frequency of lower limbs 0.48 0.42 0.35 1 Yes Yes Skeletal (n=27) Small G High ratio of dense : less dense 1.83 0.57 1.19 2 Yes Yes Skeletal terrestrial elements mammal (n=18) Bird H High ratio of dense : less dense 0.90 1.09 1.07 1 No Yes Skeletal (n=34) elements Fish I Intermediate frequency of 0.4 N/A 0.48 2 N/A Yes Comparative (n=10) vertebrae N Low frequency of fish remains 0.005 N/A 0.02 4 N/A Yes Comparative General J Low frequency of butchery 0.003 N/A 0.003 3 N/A No Comparative (n=2190) marks K Intermediate total number of 0.10 N/A 0.2 4 N/A Yes Comparative specimens (NSP) L Intermediate frequency of 0.22 N/A 0.54 2 N/A Yes Comparative unidentifiable bones M Intermediate frequency of 0.02 N/A 0.01 4 N/A Yes Comparative carnivore marks Table 19: Tunnel expectation results. Expectations are placed in taxonomic categories regardless of their alphabetical label. See Chapter 5 for a detailed discussion of the construction of the expectations.
233 impacted by tertiary and quaternary stage activities. These activities were expected to significantly destroy less dense bone, leaving only dense bird specimens as indicators of storage.
Although the bird expectation is not met when compared to the skeletal measure, with fewer dense elements present than in a normal skeleton, the ratio is met relatively; the ratio of dense to less dense bird elements is highest in the tunnel relative to other house areas. This suggests that birds, when stored in the tunnel, were undergoing density-mediated destruction at higher rates than other areas.
The only expectation not met for storage was the even ratio of high-utility to low-utility elements of large terrestrial mammals (Expectation E). It was expected that the tunnel would have average ratios of high-utility elements to low-utility elements. Although high-utility elements would be stored in the tunnel, they would also be attached to low-utility elements such as phalanges. In the main room, the low-utility elements would be removed and discarded, while meat and associated high-utility elements were moved into the kitchen for preparation. Thus, the main room would have lower ratios of high- to low-utility elements, while the kitchen would have higher ratios of high- to low-utility elements. The tunnel, by comparison, would have even ratios. This is not the case, with neither the absolute nor relative test being confirmed; however, as mentioned above, the small sample size inhibits strong interpretation of the results. The average skeletal ratio is 1.64, while the ratio for the tunnel is only 0.75. This is the lowest ratio in the house, indicating that the tunnel has the lowest ratio of high- to low-utility elements. Many factors may have influenced this ratio, including the fact that high-utility elements may not have been typically stored in the tunnel. Alternatively, high-utility elements may have been destroyed by tertiary and quaternary activities.
Overall, storage does seem to be well supported by the faunal remains in the tunnel, despite the small sample size. Large marine mammal appendicular elements and large terrestrial mammal lower limbs are found in expected frequencies that relate directly to storage. Large terrestrial mammal high-utility elements were not found in the tunnel in their expected ratios, but may have been more significantly affected by density-mediated attrition than predicted by the model. Birds also support the expected storage indices in the tunnel, with the highest ratios of dense elements in the house and the overall highest taxonomic frequencies of any area (Table 16). These bird
234 ratios also support the tertiary and quaternary activities that were expected to destroy many of the less dense faunal elements.
8.3.2 Expectations Relating to Density-Mediated Attrition Activities
As described above, the expectation of dense bird bones was relatively confirmed, which suggests that at least some of the activities related to density-mediated attrition were impacting the tunnel assemblage as expected. Tertiary and quaternary activities were expected to both remove less dense material and render many fragments unidentifiable. The tunnel was expected to have medium rates of density-mediated destruction over both tertiary and quaternary stage activities. Although the introduction of bones into the tunnel through cleaning activities may increase the amount of less-dense material, subsequent activities were expected to leave only the dense elements. Together, cleaning and density-mediated attrition were expected to eliminate all but the densest bones.
There are four expectations that were expected to show medium amounts of density-mediated destruction: high ratio of dense to less dense elements for small seals (Expectation C), small terrestrial mammals (Expectation G), and birds (Expectation H), and intermediate frequencies of fish vertebrae (Expectation I). Three expectations were confirmed and may represent density- mediated destruction: small terrestrial mammal dense to less dense elements (Expectation C), intermediate frequency of fish vertebrae (Expectation I), and dense to less dense bird elements (Expectation H). Although some of these expectations were confirmed, density-mediated attrition cannot account for all of the faunal signatures in the tunnel.
Small terrestrial mammals show that mostly dense elements were left in the tunnel for this taxon, which confirms the expectation both relatively and absolutely (Expectation G). It was expected that small mammals would be stored in the tunnel, but then the less dense elements would be destroyed through tertiary and quaternary activities. This appears to be the case, although again, the small sample size limits the usefulness of this skeletal ratio (n=18).
The second confirmed expectation of density-mediated destruction is the intermediate frequency of fish vertebrae (Expectation I). Relative to other areas, there are intermediate frequencies of fish vertebrae in the tunnel, which confirms Expectation I. Although the frequency is slightly lower than the house average, the frequency of vertebrae in the tunnel is the second highest in the
235 house, indicating that density-mediated destruction was potentially destroying fish specimens at a higher degree in the tunnel than most other areas. As with the small terrestrial mammals, the small sample size (n=10) hampers the interpretive ability of this comparison.
The third confirmed expectation is the density ratios of bird bones, described above (Expectation H). Although the bird dense to less dense ratio is low in comparison to the normal skeletal measure, it is the highest in the house. This suggests that less dense bird elements were being destroyed in greater frequencies in the tunnel than other areas of the house. However, in general, the low ratio of dense to less dense elements indicates that the household deposits of bird bones were not being subjected to high degrees of density-mediated destruction as most of the specimens in all areas are actually low density elements.
The only expectation not met was the ratio of dense to less dense elements for small seals (Expectation C). This ratio is slightly lower than the house average, and ranks the tunnel third in comparison to all other house areas. This ratio shows that although there were more dense than less dense small seal elements in the tunnel, this is not significantly higher than other areas. This average ranking indicates that there may have been intermediate frequencies of destruction, which follows the general expectation for the tunnel of intermediate frequencies of density- mediated destruction.
8.3.3 Summary of Tunnel Expectations
There is some support, based on the confirmation of three out of four expectations that the tunnel may retain evidence of storage behaviours. However, most of the evidence of storage activities was likely altered by contemporary and post-depositional activities. Some of the tertiary and quaternary activities likely resulted in density-mediated destruction, although in lower frequencies than predicted. However, the ratio of high- to low-density bird and small seal elements and the presence of fragile fish elements suggest that density-mediated activities such as trampling and carnivore action did not completely overwrite the secondary expectations for storage in the tunnel. These variable results in density-mediated destruction support the impact of intermediate frequencies of density-mediated destruction that was expected to occur in the tunnel.
236
8.4 Main Room Floor Results
The main room floor had the lowest percentage of discrete test expectations confirmed at 37% (19 total discrete expectations tests, only 7 confirmed) and the second lowest combined test results at 46% (13 total, 6 confirmed) (Table 12). This low level of confirmation is not affected by sample size on the main room floor more than other areas; specimens adequately represent all taxa except large marine mammals, bird, and fish. Although expectations for these taxa will be discussed, their interpretations are not as strongly supported as the other expectation results. Again, as with the tunnel, expectations that were projected to result in comparatively intermediate or average results are difficult to assess (Expectations J, L, and N). These intermediate or average indices have variable relative ranking, from second to fourth, with little distinguishing them from the other areas with similar results. This variable ranking widens the range of possibilities for confirmation, which can lead to less robust interpretations. Only expectations at either the high or low extreme are discussed. Small sample sizes of large marine mammals, birds, and fish, limit the interpretive ability of results based on these taxa.
Secondary butchery practices, tertiary cleaning activities, and tertiary and quaternary trampling activities are expected to have the biggest influence on the assemblage. However, as seen by the low overall confirmation and further discussed below, none of these expectations is strongly supported.
8.4.1 Expectations Relating to Butchery Activities
The most typical secondary activity on the floor was expected to be butchery of animals brought into the house. Five expectations were proposed that would show the butchery practices, even after being affected by tertiary and quaternary activities. All five expectations were related to the processing of animal bones prior to their consumption or preparation. In this case, the documentary record indicates that low utility bones would be removed and discarded. The high- utility bones were expected to be moved into the kitchen for preparation or onto the benches for consumption. For many animals, heads, distal elements, and extremities are considered low- utility elements. Small seals (Expectation C), large terrestrial mammals (Expectation E and F), small terrestrial mammals (Expectation G) and birds (Expectation H) were predicted to show evidence of butchery practices.
237
Normal Confirmation Taxonomic Skeletal Average Relative Type of Category Final Expectations Results Measure Results Ranking Absolute Relative Expectation Large A Low frequency of appendicular 0.27 0.62 0.52 5 Yes Yes Skeletal marine elements mammal B Bacula, whale ribs and mandibles 0.41 N/A 0.18 1 N/A No Comparative (n=22) not expected to be present on main room floor Small seal C High frequency of heads and hind 0.29 0.28 0.31 3 Yes No Skeletal (n=1090) flippers D Overwritten by Expectation C N/A N/A N/A N/A N/A N/A N/A (not tested) Large E Low ratio of high-utility : small, 1.86 2.06 1.65 3 Yes No Skeletal terrestrial low-utility elements mammal F High ratio of phalanges : whole 0.24 0.40 0.29 4 No No Skeletal (n=50) limbs Small G Slightly high frequency of 0 0.006 0.01 5 No No Skeletal terrestrial mandibles mammal (n=150) Bird H High frequency of lower limbs 0.21 0.27 0.17 2 No Yes Skeletal (n=25) Fish I Intermediate frequency of 0.22 N/A 0.36 4 N/A No Comparative (n=27) vertebrae L Low frequency of fish 0.01 N/A 0.02 2 N/A No Comparative General J Intermediate frequency of 0.01 N/A 0.003 1 N/A No Comparative (n=2350) butchery marks K Intermediate total number of 0.10 N/A 0.2 3 N/A Yes Comparative specimens (NSP) M High frequency of unidentifiable 0.11 N/A 0.54 4 N/A No Comparative bones N Intermediate frequency of 0.02 N/A 0.01 2 N/A Yes Comparative carnivore marks Table 20: Main room floor expectation results. Expectations are placed in taxonomic categories, regardless of their alphabetical label. See Chapter 5 for a detailed discussion of the construction of the expectations.
238
Small seals, if brought back to the site whole, were expected to have their heads and hind limbs removed and discarded on the main room floor, which would result in a high frequency of these elements on the main room floor (Expectation C). This expectation was confirmed based on absolute measures, but not confirmed based on relative measures. This result was only slightly higher than the normal skeletal frequency, and was lower than both the house average and the relative expectation (Table 20). The lower than expected relative result is likely influenced by the surprising number of hind flippers present in the house as a whole. As described in Chapter 7: General Subsistence, hind flipper elements are the most abundant in the house, suggesting that they were not removed prior to entering the site. Thus, their increased presence in the house would potentially alter the frequency of hind flipper elements in other areas of the house. If other areas in the house had higher frequencies of hind flipper specimens (e.g. for consumption on the benches or storage in the tunnel), the frequency of hind flipper proportions on the main room floor might be relatively lower than expected. The lack of data on hind flippers in the documentary record makes it difficult to determine how or where these elements would be discarded. The high frequency of small seal hind flipper elements on the benches (0.33) may indicate that these elements were being treated like large marine mammal flippers, with high discard rates on the bench and intermediate frequencies in the tunnel. However, further investigation of this pattern is needed in other Thule dwellings.
Large terrestrial mammals were also expected to have a low ratio of high- to low-utility elements (Expectation E) and a high ratio of phalanges to whole limbs (Expectation F). Although only high-utility packages were expected to be brought back to the site for large terrestrial mammals, these packages often had lower utility elements attached to them, such as metapodials and phalanges. These are expected to be removed and discarded on the main room floor, with high- utility elements being moved into the kitchen for more preparation. Neither of these expectations is met absolutely or relatively. Although the ratio of high- to low-utility elements is lower than the normal skeletal measure in comparison to the rest of the house, it is overall about average (Expectation E). Neither of the phalanx expectations is met, with a low ratio of phalanges to whole limbs when compared to both the normal skeletal measure and the rest of the house.
Small terrestrial mammals were also expected to have their heads removed in the main room prior to being moved to the kitchen for preparation or the benches for use in manufacturing. However, because there was also expected to be high degrees of trampling and cleaning on the
239 main room floor, only the densest cranial elements were expected in slightly higher frequencies: the mandibles (Expectation G). This expectation was not met either in comparison to a normal skeletal measure or in comparison to the rest of the house. There were no mandibles recovered on the main room floor. For the whole house, there were only four mandibles in total, all on the benches. The lack of mandibles in the house may indicate that heads were removed outside. However, more evidence would be needed to support this interpretation.
Birds, which were also expected to enter the house whole and have their low-utility parts removed on the main room floor, were expected to show evidence of butchery practices with high frequencies of lower limbs (Expectation H). These elements do not have meat associated with them and are not as useful in manufacture as other elements. Although this index was not confirmed in absolute terms, as it was slightly lower than the normal skeletal ratio, it was somewhat confirmed in relative terms, with the second highest frequency of lower limbs in the house. However, the small sample size (n=25) limits the usefulness of this interpretation.
Taken together, none of the five taxonomic expectations indicate that there is a distinct patterning of butchery on the main room floor. Butchery marks may give some support to the fact that butchery may have taken place in the main room. Butchery marks such as impact scars and cut marks were expected to be present in intermediate frequencies (Expectation J), with both the kitchen and benches expected to have higher frequencies of butchery marks due to further preparation, consumption, and use in manufacture. However, this is not the case, as butchery marks are in the highest frequencies in the main room. Despite this unexpected result, the high frequency of butchery marks (bone scars and cut marks) may be one of the only strong indications that butchery took place on the main room floor.
8.4.2 Expectations Relating to Cleaning Activities
Cleaning activities during the tertiary stage were expected to deposit bones on the floor from the benches, while removing large specimens from the floor. Most of the deposited bones also would be affected by other tertiary and quaternary activities, such as trampling and dog gnawing described below. However, some faunal specimens were expected to represent cleaning activities. Four expectations were projected to show cleaning activities: low frequency of large marine mammal appendicular elements (Expectation A), intermediate frequency of fish vertebrae (Expectation I), low frequency of fish (Expectation L), and high numbers of unidentifiable
240 specimens (Expectation M). All four expectations were expected to be added to the floor in greater quantities during tertiary cleaning than would normally have been present during secondary butchery activities. Only one of the expectations was confirmed.
Large marine mammal appendicular elements are in lowest frequency on the main room floor (Expectation A). This was confirmed both absolutely and relatively, with a frequency of 0.27, compared to a normal skeletal frequency of 0.62; this frequency was also the relatively lowest frequency (rank=5). However, this result is based on a small sample size (n=22), which limits the interpretive capability of this result. This suggests that large marine mammal flipper elements were present on the main floor, although there is no evidence to suggest that they were added during cleaning rather than during secondary activities.
Fish, which were not expected on the floor during butchery practices, were expected to be added to the floor during cleaning and then destroyed by density-mediated activities. Only intermediate frequencies of vertebrae were expected (Expectation I), with low overall taxonomic frequencies of fish (Expectation J). Neither expectation was confirmed, with the second lowest frequencies of vertebra and the second highest overall taxonomic frequency in the house. This relatively high frequency of fish specimens indicates that fish may have added to the floor more frequently in cleaning activities, or that density-mediated activities did not affect the fish as much as expected, which is discussed in the following section. Alternatively, the main room floor may have been used for other purposes not recorded in the documentary record.
The other expectation that was not confirmed was the high number of unidentifiable bones (Expectation K). It was expected that large, identifiable bones would be removed, leaving relatively high frequencies of unidentifiable bones. This is not supported by the archaeology, with the main room floor having the second lowest frequency of indeterminate specimens in the house (Table 20). The low expectations related to cleaning are affected by density-mediated attrition activities, described below. A summary of interpretations of expectation results considers both the cleaning and density-mediated activities together, as they are difficult to disentangle.
241
8.4.3 Expectations Relating to Density-Mediated Attrition Activities
Density-mediated attrition from trampling and carnivore destruction in both tertiary and quaternary stage activities was expected to affect the main room floor assemblage. The wood planked floor and high-traffic area of the floor was expected to result in high degrees of destruction through trampling during the house occupation. Also, carnivore activity was expected in intermediate frequencies in the main room after the house was abandoned as carnivores scavenged bones within the house. Three expectations were established to test these assumptions: small terrestrial mammal mandible frequencies (Expectation G), fish vertebrae frequencies (Expectation I), and overall fish frequencies (Expectation L). Most of these overlap with the expected butchery activity results and cleaning activity results described above. None of the expectations were confirmed.
Neither small terrestrial mammal (Expectation G) nor fish vertebrae expectations (Expectation I) support the expectation that this assemblage was heavily impacted by density-mediated attrition activities. Small terrestrial mammal heads were expected to be left on the main room floor. Density-mediated destruction through high-traffic trampling was expected to leave only mandibles (Expectation G). The results show that there were no mandibles on the main room floor, despite a high sample size of 150 for small terrestrial mammals. Fish, expected to be cleaned onto the floor and then all but their vertebrae destroyed, have frequencies that indicate that most of their elements were not the dense vertebrae (Expectation I), but rather fragile cranial elements. Additionally, there is a high frequency of fish specimens overall on the main room floor, which was not expected in a high-traffic area (Expectation L). Fish results show that less dense elements survived more frequently than dense elements.
8.4.4 Summary of Main Room Floor Expectations
Very few of the expectations derived from the documentary record were supported for the main room floor. None of the five expectations for butchery were strongly supported; the only support for butchery on the main room floor was the slightly high frequency of butchery marks, which was not actually an expectation for butchery practices compared to the rest of the house. Similarly, very few of the expectations related to cleaning activities were supported. Only the low frequency of large marine mammal appendicular elements supported the expectation that cleaning would move some elements from the benches to the main room floor; these low
242 frequencies may have also resulted from other activities such as trampling, general movement, or a few instances of discard of flippers on the floor. Although fish are present on the floor, which supports the fact that some cleaning may have deposited them there, fish frequencies are much higher than expected. These fish frequencies, along with the absence of small terrestrial mammal mandibles, suggests that activities affecting bone density, such as trampling and carnivore gnawing, did not affect the floor assemblage as significantly as expected.
The low confirmation of expectation indices suggests that the archaeological understanding of the main room floor, based on documentary and taphonomic assumptions, is not accurate. This lack of correlation between the documentary and archaeological records may be a result of many factors. One of the main factors may be the mixing of floor material with bench material during house occupation. Benches, as described in the documentary record, were cleaned, which would move much of the bench material onto the floor. This would obscure many taxonomic and taphonomic patterns. One of the expectations that were not met was the high frequency of indeterminate bone. It was expected that there would be greater density-mediated destruction on the floor due to trampling than the faunal record indicates, reducing bones into small bits and increasing the number of unidentifiable specimens. This is not the case for the main room floor in Feature 87. The structure of the floor of Feature 87 may play a part of that; in other houses, the spaces between and under the floorboards are often repositories of small, identifiable bones. In Feature 87, the floorboards were exceptionally well made, with very little space between them. Additionally, they sat on mostly sterile sand, with very little material under the floorboards themselves. These tight-fitting, well-made planks likely prevented much material, even very small bones, from getting under the floor. General movement may have pushed much of this material to the edges of the floor, which would have ended up under the benches or in the tunnel entrance. Alternatively, cleaning such well-made floors may have allowed for the removal of material into the tunnel or midden rather than the accumulation of material.
Another potential explanation for the lack of confirmation of the expectations is that the expectations are a drastic simplification of the actual activities that took place on the main room floor. As can be seen in Table 2, there were many activities that took place on the main room floor based on the documentary record; by selecting only butchery as the main secondary activity, this likely reduced the applicability of the expectations to the reality of the archaeological record. Related to this is the fact that the main room floor is a nexus of many
243 different activities, many of which may not be described in the documentary record. The main room of Feature 87 is constructed differently than other houses in the documentary or archaeological record, which likely affected the use of space. Unlike other areas such as the tunnel or kitchen, the difference in the construction of the main room floor between documentary records and archaeological evidence suggests that the behaviours and activities on the main room floor could not be directly correlated between the two.
8.5 Kitchen Results
The kitchen had the highest percentage of expectations confirmed of any area in the house at 77% for combined expectations (13 total expectations, 10 confirmed) and 72% for separated test expectations (18 total expectations, 13 confirmed) (Table 12). As discussed in Chapter 7 on general subsistence, and in detail below, the kitchen is distinctly different from any other area, mostly due to the high number of small, burnt, indeterminate bones that make up the great majority of the assemblage. Of the 12,711 bones counted in the kitchen, only a small proportion of these bones were identified, mostly to small seal (n=493) and fish (n=128). All the other taxa had extremely low representation (Table 21). Large terrestrial mammal (n=3), small terrestrial mammal (n=2), and birds (n=2) will be discussed briefly, but the small sample size limits any interpretations based on these expectations and results.
Faunal material was expected to show activities related to the preparation of food including discard of low-utility elements and secondary burning in fires. Trampling, burning, and carnivore activities were expected to result in some density-mediated attrition.
8.5.1 Expectations Relating to Preparation Activities
The kitchen was expected to have evidence of preparation activities where larger packages were further reduced and where bones were completely removed from the meat. Five expectations were created to test preparation activities: absence of large terrestrial mammal lower limbs (Expectation F), high frequency of distal limbs for small terrestrial mammals (Expectation G), absence of bird specimens (Expectation H), high frequency of fish vertebrae (Expectation I), and intermediate frequency of fish (Expectation J). Of the five expectations that directly relate to preparation of food for cooking and consumption, four were confirmed. Only the high frequency of fish vertebrae was unconfirmed (Expectation I).
244
One of the specific preparation techniques in the region was the production of bone grease by smashing large terrestrial mammal lower limb bones. This was expected to be done in the kitchen, with storage of these lower limbs in the tunnel. The lack of lower limbs of large terrestrial mammals in the kitchen suggests that this may have occurred here. The high frequency of indeterminate bones (Expectation K) may also support this interpretation. However, as mentioned above, there are very few large terrestrial mammal bones in the kitchen (n=3), and so any interpretation is limited. Although there is an extremely high relative frequency of indeterminate bones in the kitchen, other activities from tertiary and quaternary stages discussed below greatly impacted this frequency.
The second expectation that was proposed to support preparation activities was the high frequency of distal limbs of small terrestrial mammals (Expectation G). Heads of small terrestrial mammals were expected to be removed in the main room, while the rest of the animals were moved into the kitchen for preparation. Here, distal elements would have been removed, as they do not have much meat associated with them. This expectation was confirmed, both absolutely and relatively, with only distal elements of small terrestrial mammals present. This may indicate that these bones were removed prior to cooking; however, the limited number of specimens (n=2) inhibits any interpretation.
As with large terrestrial mammals, the preparation of bird specimens was not expected to leave bones in the kitchen. It was expected that birds would be processed in the main room, and when they entered the kitchen, they were mostly placed in bags to be boiled. Although there are two bird specimens in the kitchen, the frequency of these is below 0.001, which is the lowest in the house. The lack of bird bones could suggest that the expectation that bird bones were not discarded in the kitchen during preparation is supported. The absence of evidence is not sufficient to confirm the expectation; as with large terrestrial mammals, birds may have been destroyed by tertiary and quaternary activities, rendering them unidentifiable. As discussed below, the evidence for destruction through burning or other activities is variable among taxa.
245
Normal Confirmation Taxonomic Skeletal Average Relative Type of Category Final Expectations Results Measure Results Ranking Absolute Relative Expectation Large A LMM not expected to be 0 N/A 0.005 5 N/A Yes Comparative marine present in the kitchen mammal B Overwritten by N/A N/A N/A N/A N/A N/A N/A (n=0) Expectation A Small seal C Slightly high ratio of dense 0.57 1.30 1.18 4 No No Skeletal (n=493) : less dense elements D Heads and hind flippers 0.22 0.28 0.31 5 No Yes Skeletal not expected to be present in the kitchen Large E High ratio of dense : less All dense 1.05 3.00 1 Yes Yes Skeletal terrestrial dense elements mammal F Lower limbs will be 0 0.42 0.35 5 Yes Yes Skeletal (n=3) smashed and rendered unidentifiable Small G High frequency of distal All distal 0.67 0.95 1 Yes Yes Skeletal terrestrial limbs mammal (n=2) Bird H Whole bird bodies cooked <0.001 N/A 0.005 5 N/A Yes Comparative (n=2) in bags with no discarded remains in kitchen Fish I High frequency of 0.49 N/A 0.93 3 N/A No Comparative (n=128) vertebrae L Intermediate frequency of 0.01 N/A 0.02 3 N/A Yes Comparative fish General J Low frequency of butchery <0.001 N/A 0.003 5 N/A Yes Comparative (n=12711) marks K High frequency of 0.81 N/A 0.54 1 N/A Yes Comparative indeterminate bone M Highest frequency of burn 0.92 N/A 0.61 1 N/A Yes Comparative marks N Even ratio of burnt bones Only small N/A Benches N/A N/A No Comparative across taxa seals burnt and tunnel more even Table 21: Kitchen expectations results. Expectations are placed in taxonomic order regardless of their alphabletical label. See Chapter 5 for a detatiled discusion of the construction of the expectations.
246
The final two preparation expectations relate to fish; only one of them was confirmed. Based on the documentary record, fish were mostly prepared in the kitchen, where heads and vertebrae were removed and discarded. Based on tertiary and quaternary activities that were expected to destroy all but the densest elements, vertebrae were expected in high frequencies (Expectation I). Fish vertebrae are present in relatively intermediate frequencies, with a low result in comparison to the house average. Most of the fish elements were not vertebrae, but less dense elements. This unexpected result may indicate that density-mediated attrition was not as important a factor as expected, which is discussed below, or it may indicate that fish were being processed differently than in the documentary record. Fish are expected in intermediate frequencies; slightly lower frequencies than the bench consumption areas and slightly higher than the midden discard areas (Expectation L). This is confirmed, with the frequency of fish in the kitchen close to the house average and relatively ranked third.
Although the absence of evidence is difficult to interpret, the lack of large marine mammals may support the expectation that the preparation techniques for these taxa are the same in the archaeological record as they are in the documentary record. Large marine mammals were not expected to be processed in the kitchen, as the meat that entered the house would not be attached to bones, or if they were, the bones that entered were flippers that did not undergo any additional preparation other than fermenting in storage (Expectations A and B). No large marine mammals are present in the kitchen; this is the only area without any large marine mammal specimens. The absence of large marine mammals may suggest that they were not processed in the kitchen, which supports the expectation. However, these bones may also have been destroyed due to tertiary and quaternary activities, thus rendering them unidentifiable.
8.5.2 Expectations Relating to Density-Mediated Attrition Activities
Due to high frequencies of burning during secondary and tertiary activities, and trampling and carnivore gnawing during tertiary and quaternary activities, all bones were expected to be exposed to medium-high frequencies of density-mediated attrition. Density-mediated attrition due to destruction during burning, trampling, and carnivore activities was expected to be apparent in three expectations: small seal and large terrestrial mammal high ratios of dense to less dense elements (Expectation C and E), and high frequency of fish vertebra (Expectation I).
247
Only the large terrestrial mammal expectation is confirmed; small seals and fish do not show typical evidence of density-mediated attrition.
Only dense elements of large terrestrial mammals are found in the kitchen; however, there is only one dense and no less dense element (Table A2.3). Overall, there are only three large terrestrial mammal elements in the kitchen. This low number negates both the absolute and relative confirmation for Expectation E. Additionally none of the other expectations are confirmed for density-mediated attrition.
Expectation C for small seals proposes that tertiary and quaternary activities would remove some of the less dense elements, leaving a slightly high ratio of dense to less dense elements. This is not confirmed either absolutely or relatively. The ratio of dense to less dense elements (0.57) is the second lowest in the house, and is significantly lower than either the normal skeletal measure (1.30) or the house average (1.18). This ratio shows that the majority of the small seal specimens in the kitchen are the least dense elements. This does not support the presence of activities that destroy the least dense bones of assemblages.
The fish expectation related to density-mediated attrition also does not support high or intermediate levels of density destruction (Expectation I). As discussed above, high frequencies of vertebrae were expected, as less dense cranial elements were expected to be destroyed; this is not the case, with only a low frequency (0.49) of vertebrae and with over half the specimens represented by less dense elements. This is lower than the house average, and ranks the kitchen third among the five areas. These unexpected fish results correlates with the small seal results that indicate that density-mediated attrition did not affect the assemblage in predictable or measurable ways. None of the results support the expectation that the kitchen was affected by activities that resulted in medium-high degrees of density-mediated attrition.
8.5.3 Expectations Relating to Burning Activities
Burning, both during cooking preparation and as a secondary activity as fuel (either intentionally or unintentionally), was expected to in the highest frequencies of any area in the house. The amount of burning in the kitchen was greater than expected; at 92% burnt, the kitchen is significantly higher than the second ranked area (tunnel at 25%), which supports Expectation M. However, burning was not even across taxa (Expectation N); small seals were the only taxa
248 burnt. Large marine mammals, small seals, large terrestrial mammals, and fish were burnt on the benches; the tunnel also had more taxa burnt than the kitchen with small seals and small terrestrial mammals both burnt (Table A2.7). This unexpected result may indicate that when burnt, all other taxa were rendered unidentifiable; it may also indicate that the kitchen burning was targeted to specific taxa. There is no clear support for either assumption, but the analysis below shows evidence to support more targeted burning.
The high degree of burning and the uneven burning raises the question as to what types of bones were burnt and for what purpose. Either bone was mostly burnt incidentally, as a by-product of cooking and fire in the kitchen, or was mostly burnt intentionally, as a means to fuel the fire. In the first instance, incidental burning, bones from all taxa should be equally burnt. In the second instance, bones may be more targeted in their burning and also may have been further processed to make them more suitable for burning. Evidence for further processing is discussed below.
