Material Cultural Correlates of the Athapaskan Expansion


Material Information

Material Cultural Correlates of the Athapaskan Expansion A Cross-Disciplinary Approach
Physical Description:
1 online resource (353 p.)
Wilson,Joseph Andrew Park
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Committee Chair:
Collings, Peter F
Committee Members:
Moore, John H
Sassaman, Kenneth E
Barr, Juliana


Subjects / Keywords:
apache -- archaeology -- archery -- athapaskan -- denaina -- ethnohistory -- ethnology -- genetics -- gwichin -- hupa -- ket -- linguistics -- navajo -- sekani -- tsuutina -- yeniseian
Anthropology -- Dissertations, Academic -- UF
Anthropology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation


Contrary to stereotypes of proto-Athapaskan culture as simplistic and archaic, evidence points to a sophisticated web of late prehistoric Asian-Athapaskan interactions. A holistic assessment of Athapaskan migrations in the context of the transpacific Dene-Yeneseian phylum (the largest, fastest pedestrian language spread on earth) sees Athapaskan-Asian connections (in language, technology, DNA, social organization, etc.) as reflecting profound large-scale cultural-historical processes whose implications have yet to be grasped. Current understanding is that Athapaskans slowly migrated south in response to volcanic eruptions in southwest Yukon Territory after circa 200 and 800 CE. Yet problems remain, notably the archaeological invisibility of migrants on their long trek southward, and their possession of Asiatic strong complex bows which were not introduced to Northern Athapaskan territory until after these two eruptions. Linguistics, archaeology, biology, and data from ethnographic archery collections suggest Athapaskans carried sinew-backed bows to California and the Southwest. Both Apacheans and Northern Athapaskans uniquely possessed both ?Arctic?- and ?Plains?-style sinew-backed bows. Migration with retention (not diffusion through existing populations) is the best explanation. The Athapaskan expansion was faster than generally supposed, quite similar to the contemporaneous Punuk/Thule Neo-Eskimo expansion in the far north. Such a model helps to explain archaeological invisibility in the intervening space, as the impact made by the migrants was small.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by Joseph Andrew Park Wilson.
Thesis (Ph.D.)--University of Florida, 2011.
Adviser: Collings, Peter F.

Record Information

Source Institution:
Rights Management:
Applicable rights reserved.
lcc - LD1780 2011
System ID:

This item is only available as the following downloads:

Full Text




2011 Joseph Andrew Park Wilson 2


To my life partner, Michelle Trim, without whom none of this would be possible 3


ACKNOWLEDGMENTS I thank Peter Collings for his stellar mentorship, and John Moore, Ken Sassaman and Juliana Barr for their moral support and encouragement as the ideal members of my supervisory committee. I thank Victor Golla, Jack Ives, Roland Bohr, Jim Hamm and Don Dumond for helpful communication and permission to reproduce images. I thank many museum curators, collections managers, and staff for accommodating me, and apologize to those whose names I have forgotten. In chronological order of my fieldwork, I thank Kyle E. Bryner, Elise V. LeCompte, Michael H. Logan, Gerald R. Shroedel, Jefferson Chapman, Robert Pennington, anonymous and spouse, Ka thryn Barr, Daniel Swan, Eric Singleton, Randy Ramer (to the last two names, special thanks for facilitating a visit on short notice), Patricia L. Nietfeld, Thomas E. Evans, Laila Williamson, Elise Alexander, Roger Colton, Gordon Ambrosino, Judy Thompson, Dawn Scher Thomae, Nancy O. Lurie, Karen Lacy, Natasha Johnson, Bradley Marshall, Brian Seymour, Yuri Berezkin, Kelly Fenn Sparks, Felicia Pickering, Michael Frank, Deni J. Se ymour, David V. Hill, Kalinovskaya Ekaterina, Candace Sall, and Janaki Krishna. I thank my spouse, Dr. Michelle Trim for encouraging me to apply to the University of Florida in the first place. I gratefully acknowledge the University of Florida for the provision of a four-year Alumni Fellowship, furnishing me the financial resources necessary to complete my degree on schedule. I thank the libraries (and particularly the interlibrary loan staff) at both the University of Florida and Elon University (NC), for providing me with invaluable assistance tracking down obscure publ ications. And finally I thank the folks at Google Books and the Internet Archive for their free public service and pioneering efforts of digitally archiving and making searchable many obscure ethnographic resources published during the last 150 years. 4


TABLE OF CONTENTS page ACKNOWLEDG MENTS .................................................................................................. 4LIST OF TABLES ............................................................................................................ 9LIST OF FI GURES ........................................................................................................ 10LIST OF ABBR EVIATION S ........................................................................................... 13ABSTRACT ................................................................................................................... 14 CHAPTER 1 INTRODUC TION .................................................................................................... 16The Research Questi on .......................................................................................... 16Cultural Evolutionary versus Historical Models ................................................. 19Athapaskan-Asian Connections ....................................................................... 21Ethnographic Ev idence .................................................................................... 23The Structure of the St udy ...................................................................................... 262 THE HISTORY OF ATHAPASKAN ST UDIES ........................................................ 27The Linguistic Foundati ons ..................................................................................... 27Na-Dene and SinoTibetan ............................................................................... 29Timing the Athapaskan Expansion using Lexicostatistics or Glottochronology ........................................................................................... 31The North-South Direction of the Athapaskan Ex pansion ................................ 32Chronology of Athapa skan Orig ins ................................................................... 34Location of the ProtoAthapaskan Home land ................................................... 43The Alternative Long Ch ronology Model for At hapaskan Orig ins. .................. 44Archaeological Evidence fo r Athapaskan Origins ................................................... 46Evidence for Interi or Orig ins ............................................................................. 46The Problem of Archaeol ogical Visi bility .......................................................... 47A Yellowknife Chipew yan Except ion? .............................................................. 51The White River Ash Fall .................................................................................. 54Ethnographic Reconstruction of Proto-Athapaskan Culture .................................... 57Eschatology: Deat h and Rebirth ....................................................................... 59Residence Patterns and Familial De scent ........................................................ 61Female Fortitude and Spirit ual Potency in Menstura tion and Childbirth ........... 63Physical Anthropology and the Athapa skans .......................................................... 66Backgroun d ...................................................................................................... 66Biometrics and Craniofac ial Analys is ............................................................... 71 5


3 THE DENE-YENISEIAN FAMILY IN ANTHROPOLOGICAL PERSPECTIVE ........ 77Pushing the Linguistic Boundaries: the Dene-Yeniseian Phylum ............................ 77Yeniseian, Tibetan and Na-D ene ..................................................................... 78Implications Regarding t he Athapaskan Ex pansion ......................................... 81The Historical Range of Yeniseian S peakers ................................................... 82Asian-Athapaskan Eth nological Links ..................................................................... 86Eschatology ...................................................................................................... 86Matrilineal-Matrilocal So cial Organi zation......................................................... 92Menstruation and Childbirth .............................................................................. 92Transpersonal Extrasomatic Powe r of Objects and Animal s ............................ 94Eursian Parallels for Nbi'bibish, the Lightnings Knife ..................................... 95A Siberian Origin for Athapask an Copper Metallu rgy? ................................... 100Molecular Genetics and the D ene-Yeniseian Fa mily ............................................ 102Human Leukocyte An tigen Marker s ............................................................... 104Autosomal Recessive Diseases ..................................................................... 107Albumin Nask api ............................................................................................ 109Mitochondrial DNA ......................................................................................... 112An Earlier Study of Dene-Yeni seian Population Genetics .............................. 115Mitochondrial Hapl ogroup A2a ....................................................................... 118Male Mediated Mi gration ................................................................................ 120Y-Chromosomes ............................................................................................ 122Haplogroup Q ................................................................................................. 123Haplogroup C ................................................................................................. 124Haplogroup-C among Greenl andic Eski mos .................................................. 125Haplogroup C3 among the Cheyenne, Arapaho, Sioux and Ojibwa ............... 127C3 in the S outheast? ...................................................................................... 129C3*-M217 in Sout h Americ a ........................................................................... 132Haplogroup R1 and Post-Col umbian Admixt ure ............................................. 133Summary .............................................................................................................. 1414 SELECTION OF FIELD-SITES A ND METHODS OF ANALYSIS ......................... 149Premise for Collectio ns Resear ch ......................................................................... 149Overview ........................................................................................................ 149Evolutionary Anthropology: Ethnogeneti cs versus Phyl ogenetics .................. 150Basic Method Em ployed ....................................................................................... 154Sample Size and Number of Institutions......................................................... 155Ethnographic and Archaeological Material s in Museum Co llections .............. 158Strong Complex Bow Document ation and Measur ement ..................................... 159Basic Bow Features and Vocabula ry .............................................................. 161Wood Type ..................................................................................................... 163Units of Measure and Measurement Protocols ............................................... 164Bow Typol ogy ....................................................................................................... 167Museums Visi ted .................................................................................................. 169Wake Forest University, Museum of Anthropol ogy (WFU) ............................. 170Florida Museum of Natu ral History (UFL) ....................................................... 170 6


Frank H. McClung Mu seum (UTK ) ................................................................. 170Sam Noble Oklahoma Museum of Natural Histor y (SNO) .............................. 171Gilcrease Museum of t he Americas (GMA) .................................................... 171National Museum of the American Indian (N MAI) .......................................... 172American Museum of Natura l History (A MNH) ............................................... 173Peabody Museum of Natura l History-Yale (YPM) .......................................... 174Field Museum of Natural History (FMNH) ....................................................... 175Milwaukee Public Mu seum (MPM) ................................................................. 176San Diego Museum of M an (JES or SD M) ..................................................... 176Phoebe A. Hearst Museum of Anthropology (PAH) ........................................ 178Museums Not Vi sited ............................................................................................ 180University of Missouri, Museum of Anthropology (MAC) ................................ 180National Museum of Natura l History, (N MNH) ................................................ 181Peter the Great Museum of Anth ropology and Ethnogr aphy (PGM) .............. 182Royal British Columbia Museum (RCBM) ....................................................... 182State Hermitage Mus eum (SHM ) ................................................................... 182Bow Morphology, Material Cult ure, and Athapaskan Migrations ........................... 183Summary .............................................................................................................. 1855 COMPLEX ARCHERY AMONG ATHAPAS KANS AND THEIR NEIGHBORS ..... 192Overview ............................................................................................................... 192Constitution of the Dataset .................................................................................... 193The Reliability of Cult ural Identifications in Museum Re cords .............................. 194Basic Typological Divisions of the Da taset ........................................................... 196Unfinished and Fragment ary Bows ................................................................ 197Complex Bows versus Self Bo ws ................................................................... 198Horn Bows, Compound Bows, and True Composite Bows ............................. 199The Old World Origins of the Composit e Bow ................................................ 201Athapaskan Guard Bows ................................................................................ 203The Inclusion of Self-Bows in a Study of Comp lex Bows ...................................... 204Cross-sectional Shape Pr ofiles for Bo ws .............................................................. 205Intermediate Forms ........................................................................................ 205Tri-local Classification System ........................................................................ 207Morphological Data for North American Comple x Bows ....................................... 207The Abundance of Athapaskan Glue-Type Complex Bows ................................. 208Demographic Ad vantage ................................................................................ 210The Possibility of Deliberate Misidentification ................................................. 211Longer Retention of Traditi onal Lifestyles ...................................................... 213Entrepreneurism or Mercant ile Opportuni sm .................................................. 216Discussion of Complex Archery Da ta ................................................................... 217Southern Athapaskan Co mplex Bows ............................................................ 217Pacific Coast Athapaskan and Al gic Complex Bows ...................................... 223Athapaskan-Numic Cultural Reticulation and Complex Archery Diffusion ...... 225Great Basin Archaeologica l Bow Fragments: Numic or Athapa skan? .................. 227Promontory Poin t, Utah .................................................................................. 227Lovelock Cave, Nevada.................................................................................. 230 7


Northern Athapaskan Co mplex Bows ................................................................... 232Denaina Bo ws ............................................................................................... 232Other Northern Athapaskan Bows .................................................................. 235Summary .............................................................................................................. 2386 CONCLUSION: PROTO-ATHAPASKAN AN D DENE-YENISEIAN MATERIAL CULTURE ............................................................................................................. 259Cross-Disciplinary Synthesis ................................................................................ 259Genes, Languages, and Materi al Cultur e ....................................................... 259Historical Scenarios Accounting for Male Mediated Mi gration ........................ 261Linguistic Horizons: Proto-Dene-Yenise ian Words for Technologies of Late Holocene Acquisi tion ......................................................................................... 265Proto-Athapaskan Terms fo r Metal and Kn ife ....................................................... 267Archeological Ho rizons ......................................................................................... 275Proto-Athapaskan Pott ery Word s ................................................................... 282Hair-tempered Pottery in the New Wo rld ........................................................ 285Corroborating Evidence for Late Disper sals from Wester n Alaska ................. 288Maping Vestigial Traits: the Att enuation of Comple x Bows ................................... 290Final Thou ghts ...................................................................................................... 296APPENDIX: INTERNAL CLASSIFICATION OF THE ATHAPASKAN LANGUAGES .. 306LIST OF REFE RENCES ............................................................................................. 308BIOGRAPHICAL SK ETCH .......................................................................................... 353 8


LIST OF TABLES Table page 4-1 Summary of bow s studied. ............................................................................... 1875-1 Objects whose provenances were substantially revised or corrected as a result of this study. ............................................................................................ 2415-2 Objects whose provenances were substantially revised or corrected as a result of this study. ............................................................................................ 2425-3 Southern Athapaskan bow morphol ogy. ........................................................... 2435-4 Southern Athapaskan bow material compon ents. ............................................ 2445-5 Southern Athapaskan bow length measurements (i nches). ............................. 2445-6 Pacific Coast Athapaskan, Yurok and Pomo Bow length and maximum width (inches). ............................................................................................................ 2455-7 Pacific Coast Athapaskan, Yuro k and Pomo paint colors ................................. 2455-8 Pacific Coast Athapaskan, Yurok and Po mo cross-sectional bow shape. ........ 2465-9 Pacific Coast Athapaskan, Yurok and Pomo bow grip tr eatments. ................... 2465-10 Numic bow length and wid th distribut ions. ........................................................ 2475-11 Numic bow cross-secti onal shape prof iles. ....................................................... 2476-1 Metal/Knife terms in Proto-Athapask an, Eyak and Eurasian languages.. ......... 2996-2 Muskogean, Yukian and P enutian archery terms. ............................................ 3006-3 Siberian Uralic and California Penutian arch ery lexicons ................................. 3006-4 Athapaskan pottery terms ................................................................................. 300 9


LIST OF FIGURES Figure page 2-1 Na-Dene in t he Northwest .................................................................................. 763-1 Tentatave (Sino?)-DeneYeniseian dendr ogram. ............................................. 1433-2 Map of Yeniseian territory in eighteenth century (green), and the Xiongnu Empire circa 2000 year s ago (orange) .. ........................................................... 1433-3 Apache umbilical pouch stitched with l eather and horsehair, white shell and turquoise pendants, early twentieth-cent ury. .................................................... 1443-4 Comparison of forged daggers of Athapaskan Subarctic Canada and the Upper Yenisei Riv er, Siber ia............................................................................. 1443-5 Comparison of tanged-type forged me tal arrowheads of Eurasia and the Athapaskan Subar ctic ....................................................................................... 1453-6 Native Siberians in t he mid-twentieth -century................................................... 1453-7 Athapaskan/Algic Interface in Northwest Ca lifornia .......................................... 1463-8 Na-Dene language distribution in comparison with Haplogroup C3 distributi on. ....................................................................................................... 1463-9 Median-joining microsatellite netwo rk for haplogroups CM217* and C3b. ........ 1473-10 Athapaskan and Southwestern Y-chromosome haplogroup frequency distributi on. ....................................................................................................... 1484-1 The tree of life and the tree of knowledge of good and evilthat is, of human culture ............................................................................................................... 1884-2 Three common sinew-backed bow s hapes observed by t he author.. ............... 1884-3 Six common cross-sectional shapes of the bow ............................................... 1894-4 Sketches of several different nocktypes, grouped by the culture areas where they commonly appear. .................................................................................... 1894-5 Strong complex bows in North Americ a, in comparison to Na-Dene language distributio n ........................................................................................................ 1904-6 Catalog entry for mail-order purc hase of Hupa or Yurok bows from the general store of the Hoopa Va lley Reservat ion. ............................................... 1914-7 WFU 1993.02.E.3. Navajo or Pueblo arrowhead, collected in the 1940s ......... 191 10


4-8 UTK MM.6.B4.B1.1:21G/ 46. Nineteenth-century metal-tipped Apache arrow .. 1915-1 Three common surface treatments for t he backing material of complex bows. 2485-2 Comparison of archaeological bow to regional ethnographic specimens. ........ 2495-3 Detail of Peter Pond s Map of 1785. ................................................................. 2505-4 RBCM 6563. Central Canadian Cree or Athapask an sinew-backed compound bow. ................................................................................................ 2505-5 Spliced antler belly laths from compos ite segment bow. Neolithic, late third millennium BCE. Angara River, Yenisei Basin, Lake Baikal region, Siberia. .... 2505-6 NMAI 13/1513. Shoshone sinew-backed true composite bow .......................... 2515-7 PGM 2667-20. Denaina Athapaskan sinew-backed bow from Kenai Peninsula, Alaska, with hair fr inge and attached wr ist guard ............................ 2515-8 Western versus eastern style atta ched wrist guards for Northern Athapaskan bows ................................................................................................................. 2515-9 JES 141; Pomo sinew-backed bow from west-central California ...................... 2525-10 FMNH 17096. Apache double-curved si new-backed bow; detail of barberpole sinew-spiral grip-wrap. ............................................................................. 2525-11 Sixty-six inch double-curved Sek ani Athapaskan sinewbacked bow from central British Columbia, with barber-s piral wrapping on bowlimbs and nearly rectangular in cro ss-section .............................................................................. 2535-12 JES 858; Seventy-three inch doubl e-curved Apache sinew-backed bow with nearly-rectangular cross-section ....................................................................... 2535-13 Two Athapaskan-attributed trus sed cable-backed short bow s ......................... 2545-14 JES 365; Unusual double-curv ed Northern Athapaskan self-bow with attached wrist guard, upper Yukon River. ......................................................... 2555-15 Outline of a Qum-Dar ya (Xiongnu) bow ............................................................ 2555-16 Comparison of Apache and Slavey l ongbows (detail of midlimbs). .................. 2555-17 2.5-inch long juniper sinew-back bow fragment excavated by Julian Steward from Promontory Cave No. 1.; northern Utah (specimen now lost). ................. 2565-18 Comparison of Promontory Cave bow fragment to decayed ethnographic Shoshone sinew-back ed bow tip .. .................................................................... 256 11


5-19 PAH 1/21418, back-view of 4.3 in ch Lovelock Cave sinew-backed bow fragment, with longitudinal sinew-strands still attached, visible on lower edge 2575-20 Comparison of southern, Kodiak, and western style used to secure the cable backing to the bow .................................................................................. 2575-21 Comparison of belly-faces of Yu kon River and Mackenzie River Athapaskan complex bows ................................................................................................... 2586-1 Haplogroup C3 Y-DNA distribution overlaid with comple x archery distribution. 3016-2 Comparison of complex bow distribution in the contiguous 48 states with mean annual precip itation. ............................................................................... 3026-3 Comparison of Tlingit and Scythian scale armor.. ............................................ 3026-4 Estimated chronology fo r the adoption of the bow in North Am erica ................ 3036-5 Comparison of Tsuutina and Alaskan flat-bottomed pottery vessels. .............. 3036-6 JES 2649; Detail of tensioner (cabl e tension adjuster), late nineteenthcentury Yuit hide cable-backed bo w, East Cape, Siberia. ................................ 3046-7 MPM 33610; detail of vestigial siyah on Cop per Inuit bow ............................... 3046-8 SDM 27257; detail of cott on cable backing near the siyah of a Tlingit bow. ..... 3046-9 Yukaghir composite bow Northeast Si beria ..................................................... 3056-10 GMA 7336.284; Apache double-curved self bow. ............................................. 305 12


LIST OF ABBREVIATIONS AMNH American Museum of Natu ral History; New York, New York FENN Private collection of Fo rrest Fenn; Santa Fe, New Mexico FMNH Field Museum of Natu ral History, Chicago, Illinois GMA Gilcrease Museum of the Americas, Tulsa, Oklahoma MAC University of Missouri Museum of Anthropology; Columbia, Missouri MPM Milwaukee Public Museum; Milwaukee, Wisconsin NMAI National Museum of the Am erican Indian, Washington D.C. NMNH National Museum of Natu ral History, Washington D.C. PAH Pheobe A. Hearst Museum of Anthropology; Berkeley California PGM Peter the Great Museum of Anthropology & Ethnography; St Petersburg, Russia RBCM Royal British Columb ia Museum; Vancouver, Canada SDM San Diego Museum of Man; San Diego, California SNO Sam Noble Oklahoma Museum of Natural History; Norman, Oklahoma UFL Florida Museum of Natural History; Gainesville, Florida UMNH Utah Museum of Natura l History, Salt Lake City, Utah UTK Frank H. McClung Museum; Knoxville, Tennessee WFU Wake Forest University Museum of Anthropology, Winston-Salem, North Carolina YPM Yale Peabody Museum of Natural History; New Haven Connecticut 13


Abstract of Dissertation Pr esented to the Graduate School of the University of Florida in Partial Fulf illment of the Requirements for t he Degree of Doctor of Philosophy MATERIAL CULTURAL CORRELATES OF THE ATHAPASKAN EXPANSION: A CROSS-DISCIPLINARY APPROACH By Joseph Andrew Park Wilson August 2011 Chair: Peter Collings Major: Anthropology Contrary to stereotypes of proto-Athapaskan culture as simplistic and archaic, evidence points to a sophisticated web of late prehistoric Asian-Athapaskan interactions. A holistic assessment of At hapaskan migrations in the context of the transpacific Dene-Yeneseian phylum (the la rgest, fastest pedestrian language spread on earth) sees Athapaskan-Asian connections (in language, technology, DNA, social organization, etc.) as reflecting profound largescale cultural-historical processes whose implications have yet to be grasped. Current understanding is that Athapaskans slowly migrated south in response to volcanic eruptions in southwest Yukon Territory after circa 200 and 800 CE. Yet problems remain, notably the archaeological invi sibility of migrants on their long trek southward, and their possession of Asiati c strong complex bows which were not introduced to Northern Athapaskan territory unt il after these two eruptions. Linguistics, archaeology, biology, and data from ethnographic archery collections suggest Athapaskans carried sinew-backed bows to Ca lifornia and the Southwest. Both Apacheans and Northern Athapaskans uniquely possessed both Arcticand Plainsstyle sinew-backed bows. Migration with retention (not diffusion through existing 14


15 populations) is the best expl anation. The Athapaskan expansion was faster than generally supposed, quite similar to the contemporaneous Punuk/Thule Neo-Eskimo expansion in the far north. Such a model helps to explain archaeological invisibility in the intervening space, as the impact made by the migrants was small.


CHAPTER 1 INTRODUCTION The Research Question This study is an examination of west ern North American ethnographic materials from US museum collections, primarily sinew-backed bows. These materials are interpreted in the context of a four-field anthropological study of the expansion of the Athapaskan language family in we stern North America, in an e ffort to determine the role that material culture, particula rly weapons technology, has played in this migration. Like many North American ethnonyms, Athapaskan is not actually an Athapaskan name, but is derived from an old Algonkian toponym, athapaskaw, grassy place, used by the Cree to describe the Peace-At habasca Delta, on the wester n shore of Lake Athabasca, in Chipewyan territory (Ha rrington 1940:506). Most (but not all) Athapaskan speakers refer to themselves by some variant of Dn the Athapaskan word for people. My interest in questions of Athapaskan origins and migrations originates in a 24-week period of volunteer service with the Student Conservation Association in 1997-1998, when I worked in the Flagstaff Area National Monuments in northern Arizona. Although I worked on a variety of projec ts at three different monum ents, most of my time was spent helping to monitor and stabilize sensit ive backcountry archaeological sites and architectural features in and around Wupatki National Monum ent. Wupatki is located on the west bank of the Little Colorado River, adjacent to the Navajo Reservation, very near the Hopi Reservation, and surrounded on three sides by the Coconino National Forest. I was then only vaguely aware of the story of the Athapaskan migrations into the Southwest, but I was soon exposed to several co mpeting historical na rratives, as older 16


academically-trained Anglo archaeologists and interpreters, together with younger more-or-less traditional Navajo and Hopi comrades engaged in spirited discussion of indigenous culture history in w hat is still a hotly contest ed geographic realm. I quickly learned of the great philoso phical impasses between the Navajo and Hopi perspectives on Navajo origins. My Hopi colleagues were generally adamant that the Navajo were newcomers who had arrived only three or four centuries ago. My Navajo colleagues on the other hand were equally insistent that they had been there for twice that long or more, indeed for the better part of a millenni um. Scholars too are sharply divided on these questions. Disagreement over t he nature and timing of the Athapaskan migrations is longstanding, and forms part of the backdrop for other contentious issues such as the Navajo-Hopi land dispute (a bi tter decades-old political disagreement over native settlements), which is another w edge issue for native communities and the anthropologists who work with them (Schwarz 1997; Washburn 1989). It is very often the case that anthropologists and archaeologis ts political sentiments reflect those of their particular indigenous collaborators; scholars of Ho pi and/or Pueblo cultures advocate for Puebloan political viewpoints and Navajo specialists advocate for the Navajo ones. I seek to understand how and why the itinerant Navajo and Apache, originally from the far north, came to demographically ov erwhelm earlier southwestern Puebloan societies in a relatively short amount of ti me, despite an extreme difference in cultural ecology between the two regions. What coul d possibly have spurred the fastest known linguistic expansion of this scale in the We stern Hemisphere, and why did it begin only during the last 2000 years or so, when there were no obvious barriers to population 17


movements for thousands of years prior? There is widespread agreement among scholars that the movement of Athapaskan speakers was primarily north to south and commenced during late prehist ory, but that is about w here the consensus ends. Considering the vast amount of scholarship which has been produced in an effort to address these questions there is a remarkab le lack of concurrence regarding the precise chronology and how the migr ations occurred (Matson and Magne 2007:138149; Upham 1982:49). On the re lated question of why they occurred there is even less agreement. In order to address the question of the timing of this ex pansion, there are several related questions that must also be considered. These questions will be addressed in the literature review (chapters two and three). First, as the Athapaskan expansion is a linguistically defined phenonmenon, so linguistic data must necessarily come to bear heavily upon the question of the timing and t he source of the population movements involved. My assessment is that linguistic data is indeed consistent with a recent, rapid expansion. Second, human biology is another key area. As with t he linguistic data, my interpretation of the genetic data is consist ent with the rapid pace and late timing of the Athapaskan expansion. In particular, a number of genetic disorders which are specific to Athapaskan speakers indicate a series of recent genetic bottlenecks consistent with population founder effects and likely also a connection to Siberian populations in relatively recent timeframes (Erickson 2009). Third, a review of the archaeological literature is also consistent with the model of a recent, rapid expansion. All three of the above areas provide the context for a discussion of material cultural data collected in North American museums. 18


The Athapaskan expansion may be directly correlated to the late prehistoric diffusion of Asiatic sinew backed bow technology in western North America (Baldwin 1997; Downs 1972:6; Paper 1993). If it can be demonstrated to be valid, then this hypothesis represents a major piece of supporting evidence for a cultural connection between Athapaskans and Asians, as well as helping to explain how the Southern Athapaskans, initially few in number, were ab le so quickly to expand and encompass a vast territory occupied by others and far re moved from their orig inal homeland. My visits to ethnographic collect ions to examine North American sinew-backed bows were primarily conceived of to assess the mate rial support for an Athapaskan role in the diffusion of complex archery technology. Complex bows made by Athapaskan speakers and their neighbors from all over western North America were closely compared to assess the interrelationship of bow technol ogy with other ethnological data, providing information relevant to a discussion of Athapaskan migrations. Cultural Evolutionary ver sus Historical Models A language expansion of this speed and sca le beginning from a nucleus in the remote Western Subarctic raises the questi on: how could a group of societies classed among the simplest in the hemisphere give ri se to such a phenomenon? The answer is complicated. The extreme climates of interior northwestern Canada and Alaska demand versatility, efficiency and austerity in t he face of unforgiving nature. Subarctic dwellers have demonstrated effective adaptation to changing resource patterns, favoring generalized subsistence strategies and correspondingly utilitarian toolkits. The dominant cultural-evolutionary model in anthr opological theory since the mid-twentiethcentury has often presented Northern Athapaskans and t heir neighbors as typespecimens for pre-state hunter gatherer societ ies, and as direct analogs to much earlier 19


archaeological cultures in the region. But the evolutionary paradigm in North American ethnology is challenged by the late Holocene introduction of archery (presumably) from the Asian continent. John Blitz notes: [i] n the popular stereotype, the bow and arrow is closely linked to the Native Americans, yet the archaeological evidence reveals a surprisingly shallow time depth (Blitz 1988:123). Although the bow was ubiquitous throughout the Americas at European contact, it was virtually absent throughout most of both continents as late as the beginning of the Common Era. Yet Azar Gat is unconcerned by the impact of archery upon his cultural-e volutionary model for the emergence of violence in small-scale societie s. His argument for the universality of primitive warfare require s independent confirmation in mu ltiple independent smallscale societies, and Northwest America is taken as one of these test-cases. [T]he pattern of "primitive warf are" manifests itself independently everywhere. The . [bow and arrows ] only effect was to increase the range of engagement even further. The American Northwest is another vast "laboratory" of "pure" hunte r-gatherers (Gat 1999:567). Gat cites nineteenth-century descriptions of Tlingit raids and feuds to support his model of warfare in simple societies. But these allegedly independent ethnological analogs for evolutionary theory are very weak. The accounts all well postdate the dramatic escalation of violence in these societies that ensued in the wake of the rise of the European fur trade. The northwest quadr ant of the conti nent was already a protocolonial theatre of pow er by the mid-nineteenth-c entury (Harris 1995:131). Further, many of the basic assumptions under lying the cultural-evolutionary view of primitive warfare in the Northwest have been severely challenged by the revisionist approach to hunter-gatherer studies. As Kenneth Sassaman succinctly points out: the evolutionary validity of ethnogr aphic hunter-gatherers has been fully undermined by historical insight s on the connections between these 20


presumed pristine primitives and t he food-producing societies and nationstates in which they are encapsul ated (Guenther 1996; Headland and Reid 1989). With this revisionist thinking archaeologists lost the authority to blithely use ethnographic observations for analogical purposes, but they gained an arsenal of concepts for interp reting hunter-gatherer variation and change in historical, as opposed to evol utionary, terms. . This revised way of thinking is informed by theories and paradigms that we can gloss as historical (Sassaman 2010:4-5). Consistent with this revisionist paradigm is the rapid pace of the adoption of archery technology observed by archaeologists. This does not support the recently prevalent cultural evolutionary models for the adoption of new technology by Native Americans. John Blitz further argues. The large-scale pattern of dispersa l and adoption reveals processes not directly attributable to local environm ental circumstances. Instead, the rapid dissemination of the bow across major ecological boundaries is interpreted as the result of a contagious com petitive advantage in intergroup conflict (Blitz 1988:124). Yet despite the recent advances of revi sionist (or historicist) scholarship, subarctic cultures such as the Athapaskans are still widely regar ded as relicts of prehistory, rather than as hi storical actors engaged with a wider world. Headland and Reid (1989:49) regard subarctic peoples as geographically isolated and therefore as possible exceptions to the rule of widespr ead late Holocene engagement in interethnic trade between hunter gatherers and their neighbors. This study will help to demonstrate that the Athapaskan expansion du ring late Prehistory is far fr om exceptional, but could arguably be a type case for the historicist position. Athapaskan-Asian Connections While in Arizona I also became aware of apparent strong cultural affinities between American Indians and East Central Asians (particularly Athapaskans, Tibetans and Mongolians), and of ongoing intercultu ral exchanges between indigenous peoples 21


of Asia and southwestern North America (G old 1994; Meredith 2002). As James Downs notes [p]ersons who have associated with both Athapaskan, particularly Navajo and Apache, and Central Asian peoples such as Mongolians and Tibetans have often remarked at similarities of attitude and general behavior (Downs 1972:6). While some of these similarities can be observed in Native Americans in general owing to remote common origins, I could not help but wonder if there was a stronger historical connection at play in the Southern Athapaska n case, because the number and nature of similarities struck me as mo re than would reasonably expect by chance after millennia of separation. For example, my curios ity was piqued upon learning traditional Navajo weavers will incorporate a si ngle deliberate error in each piece, made to symbolize human imperfection in the face of divine perfection (Carmean 2002 :xx; Villaseor 1963:46). The identical practice and identical justification for this practice is found among Altaic weavers in Central Asia (Mellaar t 1980:93). Navajo cultural delegates to Tibet have lately remarked on the incredible similarity between specific Navajo and Tibetan woolen blanket and hat designs, along with the common personal names shared between the members of the two groups, physica l appearance and manner of prayer (Norton-McBride 2003). One such co incidence is unremarkable, but there are very many of them. And they often rela te practices (like weaving) which were presumably developed long after the time when the First Americans left Asia. Delving into the old anthropological lit erature for Athapa skan-Tibetan language links, a field pioneered by an earlier generation of eminent linguists like Edward Sapir (1925, 1991), Morris Swadesh (1952) and Robert Shafer (1952, 1957), I learned that additional Central Asian langu ages, particularly Yeniseian, are considered in this 22


context by such scholars as John Bengtso n (Blaek and Bengtson 1995), Merritt Ruhlen (1998), Michael Fortescue (1998) and more re cently by Edward Vajda (2010a, 2010b). Beyond the linguistic evidence, there are a number of anthropolog ical studies that address the profound similarity between Navajo and Tibetan religion (Chiao 1982; Karnakova 1913; Klein 1983; Kri ppner 1997; Samuels 1995). Ethnographic Evidence Lidscha Arbakow, an Alaska Native elder Vietnam war veteran, PhD candidate, and fluent Athapaskan speaker told me [m]y ancestors sailed across the Bering Sea from Asia (Wilson 2005:68). I soon discovered a large corpus of Athapaskan oral traditions recorded over the past century, ea ch making explicit cl aims to the peoples cultural roots in Asia. For example, the origin narrative of the Navajo Tsinajinii (black streaked wood) clan says that the Navajo peopl e originated in the Old World across the Pacific Ocean, and that overpopulation eventually forced them to emigrate to the New World (Weisiger 2004:253). Donald Cole (1988) records a story that was old a century ago among the Southern Chirica hua. This tale maintains that the Apaches were originally the slaves of Central Asian horti culturalists who rode horses and used bows and arrows. Legend told of a revolt and f light to the east through dark forests and across water larger than any river. Having left the Old World and entered a New World, the people moved east and south (Cole 1988: 2). Some educated Chiricahua even go so far as to claim that thei r former Asian residence was Xi njiang, in the heart of East Central Asia (Cole 1981:11). Virtually all the eighteenth and nineteenth-century records of Northern Athapaskan oral tradi tion make similar explicit claims for Asian origins in a time of great strife and warf are. Alexander Mackenzies journal from 1793 mentions: 23


the Chepewyans, and the numerous tribes who speak their language. . Their progress is Easterly, and, according to their own traditions, they came from Siberia; agreeing in dre ss and manner with the people now found upon the coast of Asia (Mackenzie 1902: v.2, 353). The Asian homeland of the At hapaskan Chipewyan, in t he account recorded by Mackenzie, was a place where they suffered great misery at the hands of a very wicked people; upon their arrival in America, the Chipewyan told Mackenzie they soon disovered and settled upon Coppermine River w here they were able to immediately make use of the abundant native copper resour ces (Mackenzie 1902: v.1, clxxiii). Asian-origin stories are found in Athapaskan folklore all over Canada, from the Rocky Mountains to the Mackenzie River Delta. Mo st versions are simila r to the contemporary Alaskan and southwestern accounts cited above, describing distant homelands on a western continent with numer ous people and exotic anima ls, where the ancestral Athapaskans were forced by their ferocious enemies to flee the continent and move to North America (Morice 1914:148-150). The expl orers who recorded t hese stories often took them at face-value, as testimony of a major migration from As ia in the relatively recent past. In contrast, later nineteenth and twentieth-century anthropologists of a cultural-evolutionary persuasion have been skeptical of such accounts, casting aspersions on their veracity: Most of these accounts seem to be in accord in placing their earlier home far to the west, either across the sea or on the other side of a long lake full of islands. From this western land they were driven by the cruelty and fierceness of their neighbors, and afte r long travel and many difficulties came into their historical habitat. . The most that may be said is that attempts to derive the northern Athapa scans from Asia on the basis of these traditions are absurd (Swanton and Dixon 1914:399-400). The work of Alice B. Kehoe (1981, 2003, 2010) questions the wisdom above. This basic Asian migration narrative has been re iterated independently at least half a dozen 24


times by closely related peoples residing thousands of miles apart. This should give one pause before assuming that these accounts are purely mythological and have no historical value. Kehoe (2010:200-201) maintains that ant hropologists over the past century have disregarded the im portance of pre-Columbian trans-oceanic voyages in favor of a paleontological (natur al-science based) model of Am erican cultural-evolution. Widespread Athapaskan oral traditions describ ing a journey from Asia by boat are incompatible with this paradigm, and have been dismissed as fanciful mythology for the better part of a century. Some Athapaskan tales of flight from persecution in Asia appear to have been forgotten by the peoples who once reported t hem. For example, in the 1860s, Emile Petitot recorded in great detail Hare and Gwichin Athapaskan accounts of migrations of persecuted ancestors across the western ocean from the continent we have left (Petitot 1878:50). But contemporar y Gwichin elders seem to have forgotten these tales, and may alternatively claim autochthonous or igins in North Amer ica (Kari 2010:218). Some indigenous activists now assert that transoceanic migration stories undermine native land claims; their views resonate with those of like-minded an thropologists. But native intellectuals are divided on the merits of this approach, just as scholars are. The late Vine Deloria argued that ancient Am erica harbored shiploads of refugees from historical conflicts. His statements echo the details of the ol d Athapaskan accounts. [M]any people feel they cannot advocate Precolumbian contact for to do so would mean demeaning the Indians and suggesting that they could not have made discoveries on their own. Strangely this debate also rages in Indian circles, and a few of my best friends are adamant about maintaining the theory of isolation in order to enhance the achievements of our ancestors. . Unpleasant though it ma y be to some In dians, we need to know the truth about North American prehistory, and indeed that of the Western Hemisphere. I personally feel that unless and until we are in some 25


way connected with world history as early peoples, perhaps even as refugees from Old World turmoils and persecutions, we will never be accorded full humanity. We cannot be primitive peoples who were suddenly discovered half a millennium ago (Deloria 1992:597). The Structure of the Study The next two chapters constitute a large cross-disciplinary review of relevant scholarship (including linguistics, genetics, ethnology and archaeology). Chapter 2 is a historical summary of more than a cent ury of Athapaskan studies in anthropology, focusing on conventional wisdom in the field. Chapter 3 brings this picture up to date with a review of recent work, especially breakthrough linguistic and molecular genetic studies. Some of these newer studies pose distinct challenges. A recently forged linguistic consensus surrounds the Dene-Yeni seian language family, meaning that the existence of Central Asian cultural ties with Athapaskans (and also their Eyak and Tlingit relatives) has left t he realm of speculation, and mu st now be directly confronted by non-linguists. Chapter 4 describes the me thods employedwhat exactly I did, which museums I visited, which bows and other obj ects I examined, how measurements were recorded and which ones, and why. Chapter 5 describes the findings based on my observations and field data, including representat ive tables of results grouped by culture area and/or language group, and expl anations of the categories in tables. Chapter 6 explores the broader significance of this st udy; what the findings mean and what the results point to in regards to the resear ch question; the conclusion, addressing the contribution of the present wo rk to the larger field of study, and suggesting the next steps for further research. 26