The results from the general subsistence analysis, these distinct burning signatures, and the lack of density-mediated attrition for most taxa suggest that most of the burnt bones were broken into bits and then burnt. Based on the general subsistence analysis, most of the burnt material was shown to be small (<1cm), indeterminate or indeterminate mammal bones; 51% of the burnt material was cancellous, while 42% was indeterminate (Figure 41). These numbers suggest that burnt bone consists of small, broken specimens, with a large proportion of cancellous bone. However, the general subsistence analysis was unable to determine if the bone was broken prior to being burnt, during, or after burning.
A few lines of evidence indicate that the bones were broken prior to burning. The first is that 98% of the bones are 100% burnt (100% burnt=3167, total burnt=3242). This completeness of burning indicates that bones either were in their small pieces when they went into the fire, or they broke into pieces in the fire, where they continued to be burnt. If bones were broken after burning, more of them were expected to show less than 100% burning, as the insides of whole bones would not have been exposed to the fire.
Although burning may break bones into small pieces, then continue to be burnt, resulting in 100% burnt bones, the bones that remained would likely be high-density cortical bones. Burning reduces bones to small, indeterminate fragments; however, it does so according to bone density (Vaneeckhout et al. 2010; Vaneeckhout et al. 2013), specifically destroying porous, cancellous
249 bone first (Costamagno et al. 2010). Thus, if whole or mostly whole bones were put into the fire and then reduced into indeterminate pieces by burning, the expected bone left should indicate density-mediated attrition: high frequencies of high-density bone and low frequencies of low- density elements. Additionally, there should be a high frequency of cortical bone, with a very low frequency of cancellous bone. However, these are not the faunal signatures present at Feature 87. Here, burning did not remove all of the less dense elements for either small seals or fish; these taxa have higher frequencies of less dense elements than dense elements. Additionally, most of the burnt specimens in the kitchen are cancellous bones, indicating that cancellous bones from either terrestrial mammals or general marine mammal bones were disproportionately put into the fire. Although cancellous bone is only found in the long bone ends and axial elements of terrestrial mammals, it is found throughout most marine mammal bones. These lines of evidence, along with the fact that only specific taxa were actually burnt in the kitchen indicate that bone burning was a specific, targeted activity that only affected some remains.
Experimentally, the types of bones that burn best are terrestrial mammal bones and cancellous bones (Costamagno et al. 2010; Vaneeckhout et al. 2010; Vaneeckhout et al. 2013). Cancellous bone was likely a large part of burning assemblage, as the high frequency of cancellous bone indicates. However, this cancellous bone likely came from marine mammals, in particular small seals. This is the only taxon that is burnt in the kitchen (Expectation N). Although absence of evidence is difficult to assess, the absence of large terrestrial mammal bones may suggest that they were being used to fuel the fire in the kitchen for cooking and were likely broken into small pieces prior to entering the fire, thus reducing their ability to be identified (Expectation F). The fact that the only taxa to be burnt was small seals, along with the general subsistence evidence from Chapter 8, indicates that comminuted small seal, and potentially large terrestrial mammal bone, was being used to fuel the kitchen fires. Although small seals do not have as high of fat content as terrestrial mammal bones due to the fact that their marrow cavity is filled with cancellous bones, the marrow of large terrestrial mammals was likely being removed and eaten, leaving the long-bone cortical surfaces with little fat content and little use in fuelling cooking fires. Marrow could not be extracted from marine mammal bones, and so the use of these bones, still filled with fat, in fires is likely.
250
8.5.4 Summary of Kitchen Expectations
The kitchen is distinctive; its faunal sample conforms relatively closely to expectations. These expectation confirmations were based on overall taxonomic expectations and modification frequencies, specifically burning. Overall, taxonomic frequencies support preparation activities in the kitchen; the absence of large marine mammals and low frequencies of large terrestrial mammals and birds corresponds to the expectation that these bones were not deposited in the kitchen during preparation. The high frequency of fish and small seals follows the expectation that these taxa were prepared in the kitchen. The expected tertiary and quaternary activities are variably confirmed; most general expectations are confirmed. Burning affected 92% of the assemblage, while indeterminate bones made up 81% of the assemblage; although these were expected to be high, they were higher than anticipated, adding to the distinctness of the kitchen assemblage. The distinct taxonomic representation, modification frequencies, and fragmented nature of the kitchen assemblage also suggest that it is an important part of understanding the activities at an early Thule house: kitchens should be excavated and included in faunal analyses. However, they may skew some of the general results, and so should be analysed as a separate area, which can be removed from or added to analyses for comparative purposes.
8.6 Benches Results
Benches have the second lowest confirmation of expectations at only 43% confirmed of the separated expectations (23 total, 10 confirmed), and the lowest at 41% for combined expectations (17 total, 7 confirmed) (Table 12). This limited confirmation is likely due to the differences between the expected location of the documentary practices and the archaeological reality, combined with the fact that the bench structures in Feature 87 were different than anything described in the documentary record. People sitting or sleeping on the benches were the basis of the documentary observations on the use of the bench space. However, the archaeological reality was that there was very little material, if any, on the benches, and most of the faunal specimens were taken from under the benches themselves, which may indicate different activities and post- depositional practices. The under bench material likely represents the expected activities to some degree, since bone material would have dropped through the bench planks; however, this is swamped by other actions that lead to the accumulation of under bench material, as seen below. The bench structures in Feature 87, with a wide back platform and two narrow side benches, are
251
Normal Confirmation Taxonomic Skeletal Average Relative Type of Category Final Expectations Results Measure Results Ranking Absolute Relative Expectation Large A High ratio of appendicular : Yes Yes Skeletal 2.1 1.62 1.89 2 marine axial elements mammal B High frequency of bacula, N/A No Comparative 0.1 N/A 0.18 4 (n=70) unidentifiable, and ivory Small seal C Intermediate frequency of Yes Yes Skeletal 0.33 0.30 0.33 3 (n=2096) high-utility elements D High frequency of front No No Skeletal 0.15 0.27 0.17 4 flippers Q High ratio of small : large No Comparative 2.27 2.89 5 elements Large E Low frequency of high-utility Yes Yes Skeletal 0.41 0.62 0.43 4 terrestrial elements mammal F High frequency of lower limbs 0.33 0.42 0.35 3 No No Skeletal (n=97) L High frequency of antler 0.04 N/A 0.03 2 N/A Yes Comparative Small G High ratio of small : large N/A No Comparative terrestrial elements 1.54 N/A 2.51 5 mammal (n=150) Bird H High frequency of wings and No No Skeletal 0.37 0.36 0.34 3 (n=57) legs P High ratio of small : large bird N/A No Comparative 5.00 N/A 8.08 4 elements Fish I Even distribution of cranial : N/A No Comparative 0.24 N/A 0.36 3 (n=189) post-cranial elements K Highest frequency of fish 0.04 N/A 0.02 1 N/A Yes Comparative General J Intermediate frequency of N/A Yes Comparative 0.01 N/A 0.003 2 (n=4700) butchery marks M Intermediate frequency of N/A Yes Comparative 0.22 N/A 0.61 3 burn marks N Low total number of N/A No Comparative 0.21 N/A 0.2 2 specimens (NSP) O High ratio of small : large N/A No Comparative 4.36 N/A 12.92 5 bones Table 22: Benches expectation results. Expectations are placed in taxonomic categories regardless of their alphabetical label. See Chapter 5 for a detailed discussion of the construction of the expectations.
252 different than anything described in the documentary record for this area and this type of feature. Expectations were derived from the normal documented uses of the back platform for sleeping and sitting during consumption activities, socializing activities, and tool or garment making practices. Due to the narrow width of the side benches, they were likely not used for sleeping, but were potentially used for sitting and participating in the above activities. However, without direct comparisons, the expectations from the documentary record were solely an estimation of the activities that may have occurred there. This, along with the fact that the archaeological material mostly came from under the benches likely limited the confirmation of expectations.
Faunal material from benches was expected to show evidence of two secondary activities: consumption and use in manufacture. Tertiary activities expected on the benches were mostly cleaning and some trampling, with negligible amount of quaternary activities.
8.6.1 Expectations Relating to Consumption Activities
Five expectations were related to consumption activities; all but one of these was confirmed. Evidence of consumption was expected in the discard of large marine mammal appendicular elements (Expectations A), intermediate frequencies of small seal high-utility elements (Expectations C), low frequencies of high-utility large terrestrial mammals (Expectation E), even ratio of cranial to post-cranial elements, and high overall frequency of fish (Expectation I, and K). Only Expectation I, the even skeletal discard of fish elements, was not confirmed.
Large marine mammal appendicular elements were expected to be in high frequencies relative to the axial elements (Expectation A). Most large marine mammal bones would not be associated with consumption as the meat could be completely removed from the bone; however, flippers of bearded seal and walrus were often brought back to the site and eaten. It was expected that the appendicular to axial ratio would be high on the bench due to discard after consumption; this is confirmed both absolutely and relatively (Table 22). Large marine mammal appendicular to axial ratios on the benches are much higher than either the normal skeletal measure or the average house measure; it is second highest among the five areas, which supports the expectation. This result is considered meaningful based on a large sample size (n=70) of large marine mammals.
Consumption practices were expected to leave intermediate frequencies of small seal high-utility elements after some of the large elements were cleared during cleaning activities (Expectation
253
C). Small seals were likely consumed on the benches after undergoing butchery on the main room floor, and potentially some preparation in the kitchen; the high-utility parts with attached bones were expected to be moved to the benches where the meat was consumed and bones discarded. This expectation is confirmed in relation to both the normal skeletal measure and the average household frequency; it ranks third out of the five areas. As discussed in other sections, intermediate frequencies are more difficult to assign meaning to, as there are a number of factors that could influence an average result. Although the sample size is large (n=2906) and the expectation is confirmed, other factors likely are influencing this average result as well.
It was expected that high-utility bones for large terrestrial mammals would be in high frequencies on the benches, as they were discarded after consumption, but that cleaning would likely remove most of these large bones, leaving a relatively low frequency in comparison to the rest of the house (Expectation E). This is the case when the high-utility frequency (0.41) is compared to the normal skeletal measure (0.62) and to the house average (0.43). The result ranks the benches fourth, which also confirms the expectation. Again, this low frequency is influenced by cleaning, and may actually be more representative of this tertiary activity than the secondary activity of consumption.
Fish were expected in the highest frequencies on the benches, where consumption would increase their numbers and their small size would make them less likely to be cleaned. This expectation was met with fish proportions of 0.04 for the benches (Table 22). This was higher than the house average, and ranks the benches first among the five areas.
The only expectation related to consumption that was not confirmed was the even representation of fish cranial and post-cranial elements. The palimpsest of different types of fish consumption (heads in soup, summer-caught fillets with ribs, frozen whole fish) was expected to result in an even cranial to post-cranial distribution (Expectation I). This is not the case, with a low ratio of cranial to post-cranial elements, which is slightly lower than the house average and ranks the benches third. Specifically, there are more vertebrae and ribs than cranial elements (Table A2.6). This may indicate that fish were prepared differently than expected, or that there were more tertiary or quaternary activities that affected the benches than expected. These are discussed below. Despite the fact that the fish element representation was not as expected, there is strong support from multiple lines of evidence that consumption occurred on the bench.
254
8.6.2 Expectations Relating to Manufacturing Activities
Bone use in manufacture was the other secondary activity that was expected to contribute to the faunal assemblage on the benches. Five expectations were tested to see if documentary records could be used to interpret bone use in manufacturing tools or garments on the benches of Feature 87: high frequency of large marine mammal bacula, indeterminate, and ivory (Expectation B), high frequency of small seal front flippers (Expectation D), high frequency of large terrestrial mammal lower limbs and antler (Expectations F and L), and high frequency of bird wings and legs (Expectation H). As described in Chapter 5, these elements are all associated with tool, garment, or other types of manufacture, typically done in dwelling structures. Unlike the high degree of confirmation for consumption activities, only one expectation was confirmed for use in manufacturing: high frequency of antler (Expectation L).
Un-worked antler was expected to be discarded in high frequencies on the bench as it was used in many types of tools (Expectation L). This is partly supported, with the benches exhibiting the second highest frequency of antler in the house (Table 22). However, the lack of evidence from other expected faunal specimens indicates that manufacturing was either not conducted on the benches or, more likely, that other activities obscured any manufacturing signatures. Also, manufacturing is more easily interpreted using tools and artefacts rather than faunal remains; future work will need to examine the artefact assemblage. All worked antler, including antler debitage, is part of the artefact assemblage.
8.6.3 Expectations Relating to Cleaning Activities
Cleaning, a tertiary activity was expected to impact the bench assemblage, as people were likely to clean large and small bones off the benches in order to sit and sleep. However, some small bits of bone were expected to fall down the sides of the benches or get trapped between bench planks; therefore, cleaning was expected to preferentially remove large bones. This was expected to result in overall higher ratios of small to large bones (Expectation O), and taxonomically specific high ratios of small to large elements for small seal (Expectation Q), small terrestrial mammals (Expectation G), and birds (Expectation H). The overall removal of both large and small bones was expected to also leave a low number of specimens on the benches (Expectation N). Of these five expectations, none were met.
255
Although all small to large ratios (small seals, small terrestrial mammals, birds, and overall) all showed that there were often significantly more small bones than large bones, these ratios were relatively low in comparison to other areas of the house. Even if the kitchen is removed due to the high number of small, indeterminate bones present there, the ratios are still relatively low. For all taxa and even the overall analysis, the benches have the lowest frequencies of small bones of any area in the house (Tables A2.2, A2.4, A2.5, and A2.7). This is likely a result of the fact that in the documentary records, cleaning was expected to mostly impact the surface of benches; archaeologically, the expectations were tested on material from under the benches.
8.6.4 Summary of Benches Expectations
As with the main room floor, confirmations of expectations for the benches based on documentary records are limited. The one activity that does seem to be confirmed is consumption practices. The high degree of confirmation of this activity is likely related to a number of factors including the fact that consumption practices are most directly visible in the zooarchaeological record; other activities such as manufacturing or cleaning are not as directly linked to the faunal record. Even with the high confirmation of consumption activities, many of the expected faunal frequencies were mitigated by tertiary and quaternary activities, which may have affected these results in ways not examined by this analysis. Further work on consumption in dwellings will need to be conducted to see if this activity can be identified in spatial areas of other Thule semi- subterranean dwellings.
The benches in Feature 87 were structurally different than those recorded from documentary sources in northwest Alaska, which likely contributed to the low confirmation of expectations. There was no direct-historical correlate to the archaeological record, so the bench expectations are an interpretive step more distant from the documentary accounts than other areas.
Also, unlike the other areas, the benches do not represent a floor area in the documentary record, but rather a surface. In all the other areas, the documentary expectations were based on activities that occurred mostly on the floor; the bench activities are all conducted on the surface. Despite the fact that the expectations were situated on the bench, most of the faunal material comes from under the bench. This discrepancy between the documentary and zooarchaeological record is not found in other areas, and may account for the low confirmation of some activities. The faunal material under the bench may represent activities other than those recorded in the documentary
256 record. Further work will need to be conducted to explore the types of bones found under a bench; below is a brief suggestion.
For the back bench or platform, the area under the bench may represent storage practices. In the simplified model from which documentary expectations were created, storage was not considered one of the key activities on the bench due to the warmth of the house and limited accessibility. The faunal frequencies of the benches can be compared to the tunnel, where expectations were developed to test storage practices. However, the final expectations from the tunnel need to be carefully compared to the bench material as different tertiary and quaternary activities would differentially affect the assemblage. Evidence of storage practices may be seen in the ratio of appendicular to axial elements for large marine mammals; both the tunnel and the benches have the highest and most similar ratio (Appendix B; Table B.1). However, other storage expectations that were confirmed in the tunnel are not similar to the bench material. For example, the high frequency of large terrestrial mammal lower limbs in the tunnel is much lower on the benches (Appendix B; Table B.3). Construction of expectations for under the benches would be needed before further interpretations could be generated.
The main room floor material and bench material in Feature 87 come from similar floor surfaces. The similarity in butchery marks (Expectation J), frequency of small and large bones (Expectation O), and some element frequencies show that the main room floor and the benches were fairly similar. Thus, it may be that under the benches should be considered more closely related to the main room floor than to storage areas such as the tunnel. In either case, the architectural structure of the floor under the bench versus the main room floor will prescribe many of the activities and post-depositional processes that could affect the faunal assemblage.
The lack of confirmation of expectations in the bench faunal material is evidence of a disconnect between the often simplified expectations drawn from the documentary record and the blurred, ambiguous nature of the archaeological record. One example of this is the much more complicated division between the benches and the main room floor in reality than is outlined in the ethnographic accounts or expectations model. The division between bench and floor is complicated both through a grid-style excavation strategy and the fact that in reality, these areas were not separate rooms, unlike the kitchen or tunnel. Thus the low frequency of confirmation is
257 not especially surprising. What is interesting is the fact that consumption practices may be visible through the noise.
8.7 Midden Results
The midden had 10 of 20 separated expectations confirmed (50%) and 8 of 15 combined expectations confirmed (53%) (Table 12; Table 23); for both combined and separated expectations, the midden ranks third of the five areas in percentage of confirmed expectations. Although large marine mammal (n=4), large terrestrial mammal (n=7), small terrestrial mammal (n=3), and bird (n=3) expectations are discussed, they all represent very few specimens; the interpretations based on these taxa are not well supported due to sample size.
Faunal material from the midden was expected to show low frequencies of primary butchery activities, high amounts of bone deposition from cleaning activities, and evidence of carnivore gnawing, weathering, and human activity that would result in density-mediated attrition.
8.7.1 Expectations Relating to Primary Butchery Activities
The midden expectations were based on the assumption that some primary butchery discard may occur outside the house, but that it was formed mostly by discard of materials from within the house through cleaning activities. Some of the primary butchery that may have occurred in the midden was the removal of heads and hind flippers of small seals; however the less dense cranial elements were expected to be destroyed by contemporary and post-depositional activities, leaving an indeterminate frequency of hind flippers (Expectation D). This indeterminate frequency of hind elements for small seals was confirmed through comparisons with both the normal skeletal ratios and with the other areas. This confirmation indicates that some primary butchery of small seals may have occurred outside the house, and that density-mediated destruction was removing less dense small seal elements. Other expectations relating to tertiary and quaternary activities that may create density-mediated faunal signatures are discussed below.
The presence of non-subsistence large marine mammal elements was not expected, and these elements are in the second highest frequency of any area (Expectation B). This may indicate that the production of tools from elements such as ivory or whale bone occurred outside, where they could be directly discarded in the midden. Alternatively, these elements were too large to be discarded in the house around the benches, and so were frequently cleaned into the midden.
258
Normal Confirmation Taxonomic Skeletal Average Relative Type of Category Final Expectations Results Measure Results Ranking Absolute Relative Expectation Large A Intermediate frequency of 0.5 0.62 0.52 3 Yes Yes Skeletal marine appendicular elements mammal B Bacula, whale ribs and mandibles 0.25 N/A 0.18 2 N/A No Comparative (n=4) not expected to be present in the midden Small seal C High ratio of high-density : low- 0.69 1.30 1.18 4 No No Skeletal (n=170) density elements D Intermediate frequency of hind 0.27 0.27 0.30 4 Yes Yes Skeletal flippers Large E High ratio of high-density : low- 2.00 1.05 3.00 4 Yes No Skeletal terrestrial density elements mammal F Lower limbs not expected to be 0.29 0.48 0.35 4 No Yes Skeletal (n=7) present in the midden Small G Slightly high ratio of cortical : Neither N/A 0.5 5 N/A No Comparative terrestrial cancellous bone present mammal (n=3) Bird H Slightly high ratio of cortical : No N/A 0.23 5 N/A No Comparative (n=3) cancellous bone cortical Fish I Fish not expected to be present in None N/A 0.36 5 N/A Yes Comparative (n=0) the midden General J Low frequency of butchery marks 0.003 N/A 0.003 4 N/A Yes Comparative (n=357) K High total number of specimens 0.01 N/A 0.20 5 N/A No Comparative (NSP) L Low ratio of large : small bones 0.12 N/A 0.08 4 N/A Yes Comparative M Low frequency of burnt bone 0.12 N/A 0.61 5 N/A Yes Comparative N High frequency of carnivore 0.02 N/A 0.01 3 N/A No Comparative marks O High frequency of weathering 0.01 N/A 0.01 4 N/A No Comparative Table 23: Midden expectations results. Expectations are placed in taxamonic categories regardless of their alphabetical label. See Chapter 5 for a detailed discussion of the construction of the expectations.
259
However, there are only four specimens of large marine mammal, which means that only one element is represented in Expectation B. This small sample size hampers any concrete conclusions.
All other taxa that may have undergone primary butchery with bones discarded in the midden affected by tertiary and quaternary activities, masking their original primary butchery signatures. This is the case for large terrestrial mammals, small mammals, and small birds (Table 7, Column Stage 2).
8.7.2 Expectations Relating to Cleaning Activities
The addition of bones from the house through cleaning activities was expected to add many specimens to the midden; at the same time, contemporary carnivore activity is expected to have destroyed many of the fragile bones that may have been added, such as fish (Expectation I). Five expectations reflect the addition of dense material cleaned from the other areas: the medium frequency of appendicular elements of large marine mammals (Expectation A), the high ratio of high-density to low-density elements of small seals (Expectation C) and large terrestrial mammals (Expectation E), and the slightly high ratio of cortical to cancellous bone for small terrestrial mammals (Expectation G) and birds (Expectation H). Of these, only the large marine mammal frequency of appendicular elements (Expectation A) was confirmed both absolutely and relatively; the large terrestrial mammal high ratio of high- to low-density elements was confirmed absolutely but it was not confirmed relatively (Expectation E) (Table 23). The lack of confirmation may indicate that the bones expected to be discarded based on the documentary record were not cleaned out of the house into the midden; it may also indicate that contemporary and post-depositional activities affecting the midden were different than expected.
8.7.3 Expectations Relating to Density-Mediated Attrition Activities
Carnivore activity in both tertiary and quaternary stages, weathering, and human actions were expected to destroy less dense bones in the midden. Four expectations discussed above as indicators for cleaning are also indicators for density-mediated destruction: high ratios of high- to low-density elements for small seals (Expectation C) and large terrestrial mammals (Expectation E) and slightly high ratios of cortical to cancellous bone for small terrestrial mammals (Expectation G) and birds (Expectation H). As discussed above, only the large terrestrial
260 mammal high ratio of high-density to low-density elements (Expectation E) was absolutely confirmed. This indicates not only that these bones were not added to the midden in expected ways during cleaning activities, but also that density-mediated activities did not affect the midden as significantly as expected. However, there were very few bird, large terrestrial mammal, or small terrestrial mammal bones in total, so all expectations must be considered as potential indicators rather than as evidence.
Additionally, five other expectations were expected to show evidence of these density-mediated activities: absence of fish (Expectation I), high total number of specimens (NSP) (Expectation K), low ratio of large to small bones (Expectation L), high frequency of carnivore marks (Expectation N), and high frequency of weathering (Expectation O). Two of these five expectations are confirmed (Table 23). Fish are completely absent in the midden, indicating that they were not discarded in this location or that they were completely destroyed by density- mediated attrition. The low ratio of large to small bones was also confirmed; however, when compared to the results from the other areas, this ratio (0.12) is the same as that of the tunnel (0.12), and only second to the main room and benches (0.23 for both). However, only the kitchen has a lower frequency at less than 0.001. Thus, this expectation is difficult to confirm. With such a low number of expectations confirmed for density destructive activities (three of ten expectations), the entire model of contemporary and post-depositional activities should be questioned. These mixed results for bone density attrition indicate that the midden may have been affected by density-mediated attrition, but that not all of the bones were destroyed uniformly, and many were likely protected as the midden accumulated.
The variable evidence for carnivore activity, weathering, and other general density-mediated destruction may indicate that the midden was buried quickly. The low frequency of weathering in comparison to other areas, the intermediate frequency of carnivore marks, and the low ratios of dense to less dense elements for small seals or cortical to cancellous bones for small terrestrial mammal and birds suggest that the expected post-depositional activities did not affect the midden as expected. The presence of less dense elements for many of the taxa and the low degree of weathering, second lowest in the house, suggest that the midden was not left exposed to the elements or animals for any length of time. Rapid burial of middens is not uncommon; refuse can build up rapidly, and when placed in a midden, new refuse can cover old refuse, preserving the midden material and potentially causing this material to enter into the permafrost. This excellent
261 preservation and quick burial is why middens are often excavated in Arctic archaeological sites. This is supported in the documentary record indirectly, as Burch (2006:268) comments on the massive amount of waste produced at a settlement. This amount of waste would cover the midden piles quickly, leading to both good preservation and limited modifications from external agents.
8.7.4 Summary of Midden Expectations
The midden shows an intermediate confirmation of results. When broken down, the midden does not show strong confirmation of primary butchery activities, cleaning activities, or contemporary or post-depositional activities. The material analyzed for the midden only came from one quadrant of a midden square. This small sample may cause distortion in the expected frequencies. Future work would need to look at the differences of material within a midden in order to investigate nuances in discard practices. Additionally, evidence for activities that affected the midden was extremely limited in the documentary record. Further archaeological research on midden and external areas of semi-subterranean Thule houses is needed to understand the use of space outside of houses (see Newell 1990; Polglase 1990). The little that is known from the documentary record indicates that it was not a monolithic, random accumulation of refuse.
8.8 Summary
This chapter has examined the ratio and frequency results based on the expectations outlined in Chapter 5. As described above, the kitchen is a very distinct area, with most of the expectations met. Although the midden also has about half of the expectations met, these expectations are distributed among primary butchery, cleaning, and density-mediated attrition activities; this limits the interpretive strength of the conclusions of all the activities. The main room floor and the benches had fewer than half of their expectations met, and were similar in some skeletal ratios, modification frequencies, and size proportions. Neither was highly distinctive from the other. The tunnel had just over half of the expectations met, with a few lines of evidence suggesting storage practices. In the tunnel, benches, and kitchen secondary activities of storage, consumption, and preparation (respectively) were all generally confirmed. This result indicates that even with the contemporary and post-depositional activities that obscure earlier activity, some of the activities described in the documentary records are visible archaeologically. Further
262 testing will need to be done to evaluate how representative these results are in relation to other houses in the Arctic. Although some tertiary and quaternary expectations were confirmed in some areas, such as the kitchen, the limited confirmation is likely a result of incomplete understanding of these activities. Further research into these types of documentary and taphonomic activities must be done in order to develop a full understanding of the history of the house.
Overall, the skeletal expectations were confirmed to a greater degree than the comparative expectations (Table 13). When divided into their individual tests, the overall confirmed proportion is very similar, although the confirmation varies greatly among areas (Table 15). Additionally, the expectations based on frequencies were confirmed in greater proportion that those tested using ratios (Table 14). These differential confirmations indicate that when using the documentary record, different lines of evidence should be used for interpretation.
263
Chapter 9 Conclusion
This research had three major interrelated goals. The main goal was to assess whether the documentary record, combined with an understanding of taphonomic processes, could be used to accurately predict the spatial patterning of faunal remains within a cold-season Thule dwelling in order to identify behaviours, activities, and practices associated with these patterns. When the expectations were compared with the faunal record, some expectations were not met, which indicates that some parts of the model are not good predictors of the archaeological material, at least in this case. Despite the fact that various expectations remained unconfirmed, this model highlights a number of behaviours that are visible in the zooarchaeological record that were expected based on the documentary record. A secondary goal was to understand the behaviours and activities within features using archaeozoological assemblages. Exploring the correlation between the documentary and archaeological records and analyzing spatial patterning through faunal material contributed to the third goal: a fuller understanding of Thule culture history, specifically the subsistence economy and use of space within a semi-subterranean dwelling at Cape Espenberg. The analyses and interpretations described in the preceding two chapters were contextualized through discussions of analogical, zooarchaeological, and spatial method and theory, through descriptions of culture history of the region, and in relation to the specific assemblage.
Chapter 1 situated the goals of the thesis within a broader discussion of analogy, spatial patterning, and household analyses. The first section of Chapter 1 outlined discussions and critiques of the use of analogical reasoning to interpret archaeological material. Many analogical interpretations of archaeological material use the documentary record as a source for understanding the archaeological record; it is from descriptions and critiques of documentary analogy that this study was constructed. The documentary record has been used to interpret many aspects of the Thule way of life. Unlike some aspects of Thule lifeways, the use of space in houses has seldom been analyzed, with most interpretations relying heavily on the documentary record. Globally, intra-feature spatial patterning has been an important part of understanding lifeways across many regions, but has rarely been interpreted using faunal material. The second section in Chapter 1 briefly outlined the history of spatial studies, and specifically described the
264 ways in which faunal material has been used to study space within and among features. There is a distinct absence of clear methodology in the zooarchaeological literature for understanding space in houses using animal remains. The third part of Chapter 1 described the unit of analysis for this study, specifically the house. Although household archaeology has a long and important history, it is argued that houses in northwest Alaska during the Thule period likely do not represent entire households, but that the bounded unit of the house is itself an important unit of analysis.
Chapters 2 and 3 situated the research within its cultural context. Chapter 2 placed the western Thule culture within the culture history of the region; Chapter 3 discussed important interpretations of the Thule way of life based on archaeological and documentary data. Chapter 3 also discussed the behaviours and actions that create space within houses as described in the documentary record. These reconstructions summarized the documentary information used in creating expectations, which are further developed in Chapter 5. Prior to developing these expectations, the specific context of Feature 87 was described in Chapter 4. Here, the geographical, environmental, social, and archaeological context of the site was outlined, with specific descriptions of the house architecture. The house context influenced the expectations developed directly from the documentary record of the region in Chapter 5. Prior to moving to the results, Chapter 6 described the terminology and methodology used in the faunal and spatial patterning analyses. Chapter 7 first analyzed the differences between floor and fill material, then used the floor material to interpret how people were using animals at the site. Basic subsistence practices, seasonality, and interaction within the region were explored. After the general subsistence patterns of early Thule at Feature 87 had been outlined, the results of expectations developed in Chapter 5 were interpreted. Chapter 8 detailed the expected results and interpreted them by area. The different activities, behaviours, and processes that were expected to affect each of the five areas were compared to the archaeological record. The current chapter discusses the different ways that the results from Chapters 7 and 8 have contributed to broader archaeological and anthropological conversations.