CHAPTER 2 THE HISTORY OF ATHAPASKAN STUDIES The Linguistic Foundations Since Athapaskan is foremost a linguisti c classification, the field of historical linguistics necessarily forms the framewor k through which all non-linguistic data are interpreted. Athapaskan is the largest family within the Na-Dene linguistic stock. The stock was formally proposed and named by Fr anz Boas student Edward Sapir (1915), although Michael Krauss notes that this proposal has much earlier roots. He traces the development of the hypothesis from the eighteenth century onward, particularly through nineteenth-century Russian lit erature (Adelung and Vater 1816; Radloff 1857; Wrangell 1839:101-103, 259; see Krauss 1973:953-954;1964:127-128) Na-Dene is the largest and most widespread of the generally acc epted American Indian language-groupings. Of the four proposed genet ic units within Na-Dene, three of them (Haida, Tlingit and Eyak) are single languages almost exclusivel y confined to the Northwest Coast, while the largest, Athapaskan, is a widespread group of very closel y related languages almost exclusively limited to the continental interior. The precise number of Athapa skan languages is difficult to determine, because broad dialect continua and mutual intelligi bility between so-called languages are commonplace. Doubts about the validity of the Na-Dene phylum and the inclusion of both Haida and Tlingit persisted for many dec ades (Krauss and Golla 1981:67). Sapirs mentor, Franz Boas did not accept the valid ity of Na-Dene as a whole, although he did consider Tlingit-Haida to be a genetic groupi ng (Boas 1894). He thought Tlingit-Haida was unrelated to Athapaskan but conver gent with Athapaskan due to diffusion and borrowing as a result of historical contact (Swadesh 1951). Boas views were shared by 27


Goddard (1920). These questions have re mained open. Some have harbored doubts about even Tlingit until fairly recently (Campbell 1997:286), while others still embrace Na-Dene according to Sapirs original us age, including both Tlingit and Haida (Drr and Renner 1995; Enrico 2004; Pinnow 1985; Ruhlen 1998). Most Americanists now occupy a middle ground, rejecting Haida, while affirming the affiliation of Tlingit. They have thus split apart Boas old Tlingit-Haid a unit which was widely accepted by an earlier generation of linguists, regardless of where they stood on the Na-Dene issue (Hymes 1956:626-632). Na-Dene is now most commonly used to describe only the relationship between Athapaskan, Eyak and Tlingit (or AET; excludi ng Haida). Eyak refers to small native culture closely related to Athapaskan, once residing in the Copper River Delta. The relationship between Athapaskan and Eyak is so clos e in fact that the distinct status of the Eyak branch was recognized by Anglo-Am ericans only in the mid-twentieth-century, when Eyak language and culture were already severely endangered (Birket-Smith and de Laguna 1938). Athapaskan and Eyak together form one majo r sub-unit of Na-Dene, Athapaskan-Eyak (the other being simply Tlingi t). In recent centuries the Eyak were largely absorbed and assimilated by the nor thwesternmost Tlingit who have steadily encroached upon their territor y (de Laguna 1990). Eyak language is now extinct, as the last native speaker died in 2008. So the only living Na-Dene languages remaining are a number of Athapaskan l anguages and Tlingit ( and possibly Haida). Michael Krauss speculates that Tlingit itself may be a hybrid language resulting from the merger of an ear ly Athapaskan-Eyak speech community with an unknown, unrelated group (Krauss 1973:960-963). This woul d account for the paradox of Tlingits 28


remarkably Athapaskan-like verbal morphology despite fewer than expected lexical cognates (in other words the stem-inventory of Tlingit could reflect a distinct strand of non-Athapaskan-Eyak genetic hist ory). If the unrelated stock was indigenous to the region and a bona fide Na-Dene linguistic stratum was later introduced by immigrants, then the Tlingits coastal adaptation and mate rial culture could have been borrowed en masse along with the associated foreign lexicon, as the arrivals were integrated into a community with deep roots in t he region. It might be that this non-Na-Dene component was a relative of Haida, and that Tlingit is genetically related (by mixed marriage) to both Haida and Athapaskan-Eyak. An artificially mo re-ancient Na-Dene follows in two different ways: (1) through an illusion of c ontinuity with stable local archaeological traditions, and (2) because hybridization has a ccelerated Tlingits linguistic divergence from kindred interior-dwelling Athapaskans. Na-Dene and Sino-Tibetan Linguistic claims for Athapaskan cultural root s in Central Asia are nearly as old as the field of Na-Dene historical linguistics it self. The earliest published reference to a possible link between Athapaskan and Tibetan languages was made around 1876 by the Catholic missionary and ethnologist, Father mile Petitot, who was a fluent speaker of multiple Canadian Athapa skan languages. In a chan ce meeting with a missionary returning from Tibet (lAbb Fage), both cler ics were astonished to discover several common vocabulary words in addition to numerous structural and grammatical similarities shared by Ti betan and Athapaskan languages (Petitot 1878:61). Later, Edward Sapir, his student Morris Swadesh and the eminent Sino-Tibetanist Robert Shafer among others have sought to establis h a transpacific stock, joining Na-Dene with Sino-Tibetan, and forming the Sino-Dene or Dene-Tibetan grouping (Sapir 1925, 29


1991; Shafer 1952, 1957; Swadesh 1952). Sapir wr ote that the similarity in feeling between Tibetan and Nadene is at least as close as between Latin and English, probably closer (Sapir 1991:134). Sapirs work incudes Haida in this phylum, but Shafer subscribed to a more conservative approach, excluding Haida and including only Athapaskan and Sino-Tibetan materials. The great bulk of evidence for Sino-Dene language links comes from At hapaskan and Tibeto-Burman lang uages. The fact that Sapirs work was yet unpublished when Shafer di d his work is significant, according to Sapirs student Morris Swadesh, who writes that Shafer's work constitutes an independent corroboration of the correctness of the theory (Swadesh 1952:178). Sapir and Shafer are considered to be the founding fathers of Na-Dene and Sino-Tibetan linguistics respectively, but their disciples have been generally unreceptive (even highly critical) towards efforts to bridge the gul f between geographically distant families (Kaye 1992). Some linguists have nonetheless proposed an even larger Dene-Sino-Caucasian super-stock, combining Na-Dene, Sino-Tibet an, Yeniseian, and Caucasian (Blaek and Bengtson 1995; Sarostin 1991). The D ene-Caucasian proposal, though rooted in Sapirs work, relies heavily upon the unconventi onal multilateralist comparative method, a.k.a. mass comparison, pioneered by Joseph Greenberg (Greenberg 1987), and furthered by his protg Ruhlen, John Bengts on, and others. Mass comparison is not recognized as a valid method by the majority of Americanist linguists However, Sapirs original Sino-Dene proposal did not use unconventional methods. Whatever its merits may be, Sino-Dene is argued on the basis of the traditional comparative method and . subject to refutation on the same basis (Michalove et al. 1998:468). 30


Timing the Athapaskan Expansion using Lexicostatistics or Glottochronology Chronological estimates for language divergence are based on lexicostatistics, extrapolated to a timeframe known as glottochronology. Morris Swadeshs (1951) classic article Diffusional Cumulation and Archaic Residue as Historical Explanations is recognized as a foundational text in anthropological linguistics because it establishes the ground-rules. Swadeshs basic premise is widely accepted. Each language possesses a stable core vocabulary. Some of these core words are nevertheless lost through time. Core word lists can be compar ed across languages to reveal words of common origin (cognates). A uniform rate of change in core vocabulary between two related language-communities would (in th eory) allow one to date the temporal divergence of the two populations, by calibrati ng the loss rate for related languages of known historical depth. In possibly the earlies t effort to establish the time of divergence between different Na-Dene la nguages, Swadesh compared 82 core-words for Tlingit and Athapaskan, finding 36 (44%) cognates and thus proposed that the most-recent common ancestor of Tlingit and Athapaskan was at least 2000 years ago (Swadesh 1954:362). Since that time, other scholars have adjusted the method and increased that age estimate, but the basic method has remained the same. It has long been assumed that the Na-Dene languages di verged gradually, moreor-less in situ, in early Holocene North America (G reenberg et al. 1986). This is not a valid assumption, as members of distinct yet related language s often migrate together. One needs only to look at the recent col onial expansions of closely related IndoEuropean languages (e.g. Spanish and Portugese, or English and Dutch) to see this is the case. Indigenous North America has other examples of this phenomenon; notably, the late expansion of Uto-Az tecan speakers in the Great Basin most likely involved 31


several distinct, closely affiliated l anguages (Miller 1986:100-104). Individual languages with a common pedigree can be completely differentiated long before a shared migration event brings them toget her to a new home. As Kenneth Weiss and Ellen Woolford argue: the distance between the . Na-Dene gr oups could not be used to date the divergence of this group in the New Worl d . since if there was pressure on one group to migrate, there could as well have been pressure on two nearby groups to move at about the same time (1986:492). Or as Richard Perry succinctly put it [o]n the basis of present evidence, for that matter, there is no compelling reason to a ssume that the division between Eyak and Athapaskan occurred in North Americ a at all (Perry 1983:720) Even assuming that the depth of divergence between Na-Dene languages would conform approximately to migration history, the estima tion of linguistic divergence rates is an inexact science fraught with difficulties distorting its conclusions; linguistic data are quite compatible with any date (Nettle 1999:3328). As a relative dating method, glottochronology must be calibrated using non-linguistic data. The next sections will review the chronological estimates of the Athapaskan migrations, based on linguistic evidence. The North-South Direction of the Athapaskan Expansion Edward Sapir (1936) described two types of linguistic evidence for Athapaskan migrations; internal evidence refers to s pecific lexemes having remote, identifiable origins, and/or neologisms reflecting relative ly recent adaptations to new surroundings. External evidence includes the patterning, diversity, and depth of differentiation (cleavage) between dialects or languages, wh ich may also hint at the direction and relative order of migr ations. The northern urheimat (linguistic homeland) or nucleus of the Athapaskans is demonstrated based on both types of evidence; there is more 32


linguistic diversity and dialect cleavage in the north than in within ei ther the Apachean or Pacific Coast branch, and there is a clear patte rn of neologisms related to southwestern foodstuffs and horticultural products (Sapir 1936: 224-225). Furthermore, the cleavage between individual Northern Athapaskan and Southern (or Pacific Coast) Athapaskan languages (e.g. Navajo and Chipewyan) is not as great as that between particular Northern languages, (in other words Denaina and Chipewyan). This establishes the north as the more ancient l anguage-bloc and the two sout hern blocs as peripheral. Sapir presented internal linguistic evidenc e which conclusively proved the recent northern origin of the Navajo; e.g. t he Navajo word for corn or maize, n d now the most sacred plant in the Navajo cosmos, or iginally meant alien-food or food of the enemy (Sapir 1936:231). It less well know n that a remarkably similar plant-name construction, made from the same na root-word (for enemy, a lien, or foreigner), is found in the language of the Chilcotin, in the southernm ost contiguous extension of the nuclear Northern Athapaskan bloc, and can likewise be used to demonstrate a southward migration, as report ed by A.G. Morice: They call the particular kind of grass (Poa tenuifolia) known as bunch-grass, which is one of the most valued possessions of their present country, nna-tl, which means grass of the Foreigners, that is the Shushwaps (Morice 1914:153). Further, the Athapaskan names for a row of hills in the Great Bear Lake basin are arranged in a geos patial-chronological template following a northwest-to-southeast migrat ion itinerary, with First Promontory being the northernmost one, and Last Mountain and Last Steppe respectively being the southernmost two (Morice 1914:154). 33


While the primary movements have been north to sout h, there have been other migrations back and forth between the thre e major geographic blocs. The nineteenthcentury witnessed the annihilation of tens of thousands of California natives in massacres and through disease in the wake of white settlement, and the Pacific Coast Athapaskan groups were not spared (Thornton 1986). Navajos during this era encountered a stream of Pacific Coast Athapaskan refugees who made the arduous journey by foot across t he Rocky Mountains, upon learning of the existence of their linguistic kin to the east. The Navajo gr eeted the California Athapa skans as Western immigrants and long lost brethr en and readily assimilated them into their nation in a testament to the mutual inte lligibility of their respecti ve languages (Matthews 1897:3132). The Navajo also have a native term for Northern Athapaskans, Din nahodloni, also Navajo. Nineteenth-century Navajo lore describes a long trek to Canada by a party of Navajos, to visit the Din nahodloni whose existence they still dimly recalled in legend (Morice 1914:155). Thes e examples are indications of the affinities between Athapaskan speech communities in different geographic areas, affinities that suggest that the time of separation between t he three branches cannot be immense. Backmigration of Apaches north may also be indi cated by the presence of strong Apachean folklore motifs among the Northern At hapaskan Tsuutina (Cur tis 1928:136-144). Chronology of Athapaskan Origins The relative chronology of the Na-Dene languages is widely accepted. Eyak diverged from undifferentiated proto-At hapaskan prior to the divergence between any particular Athapaskan languages. However the e ffort to chart this divergence process in calendar years is more of an art than a science, demanding the liberal incorporation of non-linguistic data (usually archaeology or history). Depending on what interpretation 34


of the prehistoric record is applied, one can st retch or compress the relativistic linguistic chronology to fit a particu lar historical narrative. The most often cited lexicostatistical es timates for Na-Dene languages are those of Michael Krauss (1973); y ears for Athabascan, and 3400 years between Athabascan and Eyak and about 4500 y ears for Tlingit (Kari 2010:210). Krauss estimates are based in the conventional application of Morris Swadeshs standard lexicostatistical methods, with only mi nor adjustments. There is considerable variation between the chronological estimates of different scholars, and much of this variability in the literature is conditioned by the application of non-linguistic data. For example, Kaufman and Golla (2000:51-52) estimate 3500 years divergence between Tlingit and Athapaskan (1000 years less than Krauss), while Swadeshs (1954:362) calculations for this divergence (as menti oned) were as few as 2000 years; less than half of Krauss canonical estimate. Kr auss (1973:953) wrote t hat: Kroeber ([1955] on extralinguistic grounds, and certainly correct ly) thought these dates too low. In Kroebers own words: In spite of its wide geographic spread it is obvious that Athabascan has well-marked consistency; and yet I should expect at least 20 centuries of separation, quite probably up to fully 30, for some of the mo re divergent of its branches, rather than for Tlingi t from Athabascan (1955:92-93). Skepticism of Swadeshs unexpectedly s hallow 2000 year date is understandable on linguistic grounds, considering that t he genetic relationship between Tlingit and Athapaskan itself has only recently become uni versally accepted. The asymmetrical nature of the relationship between Tlingi t and Athapaskan (more morphological than lexical) suggests that some combination of genetic and non-genetic (or contact-induced) factors may be at play, clouding the historic al perspective; the estimated time depth 35


would vary depending on which li nguistic features were considered diagnostic, and that is not clear. Ignoring the Tlingit issue, the widespread anthropological opposition to the youngest dates (<2000 years) for proto-At hapaskan and/or pr oto-Athapaskan-Eyak (excluding Tlingit) are not ju stified on linguistic grounds so much as they are rooted in basic incredulity that so massive, far-flung, and culturally diverse a family could be recent in origin. Harry Hoijer also had reservations about his own findings of a remarkably truncated chronology similar to Swadeshs: It is somewhat surprising to find how littl e time is involved in the break-up of the Athapaskan languagesthe earliest time of divergence is only about 1300 years ago. The movement of the Pacific Coast languages to the south apparently begins almost at once and is essentially complete at a date roughly 1000 years ago. The movem ent of the Apachean languages southward appears to have begun somewhat later, about 1000 years ago and, if our dating is accurate, wa s not entirely complete until about 600 years ago (Hoijer 1956:232). Hoijers glottochronological estima tes for Athapaskan diversification are problematic from the perspective of ethnographers like Kroeber, vi ewing Athapaskans as having deep cultural roots in various landscapes; Athapaskans occupy about the same sized territory as the Uto-Aztecans do, so therefore it follows they should be about the same age. As a means to address th is paradox, Kroeber may have been the first to suggest that the Athapaskan languages are mo re radically conservative than other languages, resisting lexical change and ther efore appearing younger than they are: But obviously if one set of languages were twice as resistant to alteration as another set, and the same formula were applied to them, the resistant set would emerge from comput ation seeming only half as "old" as the second set. The Athabascan and Uto-Aztecan groups have been recognized as stocks about equally long and with an analogous and approximately equal spread; it probably has been tacitly assu med that their pr incipal component languages diverged about equally. It was on this generic parallelism . that I based my expectation that the At habascan split-up would probably prove to be datable up to 30 centuries back. Ho ijer's 13 centuries thus comes as a shock; and the possibility that t he Athabascan language s are inherently 36


more resistive to change suddenly confronts us with a certain insistence (Kroeber 1959:241). Most of the revisions of Swadeshs and Ho ijers shorter-than-expected dates have involved the systematic purging of items in their word lists to exclude questionable matches resulting in a lengthen ing of the chronology by appr oximately 1000 years in the decades that followed their publications. A neutral assessment recognizes that the expectation of increased age has been the driving force of this revision; a different set of expectations would favor diffe rent results. It is equally possible that the Athapaskan languages are indeed very young and have had a remarkably succesfull expansion. Krauss canonical el ongated chronology is based on the fact that Hoijer and Swadesh both later increased their estimate s for the breakup of Athapaskan back to about 2000 years BP, in response to dat a showing greater Athapaskan language diversity in Alaska than anywhere else. Twelve Northern Athapaskan languages were not adequately represented in the early calcul ations, with the Dena ina of far western Alaska being the most radically divergent of these (Krauss 1973:952). It is clear that Athapaskan language diversific ation began in the far north west. Nonetheless, the precise chronology is speculative and open to debate. It is possible that this timeline is indeed too short, as many assume. It is equally plausible that it is too long; the tumult of Athapaskan contact history could just as well be responsible for hastening lexical erosion, and thus the geographic spread of Athapaskan ma y even less than 1300 years, especially if the widespread assumption of linguistic conservatism is incorrect. Another early critic of Ho ijers chronology was Wilhelm Milke (1959), who like Kroeber, primarily sought to lengthen the ch ronology with adjustments of Hoijers 37


methods. Yet his work was an exception to the rule with respect to the Pacific Athapaskan subdivision, arguing for an even shorter chronology, as Kroeber notes: Milke points out that Hoijer appears to be conservative in recognizing cognates, especially on the Pacific coast. . [Then] the time of divergence would fall from 858 to 660 or 597 y earsin far better accord with the distance separating them, and leaving the Pacific division much less anomalous in its internal or ganization (Kroeber 1959:257-258). Standard estimates (following Ho ijer) tend to place the Pa cific Coast migration(s) in the late first millennium CE, just prio r to the Apachean migration which in turn commenced in the early second millennium. However, Milkes truncated Pacific Coast Athapaskan chronology is notably similar to the standard estimates for the divergence within the Apachean branch, suggesting that both the Sout hern and Pacific migrations may have happened at nearly the same time, in the early second millennium. This is not to claim that Apacheans and Pacific Coast Athapaskans were one undifferentiated unit 1000 years ago; they were more likely distinct communities at that point, but they could have occupied different geographical positions in a vast Canadian Athapaskan dialect continuum. Multiple distinct speech communi ties were likely involved in subsequent migrations out of the urheimat. The questi on of when each departed is vexing because the chronological differences between the various non-Alaskan branches of the family appear trivial; it is difficult to ascertain, for example, whether Navajo is closer to Chipewyan or to Tsuutina, becsause C anadian Athapaskans languages would not have been deeply differentiated in late prehistory. The heterogeneity of linguistic innovation in the north is suggestive that Athapaskan dialects have cross-pollinated each other, and Athapaskan bands have likely fissioned and reformed since the the Southern Athapaskan migration, meaning that the use of modern ethnic gr oups to infer family-tree lineages is not straightforward. 38


Despite the presence of deep internal di fferentiation in western Alaska, the Athapaskan expansion may be seen as a single coordinated event rather than multiple independent migrations in succession. D on Dumond discusses the homogeneity of the Candadian Athapaskan langua ges (Figure 1-1): [R]ivers debouching into the Pacific w ould have tempted some early NaDene to work their way coastward. . But in this case any linguistic evidence of early passage actually within the corridor has been blotted out by a later spread of specifically Canadian Athapaskan speakers from the north (Dumond 1969:859). The path of proto-Na-Dene grou ps in the region is speculative. Dumonds mention of a later spread of specifically C anadian Athapaskan speakers is what merits discussion and scrutiny. It is from this zone that both Southern and Pacific Coast Athapaskan speakers emerged although the Apachean/Sout hern branch likely budded off of this bloc to the east of where the Paci fic branch emerged. It is hard to say, yet the same regional demographic expansion process likely accounts for both excursions, if one considers that the Pacifi c Coast migrants arrived at their destination by a more direct and expedient route t han the Apacheans, and thus have resided in their current home for slightly longer. And if the Pacific branch remained in contact with the Northern branch longer than they did with the Apac heans, then this would make them appear closer to Northern Athapaskans in glottoch ronological terms even if their expansive push commenced at the same time or earlie r than the Apacheans. This could explain how the lexical diversity (represented as time-depth) between Pacific Coast Athapaskan (PCA) languages is greater than between the Apachean langua ges, while PCA taken as a whole unit is apparently less divergent from the Nort hern bloc. A Northern Athapaskan-PCA dialect chain stretching from the Columbia Plateau to Oregon may have existed for several centuries after the Apacheans became isolated from their 39


Canadian forebears. This would agree with M ilkes observation that it seems plausible that the similarity between Pacific and Nort hern languages is greater than that between the Pacific and Southern languages (Milke 1959:188; my translation). There is considerable chronological overlap between the standard migration models for both groups, with later dates for the PCA movements corresponding with the early dates for the Apachean movements. If this was an extended process involving multiple closely engaged dialects in a vast territory, it is plausible to envision the westernmost migrants moving rapidly southwar d along the coast in a contiguous strip remaining in contact with the Northern bloc. Meanwhile, neighboring groups moved east or southeast, initially remaining in c ontact with fellow Athapaskans but eventually turning sharply south further inland, ultima tely forming the Southern/Apachean bloc. The distribution of the numer ous PCA bands is curiously not found in prime real estate, with very little presence on the c oast itself, but primarily in the remote and sparsely populated uplands, away from the lush river-mouths. This is intelligible as the effect of gradual movement out from the in terior, a movement essentially a matter of speech boundary advance from village to villa ge (Jacobs 1937:61). This is not migration into virgin territory, nor is it population replacement, but something in betweenthe progressive assimilation of smaller bands as the expansion frontier commences. The languages of streams with fe wer and smaller villages were less able to hold out against the numbers of aliens speaking upriver or interior languages (62). By this reckoning, the PCA population is expected to be a fairly uniform admixture between Athapaskan migrants from the north, and Athapaskanize d local villages. The commonly held view of PCA languages as prestige languages may support this notion. 40


Whistlers assessment is credible, favoring a slightly later thirteenth century CE date of entry for the California Athapaskans: Athapaskan entry was more recent than Jacobs implied. . Golla notes various cultural and linguistic evidence for recency of Athapaskan adaptation to California. Al so, as mentioned, some of the linguistic diversity of California Athapaskan may have been the result of a spread of Athapaskan through asymmetric bili ngualism with the language spreading to non-Athapaskan ethnic groups who chose to speak it rather than their former language. This phenomenon might also lead to overestimation of the date of entry of Athapaskan or iginally (Whistler 1979:74). These same kinds of arguments are applicab le to the late prehistoric Athapaskan movements into other regions as well, e.g. Denaina movements westward out of central Alaska onto the Yupik-dominated coast; Beaver and Chipewyans moving eastward into Algonkian territory southeast of Lake Athaba sca and west of Hudsons Bay; and most famously, Apachean penetration of the Puebloan Southwest. Local archaeological cultures and regional genetic traits may tell one story, even as la nguages have rapidly changed due to the influence of shifting patterns of multilingualism through alliance formation. Multiple intertwined strands of demographic history are often present within a single speech community. The Pacific Coast bloc is interesting because there is an intermediate group, a probable residue of the Athapaskan migrati ons in the now-extinct KwalhioquaClatskanie-Owilapsh (KCO) Athapaskans who dwelled around t he mouth of the Columbia River in Oregon and Washington State. KCO is a single language of three dialects, grouped with PCA on geographical grounds but technically distinct from the other PCA languages in southern Oregon and nor thern California (Krauss 2005a:114). KCO is linguistically about equidistant fr om its nearest Northern Athapaskan and PCA 41


neighbors, and thus could be composed of remnant-bands in a chain of migrations connecting the Northern and Pacific Coast blocs. Plains Apache (a.k.a. Kiowa Apache) in relation to southwestern Apachean is similar to KCO in relation to PCA, in that Plains Apache is too a single language distinct from all the other Apachean dialects including Navajo, which are themselves merely dialects of one Southern Athapaskan langua ge (Hoijer 1971:4). Plains Apache, like KCO, has also been proposed as an interm ediary or remnant in the Athapaskan migrations, between the Ttsuutina (a.k.a. Sars i) of the Canadaian Pl ains and the other Apacheans to the southwest (Gunnerson and Gunnerson 1971:18-19). The Plains Apache resided in the Dakotas with the Kiow a in the eighteenth century (Brant 1951:9), and were known to have friendly relations and even occasional intermarriage with the Tsuutina through the Kiowa alliance with the Blackfeet confederacy to the north (Brant 1951:9; Hyde 1959:26-27). However, it seems more likely that the Plains Apache are a reflux migration of Apacheans back to the north, because they have much more linguistically and culturally in common with their Apachean brethren to the south, than with their Northern Athapaskan allies (Brant 1951:129; Opler 1983). Evidence of eighteenth century Tsuutina-Apache intercourse nonetheless indicates Athapaskan interconnectedness and unity through time. Hoijers early work (1938) lumped Lipan and Jicarilla Apache dialects into the same Eastern or Plains Apachean subgrou p as Plains (Kiowa) Apache proper, but Hoijers later refinements (1971) were inst rumental in establishing a Southwestern Apachean language including a ll Apachean dialects except Plains Apache. Nonetheless, one should not overestimate the differences between the languages 42


concerned. There may have been intervening historical factors which have distorted this chronology somewhat. As Morris Opler writes: Lexicostatistical data show that Kiow a-Apache did indeed differentiate from Jicarilla and Lipan earlier than these la tter two dialects separated from each other. Yet the gap in time is not so great as may be supposed; the divergence time of Jicarilla and Lipan is a little over 200 years (Hoijer 1956; Hymes 1957). The Kiowa Apache diver gence from Lipan and Jicarilla . occurred about 200 years earlier (429 and 401 years, respectively). The time difference may even be consider ably narrower than available evidence can reveal. The federal government po licy of consolidating southwestern Apacheans on a few reservations, begun in the 1870s, forced these six tribes into common speech communities and may, as Ho ijer (1956:226) has pointed out, "account in part for the generally low times of divergence" (Opler 1983:368). Intertribal contact situations on reservati ons might also have caused Eastern Apache tribes (especially the demographically chal lenged Lipan) to become more westernized in their speech patterns, clouding the histor ical picture. Al so, pre-contact-era Apacheans were certainly less divergent fr om the Northern bloc and likely retained stronger memories of thei r own northern origins. Location of the Proto-Athapaskan Homeland Using mean retention ratios (percentage of shared cognates) to infer a relative divergence chronology and also ancient distributional patterns of the languages, Fowler (1977:103-104) suggested that the Pacific and Southern groups were located near and closely related to Ingalik [Deg Hit an] before they di verged. Further, he suggests that Tsuutina also has its root s in west-central Alaska, and that the Athapaskan expansion was a veritable explosi on of people out of this region within a relatively short timeframe. These data further suggest that the proto-Athapaskan homeland was located in close proximity to the deepest lexicostatistical cleavage 43


between local Athapaskan-Eyak languages (Denaina, Aht na and Eyak) in southwestcentral Alaska (Kari 1989). The other school of thought on this matter, embodied by Krauss and Golla (1981:68), is that the Proto-Athapaskan homeland is in a less diverse region, eastern interior Alaska or the wester n British Columbian cordillera. This is argued on the basis of the surprising lack of Eskimoan loanwords in all Athapaskan languages except for Denaina and Deg Hitan, which suggests that Proto-Athapaskan could not have been subjected to a long history of contact with Yupik the way that Denaina and Deg Hitan have been. It therefore must have been locat ed considerably to the east, in the Upper Yukon drainage and/or nort hern British Columbia. While th is view is the most widely accepted in the literatur e, it is somewhat lacking in co rroboration from other forms of evidence, and based on the assumption that the Proto-Athapaskans were present in their original homeland for a long time. If the Athapaskan arrival in Alaska was immediately followed by rapid migration and dispersal toward the interior, then the Yupik influence upon Denaina and Deg Hitan could simply reflect the last millennium or so of contact history. Eskimo influence upon Dena ina is relatively slight, and could be explained as a product of loanwords transpi ring after the major expansion of the Athapaskan languages out of western Alaska had commenced (Kari 1989:552). The Alternative Long Chronology Model for Athapaskan Origins. Ever since twentieth-century scholars recogni zed that Native American occupation of the Americas began during the termi nal Pleistocene, there have been many concerted efforts to characterize the m ap of the Athapaskan languages in terms of presumed in situ post-Pleistocene growth (Harri ngton 1940; Kari 2010; Krantz 1977; Rogers 1985; Rogers et al. 1990) Such efforts, at leas t with regard to Athapaskan 44


proper, must be characterized as marginal to mainstream historical linguistics, for assuming glacially slow rates of languag e change without historical precedent. Nonetheless, there has been a marriage of convenience between long-chronology advocates and prominent biolog ical anthropologists who favor single-wave migration scenarios (Rubicz et al. 2002); likewise with archaeologists who anachronistically characterize the Northern Archaic mi croblade users of Alaska and Canada as Athapaskan. Cross-disciplin ary syntheses have thus giv en the long-chronology undue influence within anthropology in general, cons idering its relatively poor support on linguistic grounds. But it is the demonstrable lack of arc haeological continuity in late prehistory througout most of the Athapaska n heartland that is the str ongest support to justify the short chronology (Shinkwin 1977). This rec ourse is natural cons idering the inherently relativistic nature of linguistic dating. It might even be argued that standard shortchronology estimates of 2000-3000 years divergence between Athapaskan languages (Krauss 1973:953) are rather slightly too long to be reconciled with archaeological evidence for Athapaskan presence which is s ubstantially younger than 2000 years in most (if not all) cases (Perry 1983:720). These data are compatib le with a number of different short-chronology scenarios; language diversification may have already begun prior to settlement, or contact-induced chan ge involving extinct local languages has resulted in deepening of differentiation, or a combination of both factors (Weiss and Woolford 1986). Some long-chronology advocates maintain that the original Na-Dene homeland was in an ice free refuge on the northwest coast during the Pleistocene, because all 45


three generally recognized subdivisions of the Na-Dene family were found occupying coastal terrain in the contact era, from C ook Inlet to California (Rogers et al. 1990:133). This model is based on a standard convention of historical linguistics; it is assumed that the development of language diversification is proporti onal to time depth of human occupation of an area (G ruhn 1988: 77). While this is logi cal on the surface, it requires that the diversification has o ccurred in-situ, which is not necessarily a valid assumption. In the Na-Dene case this assumption is unwarranted, because of strong evidence (both archaeological and ethnographic) that the groups in questi on have migrated from other locations in recent centuries. As William R. Fowler Jr. put it: [ o]pinion on these issues has diverged sharply and been polari zed by the unnecessary and unfounded assumption that present distribution reflec ts ancient distributi on (Fowler 1977:102). Archaeological Evidence for Athapaskan Origins Evidence for Interior Origins The few examples of coastally-adapted Athapaskans (Denaina in Alaska and a small minority of PCA languages) are all la te-Prehistoric intruders in the region (Dzeniskevich 1981; Whistler 1979). Moreover, Eyak is closer to Athapaskan than any other language, and most likely a late coasta l intruder as well, for they had a landbased economy and, unlike the Eskimo or Tlingit, never became sea-mammal hunters (Krauss & Golla 1981:68). Even Tlingit s ea-mammal hunting may be a comparatively late adaptation; oral traditions provide compelling evidence for interior origins, and a number of early ethnographers view ed Tlingit settlement of s outheast Alaska as recent, perhaps within the last 500 years (Moss 2004: 184). Archaeological evidence suggests somewhat greater (but still m oderate) antiquity for Tlingit settlements, perhaps between 500-1000 C.E. (Moss et al. 1989:540). This ~500-1000-year discrepancy between 46


archaeologically and ethnographically derived es timates for Tlingit origins could be readily accounted for by Krauss model which pos tulates a hybrid origin for the Tlingit; the non-Na-Dene segment of t he population would likely be the one with longstanding roots in the coastal region, while the Na-Den e linguistic culture was grafted onto this trunk somewhat later (Krauss 1973:960-963). It is not necessarily the case that continuity in the archaeological record is indi cative of linguistic c ontinuity. Furthermore, the view that Proto-Athapask an and Proto-Na-Dene cultures were interior-adapted does not necessarily mean that this interior adapt ation was first developed in North America. The interior-adaptation of Na-Dene soci ety could have been developed in Asia and transferred to North America by the migrants (Murdock 1955:86). The Problem of Archaeological Visibility The chronological framework of the At hapaskan prehistory of North America for the millennia prior to the second millenniu m of the Common Era is one area where the archaeologists are decidedly more beholden to linguists than the other way around. Given the inherent guesswork of glottochronol ogy, this does not instill confidence. For glottochronological estimates, archaeologist s rely mostly on the opinion of linguist Michael Krauss (1973), which is in turn condi tioned by the earlier opinions of Hoijer and Swadesh. The entire field is rife with specul ative musings and tautologies, as John Ives notes. Assessments of conti nuity and discontinuity lie too often within the realm of opinion and intuition, lacking the clear defin ition required for thei r evaluation (Ives 1990:55). Moreau Maxwell accura tely states that several archaeologically distinct complexes from across the subarctic have contributed to the development of Athapaskan people (Maxwe ll 1980:178). But this does not imply that a contribution to the development of a culture (however significant) is identic al to linguistic ancestry. 47