9.1 Contributions to Method and Theory of Analogical Reasoning
By determining which behaviours and activities could be correlated with the documentary record for Feature 87, this research expands on the use of analogical reasoning for interpreting
265 archaeological material in Thule houses. Analogy is a type of reasoning that uses established similarities between a source and a subject to propose more extensive similarities (David and Kramer 2001b; Gould and Watson 1982; Roux 2007; Shelley 1999; Wylie 1982; 1985). Thule archaeological material (the subject) has often been interpreted through comparisons with the detailed documentary record (the source) as seen in the cultural setting of the Thule Inuit in Chapter 2 and in the description of the Thule way of life in Chapter 3. In most interpretations of Thule archaeological material, similarities between the documentary record and archaeological material have been identified, and further interpretations have been proposed that draw from the documentary record. The use of analogy that finds similarities in the documentary record and proposes further similarities is most commonly used in archaeological interpretation. In this study, it was used to reconstruct subsistence patterns of early Thule at Feature 87. Another use of analogy is to create expectations from the documentary record which are then tested against the archaeological material. This study used this approach when creating and testing expectations on the use of space within dwellings. In both uses of analogy, it has been shown that analogies can only claim probability of interpretation, rather than validating or invalidating an interpretation (Wylie 1982); known similarities between the source and subject do not indicate with certainty similarity in all other areas or in actions, behaviours, or beliefs that created that similar material culture (David and Kramer 2001b; Gould and Watson 1982; Wylie 1982). However, interpretations can be strengthened in a number of ways, such as by establishing more similarities between the source and subject or by using multiple lines of evidence to corroborate an interpretation, what Wylie (2002b) terms ‘bootstrapping’.
In this research, analogy that is used to interpret spatial patterning in dwellings was strengthened by establish strong similarities and exploring the differences between the subject and source material prior. This research builds on previous ethnoarchaeological, experimental, and archaeological research that attempts to identify behaviours, patterns, and actions that can be interpreted from archaeological material (e.g. Allison 1999b; Binford 1978b; 1980; Cutting 2006; Daviau 1993; Kent 1981; LaMotta and Schiffer 1999; Lightfoot 1993; Schiffer 1987). Results from this one house at Cape Espenberg have shown that certain behaviours and activities are very similar in the documentary record and the faunal record and are associated with particular areas of the house.
266
One of the most significant results is that the strongest confirmations were from those expectations generated directly from the documentary record of the region (secondary expectations), not from the contemporary activities (tertiary expectations) or post-depositional activities (quaternary expectations). This means that at Feature 87, the faunal record mainly reflects the behaviours and activities in which archaeologists are most interested and which ethnographers and other observers would have recorded. The benches, kitchen, and tunnel all showed strong evidence of activities derived directly from the documentary record. Although the confirmation of the bench expectations were generally low, consumption practices were fairly accurately represented in the faunal material. Five expectations were constructed to test for consumption practices on the bench, and all but one of them was confirmed (Table 8.11). Many of the kitchen preparation expectations were also confirmed. Preparation expectations that reduced meat packets by removing distal limb elements and de-boned large high-utility packages were strongly supported. Again, five expectations were constructed to test for preparation activities in the kitchen, and all but one of them was confirmed (Table 8.10). The tunnel had an intermediate frequency of expectations confirmed overall, but had three of four storage-related expectations confirmed. Even with evidence of some density-mediated destruction, cleaning, and other modifications, consumption on the benches, preparation in the kitchen, and to a lesser extent storage practices in the tunnel were all visible in the faunal record for Feature 87. Neither the main room floor nor the midden showed strong evidence of secondary activities and were more likely affected than the benches, kitchen, and tunnel by other secondary activities and tertiary and quaternary processes.
Strong similarities between the source (documentary record) and subject material (faunal remains) for consumption, preparation, and storage activities in their respective areas increases the probability that other interpretations based on the documentary record related to these archaeozoological materials may also be supported. For example, fish are present in high frequencies on the benches; this may not only support consumption activities on the benches, but may also support interpretations of seasonality, hunting methods, and transport decisions. In the documentary records, fish in the winter were caught, frozen, transported whole to the site, and consumed whole on the benches. In contrast, fish caught in the warm season were processed by removing most of their bones; confirmation of a high number of bones on benches for consumption practices may suggest that archaeologically people used fish in similar ways as
267 described in the documentary record. In order to strengthen this interpretation, other lines of evidence would need to be provided; for winter fishing, this may come in the form of specific artefacts used for fishing on the ice, or through similar fish bone patterns on the benches in other early Thule semi-subterranean houses. Without this corroboration (or bootstrapping), any interpretation using the same documentary record for interpretation that was used to establish the expectations is circular (Wylie 2002b). Through an iterative process of creating, testing, and interpreting, expectations based on the documentary record can interpret with greater certainty the actions, behaviours, and potentially ideology behind the patterning of archaeozoological material. This type of research that strengthens the use of the analogy is necessary in archaeological interpretation, since analogy is the fundamental tool archaeologists use to interpret past remains.
Although this is one study in one region, it shows that, at least here, the direct-historical approach can be used to interpret and extrapolate on both the subsistence practices and spatial patterning of faunal material relating to subsistence activities. From this study, the practices surrounding the preparation, consumption, and storage of faunal material identified in the documentary sources can be used carefully to interpret further the early Thule material in the region.
Archaeologists are concerned with the influence of their prior knowledge of the documentary record, both conscious and unconscious, on their interpretations of the archaeological material. This is a salient concern when using any type of analogy, but can be explored and tested. This study has shown that for one early Thule feature in northwest Alaska, the activities primarily observed and written about in documentary records are also observed in the deposition of faunal remains. Consumption on the benches, preparation in the kitchen, and storage in the tunnel were all confirmed as hypothesized based on the documentary record for the area. Other factors such as cleaning, trampling, carnivore gnawing, and burning did not drastically alter the expected frequencies for these three activities. This supports the archaeological interpretations of the benches, kitchen, and tunnel designations, and also validates the use of the documentary record in understanding these activities. If this pattern holds true for other houses and with other lines of evidence, the interactions and beliefs that underlie these actions may also be interpreted. Despite the limitations of the documentary record outlined in Chapters 1 and 3, the Arctic documentary record can be used to nuance archaeological inferences and theorize about the patterning of
268 subsistence and processing activities within a house. Although further study is needed within the Arctic and beyond, this study has indicated that the documentary record is useful in interpreting the activities within houses that are related to subsistence practices.
9.2 Contributions to Spatial Studies
This research proposed a method for using faunal analysis to interpret spatial patterning within a feature, and then used this method to interpret spatially distinct activities in an early Thule house. Previous work in the Arctic has looked at basic modification, taxonomic, and element frequencies (Darwent and Foin 2010; Friesen and Betts 2006; Howse 2008); this research expanded on these analyses. First, areas were defined within the house; faunal material within these areas was then compared and contrasted using skeletal and comparative indices. Although frequencies were used in understanding the distribution of taxa and elements among the areas, as had been the case in the previous analyses, ratios were also used to understand the distributions of elements. Ratio analysis of specific element sets or bone types allows a more fine-grained approach to understanding the distribution of fauna. Both frequency and ratio indices are useful in analyzing faunal data; this type of relational data will allow faunal patterns to be compared among other early Thule houses, and can be used as a model for other intra-feature spatial analyses. The methods developed during this research used large amounts of faunal data to identify different faunal patterns in different areas of the house. Faunal remains have yet to be used extensively in understanding how people created, used, maintained, and conceptualized space, and this is one way to generate data to investigate these questions.
This research applied these methods to the Feature 87 faunal assemblage in order to interpret small-scale activities and behaviours that occurred in a dwelling feature. Faunal data may reveal different information about the use of space in houses in comparison to architectural or artefactual data. As discussed above, the faunal material showed that in Feature 87, certain animals were likely consumed on the benches, much of the preparation was probably done in the kitchen, and some meat packages with associated bones were likely stored in the tunnel. Other interpretations, such as modification frequencies, also revealed interesting patterns. Although burning was expected to be present in the kitchen, the extremely high percentage of burnt bone (92%) was unanticipated. Examination of the burnt material in the kitchen in comparison to the burnt material in the rest of the house revealed that burning in this area was quite different than
269 burning in other areas. Burning in the kitchen was very specific, with only small seal and mostly small pieces of cancellous bone showing evidence of being burnt. In the rest of the areas, burning was more evenly distributed, and was not as significantly dominated by small, cancellous, or indeterminate bone. It was suggested that the burnt bone in the kitchen produced as a result of bone being used as a fuel source, where marine mammal and terrestrial mammal cancellous bone was broken into small pieces prior to entering the fire. The limited amount of burning in other areas with a different signature suggests that burnt bone in other areas is not produced from the use of bone as fuel, but is likely a result of unintentional secondary burning. Another interesting spatial difference was the low frequency of modifications on the midden material. Unlike the results from Friesen and Betts’ (2006) study of spatial patterning of a Thule house at Cache Point in the Mackenzie Delta, the midden material at Feature 87 in Cape Espenberg showed very limited amounts of exposure to weathering or animal activity, which may indicate different uses, or depositional histories of these middens.
In Feature 87, the difference among areas was most apparent when space was delineated by architecture. Areas used for multiple purposes, such as the main room floor or midden, have less distinctive signatures and fewer correlations with the documentary record. Areas with sharply delineated space, such as the tunnel and kitchen, had the highest confirmation of expectations and the most distinct assemblage composition. This interpretation should be tested to see if it holds true across different regions beyond the early Thule semi-subterranean dwellings.
This research has expanded the methodology for understanding spatial patterning in the faunal record, and has added to the archaeological interpretations of spatial patterning within dwellings. The results from Chapter 8 identified spatially-defined practices in specific areas in early Thule households through the use of the faunal record. Faunal material can and should be used to interpret small-scale activities, behaviours, and patterns since it represents different aspects of human lifeways than either architecture or artefact spatial patterning. This analysis is one of a limited number of studies from around the world that has looked at the differences in space within dwellings using faunal material. Methods such as those suggested here can be used to understand spatial patterning across different societies in different regions. Through high- resolution excavation and application of frequency and ratio indices in both skeletal and relative comparisons, faunal material can be used to understand different uses of space within houses.
270
9.3 Contributions to Thule Subsistence
This research has also contributed to understanding early Thule subsistence. The faunal material at Feature 87 indicates that small seals were the focus of subsistence throughout the cold season at Cape Espenberg during the fourteenth century. Ringed seals (Pusa hispida) were the most common small seal in the assemblage; spotted seals (Phoca largha) contributed a fairly sizable proportion of identified seal elements (19% of the small seals by NISP), and ribbon seals (Histriophoca fasciata) were rare (2% of small seals by NISP). Ringed seals, a ubiquitous and important prey species for Thule people across the Arctic, are ice dependent, maintaining breathing holes in the land fast ice throughout the winter. Ringed seals were likely the staple of the cold season economy, which is supported by the age at death data analyzed by epiphyseal fusion. Spotted and ribbon seals are not present in the area during the winter, but follow the ice edge south. These seals were likely hunted in the warm season, either via boats on the open water or on the ice-edge by stalking, harpooning, or netting.
Although small seals were likely the staple of the cold season, the presence of large baleen whale bones, the cut marks on these bones, and the elements present suggest that the occasional large bowhead whale was hunted. Although the number of specimens is low, the meat weight analysis shows that bowhead whales contributed the most to the diet. Even though the evidence does not suggest that whales were regularly hunted, the successful hunt of just one large whale would drastically impact the community’s economy for the season.
Other large marine mammals also contributed to the diet in terms of meat weight. Beluga whales, walrus, and bearded seals all contribute low proportions to the %NISP and %MNI, but make fairly substantial contributions in terms of meat weight (though caution must be exercised in interpreting meat weight calculations). Like large baleen whales, the low bone frequencies suggest that these animals were not targeted at Cape Espenberg, but were hunted when they were encountered. Alternatively, it could indicate that whales and other large marine mammals were hunted far from the site, with only specific elements brought back to the site.
Among terrestrial mammals, caribou ranked the highest in both NISP and meat weight measures. Caribou were an important species to Thule groups, even those who lived on the coast. They provided much of the raw material such as fur, sinew, bone, and antlers. Caribou could be occasionally taken on the coast, but were likely hunted inland. Element frequencies and
271 epiphyseal fusion show that these caribou were probably hunted away from the site, likely in the fall, with only certain elements being brought back to the site. The positive correlation with the unsaturated marrow index and fracture freshness index suggests that the remains in the house may have been curated for marrow and bone grease processing.
The presence of animals most typically associated with the forest may indicate seasonal movement, hunting parties, or trade with the interior. Moose, beaver, and mink are all present in low frequencies. The elements that make up the assemblage of beaver and mink suggest that they were brought whole to the site, which may indicate hunting rather than trade for skins. In either case, the presence of these unexpected species serves as a reminder that people were mobile, with connections beyond the immediate region.
Although dog or wolf (Canis lupus) specimens make up less than 3% of the NISP assemblage and only about 1% of the meat weight, they were likely an important species to the people of Feature 87. Age at death, skeletal element distribution, and cut mark locations suggest that older dogs that died were skinned and discarded whole, rather than undergoing butchery and preparation typically associated with consumption. Although they may have been eaten in times of famine, neither dogs nor wolves were regularly eaten; often their fur was used to trim clothing. Other furbearers such as foxes, arctic ground squirrels, and hares likely also contributed to clothing production, and potentially to break up the monotony of the consumption of seal meat.
Birds, fish, and molluscs all contributed to the faunal assemblage in low frequencies. There are a wide array of different birds such as geese, ducks, terns, loons, raptors, and dunlins. Waterfowl make up the majority of the remains, followed by ptarmigans and shorebirds. A number of bird bones contained medullary bone, which indicates that they were caught in the egg-laying season, between May and June. These birds were either stored for the winter, or more likely caught at the end of the cold season. A small range of species including herring, salmon and tomcod make up the fish assemblage. The presence of both cranial and postcranial remains may suggest that the fish were caught in the winter. People likely collected molluscs on the beach or intertidal zone during the course of their daily activities.
The general subsistence analysis of Feature 87 has outlined subsistence practices for early Thule Inuit at Cape Espenberg in western Alaska. This area does not have extensive archaeozoological
272 data and so this analysis begins to detail the subsistence practices in the area. This analysis also has contributed to the knowledge of the larger Alaskan Thule subsistence strategies as very few detailed studies have been published on the faunal material. This study is likely one of the most accurate analyses to date of coastal Thule faunal material in Alaska since it used the entire floor assemblage. Many older analyses of Thule material are very cursory, and most were based on samples taken only from parts of the house floor. As has been demonstrated both in this study and in previous studies, different areas within dwellings have different taxonomic and modification frequencies (Friesen and Betts 2006), which would only outline a small portion of the overall subsistence strateiges. Additionally, many other analyses have used the entire assemblage without separating fill from floor material. In this analysis, the floor was defined through archaeological context, but was also tested through a comparison of fill and floor faunal material. As with differences among the floor areas, the fill and floor materials were also distinct. The difference in fill and floor material highlights the need for defining and testing sample contexts in order to carefully reconstruct the lifeways of past peoples. General subsistence strategy reconstructions would change if the samples were not restricted to floor material. Detailed archaeozoological analyses help to expand the knowledge of early Thule people beyond artefactual remains to explore the different ways people interacted in local areas and with other regions.
9.4 Limitations and Future Work
The high-resolution excavations at Feature 87 have supplied a robust faunal assemblage on which to base interpretations of early Thule subsistence and spatial patterning, to test methods for understanding spatial patterning of faunal material, and to analyze expectations based on the direct-historical method in order to strengthen the use of analogy. This is only one step in understanding early Thule culture history of the region, spatial patterning of Thule houses, and analogical reasoning using the Iñupiat documentary record. Further work will need to be conducted on all aspects of these interpretations in order to explore similarities and differences.
As discussed in Chapter 1, individual houses in the Thule period likely only represent part of a household. The functionalist ethnographic description of a household as an activity group that engages in one or more practices of production, consumption or distribution, reproduction, co- residence, and transmission (Ashmore and Wilk 1988; Wilk and Netting 1984) can be equated to
273 the compound family (Burch 1998a; 2005; 2006) or the small village (Ray 1964) in northwest Alaska. Thus, the description of subsistence practices for one house limits a complete understanding of the household. However, establishing contemporaneity of houses or delineating all the spaces used in a cold season village by a single household in order to establish and analyze a household is extremely difficult, which is why the house, as a bounded, archaeologically visible feature, is used as a discrete unit of analysis that can be compared and contrasted within a site and across a region. Understanding that these interpretations of subsistence and spatial patterning are limited to one part of a household unit during one season of occupation is very important; people and subsistence resources likely moved both within the village itself and throughout the region. Evidence of some of this comes in the form of unexpected animal remains from other regions such as mink or beaver, and the uneven distribution of animal remains that may have been shared. Future work that explores contemporaneous houses within a small village and the external areas between these houses will be needed in order to gain a more complete, detailed picture of early Thule life in a cold-season coastal village. For now, this research is able to interpret the subsistence practices and spatial patterning of one domestic dwelling within the village.
This research has presented methods for understanding spatial patterning and testing expectations derived from the documentary record. It worked from expectations to analysis to understand the use of space within houses, and also to establish expectations in order to test the correlation between the documentary record and the archaeological assemblage. For this, areas were defined archaeologically and differences in faunal material were interpreted among these areas. In future, a comparison between this approach and an approach that looks for distinct areas based on the faunal material may provide different interpretations of the divisions of space within the house. This type of exploratory approach may identify patterns not present in the documentary record.
A caveat that has been present throughout the analysis and discussion is that the model of expectations is a simplification and extrapolation of the reality of activities, behaviours, and processes that occurred in the past. Most of the documentary accounts do not specifically describe how bones were placed within the houses or the activities that distributed bones; expectations were created based on extrapolation of these activities. However, there was also simplification; many activities occurred in many areas, and so only the most prominent activity was detailed for each area. Both through extrapolation of the documentary accounts and through
274 simplification of the activities that occurred in the different areas, there are errors and inaccuracies introduced. However, this type of expectation testing is a necessary step in creating reliable and accurate interpretations of the archaeological record. Nuance can and should be added back into all interpretations based on the archaeological context. Future work will help establish the broader patterns and regional or temporal distinctions, which may allow for detailed and more extensive use of the documentary record in interpreting archaeological remains.
One issue that is important in discussing confirmation of expectations is the issue of equifinality; expectations, even when confirmed, cannot be assumed to have direct relationships with the documentary behaviours from which they are derived. Other processes may have impacted the assemblage, which produce similar results. Equifinality can be mediated by careful examination of the archaeological context to understand how and why the expectations were confirmed or denied. Through this examination, multiple lines of evidence can be used to bootstrap an interpretation (Wylie 2002b). For the confirmed secondary activities, consumption, preparation, and storage, other lines of faunal evidence, architectural evidence, and taphonomic evidence were used to strengthen the interpretations. More work on and testing of these correlations between the documentary record and the faunal record will be needed to further support or refute these claims. If future work does corroborate the findings at Feature 87, these important correlations between the archaeological record and the documentary record can be used to expand interpretations about the behaviours and motivations around these activities.
Finally, the documentary record itself is a limiting factor in establishing expectations to test in the faunal record. Multiple factors may lead to an incomplete documentary record, including the inability to record all activities, even for one group for an entire year. Another factor that contributes to the absence of data in documentary accounts is the variability among groups, regionally, seasonally, and through time. Using other cross-cultural sources, experimental tests, and generalizations on human actions can help fill in the gaps. Further careful understanding of site formation processes, from deposition through excavation and analysis, is needed to identify cultural impacts on dwelling assemblages. Comparison with other assemblages and other documentary sources may show patterns that indicate what types of activities affected the assemblage that were not recorded in the documentary record. This type of interpretation must move beyond the direct-historical or documentary analogical models while staying grounded in those analogies that have been tested and strengthened. As documentary sources are strengthened
275 based on the archaeological record, expectations based on other types of sources can be proposed and tested that explore activities not present in the documentary record of the region. It is from these types of strong analogical interpretations that the past can be constructed.
9.5 Concluding Remarks
Future work will be necessary to test the results and methods defined above. Other early Thule houses need to be excavated in high-resolution grids and compared to the results of Feature 87. With a larger sample, commonalities among early Thule dwellings can be established, and differences can be explored. Spatial patterning also needs to be interpreted across time, compared and contrasted with the predecessors of the early Thule Inuit and with their descendants in the same region. These types of analyses will allow a greater understanding of the people who occupied the houses, how they ordered their house space, and potential changes through time and across the northwest Alaskan region. Once correlations with the documentary record have been strengthened, further investigations into activities, practices, and behaviours outside of the documentary record can be interpreted. Clarifying and deconstructing the analogies used to interpret archaeological remains is a fundamental part of the discipline; strong analogical reasoning is the basis for constructing the past from archaeological material.
276
References
Ackerman, Robert E.
1961 Archaeological Investigations into the Prehistory of St. Lawrence Island, Alaska. University of Pennsylvania.
1984 Prehistory of the Asian Eskimo Zone. Handbook of North American Indians. Smithsonian Institution.
1988 Settlements and Sea Mammal Hunting in the Bering-Chukchi Sea Region. Arctic Anthropology 25(1): 52–79.
1998 Early Maritime Traditions in the Bering, Chukchi, and East Siberian Seas. Arctic anthropology 35(1): 247–262.
Alix, Claire
2009 Persistence and Change in Thule Wood Use. In The Northern World AD 900-1400, edited by Herbert D.G. Maschner, Owen K. Mason, and Robert McGhee, pp. 179–205. University of Utah Press, Salt Lake City.
Allison, Penelope M.
1999a Introduction. In The Archaeology of Household Activities, edited by Penelope M. Allison, pp. 1–18. Routledge, London.
Allison, Penelope M. (editor).
1999b The Archaeology of Household Activities. Archaeology. Routledge, London.
Amorosi, Thomas, James Woollett, Sophia Perdikaris, and Thomas McGovern
1996 Regional Zooarchaeology and Global Change: Problems and Potentials. World Archaeology 28(1): 126–157.
Anderson, Douglas D.
1984 Prehistory of North Alaska. Ed. David Damas. Handbook of North American Indians. Smithsonian Institution.
Anderson, Lesleigh, Mark B. Abbott, Bruce P. Finney, and Stephen J. Burns
2005 Regional atmospheric circulation change in the North Pacific during the Holocene inferred from lacustrine carbonate oxygen isotopes, Yukon Territory, Canada. Quaternary Research 64(1): 21–35.
277
Anderson, Shelby L., and Adam K. Freeburg
2013 A High-Resolution Chronology for the Cape Krusenstern Site Complex, Northwest Alaska. Arctic Anthropology 50(1): 49–71.
Andrews, Peter
1995 Experiments in Taphonomy. Journal of Archaeological Science 22(April 1994): 147–153.
1990 Owls, Caves and Fossils: Predation, Preservation and Accumulation of Small Mammal Bones in Caves. University of Chicago Press, Chicago.
AOU
2014 AOU Check-list of North American birds, including 54th supplement. 7th ed. Buteo Books, Arrington, VA.
Arnold, Charles D.
1994a Archaeological Investigations on Richards Island. In Bridges Across Time: The NOGAP Archaeological Project, edited by Jean-Luc Pilon, pp. 86–93. Occassional Paper. Canadian Archaeological Association.
1994b The Importance of Wood in the Early Thule Culture of the Western Canadian Arctic. In Threads of Arctic Prehistory: Papers in Honour of William E. Taylor Jr., edited by David Morrison and Jean-Luc Pilon, pp. 269–280. Mercury Series, Archaeological Survey of Canada Paper. Canadian Museum of Civilization, Hull, Quebec.
Arnold, Charles D, and Elisa J Hart
1992 The Mackenzie Inuit Winter House. Arctic 45(2): 199–200.
Arroyo, A. B. Marín
2009 Assessing What Lies Beneath the Spatial Distribution of a Zooarchaeological Record: the Use of Gis and Spatial Correlations At El Mirón Cave (Spain). Archaeometry 51(3): 506–524.
Ascher, Robert
1961 Analogy in Archaeological Interpretation. Southwestern Journal of Anthropology 17(4): 317–325.
Ashmore, Wendy
2002 “Decisions and Dispositions”: Socializing Spatial Archaeology. American Anthropologist 104(4): 1172–1183.
278
Ashmore, Wendy, and Richard R. Wilk
1988 Houshold community in the Mesoamerican past. In Houshold and Community in the Mesoamerican Past, edited by Richard R. Wilk and Wendy Ashmore, pp. 1–27. University of New Mexico Press, Albuquerque.
Asmussen, Brit
2009 Intentional or incidental thermal modification? Analysing site occupation via burned bone. Journal of Archaeological Science 36(2): 528–536.
Atalay, Sonya, and Christine A. Hastorf
2006 Food, meals, and daily activities: Food habitus at Neolithic Catalhoyuk. American Antiquity 71(2): 283–319.
Aydin, Kerim, and Franz Mueter
2007 The Bering Sea—A dynamic food web perspective. Deep Sea Research Part II: Topical Studies in Oceanography 54: 2501–2525.
Bailey, Geoff
2007 Time perspectives, palimpsests and the archaeology of time. Journal of Anthropological Archaeology 26(2): 198–223.
Balikci, Asen
1970 The Netsilik Eskimo. Natural History Press, Garden City, N.Y.
Bandi, Hans-Georg
1995 Siberian Eskimos as Whalers and Warriors. In Hunting the Largest Animals: Native Whaling in the Western Arctic and Subarctic, edited by Allen P. Mccartney, pp. 165–183. Occassional Paper. The Canadian Circumpolar Institute, Edmonton.
Bandi, Hans Georg, and Reto Blumer
2002 Investigations by Swiss Archaeologists on St. Lawrence Island. In Archaeology in the Bering Strait Region: Research on Two Continents, edited by Don E. Dumond and Reto Blumer, pp. 25–60. University of Oregon Anthropological Papers. Department of Anthropology and Museum of Natural History, Eugene.
Banfield, Alexander W. F.
1974 The mammals of Canada. University. Toronto.
279
Banning, Edward B.
2002 The Archaeologist’s Laboratory: The Analysis of Archaeological Data. Kluwer Academic Publishers, New York.
Barile, Kerri S., and Jamie C. Brandon (editors).
2004 Household Chores and Household Choices: Theorizing the Domestic Sphere in Historical Archaeology. The University of Alabama Press, Tuscaloosa, Alabama.
Bar-Oz, Guy, and Natalie D Munro
2004 Beyond Cautionary Tales: A Multivariate Taphonomic Approach for Resolving Equifinality in Zooarchaeological Studies. Journal of Taphonomy 2(4): 201–222.
Barrett, James H
1993 Bone Weight, Meat Yield Estimates and Cod (Gadus morhua): A preliminary study of the weight method. International Journal of Osteoarchaeology 3: 1–18.
Bartram, Laurence E Jr., Ellen M. Kroll, and Henry T. Bunn
1991 Variability in camp structure and bone food refuse patterning at Kua San camps. In The Interpretation of Archaeological Spatial Patterning, edited by Ellen M. Kroll and Henry T. Bunn, pp. 77–148. Plenum Press, New York.
Beechey, Frederick William
1831 Narrative of a Voyage to the Pacific and Beering’s Strait to Co-Operate with the Polar Expeditions; Performed in His Majesty's Ship “Blossom”...in the Years 1985, 26, 27, 28. Vol. 2 vol. Colburn and Bently, London.
Behrensmeyer, Anna K.
1975 The taphonomy and paleoecology of Plio-Pleistocene vertebrate assemblages east of Lake Rudolf, Kenya. Bulletin of the Museum of Comparative Zoology. Harvard University Press, Cambridge, MA, MA.
1978 Taphonomic and Ecologic Information from Bone Weathering. Paleobiology 4(2): 150– 162.
Betts, Matthew W.
2005 Seven Focal Economies for Six Focal Places: The Development of Economic Diversity in the Western Canadian Arctic. Arctic Anthropology 42(1): 47–87.
2007 The Mackenzie Inuit Whale Bone Industry: Raw Material, Tool Manufacture, Scheduling, and Trade. Arctic 60(2): 129 – 144.
280
Betts, Matthew W., and T. Max Friesen
2004 Quantifying hunter-gatherer intensification: a zooarchaeological case study from Arctic Canada. Journal of Anthropological Archaeology 23(4): 357–384.
2006 Declining foraging returns from an inexhaustible resource? Abundance indices and beluga whaling in the western Canadian Arctic. Journal of Anthropological Archaeology 25(1): 59–81.
2013 Archaeofaunal signatures of specialized bowhead whaling in the Western Canadian Arctic: a regional study. Anthropozoologica 48(1): 53–73.
Bevan, Andrew, and James Conolly
2006 Multiscalar Approaches to Settlement Pattern Analysis. In Confronting Scale in Archaeology: Issues of Theory and Practice, edited by Gary Lock and Brian Leigh Molyneaux, pp. 217–234. Springer, New York.
Bigelow, Nancy H., Owen K. Mason, Caire M. Alix, and John F. Hoffecker
2013 Dating wood and other plant bits at Cape Espenberg: Building a landscape chronology and avoiding pitfalls. In Alaska Anthropological Association 40th Annual Meeting. Anchorage.
Binford, Lewis R.
1967 Smudge pits and hide smoking: The use of analogy in archaeological reasoning. American Antiquity 32(1): 1–12.
1968 Archaeological perspectives. In New Perspectives in Archaeology, edited by Sally R. Binford and Lewis R. Binford, pp. 5–32. Aldine Publishing Company, Chicago.
1972 Directionality in archaeological sequences. In An Archaeological Perspective, edited by Lewis R. Binford, pp. 314–326. Seminar Press, New York.
1977 For Theory Building in Archaeology. Academic Press, New York.
1978a Nunamiut Ethnoarchaeology. Ed. Stuart Struever. Studies in Archaeology. Academic Press, New York.
1978b Dimensional Analysis of Behavior and Site Structure: Learning from an Eskimo Hunting Stand. American Anthropologist 43(3): 330–361.
1980 Willow Smoke and Dogs’ Tails: Hunter-Gatherer Settlement Systems and Archaeological Site Formation. American Antiquity 45(1): 4–20.
1981a Ancient Men and Modern Myths. Academic Press, New York.
1981b Behavioral Archaeology and the “Pompeii Premise.” Journal of Anthropological Research 37(3): 195–208.