Many non-Athapaskan speakers have contri buted to the development of known Athapaskan cultures. Similar ev ents in prehistory would help explain the diversity. Despite the fact that Northern Athapaskans occupied most of the land of northwest interior North America at t he time of European contact, the evidence for the prehistory of Athapaskan societies is remarkably sparce There have been numerous efforts to connect archaeological cultures to those of ethnographic Athapa skans, with mixed results. In western Alaska, for example, Frederica de Laguna found a tremendous quantity of historical Athapaskan material, and some late prehistor ic material, but no strong continuity with earlier cult ures in the region. She not es that the material from the middle and lower Yukon is for the most part so modern that it might with some justice be called ethnological rather than archaeological (de Laguna 1937:154). One may attempt to draw conclusions about lin guistic continuity based on a few common threads in the toolkits of ethnographic Athap askans and earlier prehistoric cultures. Athapaskans did use some of the same to ols that have been present for millennia in Alaska; for example chi-thos (simple stone choppers or hide-scrapers) are one category of artifact which connects ethnographic and la te prehistoric cultures with earlier prehistoric cultures in the region. It ma y nonetheless be problematic to infer an Athapaskan linguistic pedigree on the basis of such implements, regardless of context. J.L. Giddings makes essentially this point ab out the prehistoric Ko buk River culture of northwest interior Alaska: A close comparison of Kobuk archaeology with inland Athapascan ethnology might then bring out striking similarities, but these could only prove that Athapascans have recently made us e of ideas that were once in vogue along the Kobuk. It would not mean that Kobuk people of a remote time spoke an Athapascan language, or practiced an Athapascan form of culture. The significant fact re mains that we have practically no 48


archaeological knowledge of the Athapascan-speakin g peoples who have in historic times occupied nearly all of the American boreal forests (Giddings 1952:114; emphasis original). The strongest case for a Northern Athapaskan type-site may be Klo-kut in northwest Yukon Territory, near the Alaska border. There the historic phase is associated with Gwichin peoples. Indeed, lozenge-shaped Klo-kut (a.k.a. Kavik) type arrowheads are associated with late prehi storic Athapaskan societies in numerous northern locales, from historic Carrier and Chilc otin territory in central British Columbia all the way to central Alaska. Yet the si mple use of Klo-kut type points to diagnose Athapaskan presence is not without its problems they are also disturbingly similar to late prehistoric Inuvialuit Eskimo arrow heads found near Tuktoyuktuk, and to probable Yupik points from southwest Alaska (Morl an 1970b:29). Given the widespread interest in Athapaskan prehistory in the north, Richard Morlan called it an intemperate reflection when Frederica de Laguna (in her concluding address to a symposium of the Canadian Archaeological Associ ation) pointed out the elephant in the room: that there is no entity identifiable as Athapaskan prehistory (Morl an 1970a:2). Kavik points may be associated with the spread of particular ar chery technologies which were often, but not exclusively, used by Athapaskans. But the Klo-kut site itself demonstrates only about one millennium of Athapaskan presence in Gwichin te rritory, not more: In short, the Kavik type stone point ma y be one of our best hallmarks of late Athapaskan sites, and the absence of the type in the early period at Klo-kut may indicate that other diagnostics must be found for somewhat earlier horizons. The Klo-kut sequence, on the other hand, spans little more than the last millennium (Morlan 1970b:29). Such evidence has been used to argue that the Gwichin are late prehistoric migrants into their current te rritory (Hall 1969). But a similar lack of time-depth is 49


observed at numerous Athapaskan sites throughout the north, so pinpointing the source of the Athapaskan expansion has continued to be a challenge. Of the several sites with late Athapaskan phases noted by Edwin Ha ll only one, Aishihik, (Southern Tutchone territory, southwest Yukon) has any claim to being more than 1000 years old, and thus is the main evidence for tenative claims t hat Proto-Athapaskan culture descended from that of Archaic microblade users. Yet early Aishihik lacks many traits of the later, definitely Athapaskan phases (Ha ll 1969:317-318). So the notion that early Aishihik represents Proto-Athapaskan cu lture is not well support ed. John Ives writes: For later prehistory, essentially the la st 2000 years in interior northwestern North America . there is no single archaeological culture, no set of key artifacts or elements that can be conv eniently tagged as Athapaskan. . It is somewhat of an irony that, in virt ually every case given here, there is a fair degree of uncertainty about whether or not an Athapaskan prehistory has been described. . There does not exist a means to determine from lithic assemblages which language a prehistoric people spoke. And finally, the best correlations for individual artifacts or assemblages in northwestern North America are with gro ss environmental regions, and not maps of language distribution (Ives 1990:54-55). Most archaeologists think that Northern Athapaskan groups achieved their modern distribution by 1000 CE (Matson and Magne 2007). Some exceptional areas were populated slightly later; thes e include parts of southwest Alaska and central British Columbia which were settled by the thirteenth century (Dumond 1980:34; Dzeniskevich 1981; Wilmeth 1977). But to say t herefore that the Athapaskan pr esence in the center of this spread-zone is firmly established by 1000 CE is an exaggeration; in many cases this date is simply assumed due to a lack of firm evidence to the contrary. Don Dumond (1980:34) says [s]ites clearly representing pre-contac t age Athapaskan occupations are of almost legendary scarcity, however, and it is by no means certain that they were anywhere in Alaska before the late first m illennium BC. Even to say, as Dumond does, 50


that some Athapaskan sites in Alaska are as old as 2000 years (or more) reflects his abiding faith in the linguistic estimates of Michael Krauss. Investigating the cultural remains of Tanana Athapaskans of central Alaska, Anne Shinkwin expresses confidence in an Athapa skan cultural phase spanning most (but not all) of the second millennium CE, which canno t be linked to the earlie r cultures identified in the region during the first half of the first millennium CE (Shinkwin 1977:44). In other words, there is something like a 600-year gap of cultural indeterminacy in central Alaska during the middle of the Common Era. It may come as a su rprise that the earliest firm evidence for Athapaskan occupation of central Ala ska is (give or take a century or two) roughly the same time as that of the Denaina ex pansion into Cook Inlet, the Chilcotin penetration of the Columbia Plateau, or the Pacific Coast Athapaskan push into California (Dzeniskevich 1981; Whistler 197 9; Wilmeth 1977). There are even some credible recent efforts to dem onstrate that the A pacheans arrived in both the northern and southern portions of the Sout hwest only slightly later, as early as the late thirteenth and early fourteenth centur ies (Seymour 2008; Towner 2003). Athapaskans were virtually invisible in the first millennium CE, and then they appeared suddenly in the second millennium CE, rapi dly attaining a geographic distribution of unparalleled immensity for a New World family. This is a remarkable fact that begs some historical explanation. A Yellowknife Chipewyan Exception? The position outlined above reflects an adherenc e to the direct historical approach to archaeology. This method dictates that the only way to make linguistic inferences about prehistoric cultures is through direct historical observations of the ethnographic present and the recent historical past, which ma y form a bridge to late prehistory which 51


can be followed only as long as there is no majo r discontinuity in th e record. Otherwise, any case for the linguistic affiliation of arc haeological remains is purely speculative. The direct historical method has ye lided only one formidable case for unbroken continuity between Athapaksan speakers and archaeological cultures more than 2000 years old. William Nobles (1975, 1977) work in the cent ral district of Mackenzie, Northwest Territories has apparently revealed 2200 years of continuity using the direct historical approach, measured back in time from settlements of the historical Yellowknife band of Chipewyan. Noble conn ects the dots between about nine distinct archaeological cultures, with t he Taltheilei Shale tradition of circa 200BCE being the oldest and representing Proto-Athapaskan culture in Nobles view. Interestingly, he posits that Taltheilei may derive at least in part from the (presumably-Eskimo) Arctic Small Tool (AST) tradition, which is possi bly an admixed Eskimo-Athapaskan culture according to Noble (1975:777). An Athapaskan identity for Taltheilei does not seem obvious on the basis of apparent similarities with late AST; if correct, then it forces revision of the standard view that AST equates to Eskimo prehistory in the region. There are several problems with this work, however. Historical continuity back to Taltheilei times is evident only in the remo te and geographically isol ated region north of Great Slave Lake (adjacent to Eskimo terri tory to the east), along the northeastern periphery of the Athapaskan bloc as a whole. Similar historical continuity is not found among any of the other near by Athapaskan bands (Dogrib, Chipewyan proper, and Slavey). Only Chipewyan proper have anywhere near 1000 years of occupation in the region; most bands have substantially less. Ye llowknife oral tradition furthermore states that the people orig inally inhabited the south side of Great Slave Lake (near other 52


Chipewyan speakers) moving north only later. Yet the very old sites are exclusively on the north side of the lake. Noble is fo rced to dismiss these problems when he says Yellowknife origin mythology helps li ttle and, at present, appears to be highly unreliable (Noble1975:777). The Yellowknives are now extinct, suffering depopulation due to influenza and other diseases. The few survivors were absorbed by neighboring Dogrib and Chipewyan groups in the early twentieth-century (Gillespie 1981:288). Given that Yellowknife history is so sparse, it is diffi cult to rule out the possibility that descendants of Taltheilei shale people were not themse lves Athapaskans, but were simply amalgamated with intrusive Athapaskans in the second millennium CE, resulting in apparent material-cultural continuity masquerading as linguistic continuity. Native copper implements appear suddenl y during the Frank Channel Complex, 1300-1500 CE (Noble 1975:774). The ethnonym Yellowknife indicates that this particular band of Chipewyan was involved in the copper trade. Integration of Athapaskan speakers into Taltheilei descendant cultures after 1300 CE, followed by linguistic replacement, is another possibility. Nobles work suggests Yellowknives had a distinctive material culture (not shared by other Chipewyans), despite the fact that the two groups spoke closely related dialects and were seen as in distinguishable by t he earliest authorities. The material culture of these two Chipewyan groups appears to be operating independently of the linguistics. . Lex ico statistical estimates presently offer little help in dating the eastern Chipewyan sequence. . Hoijer (1956:228) once suggested that the eastern Chipewyan-Beaver Indian divergence occurred about 475 y ears ago, and that the Southern Athapaskans diverged from the No rthern Athapaskans (including the modern Chipewyan) less than 1000 years ago. Clearly the estimate . does not align with the archaeological date for Athapaskan assemblages in either northern Manitoba or central District of Mackenzie (Noble 1975:778779). 53


But neither does this archaeological evidence align with late eighteenth and early nineteenth-century Chipewyan an d Yellowknife oral traditions suggesting that the two groups were until recently one undifferentiat ed unit, and that the Ye llowknives migrated to north of Great Slave Lake at no very di stant period (Franklin 1824: v.2, 76). Chipewyan origin traditions stated that the group had migrated from the far west, originating on a continent on the other side of the Pacifi c Ocean (Mackenzie 1902: v.1, clxxiii). The linguistic and ethnological ev idence are in agreement; it is only the archaeological evidence that appears to contr adict them. The error could be in the interpretation of the evidence. To phrase Giddings (1952:114), I woul d suggest that the Yellowknife-Chipewyan simply made use of ideas that were once in vogue among the Taltheilei daughter cultures. It does not make sense that an area of low language diversity on the easternmost fringe of the northern Athapaskan expansion zone would have the oldest Athapaskan tenure at more than 2000 years, when areas of much higher Athapaskan language density in central Alaska have generally less than half this time-depth in direct historical sequence. The White River Ash Fall Given the extreme paucity of evidence for Athapaskans in the first millennium CE, there has been much interest to l earn what could have prompted the sudden expansion of Athapaskan speakers from a presumably small territorial nucleus. One popular theory was independently advanced by William Workman (1974) and David Derry (1975), and has been favored by severa l others since (Ives 2003; Moodie et al. 1992). Two eruptions of the White River volc ano in southeastern Alaska during the first millennium CE were responsible for spreading a white volcanic ash horizon several inches thick, northward and eastward over more than 125,000 square miles of territory 54


into the upper Yukon basin of the Yukon Territory and adjacent portions of Alaska. The first of these two eruptions occurred in the early second centrury of the Common Era, and the second occurred in the late eighth or early ninth century (Ives 2003:266). The epicenter of this volcanism is roughly the sa me as Michael Krauss (1973) opinion of the likely location of the Proto-Athapaskan linguistic homeland, and these events occurred prior to the early-second millennium arriva l of Athapaskan speakers in most of their current territories. Therefore, many sc holars have chosen to regard these two successive cataclysms, post hoc, as the events which precipitated the Athapaskan expansion in its various manifestations. Derry (1975:144) maintains that the second er uption of the volcano (in the late first millennium CE, responsible for the eastern ash field) spurred both the northward and southward movements in the Athapaskan expans ion. By his reckoning, the Gwichin penetration of North Alaska was one major result, as was the spread of Klo-kut arrowheads and bone and antler projectile points throughout Northern Athapaskan territory. Moodie and colleagues have similar views to those of Derry, but also attribute the rise in native copper exploitation among Athapaskan cultures to the flight of the volcanic refugees, who allegedly brought kno wledge of the metal with them from the White River, preserved as oral tradition and spreading as far as Yellowknife/Chipewyan territories in the east (Moodie et al. 1992:163). Workman (1975:254) suggests that this second (eighth/ninth centur y) eruption spurred both the Pacific Coast Athapaskan and the Apachean migrations, while Ives (2003:267) suggests that the PCA movement was spurred by the first eruption (second cent ury) and only the Apachean movement was 55


spurred by the second erupti on. Thus various authors hav e attributed almost all the Athapaskan migrations to the effect of White River ashf all, in one way or another. The actual expansion according to this model would be relatively fast but gradualistic, with displaced bands displacing other bands in a domino effect over the course of centuries, with subtle cult ural transformations ensuing which would presumably have masked the actual traces of t hese migrations in the intervening space. This all is necessary because there is no firsthand evidence of migration in the wake of the volcano. The model is based on the coincidence of Krauss (1973) linguistic hypothesis with the location and timing of a ca taclysm which took place prior to the dispersal ( post hoc ergo propter hoc). Workman descri bes the movements: The primary refugees, those affected di rectly by the ashfall, if successful in establishing themselves on the per iphery of the affected area, would have displaced in turn other groups of Athabaskan speech. . Eventually, at the southern periphery of the Athabaskan world, some groups would have been detached to the south into the domain of alien peoples. . It does not seem unreasonable that such small detached groups might be obliged to move a long distance before they found a land in which they could settle. . Attempts to trace this hypothesized movement south by archaeological techniques would seem premature. . This may never be possible in great detail as the materi al culture of these peoples may have been transformed as they spread (Workman 1974:255). A gradualistic model is necessary to ex plain the great demographic success of the refugees and the steady population increase as the people moved south. Considering the very low population density in the primary affected zone, as well as the fact that the most severe ashfall was in only a small fracti on of the affected area, it is counterintuitive that the Apachean descendant cult ures thousands of miles away from the event came to numerically overrun all of the well-established local cult ures in the Southwest. The notion that the primary refugees stopped on the edge of the affected zone and displaced others (who also expanded outward cascading fashion) makes the model 56


seem more demographically feasible, but it does not harmonize with the Chipewyan and Yellowknife oral traditions for the origin of copper metallurg y which place their cultural origins in Alaskas copper districts (Moodi e et al. 1992). Based on their argument of oral tradition related to the volcano itself, there must have been primary refugees well outside of the periphery, a dilemma which re introduces the paradox: how to explain the demographic and cultural success of a few displaced refugees? In its favor the ashfall displacement model is holistic, attempting to come to grips with the speed and scale of the Athapaskan expansion in it s entirety. Other proposed ar chaeological models of the Athapaskan expansion merit discussion (Wil cox 1981; Perry 1979), but space does not permit. Ethnographic Reconstruction of Proto-Athapaskan Culture Local manifestations of Athapaskan cultur e (particularly the divergent Southern and Pacific Coast Athapaskan offshoots) oft en closely mirror non-Athapaskan neighbor cultures. It is thus commonl y held that they borrowed t heir culture almost en masse from their neighbors (Tsuchiy ama 1947:1). In other words: Somewhere in the course of their migrations Athapascan blood became mixed with that of the Tanoan, Kere sian, Yunian, Zunian, and Shoshonian peoples . the strength of whic h may have allowed little but the Athapascan language to filter through (Wagner and Travis 1979:481). No doubt the incorporation of foreign cult ural elements throughout the duration of the Athapaskan expansion has accelerated t he obvious cultural divergence between Northern, Southern and Pacific Coast blocs. But no less problematic is the widespread view that all distinctive cultural el ements held in common between neighbors are borrowed by the Athapaskans from non-At hapaskans (rather than vice versa or independently reintroduced). This can be attributed largely to the tenuous 57


presupposition that proto-Athapask an culture is similar to t hat of the least developed tribes of the northern area (Tsuchiyama 1947:1), mean ing those complex traits observed in any particular Athapaskan society must have been derived from linguistically unrelated neighbor s via historical borrowing. This seems especially remarkable given another widespread presupposition; that Athapaskan languages themselves are extremely conservative and resistant to change. Oddly polar opposite tendencies are assumed for different por tions of the Athapaskan sociocultural spectrum; linguistic conservatism c ontrasts with cultural gregariousness. A shorter linguistic chronology for the beginning of the expansion (less than 2000 years) reduces the degree to which Athapaskan linguistic conservatism is noteworthy; shorter duration equals fewer opportunities for contact-induced change with neighbors en route. Likewise, a model which posits ex tremely rapid movements, particularly for the Southern Athapaskan migrants, would maximize their term of occupancy in their current homeland (to between circa 500 and 800 years BP), allowing for greater opportunities for acclimatizati on and acculturation. Rapid mi gration also greatly reduces the likely impact of their physical presence on the intervening landscape. The disparate trends (of linguistic homogeneity and cultural heterogeneity) are thus explicable without recourse to Athapaskan exceptionalism. Further corroboration of the common cult ural (rather than simply linguistic) template for the three disparate Athapaskan blocs is observable through the underutilized technique of ethnographic recons truction. Such work is pioneered by Perry (1983), to whom I am here indebted. T he method is straightforward in theory if not always in practice. Where distinctive cultural elements are found frequently in 58


multiple different Athapaskan contexts but rarely or never in intervening nonAthapaskan contexts, then these f eatures are strong candidates for inclusion in a list of Proto-Athapaskan culture traits When general trait catego ries are widespread and not ethnically bounded, reconstruction difficult, but still possible, if particular patterns of traits are frequent in Athapaskan contexts but ra re or absent elsew here. Such patterns could take the form of unique det ails or sequences of cultural motifs in close association in far-flung but linguistically linked contexts. Such occurrences can point to an early Athapaskan presence of features in question, and possibly to an Athapaskan origin for the system or template as a whole. Perrys (1983) work forms the starting point for this discussion. Eschatology: Death and Rebirth Na-Dene mortuary customs are shared in al l three geographic blocs. Cultures as far removed as the Tlingit, Upper Tanana, Western Apache, Navajo and Hupa all mandate removing a recently deceased pers on from a dwelling through some opening other than the door (Brugge 1978:313; Goddard 1904:70; Perry 1983:726). This is done to prevent the living and the dead from ever using the same path, which could provide a path for the ghost to return. This ghost is conc eived of only as the malevolent portion of a two-part entity; the benign or beneficent portion of which is not a threat to the living (Honigmann 1945:467-468). Nume rous Athapaskans residing in all three culture areas practiced strict name avoidance of the deceased (to prevent any invocation of malevolence) and frequently abandoned and/or destroyed the deceaseds possessions and dwelling (Goddard 1904:39, 7173; Honigmann 1945:469; Perr y 1983:726). Tlingit, Hupa, Tsuutina and Navajo all allowed a four-day period for the spirit to continue to occupy the corpse prior to the fu neral (Brugge 1978:315316; Honigmann 1945:46859


469). Elevated surface burials on a tree or platform (often reserved for specific categories of individuals like infants or shamans) have a wide distribution among NaDene peoples, reported for Tlingit, Nort hern and Southern Athapaskans (Brugge 1978:315; Hrdli ka 1905:493). The notion that o wls are fluent Athapaskan speakers and harbingers of death is f ound in Athapaskan contexts as far removed as Denaina and Apache (Hrdlika 1905:489; Perry 1983:726). The extreme physical danger posed by cont act with corpses is well known among Southern and Pacific Coast Athapaskans, and some Northern Athapaskans like the Tsuutina and Deg Hitan; the danger is particularly acute for relatives of the deceased (Goddard 1904:72-73; Honigmann 1945:469; Opler 1936:222223). Some Canadian Athapaskans are an exception. For example, Carrier widow s carried the bones of their dead husbands with them (hence the name Carriers). Morice (1925:576) suggests that these anomalous mortuary practices were recently borrowed from neighboring Tsimshianic coast-dwellers, as they are not practiced by more interior-dwelling folk. This explanation appears plausible; much fa rther north, Athapaskan eschatology falls into a pattern more reminiscent of the two southerly branches. Alaskan Athapaskans, like the Deg Hitan, dem onstrated an extreme fear of corpse s rivaling that of their distant Apachean kin (Dall 1897:67). Perry notes: it is likely that in Proto-Athapaskan ideology the human spirit was believed to have two components, one of which was associated with breath, vapor, wind, or air, and the other conceived of as a shadowlike entit y. One of these continued to be associated with the corp se for a time, r epresenting danger to the survivors. Practices intended to deal with this included avoiding the name of the deceased, destroying or abandoning his property, removal of the corpse through an opening other t han the doorway of the dwelling, and destruction or abandonment of the dwel ling in which death had occurred (Perry 1983:727). 60


Residence Patterns and Familial Descent Matrilocal residence is preferred by the va st majority of Northern and Southern Athapaskans; Perry (1989) identified 27 different Athapaskan ethnolinguistic subdivisions, distributed uniformly throughout both major geographic blocs, for whom this is the predominant residence pattern. Furthermore, at leas t twelve Athapaskan groups (clustered in the western halves of thei r respective blocs) primarily reckon their kinship and descent through the mothers line (matrilineally). The Pacific Coast Athapaskans are the exce ption to this rule, fitting wit hin the California pattern of patrilineal/patrilocal kinship and residence. Early twentieth-century anthropologists were uniformly of the opinion that the western Athapaskan kinship system was a product of diffusion from the Northwest Coas t, but more recent scholarship recognizes that the diffusion explanation is too simplistic (VanStone 1974:52). Although it is often assumed that the matrilin eal organization of the Athapasca ns . and of the Eyak, represented coastal influence, the opposite may be true (Drucker 1963:198). The high frequency of matrilineal organization among Athapaskans in Alaska, the Canadian Cordillera, and the Southwest are indicative that these patterns were already prevalent at the Proto-Athapaskan stage of development, and possibly earlier (Jett 1978; Murdock 1955; Perry 1989). The easternmost groups, both north and south, emphasize bilateral descent systems, possibly as an adaptation to specific environmental circumstances (Perry 1989:40; VanStone 1974:53). Earlier scholarship attributed the develop ment of matrilineal-matrilocal patterns among Northern Athapaskans to a dependence on salmon, and the importance of womens work in riverine economies. But Athapaskan specialists no longer believe that most groups were as dependent on fi sh as had been previously supposed 61


(VanStone 1974:52). Some matrilocal Northern Athapaskans even practice strict fish avoidance, like the Sekani who disdain fish of any kind and regard fishing as a degrading occupation unworthy of a hunter (Morice 1889:130). In their juxtaposition of a strong disdain for fish with matrilocalit y, Sekani are like all of the Southern Athapaskan groups (Landar 1960). Richard J. Perry (1989:43-44) favors a sc enario where matrilocality and matrilineal descent patterns were reinforced among mount ain dwelling Athapask an groups of the west (both north and south) who featured semi-permanent or permanent femaledominated base camps and high rates of male absenteeism during seasonal hunting trips; matrilineality was subsequently lost among more mobile eastern groups where women traveled with the men. Yet matriloca lity was largely preser ved in the eastern Athapaskan range, even as matrilineal des cent was apparently abandoned. Perrys model favors the preservation of matrilocality, but does not quite explain matrilineality. VanStone (1974:53) suggests that the shift to bilateral kinship reckoning in the easternmost bands may have been a result of more extreme environmental factors, as a larger number and variety of kinship affiliations is an adaptive; more kinfolk is a larger support network increasi ng the odds of an individuals survival. Regardless of the ultimate cause, matrilineality is an anomaly among foragers, with only thirteen of 101 hunter-gatherer groups in Murdocks ( 1957) world ethnographic sample being matrilineal, six (or seven) of the thir teen are Na-Dene speakers (Perry 1989:34). Some mid-twentieth-century scholars suggested that the original Na-Dene migrants from Asia may have brought a matr ilineal descent system with them from the Old World, and this system was subsequently lost among the easternmost Athapaskans 62


(Birket-Smith and De Laguna 1938: 449; Garfield 1953:61; Murdock 1955:86-87). This would help to explain why forager social-structure is so decidedly un-forager-like in this case (Kelly 1995:270-278). The Asian-origin s view was conditioned to some degree by the linguistic evidence for a relationshi p between Na-Dene and Si no-Tibetan languages (Sapir 1925, 1991; Shafer 1952). Some culturally Tibetan groups (like the Naxi) were matrilineal up until the twentieth -century, as were some Al taic dynasties of Inner Mongolia during the early Common Era. The Chinese are today patrilineal, but philological and archaeological evidence su ggest that matrilineal descent may have been the norm among horticulturalists in Neo lithic China (Pearson and Underhill 1987). Criticism of the Asian-origin hypothesis fo r northwestern North American matrilineal social structure rests mainly upon the extreme distances between known matrilineal societies of interior Asia, and those of No rth America (Garfield 1953:59). Furthermore, the old pervasive skepticism toward theorie s of Athapaskan-Asian language relations is easily extended to theories of the importation of matriliny, as complex, speculative explanations for which there is li ttle empirical basis (Perry 1989:47). Female Fortitude and Spiritual Pote ncy in Mensturation and Childbirth Athapaskan societies in all three geographic blocs (Northern, Southern and Pacific Coast) shared a common belief in an extr eme cosmic danger posed by menstruating women, and in the particular importance of female puberty rituals. These sorts of beliefs are not unique to Athapaskans, but are nearly ubiquitous in the American West. But Athapaskan examples include distinctiv e features and are not eworthy both for the intensity of these beliefs and the degree of importance placed on associated ritual observances. The emphasis on public celebrat ion during puberty rituals is more intense in the Southern and Pacific Coast blocs, wh ile the emphasis on seclusion of the new 63


menstruant is more intense as one moves north, however this Athapaskan puberty/menstruation/childbirth ri tual-observance pattern incl udes a series of similar public and private elements found in multiple geographic areas. The isolation phase of the puberty ritual lasted as long as a year among some Gwichin, a month among the Upper Tanana in Southern Alaska, and as few as ten days among the Kaska in British Columbia (Perry 1983:724). Both extremes were present among the Denaina, as seclusion could last anywhere from one week to one year, depending on who was asked (Osgood 1937:162). Among t he Chiricahua and Western Apache, this entire phase is condensed into a period of a few preparatory days leading up to a major public ceremony. Yet in all these disparate groups (plus Navajo) the girl is forbidden from touching herself and is given a scratching stick to remedy itches. All of these (plus Mescalero and with the exception of Kaska) likewise forbid the girl from directly touching water, providi ng her with a drinking tube. The gaze of the menstruant is especially hazardous, and many Northern groups compel her to look downward and utilize special face covering hoods to prevent inadvertent eye contact. The Upper Tanana and Mescalero Apache forb id her from looking at the sky, and Eastern Navajo also cover her face (Driver 1941:33; Perry 1983:724). In many respects femaleness entailed a sense of danger. Practices designed to avoid this danger included menstrual seclusion, the use of scratching sticks and a drinking tube, and seclusion or purification associated with childbirth. Although girls' puberty recognition and menstrual seclusion are almost ubiquitous in western North America, the Athapaskan pattern again shows a number of dist inctive aspects (Perry 1983:723-724). Throughout much of Northern California, puberty ceremonies are frequently the most important social gatherings held (akin only to Apache puber ty rites in that regard). Among native Northern Californian s, the immediate family of the pubescent girl is 64


imperiled by ghosts upon the commencement of t he public ritual. In great contrast to this feeling of fear of ghosts on the part of the girl and her family is the spirit of merriment enjoyed by the crowd (Driver 1941:33). Like both Southern and Pacific Coast Athapaskans, central Alaskan Athapaskans also have a public celebration as a major component of the puberty rite (Pe rry 1983:725). Southern Athapaskans believe that the pubescent girl has access to tremendous healing power which may be transferred to the participants in the public celebration; this is an unusual characteristic of the Apachean rite that is shared with only the North Pacific Coastindicating a likely historical connection between t hese culture areas (Driver 1941: 34). In particular, the similar public puberty rituals of the Souther n and Pacific Coast Athapaskans are highly suggestive of a common origin in the Canadi an Cordillera within the last ~1000 years, as Driver maintains. The predominant type of public puberty ceremony originated on the Northwest Coast and spread southward mainly by migration, perhaps entirely in the custody of Athabascans (Dri ver 1941:62). He elaborates: Athapaskans in the Southwest brought their puberty rites with them in their migrations from north to south. T hese rites share more with northern California rites than with those of any other area. The presence of Athapaskans in northern California fits into the picture. . I believe . that at least some of the puberty details shared by northern Californians and Apacheans are heritages from their common protoculture (Driver 1969:906). To Athapaskan women is attributed great spiritual fortitude allowing them to withstand the spiritual danger s posed by menstruation and childbirth. Among the Tagish and Southern Tutchone, women are disc ouraged from hunting, not for lack of requisite skill, but because the spiritual power of menstruation is frightening to the prey (Cruikshank 1979). Similarly, menstruation is considered disruptive to Navajo ceremonial efficacy; a menstr uating chanter or singer (hataahi) attempting to propitiate 65


beneficent deities would risk driving them away. Only men and postmenopausal women are permitted to learn and practice the Navajo chantways and holiness rites (Aberle 1982:227). A similar situation per tains for Pacific Coast Athapaskans, as among the Hupa the chief thing that contami nates the world is menstrual blood. This belief in the contaminating influence of menstrual blood is a fundamental concept running through much of Hupa re ligion (Driver 1941:29). Strict male avoidance of childbirth is a ssociated with a similar concept of spiritual contagion (capable of causing physical disease to men) among the Denaina, Upper Tanana, Gwichin, Chipewyan, Slavey, and Chiricahua Apache. Women in these societies are respected for their spiritual strength allowing them to withstand this natural danger (Perry 1983:725). Among the Chipewyan, this belief extends to the physical realm, as one male informant maintained that men are the weaker sex and that one woman has the strength and endur ance of two men (Abel 1993:22). Although European observers later interpreted these taboos as evidence of womens inferior status in Dene society, Dene women themselves did not see them this way. Menstruation was the sign of a potent power particular to women, and from a sens e of responsibility to t he community, women felt obliged to be cautious and to avoid causing harm at those times (Abel 1993:21). Having summarized the ethnological ev idence for proto-Athapaskan culture, Chapter 2 will now conclude with a brief review of the classical literature in Athapaskan physical anthropology. Physical Anthropology and the Athapaskans Background The distinctively Asian appearance of many Athapaskan speakers has been recognized for many years, and early bi ological anthropologist Ale Hrdli ka (1925:494) 66


was quick to suggest that this virile brach ycephalic type represented a later wave of migration from Asia than the other so-ca lled Indian groups. Yet biological studies attempting to differentiate Athapaskan populat ions from their non-Athapaskan neighbors falter on the premise that cont emporary ethnolinguistic cultures are static, fixed in their ancient residence, and possessing of an enduring racial component. On the contrary, it is abundantly clear that multiple languages with fundamentally different geographical origins may co-exist within a single biol ogical population, and simultaneously peoples with different biological histories may share in common a single, roughly homogenous ethnolinguistic culture. W herever Athapaskans have resided in close proximity to nonAthapaskans, there has been frequent interma rriage across linguistic boundaries and diffusion trends which are subject to change generationally. For example, in British Columbia, wher e Nuxalk (Salish and Bella Coola) peoples and Athapaskans have been neighbors for centur ies, a significant number of protoAthapaskan loanwords in Nuxalk languages suggest the prestige of Athapaskan languages in the region was very high in prehistory, and asymmetric bilingualism favored the expansion of Athapaskan lang uages into non-Athapaskan populations. Subsequently, the status of two major ethnic groups has shifted radically, so that the acculturation gradient essentiall y reversed itself in recent centuries, with the prestige associated with Nuxalk languages becoming much higher than Athapaskan. Thus, in later generations, the dominant flow wa s of Salishan loanwords into Athapaskan languages like Carrier and Chilcotin, and the b ilingualism differential began to favor the expansion of Salishan groups over Athapaskan ones (Nater 1994). The implications for 67


human biology are clearthe Subarctic-Plat eau boundary does not def ine biologically distinct populations, but is a fluid cu ltural frontier within a single gene pool. Even as ethno-linguistic fr ontiers appear to be sharply defined by different physiographic zones (Arctic, Subarctic, Pl ateau, etc.), there have invariably been shifting patterns of multilingualism across these frontiers. Ernest Burch and Thomas Correll posed the question regarding the natives of Northern Alaska. Are individuals with an Athapaskan father and an Eskimo mother Indian or Eskimo? Are groups who spend half the year inland and the other half on the coast inland or coastal? (Burch and Correll 1972:19; emphasis original). There are profound cultural differences between language communities, but these are subj ect to change rapidly; innumerable discontinuities in the late prehistoric archaeological record b ear this out. Prior to the Athapaskan expansion, the protolanguage occupied a small homeland (or beachhead) whose periphery may have incl uded any number of obscure American Indian neighbors whose languages are unknowabl e (apart from an opaque substrata) but whose peoples influenced the eventual biologi cal and cultural makeup of the mobile Athapaskans. Burch and Correll note, regarding the Athapask an-Eskimo frontier in North Alaska: If someone travelled to another region, he could communicate through the local dialect, or use his "home" dialect with nearly equal fluency as he chose. He thus maximized or minimi zed social barriers through linguistic devices according to the requirements of the situation. . For example, offspring of mixed Koyukon (Athapask an) Inupik (Eskimo) marriages on the upper Kobuk River typically had the choice of passing as Indian or Eskimo simply on the basis of the language that they decided to speak, and this could vary from one set of ci rcumstances to another (Burch and Correll 1972:24; emphasis original). Recent generations have witnessed the conversion of entire bands of Athapaskans to the native linguistic culture of their neighbors. For example, by 1900, the Athapaskan residents of t he bilingual village of Kvygym paynagmyut overwhelmingly 68


favored Eskimo language for everyday use outside the home, and frequently intermarried with their Eskimo neighbors; there were consequently fewer than 20 speakers of their Kuskokwim dialect of Deg Hitan remaining in 197 8 (Snow 1981:602). The same situation pertains for the Tagi sh language (or dialect of Nahanni); by the 1880s Tagish were essentially considered to be an Inland Tlingit band, Athapaskan in its cultural roots but no l onger in language. By 1974, onl y one or two Tagish elders could remember any of their native At hapaskan language, while fluency in the distinctive local dialect of Tlingit was commonplace (McClellan 1981:481). Such band- conversions must have occurred in prehistor y too, introducing genes from Athapaskan speakers into nearby non-Athapaskan populations. The protracted circumference of any major Athapaskan bloc may have seen par ticular groups shift linguistic identity multiple times. If the cultural capital of Northern Athapaskan speakers was much higher in the past (as it seems was definitely the ca se on the Subarctic-Plateau frontier) then it is very likely that non-Athapaska n peoples were assimilated to the linguistic culture of the intrusive Athapaskans, in the reverse of the process described for the Tagish and Deg Hitan above. Such appears to have been the case for the Southern Athapaskans. Apacheans have expanded their ranks by liberally assimilating local peoples; their southern incursion apparently served to amalgamate a loose network of preAthapaskan hunter-gatherer so cieties in the greater desert Southwest under one cultural banner; this is evidenced by the in credible speed with which they inflated their ranks from modest foundations and cl aimed vast territories which had been continuously occupied by others (Upham 1984: 250). This is borne out in recent molecular genetic studies which suggest that the Apacheans liberally recruited and 69


assimilated small groups of socially and economically marginalized peoples . living on the Puebloan fringe (Malhi et al. 2008:419). A similar means of demographic expansion was apparently employed by Pacific Coast Athapaskans, albeit less rapidly and on a slightly smaller scale (Jacobs 1937). Regarding the inherent biological or racial character (or the lack thereof) of these Athapaskan assimilators, the assessment of Franz Boas is still as relevant today as it was seven decades ago. It is true that wherever we find tw o tribes speaking affiliated languages there must have existed blood-relati onship; but we have abundant proof showing that by infusion of foreign blood the anatomical types have changed to such an extent that the original type has been practically swamped by the intruders. Such is the case in North America among the Athapascan tribes of the Southwest. . The laws according to which anatomical types are pres erved are not the same as those according to which languages are preser ved, and for this reason we must not expect to find the results of classifications based on these two considerations to coincide (Boas 1940:153). The effect of cross-cultural unions has been to alloy the local populations of two different linguistic stocks, hence appearing biolog ically much closer to each other than either one is to far flung me mbers of the same stock (e .g. accelerating biological divergence between Athapaskans in differ ent geographic blocs), even as cultural differences between them may be preserved or even strengthened. To the extent that biology and language are correlated, the signal is thus weak and not easily detectable at large geographic scales; linguistic boundaries are no significant barrier to gene flow (Hunley and Long 2005; Wang et al. 2007). Do these conf ounding scenarios prevent any rational discussion of the distinct biolog ical origins of the Athapaskan languages? Not necessarily, if we are mindful of all the historical uncertainties. The relationship between genes and languages may be subtle and complex (Wilson 2008:271). Intense interethnic admixture appears to be the rule ra ther than the excepti on (at least in the 70


North American case), and so researchers should learn to expect it, rather than to treat it as an anomaly or a confounding variabl e. It is to be expected should be acknowledged as such. Regarding genetic differentiation (or la ck thereof) along the Athapaskan/Algonkian continuum in the Subarctic, Christopher Meik lejohn (1977) notes that the internal structure of biological units in the Subarctic is only of a fleeting reality. Gene flow immediately overrides incipient deme forma tion, reflecting the existence of an: apparent paradox with populations t hat are both locally inbred and regionally outbred at the same time! . The implications for long term deme structure among Athapaska ns are considerable. I severely question whether effective demes beyond those ephemerally observable at one point in time occur. . The operation on group size constraints in conjunction with interband social cohesion will ensure t hat local isolates fail to survive. Beyond this, restrictions upon group cohesion wrought by the boreal forest ecosystem will ensure furt her disruption and effective gene spread. . Short term genetic study results must be viewed as a narrow slice in a complicated temporal continuum in wh ich long term gene flow will override the fleeting face of differentiation (Meiklejohn 1977:109-110). The implication of Meiklejohns argument is that, where population density is low, cultures easily shift back and forth bet ween endogamy and exogamy as circumstances demand. Biometrics and Craniofacial Analysis In pioneering work, Harry Shapiro (1931:378) observed that within-group variation among different so-called Eskimos was much greater than the between-group variation between specific bands of Eskimo and At hapaskan Indians. Anthropometric data suggested that the Chipewyan near Huds ons Bay (central Canada) and the distant Eskimo of Seward Peninsula (northwest Alaska) were essentially one biological population, even though they have no linguistic heritage in common. George Neumann (1952:27-29) in a broad survey of biometric data, postulated a Deneid physical type 71


which included Northern and Southern At hapaskans, Northwest Coast and Plateau Indians, and Yupik Eskimos. Aleutian Isl anders represent one ex treme in the range, and ultimately diverge from the Deneid c ontinuum, suggesting that southwest Alaska could be the beachhead for this physical ty pe within the continent. Neumanns data excluded a small percentage of Athapaskan and Haida crania which fell within the range of variation for eastern woodlands populati ons, suggesting the possibility of horizontal language transmission (i.e. the spread of Na-Dene languages to non-Na-Dene peoples), and that earlier, less specialized populations were absorbed by later more specialized ones. However, the vast majo rity of his Deneids are Athapaskan speakers and the vast majority of Athapaskan speakers are classified as Deneids. Furthermore, the Deneid type is more st rongly associated with a putat ive Central Asian source population than is any other group: the bulk of this group is associated with the expansion of speakers of languages of the Athabaskan st ock. It is hardly necessary to point out that a linguistic and physical correlation will not prove to be too high. There are also Skittagetan, Wakashan, Penutian, and other groups that are Deneid. . It represents one of the last of the ma jor migrations to the New World, it is the group that exhibits the most ma rked Asiatic connections, it exercised profound influences on both North Pacifi c Coast and Southwestern cultures, . from perhaps A.D. 1200 onward (Neumann 1952:28). Neumanns Deneid physical type corres ponds almost exactly to the distinct Northwest Coast/ Na-Dene dental pattern noted by Greenberg, Turner and Zegura (1986). Building upon earlier work, Christy Tu rner and Richard Scott (2007) maintain that the sinodont dentition of Eskimo-Aleuts and Na-Dene/Subarctic/Northwest Coast peoples are distinct enough from each other and fr om the rest of Native Americans to suggest that the Na-Dene and Eskimo-Al eut must represent two independent 72


migrations, more recent than others in the continent. Regarding the physical similarity of Deneid crania to those of Lake Baikal in Siberia, Neumann (1952) continues: It is quite certain that a considerable part of the wave of people that became the Deneids in North America must hav e remained in the Transbaikal area. . This would explain the existence of a complex of morphological traits common to many groups extending fr om Lake Baikal to the Aleutians, and from there south to the Apache of Arizona and New Mexico (Neumann 1952:29). As the history of the relev ant populations has become be tter defined by archaeology, a paradox has emerged. The probl em is that the biological populations in which these profound phenotypic similarities appear (that is, the comm on traits of Deneids and Southwest Siberians) are not shown to have gr eat antiquity in their present locations. This dilemma is cogently express ed by Alexander Kozintsev et al.: Contrary to expectation, the earliest known inhabitants of the Baikal area, who lived there in the sixth and fifth m illennia BC, were extremely flat-faced Mongoloids. Their descendants began to look less Mongoloid (and accordingly more Amerindian) only in the fourth millennium BC, and this tendency continued in the 3d millennium BC . . Clearly, the only possible explanation is gene flow fr om the more western regi ons of southern Siberia inhabited by the Caucasoids. The resulting similarity of late Baikalians with the Amerindians, then, is superfici al (Kozintsev et al. 1999:194). There are two competing t heories regarding the human bi ology of Central Asia, whose populations (like Amer ican Indians) are physically intermediate between the socalled mongoloid and caucasoi d phenotypic extremes of east and west Eurasia. The admixture hypothesis suggests that t he Bronze Age witnessed a convergence of specialized eastern and western types in the Eurasian heartland, and this convergent/admixed Inner Asian phenotype is fortuitous and only superficially similar to that of many of Americas nat ive peoples. Alternatively, t he hearth hypothesis reverses this hypothetical gene-flow gradient, suggesting the eastern and western phenotypic extremes represent the culmi nation of long processes of di vergence and diversification 73


from a centrally located intermediaries in a largely undifferentiated Eurasian source population (i.e. the hearth), ancestral to all three groups (so-called caucasoids, mongoloids and americanoids). Doubtless both hypotheses have kernels of truth, and the actual process may have been one of pendul um-like movements of peoples in and out of the Eurasian heartlandjust as back and forth movements in the north Pacific Rim are also feasible. Kozintsev and co lleagues ultimately favor an intermediate scenario, whereby the Bronze Age Okunev and Sopka peoples of the Yenisei and Ob headwaters of Southwest Siberia are collateral relatives of American Indians, sharing a direct common ancestor, yet who nonetheless appear (unlike their North American kin) to have been thoroughly admixed with west Eu rasian caucasoids during the Bronze Age (Kozintsev et al. 1999). The phenotypes of Northwest North Amer icans and Central Asian populations evidently developed in tandem since the Bronze Age. Some of this development can be attributed to uniform responses to environment and the plasticity of phenotypes in both America and Siberia: ancient groups [are] mostly characteri zed by broad face and wide orbits, . [more so] than modern ones . regardless of origin. . This vector, then, reveals a universal diachronic tendency (evidently related to the gracilization phenomenon) and is accordingl y irrelevant for tracing genetical affinities (Kozintsev et al. 1999:200). However, such a parallel drift of phenotyp es in response to environmental stimuli is not sufficient to account for the overa ll degree of physical similarity between Late Holocene East Central Asian (ECA) peoples and the peoples of the Americas, and for the suddenness of its appearance. But efforts to connect Deneid and Late Holocene Baikallian biometrics are complicated by the hi storical heterogeneity of populations in both regionsuncertainty is pervasive and a satisfactory resolution is not immediately 74


forthcoming. Nonetheless, craniometric analyse s using traditional methods yield a clear consensus that some degree of Late Holoce ne intercontinental migration is the best explanantion for the post-N eolithic convergence between Asian and Native American phenotypes; the suddenness of the Late-Holocene morphological changes are more consistent with dispersal from Asia rather than evolution in situ, as Gonzlez-Jos et al. point out: In summary, craniofacial studies s eem to support a scenario in which America was successively occupied by two morphologically differentiated human populations, with the generaliz ed (e.g., Paleoamerican) morphology first entering the New World and being replaced or assimilated by groups carrying derived traits (Gonz lez-Jos et al. 2008:177). Physical anthropologists for the better par t of a century hav e agreed that the biometric data support multiple migrations out of Asia into America. It has only been with the advent of molecular gen etics that this view has been significantly challenged by contemporary scholars. Chapter 3 will address cutting-edge linguistic and molecular genetic evidence, in the contex t of comparative ethnology. 75