281
1984 Butchering, sharing, and the archaeological record. Journal of Anthropological Archaeology 3: 235–257.
Binford, Lewis R., and J. B. Bertram
1977 Bone frequencies--and attritional processes. In For theory building in archaeology, edited by Lewis R. Binford, pp. 77–153. Academic Press, New York.
Bird-David, Nurit
2006 Animistic epistemology: Why do some hunter-gatherers not depict animals? Ethnos 71(1): 33–50.
Birket-Smith, Kaj
1929 The Caribou Eskimos: Material and Social Life and Their Cultural Position. Report of the Fifth Thule Expedition 1921-24: The Danish Expedition to Arctic North America in Charge of Knud Rasmussen, Ph.D. Vol. 5, Part 1. Report of the Fifth Thule Expedition 1921-24. Gyldeddalske Boghandel, Nordisk Forlag, Copenhagen.
1959 The Eskimos. Ed. W E Calvert. Vol. [New enl. Methuen & Co. Ltd., London.
Blanco-González, Antonio, and John Chapman
2014 Revisiting the Chalcolithic site of El Ventorro (Madrid, Spain). Ceramic Re-fitting and Taphonomy. Proceedings of the Prehistoric Society 80: 87–103.
Blanton, Richard E.
1994 Houses and Households: A Comparative Study. Plenum Press, New York.
Blumer, Reto
2002 Radiochronoloigcal Assessment of Neo-Eskimo Occupations on St. Lawrence Island, Alaska. In Archaeology in the Bering Strait Region: Research on Two Continents, edited by Don E. Dumon and Richard L. Bland, pp. 61–106. University of Oregon Anthropological Papers. Department of Anthropology and Museum of Natural History.
Boas, Franz
1888 The Central Eskimo. Smithsonian Institution, Washington.
Bockstoce, John R.
1976 On the Development of Whaling in the Western Thule Culture. Folk 18: 41–45.
1977 Eskimos of Northwest Alaska in the Early Nineteenth Century: Based on the Beechey and Belcher Collections and records compiled during the voyage of H.M.S. Blossom to Northwest Alaska in 1826 and 1827. Ed. T. K. Penniman. Monograph Series. Pitt Rivers Museum, Oxford.
282
1979 The Archaeology of Cape Nome, Alaska. University Museum Monograph. University of Pennsylvania, Philadelphia.
2009 Furs and frontiers in the far north: the contest among native and foreign nations for the Bering Strait fur trade. Yale University Press, New Haven.
Bokonyi, S.
1980 A new method for the determination of the number of individuals in animal bone material. American Journal of Archaeology 74: 291–292.
Borré, Kristen
1994 The healing power of the seal: the meaning of Inuit health practice and belief. Arctic Anthropology 31(1): 1–15.
Bourdieu, Pierre
1977 Outline of a Theory of Practice. Cambridge University Press, Cambridge.
Bradley, Richard, and Elise Fraser
2010 Bronze Age Barrows ont he Heathlands of Southern England: Construction, Forms amd Interpretations. Oxford Journal of Archaeology 29(1): 15–33.
Brain, C. K.
1969 The contribution of Namib desert Hottentots to an understanding of Australopithecine bone accumulations. Scientific Papers of the Namib Desert Research Station 39: 13–22.
1976 Some principles in the interpretation of bone accumulations associated with man. In Human origins: Louis Leakey and the East African evidence, edited by G. L. Isaac and E. R. McCown, pp. 97–116. W. A. Benjamin, Inc., Menlo Park.
1981 The hunters or the hunted? An introduction to African cave taphonomy. University of Chicago Press, Chicago.
Brink, Jack W
2005 Inukshuk: Caribou drive lanes on Southern Victoria Island, Nunavut, Canada. Arctic Anthropology 42(1): 1–28.
Bronshtein, Mikhail M., and Kirill A. Dneprovsky
2002 The northeastern Chukchi Peninsula during the Birnirk and Early Punuk periods. In Archaeology in the Bering Strait Region: Research on Two Continents, edited by Don E. Dumond and Richard L. Bland, pp. 153–165. University of Oregon Anthropological Papers. University of Oregon.
283
Broughton, Jack M., Daniel Mullins, and Tamara Ekker
2007 Avian resource depression or intertaxonomic variation in bone density? A test with San Francisco Bay avifaunas. Journal of Archaeological Science 34: 374–391.
Brumfiel, Elizabeth M
1992 Distinguished Lecture in Archaeology: Breaking and Entering the Ecosystem-Gender, Class, and Faction Steal the Show. American Anthropologist 94(3): 551–567.
Bunn, Henry T., and Ellen M. Kroll
1986 Systematic Butchery by Plio/Pleistocene Hominids at Olduvai Gorge, Tanzania. Current Anthropology 27(5): 431–452.
Burch, Ernest S. Jr.
1972 The Caribou/Wild Reindeer as a Human Resource. American Antiquity 37(3): 339–368.
1980 Traditional Eskimo Societies in Northwest Alaska. Senri Ethnological Studies 4: 253–304.
1984 Kotzebue Sound Eskimo. Ed. David Damas. Handbook of North American Indians. Smithsonian Institution.
1988a Modes of exchange in north-west Alaska. In Hunters and Gatherers: Property, Power and Ideology, edited by Tim Ingold, David Riches, and James Woodburn, pp. 95–109. Berg Publishers, Oxford.
1988b War and trade. In Crossroads of Continents: Cultures of Siberia and Alaska, edited by William W. Fitzhugh and Aron Crowell, pp. 227–240. Smithsonian Institution, Washington, DC.
1998a The Inupiaq Eskimo nations of northwest Alaska. University of Alaska Press, Fairbanks.
1998b Boundaries and Borders in Early Contact North-Central Alaska. Arctic Anthropology 35(2): 19–48.
2005 Alliance and Conflict: The world system of the Inupiaq Eskimos. University of Calgary Press, Calgary.
2006 Social Life in Northwest Alaska: The Structure of Inupiaq Eskimo Nations. University of Alaska Press, Fairbanks.
2010 The method of ethnographic reconstruction. Alaska Journal of Anthropology 8(2): 123– 140.
2012 Caribou herds of northwest Alaska, 1850-2000. University of Alaska Press, Fairbanks.
2013 Inupiaq Ethnohistory: Selected Essays. Ed. Erica Hill. University of Alaska Press, Fairbanks.
284
Butler, Virginia L.
1993 Natural versus Cultural Salmonid Remains: Origin of the Dalles Roadcut bones, Columbia River, Oregon, U.S.A. Journal of Archaeological Science 20: 1–24.
Butler, Virginia L, and James C Chatters
1994 The role of bone density in structuring prehistoric salmon bone assemblages. Journal of Archaeological Science 21: 413–424.
Cannon, Michael D
2001 Archaeofaunal Relative Abundance, Sample Size, and Statistical Methods. Journal of Archaeological Science 28(2): 185–195.
Carballo, David M., and Thomas Pluckhahn
2007 Transportation corridors and political evolution in highland Mesoamerica: Settlement analyses incorporating GIS for northern Tlaxcala, Mexico. Journal of Anthropological Archaeology 26(4): 607–629.
Carr, Christopher
1984 The Nature of Organization of Intrasite Archaeological Records and Spatial Analytic Approaches to Their Investigation. Advances in Archaeological Method and Theory 7: 103–222.
Chang, K. C.
1967 Major Aspects of Archaeology and Ethnology. Current Anthropology 8(3): 227–243.
Chaplin, R. E.
1971 The Study of Animal Bones from Archaeological Sites. Seminar Press, London.
Chapman, John, and Bisserka Gaydarska
2007 Parts and wholes: Fragmentation in preshistoric context. Oxbow Books, Oxford.
Chard, Chester S.
1955 Eskimo Archaeology in Siberia. Southwestern Journal of Anthropology 11(2): 150–177.
Church, Robert R, and R Lee Lyman
2003 Small fragments make small differences in efficiency when rendering grease from fractured artiodactyl bones by boiling. Journal of Archaeological Science 30(8): 1077–1084.
285
Clark, J. Desmond
1968 Studies of hunter-gatherers as an aid to the interpretation of prehistoric societies. In Man the Hunter, edited by Richard B. Lee and Irven DeVore, pp. 276–280. Aldine Publishing Company, Chicago.
Clark, Jamie L., and Bertrand Ligouis
2010 Burned bone in the Howieson’s Poort and post-Howieson’s Poort Middle Stone Age deposits at Sibudu (South Africa): behavioral and taphonomic implications. Journal of Archaeological Science 37(10): 2650–2661.
Clark, Martyn P., Mark C. Serreze, and Roger G. Barry
1996 Characteristics of Arctic Ocean climate based on COADS data, 1980-1993. Geophysical Research Letters 23(15): 1953–1956.
Clark, Tiffany C.
1998 Assessing room function using unmodified faunal bone: A case study from East-Central Arizona. Kiva 64(1): 886–898.
Clarke, David L. (editor).
1977 Spatial archaeology. Academic Press, London.
Collins, Henry B. Jr.
1929 The Ancient Eskimo Culture of Northwestern Alaska. Explorations and Field-Work of the Smithsonian Institution: 141–150.
1937 Archeology of St. Lawrence Island, Alaska. Smithsonian Miscellaneous Collections. Vol. 96, No. 1. Smithsonian Institution, Washington.
1940 Outline of Eskimo Prehistory. In Essays in Historical Anthropology of North America: Published in Honor of John R. Swanton, 100:pp. 533–592. Smithsonian Miscellaneous Collections. Smithsonian Institution, Washington, DC.
1964 The Arctic and Subarctic. In Prehistoric Man in the New World, edited by Jesse D. Jennings and Edward Norbeck, pp. 85–114. University of Chicago Press, Chicago.
Coltrain, Joan Brenner, M. Geoffrey Hayes, and Dennis H. O’Rourke
2004 Sealing, whaling and caribou: the skeletal isotope chemistry of Eastern Arctic foragers. Journal of Archaeological Science 31(1): 39–57.
286
Conard, Nicholas J, Steven J Walker, and Andrew W Kandel
2008 How heating and cooling and wetting and drying can destroy dense faunal elements and lead to differential preservation. Palaeogeography, Palaeoclimatology, Palaeoecology 266(3-4): 236–245.
Cook, J.
1986 The application of scanning electron microscopy to taphonomic and archaeological problems. In Studies in the Upper Paleolithic of Britian and northwest Europe, edited by D. A. Roe, pp. 143–163. British Archaeological Reports International Series.
Costamagno, Sandrine, Christophe Griggo, and Vincent Mourre
1999 Approche experimentale d’un probleme taphonomique: Utilisation de combustible osseux au Paleolithique. Prehistoire Europeenne 13: 167–194.
Costamagno, Sandrine, Isabelle Thery-Parisot, Jean-Philip Brugal, and Raphaele Guibert
2005 Taphonomic consequences of the use of bones as fuel. Experimental data and archaeological applications. In Biosphere to Lithosphere: New Studies in Vertebrate Taphonomy, edited by Terry O’Connor, pp. 51–62. Proceedings of the 9th Conferene of the International Council of Archaeozoology, Durham, August 2002. Oxbow Books.
Costamagno, Sandrine, Isabelle Thery-Parisot, Delphine Kuntz, François Bon, and Romain Mensan
2010 Taphonomic Impact of Prolonged Combustion on Bones used as Fuel. In The Taphonomy of Burned Organic Residues and Combustion Features in Archaeological Contexts, edited by Isabelle Thery-Parisot, L Chabal, and Sandrine Costamagno, 2:pp. 169–183. Proceedings of the round table, Valbonne, May 27-29 2008. P@lethnologie 2.
Coupland, Gary, Terence Clark, and Amanada Palmer
2009 Hierarchy, communalism, and the spatial order of Northwest Coast plank houses: A comparative study. American Antiquity 74(1): 77–106.
Cowgill, George L., J. H. Altschul, and R. S. Sload
1984 Spatial analysis of Teotihuacan: a Mesoamerican metropolis. In Intrasite Spatial Analysis in Archaeology, edited by H. Hietala, pp. 154–195. Cambridge University Press, Cambridge.
Crawford, Laura J
2012 Thule Plant and Driftwood Use at Cape Espenberg, Alaska. University of Alaska Fairbanks.
287
Crockford, Susan J., and S. Gay Frederick
2011 Neoglacial Sea Ice and Life History Flexibility in Ringed and Fur Seals. In Human Impacts on Seals, Sea Lions, and Sea Otters: Integrating Archaeology and Ecology in the Northeast Pacific, edited by Todd J. Braje and Torben C. Rick, pp. 65–91. University of California Press, Berkeley.
Csonka, Yvon
2003 Ekven-Prehistoric Whale Hunters’ Settlement on the Asian Shore of Bering Strait. In Indigenous Ways to the Present: Native Whaling in the Western Arctic, edited by Allen P. McCartney, pp. 109–136. Studies in Whaling. Canadian Circumpolar Institute, Edmonton.
Curtis, Edward C.
1930 The North American Indian, vol. 20. New York.
Cutting, Marion
2006 More than one way to study a building: Approaches to prehistorica household and settlement space. Oxford Journal of Archaeology 25(3): 225–246.
D’Arrigo, Rosanne, Erika Mashig, David Frank, Rob Wilson, and Gordon Jacoby
2005 Temperature variability over the past millennium inferred from Northwestern Alaska tree rings. Climate Dynamics 24(2-3): 227–236.
Damas, David
1969 Characteristics of Central Eskimo band structure. In Contributions to Anthropology: Ecoloigcal Essays. Bulletin 230. Proceedings of the Conference on Cultural Ecology, Ottawa, August 3-6, 1966, edited by David Damas, pp. 116–141. Anthropological Series. National Museums of Canada, Ottawa.
1972 Central Eskimo Systems of Food Sharing. Ethnology 11(3): 220–240.
Darwent, Christyann M., and Jeremy C. Foin
2010 Zooarchaeological Analysis of a Late Dorset and an Early Thule Dwelling at Cape Grinnell, Northwest Greenland. Geografisk Tiddsskrift-Danish Journal of Geography 110(2): 315–336.
Darwent, John, Owen K. Mason, John F. Hoffecker, and Christyann M. Darwent
2013 1,000 Years of House Change at Cape Espenberg, Alaska: A Case Study in Horizontal Stratigraphy. American Antiquity 78(3): 433–455.
288
Daviau, M.
1993 Houses and their furnishing in Bronze Age Palestine: Domestic activity areas and artefact distributions in the Middle and Late Bronze Ages. Journal for the Study of the Old Testament. Sheffield Academic Press, Sheffield.
David, Nicholas, and Carol Kramer (editors).
2001a Ethnoarchaeology in Action. Cambridge University Press, Cambridge.
David, Nicholas, and Carol Kramer
2001b Theorizing ethnoarchaeology and analogy. In Ethnoarchaeology in Action, pp. 33–62. Cambridge University Press, Cambridge.
Dawson, Peter C.
1995 “Unsympathetic Users”: An Ethnoarchaeological Examination of Inuit Responses to the Changing Nature of the Built Environment. Arctic 48(1): 71 – 80.
1997 Variability in traditional and non-traditional Inuit architecture, AD. 1000 to present. The University of Calgary.
2001 Interpreting Variability in Thule Inuit Architecture: A Case Study from the Canadian High Arctic. American Antiquity 66(3): 453–470.
2006 Seeing like an Inuit family: The relationship between house form and culture in northern Canada. Etudes/Inuit/Studies 30(2): 113–135.
Dawson, Peter C., Richard Levy, Don Gardner, and Matthew Walls
2007 Simulating the behaviour of light inside Arctic dwellings: implications for assessing the role of vision in task performance. World Archaeology 39(1): 17–35.
Dekin, Albert A. Jr., and John N. Kilmarx
1990 Stratigraphy, Depositional Processes, and Archaeological Integrity at Mound 44. In The Utqiagvik Excavations, Volume 2: Additional Reports of the 1982 Investigations by the Utqiagvik Archaeology Project Barrow, Alaska, edited by Edwin S. Jr. Hall and Lynne Fullerton, pp. 22– 37. The North Slope Borough Commission on Iñupiat History, Language and Culture, Barrow.
Denys, C
2002 Taphonomy and experimentation. Archaeometry 44(3): 469–484.
Diab, Mark C
1998 Economic Utility of the Ringed Seal (Phoca hispida): Implications for Arctic Archaeology. Journal of Archaeological Science 25(1): 1–26.
289
Dneprovsky, Kirill A.
2002 Ekven House H-18: A Birnirk- and Early Punuk-Period Site in Chukotka. In Archaeology in the Bering Strait Region: Research on Two Continents, edited by Don E. Dumond and Richard L. Bland, pp. 166–206. University of Oregon Anthropological Papers. University of Oregon.
Drennan, Robert D
2009 Statistics for Archaeologists: A Commonsense Approach. Second Edi. Springer, New York.
Driver, Jonathan C.
2011 Identification, Classification and Zooarchaeology. Ethnobiology Letters 2: 19–39.
Dumond, Don E.
1977 The Eskimos and Aleuts. Thames and Hudson, London.
1984 Prehistory of the Bering Sea Region. Handbook of North American Indians. Smithsonian Institution.
1987 The Eskimos and Aleuts: Revised Edition. Thames and Hudson, London.
2002 The Legacy of Henry B. Collins. In Archaeology in the Bering Strait Region: Research on Two Continents, edited by Don E. Dumond and Richard L. Bland, pp. 9–24. University of Oregon Anthropological Papers. Department of Anthropology and Museum of Natural History, Eugene.
2009 The “Arctic Maritime” Expansion: A View from the South. In The Northern World AD 900-1400, edited by Herbert Maschner, Owen Mason, and Robert McGhee, pp. 58–75. The University of Utah Press, Salt Lake City.
Dunnell, Robert C.
1978 Archaeological potential of anthropological and scientific models of function. In Archaeological Essays in Honor of Irving B. Rouse, edited by Robert C. Dunnell and Edwin S. Hall, pp. 41–73. Mouton, The Hague.
1992 Archaeology and evolutionary theory. In Quandaries and Quests, edited by LuAnn Wandsnider, pp. 209–224. Southern Illinois University, Carbondale.
Eldridge, Kelly Anne
2012 Archaeofaunal representation of late Western Thule regionalization: Insights from the Snake River Sandspit site in Nome, Alaska. University of Alaska Anchorage.
290
Enloe, James G, and F. David
1992 Food sharing in the Paleolithic: Carcass refitting at Pincevent. In Piecing Together the Past: Applications of Refitting Studies in Archaeology, edited by J. L. Hofman and James G Enloe, pp. 296–299. International Series. British Archaeological Reports, Oxford.
Enloe, James G, Francine David, and Timothy S Hare
1994 Patterns of faunal processing at Section 27 of Pincevent: The use of spatial analysis and ethnoarchaeological data in the interpretation of archaeological site structure. Journal of Anthropological Archaeology 13: 105–124.
Fair, Susan W., James Creech, Gideon K. Barr, Sr., and Edgar Ningeulook
1996 The Hope and Promise of Ublasaun. In Ublasaun: First Light, edited by Jeanne Schaaf, pp. 62-93. National Park Service, Anchorage, AK.
Faith, J Tyler, Curtis W Marean, and Anna K Behrensmeyer
2007 Carnivore competition, bone destruction, and bone density. Journal of Archaeological Science 34(12): 2025–2034.
Faith, J. Tyler, and Adam D. Gordon
2007 Skeletal element abundances in archaeofaunal assemblages: economic utility, sample size, and assessment of carcass transport strategies. Journal of Archaeological Science 34(6): 872– 882.
Falconer, Steven E
1995 Rural Responses to Early Urbanistn: Bronze Age Household and Village Economy at Tell el-Hayyat, Jordan. Journal of Field Archaeology 22: 399–419.
Fenton, William N.
1962 Ethnohistory and Its Problems. Ethnohistory 9(1): 1-23.
Fieller, N. R. J., and A. Turner
1982 Number estimation in vertebrate samples. Journal of Archaeological Science 9: 49–62.
Fienup-Riordan, Ann
1990 Eskimo Essays: Yup’ik Lives and How We See Them. Rutgers University Press, New Brunswick and London.
Flannery, Kent V.
1976 The early Mesoamerican village. Academic Press, New York.
291
Foote, Don Charles
1965 Explorationand Resource Utilization in Northwestern Arctic Alaska Before 1855. McGill University.
Ford, James A.
1959 Eskimo prehistory in the vicinity of Point Barrow, Alaska. Anthropological Papers of the American Museum of Natural History, New York.
Franklin, M.
2004 Foreward. In Household Chores and Household Choices: Theorizing the Domestic Sphere in Historical Archaeology, edited by K. S. Barile and J. C. Brandon, pp. xiii–xiv. University of Alabama Press, Tuscaloosa.
Friesen, T. Max
1995 “Periphery” as Centre: Long-Term Patterns of Intersocietal Interaction on Herschel Island, Northwern Yukon Territory. McGill University.
1999 Resource Structure, Scalar Stress, and the Development of Inuit Social Organization. World Archaeology 31(1): 21–37.
2000 The Role of Social Factors in Dorset-Thule Interaction. In Arctic Identities and Culture Contacts, edited by Martin Appelt, Maria Hinnerson Berglund, and Hans Christian Gulløv, pp. 206–220. Danish Polar Center, Copenhagen.
2001 A zooarchaeological signature for meat storage: re-thinking the drying utility index. American Antiquity 66(2): 315–331.
2012a Alaskan Analogues and Eastern Uncertainties: Reconstructing Thule Inuit Interaction Networks in the Eastern North American Arctic. In Networks, Interaction and Emerging Identities in Fennoscandia and Beyond: Papers from teh conference held in Tromso, Norway, October 13-16 2009, edited by Charlotte Damm and Janne Saarikivi, pp. 3–26. Suomalais- Urilaisen Seuran Toimituksia=Memoires de la Societe Finno-Ougrienne. Helsinki.
2012b The Importance of Reading Ernest: Applying Burch’s Study of Interregional Interaction to Inuvialuit Ethnohistory. Arctic Anthropology 49(2): 29–40.
2013a When Worlds Collide: Hunter-Gatherer World-System Change in the 19th Century Canadian Arctic. University of Arizona Press.
2013b North America: Paleoeskimo and Inuit archaeology. In The Encylopedia of Global Human Migration, edited by Immanuel Ness, pp. 1–8. Blackwell Publishing.
2013c The impact of weapon technology on caribou drive system variability in the prehistoric Canadian Arctic. Quaternary International 297: 13–23.
292
Friesen, T. Max, and Charles D. Arnold
1995a Zooarchaeology of a Focal Resource: Dietary Importance of Beluga Whales to the Precontact Mackenzie Inuit. Arctic 48(1): 22–30.
1995b Prehistoric Beluga Whale Hunting at Gupuk, Mackenzie Delta, Northwest Territories, Canada. In Hunting the Largest Animals: Native Whaling in the Western Arctic and Subarctic, edited by Allen P. McCartney, pp. 109–125. Occassional Publication. The Canadian Circumpolar Institute, Edmonton.
2008 The timing of the Thule migration: new dates from the Western Canadian Arctic. American Antiquity 73(3): 527–538.
Friesen, T. Max, and Matthew W. Betts
2006 Archaeofaunas and Architecture: Zooarchaeological Variability in an Inuit Semi- Subterranean House, Arctic Canada. In Integrating Zooarchaeology: Proceedings of the 9th Conference of the International Council of Archaeozoology, Durham, August 2002, edited by Mark Maltby, pp. 65–76. Oxbow Books, Oxford.
Friesen, T. Max, and David Morrison
2002 Regional variability in Mackenzie Inuit beluga whale use. International Journal of Osteoarchaeology 12(1): 23–33.
Friesen, T. Max, and Andrew Stewart
2013 To freeze or to dry: Seasonal variability in caribou processing and storage in the barrenlands of Northern Canada. Anthropozoologica 48(1): 89–109.
Frink, Lisa M.
2007 Storage and Status in Precolonial and Colonial Coastal Western Alaska. Current Anthropology 48(3): 349–374.
Froese, R., and D. Pauly
2014 FishBase.
Gautier, A.
1993 Trace fossils in archaeozoology. Journal of Archaeological Science 20: 511–523.
Gerlach, S. Craig
1996 Historical archaeology and the early twentieth century reindeer herding frontier on the northern Seward Peninsula. In Ublasaun: First Light, edited by Jeanne Schaaf, pp. 94-109. National Park Service, Anchorage, AK.
293
Gerlach, S. Craig, and Owen K Mason
1992 Calibrated radiocarbon dates and cultural interaction in the Western Arctic. Arctic Anthropology 29(1): 54–81.
Giddens, Anthony
1984 The constitution of society: Outline of a theory of structuration. Cambridge.
Giddings, J. Louis
1952 The Arctic Woodland Culture of the Kobuk River. University Museum Monographs. University of Pennsylvania, Philadelphia.
1964 The archaeology of Cape Denbigh. Brown University Press, Providence.
1967 Ancient Men of the Arctic. University of Washington Press, Seattle.
Giddings, J. Louis, and Douglas D. Anderson
1986 Beach Ridge Archaeology of Cape Krusenstern: Eskimo and Pre-Eskimo Settlements around Kotzebue Sound, Alaska. Publications in Archeology. National Park Service, U.S. Department of the Interior, Washington, DC.
Gidney, L.
2000 Economic trends, craft specialization and social status: bone assemblages from Leicester. In Animal Bones, Human Societies, edited by Peter Rowley-Conwy, pp. 170–178. Oxbow Books, Oxford.
Goebel, Ted, Heather L. Smith, Lyndsay DiPietro, Michael R. Waters, Bryan Hockett, Kelly E. Graf, Robert Gal, Sergei B. Slobodin, Robert J. Speakman, Steven G. Driese, and David Rhode
2013 Serpentine Hot Springs, Alaska: results of excavations and implications for the age and significance of northern fluted points. Journal of Archaeological Science 40(12): 4222–4233.
Gotfredsen, Anne Birgitte
2010 Faunal Remains from the Wollaston Forland–Clavering Ø Region, Northeast Greenland– Thule Culture Subsistence in a High Arctic Polynya and Ice-edge Habitat. Geografisk Tidsskrift- Danish Journal of Geography 110(2): 175–200.
Gould, Richard A.
1978 Beyond analogy in ethnoarchaeology. In Explorations in ethnoarchaeology, edited by Richard A. Gould, pp. 249–293. University of New Mexico Press, Albuquerque.
1980 Living archaeology. Cambridge University Press, New York.
294
Gould, Richard A., and Patty Jo Watson
1982 A Dialogue on the Meaning Ethnoarchaeological and Use of Analogy Reasoning. Journal of Anthropological Archaeology 1: 355–381.
Grayson, Donald K.
1973 On the methodology of faunal analysis. American Antiquity 39: 432–439.
1978 Minimum Numbers and Sample Size in Vertebrate Faunal Analysis. American Antiquity 43(1): 53–65.
1979 On the quantification of vertebrate archaeofaunas. In Advances in Archaeological Method and Theory, edited by Michael B. Schiffer, pp. 199–237. Academic Press, New York.
1981 The effects of sample size on some derived measures in vertebrate faunal analysis. Journal of Archaeological Science 8: 77–88.
1984 Quantitative Zooarchaeology. Academic Press, New York.
Grayson, Donald K, and Francoise Delpech
1998 Changing Diet Breadth in the Early Upper Palaeolithic of Southwestern France. Journal of Archaeological Science 25(11): 1119–1129.
Grier, Colin, and James M. Savelle
1994 Intrasite Spatial Patterning and Thule Eskimo Social Organization. Arctic Anthropology 31(2): 95–107.
Grønnow, Bjarne
1986 Recent Archaeological Investigations of West Greenland Caribou Hunting. Arctic Anthropology 23(1&2): 57–80.
Grønnow, Bjarne, Hans Christian Gulløv, Bjarne Holm Jakobsen, Anne Birgitte Gotfredsen, Laura Hauch Kauffmann, Aart Kroon, Jørn Bjarke, Torp Pedersen, and Mikkel Sørensen
2011 At the edge: High Arctic Walrus hunters during the Little Ice Age. Antiquity 85: 960–977.
Groot, C., and L. Margolis (editors).
1991 Pacific salmon life histories. University of British Columbia Press, Vancouver.
Guadelli, Jean-Luc
2008 La gélifraction des restes fauniques. Expérimentation et transfert au fossile. Annales de Paléontologie 94(3): 121–165.
295
Gubser, Nicholas J.
1965 The Nunamiut Eskimos: Hunters of Caribou. Yale University Press, New Haven, CT.
Gullason, Lynda
2005 Canadian Arctic Historical Archaeology in Review. Revista de Arqueologia Americana 23: 7–93.
Gulløv, Hans Christian, and Robert McGhee
2006 Did Bering Strait People Initiate the Thule Migration? Alaska Journal of Anthropology 4(1-2): 54–63.
Habu, Junko, and James M Savelle
1994 Construction, Use, and Abandonment of a Thule Whale Bone House, Somerset Island, Arctic Canada. The Quaternary Research 33(1): 1–18.
Hachem, Lamys
2000 New observations on the Bandkeramik house and social organization. Antiquity 74: 308– 312.
Hall, Edwin S.
1984 A Clear and Present Danger: The Use of Ethnohistoric Data for Interpreting Mound 44 at the Utqiagvik Site. Arctic Anthropology 21(1): 135–139.
Hall, Edwin S. Jr., and Lynne Fullerton (editors).
1990a The Utqiagvik Excavations: Volume 3. Excavation of a Prehistoric Catastrophe: A Preserved Household from the Utqiagvik Village, Barrow, Alaska. The North Slope Borough Commission on Iñupiat History, Language and Culture, Barrow.
Hall, Edwin S. Jr., and Lynne Fullerton (editors).
1990b The Utqiagvik Excavations: Volume 2. Additional Reports of the 1982 Investigations by the Utqiagvik Archaeology Project Barrow, Alaska. The North Slope Borough Commission on Iñupiat History, Language and Culture, Barrow.
Hammer, O., D. A. T. Harper, and P. D. Ryan
2001 PAST: Paleontological STatistics Software Package for Education and Data Analysis. Palaeontologia Electronica 4(1): 9pp.
Harritt, Roger K.
1994 Eskimo Prehistory on the Seward Peninsula, Alaska. Anchorage.
296
1995 The Development and Spread of the Whale Hunting Complex in Bering Strait: Retrospective and Prospects. In Hunting the Largest Animals: Native Whaling in the Western Arctic and Subarctic, pp. 33–50. Occasional Papers. The Canadian Circumpolar Institute, Edmonton.
2004 A Preliminary Reevaluation of the Punuk-Thule Interface at Wales, Alaska. Arctic Anthropology 41(2): 163–176.
2010 Variations of Late Prehistoric Houses in Coastal Northwest Alaska: A View from Wales. Arctic Anthropology 47(1): 57–70.
Harry, Karen, and Liam Frink
2009 The Arctic Cooking Pot: Why Was It Adopted? American Anthropologist 111(3): 330– 343.
Hayes, M. Geoffrey, Joan Brenner Coltrain, and Dennis H. O’Rourke
2005 Molecular Archaeology of the Dorset, Thule, and Sadlermiut: Ancestor-Descendant Relationships in Eastern North American Arctic Prehistory. In Contributions to the Study of the Dorset Palaeo-Eskimos, edited by Patricia D. Sutherland, pp. 11–32. Archaeolog. Mercury Series, Archaeological Paper. Canadian Museum of Civilization, Hull, Quebec.
Haynes, Gary
1983 Frequencies of Spiral and Green-Bone Fractures on Ungulate Limb Bones in Modern Surface Assemblages. American Antiquity 48(1): 102–114.
Hegmon, Michelle
2003 Setting Theoretical Egos Aside: Issues and Theory in North American Archaeology. American Antiquity 68(2): 213–243.
Helgason, Agnar, Gísli Pálsson, Henning Sloth Pedersen, Emily Angulalik, Ellen Dröfn Gunnarsdóttir, Bryndís Yngvadóttir, and Kári Stefánsson
2006 mtDNA variation in Inuit populations of Greenland and Canada: migration history and population structure. American Journal of Physical Anthropology 130(1): 1–12.
Helmer, James W.
1992 Prehistoric Site Location Strategies in the North Devon Lowlands, High Arctic Canada. Journal of Field Archaeology 19(3): 291–313.
Hendon, Julia A.
1996 Archaeological Approaches to the Organization of Domestic Labor: Household Practice and Domestic Relations. Annual Review of Anthropology 25(1): 45–61.
297
Henshaw, Anne S.
2003 Polynyas and Ice Edge Habitats in Cultural Context: Archaeological Perspectives from Southeast Baffin Island. Arctic 56(1): 1 – 13.
Hietala, Harold J. (editor).
1984 Intrasite Spatial Analysis in Archaeology. Cambridge University Press, Cambridge.
Hill, Erica
1998 Gender-Informed Archaeology: The Priority of Definition, the Use of Analogy, and the Multivariate Approach. Journal of Archaeological Method and Theory 5(1): 99–128.
2011 Animals as Agents: Hunting Ritual and Relational Ontologies in Prehistoric Alaska and Chukotka. Cambridge Archaeological Journal 21(03): 407–426.
2013 Archaeology and Animal Persons: Toward a Prehistory of Human-Animal Relations. Environment and Society: Advances in Research 4(1): 117–136.
Hill, James N.
1968 Broken K Pueblo: Patterns of Form and Function. In New Perspectives in Archaeology, edited by Sally R. Binford and Lewis R. Binford, pp. 103–142. Aldine, Chicago.
Hilton, Michael R.
2003 Quantifying Postdepositional Redistribution of the Archaeological Record Produced by Freeze–Thaw and Other Mechanisms: An Experimental Approach. Journal of Archaeological Method and Theory 10(3): 165–202.
Hirth, Kenneth G.
1993 The household as an analytical unit: Problems in method and theory. In Prehispanic Domestic Units in Western Mesoamerica: Studies of the Household, Compoud, and Residence, edited by R. Santley and Kenneth G. Hirth, pp. 21–36. CRC Press, Boca Raton, FL.
Hockett, Bryan Scott
1998 Sociopolitical Meaning of Faunal Remains from Baker Village. American Antiquity 63(2): 289–302.
Hodder, Ian (editor).
1982 Symbolic and Structural Archaeology. Cambridge University Press, Cambridge.
Hodder, Ian
1987 The Archaeology of Contextual Meanings. Cambridge University Press, Cambridge.
298
Hodder, Ian, and Scott Hutson
2003 Reading the past: current approaches to interpretation in archaeology. 3rd ed. Cambridge University Press, Cambridge.
Hodder, Ian, and Clive Orton (editors).
1976 Spatial Analysis in Archaeology. Cambridge University Press, Cambridge.
Hodgetts, Lisa Maye
1999 Animal bones and human society in the Late Younger Stone Age of Arctic Norway: Volume 2 of 2. University of Durham.
Hoffecker, John, Owen K. Mason, Scott A. Elias, Diane Hnason, Claire Alix, Georgeanne L. Reynolds, and Karlene Leeper
2012 Uivvaq: A Stratified Inupiaq Occupation at Cape Lisburne, Northwest Alaska. Alaska Journal of Anthropology 10(1&2): 143–172.
Hollinger, R. Eric, Stephen Ousley, and Charles Utermohle
2009 The Thule Migration: A New Look at the Archaeology and Biology of the Point Barrow Region Populations. In The Northern World AD 900-1400, edited by Herbert D. G. Maschner, Owen K. Mason, and Robert McGhee, pp. 131–154. The University of Utah Press, Salt Lake City.
Hollowell, Julie
2006 Moral arguments on subsistence digging. In The Ethics of Archaeology: Philosophical Perspectives on Archaeological Practice, edited by Chris Scarre and Geoffrey Scarre, pp. 69–93. Cambridge University Press, Cambridge.
Holtved, Erik
1944 Archaeological investigations in the Thule District, I-II. Meddelelser om Gronland 141(1- 2).
Howse, Lesley
2008 Late Dorset Caribou Hunters: Zooarchaeology of the Bell Site, Victoria Island. Arctic Anthropology 45(1): 22–40.
Hrdlicka, Ales
1930 Anthropological Survey in Alaska. In 46th Annual Report of the Bureau of American Ethnology (1928-1929), pp. 19–374. U.S. Government Printing Office, Washington, DC.
299
Hu, F S, E Ito, T a Brown, B B Curry, and D R Engstrom
2001 Pronounced climatic variations in Alaska during the last two millennia. Proceedings of the National Academy of Sciences of the United States of America 98(19): 10552–6.
Ingold, Tim
2000 The Perception of the Environment: Essays on livelihood, dwelling and skill. Routledge, London.
Ingraham, W. James Jr.
1981 Shelf Environment. In The Eastern Bering Sea Shelf: Oceanography and Resources. Volume One, edited by Donald W. Hood and John A. Calder, pp. 455–470. Office of Marine Population Assessment of the National Oceanic and Atmospheric Administration, Seattle.
Irving, William N.
1962 1961 Field Work in the Western Brooks Range, Alaska: Preliminary Report. Arctic Anthropology 1(1): 76–83.
Jackson, H. Edwin
1989 The Trouble with Transformations: Effects of Sample Size and Sample Composition on Meat Weight Estimates Based on Skeletal Mass Allometry. Journal of Archaeological Science 16: 601–610.
Jacobs, John D, and George III Sabo
1978 Environments and Adaptations of the Thule Culture on the Davis Strait Coast of Baffin Island. Arctic and Alpine Research 10(3): 595–615.
Jacobs, John D, and Douglas R Stenton
1985 Environment, Resources, and Prehistoric Settlement in Upper Frobisher Bay, Baffin Island. Arctic Anthropology 22(2): 59–76.
Jenness, Diamond
1922 The Life of the Copper Eskimos. A Report of the Canadian Arctic Expedition 1913-18. Vol. 11, Part A. A Report of the Canadian Arctic Expedition 1913-18. F.A. Acland, Ottawa.
1928 Archaeolgoical investigations in Bering Strait, 1926. National Museum of Canad Bulletin 50: 71–80.
Jenness, Stuart E.
1990 Diamond Jenness’ s archaeological investigations on Barter Island, Alaska. Polar Record 26(157): 91–102.
300
Jensen, Anne M.
2009 Nuvuk Point Barrow, Alaska: The Thule Cemetery and Ipiutak Occupation. Field Methods. Bryn Mawr College.
2012 The Material Culture of Iñupiat Whaling: An Ethnographic and Ethnohistorical Perspective. Arctic Anthropology 49(2): 143–161.
Jones, Anore
2010 Plants that we eat: Nauriat niginaqtuat: from the traditional wisdom of the Iñupiat elders of northwest Alaska. University of Alaska Press, Fairbanks.
Jordan, James W.
2009 Arctic Climate and Landscpae ca. AD 800-1400. In The Northern World AD 900-1400, edited by Herbert Maschner, Owen Mason, and Robert McGhee, pp. 7–29. The University of Utah Press, Salt Lake City.
Junker, L. L., K. Mudar, and Schwaller M.
1994 Social stratification, household wealth, and competitive feasting in 15th/16th century Philippine chiefdomes. Research in Economic Anthropology 15: 307–358.
Kankaanpaa, Jarmo Kalevi
1996 Thule Subsistence. Brown University.
Kent, Susan
1981 The Dog: An Archaeologist’s Best Friend or Worst Enemy--the Spatial Distribution of Faunal Remains. Journal of Field Archaeology 8(3): 367–372.
1984 Analysing Activity Areas: An Ethnoarchaeological Study of the Use of Space. University of New Mexico Press, Albuquerque.
Kent, Susan (editor).
1987 Method and Theory for Activity Area Research: An Ethnoarchaeological Approach. Columbia University Press, New York.
1990 Domestic Architecture and the Use of Space: An Interdisciplinary Cross-cultural Study. University of New Mexico, Albuquerque.
Kessel, Brina
1989 Birds of the Seward Peninsula, Alaska: Their Biogeography, Seasonality, and Natural History. Universtiy of Alaska Press.
301
King, Stacie M
2008 The spatial organization of food sharing in Early Postclassic households: an application of soil chemistry in Ancient Oaxaca, Mexico. Journal of Archaeological Science 35(5): 1224–1239.
Klein, Richard G, Kathryn Cruz-Uribe, and Richard G Milo
1999 Skeletal Part Representation in Archaeofaunas: Comments on “Explaining the ‘Klasies Pattern’: Kua Ethnoarchaeology, the Die Kelders Middle Stone Age Archaeofauna, Long Bone Fragmentation and Carnivore Ravaging” by Bartram & Marean. Journal of Archaeological Science 26(9): 1225–1234.
Klein, Richard G., and Kathryn Cruz-Uribe
1984 The Analysis of Animal Bones from Archaeological Sites. University of Chicago Press, Chicago.
Kleivan, Inge
1984 West Greenland before 1950. In Handbook of North American Indians, edited by David Damas, pp. 595–621. Smithsonian Institution, Washington.
Knudson, Kelly J, Lisa Frink, Brian W Hoffman, and T Douglas Price
2004 Chemical characterization of Arctic soils: activity area analysis in contemporary Yup’ik fish camps using ICP-AES. Journal of Archaeological Science 31(4): 443–456.
Knudson, Kelly J., and Liam Frink
2010a Ethnoarchaeological analysis of Arctic fish processing: chemical characterization of soils on Nelson Island, Alaska. Journal of Archaeological Science 37(4): 769–783.
2010b Soil chemical signatures of a historic sod house: activity area analysis of an arctic semisubterranean structure on Nelson Island, Alaska. Archaeological and Anthropological Sciences 2(4): 265–282.
2011 Nitrogen isotope analysis in the Arctic: Identifying fish processing and marine resource use through ethnoarchaeological soil analysis on Nelson Island, Alaska. Alaska Journal of Anthropology 9(2): 17–54.
Von Kotzebue, Otto
1821 A voyage of discovery, into the south sea and Beering’s [sic] straits: for the purpose of exploring a north-east passage, undertaken in the years 1815-1818, at the expense of His Highness the Chancellor of the Empire, Count Romanzoff in the ship Rurik und. Longman, Hurst, Rees, Orme, and Brown, London.
Kovacik, Joseph J.
1998 Collective memory and pueblo space. Norwegian Archaeological Review 31(2): 141–152.
302
Kramer, Carol (editor).
1979 Ethnoarchaeology: Implications of Ethnography for Archaeology. Columbia University Press, New York.
Kramer, Carol
1982 Village Ethnoarchaeology: Rural Iran in Archaeological Perspectives. Academic Press, New York.
Kroeber, Alfred Louis
1900 The Eskimo of Smith Sound. Science (New York, N.Y.). Vol. 11. Bulletin American Museum of Natural History. Knickerbocker Press, New York, April 13.
1916 Zuñi potsherds. Anthropological Papers of the American Museum of Natural History XVIII(I): 1–38.
Kroll, Ellen M., and Glynn L.I. Isaac
1984 Configurations of artifacts and bones at early Pleistocene sites in East Africa. In Intrasite Spatial Analysis in Archaeology, edited by Harold J. Hietala, pp. 4–31. Cambridge University Press, Cambridge.
Kroll, Ellen M., and T. Douglas Price (editors).
1991 The Interpretation of Archaeological Spatial Patterning. Plenum Press, New York.
Krupnik, Igor, and Michael A. Chlenov
2009 Distant Lands and Brave Pioneers: Original Thule Migration Revisited. In On the Track of the Thule Culture from Bering Strait to East Greenland: Proceedings of the SILA Conference “The Thule Culture-New Perspectives in Inuit Prehistory” Copenhagen, Oct. 26th-28th, 2006. Papers in Honour of Hans Chirstian Gullov2, edited by Bjarne Grønnow, pp. 11–23. Studies in Archaeology & History. Publications from the National Museum, Copenhagen.
Kuckelman, Kristin A
2010 The depopulation of Sand Canyon Pueblo, a large ancestral pueblo village in southwestern Colorado. American Antiquity 75(3): 497–525.
De Laguna, Frederica
1947 The Prehistory of the Northern North America as Seen from the Yukon. Memoirs of the Society for American Archaeology. The Society for American Archaeology, Menasha, Wisconsin.
303
Lam, Yin M., Xingbin Chen, and O. M. Pearson
1999 Intertaxonomic Variability in Patterns of Bone Density and the Differential Representation of Bovid, Cervid, and Equid Elements in the Archaeological Record. American Antiquity 64(4): 343–362.
Lam, Yin M., and O. M. Pearson
2004 The Fallibility of Bone Density Values and Their Use in Archaeological Analyses. Journal of Taphonomy 2(2): 99–115.
2005 Bone Density Studies and the Interpretaion of the Faunal Record. Evolutionary Anthropology 14: 99–108.
Lam, Yin M., O. M. Pearson, Curtis W. Marean, and Xingbin Chen
2003 Bone density studies in zooarchaeology. Journal of Archaeological Science 30(12): 1701– 1708.
LaMotta, Vincent M., and Michael B. Schiffer
1999 Formation processes of house floor assemblages. In The Archaeology of Household Activities, edited by Penelope M. Allison, pp. 19–29. Routledge, London.
Lantis, Margaret
1938 The Alaskan Whale Cult and its Affinities. American Anthropologist 40: 438–464.
Larsen, Helge
1954 The Position of Ipiutak in Eskimo Culture. American Antiquity 20(1): 74–79.
1968 Trail Creek: Final report on the excavation of two caves on Seward Peninsula, Alaska. Acta Arctica. Copenhagen.
2001 Deering: A men’s house from Seward Peninsula, Alaska. Ed. Martin Appelt. National Museum of Denmark, Copenhagen.
Larsen, Helge, and Froelich G. Rainey
1948 Ipiutak and the Arctic Whale Hunting Culture. Anthropological Papers of the American Museum of Natural History. New York.
Laurence, R.
1994 Roman Pompeii: Space and Society. Routledge, New York.
Lawrence, Denise S., and Setha M. Low
1990 The built environment and spatial form. Annual Review of Anthropology 19: 453–505.
304
Lee, Molly, and Gregory A. Reinhardt
2003 Eskimo Architecture: Dwelling and Structure in the Early Historic Period. University of Alaska Press, Fairbanks.
LeMoine, Genevieve M., and Christyann M. Darwent
2010 The Inglefield Land Archaeology Project: Introduction and Overview. Geografisk Tidsskrift-Danish Journal of Geography 110(2): 279–296.
Leonova, N. B., and E. V. Min’kov
1988 Spatial Analysis of Faunal Remains from Kamennaya Balka II. Journal of Anthropological Archaeology 7: 203–230.
Levy, Richard, and Peter Dawson
2009 Using finite element methods to analyze ancient architecture: an example from the North American Arctic. Journal of Archaeological Science 36(10): 2298–2307.
Lightfoot, Kent G.
1993 Abandonment processes in prehistoric pueblos. In Abandonment of Settlements and Regions: Ethnoarchaeological and Archaeological Approaches, edited by Catherine M. Cameron and S. A. Tomka, pp. 165–177. Cambridge University Press, Cambridge.
Logan, Brad, and Matthew E. Jr. Hill
2000 Spatial Analysis of Small Scale Debris from a Late Prehistoric Site in the Lower Missouri Valley, Kansas. Journal of Field Archaeology 27(3): 241–256.
Longacre, William A.
1970 Archaeology as Anthropology: A Case Study. Anthropological Papers. University of Arizona, Tuscon.
1978 Ethnoarchaeology. Reviews in Anthropology 5: 357–363.
Longacre, William A. (editor).
1991 Ceramic Ethnoarchaeology. The University of Arizona Press, Tuscon.
Longacre, William A., and James M. Skibo (editors).
1994 Kalinga Ethnoarchaeology: Expanding Archaeological Method and Theory. Smithsonian Institution Press, Washington, DC.
Lucas, Gavin
2012 Understanding the archaeological record. Cambridge University Press, Cambridge.
305
Lucier, Charles V., and James W. VanStone
1995 Traditional Beluga Drives of the Iñupiat of Kotzebue Sound, Alaska. Fieldiana Anthropology. Field Museum of Natural History, Chicago.
Lutz, Bruce J.
1973 An Archaeological Karigi at the Site Ungalaqliq of Western Alaska. Arctic Anthropology 10(1): 111–118.
Lyman, R. Lee
1984 Bone density and differential survivorship of fossil classes. Journal of Anthropological Archaeology 3(4): 259–299.
1993 Density-Mediated Attrition of Bone Assemblages: New Insights. In From Bones to Behavior: Ethnoarchaeological and Experimental Contributions to the Interpretation of Faunal Remains, edited by Jean Hudson, pp. 324–341. Centre for Archaeological Investigations. Souther Illinois University Carbondale, Carbondale.
1994 Vertebrate Taphonomy. Cambridge Manuals in Archaeology. Cambridge University Press, Cambridge.
2005 Analyzing cut marks: lessons from artiodactyl remains in the northwestern United States. Journal of Archaeological Science 32(12): 1722–1732.
2008 Quantitative Paleozoology. Cambridge Manuals in Archaeology. Cambridge University Press, Cambridge.
Lyman, R. Lee, Lori E. Houghton, and Anne L. Chambers
1992 The Effect of Structural Density on Marmot Skeletal Part Representation in Archaeological Sites. Journal of Archaeological Science 19: 557–573.
Lyman, R. Lee, and Michael J. O’Brien
2001 The Direct Historical Approach, Analogical Reasoning, and Theory in Americanist Archaeology. Journal of Archaeological Method and Theory 8(4): 303–342.
Lyman, R. Lee, James M. Savelle, and Peter Whitridge
1992 Derivation and Application of a Meat Utility Index for Phocid Seals. Journal of Archaeological Science 19: 531–555.
MacDonald, Stephen O., and Joseph A. Cook
2010 Recent Mammals of Alaska. University of Alaska Press, Fairbanks.
306
MacNeish, Richard S.
1959 A Speculative Framework of Northern North American Prehistory as of April 1959. Anthropologica 1(1/2): 7–23.
Marchani, E. E., A. R. Rogers, and D. H. O’Rourke
2007 Brief Communication: The Thule Migration: Rejecting Population Histories Using Computer Simulation. American Journal of Physical Anthropology 134: 281–284.
Marciniak, Arkadiusz
2005 Placing Animals in the Neolithic: Social Zooarchaeology of Prhistoric Farming Communities. UCL Press, London.
Marean, Curtis W., and Leanne Bertino
1994 Intrasite Spatial Analysis of Bone: Subtracting the Effect of Secondary Carnivore Consumers. American Antiquity 59(4): 748–768.
Marean, Curtis W., and Carol J. Frey
1997 Animal Bones from Caves to Cities: Reverse Utility Curves as Methodological Artifacts. American Antiquity 62(4): 698–711.
Marean, Curtis W., and Lillian M. Spencer
1991 Impact of Carnivore Ravaging on Zooarchaeological Measures of Element Abundance. American Antiquity 56(4): 645–658.
Marean, Curtis W., Lillian M. Spencer, Robert J. Blumenschineh, and Salvatore D. Capaldob
1992 Captive Hyaena Bone Choice and Destruction , the Schlepp Effect and Olduvai Archaeofaunas. Journal of Archaeological Science 19: 101–121.
Maschner, Herbert D. G., and Robert McGhee
2009 Prologue and Introduction. In The Northern World AD 900-1400, edited by Herbert D. G. Maschner, Owen K. Mason, and Robert McGhee, pp. 1–6. University of Utah Press, Salt Lake City.
Mason, Owen K.
1990 Beach Ridge Geomorphology of Kotzebue Sound: Implications for Paleoclimatology and Archaeology. University of Alaska Fairbanks.
1998 The Contest between the Ipiutak, Old Bering Sea, and Birnirk Polities and the Origin of Whaling during the First Millennium A. D. along Bering Strait. Journal of Anthropological Archaeology 17: 240–325.
307
2000 Archaeological Rorshach in Delineating Ipiutak, Punuk and Birnirk in NW Alaska: Masters, Slaves or Partners in Trade? In Identities and Cultural Contacts in the Arctic: Proceedings from a Conference at the Danish National Museum, Copenhagen, November 30 to December 2, 19992, edited by Martin Appelt, Joel Bergland, and Hans Christian Gulløv, pp. 229–251. Danish Polar Center Pulications. Danish National Museum & Danish Polar Center, Copenhagen.
2009a “The Multiplication of Forms:” Bering Strait Harpoon Heads as a Demic and Macroevolutionary Proxy. In Macroevolution in Human Prehistory, edited by Anna Prentiss, Ian Kuijt, and James C. Chatters, pp. 73–103. Springer, New York.
2009b Flight from the Bering Strait: Did Siberian Punuk/Thule Military Cadres Conquer Northwest Alaska? In The Northern World AD 900-1400, edited by Herbert D. G. Maschner, Owen Mason, and Robert McGhee, pp. 76–128. The University of Utah Press, Salt Lake City.
2012 Memories of Warfare: Archaeology and Oral History in Assessing the Conflict and Alliance Model of Ernest S. Burch. Arctic Anthropology 49(2): 72–93.
Mason, Owen K., and Valerie Barber
2003 A Paleo-Geographic Preface to the Origins of Whaling: Cold is Better. In Indigenous Ways to the Present: Native Whaling in the Western Arctic, edited by Allen P. McCartney, pp. 69–107. Studies in Whaling. Canadian Circumpolar Institute, Edmonton.
Mason, Owen K., and Nancy H. Bigelow
2008 The Crucible of Early to Mid-Holocene Climate in Northern Alaska: Does Northern Archaic Represent the People of the Spreading Forest? Arctic Anthropology 45(2): 39–70.
Mason, Owen K., and Peter M. Bowers
2009 The Origin of Thule is Always Elsewhere: Early Thule within Kotzebue Sound, “Cul-de- Sac” or Nursery? In On the Track of the Thule Culture from Bering Strait to East Greenland: Proceedings of the SILA Conference “The Thule Culture-New Perspectives in Inuit Prehistory” Copenhagen, Oct. 26th-28th, 2006. Papers in Honour of Hans Chirstian Gullov, edited by Bjarne Grønnow, pp. 25–44. Studies in Archaeology & History. Publications from the National Museum, Copenhagen.
Mason, Owen K., and S. Craig Gerlach
1995a Chukchi Hot Spots, Paleo-Polynyas, and Caribou Crashes: Climatic and Ecological Dimensions of North Alaska Prehistory. Arctic Anthropology 32(1): 101–130.
1995b The Archaeological Imagination, Zooarchaeological Data, the Origins of Whaling in the Western Arctic, and “Old Whaling” and Choris Cultures. In Hunting the Largest Animals: Native Whaling in the Western Arctic and Subarctic, edited by Allen P. McCartney, pp. 1–31. Occassional Publication. The Canadian Circumpolar Institute, Edmonton.
308
Mason, Owen K., and James W. Jordan
1993 Heightened North Pacific Storminess during Synchronous Late Holocene Erosion of Northwest Alaska Beach Ridges. Quaternary Research 40: 55–69.
2002 Minimal late Holocene sea level rise in the Chukchi Sea: arctic insensitivity to global change? Global and Planetary Change 32(1): 13–23.
Mason, Owen K., and Stefanie L. Ludwig
1990 Resurrecting Beach Ridge Archaeology: Parallel Depositional Records from St. Lawrence Island and Cape Krusenstern, Western Alaska. Geoarchaeology: An International Journal 5(4): 349–373.
Masson, Bertrand
2010 Combustion features and periglacial structures: a new taphonomic analysis of mousterian combustion features at Saint-Vaast-La-Hougue (50). P@lethnologie 2: 5–23.
Mathiassen, Therkel
1927a Archaeology of the Central Eskimos. I. Descriptive Part. Report of the Fifth Thule Expedition 1921-42.
1927b The Thule Culture and its Position within the Eskimo Culture: Archaeology of the Central Eskimos II. Ed. Thule Expedition. Gyldendal, Copenhagen.
1929 Some specimens from the Bering Sea culture. Indian Notes 6(1): 33–56.
Matthiesen, H., J. B. Jensen, D. Gregory, J. Hollesen, and B. Elberling
2014 Degradation of Archaeological Wood Under Freezing and Thawing Conditions-Effects of Permafrost and Climate Change. Archaeometry 56(3): 479–495.
Maxwell, Moreau S.
1981 A Southeastern Baffin Thule House with Ruin Island Characteristics. Arctic 34(2): 133– 140.
1985 Prehistory of the Eastern Arctic. Academic Press, Inc., New York.
McCartney, Allen P.
1977 Thule Eskimo Prehistory along Northwestern Hudson Bay. Mercury Series, Archaeological Survey of Canada. National Museum of Man, Ottawa.
1979 A processual consideration of Thule Whale bone houses. In Thule Eskimo Culture: An Anthropological Retrospective, edited by Alan P. McCartney, pp. 301–323. Mercury Series, Archaeological Survey of Canada. National Museum of Man, Ottawa.
309
1980 The Nature of Thule Eskimo Whale Use. Arctic 33(3): 517–541.
1995 Whale Size Selection by Precontact Hunters of the North American Western Arctic and Subarctic. In Hunting the Largest Animals: Native Whaling in the Western Arctic and Subarctic, edited by Allen P. McCartney, pp. 83–108. The Canadian Circumpolar Institute, Edmonton.
McCartney, Allen P., and D. J. Mack
1973 Iron Utilization by Thule Eskimos of Central Canada. American Antiquity 38(3): 328–339.
McCartney, Allen P., and James M. Savelle
1985 Thule Eskimo Whaling in the Central Canadian Arctic. Arctic Anthropology 22(2): 37–58.
1993 Bowhead whale bones and Thule Eskimo subsistence–settlement patterns in the central Canadian Arctic. Polar Record 29(168): 1–12.
McCartney, Peter H., and Margaret F. Glass
1990 Simulation Models and the Interpretation of Archaeological Diversity. American Archaeology 55(3): 521–536.
McCullough, Karen M.
1986 The Ruin Island Phase of Thule Culture in the Eastern High Arctic. University of Toronto.
1989 The Ruin Islanders: Thule Culture Pioneers in the Eastern High Arctic. Archaeological Survey of Canada Mercury Paper. Canadian Museum of Civilization, Ottawa.
McGhee, Robert
1969 Speculations on Climatic Change and Thule Culture Development. Folk 11/12: 173–184.
1972 Copper Eskimo Prehistory. National Museum of Man Publications in Archaeology. Vol. 2. National Museums of Canada, Ottawa.
1974 Beluga Hunters: An Archaeological Reconstruction of the History and Culture of the Mackenzie Delta Kittegaryumiut. Institute of Social and Economic Research, Memorial University of Newfoundland, St. John’s, Newfoundland.
1977 Ivory for the Sea Woman: The Symbolic Attributes of a Prehistoric Technology. Canadian Journal of Archaeology 1(1): 141–149.
1984a The Timing of the Thule Migration. Polarforschung 54(1): 1–7.
1984b Thule Prehistory of Canda. Ed. David Damas. Handbook of NorthAmerican Indians. Smithsonian Institution, Washington, DC.
1997 Meetings between Dorset culture Palaeo-Eskimos and Thule culture Inuit: Evidence from Brooman Point. In Fifty Years of Arctic Research: Anthropological Studies from Greenland to
310
Siberia, edited by R. Gilberg and Hans Christian Gulløv, pp. 209–213. Ethnograph. Publications of The National Museum Ethnographical Series. The National Museum of Denmark, Copenhagen.
2000 Radiocarbon Dating and the Timing of the Thule Migration. In Identities and Cultural Contacts in the Arctic: Proceedings from a Conference at the Danish National Museum, Copenhagen, November 30 to December 2, 1999, edited by Martin Appelt, Joel Berglund, and Hans Christian Gulløv, pp. 181–191. Danish National Museum & Danish Polar Center, Copenhagen.
2009a When and Why did the Inuit Move to the Eastern Arctic? In The Northern World AD 900-1400, edited by Herbert Maschner, Owen Mason, and Robert McGhee, pp. 155–164. University of Utah Press, Salt Lake City.
2009b The Population Size and Temporal Duration of Thule Culture in Arctic Canada. In On the Track of the Thule Culture from Bering Strait to East Greenland: Proceedings of the SILA Conference “The Thule Culture-New Perspectives in Inuit Prehistory” Copenhagen, Oct. 26th- 28th, 2006. Papers in Honour of Hans Chirstian Gullov, edited by Bjarne Grønnow, pp. 75–89. Studies in Archaeology & History. Publications from the National Museum, Copenhagen.
McGuire, Randall H.
1982 The study of ethnicity in historical archaeology. Journal of Anthropological Archaeology 1(2): 159–178.
McNiven, Ian J.