Figure 2-1. Na-Dene in the No rthwest (from Dumond 1969:859). 76


CHAPTER 3 THE DENE-YENISEIAN FAMILY IN ANTHROPOLOGICAL PERSPECTIVE Pushing the Linguistic Boundaries: the Dene-Yeniseian Phylum The cutting edge in Athapaskan studies is embodied by the work of the February 2008 Dene-Yeniseic Symposium at the Univ ersity of Alaska, Fairbanks. This symposium included a large contingent of th e worlds leading experts on Athapaskan languages and Athapaskan archaeol ogy, and was convened for the sole purpose of evaluating nearly a decade of research conducted by linguist Er ic Vajda (at University of Western Washington) demonstrating the possibility of a hist orical relationship between Athapaskan, Eyak and Tlingit on the one hand, and the Yeniseian languages on the other. Dene-Yeniseian is now positioned to gain widespread recognition as the first Asian-Amerindian language family, a remarkable feat considering t hat Yeniseian people reside in South Siberia and Central Asia, ~ 7000 miles away from the nearest Alaskan Athapaskans. The symposium and the publicati on surrounding it arguably represent the beginnings of a paradigm shift in Americani st anthropology, as self-described skeptics are seriously considering long-distance genetic relationships between particular Old World and New World cultures, after decades of staunch opposition to similar proposals. Bernard Comrie introduces the sympos iums work in a recent volume of the Anthropological Papers of the University of Alaska: In the literature there have been vari ous proposals concerning genealogical relations between languages of the Americas and of Siberia. So far, these proposals have not succeeded in convinci ng the majority of specialists working on languages of t he Americas nor those wo rking on languages of Siberia, nor those working in histor ical-comparative linguistics. The importance of Vajdas work on the Dene-Yeniseian hypo thesis is that for the first time evidence for a geneaological relationship between languages of the America and of Siberia has been presented (excluding, of course, the case of Eskimo . .) which not only satisfies the methods of historical77


comparative linguistics but also su cceeds in convincing skeptics like me (Comrie 2010:35). A global literature review of the backgr ound to the Dene-Yeniseain hypothesis is presented by Vajda (2010b), so there is no pressing need to cover the same ground. Nonetheless, a brief sketch of some importa nt milestones in Engl ish-language literature is helpful, to illustrate that Dene-Yenise ian represents a refinem ent (rather than a replacement) of earlier proposals, and is the cu lmination of a long hi story of scholarship. Yeniseian, Tibetan and Na-Dene Yeniseian language is now cr itically endangered; the Ke t of western Siberia are the only remaining members of this previo usly wide-ranging family, and the number of fluent speakers is less than 100. The question of the geneaological st atus of Yeniseian languages in Eurasia has intruigued scholars for generations. As a linguistic isolate, they are often grouped for convenience with the Paleo-Asiatic languages of the Russian Far East (Chukotko-Kamchatkan, Yukaghir and Nivkh), although they have virtually nothing in common with them. Paleo-Asiatic (o r Paleo-Siberian) is really a dustbin category for unclassified languag esthe designation simply indicates a possibility that Yeniseian languages were present in Siberia before the expansio n of Neo-Siberian reindeer pastoralism that took place during t he Common Era. In te rms of their possible geneaological affinities, a stronger case can be made for a relationship between Yeniseian and the languages of East Central Asia to the south, than between Yeniseian and other languages of Siberia, paleo or otherwise. Affiniti es with Tibetan in particular have been remarked upon by many. One of the earliest to do so in print was James Byrne, who wrote using data collected when multiple Yeniseian languages were still spoken. Describing Ket and Ko tt (now extinct), he wrote: 78


The internal changes which the stems of nouns and verbs undergo in these languages, and which make them so unlik e the other [Siberian] languages . are points of resemblance to the [Chinese and Tibetan] languages . ; those of the verb especially having a ce rtain resemblance to the structure of the Tibetan verb. And it is remarkabl e, that the numerals of both languages [Ket and Kott] resemble the Tibetan numer als. . It se ems most probable, therefore, that these l anguages are originally akin to the Tibetan, and have been altered by mixture with the Siberian languages (B yrne 1885:472). Sapir (1925, 1991) and Shafer (1952), usin g the historical-comparative method, both suggest geneaological relations between Na-Dene and Tibetan based largely on similar verb morphology as described abov e. There is three-way structural resemblance between Yeniseian, Tibetan and Na-Dene verbs. A cadre of midtwentieth-century Soviet linguists were inst rumental in advancing the hypothesis of Yeniseian-Tibetan language links under the auspices of the multilateralist SinoCaucasian proposal, which included also the North Caucasic languages and eventually Na-Dene too (Starostin 1991). But Americ an archaeologst Henry Co llins (1954:35-36) was likely the first to suggest that the Y eniseian languages could be a Siberian intermediary family between Sino-Tibet an and Na-Dene, and that Na-Dene and Yeniseian could have a special relationship. A similar theory was advanced by Merritt Ruhlen (1998), who formally propos ed the Dene-Yeniseian family: Although I do not question that Na-Dene is related to both Sino-Tibetan and Caucasian (within the la rger Dene-Caucasian family), . the evidence presented below indicates that Na-D ene is more closely related to Yeniseian than to either of these other two families (Ruhlen 1998:13994). Ruhlens publication in a high-profile venue (Proceedings of the National Academy of Sciences ) generated considerable attention in t he scientific community, but his work was not well received by the linguists of the historical-comparative school, because Ruhlens work is based in the discredited mu ltilaterial-comparative method prescribed by Ruhlens mentor Joseph Greenberg (1987). 79


However, Ruhlens work did inspire Eric Vadja to dig further into the problem using more rigorous orthodox methods, and Vajdas (2010a, 2010b) work has since been well received by a majority of specialists. Whereas Sino-Dene and Dene-Sino-Caucasian both met with controversy, it is now safe to say (thanks to Vajdas exacting work) that Dene-Yeniseian is rapidly gaining widesprea d acceptance. As fa r as validating the multilateral-comparative method, the question remains open. A judicious statement would be that this method can generate inte resting hypotheses (like Dene-Yeniseian), but that traditional methods mu st still be used to prove them (Michalov et al. 1998:465). In contrast to Ruhlens proposal (w hich suggested a Dene-Yeniseian branch, including Haida, on the Dene-Caucasian tree) Vajdas proposal does not assume that Dene-Yeniseian is necessarily related to Hai da or Sino-Tibetan. This narrower view of Dene-Yeniseian includes only Athapaskan, Eyak, Tlingit and Yeniseian, although Vajda does report one triple-resemblance in protoforms which he calls striking; the word liver, Yeniseian: *s Na-Dene: *s nt Sino-Tibetan *m-sin (Vajda 2010b:114); for additional Sino-Yeniseian comparisons, see Sedl ek (2008). Regarding the possible inclusion of Si no-Tibetan and Haida in par ticular, Vajda writes: Though I have excluded Sino-Tibetan from the present study due to difficulty in assessing historical Ti beto-Burman verb morphology, I would single it out as a promising potential rela tive of Yeniseian and Na-Dene. . My comparison of Yeniseian and Na-Dene failed to turn up new evidence in favor of linking either family with Haid a. If Haida is indeed related to Athabaskan, it would appear to be at ti me depth older than the link between Yeniseian and Na-De ne (Vajda 2010b:114-115). Sino-Tibetan and Haida remain the two most plausible candidates for more distant relatives of the established Dene-Yeniseian fa mily, but it is unclear if Haida or SinoTibetan is closer in this regard (Figure 3-1). 80


Implications Regarding the Athapaskan Expansion The great geographic gulf between Yeniseian and the other members of the family means we can no longer assume that Na-D ene familial diversification has taken place entirely within the New World. The poi nt of departure between nearest common ancestors may be anywhere on the long (possi bly indirect) journe y between southwest Siberia and Alaska. When this occurred is an open question. Tr aditionalist linguists generally argue that the maximum time-dep th for the useful application of the comparative method is between 4000 and 6000 years, a range withi n which most well established linguistic stocks are comfortably placed (Matson and Magne 2007:133). Few anthropologists would place stock in lexicostatistics as a means for establishing firm dates, but the approximate cap of 6000 years is still widely assumed as a terminus ante quem; when the loss of core vocabulary becomes too great to allow determination of genetic relationship. The strength of the Dene-Yeniseian conn ection (now generally acknowledged) is an apparent exception to th is rule, because the physical gulf seems too great for a historical relationship withi n these timeframes. But this opinion is conditioned by non-linguistic evidence (both archaeological and genetic) which is subject to a variety of interpretations. It is also feasible that Dene-Yeniseian culturalhistorical relations are within the estab lished 4000-6000 year windo w for the utility of lexicostatistical methods, but that the material culture trac es of this relationship remain elusive. The geographic span between the two branches (Yeniseian and Na-Dene) constitutes the largest (and possibly the fast est) pedestrian language spread on earth. Johanna Nichols thus claims the Dene-Yeniseian family appears plausible on statistical grounds but is nonetheless geographically pr oblematic (Nichols 2010:299). Some 81


regard a very slow migration rate as more likely and see the spatial gulf in the DeneYeniseian family as necessitating many mille nnia to transpire. James Kari asserts: [t]he Dene-Yeniseian language stock renders the short Na -Dene chronology obsolete (Kari 2010:210). But Kari also insists t hat the Na-Dene languages have been present in North America for more than 10,000 years. This hypothesis of in-situ American diversification of Na-Dene great ly exacerbates this implausib ility in geographical terms. Joseph Greenberg included Haida in hi s 9000-year chronology for the development of Na-Dene (Greenberg et al. 1 986). This long chronology was adopted by archaeologists to support t he identification of Archaic Am erican microblade users as Na-Dene prior to 7000 years ago (Sutton 2008:84). The recognition that Yeniseian is closer to Na-Dene than either language family is to Haida challenges this model, as lexicostatistical divergenc e between Tlingit and Athapaskan may be less than 3500 years (Kaufman and Golla 2000:52). Even th is divergence process could have begun anywhere between interior Eurasia and Alaska If Na-Dene and/or Yeniseian peoples were highly mobile latecomers into one or both of their respective territories, then the geographical implausibility of the family is mitigated rather t han exacerbated. There is a strong possibility that the historical range of the Yenise ian languages was much greater in Eurasia during the past. The Historical Range of Yeniseian Speakers Although recent Ket have resided in central Siberia northward to the Arctic Circle, the six or seven extinct Yeniseian languages were spoken considerably to the south as far as the headwaters of the Yenisei and Ob ri vers, northwest of Lake Baikal, all within the last 300 years. The few hundred remaining Ket reside on the middle reaches of the river, but at least four of the extinct Southern Yeniseian languages (Kott, Assan, Arin, & 82


Pumpokol) were distributed considerably to the south along the upper reaches of the Yenisei as early as the mid eighteenth century; these headwat ers and tributaries immediately northwest of Lake Baikal are precisely where hundreds of americanoid Okunev burials were found, all from t he Bronze Age between 3500-4500 years old (Kozintsev et al. 1999:197). Furthermore, a gr eat swath of territory to the south and west of these lands (in modern Finno-Ugric, Samoyedic, Turkic and Mongolian territory) shows frequent or occasional Yeniseian toponyms and evidence of former widespread bilingualism between Yeniseian and Turkic. In the Lena Basin and the considerable territory to the west, recent incursions by Neo-Siberian pastoralists have obscured any such toponyms and/or linguistic substrates; the possible Yeniseian language history in this greater region is unclear (Kari and Po tter 2010:8-10). The southerly origin of Yeniseian languages is likely based on present data; the pre-Metal Age peoples of the region do not posess strongly americanoid physical traits. Furthermore, there is reason to suspect that a southern Yeniseian language may have been the elite language of t he Xiongnu Empire (a.k.a. t he eastern Huns) since the only surviving Xiongnu materials in Old Chinese transcription (consisting of one sentence and about two-dozen scattered words) show remarkable affinities to Yeniseian grammar and basic vocabulary (Pulleyblank 1962:239-265; Vovin 2000, 2003). At the very least, the Xiongnu likely included Yenise ian tribes in their confederacy along with Scythians, Turks and Mongols. The ancient range of Yeniseians as equestrian nomadic confederates of the Xi ongnu would have been enormous encompassing southern Siberia, Chinese Turkestan, Inner and Oute r Mongolia eastward nearly to Manchuria and the maritime provinces of the Russian Far Ea st (Figure 3-2). Gene rationally shifting 83


patterns of multilingualism are the norm in such large, multiethnic tribal confederacies. Michael Fortescue (with a hint of skepticism) summarizes conventional wisdom on the subject of Yeniseian origins: Yeniseians, today hunters of the taiga, were supposedly drawn into the Hun union some two thousand years ago and subsequently displaced into the Sayan region [the Altai Mountians], where they rema ined until the expansio n of the Mongolian empire pushed them still further north . peripheral to the spread zone highway of the Eurasian steppes, . dr awn into the fast traffic, later to be spun off again. . The linguistic links to Na-Dene . suggests rather that they have been present in Siberia for a very long time indeed. Certain Americanoid physical traits of the Yenise ian people have been remarked upon (Fortescue 1998:56, n.23). However, contrary to Forte scue, I suggest the conventional wisdom in this case is consistent with a preponderance of the physical traits, which shows southern Siberian americanoid phenotypes to be a product of Bronze Age admixture (Kozintsev et al. 1999), and consistent with demon strable processes of Cent ral Asian state formation and disintegration. Despite the romantic desire of tw entieth-century anthropologists (like Mircea Eliade) to find concrete ethnol ogical ties between contemporary Siberian natives and their Paleo-Asiatic forebears, it appears that many Native Siberians have concrete, historical roots in the co smopolitan Eurasian heartland, which was a population reservoir. It was t he shattered relicts of failed st ates and empires which flew cyclically into the unforgiving climate of North Asia: The devolution, or more accurately the disintegration, of Inner Asian states in this region is well documented. It is not always realized what follows from this, that many of the tribal peoples of Elia des time had been imperial rulers some centuries before. The M ongols are the most fa mous of these. There also are the examples of the Kh itans and Jurchens, the ruling elites of the Liao (tenth through twelfth cent uries) and Jin (twelfth and thirteenth centuries) empires respectively, whose descendants became the impoverished peripheral tribes of early twentieth-century ethnography (Humphrey 1994:195). 84


The southern origins of the Yeniseians may likewise be inferred on the basis of linguistic evidence, particularly loanwords into Yeniseian from southern languages which are no longer contiguous with them. Just as the physical anthropological evidence points to South Siberia as a mult iethnic melting pot, so too an admixture scenario may well explain many disparate features shared between Yeniseian and Sino-Tibetan languages, which could reflec t ancient and intensive borrowing (Behr 2003:176) as easily as pre-Neolithic Ti betan-Yeniseian comm on genealogical descent (Ruhlen 1998; Sedl ek 2008). Both phonologically and morphologically the contrast between Kettish and Kottish on the one hand and Chinese or Tibetan on the other coul d scarcely be greater. . There are a few striking word comparisons but these may well be explicable by early borrowings, especially if it should turn out that the Yenisseians and the Chinese were once contiguous (Pulleyblank 1962:243). Hundreds of apparent Sino-Tibetan cognates have also been identified in NaDene (Bengtson 1994; Sapir 1991; Shafer 1952). The most striking Sino-Dene typological parallelism identifie d by Sapir is the well developed tone system found in Athapaskan, Tibetan and Chinese (Sapir 1925); Yeniseian is also a tonal language (Comrie 2010:31). However these tone systems are not particularly ancient and have not been reconstructed in any of the rele vant protolanguages (Bengtson 1994:210). The genesis of the tone systems in at least some of these languages may have been contact-induced, indicating some degree of nongenetic mutual influences. Ethnological parallels between Athapaskans, Si berians, and East Central Asians also seem to favor strong late prehistoric ties between the continents (Mortensen 2006). Dene-Yeniseian languages should be viewed in this context. 85


Asian-Athapaskan Ethnological Links Chapter 2 outlined some features of Pr oto-Athapaskan society, using the method of ethnographic reconstruction (building upon Perry 1983, 1989). These areas include pervasive beliefs about eschatology (deat h and the afterlife), femaleness (female puberty rituals and childbirth practices), and social organization (matriliny and matrilocality). In this section, I will draw parallels in each of these areas to similar practices in Asia, found among Native Siber ians and Central Asians. Yeniseian data will be included whenever possibl e, although in Siberia (as in North America) culture traits are rarely neatly bounded by language fam ily. In addition to the above categories, I will address another one of Perr ys (1983) proto-Athapaskan cu ltural traits; specific forms of extrasomatic power, especially rele vant to transpacific links, with further relevance to the cultural importance of projectile weapons technology. Eschatology Siberian antecedents for Proto-Athapa skan eschatology are numerous and detailed. The multipart soul is a widespr ead feature of Siberi an shamanism from the Altai to the Chukchi Peninsula (with the component parts generally numbering between three and six). Particularly the Samoyedic N entsi of the lower Yenisei valley in North Siberia possessed a belief in a three-component soul where two of these components are virtually identical to the proto-Athapaskan shadow-soul and breath-soul (the third part being the intellect). Just as among mo st Athapaskans, the Nentsi shadow-soul is considered malign (capable of returning to molest the family) and the breath benign. Furthermore, Nentsi practiced strict name taboos and destroyed the possessions of the deceased (leaving them at t he gravesite) and burned the deceaseds house in a manner precisely analogous to common Athapaskan practices (Perry 1983:728). Among 86


Southern Athapaskans, the thorac ic organs (heart-lungs-liver) are considered to be just one super-organ, so the action of breath is a function of the working heart (Baldwin 1997:27). The Northeast Siberia n Yukaghir and neighboring Polar Yakut both have the same three-soul model as the Nentsi, with the intellect, shadow-soul, breath-soul/heartsoul. Like the Navajo and Apache, they concei ve of breath as the action of the working heart. Furthermore, these two neighboring Si berian groups both believe in human-tohuman reincarnation within family lin eages (Jochelson 1910:158-161). Similar reincarnation beliefs are found among many Athapaskans. Most Na-Dene (like some Eskimo groups) al so believed that at least a portion (usually the breath/wind) of the human soul was capable of r ebirth or reincarnation in the terrestrial sphere. This by itself is unremarkable, as rebirth/reincarnation eschatologies are common the world over including many indigenous groups and small-scale societies. The importance of Athapaskan, Northwest Coast and Eskimo reincarnation beliefs lies not in their mere existence, but rather in their pervasiveness and complexity in comparison with those of si milar small-scale societies; in both of these respects they show unexpectedly str ong similarity to the complex societies of Central Asia. Generalized animistic belie fs are widespread among indigenous peoples, but they are less commonly conjoined wi th reincarnation beliefs. Gananath Obeyesekere examines the reincarnation beliefs of several northwestern American Indian societies, including Carrier and Beaver Athapaskans, Tlingit, Haida and (Tsimshianic) Gitxsan. He observes T heir theories have an underlying ideological resemblance to South Asian religions . in the view that animals and even plants are 87


endowed with consciousness (Obeyesekere 2002: 38-39). They are thus distinct from the reincarnation beliefs of other small-scale societies: the only empirical cases similar to t he Buddhist among small-scale societies outside the Indic area are those of the Northwest Coast Indians [including Athapaskans] and Inuit, because the Igbo and other West African groups and the Trobriand Isanders have no notion of animal rebirths. . One cannot therefore totally discount the idea that the circumpolar distribution of these rebirth eschatologies might we ll have extended from Siberia down into Central Asia and then to the Indian subcontinent (Obeyesekere 2002:89). Obeyesekere does note one difference bet ween northwestern Native American reincarnation eschatology and t hat of the Buddhists Asia; w hen the great Indic religions emerged into history they had another key doctrine in addition to the doctrine of rebirth, namely that of karma, a system of ethical intentions t hat decide the natur e of rebirth (Obeyesekere 1994:xix; emphasis original). Obeyesekere maintains that Indias karmic eschatology was developed from a small-scale antecedent very similar to Athapaskan rebirth eschatology. But Buddhism and its ilk developed this basic template into a hierarchical cosmology (with animals r anked below humans) and an ultimate concern with cosmic justice, typica l of the organized religions of many complex stratified societies. In contrast, the notion that rebirth into a lower station is punishment for actions in life, is quite alien to members of small-scale animistic societies (for whom different orders of existence ar e not rigidly stratified). Religious laws governing punishment and reward after death linked to ones behavior in life are a general feature of hi ghly stratified statelevel societies. Obeseyekeres model thus distinguishes Indic eschatology from simple rebirth in most small-scale societies. However, there are a num ber of exceptions to this tendency which seem to indicate the possible influence of the organized religions of Eurasian 88


states upon the circumpolar shamanism comp lex. Complex eschatologies from the Eurasian hearth may have spread gradually to Siberia and thence to North America. Buddhist missionaries among the Xiongnu were active circa 2000 years ago, and the one surviving fragment of Xiongnu text (with strong Yeniseain affinities) was transcribed by Futo Cheng, a Buddhist monk (Vovin 200 0:95). Buddhist elements of the north Asian shamanism complex have long been rec ognized; one theory maintains that the Evenki word shaman itself is a Buddhist idiom, der ived from the Sanskrit word for monk ramana (Mironov and Shirokogoroff 1924). This view has been lately substantiated by textual evid ence showing that the word shaman could mean either medicine man or monk in ancient Turkic languages, depending on the context (Zieme 2008). Ian Stevenson argues that karma-like concepts and specific taboos among the (Dene-Yeniseian) Tlingit are probably indicative of Buddhist influence: In addition to the belief in reincarnation itself and in a concept somewhat similar to that of karma linking one life with another, the Tlingit have two other significant ideas with regard to reincarnation. First, they believe that children who remember their past liv es are fated to die young and they endeavor to discourage a child who claims to remember a previous life from doing so. An identical belief exists in India, where families of such children frequently make strenuous efforts to suppress the apparent memories of a previous life told by a child. Secondly, t he Tlingits also believe in rebirth as contrasted with reincarnation. According to the concept of rebirth, the old personality gives rise to the new as a candle burning low may light a new candle and so continue the series. In reincarnation, on the other hand, the same personality continues, although c hanged by the circumstances of the new life. Reincarnation as thus defined is a concept of Hinduism and rebirth a concept of Buddhism. . The clos e similarities between the ideas on reincarnation among the Tlingits and Buddhists also suggest that the ancestors of the Tlingits imported ra ther than invented their ideas on reincarnation (Stevenson 1974:222). That a concept similar to karma survived wit hin the socially stratified societies of the Northwest Coast is fathomable; social st ratification has existed in the region since ca. 400 BCE. But in the more egalitarian so cieties of the interior Subarctic, such 89


concepts would be unlikely to endure. Nonet heless, reincarnation eschatologies are almost universal among Nort hern Athapaskans, Eyak and Tlin git, and a number of them appear to fit Obeyesekeres karmic pattern atypical for simple societies. Widespread cognate Athapaskan terms for rein carnation (like Gwichin natli and Slavey ndaadlinha ) suggest the existence of a Proto-Athapaskan word for the phenomenon (Mills 1994:20-22). Reincarnation beliefs invo lving both humans and animals are also reported for Navajo, Apache and Pacific Co ast Athapaskans (Bourke 1892:470; Haile 1943:87-88; Mills 1994:25; Powers 1877:110). Obeyesekere (2002:89) notes tw o different types of animal reincarnation. Parallel reincarnation is the more common type in No rth America, where reincarnation occurs entirely within species boundaries. Buddhist soci eties, by contrast, all believe in the second type, cross-species reincarnation (met empsychosis), whereby different orders of existence are placed in a hierarchy and the upward or downward mobility of an individual consciousness is determined by karma. While the karmic type of rebirth is extremely rare among foragers, it does appear occasiona lly among Na-Dene and Eskimos. This is understandabl e, as [c]oncepts of reinca rnation in the New World have doubtless continued to be influenced by . Si berian shamanism (its elf affected by Buddhist belief and practice) (Mills 1994:20). Reincarnation among the Athapaskan Kaska is governed by principles of meritbased causality, as among the Buddhists. Kaska share with the Buddhists a coolness toward the prospect of reincar nation to correct lifes errors; the truly virtuous are freed from the corporeal body in reward fo r an honorable life (H onigmann 1954:136-137). Although Navajo eschatology is relatively unconcerned with rewa rd and punishment in 90


the hereafter, the length of ones life may be affected by virtuous behavior in a previous life. The wind-soul is impervious to change, but the fate of this body is linked to the actions of the wind-soul in past bodies. But, if the person has liv ed an improper life, the same wind may be sent into another human body, which will then only ha ve a short life. If such person had lived a proper life, as indicated by death of old age, that wind soul may again be dispatched into anot her human body, which also will die of old age (Haile 1943:87-88). Among the Upper Tanana, The dead may be r eborn in the form of animals. When a hunter meets an animal he is unable to kill, he regards this as evidence of the transmigration of a human spirit into anima l form (McKennan 1959:160). The Gwichin likewise may be reincarnated into a Caribou or another game animal. They consider reincarnation into another human body to be highly auspicious, but to be reborn as a bear is a lamentable tragedy (Petitot 1878: 26). The virtuous among the Mattole Athapaskans in California were said to be rewarded with rebirth in a paradise across the Pacific Ocean, while the bad Mattole were r eportedly reincarnated as grizzly bears as punishment for their misdeeds, to be hunted and despised by men (Powers 1877:110). A similar belief system is recorded for Apac hes, with snakes being the most despised animal and rebirth as a rattlesnake being the just punishment for evil humans (Bancroft 1875:527). The reincarnation of evil humans in to lemmings is recor ded for the Caribou Inuit, interior-adapted near-neighbors of t he Chipewyan on the western shore of Hudsons Bay. The presence of complex Caribou Inuit karmic eschatology belies Rasmussens regard for these folk as po ssessing religious ideas among the most primitive I have found (Igjug rjuk and Rasmussen 2001:83). Similar statements about the primitiveness of Athapaskan religious beliefs abound in t he literature, and Athapaskans likewise defy this st ereotype upon closer inspection. 91


Matrilineal-Matrilocal Social Organization Ethnographic Ket primarily reckon their ki nship patrilineally, but as the sole surviving Yeniseian group, might not adequately represent the language family as a whole. Indeed, the Ket are matriloca l (like most Athapaskans) and folklore and ethnohistory combine to suggest that they too may have been matrilineal in the past (Nikolaev 1985:80-89; Vajda 2001 :205). This would be cons istent with archaeological views of the South Siberian Ne olithic, where DNA lineages cl ustered in elite burials in the Lake Baikal region are consistent with matrilineal soci al structure (Mooder et al. 2005:631). Folklore of the Tuva and Khakas s people of South Siberia also suggest formerly prevalent matriliny (Dalkesen 2010). Khakass are a Turkic ethnic conglomerate who in recent centuries entirel y assimilated both the Arin and Yarin bands of Southern Yeniseians; Khakass cultural fu sion includes a quotient of Yeniseians who now speak a Turkic language (Wixman 1984:101) Evidence for pre-modern Yeniseian social structure is relatively scant, but what does exist is consistent with the view that the pattern among Na-Dene speakers coul d be connected to post-Neolithic South Siberia. Menstruation and Childbirth Very similar to California Athapaskans among the Northern Evenki of Siberia, women are believed to be perennially threatening polluters of men (Montgomery 1974:140; see Shirokogoroff 1933). These ideas are also prevalent among the Samoyeds and Ob Ugrians in Northern and We stern Siberia (both are Uralic-speaking neighbors of the Ket). They share s pecific menstrual t aboos with Athapaskans, including prohibitions against handling m ens hunting equipment and crossing paths with hunters (Abel 1993:21; Ba lzer 1981:851; Perry 1983:728). 92


The Ket themselves have no formal puberty rituals, but do have similar rules forbidding the presence of men during ch ildbirth. As among most Athapaskans, dangerous physical complications are feared whenever a woman is permitted to see a mans face before childbirth. The Ket also retain, long into old age, the umbilical cords of children preserved as amulets in ornately decorated pouches. Kept on the person; these amulets are believed to protect an i ndividual throughout his/her entire life (Lee 1967:44). These same customs are also practiced by Athapaskans in all three geographic blocs. The Chipewyan have the i dentical custom of preserving umbilical cords in ornately decorated pouches for life, as do the vast majority of Pacific Coast Athapaskans (Abel 1993:20; Barnett 1937:178; Curtis 1928:25-26; Driver 1939:349; Essene 1942:32). Similar or identical cust oms are found in the Plateau, Northern and Central Plains, among such groups as the Thompson Salish, Nez Perce, Klikitat, Tenino, Umatilla, Blackfoot, Cheyenne, Lak ota, Kiowa, Arapaho, and eastern Apache (Howard 1907:17; Ray 1942:199; Sutton 2008:119; Teit 1900:304-305; Wissler 1920:92). Tsuutina, Western Apache, Ute and Kwakiutl groups also preserve the umbilical cord in decorated pouches, but only for a finite time, eventually ritually disposing of it during childhood (Curtis 1928:105; Drucker 1950:206; Gifford 1940:158). Figure 3-3 shows an Apache umbilical pouch worn by an adult. The Ket dispose of the afterbirth by hanging it in a tree (Lee 1967:44). This is the same practice as Northern Athapaskan Carrier and Chilcotin; disposal of the placenta in a tree is also a universal practice among Oregon Athapaskans and southwestern Apaches (Gifford 1940:62; Jenness 1943:30; Ray 1942:195). Non-Athapaskans who follow this custom include Plains Cree, some Oregon Coast tribes, Inland Salish and 93


their confederates, and (rarely) Coast Salis h (Barnett 1937:138; 1938:253; Mandelbaum 1940:241; Teit 1900:304). Transpersonal Extrasomatic Po wer of Objects and Animals Perry notes the pervasiveness in all three Athapaskan blocs of the attribution of extrasomatic power to beings and objects (P erry 1983:723). This is an example of a common tendency among Native Americans in general which nonetheless has uniquely Athapaskan manifestations of particular intens ity; this power although it may resemble aspects of some non-Athapaskan cultur es, nonetheless is ubiquitous among Athapaskan groups with striking conformity (Perry 1983:723). This striking conformity encompasses the occurrence of essentially identical concepts of bear sickness (a dangerous spiritual affliction brought on through exposure to bears and bear detritus) among the Denaina and Western Apache, two groups found at near extremes within the geographic range of Athapaskan speakers, as well as very different ecological adaptations (Perry 1983:719). Vaguely si milar animistic belief systems are found throughout Siberia, for example among Evenki sham anists, but it is di fficult to ascribe historical significance to these more general parallels (Heyne 1999:377, 384). In the Apachean case, such extrasomatic pow er is transpersonal in the sense that the actual bear (or snake, or lightning etc.) is not the source of the power but rather the physical manifestation of a pr imordial essence. To prevent the disease (bear sickness or snake sickness or lightning sickness, et c.) requires homeopathic prophylactics. In the same way that smallpox vaccine includes a controlled form of smallpox virus, so the mastery of extrasomatic power is used as a safeguard against the very source from which it is derived. To ward off lightning, one needs lightning power; to kill bear, bear power; to cure snake sickness, snake pow er (Basso 1966:150). This power is not 94


limited to animate beings, but is transferrable to intimate objects; lightning-struck trees can be just as dangerous as lightning itself (Perry 1983:723). These same lightningstruck trees are believed to spontaneously produce bifacial lithic artifacts. Such items are viewed by Apacheans as talismans from t he sky, and antidotes to lightning power. According to one White Mountain Apache informant, Helen Crocker: And they say that when lightning is r eal hard, you know, just after one another, these rocks come from there. They land in the bark [of lightningstruck trees] or on the ground. They take those. It's used in a prayer, and it's called "from the lightning," nbi'bibish. Nbi is the lightning and bibish is his knife. It's the lightning's knife. It's one [of the] sacred things Indians use. Kept those, and medicine men used it some more in the prayers (Kessel 2005:67-68). Apachean conceptual associations between lightning and projectile points are strong. Arrowheads or blades (putatively of celestial orig in) protect against lightning and, in a complementary manner, wooden amul ets carved from lightning-struck trees protect against arrows and bulle ts (Bourke 1892:593). Apachean beesh means knife, but it also means metal or flint, that is the primary substance us ed to make the knife (Greenfeld 1973:100). It is cognate with Northern Athapaskan words, such as Carrier ps, referring to a fine-grained augite-porphyrite preferred for flintknapping, or Denaina vashla, chavash womans knife (Golla 1998:2; Mori ce 1894:53). It is also cognate with Denaina delvashi, lightning (Kari 2007:149-150). Use of grooved arrow shafts, symbolic of lightning, is found in a continuum of cultures beginning with the Sekani Athapaskans in the western S ubarctic, extending through the Great Plains and ending with the eastern Apaches in the Southw est (Mason 1907:98; Morice 1894:55-56). Eursian Parallels for Nbi'bibish, the Lightnings Knife Extreme fear of destructive spiritual power of lightning (paralleled among the Apacheans) is found throughout the Tibetan-Mongolian religious world system, and is 95


attributable to the very real danger lightning poses to nomadic tent villages in wide-open spaces of the Altai-Sayan uplands and Tibetan Plateau (likewise among the mountainand plains-dwelling Apacheans). This woul d be unremarkable, but the complex of symbolic associations with li ghtning in the shamanic cosmos are not always attributable to similar environmental conditions. Unlik e the general shamanistic concern with the extrasomatic power of dangerous animals li ke bears, North Eurasian-North American practices and beliefs related to lightning and the spontaneous generat ion of bifacial tools is a striking parallelism at the level of minor detail, and worthy of consideration. Lightning-struck tree-wood is used commonly by Siberian shamans for divination, purification, and healing of psychic diseases ; such is reported in northeast-central Siberia among the Evenki of Sakha, in the northeast among the Turkic Yakut, and the Turkic Tuvans of south-central Siberia (Bulatova 1997:240; Czaplicka 1914:195; Stevens 2004:198). Altaian peoples will not eat the meat of an animal killed by lightning, and have special funer ary protocols for people killed by lightning minimize contact with the deceased (Czaplicka 1914:162). Tibetan folk religion (in southern Siberia, Mongolia, Tibet and Nepal) includes the concept of the thog lcags, (pronounced togchag) sky iron and closely related thog rdo thunderbolt stone (Nepalese vajra dunga ), equivalent to the White Mountain Apache nbi'bibish. Tibetan thog, sky, roof, high, thunderbolt, lightning, is equivalent to Apache nbi and lcags metal, flint, is equivalent to Apache bibish ; in compounds lcags can also mean lightning or thunderbolt (Jschke 1881:148, 237, 286). Like tradition al Apaches, traditional Tibetans believe that prehistoric stone blades (or meteoric metal and ancient bronze blades), adzes and arrowheads are protective amulets spontaneously deposited by lightning into trees and 96


soils. Tibetans believe that at certain times when lightning strikes the ground it generates a spark which reacts with we t earth to produce thogchags (Bellezza 1998:44). An archaeologist reported acquiring a Neolithic hand axe from a Tibetan peasant in Darjeeling, India, the finder having seen a tree struck by lightning went to look for the thunderbolt and found it . in the ground amongst it s roots (Walsh 1904:21). The historical context for thog lcags and thog rdo traditions is further illustrated by John Bellezza: A belief prevalent among the nomads of northern Tibet is that thogchags can prevent a person from being struck by lightning, a palpable danger in the wide open spaces of th is part of the country. . The earliest historical links seem to be with Bronze Age Si no-Tibetan cultures and the Central Asian Iron Age Saka-Scythians . arrows and other metallic objects [found] at graves and megaliths in nor thern Tibet. . Tibetan Neolithic fetishes . are credited with havi ng been self-formed and have talismanic value. For example, the direct pr ecursor of dart-shaped thogchags are probably dart-shaped Neolithic stone amul ets . Neolithic arrowheads, spearheads, rings and even Paleolithic ch oppers . worn or enshrined for their purported super natural properties (Bellezza 1998:44-47). Altaic (Turkic-Mongolian) mythology suggests that the first sword was forged from indestructible meteoric sky iron by the cu lture hero Timur, whose name came to mean iron. The origination of this myth am ong Yeniseian-Xiongnu (Huns) is likely, as the sparse Xiongnu literary sources include the ear liest references to the Altaic supreme deity Tengri, sky, which appears to be a Yeniseian etymon; the Yeniseian rootword *t gVr, high, with the Altaic possessive suffix i (Vovin 2003:389, 393). The metallurgical Saka-Scythian cultures of Cent ral Asia contributed much to the mythology of the Altai-Sayan region and Northern Eura sia as a whole. Swords made from meteoric iron were called thunderbolt swords or lightning blades, and were believed to bestow divine power upon their wielder; this folk belief persisted in Central Asia from Buddhist times well into the Is lamic period (Rosedale 1891:13). 97