2010 Navigating the human-animal divide: marine mammal hunters and rituals of sensory allurement. World Archaeology 42(2): 215–230.
McNutt, S. Lyn
1981 Remote Sensing Analysis of Ice Growth and Distribution in the Eastern Bering Sea. In The Eastern Bering Sea Shelf: Oceanography and Resources. Volume One, edited by Donald W. Hood and John A. Calder, pp. 141–166. Office of Marine Population Assessment of the National Oceanic and Atmospheric Administration, Seattle.
Mecklenburg, Catherine W., T. Anthony Mecklenburg, and Lyman K. Thorsteinson
2002 Fishes of Alaska. American Fisheries Society, Bethesda, MD.
Medina, Matías E., Pablo Teta, and Diego Rivero
2012 Burning damage and small-mammal human consumption in Quebrada del Real 1 (Cordoba, Argentina): an experimental approach. Journal of Archaeological Science 39(3): 737– 743.
311
Metcalfe, Duncan, and Kevin T. Jones
1988 A Reconsideration of Animal Body-Part Utility Indices. American Antiquity 53(3): 486– 504.
Middleton, William D., Gary M. Feinman, and Linda M. Nicholas
2002 Domestic Faunal Assemblages from the Classic Period Valley of Oaxaca, Mexico: A Perspective on the Subsistence and Craft Economies. Journal of Archaeological Science 29(3): 233–249.
Miller, Daniel
1987 Material Culture and Mass Consumption. Blackwell, Oxford.
Milne, S. Brooke
2003 Identifying Pre-Dorset structural features on southern Baffin Island: Challenges and considerations for alternative sampling. Etudes/Inuit/Studies 27(1-2): 67–90.
Mohl, Jeppe
1979 Description and analysis of the bone material from Nugarsuk: An Eskimo settlement representative of the Thule culture in West Greenland. In Thule Eskimo Culture: An Anthropological Retrospective, edited by Allen P. McCartney, pp. 380–394. Mercury Series Paper. Canadian Museum of Civilization, Archaeological Survey of Canada, Ottawa.
Moody, John F., and Lisa M. Hodgetts
2013 Subsistence Practices of Pioneering Thule – Inuit : A Faunal Analysis of Tiktalik. Arctic Anthropology 50(2): 4–24.
Morin, Eugène
2007 Fat composition and Nunamiut decision-making: a new look at the marrow and bone grease indices. Journal of Archaeological Science 34(1): 69–82.
2010 Taphonomic implications of the use of bone as fuel. P@lethnologie 2: 209–217.
Morlan, Richard E.
1994 Bison Bone Fragmentation and Survivorship: a Comparative Method. Journal of Archaeological Science 21: 797–807.
Morrison, David A.
1983a Thule Culture in Western Coronation Gulf, N.W.T. Archaeological Survey of Canada, Mercury Series. Canadian Museum of Civilization, Ottawa.
312
1983b Thule Sea Mammal Hunting in the Western Central Arctic. Arctic Anthropology 20(2): 61–78.
1987 Thule and historic copper use in the Copper Inuit area. American Antiquity 52(1): 3-12.
1989 Radiocarbon Dating Thule Culture. Arctic Anthropology 26(2): 48–77.
1991 The Copper Inuit soapstone trade. Arctic 44(3): 239-246.
1994 An Archaeological Perspective on Neoeskimo Economies. In Threads of Arctic Prehistory: Papers in Honour of William E. Taylor Jr., edited by David Morrison and Jean-Luc Pilon, pp. 311–324. Archaeological Survey of Canada Paper. Canadian Museum of Civilization Mercury Series, Ottawa.
1997a An ethnohistory of the Inuvialuit from earliest ties to 1902. Revista de Arqueologia Americana 12: 29-54.
1997b Caribou hunters in the western Arctic: zooarchaeology of the Rita-Claire and Bison Skull Sites. Archaeological Survey of Canada. Canadian Museum of Civilization, Hull, Quebec.
1997c Estimating the age and sex of caribou from mandibular measurements. Journal of Archaeological Science 24: 1093-1106.
1999 The Earliest Thule Migration. Canadian Journal of Archaeology 22: 139–156.
2000 The Arrival of the Inuit: Amundsen Gulf and the Thule Migration. In Identities and Cultural Contacts in the Arctic: Proceedings from a Conference at the Danish National Museum, Copenhagen, November 30 to December 2, 1999, edited by Martin Appelt, Joel Berglund, and Hans Christian Gulløv, pp. 221–228. Danish Polar Center Publication. Danish National Museum & Danish Polar Center, Copenhagen.
2001 Radiocarbon Dating the Birnirk-Thule Transition. Anthropological Papers of the University of Alaska 1(1): 73–85.
2002 Diamond Jenness: The first Canadian archaeologist. In Honoring Our Elders: A History of Eastern Arctic Archaeology, edited by William W. Fitzhugh, Stephen Loring, and Daniel Odess, pp. 61–65. Contributions to Circumpolar Anthropology. Arctic Studies Center, National Museum of Natural History, Smithsonian Institution, Washington, DC.
Moss, Madonna L., and Peter M. Bowers
2007 Migratory Bird Harvest in Northwestern Alaska: A Zooarchaeological Analysis of Ipiutak and Thule Occupations from the Deering Archaeological District The Regulation of Migratory. Arctic Anthropology 44(1): 37–50.
313
Murdoch, John
1892 Ethnological Results of the Point Barrow Expedition. Ninth Annual Report of the Bureau of Ethnology to the Secretary of the Smithsonian Institution. Government Printing Office, Washington.
Nadal, Daniel
2003 A long continuity: The Ohalo II brush huts (19.5 ky) and the dwelling structures in the Natufian and PPNA sites in the Jordan Valley. Archaeology, Anthropology and Ethnology in Euroasia 13(1): 34–48.
Nash, D. J.
2009 Household archaeolgy in the Andes. Journal of Archaeological Research 17: 205–261.
Nelson, Edward William
1899 The Eskimo about Bering Strait. Eighteenth Annual Report of the Bureau of American Ethnology. Goverment Printing Office, Washington.
Nelson, N. C.
1916 Chronology of the Tano Ruins, New Mexico. American Anthropologist 18(2): 159–180.
Newell, Raymond R.
1990 The Intermound and Extramound Tests. In The Utqiagvik Excavations, Volume 1: The 1981 Excavations at the Utqiagvik Archaeological Site Barrow, Alaska, edited by Edwin S. Jr. Hall and Lynne Fullerton, pp. 174–209. The North Slope Borough Commission on Iñupiat History, Language and Culture, Barrow.
Nicholson, H. B. (editor).
1976 Origins of religious art & iconography in preclassic Mesoamerica. UCLA Latin American Center Publications, Los Angeles.
Norman, Lauren, and T. Max Friesen
2010 Thule Fishing Revisited: The Economic Importance of Fish at the Pembroke and Bell Sites, Victoria Island, Nunavut. Geografisk Tiddsskrift-Danish Journal of Geography 110(2): 261–278.
Novecosky, Brad J., and Peter R. W. Popkin
2005 Canidae volume bone mineral density values: an application to sites in western Canada. Journal of Archaeological Science 32(11): 1677–1690.
314
O’Brien, Matthew, and Curtis B. Storlie
2011 An Alternative Bilateral Refitting Model for Zooarchaeological Assemblages.
O’Connell, James F.
1995 Ethnoarchaeology Needs a General Theory of Behavior. Journal of Archaeological Research 3(3): 205–255.
O’Connell, James F., Kristen Hawkes, and Nicholas Blurton-Jones
1990 Reanalysis of large mammal body part transport among the Hadza. Journal of Archaeological Science 17: 310–316.
1992 Patterns in the distribution, site structure, and assmblage composition of Hadza kill- butchering sites. Journal of Archaeological Science 19: 319–345.
1988 Hadza Hunting, Butchering, and Bone Transport and Their Archaeological Implications. Journal of Anthropological Research 44(2): 113–161.
O’Connor, Terry
2000 The Archaeology of Animal Bones. Sutton Publishing, Gloucestershire.
Olson, C., N. Bjorck, and J. Stora
2009 Huts and Deposition of Refuse at Frakenronningen, a Neolithic Coastal Dwelling Site in Eastern Middle Sweden. International Journal of Osteoarchaeology 10.
Omura, Keiichi
2002 Construction of Inuinnaqtun (Real Inuit-way): Self-Image and Everyday Practices in Inuit Society. Senri Ethnological Studies 60: 101–111.
2013 The Ontology of Sociality: “Sharing” and Subsistence Mechanisms. In Groups: Evolution of Human Sociality, edited by Kaori Kawai, pp. 123–142. Kyoto University Press and Trans Pacific Press, Portland, OR.
Orton, Clive
2000 Sampling in Archaeology. Cambridge Manuals in Archaeology. Cambridge University Press, Cambridge.
Orton, David C.
2012 Taphonomy and interpretation: An analytical framework for social zooarchaeology. International Journal of Osteoarchaeology 22(3): 320–337.
315
Oswalt, Wendell H., and James W. VanStone
1967 The ethnoarcheology of Crow Village, Alaska. Bureau of American Ethnology. Smithsonian Institution, Washington, DC.
Otárola-Castillo, Erik
2010 Differences between NISP and MNE in cutmark analysis of highly fragmented faunal assemblages. Journal of Archaeological Science 37(1): 1–12.
Outram, Alan K.
2001 A New Approach to Identifying Bone Marrow and Grease Exploitation: Why the “Indeterminate” Fragments should not be Ignored. Journal of Archaeological Science 28(4): 401–410.
2002 Bone Fracture and Within-bone Nutrients: an Experimentally Based Method for Investigating Levels of Marrow Extraction. In Consuming Passions and Patterns of Consumption, edited by Preston T Miracle and Nicky Milner, pp. 51–63. McDonald Institute for Archaeological Research, Cambridge.
Outram, Alan K, and Keith Dobney and Peter Rowley-Conwy Umberto Albarella
2005 Distinguishing bone fat exploitation from other taphonomic processes: what caused the high level of bone fragmentation at the Middle Neolithic site of Ajvide, Gotland? In The Zooarchaeology of Fats, Oils, Milk and Dairying, edited by J and A Outram Mulville. Oxbow Books, Oxford.
Palsson, Gisli, and Vilhjalmur Stefansson
2001 Writing on ice: the ethnographic notebooks of Vilhjalmur Stefansson. Ed. Gâisli Pâalsson. University Press of New England, Hanover, NH.
Park, Robert W.
1988 “Winter Houses” and Qarmat in Thule and Historic Inuit Settlement Patterns: Some Implications for Thule Studies. Canadian Journal of Archaeology 12: 163–175.
1989 Porden Point: An Intrasite Approach to Settlement System Analysis. The University of Alberta.
1999 Seal use and storage in the Thule culture of Arctic North America. Revista de Arqueologia Americana 16: 78–97.
Partlow, Megan A.
2006 Sampling Fish Bones: A Consideration of the Importance of Screen Size and Disposal Context in the North Pacific. Arctic Anthropology 43(1): 67–79.
316
Patton, A. Katherine, and James M. Savelle
2006 The symbolic dimensions of whale bone use in Thule winter dwellings. Etudes/Inuit/Studies 30(2): 137–161.
Pavao, Barnet, and Peter W. Stahl
1999 Structural Density Assays of Leporid Skeletal Elements with Implications for Taphonomic, Actualistic and Archaeological Research. Journal of Archaeological Science 26: 53–66.
Petitot, Emile
1887 Les Grands Esquimaux. Librairie Plon, Paris.
Ping, C. L., G. J. Michaelson, J. M. Kimble, V. E. Romanovsky, Y. L. Shur, D. K. Swanson, and D. A. Walker
2008 Cryogenesis and soil formation along a bioclimate gradient in Arctic North America. Journal of Geophysical Research 113(G03S12): 1–14.
Pluckhahn, Thomas J.
2010 Household Archaeology in the Southeastern United States: History, Trends, and Challenges. Journal of Archaeological Research 18(4): 331–385.
Plug, C., and Ina Plug
1990 MNI counts as estimates of species abundance. South African Archaeological Bulletin 45: 53–57.
Polglase, Christopher R.
1990 The 1982 Extramound Investigations Around Mounds 7 and 8. In The Utqiagvik Excavations, Volume 2: Additional Reports of the 1982 INvestigations by the Utqiagvik Archaeology Project Barrow, Alaska, edited by Edwin S. Jr. Hall and Lynne Fullerton, pp. 105– 166. The North Slope Borough Commission on Iñupiat History, Language and Culture, Barrow.
Powers, William Rogers, J. K. Adams, A. Godfrey, J. Ketz, D. Plaskett, and G.R. Scott
1982 The Chukchi-Imuruk Report: Archaeological Investigations in the Bering Land Bridge National Preserve, Seward Peninsula, Alaska 1974 and 1975. Anthropology and Historic. Anthropology and Historic Preservation Studies Unit Occasional Paper. University of Alaska, Fairbanks.
Raff, Jennifer A., Margarita Rzhetskaya, and M. Geoffrey Hayes
2013 Patterns of Genetic Diversity and Shared Mitochondrial and Y-Chromosome Lineages among Iñupiat Communities of the Alaskan North Slope. In Alaska Anthropological Association 40th Annual Meeting. Anchorage.
317
Raff, Jennifer A., Margarita Rzhetzkaya, Justin Tackney, and M. Geoffrey Hayes
2015 Mitochondrial diversity of Iñupiat people from the Alaskan North Slope provides evidence for the origins of the Paleo- and Neo-Eskimo peoples. American Journal of Physical Anthropology.. doi: 10.10002/ajpa.22750.
Raghavan, M., M. DeGiorgio, A. Albrechtsen, I. Moltke, P. Skoglund, T. S. Korneliussen, B. Gronnow, M. Appelt, H. C. Gullov, T. M. Friesen, W. Fitzhugh, H. Malmstrom, S. Rasmussen, J. Olsen, L. Melchior, B. T. Fuller, S. M. Fahrni, T. Stafford, V. Grimes, M. a. P. Renouf, J. Cybulski, N. Lynnerup, M. M. Lahr, K. Britton, R. Knecht, J. Arneborg, M. Metspalu, O. E. Cornejo, A.-S. Malaspinas, Y. Wang, M. Rasmussen, V. Raghavan, T. V. O. Hansen, E. Khusnutdinova, T. Pierre, K. Dneprovsky, C. Andreasen, H. Lange, M. G. Hayes, J. Coltrain, V. a. Spitsyn, A. Gotherstrom, L. Orlando, T. Kivisild, R. Villems, M. H. Crawford, F. C. Nielsen, J. Dissing, J. Heinemeier, M. Meldgaard, C. Bustamante, D. H. O’Rourke, M. Jakobsson, M. T. P. Gilbert, R. Nielsen, and E. Willerslev
2014 The genetic prehistory of the New World Arctic. Science 345(6200): 1255832–1255832.
Rainey, Froelich G.
1941 Eskimo Prehistory: The Okvik Site on the Punuk Island. Anthropological Papers of the American Museum of Natural History 37(4): 453–569.
1947 The Whale Hunters of Tigara. Anthropological Papers of the American Museum of Natural History 41(2): 231–283.
Rapoport, A.
1990 Systems of activities and systems of settings. In Domestic Architecture and the Use of Space, edited by Susan Kent, pp. 9–20. Cambridge University Press, Cambridge.
Rasmussen, Knud
1925 Folklore. In the Danish Ethnographic and Geographic Expedition to Arctic America. Preliminary Report of the Fifth Thule Expedition. The Geographical Review 15(4): 525-535.
1931 The Netsilik Eskimos: Social Life and Spiritual Culture. Report of the Fifith Thule Expedition. Gyldendalske boghandel, Nordisk forlag, Copenhagen.
1932 Intellectual Culture of the Copper Eskimos. Report of the Fifth Thule Expedition 1921-24. Gyldeddalske Boghandel, Nordisk Forlag, Copenhagen.
1933 Across Arctic America: Narrative of the Fifth Thule Expedition. G.P. Putnam’s Sons, London.
1976 Observations on the Intellectual Culture of the Caribou Eskimos. AMS Press, New York.
318
Ravn, Mads
2011 Ethnographic analogy from the Pacific: just as analogical as any other analogy. World Archaeology 43(2008): 716–725.
Ray, Dorothy Jean
1964 Nineteenth Century Settlement and Subsistence Patterns in Bering Strait. Arctic Anthropology 2(2): 61–94.
1967 Land Tenure and Polity of the Bering Strait Eskimos. Journal of the West 6(3): 371–394.
1975a The Eskimos of Bering Strait, 1650-1898. University of Washington Press, Seattle.
1975b Early Maritime Trade with the Eskimo of Bering Strait and the Introduction of Firearms. Arctic Anthropology 12(1): 1–9.
1983 Ethnohistory in the Arctic: The Bering Strait Eskimo. Ed. R. A. Pierce. Alaska History. The Limestone Press, Kingston.
1984 Bering Strait Eskimos. Ed. David Damas. Handbook of North American Indians. Smithsonian Institution Press.
Reed, C. A.
1963 Osteo-Archaeology. In Science in Archaeology: A Survey of Progress and Research, edited by D. R. Brothwell and E. S. Higgs, pp. 1213–1227. Thames and Hudson, London.
Reitz, Elizabeth J., and Elizabeth S. Wing
2008 Zooarchaeology. 2nd ed. Cambridge Manuals in Archaeology. Cambridge University Press, Cambridge.
Rick, Anne Meachem
1980 Non-Cetacean Vertebrate Remains from Two Thule Winter Houses on Somerset Island, N.W.T. Canadian Journal of Archaeology 4: 99–117.
Ringrose, T. J.
1993 Bone Counts and Statistics: A Critique. Journal of Archaeological Science 20: 121–157.
Robertson, Elizabeth C., Jeffrey D. Seibert, Deepika C. Fernandez, and Marc U. Zender
2006 Space and Spatial Analysis in Archaeology. University of Calgary Press, Calgary.
Robertson, Ian G.
1999 Spatial and multivariate analysis, random sampling error, and analytical noise: Empirical Bayesian methods at Teotihuacan, Mexico. American Antiquity 64(1): 137–152.
319
Robin, Cynthia
2002 Outside of houses: The practices of everyday life at Chan Noohol, Belize. Journal of Social Archaeology 2(2): 245–268.
2003 New Directions in Classic Maya Household Archaeology. Journal of Archaeological Research 11(4): 307–356.
Rosell, Jordi, Ruth Blasco, M. Cristina Fernandez-Laso, Manuel Vaquero, and Eudald Carbonell
2012 Connecting areas: Faunal refits as a diagnostic element to identify synchronicity in the Abric Romani archaeoloigcal assemblages. Quaternary International 252: 56–67.
Rossignol, Jacqueline, and LuAnn Wandsnider (editors).
1992 Space, Time, and Archaeological Landscapes. Plenum Press, New York.
Roux, Valentine
2007 Ethnoarchaeology: A Non Historical Science of Reference Necessary for Interpreting the Past. Journal of Archaeological Method and Theory 14(2): 153–178.
Sabloff, J. A., and Wendy Ashmore
2001 An aspect of archaeology’s recent past and its relevance in the new millenium. In Archaeology at the Millenium: A Sourcebook, edited by Gary M. Feinman and T. Douglas Price, pp. 11–32. Kluwer/Plenum, New York.
Sabo, George III, and John D Jacobs
1980 Aspects of Thule Culture Adaptations in Southern Baffin Island. Arctic 33(3): 487–504.
Salmon, Merrilee H.
1982 Confirmation in Archaeology and Analogy and Functional Ascription. In Philosophy and Archaeology, pp. 31–83. Academic Press, New York.
Sapir, Edward
1916 Time perspective in aboriginal American culture: A study in method. Canada Dept. of Mines, Geological Survey, Ottawa.
Savelle, James M.
1986 Mobility and Subsistence-Settlement: An Archaeological Example from the Central Canadian Arctic. The University of Alberta.
1987 Collectors and foragers: Subsistence-settlement system change in the central Canadian Arctic, A.D. 1000-1960. BAR, Oxford.
320
1995 An Ethnoarchaeological Investigation of Inuit Beluga Whale and Narwhal Harvesting. In Hunting the Largest Animals: Native Whaling in the Western Arctic and Subarctic, edited by Allen P. McCartney, pp. 127–148. Occassional Paper. The Canadian Circumpolar Institute, Edmonton.
1997 The role of architectural utility in the formation of zooarchaeological whale bone assemblages. Journal of Archaeological Science 24: 869–885.
2000 Information systems and Thule Eskimo bowhead whaling. In Animal Bones, Human Societies, edited by Peter Rowley-Conwy, pp. 74–86. Oxbow Books, Oxford.
2005 The Development of Indigenous Whaling: Prehistoric and Historic Contexts. Senri Ethnological Studies 67: 53–58.
Savelle, James M., Arthur S. Dyke, and Allen P. Mccartney
2000 Holocene bowhead whale (Balaena mysticetus) mortality patterns in the Canadian Arctic Archipelago. Arctic 53(4): 414–421.
Savelle, James M., and T. Max Friesen
1996 An Odontocete (Cetacea) Meat Utility Index. Journal of Archaeological Science 23(5): 713–721.
Savelle, James M., and Junko Habu
2004 A Processual Investigation of a Thule Whale Bone House, Somerset Island, Arctic Canada. Arctic Anthropology 41(2): 204–221.
Savelle, James M., and Allen P. McCartney
1988 Geographical and temporal variation in Thule Eskimo subsistence economies: a model. Research in Economic Anthropology 10: 21–72.
1990 Prehistoric Thule Eskimo whaling in the Canadian Arctic Islands: Current knowledge and future reserach directions. In Canada’s Missing Dimensions: Science and History in the Canadian Arctic Islands, Vol. II, edited by C. R. Harington, pp. 695–723. Canadian Museum of Nature, Ottawa.
1994 Thule Inuit Bowhead Whaling: A Biometrical Analysis. In Threads of Arctic Prehistory: Papers in Honour of William E. Taylor Jr., edited by David Morrison and Jean-Luc Pilon, pp. 281–310. Archaeological Survey of Canada Paper. Canadian Museum of Civilization Mercury Series, Ottawa.
1999 Thule Eskimo bowhead whale interception strategies. World Archaeology 30(3): 437–451.
2002 Prehistoric Gray whale harvesting in the Bering Strait and Chukchi Sea regions of Alaska: A zooarchaeological assessment. In Archaeology in the Bering Strait Region: Research on Two
321
Continents, edited by Don E. Dumond and Richard L. Bland, pp. 307–318. University of Oregon Anthropological Papers. University of Oregon.
2003 Prehistoric Bowhead Whaling in the Bering Strait and Chukchi Sea Regions of Alaska: A Zooarchaeological Assessment. In Indigenous Ways to the Present: Native Whaling in the Western Arctic, edited by Allen P. McCartney, pp. 167–184. Studies in Whaling. Canadian Circumpolar Institute, Edmonton.
Savelle, James M., and George W. Wenzel
2003 Out of Alaska: Reconstructing the Social Structure of Prehistoric Canadian Thule Culture. Senri Ethnological Studies 63: 103–121.
Schaaf, Jeanne M.
1988 Bering Land Bridge National Preserve: An Archaeological Survey Volume II: Site Descriptions. National Park Service, U.S. Department of the Interior, Anchorage.
1995 Late-Prehistoric Inupiaq Societies, Northern Seward Peninsula, Alaska: An Archaeological Analysis AD 1500-1800. The University of Minnesota.
Schiffer, Michael B.
1972 Archaeological Context and Systemic Context. American Antiquity 37(2): 156–165.
1976 Behavioral Archaeology. Academic Press, New York.
1977 Current directions in archaeologial method and theory. American Anthropologist 79(3): 647–649.
1983 Toward the Identification of Formation Processes. American Antiquity 48(4): 675–706.
1985 Is There a “Pompeii Premise” in Archaeology? Journal of Anthropological Research 41(1): 18–41.
1987 Formation Processes of the Archaeological Record. University of New Mexico Press, Albuquerque.
2010 Behavioral Archaeology: Principles and Practice. Equinox Handbooks in Anthropological Archaeology. Equinox Publishing, London.
Schledermann, Peter
1975a Thule Eskimo Prehistory of Cumberland Sound, Baffin Island, Canada. Mercury Series Paper. Canadian Museum of Civilization, Archaeological Survey of Canada, Ottawa.
1975b A Late Dorset Site on Axel Heiberg Island. Arctic 28(4): 300.
1976a Thule Culture Communal Houses in Labrador. Arctic 29(1): 27–37.
322
1976b The Effect of Climatic/Ecological Changes on the Style of Thule Culture Winter Dwellings. Arctic and Alpine Research 8(1): 37–47.
1978 Preliminary results of archaeological investigations in the Bache Peninsula region, Ellesmere Island, N. W. T. Arctic 31(4): 459–474.
Schledermann, Peter, and Karen McCullough
1980 Western Elements in the Early Thule Culture of the Eastern High Arctic. Arctic 33(4): 833–841.
Schmitt, Dave N., and Charles D. Zeier
1993 Not by bones alone: Exploring household composition and socioeconomic status in an isolated historic mining community. Historical Archaeology 27(4): 20–38.
2010 Not by Bones Alone: Exploring Household Composition and Socioeconomic Status an Isolated Historic Mining Community. Historical Archaeology 27(4): 20–38.
Scott, Elizabeth M.
1996 Who ate what? Archaeological food remains and cultural diversity. In Case Studeis in Environmental Archaeology, edited by Elizabeth J. Reitz, Lee A. Newson, and Sylvia J. Scudder, pp. 339–356. Plenum Press, New York.
Searles, Edmund
2002 Food and the Making of Modern Inuit Identities. Food and Foodways 10(1-2): 55–78.
Shanks, M., and Christopher Y. Tilley
1987 Re-constructing Archaeology. Cambridge University Press, Cambridge.
Sheehan, Glenn W.
1990 Excavations at Mound 34. In The Utqiagvik Excavations, Volume 2: Additional Reports of the 1982 Investigations by the Utqiagvik Archaeology Project Barrow, Alaska1, edited by Edwin S. Jr. Hall and Lynne Fullerton, pp. 181–325. The North Slope Borough Commission on Iñupiat History, Language and Culture, Barrow.
1995 Whaling Surplus, Trade, War, and the Integration of Prehistoric Northern and Northwestern Alaskan Economies, A.D. 1200-1826. In Hunting the Largest Animals: Native Whaling in the Western Arctic and Subarctic, edited by Allen P. McCartney, pp. 185–206. Studies in Whaling. The Canadian Circumpolar Institute, University of Alberta.
Shelley, Cameron
1999 Multiple Analogies in Archaeology. Philosophy of Science 66(4): 579–605.
323
Sheppard, William L.
1998 Population movements, interaction, and legendary geography. Arctic Anthropology 35(2): 147–165.
Shipman, Pat, Giraud Fosterb, and Margaret Schoeninger
1984 Burnt Bones and Teeth: an Experimental Study of Color, Morphology, Crystal Structure and Shrinkage. Journal of Archaeological Science 11: 307–325.
Simpson, John
1875 Observations on the Western Eskimo and the Country They Inhabit. Reprint of 1855 report. In A Selection of Papers on Arctic Geography and Ethnology, Reprinted and Presented to the Arctic Expedition of 1875, pp. 233–275. Royal Geographic Society, London.
Smith, Michael E.
1992 Braudel’s temporal rhythms and chronology theory in archaeology. In Archaeology, Annales, and Ethnohistory, edited by A. B. Kanpp, pp. 223–234. Cambridge University Press, Cambridge.
2010 The archaeological study of neighborhoods and districts in ancient cities. Journal of Anthropological Archaeology 29(2): 137–154.
Sobelman, Sandra S.
1985 The Economics of Wild Resource Use in Shishmaref, Alaska. Fairbanks.
Souvatzi, S. G.
2008 A Social Archaeology of Households in Neolithic Greece: An Anthropological Approach. Cambridge University Press, Cambridge.
Spencer, Robert F.
1959 The North Alaskan Eskimo: A Study in Ecology and Society. Bureau of American Ethnology. Smithsonian Institution, Washington, DC.
Staab, Margie L.
1979 Analysis of faunal material recovered from a Thule Eskimo site on the island of Silumiut, N.W.T., Canada. In Thule Eskimo Culture: An Anthropological Retrospective, edited by Allen P. McCartney, pp. 349–379. Archaeological Survey of Canada, Mercury Series. Canadian Museum of Civilization, Ottawa.
Stahl, Ann Brower
1994 Change and Continuity in the Banda Area , Ghana: The Direct Historical Approach. Journal of Field Archaeology 21(2): 181–203.
324
1993 Concepts of Time and Approaches to Analogical Reasoning in Historical Perspective. American Antiquity 58(2): 235–260.
Stahl, Peter W., and James A. Zeidler
1990 Differential Bone-Refuse Accumulation in Food-Preparation and Traffic Areas on an Early Ecuadorian House Floor. Latin American Antiquity 1(2): 150–169.
Staley, David P.
1993 St. Lawrence Island’s Subsistence Diggeres: A New Perspective on Human Effects on Archaeological Sites. Journal of Field Archaeology 20: 347–355.
Stalibrass, Sue
2000 Dead dogs, dead horses: Site formation processes at Ribchester Roman fort. In Animal Bones, Human Societies, edited by Peter Rowley-Conwy, pp. 158–165. Oxbow Books, Oxford.
Stanford, Dennis J.
1973 The Origins of Thule Culture. The University of New Mexico.
1976 The Walakpa Site, Alaska: Its Place in the Birnirk and Thule Cultures. Smithsonian Contributions to Anthropology. Smithsonian Institution Press, Washington, DC.
Steensby, H P
1916 An Anthropological Study of the Origin of the Eskimo Culture. Bianco Lunos Bogtrykkeri, Kobenhavn.
Stefansson, Vilhjalmur
1913 My life with the Eskimo. The Macmillan Company, New York.
1914a The Stefansson-Anderson Arctic Expedition of the American Museum: Preliminary Ethnological Report. Anthropological Papers. The American Museum of Natural History, New York.
1914b Prehistoric and present commerce among the Arctic coast Eskimo. Canada Geological Survey, Ottawa.
1944 The Friednly Arctic: The Story of Five Years in the Polar Regions. New edition, with new material. MacMillan Co., New York.