Sequences of cognate mythological motifs involving sword-wield ing culture-heroes are cataloged in North Eurasian cultures as far apart as Japan and the British Isles; these remarkable borrowings may suggest a common origin of the tales among the blacksmithing cultures of t he Russian steppes (Littleton 1983, 1995). Likewise, the folk belief in the power of amulets or talismans made from archaeological arrowheads, axes, and knives, alleged to have been produced s pontaneously by lightning and therefore protective against the same, is found in Eu rasian cultures as far apart as Ireland, England, Tibet, China and Japan. All of these cultures have equivalent terms to Apache nbi'bibish or English thunderbolt, referred to a Stone Age arrowhead or adze worn as a talisman to protect from li ghtning (Ettlinger 1939; Reisc hauer 1940; Skeat 1912). This is strongly suggestive of a common origin for th is practice in Eurasia, either prior to dispersion of East and West Eurasians ci rca 20,000 years ago or alternatively during the Metal Age less than 4000 years ago, among the highly mobile cultures of the Eurasian heartland, like the Scythians. The term Metal Age is often used by Central Asian archaeologists, because the distinctio n between Bronze Age and Iron Age in the region is imprecise (Bellezza 1998:46). Metal Age origins seem plausible, as St one Age peoples would be unlikely to posit supernatural origins for the utilitarian t ools they made themselvesthe mythologized context for stone tools appears commonly among peoples for whom the actual production of such tools has become virtually obsolete and therefore mysterious. The Athapaskan case would seem to be an exception to this rule. But further inspection shows that the demonstrably Athapaskan peri od in western Subarctic prehistory (midCommon Era) is characterized by rapid repl acement of atlatl technology by archery 98


technology, increasing reliance on bone, ivor y and copper tools, and obsolescence of stone dart points (Clark 1991:102 -103). Late proto-Athapask an cultures are closely identified with the spread of the copper i ndustry in the north (Clark 2001:175). Athapaskans are notoriously poor flintk nappers whose stone projectile points are characteristically unrefined; finer and more distinctive specimens of lithic technology found in Northern and Souther n Athapaskan contexts are often presupposed to have been scavenged from non-At hapaskan sources (Ferg a nd Kessel 1987:50-52; Morice 1894:54). For Apaches, it is just these finer specimens of mysterious prehistoric lithic workmanship that are regarded as spontaneo usly generated by lightning. This is analogous to the Eurasian Meta l Age parallels. Oral trad itions of Athapaskan copper use suggest the Athapaskan language expa nsion involved the spread of copperworking technology (Moodie et al. 1992). For the Northwest Coast peoples lacking direct access to native metal sources, the situation was the reverse, because the farther from the supply of copper . the more mysterious the origin, as is evi dent from the myths (Emmons and De Laguna 1991:180). Athapaskan-made native copper arro wheads were kept as charms by Tlingit chiefs, later replaced by smelt ed sheet copper shield-amulets when European metals glutted the market (Keithahn 1964:7778). Tlingit mythologi zed the origin of copper and were originally dependent on neighboring Athapa skan-Eyak middlemen to supply the metal; to the ex tent that heat-treatment wa s employed by Tlingit copper workers, it did not incl ude the advanced tempering proc ess of Athapaskan or Eyak coppersmiths, and heat treatment of any sort was limited to groups like Yakutat who 99

PAGE 100

invaded Athapaskan-Eyak territory in the north of Tlingit range, assi milating local people and thus gaining knowledge of copperwork ing (Emmons and De Laguna 1991:179). A Siberian Origin for At hapaskan Copper Metallurgy? The skill of interior Athapaskan coppersmiths is significantly greater than that of Northwest Coastal peoples and Eskimos, cont rary to stereotypes. Frederica de Laguna (1947:182) first noted the striking similari ty between Athapaskan-Eyak spiral-hilted native copper daggers and the sp iral-hilted bronze daggers of the Ordos region of Inner Mongolia and North China; she postulated that some historical relationship between the cultures involved was quite likely. Larger, somewhat cruder copies of these Athapaskan blades were made by coastal Tlingit usi ng only cold hammering (in contrast to Athapaskan hot forging). The higher quality of Athapaskan-made metal daggers was necessitated by their key role in diversified subsistence strategies. Tlingit daggers, by contrast, were luxury items that evolved as part of the affluenc e of Northwest Coast cultures and poorly adapted to any purpose mo re worthwhile than fighting (Witthoft and Eyman 1969:22). Regarding Athapaskan metalwork, Witthoft and Eyman continue: Dene daggers in copper and steel, show both heat treatment and stress hardening as methods for controlling strength and edge-hardness of tools. Dene copper arrowtips, awls, and ot her tools show an equally expert technology. . [Dene] have been called one of the most primitive Indian groups in the Americas. Their use of metals was nevertheless the most sophisticated that we have seen in Nort h America. . [T]he double-spiral motif of the Dene . has no exact count erpart in North American art except in Tlingit. It has many parallels in the art of north China and Siberia. . Dene . considered peoples with simplistic technologies and a slight cultural heritage, appear in a new light as the carriers of a peculiar metalage technology. . A single copper knife in any Dene collection raises problems inconsistent with the ster eotyped phrases of the literature (Witthoft and Eyman 1969:21-23). Figure 3-4 shows an example of a mid-ni neteenth-century Tahltan (or Gwichin) Athapaskan dagger I exami ned during my fieldwork (a forged steel copy of a copper 100

PAGE 101

prototype), in comparison to a Krasnoyarsk-Tagar Siberi an forged iron dagger from the Upper Yenisei river region, mi d-first millennium BCE, like the contemporaneous bronze daggers from the same region. Tagar te rritory was home to Southern Yeniseian speakers in modern times, and R.V. Nikolae v argues that the Krasnoyarsk segment of the Tagar culture represents a northward in cursion of Yenisean speakers into the Yenisei valley from the Sayan region during the first millennium BCE (Nikolaev 1989:7072; Vajda 2001:207). If this view is correct then the two pictured weapons come from linguistically affiliated peoples. The double-sp iraled pommel of both weapons is visible (taking the form of mirrored rams heads on the Siberian dagger), but the comparable axial ridge on the American blade is obscur ed by the dark patina. The two pommels and two blades are of nearly identical widths (about two inches and one inch respectively), but this American dagger is unusually short at seven inches compared to the Siberian weapons twelve inches; At hapaskan native copper daggers as long as sixteen inches were examined by Witt hoft and Eyman (1969:21) Copper daggers are highly uniform with a broad distribution in pre-Columbian northwest Canada and Alaska. Figure 3-5 shows an asso rtment of Eurasian and Athapaskan forged metal arrowheads, diamond-shaped and hex agonal, featuring solid t angs for hafting or tubular projecting ones for insertion of the arrow s haft. Although less striki ng than the spiraled daggers, these objects show an equal mast ery of heat-treatment and edge hardening techniques (Franklin et al. 1981). No heat treatment of stone tools is known in the Western Subarctic, which would be an expec ted antecedent if Athapaskan metallurgical skill with native copper was developed locally. Its absence is suggestive of a Siberian source for copper annealing technology (Clark 1991:116). 101

PAGE 102

Molecular Genetics and the Dene-Yeniseian Family The remainder of Chapter 3 is devoted to exploring the impact of the DeneYeniseian language family upon contemporary molecular anthropology. The burgeoning field of DNA-based population genetics is ideally positioned to address questions raised by the new linguistic consen sus for Dene-Yeniseian geneaological ties. But the state of the discipli ne is rather unreceptive to l ong-distance language relations, because most experts favor onl y one major migration from As ia to North America, a scenario incompatible with the notion that Na-Dene lang uages came as a part of subsequent migration(s). The followin g discussion will suggest alternative interpretations of the large body of mo lecular genetic data, which would be more compatible with the hypothes is of Dene-Yeniseian ge neaological relatedness. The last 20 years have been the formative period of molecular anthropology, and the growth of the field as a whole has been spurred by the steady increase in the efficiency, speed, volume and sophistication of DNA analysis. Earlier generations of archaeologists, linguists and physical anthropolog ists all more or less agreed that North America had been colonized by several diffe rent Eurasian founding populations at several different times. Today it is rare to find a molecular anthropologist who favors more than two distinct migration events, and a majority of res earchers are enamored with the single-origin hypothesis, which postulates just one founding group ancestral to all Native Americans, lumping Eskimos and Na-Dene into the same Beringian cohort as the rest of indigenous Am ericans (Bonatto and Salzano 1997; Fagundes et al. 2008; Goebel et al. 2008; Kolman et al. 1996; Merr iwether et al. 1995; Mulligan et al. 2004; Rubicz et al. 2002; Stone and Stoneking 1998; Tamm et al. 2007; Tarazona-Santos and Santos 2002; Zegura et al. 2004). 102

PAGE 103

The main problem with the single origin hypo thesis is that it is primarily based on surveys of uniparentally inherited DNA. Fa r more often this has been maternally inherited mitochondrial DNA (mtDNA), but more recently, the non-recombinant portion of the Y-chromosome (NRY) has achieved prominence as well. Uniparental mtDNA and NRY data do have the advantage of potentia lly revealing sex-specific migration history which is complementary to ethnologys emphasis on marriage customs and social organization based on geographic mari tal locality (Mooder et al. 2005; Wilkins 2006). But the drawback of uniparental data is that they are limited in their capacity to reveal complex demographic histories by a smaller effective sample size than autosomal DNA (Battilana et al. 2007:64-65), and are also subject to more pronounced loss of diversity due to genetic drift. A nd because just mtDNA has been most frequently used, it has potentially di sregarded male-skewed founder effects relevant to Athapaskan migrations (Gor don 2012; Ives 2010; Wilson 2008) Based on autosomal non-coding regions with low to moderate reco mbination rates, contemporary Native Americans appear to be a representative subset of Eurasian genetic diversity (Battilana et al. 2007). Native American and Eurasian popul ations are profoundly similar, and it is difficult to sort out the signi ficance of this similarity in terms of demographic history. Because all known genetic diversity of both Eurasia and the Americas derives from the same Out-of-Afric a population, a rough maximum time-depth for the peopling of the Americas may be infe rred, but it is ultimately impossible to determine when the latest pre-Columbian migrati ons occurred, because genetic diversity within and between Old World and New World populations is nearly identical. Whether or not molecular divergence between related American lineages may predate their arrival on American 103

PAGE 104

soil is impossible to determine, and thus genetic evidence is expected to provide maximum age estimates for the peopling of t he Americas (Zegura et al. 2004:172; emphasis original). Human Leukocyte Antigen Markers In addition to the sex-specific markers, several autosomal/ bi allelic markers are informative about Athapaskan/Na-Dene demographi c history. HLA (human leukocyte antigen) evidence is one in area in particula r where the single-origin hypothesis has not been endorsed by a majority of researc hers. The HLA syst em is the major histocompatibility complex, a super-locus containing many immune-system genes, spread throughout the length of chromosome si x. Substantial genetic support for NaDene ties to Asia has come from HLA st udies. For example, HLA (DRB1 and DQB1) markers clearly distinguish between E skimos, Athapaskans, and other American Indians, and these markers can be traced to different Old World source populations (Uinuk-Ool et al. 2003). Specifically, uni que HLA molecular markers are shared by Athapaskans and Paleo-Asiatic populations of Siberia, and this has been cited as possible support of the Dene-Ye niseian language hypothesis. [T]he HLA-DRB1 and DQB1 frequencies of [Carrier and Sekani] are most similar not to those of any of the other groups of native Americans tested, but to those of the Nivkh on Sakhalin Island and in the Lower Amur region in southeastern Siberia. This grou ping suggests that Nivkh and native Americans speaking Athabascan languag es are derived from a common ancestral population, while other nativ e American groups trace their origin to other ancestral populations (Uinuk-Ool et al. 2003:242). Nivkh, like Ket, is commonly classified as a Paleo-Asiatic language. Nivkh have been residents of the Low er Amur River (Far Eastern Si beria) since the Neolithic, with an insular branch moving to Sakhalin Isl and in the last 1000-2000 years (Black 1973; Wixman 1984:145-146). Nivkh ethnic ties to Northwestern North America have been 104

PAGE 105

previously suggested on the basis of linguistic typology (Sternberg 1904). Athapaskan ties to the Lower Amur region are also s uggested by NRY work (Lell et al. 2002), while most other genetic studies c onnect Athapaskans (and Native Americans in general) to East Central Asia and southern Siberia. Th is need not be an either-or proposition; a middle-way scenario would envision the Lower Amur as a lay over on the journey from Central Asia to Alaska, or as the easte rn front of a vast S outh-Siberian population system, including all relev ant populations (Wilson 20 08:272-273). Recent highresolution mitochondrial DNA work also indi cates that the mid-lower Amur and the Sayan/Altai uplands of South Siberia represent two distinct segments of the ancestral founding gene pool for Native Americans, at odds with the interpretation of limited founding mtDNA lineages populating the Americas as a single migration (Volodko et al. 2008:1084). Given that Yeniseian and Ni vkh have no demonstrable li nguistic relationship and are not in direct communion, significant adm ixture between migrant Central Asians and local Amur fishing peoples (among Na-Dene ancestors) could help explain these HLA data. Recent work has confirmed ear lier studies connecting Athapaskans and Nivkh (and their Siberian neighbors) through high resolution HLA-DRB1 and DQB1 analysis, and furthermore has found evidence of admixture with transpacif ic colonizers; different movements of people in either direction in different times are supported by the Athabaskan population admixture with As ian-Pacific population[s] and with Amerindians (Arnaiz-Villena et al. 2010:103). [T]here is a clear genetic HLA re latedness between isolated populations close to Beringia: Eskimos, Udegeys, Nivkhs (North East coast of Siberia) and Koryaks and Chukchi . Athabaskan, Alaskan Eskimos (Yupik) and Tlingit. These results 105

PAGE 106

. . suggest that admixture occurred between extreme North East Siberian groups and North American Na-Dene (inc luding Tlingit) and Eskimo (Yupik) people (Arnaiz-Villena et al. 2010:105). Arnaiz-Villena and colleagues favor a complex model of HLA admixture involving transpacific migrants merging with preexisti ng populations. They nonetheless believe all these movements took place more t han 12,000 years ago (a long-chronology model), because the destruction caused by post-Columbian epidemics is known to have affected Eskimos, Athapaskans, and other American Indians more or less equally, suggesting that none of them had any particu lar immunity to Old World diseases provided by their HLA-based immune system. Howe ver, this view fails to recognize that the horror of Eurasian diseases was exacerbated by only a slight histo-compatibility disadvantage, and not one created by millennia of population isolation. Virgin soil epidemics (like the bubonic pl ague) have depopulated major portions of Eurasia itself in the last 1500 years, despite the fact that the infected populations were already previously exposed to various other diseases through sustained long-term contact with the infectors. It is not the fact of recent long-distance migration which causes novel pandemics to flourish, but the scale of such migration, in conjunction with a variety of sociocultural-historical factor s (like lack of quarantine, multifamily residence patterns etc.) and especially the sheer number and quantity of diseases appearing simultaneously, and the demographic imbalance between the infectors and the infected populations. There is no evidence that Native American immune response was substantially different that any otherwise healthy low-den sity population would have been under the circumstances, and the Europeans died of their own diseases at high rates too (Crosby 1976). Western Europeans simply had an inexh austible population re servoir to replace 106

PAGE 107

their own numerous casualties, and a modest disease resistance provided by centuries of residence in urban cess pools; this pattern of less populous cultures succumbing to infectious disease . endemic in larger cities, . was a common theme throughout early Eurasian history (Agger and Maschner 2009: 321). Factoring in the more effective nursing and quarantine practice s developed over centuries in toxic metropoleis, one realizes that Eurasian germ warfare was as effective as it was for reasons well beyond the HLA system. Growing evidence for the moderate presence of se veral putatively Old World pandemic diseases in pre-Columbian Am erican contexts bears this out, as Martin and Goodman point out: Osteologic data demonstrate that nativ e groups were most definitely not living in a pristine, disease-free envir onment before contact. . Different populations were affected at different times and suffered varying rates of mortality. Diseases such as treponem iasis and tuberculosis were already present in the New World, along with dis eases such as tularemia, giardia, rabies, amebic dysentery, hepatitis, herpes, pertussis, and poliomyelitis, although the prevalence of almost a ll of these was probably low in any given group (Martin and Goodman 2002:67-68). Autosomal Recessive Diseases Also relevant to Athapaskan culture history is the number of Athapaskan-specific autosomal genetic diseases stemming fr om a severe population bottleneck (or bottlenecks) during or near pr oto-Athapaskan timeframes. The frequency of otherwise rare autosomal recessive diseases among Athapaskan speakers is a strong indication of a relatively recent and rapid population expansion from an extr emely small founding group. Some of these di sorders are found exclusivel y in Northern and Southern Athapaskans; e.g. HOXA1 deficiency (At hapaskan Brainstem Dysgenesis), Navajo poikiloderma, and Athapaskan Severe Combi ned Immunodeficiency. Several other such diseases are exclusive to the Navajo. Most significant are those genetic diseases 107

PAGE 108

exclusive to Athapaskans, their Yupik Eski mo neighbors, and Siberians. A single mutation underlying metachromatic leukodystrophy has been found among Navajo and Yupik Eskimos, and these particular Yupiks have intermarried extensively with Yukon drainage Athapaskan groups. Intermarried groups of Alaskan Yupik and Deg Hitan Athapaskans, along with Southwestern Navajo and Siberian Yakuts all have a rare form of methemoglobinemia due to diaphorase defi ciency, and a specific mutation has been identified in the Siber ian population. Until such time as the Navajo mutation is described, we will not know if this is a ra re mutation shared by descent, but it seems likely (Erickson 2009: 2604). The Yakuts are Old Uyghur-Turkic (Alt aic) speakers, commonly thought to have originated among the Iron Age hor se breeders of the Altai/Sa yan region and Southwest Siberia, before having been forced to migrat e north and east along the Lena River in the thirteenth century in the wake of Mongolian expansion (Wixm an 1984:219-220). Yeniseian languages were also forced northward along a major Siberian river valley (in this case, the Yenisei) during the same tumultuous period in history (Fortescue 1998:56)both Yeniseians and Yakuts are allged to have moved from southwest Siberia to central and northern Siberia as a result of the Mongol aggression. Yakut historical range is several hundred miles fr om the nearest Ket bands to their west. Twentieth-century Yakut speakers occupied mo st of the contiguo us territory between the Lower Tunguska River (a tributary of the Yenisei) and the Kolyma River in the northeast, where their neighbor s are the americanoid Chukchi (Figure 3-6). Like the Athapaskans, Yakut were subject to severe genetic bottleneck(s) in the Common Era, and like the Southern Athapaskans, they are a very conservative 108

PAGE 109

language family highly successful at assim ilating other indigenous peoples and rapidly expanding their grogr aphical range. Yakut ethnogenesis wa s very complex; their DNA appears heavily admixed between long-distanc e ECA and indigenous northern Siberian sources within the last 1000 year s, and they also appear to have had a very small male and female founding group (Khitrinskaya et al. 2003; Pakendorf et al. 2002; Pakendorf et al. 2006; Pakendorf et al. 2003; Ricaut et al. 2006; Tarskaia et al. 2002; Zlojutro et al. 2009). Yakut DNA diversity includes all four major Native American mitochondrial founding lineages (A, B, C & D) and two Y-chromosome haplogroups (C3 and R1) common to Northern Athapaskans (Kharkov et al. 2008; Puzyrev et al. 2003). Common factors stemming from a similar genetic bottl eneck in similar timeframes (and maybe ultimately stemming from common geographic origins near the Yeniseian urheimat) may provide the context for understanding the occurance of a methemoglobinemia due to diaphorase deficiency among Athapaskans and Yakuts. Albumin Naskapi Another autosomal marker characte ristic of Athapaskan speakers is Albumin Naskapi (AL *Naskapi), an A G transition in exon 9 of the gene (on chromosome 4) which is responsible for human plasma protein. The AL* Naskapi variant is widespread among Athapaskans, northern Algonkians, and in also found in Northern India and Turkey. The languages spoken by the se veral Algonquian groups in which AL*Naskapi has been found derive from a protolanguage estimated to be about twice as old as the protoAthapaskan language . and protoAthapaskan was still spoken after the tribal groups speaking Algonquian languages were already geographically dispersed. This suggests that an Algonquian origin of AL*Naskapi is more plausible than an Athapaskan one. The dispersal of Athapaskan at a time after the hypothesized date when Algonquians abandoned the Columbia Plateau and the absence of AL*Naskapi in groups that speak languages related to Athapaskan, such as 109

PAGE 110

Tlingit, or purportedly related, as Haid a, also support this hypothesis (Smith et al. 2000:565). The Algonkian origin of this allele seems plausible. The Proto-Athapaskan founding population may have been ex tremely small, so that the assimilation of some comparably diminutive Algic band during proto-Athapaskan timeframes might have resulted in the near ubiquity of AL*Naskapi among Athapaskans due to rapid genetic drift within a few generations. There is a large Algic-Athapaskan interface in the Subarctic and Northern Plains, and Algic and Athapaskan communities are interdigitated and cooperative in California, di fficult to distinguish archaeologically from one another, possibly indicative one recent mu ltilingual migration wave from the Plateau region (Whistler 1979; see Figure 37). In short, the distribution of the AL*Naskapi allele is indeed very useful in looking at particu lar aspects of the Athapaskan expansion such as the movement of Apacheans out of the Plains (Ives and Rice 2006). However, until more comprehensive data are available (e.g ., from the east and west coasts and from Alaska), then the relevence of AL*Naskapi to discussions of the ultimate origins of the Na-Dene is somewhat limited. A final point about AL*Naskapi is worth making, related to the rare Old World occurrence of the polymorphi sm in Northern India and Turkey. Scott and ORourke (2010) assert that: As deriving Athapaskans and/or Algonq uians from Eti Turks or North Indians seems unlikely, the possibility exists that the gene arose independently in the New World and the Ol d World. . If the gene arose through mutation in the New World, it would not help corroborate Ruhlens (1998) view that Na-Dene and Yenisei groups were derived from a common Eurasian stem population. However finding AL*Naskapi in the Kets could support one Fortescue model that holds there was an east to west movement of Dene-Yenisei an-speaking populations in the Holocene (Scott and ORourke 2010:130). 110

PAGE 111

Michael Fortescue (2010) hypothesizes that genetic data might be better explained by an east-to-west reversal the direct ion of migration (from Alaska to Western Siberia) as an explanation for the commonality of DeneYeniseian languages, which would then be classified as Paleo-American intr uders into Asia. But it is not true that the hypothetical presence of AL*Naskapi among Yeniseian speakers must necessarily substantiate such a reflux (back migrati on) model, nor does AL*Naskapi having a welldocumented presence among Turks and Nort h Indians necessarily argue for the independent New World vs. Old World origin s of the gene, as Scott and ORourke (2010) assert. I have already discussed the case of shared autosomal diseases among the Athapaskans and Turkic Yakut of Siberia so another Turkic group having genes in common with North Americans is quite plausib le. The common origin of Eti Turks and American Indians in ECA is a distinct possibilit y first discussed by the discoverers of this genetic marker among theTurks: We have suggested that the restricted distribution of the Naskapi allele makes it particularly valuable in determining population affinities between [American] Indian and Asian populations. . Modern Turkey was in antiquity part of the Eastern (Byzantine) Roman Empire, and historians of the period have provided extensive documentation of nearly continuous invasion of the area by nomadic peoples of Central and East Asian stock. . However, to distinguish between the possible origins of a Naskapi-like albumin in the Eti Turks requires colle ction of more data on the albumins of geographically intervening populations in Siberia and other parts of Central Asia (Franklin et al. 1980:5481-5482). Corroboration comes from ancient DNA studi es which show that the Xiongnu were genetically intermediate betw een Modern Turks, Mongols, and Native Americans. The co-occurance of mitochondrial DNA haplogroups A,B,C, D and M among the Xiongnu is indicative of their affinity to modern Mongolians and Native Americans (Kolman et al. 1996), and Xiongnu graves also yeild DNA s hared with modernTurkey (Keyser-Tracqui 111

PAGE 112

et al. 2003:256, 259). Ancient an d modern DNA reveals that the succession over time of different Turkic and Mongolian tribes in t he current territory of Mongolia resulted in cultural rather than genetic exchanges (Ke yser-Tracqui et al. 2006:272). In other words, genetic continuity spans diffe rent ethno-linguisti c imperial regimes. The Yeniseians of protohistor y and ethnography may indeed have their ultimate southern origins in this turbulent milieu, as Asia nists have long suspected. Therefore the presence of AL*Naskapi among Turks and/or Yeniseian speakers cannot be simply written off as evidence of indepe ndent origin of the same mutation, nor can it be used to infer back migration, with no other corroborating evidence. AL*Naskapi is just another in a long series of traits held in common between these specific populations. Mitochondrial DNA Contemporary Native American mtDNA haplogroups consist of four major lineages (A, B, C & D) present in both North and South America and one minor lineage (X) present in North America only. All five of these haplogroups are also present in Southwest Siberia / East Central Asia (E CA), but are discontinuous elsewhere in Eurasia (Mulligan et al. 2004). The distribution of the fo ur [main] founding lineage haplogroups in Native Americans from No rth, Central, and South America shows a north to south increase in the frequency of lineage B and a north to south decrease in the frequency of lineage A (Merri wether et al 1995:411). The single-origin hypothesis originated from mitochondr ial DNA studies, because statistical and geographic scarcity of New World haplogroups in Asia makes it improbable that the same four haplotypes would be drawn from one geographic region more than once (Kolman et al. 1996:1321). Genet ic drift is then used to account for lower diversity in certain regions (e.g., t he far north, where haplogroup A predominates). 112

PAGE 113

But several other Eurasian mtDNA lineages and sublineages (e.g. M & D2a1) are also found in ancient far northern North America (Gilbert et al. 2008; Malhi et al. 2007; Stone and Stoneking 1998). The loss of additional rare lineages through genetic drift, genocide and disease is easily inferred. As the speed and scale of DNA sequencing technology increases, new insights abound. Just a few years ago, researchers claimed our knowledge of mtDNA and the Y chromosome is pretty well saturated (Mulligan et al. 2004:308). But since that time, the growth in the data has outpaced our ablitiy to interpret them; the recognized maternal founding lineages of Native Americans are at least 15, indicating that the overall number of Beringian or Asian founder mitochondrial genomes will probably increase ex tensively (Perego et al. 2010:1174). And rare North American lineages may have been misattributed to European admixture, especially where the lineages are common to both Eu rope and Central Asia (Wilson 2008:270). Multiple reintroductions of the same Asi an genes are possible; there could have been multiple waves of migration from a si ngle parent population in Asia/Siberia which repeatedly reintroduced the same lineages to the New World (Merriwether et al. 1995:411). Recognition of the greater than expected diversity of American mtDNA founding lineages makes the possibility of multiple migrations from the same (or similar) source population(s) seem increasingly pl ausible. High resolution data provided by recent autosomal studies suggest an endurin g connection between the continents in recent millennia. The contintent-wide similarity of Native Americans to Asians increases the closer one gets to Bering Strait (Wang et al. 2007:2049), a swat h of the continent dominated by Athapaskans and Eskimos (the tw o major Native American ethnic groups with the strongest cultural and linguistic ties to Asia, and those with the most typically 113

PAGE 114

Asiatic phenotypes). DNA and language agree with material culture as described by William Fitzhugh: Whether or not one agrees with the transoceanic mechanism . the stylistic parallels around the Pa cific Rim are more abundant and more convincing than those around the Atlantic [and] t he number of parallels increases and becomes more specific as one proceeds north from the mouth of the Columbia River to Bering Strait (Fitzhugh 1994:29). Adhering to the now-dominant single origin paradigm, Wang and colleagues attempt to explain the paralle l gradient of increasing molecu lar similarity with proximity to Asia as the result of the steady loss of the original genetic diversity as daughter populations moved south and east away from Bering Strait (Wang et al. 2007:2059). This explains private alle les (ubiquitous American polym orphisms). But the physical similarity of Native Americans to Asians is not particularly ancient, as one would expect following the single origin model. This model reverses the chronological priority of the derivation of Asian traits indi cated by Late Holocene morpholog ical discontinuities in the physical anthropology, not to mention the paralle l archaeological evidence for the steady intensification of hist orical diffusion from Asia in the last 3000 years (Fitzhugh 1994). Wang and colleagues alternative hypothesis is more consistent with the crossdisciplinary data: Alternatively, similar patterns could re sult from gene flow across the Bering Strait in the last few thousand years, together with continual interactions between neighbors on both sides of the Bering Strait. . It is also possible to envision a series of prehistoric migr ations, possibly from the same source population, with the more recent de scendants gradually diffusing into preexisting Native American populat ions (Wang et al. 2007:2059-2060). This alternative is superior because it me shes with the cross-disciplinary evidence for Late Holocene Asian-American ties and it also accounts for t he ubiquity of private American alleles (originating in the precursor populations). It largely agrees with 114

PAGE 115

Gonzlez-Jos and colleagues attempt to introduce more flexibility into the single origin model by allowing for significant late Holo cene gene flow into the existing populations; recent circumarctic gene low would have e nabled the dispersion of northeast Asianderived characters and some particular genetic lineages from East Asia to America and vice versa (Gonzlez-Jos 2008:175). But th is begs the question: is the term single origin misapplied where such a tremendous in flux of genes is evident? I propose the beginnings of a model where later genetic di spersals from Asia were absorbed by existing gene pools, maintaining the frequency of ubiquitous private alleles, while genetic drift favored the preexisting mtDNA lineages (mostly haplogroup A in the north), in disproportion to the influence of Asiatic ethnolinguistic cultures like Dene-Yenesiean. An Earlier Study of Dene-Yen iseian Population Genetics Wilson (2008) and Scott and ORourke (2010) have both written review articles addressing the genetic implicat ions of the Dene-Yeniseian hypothesis. However, a 2002 paper by Rubicz and colleagues is the only orig inal study to have collected original genetic data to purposefully test the veracity of this hypothesis as first proposed by Ruhlen (1998). Using primarily mtDNA mole cular data (along with classical genetic immunoglobulin and blood group ma rkers), they concluded: Contrary to Ruhlens in terpretation of the linguistic data, analysis of the genetic data shows that the Na-Dene cl uster with other Native American populations, while the Kets genetically resemble the surrounding Siberian groups. . However, this study does show a significant correlation between genes and language when Na-De ne and Yeniseian are treated as distinct language families. These re sults suggest that the Na-Dene and Yeniseian populations are both genetically and lin guistically unrelated (Rubicz et al. 2002:743, 755). But Vajda (2010a, 2010b) has subsequent ly shown that a Dene-Yeniseian linguistic relationship is very likely, so the series of three caveats introduced by Rubicz 115

PAGE 116

and colleagues (Rubicz et al. 2002:754) must be reconsidered, as John Ives has pointed out, it may be that: Yeniseian and Dene populations did not share a common genetic ancestry, but that there had been hor izontal language transmissi on from one of these groups to the other; or, Yeniseian and Dene did have a common origin, but languages and genes evolved at different rates; or, Yeniseian and Dene had common linguistic and genetic origins, but recent genetic differentiation [via gene flow] had obscured the genet ic relationship (Ives 2010:325). I am inclined to disregard (2), evoluti onary differential, which assumes much slower-than-expected language evolution and/or faster-t han-expected genetic evolution; these would require some form of Athapaskan exceptionalis m. Alternatively, some combination of (1) horizontal lang uage transmission thr ough demic expansion and elite dominance and (3) gene flow from surrounding populations swamping the original genetic similarity, is not just possible, but quite probable under the known circumstances. These two processes (gene flow and horizontal language transmission) are closely linked phenomena alo ng any cultural frontier, and were directly observed in ethnohistorical timeframes. Signficant gene flow across linguistic boundaries is the rule rather than the exception (Hunley and Long 2005). Rubicz and colleagues study is flawed because most of the non-Na-Dene Native American groups sampled (Blackfoot, Cree, Yupik, Ojibwa, and Papago) have had well-documented, direct, sustained interactions with Athapaskan speakers within historical timeframes to say nothing of late prehistory. The only two groups who did not have direct relationships with Athapaskan speakers (St. Lawrence Yupik and subarctic Ojibwa) nonetheless had indire ct or collateral relations with Athapaskans via their imme diate neighbors of the same linguistic family (Cree in the case of the Ojibwa; Alaskan Yupik in the case of St. Lawr ence). The Cree have a 116

PAGE 117

vast frontier-interface with Chipewyan and Beaver Athapaskans, and historical alliances with both the Chipewyan and the Ojibwa. Al askan Yupik have close ties to Denaina Athapaskans. Arizona Papago were nearly ove rrun by Apache after the seventeenth century, and the two groups formed numer ous marriage alliances between periods of sporadic conflict. Finally, the Blackfoot virtua lly assimilated the Tsuutina. In short, all the non Na-Dene North American natives sele cted for Rubicz and colleagues study were peripheral to and engaged with Na-Dene speakers in a greater cultural sphere, and many have well-known kinship ties to specif ic Athapaskan groups. Thus their blood-affinity to Na-Dene speakers cannot be used to deny Athapaskan linguistic ties to Asia. To assess the possibility of blood ki nship between Na-Dene and Yeniseian, one must acknowledge that neit her language group exists in a vacuum, but are both shaped by continuous historical affiliations with neighboring peoples. DNA evidence strongly indicates that Native American ancestors likely came originally from Central Asia, particularly the Altai region and southern Sibe ria (Derenko et al. 2001; Kolman et al. 1996; Starikovskaya et al. 2005; Zakharov et al. 2004; Zegura et al. 2004). We have not yet developed statistical methods for te asing out the stealthy signatures of later migrations from the same source populations. The molecular genetic data (in close agreement with the biometric dat a) suggest that the strong similarity between American and Central Asian gene pools has developed in tandem (Yao et al. 2004). This is impossible to explain with just one ancient migration, and some single origin proponents favor the notion that this is just a fortuitous coincidence. [O]ne must exercise caution before suggesting that Native Americans are descendants of Altai populations. The ri ch diversity of mtDNA lineages in 117

PAGE 118

the Altai suggests a pattern of gene flow with a spectrum of Eurasian populations. A genetic melting pot could easily contain all of the elements common in Native Americans and falsely give the impression of being the ancestor (Mulligan et al. 2004:302). But this begs the question: how could significant recent Asian connections go largely undetected in the realm of mtDNA? Re cent migrations are strongly implied by Dene-Yeniseian language links, skeletal morphology showing derived traits from Late Holocene Asia, and numerous cultural parallelisms at the level of minor details which could be explained only by particular histor ical links between corresponding traditions (Berezkin 2005:79). One possibility is that closely related mt-DNA lineages have been reintroduced from Asia on multiple occasions. Mitochondrial Haplogroup A2a Mitochondrial Haplogroup A is ancient and widespread throughout the Americas. But Na-Dene-speakers have extremely hi gh frequencies of a si ngle subhaplogroup, classified as the A2a clade, indicating a founder effect prio r to the Athapaskan expansion (Tamm et al. 2007; Malhi et al. 2008) This A2a clade is also shared by the Inuit and associated with the Neo-Eskimo expa nsion. Both ethnol inguistic expansions occurred during the Common Era, and both originated among neighbor ing populations in Alaska or Beringia. A2a is not found anyw here south of the historical Apachean range in Northern Mexico (Achilli et al. 2008; Gilbert et al. 2008; Perego et al. 2010). The age of modern A2a among Greenlandic Eskimos (using molecular divergence estimates) is between 1000 and 2000 years, cons istent with the hypothesis of a more recent origin and spread of the Neo-Eskimo Thule culture (Gilbert et al. 2008:1788). A2a is also found in among Samoyedic and Evenki (Tungus) speakers in the Yenisei-Ob Sayan region of Southwest Siberia, precisely the region where Bronze-Age admixed 118

PAGE 119

americanoid skeletons were found (Kozin tsev et al. 1999), and where Yeniseian speakers resided as recently 200 years ago. Tamm and colleagues see this as a reflux migration from America back to Central Asia, in keeping with the single origin hypothesis: Surprisingly, we also found a Native American sub-type of haplogroup A2 among Evenks and Selkups in souther n and western Siberia. . Previously, this HVS I motif is reported in one Yakut-speaking Evenk in northwestern Siberia. . A novel dem ographic scenario of relatively recent gene flow from Beringia to deep into western Siberia (Samoyedic-speaking Selkups) is the most likely expl anation for the phylogeography of haplogroup A2a, which is nested within an otherwise exclusively Native American A2 phylogeny (Tamm et al. 2007:4). Unlike A2a however, the D9S1120 autosomal microsatellite 275 base pair allele is not found in any South or Central Siberian groups. This ubiquitous Native American private allele is found at moderately high frequencies (10-50%) in all Native American populations and also among the Chukchis and Koryaks on the Asian side of Bering Strait (Scott and ORourke 2010:129). This absence in the Eurasian heartland is truly remarkable if A2a is indeed the result of a long distance back migration from America as Tamm and colleagues allege. Si nce the autosomes have four times the effective population size of mtDNA, some A2a mt DNA carriers would be expected to leave D9S1120 in the wake of their jour ney. If the original sour ce of A2 lineage is Alaska (Volodko et al. 2008), then it seems most probable that this source came from a segment of the Alaskan population lacking the D9S1120 275bp allele. A2a on the other hand is not universal in the Americas but is exclusive to the NeoEskimo and Na-Dene, two groups whose arc haeological footprint is shallow and whose origin is proximal to Asialess than 2500 years before present for both groups. A2a has only recently been dispersed continent-wide in North America through rapid 119

PAGE 120

territorial expansion in late prehistory. Yenisei/Ob/Sayan A2a lineages could even be ancestral to late Holocene Bering Strait A 2a lineages which have only spread as far as Greenland and Northern Mexico in the last m illennium or so. At the very least, the widespread presence of the haplogroup reflects robust historical ties between Northeast Asia and Central Asia. Achilli et al. suggest that the A2a lineage is much younger than the other Native American A mtDNA lineages, thus reflecti ng a much later expansion, A2a (Siberians, Inuits and NaDen) . is probably due to secondary expansions of haplogroup A2 from Beringia long after the end of the LGM (Achilli et al. 2008:6). Scott and ORourke discuss the the existence of A2a in Southwest Siberia: While the Selkups are not Kets and they speak a different language (Uralic), the Kets and Selkups are closely aligned in many dendrograms, perhaps reflecting the role of geographic propinquity on historical patterns of gene flow. Another possibility is that this marker [A2a] was reintroduced from the East during the Holocene as proposed . by Michael Fortescue [2010]. . Unfortunately, at this time, there are very few genes across Beringia that support this scenario (Scott and ORourke 2010:129, 133). Scott and ORourke (2010:129) refer to A2a as being common among American Indians in general, but neglect to mention that A2a is considered a specifically Athapaskan and Eskimo founding lineage, thus limited to portions of North America. Geographic propinquity is crucial, but t he likelihood of horizont al language change or ancient bilingualism on the Uralic-Yeniseian frontier is another issue The Selkups not only neighbor Kets, but were culturally infl uenced by Kets too (Wixman 1984:175). Male Mediated Migration Long-term Asian gene flow and multiple migr ations from Asia appear possible, but they are well-camouflaged by si milar background genetic diversity; thus they are difficult to measure (Hey 2005:970-971). Further, historical scenar ios can be constructed to 120

PAGE 121

account for recent migrations that would have preserved the general distribution of ubiquitous, ancient mtDNA found ing lineages and private autosomal alleles. Ives notes: Just as Moore (1994) pr edicted, we have entered an era in which grand syntheses of biological anthropologica l, linguistic and archaeological data have returned to prominence. . [A]pparent discr epancies between different forms of evidence may spur us to pronounce too quickly that one proposition or another is simply not possible (Ives 2010:331-332). Recent work raises the issue of w hether the Na-Dene founding population was sexually asymmetrical and male-skewed, in which case their mitochondrial DNA is expected to be indigenous to the Americas ev en as their Y-chromosomes may have come from historical Asia (Wilson 2008; Iv es 2010). Bryan C. Gordons (2012) synthesis of ethnology and genetics suggests that the Southern Athapaskan migration was initiated by mostly males who were dis enfranchised caribou hunters from polygamous communities practicing female infanticide. This may help to explain why Apacheans in a very short time came to possess a plethor a of local southwestern mitochondrial DNA lineages, while speaking a language of purel y Canadian derivation (Romero 1998). John Ives is particularly insightful in using Apachean male-mediated migration as an analog for the entrance of Dene-Yeniseian speakers to Alaska ; the nature of the Apachean migration might be fundamentally sim ilar to that of earlier Dene-Yeniseians. He notes that for the Apacheans, [t]hi s small founding population experienced successful growth . through extensive incorporation of neighboring peoples, particularly women (Ives 2010:330) He presciently relates th is fact to the arrival of Dene-Yeniseians in Alaska: Should members of linguistically and genetically unrelated communities regularly be incorporated into a conservative speech community, the linguistic identity would tend to survive, but the genetic signat ure would be steadily altered (Ives 2010:331). 121

PAGE 122

For the Apacheans, the male bias of t he founding population(s) may have been very high, even as much as two to one, that is if the ~50% ratio of foreign mtDNA lineages among contemporary Na vajo proportionally reflects the sex-bias in the founding population (Lorenz and Smith 1996). Closer to 100% of Northern Athapaskan maternal lineages are shared with Eskimos, suggesting that the male-bias of their founding population could have been even higher. If female founders were few enough in number, then their lineages co uld have been lost through genetic drift, as is the rule in the far north. Mitochondria l DNA diversity in the region is very low (haplogroup A is near fixation). Recently, two extinct mitochondrial haplotypes (M & D2a1) were discovered in Canadian human remains (Gilbert et al. 2008; Malhi et al. 2007 ). This suggests that the northern ecosystems could act as a filter over time, with genetic drift preventing newer, rare Asian-derived mtDNA types from gaining a foothold, just as local isolates still fail to survive in the Subarctic (Mei klejohn 1977:110). Y-Chromosomes Y-chromosome data also reveal relatively few Native American founding lineages, forming a subset of Eurasian genetic divers ity. In a survey by Stephen Zegura and colleagues, three major haplogr oups, denoted as C, Q, and R, accounted for nearly 96% of Native American Y chromosomes (Zegura et al. 2004:164). These authors favor the single-origin hypothesis, and suggest that two of these lineages (C & Q) are paleo-American founding lineages more than 10,000 years ol d, and the other one (R) is a product of post-Columbian European admixture. Other scholarship alternatively suggests that only the Q sub-branch M3 is the paleo-American lineage, and that the main branch Q-M242, C and R were each in troduced (or independently reintroduced) by Na-Dene speaking immigrants in a scenario consistent with the Dene-Yeniseian 122

PAGE 123

hypothesis (Bortoloni et al. 2003; Lell et al 2002). Each of the th ree paternal lineages (Q, C, and R) will be crit ically addressed below. Haplogroup Q A subclade of haplogroup Q (denoted Q1a3a1 and/ or Q-M3, formerly Q3) is an ancient lineage indigenous to the Americas, whose co mmon ancestor clade Q*-M242 (the defining mutation for t he Q haplogroup) is found at moderate frequency among the Kets and Altai in Central Asia (Zegura et al. 2004). The two Q-branches very likely diverged between 10,000 and 15,000 years ago. The ubiquity and antiquity of Q-M3 in America makes it unlikely to have been asso ciated with a recent expansion of DeneYeniseian speakers. Norther n Athapaskan speakers have all Q variants at a frequency of between 20-50% (Malhi et al. 2008), which would seem like a high frequency for most populations. But it is exceedingly low for Native Americans where the typical frequency of haplogroup Q (Q-M3 + Q*-M242) is substantially more than 50%the Subarctic is the American cult ure area with the lowest frequency of Q, reflecting the expansion of Athapaskan speakers from a small nucleus wh ich included a significant number of nonQ lineages. The proto-Na-Dene were a very small founding population, confirmed by a large number of bottleneck-derived autosomal genetic disorders (Erickson 2009). Furthermore, the pan-Native American derived Q-M3 is rare among Northern Athapaskansthe underived Q*-M242 is common among them. The typical Athapaskan version of Q (Q*-M242) is more lik e those surviving in Central Asia among the Kets and their neighbors, than it is like any of the Q-M3 daughter lineages which can accurately be classified as paleo-American. The low-to-moderate frequency of private American Indian Q-M3 among Northern Athapaskans can be explained as the signature of admixture with non-Athapaskans in the last millennium or more. Haplogroup Q123

PAGE 124

M242* . is the second most prevalent in the Chipewayan, where it was observed at a frequency of 25%. . Q-M242* was also de tected at a low frequency (4%) in Mongolia (Bortolini et al. 2003:528). The underived form of Q (i.e. Q-M242* ) is plausibly a signature of a Late Holocene expansion from the same s ource population, still found in Central Asia, and iclosely associated with the Na-Dene. Karafet and co lleagues found [t]he vast majority of haplogroup Q chromosomes (79.5%) occurred in onl y two Siberian populations, the Kets and the Selkups, with frequencies of 93.8% and 66.4%, respectively (Karafet et al. 2002:772). The Chipewyan Q-M242* variant is molecularly closer to Mongolian QM242* than either Chipewyan or Mongolian is to common Native American Q-M3 or South American Q-M242* variants (Bortolini et al. 2003:535), which favors the view that the Chipewyan and Mongolian lineages have a more recent common ancestor than the other groups. Na-Dene Q-M242* may have been in correctly pooled with Q-M3. Not all Q lineages are necessarily paleo-American in origin; some have very likely been reintroduced by subsequent Central Asian immigrants. Haplogroup C The highest frequency of the common Na tive American Y-DNA haplogroup C3bP39 is found among Na-Dene language speake rs and their nearest neighbors (Malhi et al. 2008; Wells 2006; Zhong et al. 2010). The contiguous zone in which haplogroup C3b is found corresponds almost perfectly to the geographical limits of the Athapaskan expansion, enveloping all three discrete blocs of the language phylum, with little extraneous territory (Figure 3-8) This chromosome is relati vely rare elsewhere in the continent and is also very similar to (pre sumably) ancestral C3*-M217 variants found in Siberia and East Central Asia. This has prompted some to suggest this chromosome 124

PAGE 125

demonstrates the Na-Dene languages arrived as part of a separate major migration from Asia, distinct from ot her American Indians and Eskimo s (Bortolini et al. 2003; Lell et al. 2002). Zegura and colleagues (2004) contrary asse rtion that haplogroup C is ancient and widespread among Native Americans is based the observation that the haplogroup has been found in all three of Gr eenbergs (1987) putative Native American linguistic phyla (Na-Dene, Amerind, and Eskimo -Aleut). The vastly different frequencies of C among different Native American families are attri buted to genetic drift in a small polymorphic founding population. This logic is suspect because the very existence of Greenbergs Amerind phylum is rejected by a substantial majority of Americanist linguists (Bolnick et al. 2004). Admixture pathways of C3-lineages from Na-Dene source populations into non-Na-Dene groups during the Common Era ar e generally straightforward and easy to recognize. The recipient populations are addressed one at a time. Haplogroup-C among Greenlandic Eskimos Two contemporary east Greenlandic Itt oqqortoormiit (Neo-E skimo) haplogroup C males have been identified (Bosch et al. 2003) The possible subhaplogroups of these Inuit C Y-chromosomes were never determi ned. Every well-documented occurrence of haplogroup C in North Americ a is C3-M217 subgroup C3b, defined by the P39 mutation (Zhong et al. 2010). It is not certain which sp ecific branch of the C-clade would contain these unresolved Greenlandic examples. Nonetheless, Zegura and colleagues (Zegura et al. 2004) use these two isolated individuals to infer the presence of haplogroup C in a proto-Eskimo-Aleut population, subjected to extreme loss of diversity through genetic drift. This view is necessary to maintain a single-origin for all Native Americans, despite the rather strict geographic segregation of C3b-P39 to Na-Dene speakers and their 125

PAGE 126

immediate neighbors. The problem with this hypothesis is that the Greenlandic Eskimos resided in close proximity to Na-D ene speakers in the not too distant past. The well documented expansion of Thule Inuit out of the Bering Strait region took place entirely during the early-to-mid Co mmon Era (Gullv and Mc Ghee 2006). These particular Inuit C-lineages may have direct ances tors in western Alaska or Bering Strait in just the last 1000-2000 years, just prior to the Thule Ex pansion. Several authors note that Eskimos and Athapaskans have enough biol ogically in common that they could be considered one population (Schurr 2004; Sz athmary and Ossenberg 1978). Yet the dissimilarity in C3 Y-DNA frequencies bet ween Eskimos and Athapaskans (very high in Na-Dene but very low in Inuit) indicates that Athapaskans may have received more gene flow from their Alaskan Neo-Eskimo contemporaries than the reverse, and thus the two families origins may be technica lly distinct, despite being inextricably interwoven and complicated by prehistoric intimacy in Alaska. Sexual asymmetry among the Na-Dene founder population in Alaska is one possible expl anation for this. The current distribution of mtDNA haplogroup A2a in Greenland is indicative of an Alaskan origin in close proximity to Na -Dene speakers within the last 800-1000 years (Saillard et al. 2000). The recent Alaskan origin of Greenlandic C Y-DNA is also probable given classical biometric studies showing Greenlandic Eskimos physically resemble western Eskimos and Athapaskans more than any of these three groups resembles the eastern Canadian Eskimos (S hapiro 1931). Origins of the rare Ittoqqortoormiit haplogroup C Y-ch romosomes are most likely to be found along the late prehistoric Eskimo-Athapaskan continuum somewhere between western Alaska and northwest Canada. 126

PAGE 127

Haplogroup C3 among the Cheyenne, Arapaho, Sioux and Ojibwa A distribution of haplogroup C3 is present in a swathe of the Northern Plains and Upper Midwest southward to Oklahoma (Bergen et al. 1999; Bolnick et al. 2006). This zone is contiguous with both major Athapaskan bl ocs, and it thus requires little more than gene-flow to explain the presence of haplogroup C3 there. There is a well documented history of Tsuutina (Sarsi) Northern Athapaskans and Plains Apache moving through the Plains corridor in the last few centuries (Gunnerson and Gunnerson 1971). Sioux are relatively recent arrivals in their current territo ry of the Dakotas, believed by a number of ar chaeologists and historians to have been a major Apachean stronghold in the eighteenth century (Grinnell 1920; Hyde 1959; Wilcox 1981). C3b-P39 is widespread among Na-Dene groups, at frequencie s as high as 25% to 45% (Wells 2006; Malhi et al. 2008). Most Plains tr ibes possessing frequent C3b Y-chromosomes until recently occupied the former territory of the itinerant Athapaskans, also sharing numerous material cultural features with t heir Athapaskan neighbors (i.e the Apacheans and Tsuutina), e.g. hard soled moccasins, si new-backed bows and bison hide shields, and domicile design, among many other things (Baldwin, 1997; Brasser 1979). The Ojibwa, Cheyenne and Sioux effectiv ely form a band of genetic continuity connecting the Northern and Sout hern Athapaskan blocs. It is likely that horizontal language transmission may have taken place here, as Apachean stragglers or remnant bands were assimilated by Sioux and Chey enne after the eighteenth century. The nineteenth-century witnessed a coalescence of pan-tribal northern Pl ains society after an extended period of populat ion decline (Taylor 1977). This situation favored gene flow and genetic drift, reduc ing overall diversity and altering gene frequencies. The Southern Cheyenne allotment census of 1892 demonstrates the Cheyenne did 127

PAGE 128

assimilate some Apaches (Moore 1994:937). A very severe disease-induced genetic bottleneck impacted the Cheyenne in the late nineteenth-century (Bergen et al. 1999). This could easily have amplified the proporti on of C3 lineages in the population through genetic drift. The Chiricahua Apache were relo cated to Fort Sill, Oklahoma in 1894. Cheyenne, Arapaho, Kiowa, (local) Plains Apache and (non-local) Chiricahua Apache tribes have all been near neighb ors or cohabitants of several reservations tightly clustered just east of Oklahoma City, likewis e as urban Indians in Oklahoma City itself, which has the largest population of off-reservation Apaches anywhere. The assumption that the gene frequencies among any groups currently residing in Oklahoma are accurate reflections of their pre-contact situation is high ly problematic. Cheyennes, Sioux, Plains Apache and Arapa ho all had similar equestrian lifestyles in the nineteenthcentury and extensive historic al range southward from base camps in the Black Hills. The presence of haplogroup C3 in any groups residing in this zone can be most easily attributed to an ultimate Na-Dene source. The virtual absence of C3 in the Great Basin is in perfect agreement with absence of the autosomal marker Albumin Naskapi (A L *Naskapi), a variant also common among the Athapaskans and their Algonq uian neighbors of the Plains and Subarctic (Smith et al. 2000). This may be a signal that the Numic Expansion (the movement of UtoAztecan speakers into the Great Basin) involved later movements of populations originating from the sout h, who did not receive appreciable gene flow from Athapaskans. Alternatively, it could mean that the Athapaskan migrants avoided the Great Basin on their long trek south (Ives and Rice 2006). 128

PAGE 129

The occurrence of C3b-P39 among westernm ost Ojibwa (Bolnick et al. 2006) is also readily attributable to gene flow th rough the Algonquian bloc on the PlainsSubarctic boundary. The Ojibwas residence in the middle Subarctic, where exogamy and endogamy co-occur by necessity (obliter ating demic boundaries), is sufficient for them to have absorbed significant Na-D ene-derived genes from the west, via their Northern Plains Algonquian and Siouan neighbors (Meiklejohn 1977). C3 in the Southeast? Unlike the straightforward presence of haplogroup C in the north and west, the rare occurrences of C3 lineages among two different Southeastern peoples would pose more of an apparent problem for the Athapas kan/Na-Dene origin of this paternal lineage, except that both of these groups were reloca ted to northeastern Oklahoma circa 175 years ago, where they would hav e been in direct, sustained contact with Plains tribes including Apaches and closely associated peoples. Bolnick and colleagues found haplogroup C3 present in one Muscogee Creek individual and one Cherokee individual residing in Oklahoma (Bolnick et al. 2006). Bolnick and Smith (2003) include the detailed sample informati on for the later study. These are both classified as Southeastern-derived C3 lin eages, presumably because the Creeks and Cherokee share an ancestral homeland in the Great Smoky Mountains and a historical legacy amongst the so-called Five Civilized Tribes in the early nineteenth-century southeast. However, both of these specific case s are readily explicable as the result of historical interactions with Plains residents after 1835-6, when the Trail of Tears brought most of the Muscogee Creek and Cherokee to re servations in northeastern Oklahoma. The indigenous Plains Apache of Oklahoma (l argely assimilated by the powerful Kiowa) 129

PAGE 130

would have spread their paternal lineages far and wide in Oklahoma territory for generations prior to the Cherokee arrival. The relocation of captive Chiricahuas in the 1890s introduced more C3 lineages to Oklahoma, where the chromosome was al ready well established among numerous Plains peoples, and among urban Indians in Tulsa and Oklahoma City. The likelihood that the single Creek and single Cherokee male inhe rited their haplogroup C3 Ychromosomes from any of a plethora of local sources in Oklahoma is overwhelming. There is virtually no chance this C3 lineage is original to Creek and Cherokee migrants. The choice by Bolnick and coll eagues to genetically test the Oklahoma Cherokee and Creek tribes (rather than the smaller units of the same tribes located in North Carolina and Alabama respectively) is per fectly understandable be cause the Oklahoma tribes are more likely to provide a larger sample of reputed full bloods, whereas the communities still remaining in the Southeas t are heavily admixed with non-natives. However, this strategy has the major drawback of being blind to the much subtler effects of admixture with other Native Amer ican groups in the pan-tribal melting pot that is the state of Oklahoma. Native gr oups have been relocated there from many hundreds of miles apart. During the nineteent h-century, peoples originally from the Southeast would have encountered other forcibly re located first nations from as far west the Columbia Plateau, and all would have be en placed in reasonable proximity to local natives (such as Plains Apache), and would have surely encountered one another in major urban centers like Tulsa (originally a Muscogee Creek settlement) during the era of the emergence of widespread PanIndian solidarity movements. 130

PAGE 131

Bolnick and colleagues (Bolnick et al. 2004, 2006; Bolnick and Smith 2003) are to be applauded for their comprehen siveness and careful attention to ethnohistorical detail in their discussions of the demographic histor y of the Southeast. This aspect of their work makes them noteworthy among molecular anthropologists who are too often inattentive to the contribution humanistic re search can make to their discipline. However, their detailed historical summa ry of the post-contact era through the relocation period is not matched by thei r rather cursory approach to the demographic realities of life in Oklahoma since 1835. [T]he geographic proximity of many re servations in Oklahoma may have increased gene flow among the southeaste rn populations. Historical events may have therefore contributed to the observed patterns of Y chromosome variation (e.g., closer paternal relationships among populations from the Southeast than among those from the Northeast). However, such events cannot explain the opposite pattern observed in the mtDNA data (Bolnick and Smith 2003), so they must have had less effect on eastern North American genetic variation than past patterns of postmarital residence (Bolnick et al. 2006:2171). In the Southeast, kinship was reckoned matrilineally and postmarital residence was matrilocal. The effect of this marital pattern on rate s of sexually asymmetrical admixture among Creek and Cherok ee cannot be denied, but it is incorrect to imply that this effect was limited to the contact-traditional period before 1835; traditional customs likely prevailed in intertribal marriages in nineteenth-century Oklahoma, especially when multiple disparate tribes were known to have similar (or nearly identical) customary kinship and residence patterns. Oklahoma Apaches have historically included several groups; local Plains Apaches, itinerant Jicarilla and Lipan Apaches, and forcibly relocated Chiricahua Apaches (residing in a concentration camp at Fort Sill between 1894 and 1913). Since circa 1835, formal marital arrangements bet ween traditional Apaches and Southeastern 131

PAGE 132

peoples are quite conceivable, as the predominance of matrilineal matrilocal clan systems and exogamous marriage customs am ong southeastern societies are quite comparable to traditional Apachean kins hip systems (Lewis and Jordan 2008; Opler 1936). The nineteenth-century progeny of A pache men married to Creek or Cherokee women would result in individuals who were considered full bloo ded Indians and fullfledged members of their maternal clans. In both southeastern and Apachean custom, the tendency would be to deemphasize the paternal cultural heritage of an individual in favor of the maternal heritage, and this could re sult in a stealthy form of intertribal malebiased admixture, even in t he most culturally conservative of groups, making their distinct genetic origins much harder to det ermine through Y-chromosome analysis. In this regard, historical European and African American admixture observed amongst the remnant Creeks in Alabama and Cherokees in North Carolina is a more straightforward problem, and less likely to produce errors in inte rpretation. This whole issue is more of a problem for scientists than it is for natives, for whom c oncepts of genetic ancestry are far less important than social identification with a particular group (Bolnick et al. 2006). C3*-M217 in South America The greatest difficulty in addressing the New World history of Y-DNA haplogroup C3 is the presence of two C3*-M217 indi viduals among the Way uu tribe of the La Guajira Peninsula, on the Atlantic Coast of Northern Columbia and Venezuela (Karafet et al. 1999); these cannot possibly be of Na -Dene origin. However, the Late Holocene saw the sudden appearance of der ived Asiatic skeletal morphology in both continents (Gonzlez-Jos et al. 2008). If Y-DNA hap logroup C3 is a signal of Late Holocene migration from Asia, what could explain its isolated presence in the northwestern quadrants of both continents? It is hi ghly improbable that Apache genes ever 132

PAGE 133

penetrated that far south of historical Apache raiding grounds in northern Mexico. Furthermore, the Wayuu C3*-M217 variant is underived with respect to the C3b-P39 variant common to Na-Dene and other North American natives. The Wayuu C3*-M217 is separated by a full 6 mutational steps from the (Na-Dene) C3b-P39 form, reflecting its marked divergence from the predominant Native American C-haplogroup (Zegura et al. 2004:169). However, C3b-P39 is rooted in the Altai/Selkup/ Ket population system (includingYeniseian forebears of the Na-Dene) while the Wayuu variant of C3-M217 is rooted in the Other Asians population system (Figure 3-9). Only two mutational steps separate Way uu C3-M217 from two different Other Asian C3*-M217 lineages, whereas Na-Dene C3b-P39 variants are separated by only two mutational steps from a pool of M 217 lineages including Yeniseian speakers (Zegura et al. 2004). Only one mutational step (in sixteen microsatellite markers) separates a particular Navajo C3 variant from an Amur Basin M217 variant in northeastern Siberia (Lell et al. 2002). This means it is concei vable that the two different versions of the M217 haplogroup were introduced as separate founding lineages, one of which is Dene-Yeniseian. Haplogroup R1 and Post-Columbian Admixture NRY Haplogroup R1-M173 is extremely common among Native Americans in both North and South America, likewise comm on among Eskimos and Athapaskans, but with highest frequency on the Northeast coast, Eastern Arctic and Subarctic. It is also extremely common in Western Europe. Lel l and colleagues, noting its presence in the Amur Delta region along wit h the apparent ancestors to Na-Dene C3 lineages, have suggested that it was part of a more recent major migrat ion that brought the Na-Dene (Lell et al. 2002). This assertion has been vociferously challenged because there is a 133

PAGE 134

clear correlation between the frequency of R1 and the amount of European malemediated admixture since 1492; the extreme northeastern concentration of R1-M173 (around the beachhead for transatlant ic colonization) is strongly indicative that many of these chromosomes were post-Columbian intr oductions (Bosch et al. 2003; Malhi et al. 2008; Zegura et al. 2004). But simply assu ming all R1 lineages are post-Columbian is not sufficient. In a follow-up ar ticle, Lell, Sukernik and Walla ce defend a Siberian origin hypothesis for Na-Dene R1: However, this [admixture] argument would require that the proposed European male input into the Native American populations not only was extensive but also brought only a limited number of Y-chromosome haplotypes. . This possibility is cont rary to the historical fact that European male admixture into Native American populations has been continuous over the past 500 years and that it has been derived from populations throughout western Europe. By contrast, the M45b [R1*] Ychromosome microsatellite markers that we found in northern North Americans are either identical to or cl osely related to those that we found in eastern Siberia. Hence, we feel that it is much more likely that the M45b [R1*] Y chromosomes, which are common in northern Native Americans, came from the Siberian Pacific, where the remnants of their exact counterparts are currently located ( Lell, Sukernik and Wallace 2002:13801381). The lowest frequency of R1 is found among the Southern Athapaskans and particularly the Navajo, considered by numerous other measures to be among the most heterozygous and therefore admixed Atha paskan populations, whereas the highest frequency is found in the relatively unadmixed Northern Athapaskans, strongly suggesting that all Native American R1 linea ges cannot be dismissed as resulting from Euro-American admixture (Figure 3-10). The century of close interaction between Chipewyan and the Mtis communities around Fort Chipewyan has resulted in a number of very old Chipewyan families with European surnames (McCormack 1988), and thus European admixture is doubtless at 134

PAGE 135

play in the fact that more than 60% of Ch ipewyan y-chromosomes are R1. However the uniformly high frequency of R1 among Northern Athapaskans, and its consistent frequency in proportion with diagnositic haplogroup C3 in Athapaskans in general, would cast doubt on the i dea that all Athapaskan R1 chromosomes are Europeanderived; it is also not intuitively consis tent with the high leve l of traditionalism and cultural vitality among many of the groups in question (Bortolini et al. 2003). And a question is implied by Lell an d his colleagues quoted above, why must R1 in particular be so common, when many other European Y-lineages were introduced after 1492? Some of the Na-Dene R1 lineages may have a Siberian and/or ECA or igin (Bortolini et al. 2003; Lell et al. 2002). If so, how might it be possible to determine which ones are pre-Columbian, and which are po st-Columbian in origin? The R1*-M173 parent clade is widespr ead throughout the whole of Eurasia (Karafet et al. 2001). Ma lhi and colleagues (Malhi et al. 2008) assume European admixture, but do not differentiate bet ween various R subclades among the Athapaskans they surveyed. Fully 76 of 79 Native American R chromosomes were (apparently) European-derived R1b (Zegura et al. 2004), although R1b also has a wide distribution in Eurasia. There was one Pima example of Haplogroup R-M124 and two Inuit Examples of R-M17 (Tatiana Karafe t, personal communication 7 January 2011). While R-M17 could also be European, it is very common among Central Asians, and RM124 is restricted to the Indian subcontinen t, Iran, and central Asia (Cordaux et al. 2004:232). Some Native American R lineages are unlikely to be a result of European admixture. Only 8 of Zegura and colleag ues 79 R1 chromosomes were from Na-Dene groups (1 out of 12 Tanana and 7 of 174 Apac hean), and the total sample of Northern 135

PAGE 136

Athapaskan speakers (a group normally very hi gh in R) was just 12 (Zegura et al. 2004). Furthermore, 28% (22/79) of all the R1 chromosomes in their sample came from just one group (Sioux) and 31% (25/79) ca me from Mesoamerica between Southern Mexico and Northern Colombia; thus the ma jority of R1 chromosomes surveyed came from an extremely limited geogr aphic area. No one (to my knowledge) has provided any recent detailed scrutiny of sequence data for the high rate of R1 in (apparently) relatively unadmixed Northern Athapaskans. A clear understanding of the proportion of various R su bclades in various native groups with known European contact histories w ould help to clarify the possible source populations for Native American R chromoso mes, pre-Columbian or otherwise. For example, in Western Eur ope, the R1b subclade is far more common, with R1a infrequent Britain, and virtually absent in t he Iberian Peninsula. Thus, the Iberian colonizers of Mesomerica have contributed virtually no R1a chromosomes to the mestizo and indigenous communities there, and the near-exclusive presence of R1b in Central and South America is consistent with the colonizers Hispanic/Iberian origins (Zegura et al. 2004). For Anglo-Saxon admix ture in Northern North America (e.g., among the Sioux), R1a would also be necessar ily very low compared to R1b. The relative frequency of R1a (compared to R1b) in creases dramatically in Eastern Europe and Northern Europe, and the hi storical migrations of Scandinavians to Greenland and Russians to Siberia and Alaska would be expected to introduce R1a chromosomes in higher percentages to admixed populations (Zlo jutro 2008; Bosch et al. 2003). Central European migrants (like Germans) would be expected have intermediate frequencies of both types (Underhill et al. 2010). 136

PAGE 137

Among the peoples of the Alta ian/Sayan region of Central Asia thought to be collateral relatives of American Indians, there is a diverse range of haplogroup R chromosomes (including ocassionally R1 and R1b3) but the frequency of the R1a subclade R1a1 is particularly high, often greater than half, i.e. higher even than in Russians or Norwegians (Kharkov et al. 2007). If these Al taian R1 frequencies were found to be similar in antiquity, the pres ence of Siberian-deriv ed R1a1 or other R1 variants among Athapaskans and Inuit would be quite possible. Whereas Bosch and colleagues were swift to dismiss the presenc e of R1a and R1* among Inuit as a product of Scandinavian admixture (Bosch et al. 2003), this is not an ironclad assumption, noting the presence of both R 1a and R1* among native Altaians. The dispersal patterns of R1a are not clearly under stood, as no marker has ye t been described that would distinguish European R1a chromosomes from Asian (Underhill et al. 2010:479). Furthermore, the proportion of the three predominant Y-chromosome haplogroups Q242, R1 and C3 are roughly similar in Athapaskans and in Altaians. The Altaian frequencies of R1 are also widely presumed to be derived from admixture, albeit millennia ago rather than in recent centur ies, as Kharkov and colleagues note: Haplogroup R1a1 prevailed in both . Southern and Northern Altaians. . This haplogroup is thought to be associ ated with the eastward expansion of early Indo-Europeans, and marks Caucasoi d element in the gene pools of South Siberian populations. . [T]he second frequent haplogroup Q* represents paleo-Asiatic marker, probably associated with the Ket and Samoyedic contributions to the Altaic gene pool. . The presence of haplogroups C3xM77, C3c . reflects the contribution of Central Asian Mongoloid groups. These haplogroups, probably, mark the latest movements of Mongolian migrants from the territory of contemporary Tuva and Mongolia (Kharkov et al. 2007:551). The fact that haplogroups Q, C and R frequently co-occur as the three major lineages in precisely the same allegedly paleo-Asiatic S outhwest Siberian populations 137

PAGE 138

should give us pause. Karafe t and colleagues found unspecified R lineages constituted 10% of their Siberian sample, but were signifi cantly higher than this in the Yenesei and Altai regions, residence of possible relatives of the Athapaskans (Karafet et al. 2002). The estimated age of R-SRY10831b (roughly 4000 years) is well after early human dispersals into Siber ia. It has been suggested that R-SRY10831b likely traces a population migr ation originating somew here in southern Russia and the Ukraine. . The presence of R-SRY10831b in western Siberia probably chronicles known migrations originating in the Altai and Sayan Mountains. The low frequency of this haplogroup in several Central and East Siberian populations is most likely due to admixture with recent migrants of European descent (Karafet et al. 2002:784). It does appear that Russian colonizers may ha ve spread a number of R1 lineages in East Siberia and the West ern Aleutians, and that Americans of Scandinavian descent may have helped spread R lineages among Alaska natives; R has a very broad distribution, but exhibits extensive s haring with European lineages based on Y-STR haplotypes (Zlojutro 2008:37; see also Zegura et al. 2004). However all R lineages exhibit extensive sharing with European line ages, even those from Bronze Age Central Asia and Siberia. The inability to dist inguish between European-derived and Asianderived R1a1 lineages is a persistent problem, considering that all these lineages share a common origin in the Bronze Age Russian steppes. No one has found a way to distinguish definitively bet ween those R1 lineages intr oduced during the Russian colonial period, and those introduced centurie s or millennia earlier under long-lived transcontinental equestrian nom adic confederacies like t he Xiongnu/Huns and others (Underhill et al. 2010). Until a large bank of high-resolution Y-chromosome sequences are available for study, the demographic hist ory of haplogroup R in the Americas will remain something of a myst ery. R1a is extremely co mmon among Uralic speakers, including Finno-Ugric and Samoyedic, both cons idered distant linguistic relatives of the 138

PAGE 139

Eskimo-Aleuts by Fortescue (1998). R 1a is also the second most common Y-DNA lineage (after Q*) among the Selkups, closely re lated (biologically, not linguistically) to the Kets (Tambets et al. 2004). Several recent ancient DNA studies suggest that the presence of haplogroup R1a1 was well established in Central Asia and South Siberia we ll before the Common Era, even long before the tenur e of the Xiongnu Empire. R 1a was nearly ubiquitous in parts of South Siberia by t he Middle Bronze Age at the latest. Ancient DNA has been examined from the Krasnoyarsk region of t he middle Yenisei River, where Yeniseian languages were likely spoken from 2500 years ago until the eighteenth century (Nikolaev 1989). A recent study found that the R1a1 lineage was the predominant, universal patrilineage in this region for the ent ire span of the the period from 4000 years ago until circa 500 CE (Keyser et al. 2009). T hese authors further suggest that these Bronze and Iron Age Scytho-Si berians and Yeniseian-Tagar were responsible for the early peopling of the Tarim Basin and the rise of the Xiongnu Empire in Northwestern China, providing genetic and archaeological support for the Xiongnu-Yeniseian linguistic hypothesis of Vovin (2000, 2003) and Pulleyblank (1962). Another recent study (Li et al. 2010) verifies the genetic continuity between Siberia and the Tarim Basin, finding that the oldest human remains ever found in the Tarim Basin, from a 4000-year-old cemetery, revealed an admixed population of exclusively R1a1 patrilineages, in the company of diverse East and West Eurasian mtDNA lineages. An older study (Keyser-Tracqui et al 2003) shows that, in agreement with the diversity of modern DNA in Mongolia (Kolman et al. 1996), a 2000-year-old Xiongnu necropolis in Mongolia includes all four ma jor American mtDNA haplogroups (A, B, C, 139

PAGE 140

and D) plus haplogroup M. Haplogroup M was recently also found in ancient American DNA (Malhi et al. 2007). Others further not e the surprising co-occurrence of Y-DNA haplogroups R1a1 and C3 within contemporaneous individuals in a 2000-year-old elite Xiongnu cemetery in Northeast Mongolia (Kim et al. 2010). This is near the Siberian provinces identified as the source for Athapa skan R1 lineages (Lell et al. 2002). Until a broad, high resolution survey of Asian and Athapaskan R1 lineages is able to prove recent Russian or American admixture, the Bronze Age Siberian s ource for prevalent Dene-Yeniseian R1 lineages rema ins a viable hypothesis. Native North American R1 lineages reveal a great amount of post-Columbian European male-mediated admixture. However, given t he incredibly vast geographic spread in association with Br onze Age pan-Eurasian empires, and the relatively youthful ~4000-year age of the R1* clade, there is presently no way of determining exactly what percentage of Native American R1 chromoso mes may have been intr oduced at the time of the sudden Late Holocene appearance of derived Asiatic phenotypes in the Americas (Gonzlez-Jos et al. 2008). Furthermore even if the Scandinavian source of Greenlandic Inuit R1a and R1* is borne out (B osch et al. 2003), there is no guarantee that all such admixture is post-Columbian. Scholars have largely neglected to consider whether or not the pre-Colu mbian Norse colonists admixed with Neo-Eskimos or PaleoEskimos (Agger and Maschner 2009; Suther land 2009). Thus, multiple different sources for haplogroup R1 among Native Am ericans appear credible in the light of present data. Sorting them out is a daunting task, but the issue must not be avoided. 140

PAGE 141

Summary In summary, the three major North Am erican Y-DNA haplogroups, like the five mtDNA haplogroups, each have complex, mult ilayered histories. Colin Renfrew (2010:136) writes that [u]ntil there is consensus about the initial colonisation of the Americas and its genetic imprint, it may be difficult to achieve a more effective archaeogenetic narrative for the remainder of the pre-Columbian period. This is not simply a question of building a consensus su rrounding the date of the earliest peopling of the continent, but a host of other ques tions about the number, timing, duration, direction and scale of movements between t he Old World and New World. The failure to achieve synthesis between linguistics, ethnology and genetics owes to a failure to consider all possible variables. Mulligan et al. write: researchers looked for consisten cy with more complicated hypotheses before trying to reject simp ler ones. . The flaw with this approach is that it places no practical li mit on the number of migrati ons that can be accepted for a data set. Moreover, when the sources of the proposed migrations are highly speculative, there is a danger of reverting to an earlier day of anthropological explanation . Nevertheless, linguis ts continue to research and debate the issue . but few geneticists have the expertise in linguistics to evaluate their debates (Mulligan et al. 2004:307). This admitted lack of linguistic expertise has not prevented molecular genetic studies from playing a hard science trump card over the less empirical subdisciplines for the last decade, even as molecular anthropology has barely emerged from its infancy. One is not a reasonable practical limit on the number of migrations, where back and forth movements of natives on opposite sided of Bering Strait have been observed with some frequency since record s were kept. The acceptance of DeneYeniseian language links by mo st linguists now appears to strongly favor at least two major migrations. As Galina Dzeniskevich notes with regard to the prevalence of Asian 141