Stein, J.
1992 Deciphering a Shell Midden. Academic Press, Orlando.
325
Stenton, Douglas R.
1983 An Analysis of Faunal Remains from the Peale Point Site (KkDo-1), Baffin Island , N.W.T. Trent University.
1987 Recent Archaeological Investigations in Frobisher Bay, Baffin Island, N.W.T. Canadian Journal of Archaeology 11: 13–48.
Steward, Julian H.
1942 The direct historical approach to archaeology. American Antiquity 7(4): 337–343.
Stewart, Henry
2005 The fish tale that is never told: A reconsideration of the importance of fishing in Inuit societies. Senri Ethnological Studies 67: 345–361.
Stiner, Mary C, Steven L Kuhn, Stephen Weiner, and Ofer Bar-Yosef
1995 Differential Burning, Recrystallization, and Fragmentation of Archaeological Bone. Journal of Archaeological Science 22: 223–237.
Stokes, Paul
2000 A cut above the rest? Officers and men at the South Shields Roman fort. In Animal Bones, Human Societies, edited by Peter Rowley-Conwy, pp. 145–151. Oxbow Books, Oxford.
Storå, Jan
2001 Skeletal development in the Grey seal Halichoerus grypus, the Ringed seal Phoca hispida botnica, the Harbour seal Phoca vitulina vitulina and the Harp seal Phoca groenlandica. Epiphyseal Fusion and Life History. In Innovations in Assesing Season of Capture, Age, and Sex of Archaeofaunas, edited by Anne Pike-Tay, XI:pp. 199–222. ArchaeZoologia. Pensee Sauvage, Grenoble.
2002 Neolithic seal exploitation on the Aland Islands in the Baltic Sea on the basis of epiphyseal fusion data and metric studies. International Journal of Osteoarchaeology 12(1): 49–64.
Sutcliffe, A. J.
1990 Rates of decay of mammalian remains in the permafrost environment of the Canadian High Arctic. In Canada’s Missing Dimensions: Science and History in the Canadian Arctic Islands, edited by C. F. Harrington, 1:pp. 161–186. Canadian Museum of Nature, Ottawa.
Taylor, William E. Jr.
1972 An archaeoloical survey between Cape Parry and Cambridge Bay, NWT, Canada in 1963. Archaeological Survey of Canada, Mercury Series. National Museums of Canada, Ottawa.
326
Teeter, Wendy G.
2004 Animal utilization in a growing city: vertebrate exploitation at Caracol, Belize. In Maya Zooarchaeology: New Directions in Method and Theory, edited by Kitty F. Emery, pp. 177–191. Monograph. Cotsen Institute of Archaeology, University of California Los Angeles.
Théry-Parisot, Isabelle, and Sandrine Costamagno
2005 Proprietes Combustibles des Ossements. Gallia Prehistoire 47: 235–254.
Théry-Parisot, Isabelle, Sandrine Costamagno, Jean-Philipe Brugel, Philippa Fossa, and Raphaele Guilbert
2005 The use of bone as fuel during the palaeolithic, experimental study of bone combustible properties. In The Zooarchaeology of Fats, Oils, Milk and Dairying, edited by J. and A. Outram Mulville, pp. 50–59. Proceedings of the 9th Conference of the International Council of Archaeozoology, Durham, August 2002. Oxbow Books, Oxford.
Thomas, Richard
2005 Animals, Economy, and Status: Integrating Zooarchaeological and Historical Data in the Study of Dudley Castle, West Midlands (c. 1100-1750). BAR British Series. Archaeopress, Oxford.
Thompson, Steven K.
1987 Sample size for estimating multinomial proportions. American Statistician 41: 42–46.
Thornton, Harrison Robertson
1931 Among the Eskimo of Wales, Alaska, 1890-93. Ed. Neda S. Thornton and William M. Jr. Thornton. John Hopkins Press, Baltimore.
Todisco, Dominique, and Hervé Monchot
2006 Bone Weathering in a Periglacial Environment: The Tayara Site (KbFk-7), Qikirtaq Island, Nunavik (Canada). Arctic 61(1): 87–101.
Trigger, Bruce G.
1991 Distinguished Lecture in Archeology: Constraint and Freedom-A New Synthesis for Archeological Explanation. American Anthropologist 93(3): 551–569.
1995 Expanding middle-range theory. Antiquity 69: 449–458.
2006 A history of archaeological thought. Cambridge University Press, New York.
327
Tuma, Michael W.
2006 Ethnoarchaeology of Subsistence Behaviors within a Rural African American Community: Implications for Interpreting Vertebrate Faunal Data from Slave Quarters Areas of Antebellum Plantation Sites. Historical Archaeology 40(4): 1–26.
Turner, A., and N. R. J. Fieller
1985 Considerations of minimum numbers: A response to Horton. Journal of Archaeological Science 12: 477–483.
Turner, Lucien M.
1894 Ethnology of the Ungava Bay District, Hudson Bay Territory. Ethnology. Bureau of Ethnology. Smithsonian Institution, Washington.
Uerpmann, Hans-Peter
1973 Animal bone finds and economic archaeology: A critical study of “Osteo-Archaeological” method. World Archaeology 4(3): 307–322.
Val, Aurore, and Jean-Baptiste Mallye
2011 Small Carnivore Skinning by Professionals: Skeletal Modifications and Implications for the European Upper Palaeolithic. Journal of Taphonomy 9(4): 221–243.
Vale, Deborah, and Robert H. Gargett
2002 Size Matters: 3-mm Sieves Do Not Increase Richness in a Fishbone Assemblage from Arrawarra I, an Aboriginal Australian Shell Midden on the Mid-north Coast of New South Wales, Australia. Journal of Archaeological Science 29(1): 57–63.
Vaneeckhout, Samuel, Juho-Antti Junno, Anna-Kaisa Puputti, and Tiina Aikas
2010 Prehistoric burned bone : use or refuse – results of a bone combustion experiment. Faravid 34: 7–15.
Vaneeckhout, Samuel, Anna-Kaisa Salmi, and Juho-Antti Junno
2013 Archaeological Refuse Fauna in Finland: Understanding the Role of Bone Combustion. Anthropozoologica 48(1): 125–134.
VanStone, James W.
1960 An Early Nineteenth Century Artist in Alaska: Louis Choris and the First Kotzebue Expedition. Pacific Northwest Quarterly 51(4): 145–158.
1970 Akulivikchuk: A nineteenth century Eskimo village on the Nushagak River, Alaska. Fieldiana. Anthropology 60: 1–123.
328
1977 A. F. Kashevarov’s Coastal Explorations in Southwestern Alaska, 1822. Trans. David H. Kraus. Fieldiana: Anthropology 69.
1984 Exploration and Contact History of Western Alaska. Handbook of North American Indians. Smithsonian Institution.
Vaquero, Manuel, and Ignasi Pasto
2001 The Definition of Spatial Units in Middle Palaeolithic Sites: The Hearth-Related Assemblages. Journal of Archaeological Science 28(11): 1209–1220.
Van der Veen, Marijke, and Nick Fieller
1982 Sampling seeds. Journal of Archaeological Science 9: 287–298.
Viveiros de Castro, Eduardo
1998 Cosmological Deizis and Amerindiat Perspectivism. The Journal of the Royal Anthropological Institute 4(3): 469– 488.
Waguespack, Nicole M.
2002 Caribou sharing and storage: refitting the Palangana site. Journal of Anthropological Archaeology 21(3): 396–417.
Walls, Matthew
2009 Caribou Inuit Traders of the Kivalliq Nunavut, Canada. International Series. BAR, Oxford.
Watson, Patty Jo
1979 The idea of ethnoarchaeology: Notes and comments. In Ethnoarchaeology: Implications of Ethnography for Archaeology, edited by Carol Kramer, pp. 277–287. Columbia University Press, New York.
1986 Archaeological interpretation, 1985. In American Archaeology Past and Future: A Celebration of the Society for American Archaeology, edited by David J. Meltzer, D. D. Fowler, and J. A. Sabloff, pp. 439–457. Smithsonian Institution Press, Washington, DC.
Wedel, Waldo R.
1938 The direct-historical approach in Pawnee archaeology. Smithsonian Miscellaneous Collections 97(7): 1–21.
Weiss, Ehud, Mordechai E. Kislev, Orit Simchoni, Dani Nadel, and Hartmut Tschauner
2008 Plant-food preparation area on an Upper Paleolithic brush hut floor at Ohalo II, Israel. Journal of Archaeological Science 35(8): 2400–2414.
329
Weissbrod, L. T., Dayan D. Kaufman, and Mina Weinstein-Evron
2005 Micromammal taphonomy of el-Wad Terrance, Mount Carmel, Isreal: Distinguishing cultural from natural depositional agents in the Late Natufian. Journal of Archaeological Science 32: 1–17.
Whallon, Robert, William A. Lovis, and Robert K Hitchcock
2011 Information and Its Role in Hunter-Gatherer Bands. Ideas, Debates, and Perspectives 5. Cotsen Institute of Archaeology Press, UCLA.
Wheatley, David, and Mark Gillings
2002 Spatial technolgoy and archaeology: The archaeological applications of GIS. Taylor & Francis, New York.
Wheeler, Alwyne, and Andrew K.G. Jones
1989 Fishes. Cambridge Manuals in Archaeology. Cambridge University Press, Cambridge.
White, Theodore E.
1953 A Method of Calculating the Dietary Percentage of Various Food Animals Utilized by Aboriginal Peoples. American Antiquity 18(4): 396–398.
Whitridge, Peter
1992 Thule Subsistence and Optimal Diet: A Zooarchaeological Test of a Linear Programming Model. McGill University.
1999a The Construction of Social Difference in a Prehistoric Inuit Whaling Community. Arizona State University.
1999b The Prehistory of Inuit and Yupik Whale Use. Revista de Arqueologia Americana 16: 99–154.
2001 Zen Fish: A Consideration of the Discordance between Artifactual and Zooarchaeological Indicators of Thule Inuit Fish Use. Journal of Anthropological Archaeology 20(1): 3–72.
2002a Gender, Households, and the Material Construction of Social Difference: Metal Consumption at a Classic Thule Whaling Village. In Many Faces of Gender: Roles and Relationships, Past and Present, in Northern Indigenous Communities, edited by Lisa Frink, Rita S Shepard, and Gregory A Reinhardt, pp. 165–192. University of Calgary Press, Calgary.
2002b Social and ritual determinants of whale bone transport at a classic Thule winter site in the Canadian Arctic. Ed. Lisa Frink, Rita S Shepard, and Gregory A Reinhardt. International Journal of Osteoarchaeology 12(1): 65–75.
2004 Landscapes, Houses, Bodies, Things: “Place” and the Archaeology of Inuit Imaginaries. Archaeology 11(2).
330
Wilk, Richard R., and Robert McC. Netting
1984 Housholds: Changing forms and functions. In Households: Comparative and Historical Studies of the Domestic Group, edited by Robert McC. Netting, Richard R. Wilk, and Eric J. Arnould, pp. 1–28. University of California Press, Berkeley.
Wilk, Richard R., and William L. Rathje
1982 Household Archaeology. The American Behavioural Scientist 25(6): 617–639.
Willey, Gordon R.
1953 Prehistoric settlement patterns in the Viru Valley, Peru. U.S. Government Printing Office, Washington.
Willey, Gordon R., and Philip Phillips
1958 American Archaeology and General Anthropological Theory; Archaeological Unit Concepts; Archaeological Integration. In Method and Theory in American Archaeology, pp. 1– 57. The University of Chicago Press, Chicago.
Willis, Lauren M., Metin I. Eren, and Torben C. Rick
2008 Does butchering fish leave cut marks? Journal of Archaeological Science 35: 1438–1444.
Wilson, Bob
1996 Spatial patterning among animal bones in settlement archaeology: An English regional exploration. BAR British Series. Tempus Reparatum, Oxford.
Wilson, Don E., and DeeAnn M. Reeder (editors).
2005 Mammal Species of the World. A Taxonomic and Geographic Reference. 3rd ed. John Hopkins Press, Baltimore.
Wilson, G. D.
2008 The Archaeology of Everyday Life in Early Moundville. University of Alabama Press, Tuscaloosa.
Wing, Elizabeth S., and Antoinette B. Brown
1979 Paleonutrition: Method and theory in prehistoric foodways. Academic Press, New York.
Wisniewski, Josh
2005 “We’re always going back and forth” Kigiqtaamiut subsistence land use: Hunting, landscapes and recreating Iñupiat identify in Bering Strait. Report to the US AC Engr., pp. 15- 30. Alaska District, Anchorage.
331
2010 Come on Ugzruk, let me win: Experience, relationality, and knowing in Kigiqtaamiut hunting and ethnography. Ed. Igor Krupnik, Claudio Aporta, Shari Gearheard, Gita J. Laidler, and Lene Kielsen Holm. SIKU: Knowing Our Ice: Documenting Inuit Sea Ice Knowledge and Use. Springer Netherlands, Dordrecht.
Wissler, Clark
1916 Harpoons and Darts in the Stefansson Collection. Anthropological. Vol. XIV. Anthropological Papers of the American Museum of Natural History. New York.
1917 The American Indian. McMurtrie, New York.
Wobst, H. Martin
1978 The Archaeo-Ethnology of Hunter-Gatherers or the Tyranny of the Ethnographic Record in Archaeology. American Antiquity 43(2): 303–309.
Woodburn, James
1980 Hunters and gatherers today and reconstruction of the past. In Soviet and Western Anthropology, edited by Ernest Gellner, pp. 95–117. Duckworth, London.
2010 Oakes Bay 1: A Preliminary Reconstruction of a Labrador Inuit Seal Hunting Economy in the Context of Climate Change. Geografisk Tidsskrift-Danish Journal of Geography 110(2): 245–259.
Woollett, James M., Anne S. Henshaw, and Cameron P. Wake
2000 Palaeoecological Implications of Archaeological Seal Bone Assemblages: Case Studies from Labrador and Baffin Island. Arctic 53(4): 395–413.
Wylie, Alison
1982 An Analogy by Any Other Name Is Just as Analogical: A commentary on the Gould- Watson Dialogue. Journal of Anthropological Archaeology 1: 382–401.
1985 A Reaction against Analogy. Advances in Archaeological Method and Theory 8: 63–111.
1988 “Simple” analogy and the role of relevance assumptions: Implications of archaeological practice. International Studies in the Philosophy of Science 2(2): 134–150.
2002a Thinking from Things. University of California Press, Los Angeles.
2002b Bootstrapping in the Un-natural Sciences-Archaeology, for Example. In Thinking From Things: Essays in the Philosophy of Archaeology, pp. 179–184. University of California Press, Berkeley.
2002c The interpretive dilemma. In Thinking From Things: Essays in the Philosophy of Archaeology, pp. 117–126. University of California Press, Berkeley.
332
Wynne, Kate
2008 Guide to Marine Mammals of Alaska. Third. Alaska Sea Grant College Program, Fairbanks.
Yellen, John E.
1967 Settlement Patterns of the !Kung. An Archaeological Persepective. In Kalahari Hunter- Gatherers. Studies of the !Kung San and their Neighbors, edited by Richard B. Lee and Irven DeVore, pp. 47–72. Harvard University Press, Cambridge.
1977 Archaeological Approaches to the Present: Models for Reconstructing the Past. Academic Press, New York.
1991 Small Mammals: !Kung San Utilization and the Production of Faunal Assemblages. Journal of Anthropological Archaeology 10: 1–26.
Yesner, David R.
1985 Cultural boundaries and ecological frontiers in coastal regions: An example from the Alaskan Peninsula. In The Archaeology of Frontiers and Boundaries, edited by S. Green and S. Perlman, pp. 51–92. Academic Press, Toronto.
Young, Christopher E.
2002 Late Western Thule House construction in Northwest Alaska: The “Kobuk-Type” house. In Archaeology in the Bering Strait Region: Research on Two Continents, edited by Don E. Dumond and Richard L. Bland, pp. 207+226. University of Oregon Anthropological Papers. University of Oregon.
Zeder, Melinda A., and Susan R. Arter
1996 Meat Consumption and Bone Use in a Mississippian Village. In Case Studeis in Environmental Archaeology, edited by Elizabeth J. Reitz, C. Margaret Scarry, and Sylvia J. Scudder, pp. 337–355. 2nd ed. Springer, New York.
Zohar, Irit, and Miriam Belmaker
2005 Size does matter: methodological comments on sieve size and species richness in fishbone assemblages. Journal of Archaeological Science 32(4): 635–641.
333
Appendix A Overall Subsistence Analysis 1 Squares Used in General Subsistence Analysis Unit Unit Unit Unit Unit Unit South East Level Quad South East Level Quad South East Level Quad North ½ NE NE 2 2 3b SW NW NW 4 2 x3 5 3 x3d NE SE SE NW SW SW 2 2 3c SE No quad NW 5 4 3 SW NE SW North ½ 4 2 x3b NW 5 4 3b NW 2 3 x3b SE SE 5 4 3c NW SW SW NE SE NE 6 2 3 SE 2 3 x3c SW NW SW 4 2 x3c North ½ SE 6 3 x3 NW, NE, SE 2 4 x3b SE SW NW, NE, SE 6 3 x3b SW NE SW 2 4 x3c SE NW NW, NE, SE 4 3 x3b 6 3 x3c NE SE SW 3 1 3 SE SW NW 3 1 3b NE NE 6 3 x3d SE 4 3 x3c 3 2 3b SW SE SW 3 2 3c SW 4 4 x3 No quad NE NE No quad NW 7 2 3 NW NE SE 3 2 3d 4 4 x3b SE NW SW SW SW 7 5 3 SW NE NW 8 2 3 NE 4 4 x3c NW SW 8 5 3 SW 3 2 3e SE 4 4 x3d NW NW, NE, SE 14 2 2e SW No quad SW 3 3 x3b SE 5 2 3b NE NE NW NW No quad 3 3 x3c SE NE 5 2 3c SW NW NE SE NW NE 3 3 x3d 5 2 3d SE SW SW 5 2 3e SE NE NE NW NW 3 4 3c 5 3 x3 SE SE SW SW NW NE NE NW 3 4 3d 5 3 x3b SE SE SW SW NE 5 3 x3c NW SW
334
2 Calculations of MNE and MNI
The minimum number of elements (MNE) for each taxon was calculated based on specimen size, element, side, portion, completion, and age. The highest MNE for the taxa of that size gave the minimum number of individuals (MNI); MNIs based on size were added together to get the total taxa MNI. The tables below show the MNE used to calculate MNI for each species. Above the species level, MNI was only calculated if the MNE exceeded what could be accounted for in the species MNE count. Fill Floor Highest MNE Size Total Highest MNE Size Total Taxon Size (element, side) MNI MNI (element, side) MNI MNI Gastropoda Medium 1 shell 1 1 2 opercula, central 2 2 Bivalvia Medium 1 shell 1 1 2 hinges, left 2 2 Clupea pallasii Small 1 vertebra, central 1 1 Salvelinus alpinus Medium 1 vertebra, central 1 1 Coregoninae Large 1 vertebra, central 1 1 Salmonidae Large 1 basipterygium, ind. 1 0 Gadidae Small 1 certaohyal, right 1 1 Anser spp. Large 1 humerus, left 1 1 Goose Large 1 mandible, ind. 0 0 Large 1 tibiotarsus, left 1 Somateria spp. 2 Small 1 tibiotarsus, right 1 Melanitta spp. Medium 1 femur, left 1 1 Clangula hyemalis Medium 2 femurs, left 2 2 Medium 1 tibiotarsus, right 0 Duck 1 Small 1 tarsometatarsus, right 1 Anatidae Small 1 carpal, left 1 1 Medium 1 tarsometatarsus, right 1 Lagopus spp. 2 Small 1 tarsometatarsus, right 1 Calidris alpina Medium 1 humerus, right 1 1 1 humerus, right 1 1 Calidris spp. Medium 1 humerus, left 1 0 Gull Medium 1 coracoid, left 1 1 Sterna hirundo Medium 1 femur, left 1 1 Medium 2 ulnae, right 2 Sterna paradisaea 3 Small 1 ulna, left 1 Sterna spp. Medium 1 ulna, left 0 0 Medium 1 humerus, left 0 Tern 1 Small 1 ulna, left 1 Large 1 cranium, central 1 Spermophilus parryii Medium 1 femur, right 1 3 Small 1 femur, left 1 Castor canadensis Medium 1 clavicle, left 1 1 Medium 1 humerus, left 1 Ondatra zibethicus 2 Small 1 humerus, left 1 Medium 1 patella, ind. 1 Lepus spp. 2 Small 1 ulna, left 1 Large 1 metatarsal 5, right 1 3 Canis lupus Medium 1 medial phalanx, ind. 1 2 1 tarsal 1st, left 1 Small 1 patella, ind. 1 1 caudal vertebra, central 1 Vulpes spp. Medium 1 proximal phalanx, 1 2 2 premaxillae, left 2 3
335
ind. Small 1 metacarpal 2, right 1 1 scapula, right 1 Medium 1 metatarsal 1, left 1 1 tooth 1 Erignathus barbatus 2 2 Small 1 tooth 1 1 metatarsal 3 or 4, left 1 Histriophoca fasciata Medium 1 rib 2, right 1 1 Large 1 tibia-fibula, left 1 Phoca largha 2 Small 1 femur, left 1 Medium 1 femur, right 1 2 2 mandibles, left 2 Pusa hispida 3 Small 1 ulna, left 1 1 mandible, left 1 Medium 1 humerus, left 0 1 femur, right 0 Phoca/Pusa spp. 1 0 Small 1 ulna, left 1 1 ulna, right 0 Large 1 metatarsal 2, right 1 2 hind medial phalanges 5, 1 Phoca/Pusa/ left 4 4 Histriophoca spp. Medium 3 axes, central 2 4 metatarsals 4, left 2 Small 3 metatarsals 2, left 1 3 metatarsals 4, left 1 Neovison vison Medium 1 mandible, right 1 1 Large 1 scapula, left 1 Rangifer tarandus 1 2 Medium 1 femur, left 1 1 tibia, left 1 Ovibos moschatus Medium 1 tooth, right 1 1
Table A.1: MNE and MNI calculations for square S3E4, fill and floor levels Size Total Taxa Size Highest MNE (element, side) MNI MNI Gastropoda Medium 33 opercula, central 33 33 Medium 16 hinges, left 16 Bivalvia 19 Small 3 hinges, right 3 Clupea pallasii Small 3 vertebrae, central 1 1 Oncorhynchus spp. Large 1 vertebra, central 1 1 Salvelinus alpinus Medium 1 vertebra, central 1 1 Salvelinus spp. Small 1 vertebra, central 0 0 Coregoninae Large 1 vertebra, central 1 1 Large 1 opercle, left 0 Salmonidae Medium 2 cleithra, left 1 2 Small 2 hyomandibulars, left 1 Large 1 palatine, left 1 Gadus spp. 2 Small 1 otolith, ind. 1 Medium 3 preopercles, ind. 2 Gadidae 3 Small 3 otoliths, ind. 1 Anser spp. Large 1 humerus, left 1 1 Large 1 mandible, ind. 0 Goose 1 Medium 1 ulna, ind. 1 Cygnus columbianus Medium 1 humerus, ind. 1 1 Somateria spp. Medium 1 adult tarsometatarsus, left & 1 juvenile tibiotarsus, right 2 2 Melanitta spp. Medium 1 tibiotarsus, left 1 1 Large 1 keel, central 1 Clangula hyemalis 2 Medium 1 tibiotarsus, left 1 Large 1 cervical vertebrae, central 0 Duck 0 Medium 1 tibiotarsus, right 0 Large 1 mandible, left 0 Anatidae 0 Medium 1 thoracic vertebrae, central 0 Lagopus muta Medium 1 tarsometatarsus, right 1 1 Lagopus spp. Medium 1 tibiotarsus, left 0 0 Gavia stellate Medium 1 ulna, right 1 1 Gavia adamsii Large 1 ulna, right 1 1
336
Large 1 humerus, left 0 Gavia spp. 1 Small 1 ulna, left; 1 ulna, ind. 1 Podiceps spp. Medium 1 premaxilla, central 1 1 Buteo lagopus Large 1 tarsometatarsus, left 1 1 Grus canadensis Medium 1 premaxilla, central 1 1 Calidris alpina Medium 1 humerus, right 1 1 Calidris spp. Medium 1 humerus, right 1 1 Larus hyperboreus Medium 1 quadrate, left 1 1 Laridae Medium 1 quadrate, left 1 1 Sterna paradisaea Medium 1 femur, right 1 1 Large 1 femur, right 1 Sterna spp. 2 Small 1 mandible, ind. 1 Large 1 radius, left 1 Spermophilus parryii Medium 1 tibia, left 1 4 Small 1 adult femur, right & 1 juvenile femur, ind. 2 Castor canadensis Medium 1 clavicle, left 1 1 Lepus othus Large 1 radius, right 1 1 Lepus spp. Large 1 maxilla, indeterminate 1 0 Large 2 calcanea, left 2 Canis lupus Medium 1 metacarpal 3, right 1 6 Small 2 adult metacarpals 3, right & 1 juvenile patella, ind. 3 Medium 1 innominate, right 1 Vulpes vulpes 2 Small 1 mandible, right 1 Medium 2 radii, left 2 Vulpes lagopus 4 Small 2 humeri, left 2 Large 1 sternal segment, central 1 Vulpes spp. Medium 2 premaxillae, left 0 1 Small 2 scapulae, right 0 Large 1 distal phalanx, ind. 0 Canidae Medium 1 caudal vertebrae, central 0 0 Small 1 scapula, right 0 Ursus spp. Large 1 rib, left 1 1 Odobenus rosmarus Medium 1 atlas, central 1 1 Large 1 cuboid, right 1 Erignathus barbatus Medium 1 cuboid, left 1 4 Small 1 adult scapolunar, left & 1 juvenile metacarpal 3, left 2 Large 1 external cuneiform, right 1 Histriophoca fasciata 2 Medium 1 scapula, left 1 Large 1 ulna, right 1 Phoca largha Medium 3 ulnae, left 3 9 Small 5 femurs, left 5 Large 2 radii, right 2 20 Pusa hispida Medium 10 tibia-fibulas, left 10 Small 8 ulnae, left 8 Phoca/Histriophoca spp. Large 1 humerus, left 0 0 Large 1 tibia-fibula, right 0 Phoca/Pusa spp. Medium 5 tibiae-fibulae, left 2 3 Small 11 tibiae-fibulae, right 1 Large 43 hind medial phalanges 2 Pusa/Phoca/Histriophoca Medium 24 metatarsals 1, right 8 26 spp. Small 30 metatarsals 3, right 16 Phocidae Large 1 sternal segment 0 0 Neovison vison Medium 1 mandible, right 1 1 Alces alces Medium 2 atlases, right 2 2 Rangifer tarandus Large 2 ulnar carpals, left 2 5
337
Medium 2 mandibles, left 2 Small 1 mandible, right 1 Ovibos moschatus Medium 1 scapula, right 1 1 Balaena mysticetus Medium 1 humerus, indeterminate 1 1 Delphinapterus leucas Medium 1 phalanx, indeterminate 1 1 Table A.2: MNE and MNI calculations for combined floor material
3 Calculations of Meat Weight
Meat weights were calculated by multiplying the MNI for each species by an average weight for that species. This total live weight was then multiplied by a constant which is the proportion of the weight thought to be edible. This normally includes usable meat and fat but excludes skin, bone, and other inedible materials (Friesen and Arnold 1995a; Lyman 2008; Reitz and Wing 2008; White 1953). Constants differ for taxa; bird and mammal constants were derived from White (1953) and fish constants were derived from Wing and Brown (1979). Taxa weights were taken from Friesen and Arnold (1995a) and Betts and Friesen (2013). However, some species were not available and so were taken from Banfield (1974), following the calculations of Friesen and Arnold (1995a). Here, the average weights of adult male and female specimens were used as the standard mammal species weight. The only terrestrial mammal not available in Banfield (1974) is the Alaskan hare (Lepus othus); for this species, snowshoe hare (Lepus americanus) was used. The only other expectation was for marine mammals. Some marine mammals, such as walrus, are very different sizes in the Canadian and Alaskan Arctic (Banfield 1974; MacDonald and Cook 2010). The differences in small seal species is significant, with none of the small seal weights, except central Canadian ringed seals, provided in Banfield (1974; MacDonald and Cook 2010). For consistency, all marine mammals except whales (walrus, bearded seals, ringed seals, spotted seals, and ribbon seals) were calculated using Wynne (2008). The small seal category (Pusa/Phoca/Histriophoca spp.) meat weight was calculated based on the proportion each seal contributed to the overall small seal assemblage; MNIs for each of the small seal species were multiplied by the weight per individual for that species and all weights were added together then divided by the total MNI. This gave a weight per individual that was then multiplied by the MNI for small seals. Finally, bowhead whale meat weights were taken from Betts and Friesen (2013), where gross weight was determined based on a nine-metre yearling, with the edible portion at 37.5%, which follows data from McCartney (1995).
338
Weight per Edible Available % of total individual tissue % meat available Taxon MNI (gm) of weight (gm) meat Clupea pallasii 1 562 85 478 0.00 Oncorhynchus spp. 1 13600 85 11560 0.11 Salvelinus alpinus 1 4500 85 3825 0.04 Salvelinus spp. 0
Coregoninae 1 3900 85 3315 0.03 Salmonidae 2
Salminoformes 0
Gadus spp. 2 1000 85 1700 0.02 Gadidae 3 1000 85 2550 0.02
Anser spp. 1 2700 70 1890 0.02 Goose 1 2700 70 1890 0.02 Cygnus columbianus 1 7200 70 5040 0.05 Somateria spp. 1 1100 70 770 0.01 Melanitta spp. 1 1100 70 770 0.01 Clangula hyemalis 2 900 70 1260 0.01 Duck 0
Anatidae 0
Lagopus muta 1 600 70 420 0.00 Lagopus spp. 0
Gavia stellate 1 4500 70 3150 0.03 Gavia adamsii 1 4500 70 3150 0.03 Gavia spp. 2 4500 70 6300 0.06 Podiceps spp. 1 1000 70 700 0.01 Buteo lagopus 1 1060 70 742 0.01 Grus canadensis 1 4150 70 2905 0.03 Calidris alpina 1 56 70 39.2 0.00 Calidris spp. 1 56 70 39.2 0.00 Larus hyperboreus 1 2000 70 1400 0.01 Laridae 1 2000 70 1400 0.01 Sterna paradisaea 1 600 70 420 0.00 Sterna spp. 2 600 70 840 0.01
Spermophilus parryii 4 744 70 2083.2 0.02 Castor Canadensis 1 20000 50 10000 0.10 Lepus othus 1 4800 50 2400 0.02 Lepus spp. 0
Canis lupus 6 36250 50 108750 1.04 Vulpes vulpes 2 5200 50 5200 0.05 Vulpes lagopus 4 3200 50 6400 0.06 Vulpes spp. 1 4200 50 2100 0.02 Canidae 0
Ursus spp. 1 354071 70 247849.7 2.37 Odobenus rosmarus 1 1043000 70 730100 6.97 Erignathus barbatus 4 226800 70 635040 6.06 Histriophoca fasciata 2 150000 70 210000 2.00 Phoca largha 9 95250 70 600075 5.73 Phoca/Histriophoca spp. 0
Pusa hispida 20 54400 70 761600 7.27 Pusa/Phoca spp. 3 74825 70 157132.5 1.50 Pusa/Phoca/Histriophoca spp. 26 72427 70 1318171.4 12.58 Phocidae 0
Neovison vison 1 1500 70 1050 0.01 Alces alces 2 401500 50 401500 3.83
339
Rangifer tarandus 5 95350 50 238375 2.28 Ovibos moschatus 1 310000 50 155000 1.48 Balaena mysticetus 1 12126000 37.5 4547250 43.40 Delphinapterus leucas 1 400000 70 280000 2.67
Total 129 10476630
Table A.3: Meat weight calculations for all combined floor material
4 Calculations of MAU for General Subsistence Analysis
Minimum animal units (MAU) were determined by calculating the MNE for each element in each taxon. In this analysis, MAUs were calculated for the following taxonomic categories: fish, birds, caribou, dog/wolf, fox, bearded seal, and small seal. Unlike the calculations for MNE above that separated taxa into species, then into size, species were kept together at the higher taxonomic level for fish, birds, fox, and small seals and divided them only by size. Then, MNEs were calculated based on side, portion, and completion. The MNEs for each element in each size class were added together by side (i.e. all the left ulnas for large, medium, and small seals). The individual sides were then added together to get a sum MNE. This value was divided by the number of times that element occurred in the skeleton to get the MAU value. Some of the indices for bone density or utility use a set of elements rather than individual elements (i.e. carpals, metacarpals, and phalanges all make up a front flipper). For these sets, the MAU used was based on the highest individual MAU for the set; elements used for MAU calculations of element sets are recorded in the table. The tables below shows the sum MNE values and calculations to get to MAU for the taxa specified above.
4.1 Fish MAU Sum # per Only the fish elements with bone density and Element MNE individual MAU Angular 2 2 1 bone mineral content values (Butler and Chatters Ceratohyal 5 2 2.5 Opercle 2 2 1 1994) were used to calculate MAU values. There Otolith 7 2 3.5 Coracoid 1 2 0.5 is no standard number of vertebrae per skeleton, Basipterygium 1 2 0.5 Vertebra 172 50 3.44 so a high average of 50 was used. Hypural 2 1 2 Table A.4: Fish MAU calculations
340
4.2 Bird MAU
All bird taxa were combined; MAU was calculated based on this combined total for the elements with given average duck bone densities (Broughton et al. 2007). Again, cervical vertebrae are not standard, so an average of 15 was used to correspond with the duck average. Sum # per Element MNE individual MAU Cranium 2 1 2 Mandible 4 2 2 Cervical 10 15 0.67 Keel 3 1 3 Pygostyle 1 1 1 Scapula 1 2 0.5 Coracoid 5 2 2.5 Humerus 8 2 4 Ulna 6 2 3 Radius 2 2 1 Carpals 3 1 3 Carpometacarpus 4 2 2 Synsacrum 1 1 1 Femur 3 2 1.5 Tibiotarsus 6 2 3 Tarsometarsus 3 2 1.5 Table A.5: Bird MAU calculations
4.3 Dog/Wolf MAU
Dog/wolf (Canis lupus) MAU frequencies are calculated below. Sum # per Element MNE individual MAU Cranium 2 1 2 Mandible 2 2 1 Atlas-Axis (Atlas) 1 1 1 Cervical 3 5 0.6 Thoracic 9 13 0.69 Lumbar 4 7 0.57 Sacrum 0 1 0 Sternum 0 1 0 Rib 8 26 0.31 Innominate 2 2 1 Scapula 1 2 0.5 Humerus 2 2 1 Radius 2 2 1 Ulna 2 2 1 Carpals (C2) 3 2 1 Metacarpal (MC3) 4 2 2 Femur 2 2 1 Tibia 3 2 1.5 Fibula 2 2 1 Tarsals (T2) 4 2 2 Metatarsal (MT5) 4 2 2
341
Phalanges (medial) 12 16 0.75 Baculum 1 1 1 Table A.6: Dog/wolf MAU calculations
4.4 Fox MAU
All fox specimens were combined to produce a general fox skeletal frequency graph. The following frequencies were calculated. Sum # per Element MNE individual MAU Cranium 5 1 5 Mandible 1 2 0.5 Atlas-Axis (Atlas) 1 1 1 Cervical 1 5 0.2 Thoracic 4 13 0.31 Lumbar 3 7 0.43 Sacrum 1 1 1 Sternum 3 1 3 Rib 15 26 0.58 Innominate 3 2 1.5 Scapula 5 2 2.5 Humerus 3 2 1.5 Radius 4 2 2 Ulna 4 2 2 Carpals (C1) 1 2 0.5 Metacarpals (MC3) 2 2 1 Femur 5 2 2.5 Patella 1 2 0.5 Tibia 3 2 1.5 Fibula 0 2 0 Tarsals (talus) 2 2 1 Metatarsals (MT2) 3 2 1.5 Phalanges (medial) 4 16 0.25 Table A.7: Fox MAU calculations
4.5 Bearded Seal MAU Sum # per Element MNE individual MAU Cranium 0 1 0 Mandible 1 2 0.5 Atlas-Axis 0 1 0 Cervical 2 5 0.4 Thoracic 1 15 0.07 Lumbar 0 6 0 Sacrum 0 1 0 Sternum 1 9 0.11 Rib 15 30 0.5 Innominate 0 2 0 Scapula 2 2 1 Humerus 0 2 0 Radius 0 2 0 Ulna 0 2 0
342
Carpals (scapholunar) 2 2 1 Metacarpal (MC2) 2 2 1 Front phalanges (distal) 3 10 0.3 Femur 1 2 0.5 Tibia-Fibula 0 2 0 Tarsals (cuboid) 2 2 1 Metatarsals (MT3) 2 2 1 Hind phalanges (proximal) 4 10 0.4 Baculum 1 1 1 Table A.8: Bearded seal MAU calculations
4.6 Small Seal MAU Sum # per Element MNE individual MAU Cranium 10 1 10 Mandible 39 2 19.5 Cervical (Axis) 22 1 22 Thoracic 156 15 10.4 Lumbar 51 6 8.5 Sacrum 8 1 8 Sternum 116 9 12.89 Rib 510 30 17 Innominate 30 2 15 Scapula 39 2 19.5 Humerus 47 2 23.5 Radius 50 2 25 Ulna 53 2 26.5 Carpals (trapezium) 31 2 15.5 Metacarpal (MC1) 51 2 25.5 Front phalanges (medial) 151 8 18.88 Femur 48 2 24 Tibia-Fibula 60 2 30 Tarsals (navicular) 38 2 19 Metatarsal (MT3) 93 2 46.5 Hind phalanges (medial) 276 8 34.5 Baculum 4 1 4 Table A. 9: Small Seal MAU Calculations
4.7 Caribou MAU Sum # per Element MNE individual MAU Cranium 2 1 2 Mandible 4 2 2 Cervical (axis) 3 1 3 Thoracic 3 13 0.23 Lumbar 1 6 0.17 Sacrum 0 1 0 Sternum 0 1 0 Rib 11 26 0.42 Innominate 2 2 1 Scapula 4 2 2 Humerus 4 2 2 Radius-Ulna 5 2 2.5
343
Metacarpal 2 2 1 Carpals (radial) 4 2 2 Femur 4 2 2 Tibia 5 2 2.5 Fibula 1 2 0.5 Metatarsal 4 2 2 Tarsals (calcaneous) 3 2 1.5 Phalanges (proximal) 17 8 2.13 Table A.10: Caribou MAU Calculations
5 Bone Density and Utility Index Values Bone Mineral Meat Utility Density Index Element (BMD)* (%MUI)** Cranium 41.34 Mandible 1.11 8.81 Atlas 0.54 Axis 0.56 Cervical 0.35 58.95 Thoracic 0.37 35.48 Lumbar 0.38 53.12 Sacrum 0.43 Ribs 0.63 100 Innominate 0.75 58.68 Scapula 0.43 31.37 Humerus 0.67 20.95 Radius 0.71 15.45 Ulna 0.79 Front flipper 2.86 Femur 0.69 10.18 Tibia-Fibula 0.9 29.97 Calcaneous 0.45 Astragalus 0.56 Navicular 0.57 Cuboid 0.56 Rear flipper 4.76 * Bone density values from Chambers (1992) (in Lyman 1994) ** %MUI values from Diab (1998) Table A.11: Small Seal Bone Density and Meat Utility Values
Bone Food Meat Unsaturated Mineral Utility Drying Marrow Density Index Index Index Element (BMD)* (FUI)** (MDI)*** (UMI)**** Cranium 235 1.9 Mandible 1.07 590 56.2 Atlas-Axis 0.62 524 88.2 Cervical 0.45 1905 186.7 Thoracic 0.53 2433 311.3 Lumbar 0.51 1706 205.8 Sternum 3422 195.2 Rib 0.96 2650 754.4
344
Innominate 1.02 2531 196.8 Sacrum 0.4 Scapula 1.04 2295 89.5 Humerus 1.12 1891 18.5 22.8 Radius-Ulna 1.09 1323 16.4 26.3 Metacarpal 1.1 795 15.5 19.6 Carpals 0.72 0.9 Femur 1.15 5139 17 34 Tibia 1.13 3225 13 51.1 Fibula 0.68 Metatarsal 1.1 1903 11.2 46.5 Tarsals 0.9 Calcaneous 0.94 2.6 Phalanx-proximal 0.92 998 67.3 3.7 Phalanx-medial 0.72 1.8 Phalanx-distal 0.48 0.9 * Bone density values from Lam et al. (1999) ** Food utility values from Metcalfe and Jones (1988) *** Meat drying values from Friesen (2001) **** Unsaturated marrow values from Morin (2007) Table A.12: Caribou Bone Density and Utility Indices Values
Bone mineral Element content/ volume* Cranium 0.37 Mandible 0.41 Keel 0.2 Furculum 0.39 Scapula 0.48 Coracoid 0.53 Humerus 0.52 Ulna 0.49 Radius 0.34 Carpometacarpus 0.57 Synsacrum 0.26 Femur 0.33 Tibiotarsus 0.45 Tarsometatarsus 0.31 * Bone density values from average duck volume density in Broughton et al. (2007) Table A.13: Bird Bone Density Values
Bone Bone Mineral Density Content Element (VD)* (BMC)* Angular 0.2 0.23 Ceratohyal 0.06 0.1 Dentary 0.19 0.12 Exoccipital 0.11 0.17
345
Maxilla 0.2 0.24 Opercle 0.07 0.02 Otolith 1.41 0.07 Pterotic 0.12 0.17 Coracoid 0.07 0.04 Pectoral fin ray 0.29 0.03 Basipterygium 0.11 0.05 Vertebra 1 0.27 0.09 Vertebra 2 0.31 0.12 Vertebra 3 0.34 0.16 Vertebra 4 0.3 0.08 Hypural 0.14 0.07 * Bone density and bone mineral content values from chinook salmon in Butler and Chambers (1994) Table A.14: Fish Bone Density Values
346
Appendix B Expectations Testing and Spatial Patterning Analysis 1 Squares Used in Spatial Patterning Analysis South East Level Quad South East Level Quad South East Level Quad 2 2 3B SW NE NE NE NW NW 4 3 X3B 7 2 3 NW SE SE 2 2 3C SE SW SW SW NE 7 5 3 SW 4 3 X3C SE SE 8 5 3 SW 2 3 X3B SW NE 14 2 2B SE 4 4 X3B NW NW, NW, SE 2 3 X3C 14 2 2E SW SW SW SE NW 2 4 X3B 4 4 X3C SW SW 2 4 X3C SE 4 4 X3D NW NE NE 3 1 3 5 2 3B SE NW 3 1 3B NE NE 3 2 3C SW 5 2 3C NW NE SE NW NE 3 2 3D 5 2 3D SE SW SW 5 2 3E SE NE NE NW NW 3 2 3E 5 3 X3 SE SE SW SW 3 3 X3B SE NE NE NW 5 3 X3B NW SE 3 3 X3C SE SW SW NE NE 5 3 X3C NW NW SW 3 3 X3D SE NE SW NW 5 3 X3D NE SE NW SW 3 4 3C SE NW 5 4 3 SW SW NE 5 4 3B NW NW 5 4 3C NW 3 4 3D SE NE SW 6 2 3 SE NE SW NW 6 3 X3C SW 4 2 X3B SE NW SW 6 3 X3D SE NE SW NW 4 2 X3C SE SW
347
2 Quadrants per Archaeologically-Defined Area South East Quadrant South East Quadrant Tunnel Benches 5 2 SW 2 2 NW 5 2 SE 2 2 NE 5 3 SW 2 2 SW 5 3 SE 2 2 SE 5 4 SW 2 3 SW 6 2 NE 2 3 SE 6 3 NW 2 4 SW 6 2 SW 2 4 SE 6 2 SE 3 1 NE 6 3 SW 3 2 NW 6 3 SE 3 2 NE 7 2 NW 3 3 NW 7 2 NE 3 3 NE 7 2 SW 3 4 NW 7 2 SE 3 4 NE 3 1 SE Main Room Floors 3 2 SW 3 2 SE 3 4 SW 3 3 SW 3 4 SE 3 3 SE 4 2 NW 4 2 NE 4 4 NW 4 3 NW 4 4 NE 4 3 NE 4 2 SW 4 2 SE 4 4 SW 4 3 SW 5 2 NW 4 3 SE 5 4 NW 5 2 NE 5 3 NW Kitchen 5 3 NE 7 5 SW 8 5 SW
Midden 14 2 SW
3 Expectation Calculations
The expectation results were calculated in ArcGIS by querying the database for the required elements, taxa, modifications, etc. The resulting layer was used to calculate the sum of bones for each area. Ratios were calculated by dividing two sets of elements. Frequencies were calculated in each taxon by dividing the set of elements by the total number of specimens for those taxa.
The following tables show the ratios and frequencies for each area for each expectation. The household averages (Total row) and normal skeletal measures (Normal row) are also shown. They are grouped by taxa, with the general expectations at the end.
348
Baculum, Total Appendicular Axial Appendicular Indeterminate, Ivory Areas NISP % NISP Frequency NISP Frequency : Axial NISP Frequency Tunnel 5 0.23% 3 0.60 1 0.20 3.00 1 0.20 Main Room 22 0.94% 6 0.27 6 0.27 1.00 9 0.41 Kitchen 0 0% 0 0 0 0 0 0 0 Benches 70 1.49% 42 0.60 20 0.29 2.10 7 0.10 Midden 4 1.12% 2 0.50 1 0.25 2.00 1 0.25 Total 101 0.45% 53 0.52 28 0.28 1.89 18 0.18 Normal 191 118 0.62 73 0.38 1.62 Table B.1: Large Marine Mammal Ratios and Frequencies
Total Dense Less Dense Dense : Hind Flippers Less Area NISP % NISP Frequency NISP Frequency Dense NISP Frequency Tunnel 807 36.85% 239 0.30 225 0.28 1.06 237 0.29 Main Room 1090 46.38% 371 0.34 256 0.23 1.45 313 0.29 Kitchen 493 3.88% 89 0.18 155 0.31 0.57 105 0.21 Benches 2096 44.60% 643 0.31 478 0.23 1.35 689 0.33 Midden 170 47.62% 33 0.19 48 0.28 0.69 46 0.27 Total 4656 20.87% 1375 0.30 1162 0.25 1.18 1390 0.30 Normal 191 48 0.25 37 0.19 1.30 52 0.27
Hind Flippers & Head Front Flippers High Utility Low Utility Area NISP Frequency NISP Frequency NISP Frequency NISP Frequency Tunnel 245 0.30 120 0.15 279 0.35 476 0.59 Main Room Floor 316 0.29 170 0.16 404 0.37 638 0.59 Kitchen 109 0.22 136 0.28 148 0.30 287 0.58 Benches 715 0.34 313 0.15 682 0.33 1321 0.63 Midden 46 0.27 54 0.32 45 0.26 122 0.72 Total 1431 0.31 793 0.17 1558 0.33 2844 0.61 Normal 55 0.29 52 0.27 58 0.30 125 0.65
Small Bones Large Bones Small : Area NISP Frequency NISP Frequency Large Tunnel 617 0.76 189 0.23 3.26 Main Room Floor 768 0.70 321 0.29 2.39 Kitchen 473 0.96 640 0.31 2.27 Benches 1453 0.69 20 0.04 23.65 Midden 143 0.84 27 0.16 5.30 Total 3454 0.74 1197 0.26 2.89 Table B.2: Small Seal Ratios and Frequencies
349
Total High FUI Low FUI High : Lower Limbs Area NISP % NISP Frequency NISP Frequency Low FUI NISP Frequency Tunnel 27 1.23% 9 0.33 12 0.44 0.75 13 0.48 Main Room Floor 50 2.13% 26 0.52 18 0.36 1.44 18 0.36 Kitchen 3 0.02% 1 0.33 0 0 All high 0 0 Benches 97 2.06% 40 0.41 39 0.40 1.03 32 0.33 Midden 7 1.96% 3 0.43 3 0.43 1.00 2 0.29 Total 184 0.82% 79 0.43 72 0.39 1.10 65 0.35 Normal 119 74 0.62 45 0.38 1.64 50 0.42
Phalanges Limbs High Density Low Density High : Phalanges Low Area NISP Frequency NISP Frequency : Limbs NISP Frequency NISP Frequency Density Tunnel 8 0.30 19 0.70 0.42 13 0.48 8 0.30 1.63 Main Room 7 0.14 29 0.58 0.24 33 0.66 10 0.20 3.30 Kitchen 0 0 0 0 0 1 0.33 0 0 All high Benches 13 0.13 49 0.51 0.27 60 0.62 17 0.18 3.53 Midden 1 0.14 3 0.43 0.33 4 0.57 2 0.29 2.00 Total 29 0.16 100 0.54 0.29 111 0.60 37 0.20 3.00 Normal 24 0.20 60 0.50 0.40 62 0.52 59 0.50 1.05
Small Bones & Small : High FUI : Small Bones Large Bones Low FUI Large Small Bones Antler Area NISP Frequency NISP Frequency NISP Frequency Elements & Low FUI NISP Frequency Tunnel 13 0.48 14 0.52 8 0.30 0.93 1.13 0 0 Main Room 27 0.54 23 0.46 14 0.28 1.17 1.86 0 0 Kitchen 3 1.00 0 0 0 0 All small All high 0 0 Benches 49 0.51 48 0.49 25 0.26 1.02 1.60 4 0.04 Midden 1 0.14 6 0.86 1 0.14 0.17 3.00 1 0.14 Total 93 0.51 91 0.49 48 0.26 1.02 1.65 5 0.03 Table B.3: Large Terrestrial Mammal Ratios and Frequencies
350
Total Dense Less Dense Dense : Distal Limbs Less Area NISP % NISP Frequency NISP Frequency Dense NISP Frequency Tunnel 18 0.82% 11 0.61 6 0.33 1.83 15 0.83 Main Room 150 6.38% 74 0.49 45 0.30 1.64 101 0.67 Kitchen 2 0.02% 2 1.00 0 0 All dense 2 1.00 Benches 150 3.19% 50 0.33 62 0.41 0.81 37 0.25 Midden 3 0.84% 0 0 2 0.67 0 2 0.67 Total 323 1.45% 137 0.42 115 0.36 1.19 157 0.49 Normal 122 63 0.20 110 0.34 0.57 82 0.67
Mandibles Small Large Small : Cortical Area NISP Frequency NISP Frequency NISP Frequency Large NISP Frequency Tunnel 0 0.00 14 0.78 4 0.22 3.50 0 0 Main Room 0 0.00 121 0.81 29 0.19 4.17 2 0.01 Kitchen 0 0.00 2 1.00 0 0.00 All small 0 0 Benches 4 0.03 91 0.61 59 0.39 1.54 3 0.02 Midden 0 0.00 3 1.00 0 0.00 All small 0 0 Total 4 0.01 231 0.72 92 0.28 2.51 5 0.02 Normal 2 0.02
Cancellous Cortical : Whole Cortical Area NISP Frequency Cancellous NISP Frequency : Whole Tunnel 0 0 None 17 0.94 All whole Main Room 10 0.07 0.20 130 0.87 0.02 Kitchen 0 0 None 2 1.00 All whole Benches 0 0 All cortical 128 0.85 0.02 Midden 0 0 None 2 0.67 All whole Total 10 0.03 0.50 279 0.86 0.02 Table B.4: Small Terrestrial Mammal Ratios and Frequencies
Total Dense Less Dense Dense Lower Limbs Wings & Legs : Less Area NISP % NISP Frequency NISP Frequency Dense NISP Frequency NISP Frequency Tunnel 34 1.55% 9 0.26 10 0.29 0.90 7 0.21 13 0.38 Main Room 25 1.06% 6 0.24 7 0.28 0.86 0 0 4 0.16 Kitchen 2 0.02% 0 0 0 0 None 0 0 0 0 Benches 57 1.21% 14 0.25 9 0.16 1.56 11 0.19 21 0.37 Midden 3 0.84% 0 0 1 0.33 0 3 1.00 3 1.00 Total 121 0.54% 29 0.24 27 0.22 1.07 21 0.17 41 0.34 Normal 110 12 0.11 11 0.10 1.09 30 0.27 40 0.36
Small Bones Large Bones Small : Cortical Whole Cortical : Area NISP Frequency NISP Frequency Large NISP Frequency NISP Frequency Whole Tunnel 33 0.97 1 0.03 1122.00 4 0.12 28 0.82 0.14 Main Room 22 0.88 3 0.12 183.33 3 0.12 15 0.60 0.20 Kitchen 2 1.00 0 0.00 None 0 0.00 2 1.00 0.00 Benches 45 0.79 9 0.16 285.00 12 0.21 36 0.63 0.33 Midden 3 1.00 0 0.00 All small 0 0.00 3 1.00 0.00 Total 105 0.87 13 0.11 977.31 19 0.16 84 0.69 0.23 Table B.5: Bird Ratios and Frequencies
351
Total Cranial Post-Cranial Cranial Ribs Vertebrae : Post- Area NISP % NISP Frequency NISP Frequency Cranial NISP Frequency NISP Frequency Tunnel 10 0.46% 2 0.20 4 0.40 0.50 0 0.00 4 0.40 Main Room 27 1.15% 2 0.07 9 0.33 0.22 0 0.00 7 0.26 Kitchen 128 1.01% 35 0.27 55 0.43 0.64 11 0.09 42 0.33 Benches 189 4.02% 31 0.16 129 0.68 0.24 10 0.05 118 0.62 Midden 0 0.00% 0 None 0 None None 0 None 0 None Total 354 1.59% 70 0.20 197 0.56 0.36 21 0.06 171 0.48 Table B.6: Fish Ratios and Frequencies
Total Butchery Gnaw Marks Weathering Area NISP % NISP Frequency NISP Frequency NISP Frequency Tunnel 2190 9.82% 7 0.00 33 0.02 62 0.03 Main Room Floor 2350 10.53% 23 0.01 42 0.02 31 0.01 Kitchen 12711 56.98% 1 0.00 1 0.00 3 0.00 Benches 4700 21.07% 36 0.01 87 0.02 41 0.01 Midden 357 1.60% 1 0.00 6 0.02 3 0.01 Total 22308 100.00% 68 0.00 169 0.01 140 0.01
Total Burnt Fish Burnt Mammal Burnt LMM Burnt Area NISP Frequency NISP Frequency NISP Frequency NISP Frequency Tunnel 542 0.25 0 None 328 0.15 0 None Main Room Floor 299 0.13 0 None 240 0.10 0 None Kitchen 11745 0.92 0 None 1808 0.14 0 None Benches 1037 0.22 2 0.00 442 0.09 1 0.00 Midden 42 0.12 0 None 42 0.12 0 None Total 13665 0.61 2 0.00 2820 0.13 1 0.00
Small Seal Burnt LTM Burnt STM Burnt Indeterminate Burnt Area NISP Frequency NISP Frequency NISP Frequency NISP Frequency Tunnel 56 0.03 0 None 1 0.00 214 0.10 Main Room Floor 23 0.01 0 None 0 None 59 0.03 Kitchen 258 0.02 0 None 0 None 9937 0.78 Benches 49 0.01 1 0.00 0 None 633 0.13 Midden 6 0.02 0 None 0 None 0 None Total 392 0.02 1 0.00 1 0.00 10843 0.49
Indeterminate Small Bones Large Bones Small : Large : Mammal Indeterminate Area NISP Frequency NISP Frequency Large Small NISP Frequency NISP Frequency Tunnel 1957 0.89 232 0.11 8.44 0.12 609 0.28 475 0.22 Main Room 1914 0.81 435 0.19 4.40 0.23 623 0.27 258 0.11 Kitchen 12688 1.00 23 0.00 551.65 0.00 1698 0.13 10300 0.81 Benches 3818 0.81 875 0.19 4.36 0.23 839 0.18 393 0.08 Midden 320 0.90 37 0.10 8.65 0.12 125 0.35 42 0.12 Total 20697 0.93 1602 0.07 12.92 0.08 3894 0.17 12014 0.54 Table B.7: General Ratios and Frequencies
352
4 Elements Used in Calculations
The elements used to calculate each of the ratios or frequencies are tabulated below. They are organized by taxa; species that make up the taxa are identified for each section. When elements are obvious (i.e. small terrestrial mammal mandibles), they are not repeated in these tables. Category Elements
Large Marine Mammals Order: Cetacea and species Erignathus barbatus and Odobenus rosmarus Appendicular Humerus, radius, ulna, carpals, femur, tibia-fibula, tarsals, metacarpals, metatarsals, and phalanges Axial Craniums, mandibles, vertebrae, sternum, sacrum, innominates, and ribs Non-Subsistence Elements Baculum, indeterminate whale, and ivory
Species: Histriophoca fasciata, Phoca largha, Pusa hispida, Phoca/Histriophoca spp., Pusa/Phoca Small Seals spp., and small seal High Density Elements Mandible, tibia, fibula, ulna, innominate, radius, femur, humerus, rib, and scapula Low Density Elements Navicular, astragalus, axis, cuboid, atlas, calcaneous, sacrum, vertebrae, carpals, and tarsals Hind Flippers Tarsals, metatarsals, and hind phalanges Small All specimens <5cm Large All specimens >5cm Front Flippers Carpals, metacarpals, and front phalanges High Utility Elements Ribs, cervical vertebrae, sacrum, innominate, lumbar vertebrae, cranium, thoracic vertebrae, and scapula Low Utility Elements Tibia, fibula, humerus, radius, ulna, femur, mandibles, sternum, carpals, tarsals, metapodials, and phalanges
Large Terrestrial Mammals Order: Artiodactyla and Family Ursidae High FUI Elements Femur, sternum, tibia, tarsals, ribs, innominates, sacrum, thoracic vertebrae, scapula, cervical vertebrae, metatarsal, humerus, and lumbar vertebrae Low FUI Elements Radio-ulna, phalanges, metacarpals, carpals, mandibles, atlas, axis, and cranium Lower Limbs Metapodials, tarsals, carpals, and phalanges Limbs Humerus, radio-ulna, femur, tibia, metapodials, carpals, tarsals, sesmoids, and phalanges Small Any specimen <5cm Large Any specimen >5cm High Density Elements Femur, tibia, humerus, metapodials, radio-ulna, mandible, scapula, innominate, ribs, tarsals, and proximal phalanges Low Density Elements Medial and distal phalanges, carpals, fibula, atlas, axis, vertebrae, and sacrum
Small Terrestrial Mammals Order: Lagomorpha and Rodentia; Family: Canidae and Mustelidae High Density Elements From canid: tibia, femur, radius, humerus, ulna, tarsals, mandible, innominate, cervical vertebrae, fibula, metapodials, and carpals Low Density Elements From canid: phalanges, lumbar vertebrae, scapula, thoracic vertebrae, sternum, sacrum, and ribs Distal Elements Tibia, fibula, ulna, radius, metapodials, carpals, tarsals, and phalanges Small All specimens <5cm Large All specimens >5cm Cortical Bone All bone types identified as cortical or diaphysis Cancellous Bone All bone types identified as cancellous Whole Bone All bone types identified as whole
Birds Class: Aves High Density Elements Carpometacarpus, coracoid, humerus, ulna, scapula, and tibiotarsus
353
Low Density Elements Mandible, furculum, cranium, radius, femur, tarsometatarsus, synsacrum, and keel Small All specimens <5cm Large All specimens >5cm Lower Limbs Carpometacarpus, carpals, tarsometatarsus, and phalanges Wings and Legs Humerus, radius, ulna, carpometacarpus, carpals, femur, tibiotarsus, tarsometatarsus, and phalanges Cortical Bone All bone types identified as cortical and diaphysis Whole Bone All bone types identified as whole
Fish Class: Actinopterygii Cranial Elements Angular, basipterygium, brachiostegal rays, ceratohyal, cleithrum, basioccipital, coracoid, cranial, hyomandibular, opercula, opsithotic, otolith, palatine, parashpenoid, preopercule, pterygoid, quadrate, and vomer Post-Cranial Elements Vertebrae, pterygiopheres, rib/ray/spine, epibrachials, and hypural
General All Taxa Small All specimens <5cm Large All specimens >5cm Indeterminate Class: Mammalia/Aves and Indeterminate Indeterminate Mammal Class: Mammalia; Order: Indeterminate