PAGE 142

material culture among Athapaskan speaker s, If the narrow strait between the continents did not obstruct the diffusion processes two or three centuries ago, then it could hardly have been insurmountable in the fifth to tenth centuries, and earlier (Dzeniskevich 1994:59). While Scott and ORourke have produced a good summary of the genetic literature for genes across Beringi a, they conclude on a pessimistic note regarding the ability to connect Na-Dene genetically to the Ket: While linguists have developed a solid case linking the North American language family Na-Dene wit h Yeniseian, we have not found comparable parallels in the biology of these groups. There are several lines of evidence that point to central Siberi a as the ancestral homeland for some portion or even all of the Native American gene pool. However, there is no specific gene, haplogroup, or dental trai t that provides a direct link between the Kets and any Na-Dene speaking populati on. . At this time, we can infer that the ancestral populations of Na-Dene are linked to central Siberian and east Asian groups but can make no claim about their specific genetic affiliation with Kets (Scott and ORourke 2010:133). It may be true that the Ket, as the last surviving Ye niseians, are themselves imperfect biological corollaries for the once-large, multiethni c Siberian/ECA language family they represent. But linguistic fa milies often behave as fleeting waves on the surface of a deep ocean of genet ic continuituity. The hi stories revealed through the surface features of languages are different, but not better or worse, than those revealed through the deep sea of nucleo-tides. A paradi gm shift resulting in a synthesis of the three datasets of genes, languages and material culture must recognize and embrace their imperfect harmony. 142

PAGE 143

Figure 3-1. Tentatave (Sino?)-Dene-Yeniseian dendrogram (authors rendition). No time dimension is depicted, although the vertical axis is meant to imply a relative chronology. Authors proposal based on discussion in Vajda (2010b). Figure 3-2. Map of Yeniseian territory in eighteenth cent ury (green), and the Xiongnu Empire circa 2000 years ago (orange). Ethnographic Ket are the black dots in the northern part of Yeniseian territory. Courtesy Wikimedia commons. 143

PAGE 144

Figure 3-3. Apache umbilical pouch stitched with leather and horsehair, white shell and turquoise pendants, early twentieth-cent ury. Image courtesy of the Arizona Antique Centre, Scottsdale AZ. A. B. Figure 3-4. Comparison of forged daggers of Athapaskan Subarctic Canada and the Upper Yenisei River, Siberia. A) Seven inch Tahltan (or Gwichin) forged steel dagger with leather-wrapped grip; 1850 or earlier. Private collection, Discovering American Indian Art Exhibit, UTK, exhibited 29 Aug 2009 through 10 January 2010. Photograph by author. B) SHM 1669/1; Twelve inch forged iron dagger from the Tagar Culture, Southern Siberia, Krasnoyarsk Region, near Minusinsk, fifth century BCE Image B from courtesy of The State Hermitage Museum, St. Petersburg, Russia. 144

PAGE 145

Figure 3-5 Comparison of tanged-type forged me tal arrowheads of Eurasia and the Athapaskan Subarctic. The three on t he left are Altaicand Xiongnu-type forged iron arrowheads. The three on the right are Athapaskan tempered copper arrowheads. Authors sketch m ade from photographs in Franklin et al. (1981:28), Witthoft and Ey man (1969:21) and Karasulas (2004:23, 49). Figure 3-6. Native Siberians in the mid-twentieth-century. Cross-hatched areas colonized by Russians (Uinuk-Ool 2003:232); courtesy Wylie-Liss Inc. 145

PAGE 146

Figure 3-7. Athapaskan/Algic Interface in Northwest California (Dixon and Kroeber 1919). A. B. Figure 3-8. Na-Dene language distribut ion in comparison with Haplogroup C3 distribution. A) Na-Dene languages (shaded area). Authors sketch. B). Distribution of haplogroup C3b-P39 in No rth America (Greenland not shown). Red asterisk marks northeastern Ok lahoma, where Cherokee and Creek were relocated after 1835. Image B Modified from illustration by Mauricio Lucioni, courtesy Wikimedia commons 146

PAGE 147

Figure 3-9. Median-joining microsatellit e network for haplogroups CM217* and C3b (oval, far left) with the position of the P39 mutation denoted by a cross-hatch within the right-most oval. Wayuu indivi duals outside P39 cluster indicated by arrow. Haplotypes are coded in white, black, or gray by population system, with haplotype sharing indica ted by pie chart divisions (Zegura et al. 2004); by permission of Oxford University Press. 147

PAGE 148

Figure 3-10. Athapaskan and Southweste rn Y-chromosome haplogroup frequency distribution (Malhi et al. 2008) ; courtesy Wylie-Liss Inc. 148

PAGE 149

CHAPTER 4 SELECTION OF FIELD-SITES AND METHODS OF ANALYSIS Premise for Collections Research Overview One of the challenges of th is dissertation is to address the relationship between strong complex bow types used by nineteenthcentury Athapaskans in the three major geographic blocs (Northern, Southern and Pacific Coast), and to assess their relationship to each other and to thos e of neighboring populations. Geographic distribution of bow types may correlate to linguistic, genetic and ethnological data. Chapter 2 reviewed the histor y of the study of the Atha paskan expansion, from the perspective of the four s ubfields of anthropology (lingui stics, archaeology, ethnology, and human biology). Chapter 3 brought this pi cture up to date with a review of recent linguistic scholarship linking Athapaskans (Na-Dene) with Southwest Siberians (Yeniseians), exploring the immanent potential impact of this linguistic paradigm shift upon ethnology and mole cular genetics. In Chapter 4 I will lay out the basic me thodological and analytic considerations involved in the analysis of the material culture. First I will provide a background and description of the basic me thod employed. Then I will pr ovide an overview of the filedwork and institutions visited. Next I will describe in detail the manner of measuring and cataloging individual comp lex bow specimens. Finally I will assess the significance of these archery artifacts upon Athapaskan mate rial culture, and whether the spread of thre technology in question is a possible corollary to the Athapaskan expansion. Strong complex bows (generally term ed sinew-backed bows) are presumed to have diffused southward from Alaska to Ca lifornia and the Southwest during the late 149

PAGE 150

first or early second millennim CE (Blitz 1988; LeBlanc 1999:101). This timeframe is very similar to the presumed timeframe of the At hapaskan expansion into both of these regions, and a number of author s have suggested this is not a fortuitous coincidence, but a reflection of the actual transmissi on of the weapon in the hands of Athapaskan speakers (Baldwin 1997; Downs 1972:6; Krantz 1977:47-49; Paper 1993; Steward 1937:83-87). Assuming a predominantly north -to-south migrati on of Athapaskan speakers during the Common Era, then the material cultur e and technology of Northern Athapaskan societies may retain some characte ristics of the ancestr al state of similar items observed among the Athapa skan daughter cultures of the Apacheans and Pacific Coast Athapaskans. I am thus using a syn chronic, spatially-distributed dataset (nineteenth-century bow forms) as a proxy fo r a diachronic analog which is the sparse archaeological footprint of sinew-ba cked bows (and bows in general). Evolutionary Anthropology: Et hnogenetics versus Phylogenetics My method draws from those used in evolutionary archaeology, as noted by Hector Neff: Like other phenotypic characteristics, t he behaviors used to make artifacts at a particular place and time (and t herefore the artifacts themselves) can be explained by reference to history (descent, inheritance of information) and selective retention (Neff 1992:141). In other words, relationships of common descent promote selective retention of variation, and these patterns provide the basis for application of evolutionary theory to material culture. But the relationship between evolutio nary theory and anthropological archaeology has been uneasy at times, as debates surrounding so ciobiology and the cultural-processual orientation of mid-tw entieth-century scholarship have ebbed and flowed. OBrien and Holland write of the un-Darwinean nature of much of this theory: 150

PAGE 151

Evolution assumed a place in the new archaeology of the 1960s, but, similar to its status in anthropology in general, it was not Darwinian in nature. . Archaeologists scrambled to incorporate the trappings of cultural evolution into their arsenal, concer ning themselves with searching for archaeological correlates of ethnographi cally known forms of kinship, sociopolitical organization, and so on. Once these were found, archaeologists then attempted to understand how one form transformed itself into another form. Implicit in these exercises was the notion of an evolutionary progression from simple to complex . .in short, directional evolution (OBrien and Holland 1990:33). One major flaw in many nominally evol utionary approaches to archaeology in the last century was the overt desire to di scover general lawlike principles governing human behavior, and the haphazard, uncritical reappropriation of concepts and terms from physical and natural sciences (OB rien and Holland 1990:33; see also Kehoe 2010). OBrien and Hollands interest in specifically Darwinian analysis of material culture improves on the above by recognizing the explicitly non-directional nature of evolutionary change; such chang e may have adaptive significance, but not all features are adaptations. They continue: features that once were adaptations may not have been adaptations under different environmental regimes. And they conclude: simply because a feature evolved as an adaptation does not imply that the feature was an endall solution to a problem (OBrien and Holl and 1992:44). I am interested in the selection and retention of bow traits thr ough time and space, which may or may not be adaptive, but which nonetheless imply some form of descent through retention of inherited information. This is a cladistic (or phylogenet ic) model of evolution. Simultaneously, I am cognizant of J ohn Moores (1994) ethnogen etic critique of cladistic theory which notes that protolanguages and protocultures are complex entities with multiple interfaces involving neighbor ing societies, and that their daughter populations are never simply evolved forms of their parents, but hybrids between 151

PAGE 152

multiple antecedent cultures. In this sense, Darwinian evolution is an imperfect analogy because a direct linguistic ancestor can never be assumed to be the source for all variation within a material cultural or technological li neage. There are numerous opportunities for contact-induced change in mate rial culture (like any aspect of culture), and there are many aspects of Apachean and Pacific Coast Athapaskan societies which are demonstrably not a product of migration with the ret ention of features found among their Northern Athapaskan kin, despite their remarkable collective linguistic retention. The pedigree of these ideas about ethnogenesis goes back to Kroeber, who writes: Cultures are always tending to equat e themselves by imparting their characteristics to one another, even whil e another set of impulses pushes each of them toward particularistic pecu liarity. . [T]he course of organic evolution can be portrayed properly as a tr ee of life, as Darwin has called it, with trunk, limbs, branches, and twigs. The course of development of human culture in history cannot be so described, even metaphorically. There is a constant branching-out, but the branches also grow together again, wholly or partially, all the time Culture diverges, but it syncretizes and anastomoses too. . A branc h on the tree of life may approach another branch; it will not normally coale sce with it. The tree of culture, on the contrary, is a ramification of su ch coalescences, assimilations, or acculturations (Kroeber 1948:260-261). Kroebers illustrations of the tree of life and the tree of culture are reproduced in Figure 4-1. This distincti on between biological and cultural evolution made here is useful. The tree of life illustrates the phy logenetic (cladistic) model appropriate for the natural sciences. The tree of culture illustrates the ethnoge netic (reticular) model most appropriate for the social sci ences and history. These tw o models are also called vertical and horizontal (Mulder et al. 2006) But it is important not to overstate the difference between the two branching patterns. Kroebers reticular network is a good model of cultural change, but it can also describe biolog ical processesadmixture and hybridization are reticular mer gers of biological traits coalescing in ethnogenetic fashion. 152

PAGE 153

Likewise, the more typically treelike model of biologic al evolution may be appropriate for instances of culture change where isolation in time and space prevents ethnogenetic coalescence. For example, in the ca se of long-distance mass migration and transmission of novel culture traits from elsewhere, the phylogenetic model may still have merit in cultural setti ngs. This is especially the case with large-scale ethnolinguistic expansion events as in the Athapaskan case. Peter Bellwood notes: Reticulate models . stress the impor tance of continuing processes of interaction between contemporary comm unities. Phylogenetic models . imply dispersal by culturally and li nguistically related populations from common origins. . Both are nece ssary: reticulate models . can overlook the large-scale patterning visible on continental and millennial scales in the linguistic and archaeological records. . Largescale and fairly integrated colonizations did happen in prehistory; human cultures and languages can, to varying degrees . be organized in phylogenetic arrays. The generation of human diversity in the past has not been entirely reticulate and dependent on processes of in situ interaction between peoples of different ethnol inguistic background. Neit her has it been entirely radiative and dependent on adaptation in isolation (Bellwood 1996:881, 888; emphasis original). Both kinds of analysis are relevant in the case of material culture and the Athapaskan expansion, but the question becomes how best to juxtapose and articulate them. In seeking to bridge this same divide, Kenneth Weiss and Frances Hayashida state [w]e have wonder ed if there might be an alternative where differences in training and theoretical outlook lead to vision ra ther than division (Weiss and Hayashida 2002:141). As others have sugg ested, to successfully integrate cladistic and ethnogenetic models we must be more creative in our use of data (Mulder et al. 2006:63). Mulder et al. continue (using horizontal to mean reticlular and vertical to mean cladistic): To date, only geographical proximity is used as a measure of the potential for horizontal transmission. If we are to uncover past opportunities for cultural diffusion, we will need other sources of evidence for historical 153

PAGE 154

interconnectedness among societies, such as trade, loan words, and trees modified to show intergroup contact histories (Mulder et al. 2006:63). In short, both reticular and cladistic model s are applicable to both biological and cultural settings. Mass movem ents of (multiethnic) societies from north to south took place during the middle Common Era, and among these, Athapaskan speakers were most prominent. During roughly the same peri od, the south to north movement of UtoAztecan speakers during the Numic Expansion also took place (Miller 1986:100-104). Despite the aforementioned cave ats, the careful analysis of spatial variation in bow morphology may help reveal cultural curr ents underlying the physical transmission of these items in the hands of mobile cultur e bearers, particularly Athapaskan speakers. Athapaskan speakers are t he group with the widest documented presence in multiple areas of the conti nent where this particular weapon was used, from the Bering Sea to Sonora; from Hudsons Bay to Hu mboldt County. Athapaskans are thus plausible candidates for the rapid transmissi on of complex archery between circa 1150 and 1400 CE. But all sinew-backed bows (not just put atively Athapaskan ones) are relevant to understanding this process, as this rapid technological diffusion was never linguistically bounded; it must have encompass ed multiple non-Athapaskan societies of North America who were likewise involved in its transmission. Basic Method Employed My method was to map the variation in weapon design th roughout a geographic zone bounded on most sides by historical Athapaskan-speaking communities. Where an in-person visit to a particularly import ant ethnographic collection was impossible or impractical, I was able to gain some relev ant artifact descriptions via correspondence with museum curators or si mply through access to online collections databases. Where 154

PAGE 155

significant gaps in the record occurred due to lack of preserved materials (especially in British Columbia), I was compelled to supplement the meager record with less-thanstandardized bow-descriptions cu lled from ethnogr aphic literature. The result is a comprehensive dataset for the sinew backed bow in western North America. Sample Size and Number of Institutions Between November 2009 and August 2010, I made periodic scheduled research visits to twelve museums and one large privat e collection in ten different states. I examined directly more than 250 Native North American bows in addition to an assortment of various ot her ethnographic artifacts. I measured and photographed 236 of these weapons (some were excluded due to inadequate provenance and/or geographic irrelevance). This large and heterogeneous dataset was supplemented using artifact descriptions from various online databases, also those generously provided to me by museum professionals in the US and Canada, and finally those culled from an extensive search of the lit erature. The addition of these extraneous sources means that a total of 242 bows are completely documented to the extent possible, with approximately 50 additional specimens in varying degrees of completeness in terms of measurement (thi s number includes only weapons not directly examined by me). Weapons incompletely m easured are still useful for comparison and reference (especially for rarer types) but ca nnot be systematically included in my tables and analyses. Russ Bernard writes: all samp les are representative of something. The trick is to make them representative of what you want them to be (Bernard 1995:96; emphasis original). From the outset, I sought to make my dataset repr esentative of something, in this case the contact-traditional-era co mplex bowyery of Athapaskan societies. 155

PAGE 156

Contact-traditional refers to the period (variable depending on location), after European contact but prior to the most extreme a ssimilation and acculturation, when material culture was still somewhat reflective of pre-contact norms. Many Athapaskan languages possess terms derived from the Proto-Athapaskan root word, *ts -t for the sinew-backing technology involved (Victor Golla, personal communication 7 November 2009). The only way to make sure that the field dataset is legitimately reflective of historical context (to the extent possible) is through adequate unbiased sample size. One could make the case that the best sampling economy could be achieved by selecting the largest, riches t and most diverse collection and just camping out there for a few weeks to exhaustively document all the data r epresented. However this would be analogous to an ethnographer atte mpting to capture the character of a society by surveying only the admittedly di verse population of one of its major city centers. While the indivi duals represented in such a survey would indeed include a broad spectrum of the society at large, t hey would likely also possess subtle unifying characteristics common which would distingu ish them from the folks who were not present in the urban hub. As Bernard further writes: There is always going to be a tra de-off between greater accuracy and greater economy in sampling. In a study of households in a county, you should take a few households from each community (cluster) rather than study many households in a few randomly chosen communities. The problem is that this may force you to spend more in both time and money on travel than your budget will allow. So your rule actually becomes: Study all the highest level clusters y ou can afford (Bernard 1995:74-75). If I had limited myself to the single highest-le vel cluster in my sample (that is the National Museum of the Americ an Indian) and if I had allowed myself two weeks to work there (rather than the actual two days), then I would have gathered a very large dataset in a relatively short amount of time. Bu t one large sample from a single high-level 156

PAGE 157

cluster would not be sufficient, because the data would then repr esent the biases inherent in that particular co llection. The character and co mposition of even large and diverse collections (like cities or towns) are shaped by the interests of particular founding collectors and/or the hi story and ethos of the people an d institutions involved. Particular types of objects may be covet ed in one time or place, and overlooked or dismissed in another. Ever y collection I visited had unique character traits and attributes distinguishing it from others like it. Some co llections may disproportionately include niche items satisfying collectors per sonal interests; others may have aspired to geographical diversity and comprehensive r epresentation, but overlooked mundane or commonplace items,while many others appear to have been amalgamated by convenience, through a variety of haphazard sources. In particular, the most comprehensive and diverse collection (NMAI) has no Northern Athapaskan complex bows whatsoever; these crucia l data (for my purposes) would have been virtually absent from my study if I had limited my scope to just a few of the most obvious major museums. So me museums (through correspondence or through online database access) did provide me with a clear idea of what to expect before I arrived. But when such pre-screeni ng data are limited (as it was to varying degrees for a number of my target institutions), then research visits are hit or miss. It requires a bit of faith to soldier on when an ex pensive research trip yields little useful data. Yet such visits may be fortuitous as I later found out. The only place where I found more than one Alaskan Athapaskan comple x bow in the same institution was at the San Diego Museum of Man, where I had absolutely no idea what to expect before I visited in person. 157

PAGE 158

By visiting as many museums as I c ould afford to between November 2009 and August 2010, I was able to get a much more representative view of Athapaskan archery technology, and Native American complex ar chery in general, than if I had taken the more economical route (both in terms of time and money) and examined a few higherlevel clusters exclusively. In another illustration of this point, my Apachean sinewbacked bow dataset is very large (about 67 bows). This includes bows from almost every commonly recognized Apachean subgr oup, including (in descending order of prevalence) Navajo, Jicarilla, Mescale ro, Western Apache (White Mountain), Chiricahua, and Plains Apache (Kiowa-A pache). The only Apachean people with sinew-backed bows not represented in my dataset are the Lipan Apache from the southernmost Texas Plains, a group whose population in 1894 was estimated at only forty individuals and was thus by far the sm allest Athapaskan subgroup recorded in that study (Morice 1894:16). No single museum possessed bows from all six of these different Apachean cultural unitsit was nece ssary to visit more than one museum to see all of them. The sample is less of a selection of a larger population than it is a snapshot of the whole population itself. T he study includes nearly every bow (of a certain type) available for examination at eac h museum I visited, with a few additional specimens culled from other sources. Ethnographic and Archaeological Mate rials in Museum Collections The vast majority of the artifact s in my database are from ethnographic collections of the late nineteenthor early twentieth-centu ries (the reservation period) owing to the fact that the obj ects of primary interest to me (sinew-backed bows) are organic and therefore not well preserved in the archaeological record. Ethnographic materials thus serve as admittedly imperfe ct proxies for nonexistent archaeological 158

PAGE 159

remains. One should be highly critical of any attempt to use modern samples to test hypotheses about ancient migrations, but it is certainly no less reasonable than the common practice of molecular anthropologists using DNA from contemporary Native Americans from as far away as Alaska and Tierra del Fuego, in order to make inferences about the Pleistoc ene population of a sm all strip of submerged land in the Bering Sea (Tamm et al. 2007). Most of the bows I analyzed were collected from living informants in the ninetheenth century or later. A sma ll number of archaeological specimens are nonetheless represented in my dat a. The number of relevant artifacts has varied substantially from institution to in stitution. For a comb ination of reasons, some visits allowed more leisurely perusal of collections, while other visits were more efficient and selective by necessity. The dec ision to examine certain objects and not others because of time constraints was very difficult. Hindsight notwithstanding, the data collected are sufficient to address my primary research question about complex archery as a unifying feature of Athapaskan material cultures in three different culture areas. Strong Complex Bow Documentation and Measurement The primary motivation for my study is to te st the hypothesis that the diffusion of the sinew-backed bow in western interior North America occurred in correlation with a late-prehistoric Asiatic derivat ion of proto-Athapaskan culture. To bridge the temporal gap from late prehistory to the ethnographic past, it will be necessary to demonstrate a distinctive physical correspondence (indicati ng a likely design-template connection) between Athapaskan weapons systems in widely removed districts. Assuming that such correspondences are found to be largely limited to Athapaskans, it will substantiate the hypothesis that Athapaskan migrants retai ned technologies originating 159

PAGE 160

in their ancestral homeland(s). My method thus represents an extension, to the realm of material culture, of the method of ethnographic reconstr uction as employed by Perry (1983). As the key material cultural element in this hypothesis, the sinew backed bow is the primary focus; nevertheless, other material cultural objects bearing on questions of Athapaskan origins and migrat ions are relevant, and these objects (e.g. footgear, basketry, ceramics, edge tools etc.) will figure in my analysis, albeit in a less systematic and more qualitative manner. The first step taken when examining a sinew-backed bow is to note the catalog number and any relevant information about its provenance (e.g., culture and place of origin, date of acquisition, donor, etc.). T he quality of information present for any given object varies substantially, from extensiv ely documented professional collections of famous anthropologists like Cornelius Osgood, James Teit, and Edwa rd Sapir, to the more idiosyncratic donations of various ob scure philanthropists from their eclectic private collections, often having few if any provenance notes other than the donors name and date of acquisition. Some of these items are grossly inaccurately cataloged; such as, a California bow at NMAI that was attributed to arctic Alaska, and a Plains or Southweststyle weapon at Ya le that was attributed to the Florida Seminole. The next step is to note any unusual or dist inguishing features. These include (for example) materials layered ov er or under the backing, the co lors and pattern of paint or other decoration, leather or textile materials wr apped around the grip, attached ceremonial objects, engravings in the wood, etc. Then I not e the general shape of the bow, its component raw materials, the conditi on of the string (if present), and the crosssectional shape of the stave at center midlimb, and tip. I developed my own 160

PAGE 161

nomenclature for expressing these formal attr ibutes (Figures 4-2 and 4-3). No two authorities agree on the classification of bows (Hamilton 1970:45), and therefore (unfortunately) the use of bow-terminology in the literature is extremely imprecise. I use the most straightforward nomenclatu re possible, following Downs (1972) in using the less-common term strong complex bow in place of the more common sinewbacked bow or composite bow. The use of the term composite bow is especially problematic and imprecise. Driver and Massey (1957) use it to refer to all sinew-backed bows, while Hamilton (1970) uses it for a narrow subset of sinew-backed bows which are made with at least thr ee distinct stave components, otherwise referred to as compound bows. The term sinew backed bow is more satisfactory, except that elastic backing materials are not limited to sinew. Babiche (i.e. rawhide or gut thongs) may be used instead, as well as other fibers (e.g. st rings and yarns of various types) all of which appear occasionally on Arctic bows and even more rarely on Athapaskan bows. Oldworld composite bows may also be backed with a flexible wood like rattan, so using the term sinew-backed to refer to the entire technological class can be misleading. The term strong complex bow may lack the general literary currency of these other terms, but at least it is accurate in its application to all of the cases under consideration. The term self bow is much less problematic, inva riably referring to bows that are simple, that is lacking longitudinal backing or other composite reinforcements. Basic Bow Features and Vocabulary I will now list the major anatomical featur es of the strong complex bow as I understand them, remember ing that there may be variation in the literature as far as how these terms are used. In addition, th is list includes other terms and nomenclature relevant to traditional bowmaking. 161

PAGE 162

BABICHE. Conditioned animal-based cordage, most often hide, gut, or sinew, often used to make bowstrings or reinforcement for the back of a complex weapon. In the present document, it refers only to materials other than sinew, which has a distinctly fibrous appearance compared to other forms of babi che. When I use babiche, I refer to hide or gut. BALEEN. Whalebone, colloquialy; a horn-like s ubstance from the jaws of filterfeeding whales, often used as a substitute for horn in true composite bows. BACK. The surface facing away from the archer when the weapon is used. BACKING. The elastic material used to reinforc e the back, usually sinew or babiche, rarely other fibers. BELLY. The surface facing toward the arc her when the bow is used, often reinforced with compression-resistant hor n or bone lathes in true compound bows. BOWSTRING. The cord which acts as a spri ng in conjunction with bow. Sinew or babiche are preferable because they provide some elasticity, although inelastic plant cordage is serviceable where resources are limited. BOWYER. One who manufactures bows, i.e. a bowmaker. BOWYERY. The bowyers craft, i.e. bowmaking. BRACE. To string the bow (verb), not to be confused with brace (noun) which is another word for brac er, i.e. wrist guard. COMPLEX BOW. A broad class of bow including virt ually any non-self bow, principally any bow with some sort of longitudinal re inforcement of the back (e.g., composite bows, compound bows, sinew-backed bows and horn bows). COMPOSITE BOW. In the most specific usage, a t rue composite bow is a complex bow featuring both an elastic backing materi al (usually sinew or babiche) and a belly reinforcing material (usually horn, ivory, bone, or baleen). An elementary composite bow lacks the belly reinforcement (i.e. a sinew-backed bow). In general usage, composite bow is a synonym for any complex bow. COMPOUND BOW. In traditional usage, a bow with multiple stave-segments spliced together. Not to be confused with cont emporary usage referring to a modern bow that uses a levering system (usually of cables and pulleys) to bend the limbs. GRIP. The region near the center of the bow, often covered with textured wrappings, where the archers non-dom inant hand grips the bow during use. 162

PAGE 163

HORN BOW. A complex bow similar to a true composite bow, only lacking the wooden core/stave. The bow itself is made of horn or ivory, et c. Horn bows are often compound bows. LIMB. The portion of the bow between the grip and the nock. NOCK (BOW). The platform-like incision or projec tion near each tip of the bow, which helps to grip the bowstring when the bow is braced. Not to be confused with the nock of an arrow, an incision on the back of the arrow gripping the bowstring prior to release. RECURVE. Outward bend of the limb tips typical of reflexed bows. REFLEX. The condition of a bow that is braced when bent in the direction counter to the natural curvature of the unbraced bow. A substantial majority of complex (backed) bows are reflexed and/or recurved. SELF BOW. The simplest and most common form of bow, made with a single-piece wooden stave with no longitudinal reinforc ements; also known as a simple bow. SINEW-BACKED BOW. The major sub-type of complex bow found in North America, consisting of a bow-stave reinforced with sinew applied longitudinally to the back surface; also known as an elementary composite bow. SIYAH. Tip extension on some bows, serving to arrest the moti on of the bowstring after release of the arrow, resulting in a more uniform release of the bows tension. Siyahs are often vestigial (nonfunctioning) in North America. STAVE. The term for the selected and prepared material (timber or occasionally horn) used for the bow itself, in its rough st ate, or the term fo r the wooden core of the bow in the case of true composite bows. TIP. The portion of the bow betw een the nock and the terminus. WRIST GUARD. A brace; an object to protect the archers wrist (the one holding the bow) from the abrasive percussion of the bo wstring as the arrow is released. The wrist guard may be a curved plate with str aps worn on the wrist like a bracelet (Eng. bracelet = small brace), or it may be a string-stopping wedge-shaped device attached to the bow itself, as it t he case with many Alaskan Athapaskan bows. Wood Type The wood-type of the bow itself is not ed only infrequently (at best) in museum catalogs, except where a single species pre dominates within a given culture-area (e.g., 163

PAGE 164

yew in California). The unique cellular structure of a species of wood can be determined with the use of a hand-magnifying lens and the appropriate field-guide for the wood-anatomy of t he species in question. While this information would be most welcome, the necessarily exacting dete rmination would slow documentation efforts considerably under circumstances where time is a primary constraint. Furthermore, I am not trained to identify specific wood-types orig inating from regions as far-flung as Alaska and Arizona. I am able to make educated in ferences about the woods most likely used in different regions based on my close reading of the relev ant literature, and this must suffice. Although material constraints may have a significant impact on the form of the bow itself (and are therefore rele vant to my analysis), I am le ss interested in raw material analysis for its own sake, because (as is the case with ceramic analysis) locally sourced raw materials are less relevant to the questi on of ultimate origins for a technology, where long-distance transmission is suspect ed. Bows made in different regions vary widely in their locally-derived raw materials, but may nonetheless de rive from a single antecedent concept or template which is adapt ed to local circumstances. I am primarily interested in long-distance relationships an d the conceptual undercu rrents of technology which may reveal them; variation in loca lly available materials is of secondary importance. Units of Measure and Measurement Protocols The decision to use English units (inches) rather than metric units (centimeters) was made ad hoc because I began my fieldwork with a cloth tape which used only inches. Subsequently, I came to prefer in ches for several reasons, and elected to continue using them. The classical anthr opological literature (and contemporary 164

PAGE 165

literature by American archery enthusiasts) almost invari ably uses English units to describe bows (Allely and Hamm 2002; Ma son 1894; Murdoch 1885; Osgood 1937; Pope 1918; Rausing 1967). Also, many bow s seem to be naturally well suited to measurement English units. For example, the vast majority of Pacific Coast Athapaskan bows I measured fall within the range of 30 to 40 inches long, and a similar majority of Apachean bows are within the range of 40 to 50 inches long. Certain other measurements (like width at grip and at tip) also seem to l end themselves to inches and fractions of inches. My initial assumption was that the bow length would vary significantly depending on its state of bracing and resu ltant curvature (that is, whet her it was measured along its curvature or along the taut bowstring). But I quickly realized this was only the case for the most heavily curved Pacific Coast weapons. The vast majority of bows vary by less than 5% depending on how the m easurement is taken. Fo r most subsequent bows, I elected for the most straightforward rout e (a tip-to-tip meas urement ignoring the curvature of the wood), with the caveat that this measurem ent slightly underrepresented the true length of the stav e for moderately curved braced bows. Severely curved braced California bows still required a measur ement taken along the curvature of the bow, because the shorter distance can be f ound to be more than 10% shorter than the actual stave length. I elected not to re cord string-length as a separate measurement, because so many bows in museums have no st ring (or a fragmentary string) that a consistent measurement could not be taken. After measuring the overall length of a bow, I then would measure width and thickness at three locations, near the grip (center), near the midlimb (midway between 165

PAGE 166

the center and the nock), and ne ar the end (just below the no ck). The next step was to examine and classify the nocks themselves, that is the platform-like incision or projection (i.e. concave vs. convex nock), usually about one inch from each tip, where the string is braced (Figure 4-4). Nocks are often incise d in the sides and/or back (never the belly) of the bow. They may also be built-up with a colla r-like piece of sinew or babiche around the whole circ umference of the tip, fre quently secured in place with the longitudinal backing (us ually underneath the backing) and/or through transverse application of a sinew reinforcement strips Indeed some authors refer to such additions as shoulder nocks (Allely and Hamm 2002:201). I prefer the term I have coined, collar nock, because it emphasizes that the nock is a 360-degree loop (on all sides) rather than a shoulder which is a projection from just the sides (not the front or back). In the Plains, Great Basin, and Southwes t, nocks may be virtually absent in some cases, with the bowstring simply looped ti ghtly around the contracting tip of the bowlimb. The nocks may be very subtle or obscured by the sinew wrappings, and individual bows may have idiosyncratic variati on in nock form, or si milarities to more than one common pattern. In nockless weapons, heavy usage may result in a groovedimpression of the bowstring in the tip of the bow, wher e a normal incised nock would appear. Hence it is difficult to establish discr ete types in many locales. Standardization of nock-shape appears much greater in some r egions than in others. I finally make a rough accounting of the number, size and general distribution of any extant transverse reinforcements (usually sinew wrappings) whic h serve the dual function of strengthening the bow itself (especially when placed ar ound an imperfection in the wood) and also help to secure the sinew backing or sculpted nocks. 166

PAGE 167

Bow Typology It is usually straightforward to make the most basic typological classifications at this point. The North American sinew-backed bow falls in two major types. In what I term the glue-type, the elasti c backing is directly molded while moist and pliable so to adhere to the back of the bow in longitudinal strips, by means of lacquer-like adhesive (usually or fish-glue, animal-glue or pitc h). This type is exclusively associated with American Indians, and is also essentially the same as the backing method for East Asian and Central Asian composite bows. The glueless-type, in contrast, is made by plaiting and wrapping dry sinew cables ont o the bow with the bulk of the cordage running longitudinally along the back, held in pl ace by a few transverse lashings. This type is mostly associated with Eskimo-Aleut speakers from Siberia to Greenland. Athapaskans (both Northern and Southern) are the only group s known to possess both the glue-type and the glueless-type of w eapon (Driver and Massey 1957:355). See Figure 4-5 for a map of the distribution of the two complex bow types, overlaid with a map of the Na-Dene languages. A rudimentary analysis can classify most complex bows according to a basic typology established in the literature. The glue-type weapon is also called the closebacked or the sinew-lined bow ; the glueless type is also called the free-backed, the cable-backed, trussed or sinew-c orded bow (Longman and Walrond 1894:47; Mason 1894:643; Wissler 1914:456). The glueless-type is further divided into at least four subtypes, with two simpler subtypes f ound in south Alaska and Cumberland Gulf (southeast Baffin Island) respectively. Murd och places these forms closer to the prototype, whereas two alleg edly more specialized or derived varieties are found in the High Arctic and Bering Strait (Murdoch 1885) Hamilton (1970:51) disagrees with 167

PAGE 168

aspects of Murdochs diachronic typology, instead regarding the Cu mberland Gulf type as another more specialized (i.e. derived) bow whose crude appearance is dictated by extreme material constraints of the region, rather than mere archaism. It is fairly straightforward to classify most Eskimo and Aleut sinewbacked bows according to the Murdoch classificatory scheme, but some of the Athapaskan glueless bows I have examined for this study deviate somew hat from this scheme, and as appearing independent derivations from the generic south Alaskan protoform. Through my analysis, I will attempt to determine if t here is cause to define a new typological category (type or subtype) for At hapaskan glueless strong complex bows. The glue-type bows fall into three approx imate categories, (1) a shorter, wider type common to the Pacific Coast Athapaskans and their neigbors, (2) a longer, narrower type common to the Great Basin, Pl ains, and Southwest, including Southern Athapaskans (3) the very short, narrow hornbows of the Plateau and Plains, which (like many Arctic bows) is correctly classified as a true compound bow, because it includes multiple pieces of varying materials, wit h spliced lathes of horn or baleen (a.k.a. whalebone) often riveted and bound together with an abundance of sinew. This type may or may not have a wooden core at all. Unfortunately, pres ervation of glue-type weapons in the Subarctic is too sparse to allow generalization, making it difficult to precisely categorize the Northern Athapaskan glue-type sinew-backed bows documented by nineteenth-century ethnographers. Through the use of categorical substitute s for continuous variables measured in the field, I intend to find the approximat e position of Northern Athapaskan glue-type strong complex bows within t he East-West (Pacific-Plain s) continuum. The late 168

PAGE 169

nineteenthand early twentieth-century sources documenting Northern Athapaskan glue-type strong complex bows include a few rough measurements and sketches of unique types, including some specimens with unusual features. These will be analyzed to the extent possible, substituting extrapol ations for the first hand measurements actual weapons themselves which were unfortunately not collected or pres erved by any of the authors who described them. The most common Northern Athapaskan bow type in ethnographic collections is a specialized type of long self bow, most commonly made of birch and featuring an attached wrist guard. This weapon is we ll studied by Foley Benson (1975). Because some South Alaskan complex bows appear to be hybrid types featuring Athapaskanstyle attached wrist guards and Eskimo-style glueless sinew backings, the Athapaskan self bows I observed will also be analyzed to determine their possible morphological continuity with the hybrid weapons, in t he effort to determine w hether the hybrids are themselves indebted more to either parent type. Museums Visited Below is a brief chronological summary of each of the collections I have visited and on what dates, followed in turn by a list of the institutions I was unable to physically visit, but which nonetheless provided data vi a the World Wide Web and/or through correspondence with helpful staff members. Table 4-1 is an alphabetized digest of this information, along with those museums that I di d not visit in person. My selection of institutions was made on the basis of a number of flexible criteria, but travel cost was the main limiting factor. For a collection to be worth a long and expensive journey there should ideally have been a signi ficant quantity of sinew-back ed bows, including those of Athapasakan manufacture. While there were a few minor disappointments (that is, 169

PAGE 170

collections yielding less than anticipated or other logistical setbacks) the vast majority of my fieldtrips were eminently worthw hile, yielding ample data for analysis. Wake Forest University, Museum of Anthropology (WFU) On November 19th, 2009, I visited t he anthropology museum on the campus of Wake Forest University in Winston Salem No rth Carolina. It is conveniently located near my family home. This collection po ssesses a number of relevant items of Athapaskan material culture, including s outhern Plains footgear, a Navajo-Pueblo arrowhead, Navajo-Ute baskets, and (reputed) Northern Athapaskan baskets. Florida Museum of Natural History (UFL) On December 10th, 2009, I visited the Florida Museum of Natural History in Gainesville, Florida. Most of the relevant objects are part of the Pearsall Collection of American Indian Art, amassed by private co llector Leigh Morgan Pearsall (b. 1874, d. 1964) and bequeathed to the University of Florida upon his death. The Pearsall Collection is the largest ethnographic coll ection in the southeast, and most of the objects were made between 1890 and 1920. Ther e are five sinew-backed bows in this museum. Two of these are the glueless type from the Arctic, and one of these is a three-piece compound bow. The other three complex bows are of the glue-type, from the Plains and/or Southwest. Finally, there are three self-bows of the type made by California Athapaskans. In addition there are some fine examples of Plains footgear, and a large collection of ornate basketry in cluding specimens from both Northern and Southern Athapaskans. Frank H. McClung Museum (UTK) I visited the museum on the campus of t he University of Tennessee, Knoxville, on December 15th and 16th 2009, where I viewed the McClung Museums extensive 170

PAGE 171

collections, including Navajo and Apache baskets, garments, footgear, and arrows. I went during the Discovering American Indian Ar t Exhibition, featuri ng a large number of objects on loan from a local private collecti on. This particular collection strongly emphasizes the Western Subarctic cultur e area of North Amer ica, and Northern Athapaskan material culture in particular. Un iversity representatives put me in contact with the anonymous collectors who generously allowed me acce ss to the bulk of their collection including a large quantity of Athapaskan textiles and utensils. Sam Noble Oklahoma Museum of Natural History (SNO) On February 25th 2010, I visited the natural histor y museum on the Campus of the University of Oklahoma, Norman. The muse um holds about two dozen American Indian bows, many from the Great Plains, and in cluding at least two Apache bows. I anticipated that some of t hese weapons would be sinew-backed, but upon arrival it turned out that they are all self bows, many being apparent tourist items made in the twentieth-century. I docum ented the most relevant se lf-bow specimens for my database. This museum has a significant quantity of footgear and basketry from the Plains and Southwest. Gilcrease Museum of the Americas (GMA) I visited Gilcrease, the art and history mu seum affiliated with the University of Tulsa, on February 26th 2010. This museum has dozens of Native American bows from all over the country, and very fine and ol d sinew-backed bows from the Southwest, Plains and Arctic are prom inent among them. I documented more than two dozen bows, which were as many bows as I could in the short time I had available (I drove from Norman to Tulsa on the same day). The lack of computerized database access prevented a thorough inventory of Gilcrease materials; each items provenance had to 171

PAGE 172

be looked up in an old card catalog. At the end of the trip, I believe I documented virtually all of the normal glue type sinewbacked bows that the museum contains, but I was unable to view any horn-bows or compo und bows (which were stored in a different room), or to view the entire exte nt of their collection. I looked at most (if not all) of the Arctic materials, and measured most of them, but was unable to obtain complete measurements for every Arctic bow due to time constraints. Without a digital database system, accurate assessment of the scope of t he collection is currently impossible. National Museum of the American Indian (NMAI) On April 8th and 9th 2010, I visited the collections of the Smithsonia n Institutions flagship museum for North American ethnology, in storage at the Cultural Resources Support Center in Suitland, Maryland. Two months advance notice is required for visits to this facility, and this was t he first collection I visited where the volume of material is so massive as to preclude a thorough examination of all relevant materials; unfortunately, I was not able to see everything I wanted to see there. I had to prioritize, first examining Athapaskan and Alaskan sinew-backed bows, with the intention of getting to Plains and Great Basin materials late r, time permitting. I did not accurately estimate the amount of work I need to do in order to document this collection, and I had curators pull more items from the shelves than I could handle in the allotted time. Smithsonian security protocols are time consuming; entrance and exit procedures at the facility are slow, and the lunch break is mandatory, meaning that the actual workday is between six and seven hours. Work was also slowed somewhat by my unbridled awe of the stupendous coll ectionthe temptation to talk shop and peruse the copious shelves with the very knowledgeable staff wa s ever present. 172

PAGE 173

As the final straw, I was forced to spend about half of the second day on lockdown in the basement bomb shelter when several armed fugitives evading police pursuit fled their captors into the facilitys campus, where they remained at large for several hours (Basch 2010). The final tally of bows fully documented at NMAI was only thirty-seven, mostly Apache and Eskimo weaponsfar fewer than I hoped for this important trip and a non-repr esentative sample of the Mu seums bow collections. A dozen or so additional weapons were noted and/or photographed but not fully documented. If I had it to do over again ( with advanced notice of the unplanned work stoppage), I would have prioritized documentation differently so as to include a handful of items with better-known provenances that I was unable to include, at the expense of some other items which I did record for which provenances are somewhat vague. Another trip in the future is merited, but requires several months advance notice. I was unable to schedule an additional visi t during the 2010-11 academic year. American Museum of Natural History (AMNH) I visited the American Museum of Natu ral History in New York on April 22nd 2010, and I succeeded in documenting most of the available glue-type sinew-backed bows in their collection, including a large number of important Apache weapons. I fully documented thirty-six weapons, or nearly as m any in just one day in New York as I did in two days in Washington D.C. The Apac he bows were notable for having among the most detailed and diverse provenances possible, often noting the specific Apache tribe of origin including rare examples of documented Mescalero and Plains Apache sinewbacked bows. Also, this collection has (to my knowledge) the worlds only surviving examples of glue-type sinew backed bows from inland British Columb ia, collected from the Thompson Salish (and/or t heir Nicola Athapaskan allies) by James Teit in 1897, and 173

PAGE 174

documented in the Memoirs of the museum (Teit 1900:239-241) Many objects in the museum were collected by reputable turn-o f-the-century ethnograph ers like James Teit and Pliny Earle Goddard; hence t heir firm documentation. The only minor disappointment at this mu seum is that some of the very finest weapons are on permanent exhibit, and are unavailable for even basic of measurements. In this ca tegory are at least two rare Canadian sinew-backed bows, one published by Teit from the Thompson, and another being a rare and unpublished Blackfeet (Blood band) snakeskin-cover ed bow, AMNH 50.1/1241A, similar to the type in vogue among Athapaskans to the west. And I unfortunately did no t have the time to fully document all of the No rthern Athapaskan self-bows, such as the one published by Goddard (1916:219). Assistant Curators Laila Williamson and Elise Alexander were especially helpful in emailing me thorough measurements for another Blackfeet bow which was not made available to me at the time of my visit. Peabody Museum of Natural History-Yale (YPM) On April 23rd 2010, I visited another particularly rich collection, this one in the Natural History museum of Yales New Haven Connecticut campus. The building is a very old one, with inadequate storage facilitie s, meaning that the objects are spread around somewhat haphazardly in different loca tions on the premise, with some bows stored in a different building and therefore inaccessible. I did my best to document as many bows as possible in the time allott ed, given that retrie ving items from the complicated multi-floor storage system was a time-consuming ta sk in and of itself. The nineteen bows I did get to examine included most (if not all) of the glue-type sinewbacked bows in the collection which we re not on permanent display or housed in another building. 174

PAGE 175

Founded in 1866, Yales museum is one of the oldest museums of its kind in the world, meaning that this collection incl udes some very old specimens whose provenances are quite murky due to their co llection by amateurs in the era before scientific ethnography was the norm. Severa l alleged provenances required correction or amendment based on comparative analysis, due to blatant inaccuracy. I did also examine a handful of Yales Arctic w eapons, but I was unable to be more thorough under the circumstances. I am especially in debted to collections manager Roger Colten for subsequently taking photographs for me of a rare Northern Athapaskan sinew backed bow, YPM 15844, which was in a different storage facility at the time of my visit. Detailed measurements for this weapon were pr ovided by its original collector, noted Athapaskanist Cornelius Osgood (Osgood 1937:87). Field Museum of Natural History (FMNH) On May 24th and 25th 2010, in Chicago, I visited another one of this countrys premier natural history museums (incorporated 1893) and documented thirty-seven bows, many of them collected before 1910. The Field Museum sample is fairly evenly distributed between California bows and Southwest bows, wit h a good number of Plains and Great Basin bows in addition. The Plains bows included most notably a fine Blackfeet (Blood band) snakeskin-type w eapon, FMNH 51662, from southwest Alberta, similar to the weapon on perm anent display at AMNH which I was unable to measure or examine. This weapon was published by Allely and Hamm (2002:138) This collection also includes seven Northern Athapask an self-bows (Deg Hitan and Southern Tutchone), four of which were published by VanStone (1996:27-29). In the time available, I also documented a few particularl y interesting Arctic weapons, but I suspect 175

PAGE 176

there are quite a few more. I am not sure even that I exhausted t he collection of gluetype bows, since there is no publically accessible online collection database. Milwaukee Public Museum (MPM) On May 26th 2010, I visited another major early North American museum (chartered 1882). The collection here is most similar to the one at Yale in many respects, as the museum is of a simila r age and the collection was assembled very early on by non-anthropologists Hence a number of t he provenances are vague and sketchy, though many of the bows are among t he oldest I have seen. Here I was able to document everything to the be st of my ability, except for a number of Southwestern bows on permanent display. Collections manager Dawn Scher Thomae sought to arrange the measurement for me of these items at a later date, but a museum visitor has since tampered with one of the display ca se locks, rendering it inaccessible for the time being. The fifteen bows I recorded in this collection include a roughly even distribution of bows from Southwest, Plains, Great Basin, and slightly more from the Arctic. Three of the Arctic bows are com pound bows (two of which are made entirely of horn and/or ivory), and one (MPM 576, from Alaska) is very sim ilar in most respects to the Denaina Athapaskan bow YPM 15844 collect ed by Osgood (Osgood 1937:86-87). On this visit I also had the rare priv ilege of meeting and having a substantive conversation with Curator Emeritus, Dr. Nancy O. Lurie, a close colleague of June Helm and one of the first ethnogr aphers of the Dogrib Athapas kans (Helm and Lurie 1961). San Diego Museum of Man (JES or SDM) On August 10th and 11th 2010 I visited the Jessop Weapons Collection in San Diego California, a very important and ear ly archery collection with particularly good coverage of western Native Americans. Joseph Jessop amassed a global archery 176

PAGE 177

collection between 1881 and 1911, and loaned it to the newly formed museum after 1915; his family formally donated the entire collection to the museum in 1974. A few bows came from donors other than Jessop (which I differentiate using the SDM code) The collection contains at least one rare Denaina Athapaskan sinew-backed bow from Iliamna, Alaska, JES 2691, and at least one addi tional rare Northern Na-Dene complex bow, JES 2687 (most likely Athapaskan or Eyak) from southeast Alaska or Yukon Territory, and another bow from the Kenai P eninsula, JES 2681, which could be either Koniag Eskimo or Denaina. To my kno wledge, this makes the Jessop Collection the only archery collection in the United Stat es (with the possible exception of the Smithsonian Institution) to possess mo re than one Northern Na -Dene complex bow. The global scope of the Jessop archery collection makes it an ideal dataset to interrogate regarding the diffusion of archery technology from Asia to the New World, and it has been put to such use in the past (Rogers 1940). Travel time substantially abbreviated one of my workdays, but I was still able to fully document 39 weapons, including most (if not all) of the Nort h American glue-type weapons. I was merely able to document a por tion of the collection of Arctic gluelesstype complex bows. In selecting which Arctic specimens to prioritize for documentation (at this and other similar museums), I have strongly favo red Western Eskimo weapons, because they are more likely to have close Athapaskan analogs by vi rtue of historical proximity; high Arctic bows are much more morphologically derived and isolated from so-called Indian counterparts. It is once again difficult to be certain of the exact coverage of this collection, because the storage facilities at SDM are antiquated and somewhat haphazard, and the 177

PAGE 178

electronic database is sorely in need of an upgrade. One major concern for this museum is a severe mold problem in their dank and antiquated wooden basement shelving system. A substantial portion of this incredibly im portant collection (including a sizable portion of the bows I personally exami ned) suffers from severe grey-green mold infestation. Extreme conservation measures are advisable in the immediate future, to prevent catastrophic loss of invaluable cultural heritage. The best first step in this regard would be to move the bows to another storage facility, but budgetary constraints might be preventing this necessary step. I recommend the museum seek advice from other institutions with simila r archer collections, particularly the Grayson Collection at the University of Missouri. Phoebe A. Hearst Museum of Anthropology (PAH) On August 12th and 13th 2010, I visited the Hearst Museum (formerly the Lowie Museum) on the University of California, Berkeley campus. Travel time by train from San Diego meant that my work began on the afternoon of the 12th. I was able to document thirty-five bows in one and a half da ys, including all of the glue-type weapons and some of the glueless Arctic weapons; mo st important of the latter category is another one of the extremely rare Denai na Athapaskan complex bow s, this one from Kenai. Unsurprisingly, this was by far the richest collection I examined for bows made by Pacific Coast Athapaskans and their imm ediate neighbors. This collection also houses a large number of Northern Athapaskan se lf-bows, but time di d not permit me to examine them. An informative study of this Berkeley collection of Athapaskan self-bows is made by Benson (1975). Except for t he one bow on permanent display, I had the privilege of examining all t he remaining sinew-backed bows made by Ishi (b.1860? d.1916), the famous Yahi Indian who reside d on the museum premise before his death 178

PAGE 179

(when the museum was located in San Franci sco); the manufacture of these weapons is described in detail by Saxton Pope (1918). I also had the distinct privilege of meeti ng and extensively conv ersing at work and over lunch with the museums Tribal Liai son, Bradley Marshall, a traditional Hupa craftsperson with intimate personal knowle dge of traditional Pacific Coast Athapaskan material culture. He downplayed the achievements of Hupa (Athapaskan) bowyers, stating that the best bowyers were those of the neighboring (Algic) Yurok; this opinion is echoed by the narrator of a classic U.C. Be rkeley-produced ethnographic film which shows the Yurok elder Homer Cooper making a traditional Hupa-Yurok-style sinew backed bow and arrows using all pre-contact era tools and materials (Smith and Barrett 1961). This fascinating film is available on DVD for purchase through the museum gift shop. Mr. Marshall also drew my attention to a turn-of-the-century mail-order catalog from a California curio shop employi ng Hupa and Yurok (then called Klamath) craftspeople to manufacture traditional crafts for sale to custom ers around the country. A finely painted sinew-backed bow and a dozen arrows could be ordered from the Hupa Reservation for $3.50 C.O.D., including ship ping (Brizard 1903:13; see Figure 4-6). One bow in the Hearst collection is apparent ly such a bow, PAH 1/20812, for which the Museum catalog says the bow was made to or der [prior to 1916] for the nephew of E.L. McLeod. By executive order in 1891, the Hupa reservation was merged with the Yurok reservation, and they were collectively known as the Hoopa Valley Reservation (Huntsinger and McCaffrey 1995: 167). Marshall wisely cautioned me to the fact that Hupa bows in various museum collecti ons purchased from the Hupa Reservation 179

PAGE 180

during this period could have easily been Yu rok bows, misattributed by their new owners on the basis of the re servations Hoopa name. Museums Not Visited I will mention in passing two major mus eums that I was lamentably unable to access due to budgetary and time constraints; the University of Pennsylvania Museum of Archaeology and Anthropology (Philadelphia, Pennsylvani a), and the Arizona State Museum (Tucson, Arizona), neither of whic h has online database access at the present time. The following five museums I did not physi cally visit, but was still able to acquire valuable data from each remotely, either through publically accessi ble online databases, or via correspondence with helpful staff members. University of Missouri, Mu seum of Anthropology (MAC) The Grayson Archery Collection at the Univ ersity of Missouri, Columbia, is one of the more important collections that I have not yet had the opportunity to visit personally. Collected over a lifetime by Dr. Charles E. Grayson (b. 1910-d. 2009), this is one of the largest collections of its kind in the world, rivaling the Jessop Collection in San Diego in both size and scope. Grayson began donat ing his massive and comprehensive collection to the museum in the early 1990s (Grayson et al. 2007:xiixiii). Many of the North American bows in this collection are cataloged in a freely accessible online database, where better-than-average measurem ents are included with low resolution photographs. This allowed me to remote ly expand my own database somewhat, for particular metric variables, although the Grayson database is not complete enough for the purposes of this study to adopt the data en masse. I treat it prim arily as a repository for comparative reference matter, with which I can use to help refine my assessment of my own data. 180

PAGE 181

National Museum of Natural History, (NMNH) The main branch of the Smithsonian has an extensive collection of sinew-backed bows, most which are cataloged online as part of the main ethnological collection, but only rarely with photographs or rudimentary measurements, and often with rather sparse descriptions and identifications. If I return to NMAI to ex amine more of that collection, it would be worthwhile to visit the NMNH collections as well, as they both occupy the same campus in Maryland. Than ks to the assistance of collections manager Felicia Pickering and contractor Michael Frank, I was able to obtain a complete set of measurements of a Yukon River Athapaska n boys miniature complex bow, NMNH E5588-0, collected on William H. Dalls expedition on the Yu kon River on behalf of the Western Union Telegraph Company in 1867. Based on the catalog information, this bow could be Deg Hitan, Koyukon, or Gwichin, but I strongly suspect it must be D eg Hitan by reference to Dalls own account of the expedition. Dall describes his habi t of bartering with Deg Hitan child-archers for their small quarry (mice and small birds), and notes that the advent of firearms had rendered adult-sized bows obsolete among t he Athapaskans. Although there was a single Koyukon shaman in their party, there were no children mentioned other than those of the Deg Hitan. The Gwichin of Fort Yukon (allies of the Canadians) were only mentioned as being hostile to the Deg Hit an (Dall 1897:67, 143). This Yukon Athapaskan boys bow is the only comple x bow of certain Northern Athapaskan manufacture in the publically accessible database. Murdoch (1885:315) mentions a rare Tlingit sinew-backed bow from Sitka in the collection, but he does not mention the item number. 181

PAGE 182

Peter the Great Museum of Anth ropology and Ethnography (PGM) This museum collection, located in St. Petersburg Russia, includes two Denaina Athapaskan complex bows collected from K enai in the 1840s (when Alaska was still Russian), PGM 2667-17 and PGM 2667-20. These bows are photographed (but not measured) in Osgood (1937:228). Thanks to Yuri Berezkin (Chair of American Ethnology), I was able to obtain photographs and a partial set of measurements. One of these bows is very similar to PAH 2/6361, also from Kenai. Royal British Columbia Museum (RCBM) I initially planned to visit this collection in Vancouver, British Columbia, because this museum (apparently) possesses two Northern Athapaskan complex bows. However, these bows are bot h of rather vague provenance (both simply Athapaskan attributed) and one of these, RCBM 6563, is currently missi ng, whereabouts unknown. The other bow, RCBM 6571, is on permanent di splay and inaccessible to visiting researchers. This means that a visit to the collection would not have been worth the great expense. I am nonetheless indebted to Brian Se ymour for sending me partial measurements and high-quality photographs of 6571, and the si ngle image and length measure available for missing 6563. State Hermitage Museum (SHM) This museum in St. Petersburg, Russia includes at least five bronze or iron daggers of the Tagar culture of southwest Siberia, on the upper Yenisei river, dating from around the mid first millennium BCE (K arasulas 2004:29). This is the probable Dene-Yeniseian urheimat (linguistic hom eland). These daggers relevance to Athapaskan material culture history has been made explicit by Frederica De Laguna (1947:182), who notes their similarity in size, shape and form (including a distinctive 182

PAGE 183

broad double-spiraled pommel) to the annealed native-copper and trade-iron daggers of the Alaskan Athapaskans. I am grateful for the museums permission to reproduce highquality images from their website. Bow Morphology, Material Cultur e, and Athapaskan Migrations In addition to establishing new qualitative typologies for the rare weapons that do not fit into existing categories, my ultimate goal is to conduct appropriate statistical analyses of morphological variation (such as chi square and mult idimensional scaling analysis) in order to quantify the changes in complex bow morphology throughout geographic space. Taken in conjunction with the linguistic and archaeological evidence for Athapaskan strong complex bo w use, inferences about the timing and direction of the Athapaskan migrations may be possible. The complex archery hypothesis for Athapa skan origins is concisely stated by James Downs, in his standard textbook on the Navajo. He suggests that several distinct yet culturally associated Na-Dene la nguages arrived in a single wave from Asia during the Common Era, in likely response to a period of imperial expansion in northeast Asia. What caused the Nadene-speaking people of Asia to migrate into the New World is unknown. This was a period of great turmoil in the Old World. All through Eurasia peoples from the interior seem to have been pushing toward the edges of the landmass. Per haps this migration to the north and east is somehow related to this general situation. . Two material culture items that they may hav e brought with them are the hard-soled moccasin and the strong, complex bow, both of which appear to have Asian origins (Downs 1972:5-6). Most of the artifacts I exam ined directly in my fieldw ork were sinew-backed bows (both Athapaskan and non-Athapask an); I am also interested in a possible Athapaskan vector for the late prehistoric diffusion of le ather footgear in west ern North America. 183

PAGE 184

However my impression (at least initially) was that most of t he native footgear in ethnographic collections has been greatly infl uenced by Euro-American culture from a number of different source s, and was widely diffused th rough the modern coalescence of pan-Indian Plains culture, not directly attributable to prehi storic causes. I was thus unable to successfully develop a systematic methodology for the analysis of moccasins, and the incorporation of footgear into my study is haphazard. At the outset of my study, I was reluctant to incorporate projec tiles and projectile points into my analysis for several interrelated reasons. First, projectiles are often backwards compatible with bows of differ ent types, and the sinew backed bow (with few exceptions) invariably replac ed the self bow as it diffused. In other wo rds, there are many cases where different arrow types ar e used with similar bows, or where similar arrow types are used with different bows, so there is no straightforward way to use projectile form to infer bow ty pe. Even the difference between arrows versus atlatl darts is in some cases ambiguous (Cattelain 1997; Hare et al. 2004); quantitative distinction between arrowheads and dartpoints is possibl e, but the separation between the two forms is imperfect (Thomas 1978). Also, most Apache arrows in collections are made with post-contact metal tips, which I initially felt were inappropriate analogs for protoAthapaskan material culture (Figure 4-8). In other words, I felt that the numerous ambiguities associated with arrows and arrowheads were sufficient to limit their utility for my study. In hindsight, t here are areas for investigat ion which may have been more useful than originally anticipated, and these will be discussed in the analysis section when addressing objects besides bows. 184

PAGE 185

Arrows often both reflect specialized functi ons, individual style, and broad regional standards (Mason 1894:660). Mo st Athapaskans are not well known for making lithic arrowpoints. Generic, non-de script side-notched arrowheads are characteristic of many groups in Southern Athapaskan territory (Figure 4-7). Style is less useful a diagnostic feature than material source. Apache arro ws often used retouched, recycled prehistoric points or fashion slender blade-like points fr om scraps of sheet-metal (Ferg and Kessel 1987:50-52). The most common Northern Athapaskan projectile point material is poorly preserved bone or antler, al though native copper was also common (Clark 1991:102103). My initial impression was that me tal arrowheads and knives among Athapaskans were a post-contact phenomenon; in hindsi ght, it would have been interesting to investigate possible morphological cont inuities or changes between Northern Athapaskan pre-contact copper points and knives, versus Southern Athapaskan postcontact steel pointsthese issues will be discussed in subsequent chapters. At various stages of my fieldwork I have looked at a haphazard assortment of objects besides bows (such as baskets, footgear, apparel, a rrows, knives etc.), and these will be qualitatively incorporated into my analysis. I did not exami ne ceramics in the field, but they are well represented in the literature and will nonetheless be included in my holistic analysis of Athapaskan material culture. The validity of t he complex archery hypothesis for Athapaskan origins can be determined only in the context of Athapaskan material culture in general. Summary Chapter 2 reviewed more than 100 years of the cross-disciplinary literature on the Athapaskan expansion, and Chapter 3 critically explored more current literature with an eye toward the revolutionary implications of the Dene-Yeniseian hy pothesis in different 185

PAGE 186

subfields of anthropology. Having laid out t he basic methods and issues relevant to the study of bows in Chapter 4, Chapter 5 will begin with a detailed description of the materials examined, and will culminate in an extended discussion of the context of strong complex bow use in western Nort h America and in Athapaskan societies in particular. 186

PAGE 187

Table 4-1. Summary of bows studied. Complete measurements obtained Incomplete measurements Museums Consulted Visited? Complex bows Simple Bows Total Complex bows Simple bows AMNH Y 31 5 36 1 1 FMNH Y 28 9 37 0 0 GMA Y 7 2 9 5 0 MPM Y 13 1 14 2 0 NMAI Y 35 1 37 2 0 NMNH N 2 0 2 0 0 PAH Y 33 2 35 0 0 RBCM N 0 0 0 2 0 SDM/JES Y 37 2 42 0 0 SNO Y 0 3 3 0 0 SHM N 0 0 0 2 0 YPM Y 18 1 19 0 0 Misc** N 4 0 0 0 1 TOTAL n/a 208 26 234 14 2 **Miscellaneous includes bows (in museums and private collections) not examined firsthand, but whose measurements were obtained from published sources 187

PAGE 188

Figure 4-1. The tree of life and the tree of knowledge of good and evilthat is, of human culture (Kroeber 1948:260). Figure 4-2. Three common sinew-backed bo w shapes observed by the author. Crude or improperly-braced bows can be diffi cult to classify. Bow diagrams appear braced (strung), but at rest; sinew-back ed bows are often reflexed (recurved) and would appear to bend backwards when unstrung. Authors sketch. 188

PAGE 189

Figure 4-3. Six common cross-sectional shapes of the bow (reckoned transversely). Because there is continuous variat ion between each shape, intermediate shapes are possible. A single North American bow can have as many as three distinct cross-sectional shapes, as reckoned from the grip, midlimb, or near the tip. Authors sketch. Figure 4-4. Sketches of several differ ent nock-types, grouped by the culture areas where they commonly appear. Backing su rface appears shaded. Authors sketch. 189

PAGE 190

Figure 4-5. Strong complex bows in North America, in comparison to Na-Dene language distribution. Modified from Driver and Massey ( 1957: Map 138) with reference to the literature (Bird and Bohr 2001; Curtis 1928; Emmons 1911; Teit 1906; Murdoch 1885) and to authors field notes. 190

PAGE 191

Figure 4-6. Catalog entry for mail-order purchase of H upa or Yurok bows from the general store of the Hoopa Valley Reservation. Deta il from Brizard (1903). Figure 4-7. WFU 1993.02.E.3. Navajo or Pueblo arrowhead, collected in the 1940s. Approximately 1 inch by 1 inch. Phot o by author, courtesy of Wake Forest University Museum of Anthropology. Figure 4-8. UTK MM.6.B4.B1.1:21G/46. Nineteenth-century metal-tipped Apache arrow (center, foreshaft). Arcshaped blood/lightning groove is faintly visible on pitch-blackened shaft. Photo by author courtesy of Frank McClung Museum. 191

PAGE 192

CHAPTER 5 COMPLEX ARCHERY AMONG ATHAP ASKANS AND THEIR NEIGHBORS Overview In Chapter 4 I explained the method I em ployed for collecting my data. In this Chapter 5, I will present my findings and disc uss their relationship to my original research question. In Chapter 1, I a sked: how and why the itinerant Navajo and Apache, originally from the far north, ca me to demographically overwhelm earlier southwestern Puebloan societies in a relatively short amount of time, despite an extreme difference in cultural ecology between the two regions. Chapter 5 will discuss the likelihood that et hnographic complex bow di stributions help to reveal historical patters relevant to the speed and timing of the Athapaskan expansion, and how such an event was possible. A goal will be to document the geographical patterning of Athapaskan complex bow typologies, and to in tegrate a discussion of these patterns with an analyss of the te mporal processes of Athapaskan migrations and to ultimately provide a greater material cultural context for issues raised in the earlier chapters. I am primarily interested in the possible diffusion of the strong complex bow, also known as the backed bow, in the hands of Athapaska n speakers and their neighbors. Simple self-bows are not the main focus of my study, although some were observed and they will be briefly discussed. I have noted that there are appr oximately equal numbers of Athapaskan-made and non-Athapaskan-made gluetype sinew-backed bows in the collections I have studied, meaning that Athapaskan-made sinew-backed bows are significantly over-represented in proportion to the population of Athapaskan speakers. This might have bearing on the question of w hether or not Athapaskan speakers had a particularly significant role in the spread of this technology. If there are any widespread 192

PAGE 193

and characteristically Athapaskan patterns of complex archery technology, such patterns could be corollaries to the migration process. Constitution of the Dataset My primary dataset consists of a series of detailed measurements and observations made of 242 bows of aboriginal manufacture. I directly measured 236 of 242 bows (97.5%). I did not m easure the remaining six mysel f. The data for these six bows were obtained, either through published sources (Allely and Hamm 2007; Osgood 1937), or through cooperation with helpful mus eum staff and other indi viduals listed in the acknowledgements section (especially Felicia Pickering and her staff at NMNH). My main body of data only includes items w here I have obtained a complete set of measurements. The 242 objects in my primar y dataset may be categorized in a number of different ways. The vast majority (240 out of 242) are more-or-less complete specimens, with the remaining two being smaller archaeol ogical fragments of bows (one from San Francisco, California and one from Lovelock Cave, Nevada). Three bows are complete bows from archaeological contexts (in Calif ornia, Colorado, and Arizona). Only the Lovelock fragment came from a professionally excavated site and not from a stratified context (Loud and Harrington 1929: vii, 97, 178). The rest of the set (238) are mostly bows from ethnographic collectio ns, with the notable caveat that the provenance for a number of these items is ultimately obscure as with most of the archaeological bows mentioned above, owing to t heir collection by a variet y of professional and nonprofessional individuals (many anonymous) ov er more than 160 years. The presence of bona fide archaeological specimens thus may nat urally have escaped attention. In the case of those bows labeled very old when originally acquir ed by ethnographers and 193

PAGE 194

collectors in the last 150 years, they ma y have originated from subterranean caches accessed by the researchers indigenous collaborators, and thus the line between archaeological artifact and et hnographic object may be blurred in these cases. Noteworthy are a Hopi horn-bow, FMNH 66383, collected from a ruin in the occupied village of Oraibi in 1899 (or before), and the Denaina At hapaskan bow, YPM 15844, collected by Osgood in the early twentieth-cent ury from a collaborator who retrieved it from an old cache (Osgood 1937:86-87). These items clearly occupy the grey area between ethnographic and arc haeological collections. The Reliability of Cultural Id entifications in Museum Records Most of the items in my sample have fairly reliable provenances, or at least as reliable as can be expected for objects colle cted mostly by non-professionals during the late nineteenth and early twentie th-century. But a significa nt number of items are only tentatively identified in the catalogs or accessioning books, and thus have questionable origins. Depending on the standards of evidence one adheres to, between about twelve and eighteen (or more) bows in my dataset we re at least initially without any firm provenance whatsoever. In cases where a va riable degree of certai nty exists about an objects origin, I have relied upon morphological similarity with other similar objects to help confirm or refute the speculative pr ovenances in museum records (which I have indicated using question-marks next to corre sponding ethnonyms in my spreadsheets). In a number of cases, even the most basic provenance information may be unavailable, and I am left to assign cultural provenance myself based on comparison to similar materials. Other weapons possessed a particula rly dubious provenance, like a Plains/Southwestern-style sinew-backed bow inexplicably attributed to the Florida 194

PAGE 195

Seminole, YPM 30856, a central-Californi a-style sinew-backed bow attributed to Mexico, YPM 15005, or a Northern California st yle bow attributed to the Arctic, NMAI 25/1667. Likewise, some catch-all regional descriptions may be employed, such as Plains, California, or Arcti c. Similarly, a laundry list of multiple cultural affiliations may be employed, such as is the case with the old bow cataloged as Sioux, Cheyenne, Crow, Blackfeet, Cree or Hi datsa, MPM 3168a. And many of the Sioux and Cheyenne sinew-backed bows I have measured are extremely difficult to distinguish from Apachean bows (unlike other groups); this si milarity is such that bows lacking firm provenance are often impossible to place wit h certainty on typological grounds into the either the Southwest or Plains categor y, and may be labeled Sioux OR Apache, Navajo OR Plains, etc. Such ambiguous weapons include MAC 1995-0661, MAC 1995-0027A, and YPM 10763. (For a summar y of items with corrected provenance information, see Table 5-1 and Table 5-2). Fine-grained measurements permit a fram ework for the comparison of poorly identified materials with those whose proven ances are well-established. This has the advantage of strengthening my dataset and increasing the va lue of the collections studied for all concerned (museum professi onals, the interested public, and the Native Americans whose cultural herit age these objects represent). In addition to the complex bows which constitute the primary focus of my fieldwork, I was also able to correct errant cataloging information for other items of Native American cultural patrimony as well, particularly splint basketry from the No rthwest Coast. I identified two such plaited splint baskets WFU 1984.E.0437, and WFU 1984.E.0668, which we re arbitrarily attributed to Subarctic Athapaskans (sometime after they were accessioned) on the 195

PAGE 196

basis of a generally credible Western Canadian provenance. But this type of basketry is relatively rare in the region, and the onl y Western Canadian natives definitively known to make plaited splint baskets are Coast Salishan and Wakashan peoples (Goggin 1949:166). Others may have made them as we ll, so an Athapaskan provenance is not unfathomable. But it is unreasonable simply to guess at Athapaskan identification on the basis of current evidence. This spec ulation was presumably made because most of Western Canada is Athapaskan territory, despite the fact that Athapaskans reside in the interior while the most prolific basketr y producers reside on the coast. Further corroboration of a likely Salis han provenance for these putatively Athapaskan baskets is made through comparison to similar Salish cedar splint baskets in the Pearsall collection at the University of Florida (Table 5-1 and 5-2). Basic Typological Divisions of the Dataset Basic typological divisions of the dataset are fairly straig htforward. A total of 213 of the bows I have classified as complex (or backed-), and 29 as simple (or self-) bows. See Figure 4-5 for a pre-contac t depiction of the approximate geographic distribution of the two major complex bow types in North Am erica. The complex bows can be divided into several categories, not all of which are mutually exclusive. The vast majority of the complex-bows are backed with si new, but at least eight of them of them are backed with non-standard materials. Si x are backed with hide (or generically babiche, which is a term that may refer to gut or hide). T hese six babiche-backed bows are exclusively from the Arctic and S ubarctic, including tw o Athapaskan bows and four Eskimo bows (two from Alaska and two from Siberia). One additional bow (probably Tlingit, attributed to the Northwes t Coast) is backed with white plant-fiber 196

PAGE 197

cordage (presumably cotton, acquired though trade). These seven bows are of the glueless or cable-backed type. Formerly, among the Tahltan Athapaskans of the Western Canadian Subarctic, the glue or lined variety of backed-bow was also known to occasionally include hide in lieu of sinew for the longitudinally applied reinforcement backing; these glue-type hidebacked bows were not collected by the early missionaries and ethnographers who observed them (Emmons 1911:65-66). Two s hort glue-type sinew-backed bows in the dataset, from Western Canada, have snakeskin over the backing, but this skin is found in addition to the normal sinew (as a waterproofi ng agent) rather than in lieu of sinew as in the Tahltan case. Another bow, AMNH 16/ 1071, is entirely encased in a glued-on sheath of cherry bark, wrapped in a transverse spiral from tip to tip, without any apparent sinew layer underneath. By increasi ng the bows elasticity, the bird-cherrybark wrapping is a functional (though less effective) substitute for a sinew back. Bows of this type were used in the Nicola Valley of British Columbia, by the local Athapaskans and Inland Salish speakers (Teit 1900:239-241). Unfinished and Fragmentary Bows Four bows in the sample are unfinished, te chnically just stav es in preparation, prior to the application of the sinew ba cking (two Jicarilla Apache, one Hupa and one Blackfeet). All four of these staves have various cl ear diagnostic features of sinewbacked bows, indicating they are not simply self-bows, but would-be backed-bows prior to the application of backing; thus they may be included without hesitation in the relevant subsets of the dat a with finished bows of the same type. One of the archaeological bow-fragments, PAH 1/174972, lacks any obvious sinew back, but has surface characteristics suggesting the presence of an adhesive residue and a 197

PAGE 198

somewhat coarse back (in contrast to a polished belly) suggesting a sinew backing originally was attached to this weapon, lost through decay. This weapon, consisting of most of one bowlimb, is also morphologically consistent with Miwok and/or Wintu sinewbacked bows from the same region, so it can be justifiably pooled with these similar sinew-backed weapons for my analysis (Figure 5-2). Complex Bows versus Self Bows Including the four unfinished and two fragmentary weapons along with complete weapons in this category, there are 170 glue-type sinew-backed bows in addition to the single glue-type bark-wrapped weapon from Brit ish Columbia already mentioned, making 171 glue-type complex weapons total. A total of 41 bows are glueless cablebacked bows of various types, and of these 40 are from either Canada, Alaska, or Siberia, while just one cable-backed weapon is from any of the coterminous forty-eight United States (in this case, Arizona), a likely Apachean cave-bow. A similar situation obtains for the presence of glue-type weapons in the north ; there are four glue-type complex bows which are dem onstrably Canadian in provenance; (three sinew-backed, one bark-wrapped), one Blood (Blackfeet) weapon from Alberta, and three Thompson/Nicola weapons from British Columbia. The rest of the glue-type weapons are from the Western half of the lower forty-eight United States. My initial plan in undertaking collections re search for this study was to compare specifically Athapaskan gluetype sinew-backed bows from British Columbia and Alberta to those made by Southern and Pacific Coast Athapaskans. There are abundant references to Canadian Athapaskan glue-type complex bows in the early anthropological literature Morice (1894:53-59), Ra y (1942:148-149), and McClellan and Denniston (1981:378) record them for the Chilcotin, Carrier, Babine, Sekani and 198

PAGE 199

Tsetsaut. Teit (1900:239-241) records them for the Nicola. Emmons (1911:65-66) and Teit (1906:343) record them fo r the Tahltan, and Curtis ( 1928:99) records them for the Tsuutina. The Canadian Cordilleran peopl es (probably Sekani) were so closely associated with sinew-reinforced bows that they were christened Strong Bow Indians on Peter Ponds 1785 map of native Cana dian groups (Gillespie 1975:376; Ives 2003:258) (Figure 5-2). I originally thought that such weapons had been collected by the ethnographers who described them, but it appear s they were not. If any such weapons survive, I have not personally found them, nor have I found an authority who knows of their existence. The closest examples are those similar snakeskin-covered sinew-backed weapons from British Columbia and Alberta (Thompson-Salish and Blood-Blackfeet) which are geographically directly peripheral to the Chilcotin/Nicol a and Tsuutina peoples respectively, AMNH 50.1/1241A, AMNH 16/1341, and FMNH 51662. Typical Chilcotin and Tsuutina sinew-backed bows also used snakesking waterproofing, and may well resembled these weapons (Cur tis 1928: 99; Ray 1942:148). Horn Bows, Compound Bows, and True Composite Bows Eight of the weapons in my sample were either horn-bows (utilizing horn or bone instead of wood as the primary stave material) or true compound bows (featuring multiple stave segments of horn, bone or antler and/or wood spliced or laminated together), or both. There is considerable overlap between these categories. Four of these bows were of the glueless-type from Canada and four were of the glue-type from the contiguous US. In the far north the compound/composite bow is closely associated with treeless climes and with the lack of read ily available bow-wood. Further south, horn bows and compound bows are associ ated with the rise of equestrian 199

PAGE 200

transhumance in the northern Plains and Plat eau after the eighteent h century. This owes to the advantage an extremely s hort bow confers upon a mounted archer, similarly to Old World contexts (Hamilton 1970:45, 51). They also have sporadic distribution among more southerly and seden tary peoples (such as Hopi and Kaibab Paiute). The limited distribution of this w eapon in regions where horn was plentiful is suggestive of a relatively recent norther n origin (Steward 1941:231). Glue-type horn/composite bows have been reported for Athapaskans, such as the Northern Athapaskan Tsuutina of the Canadian Plains (Curtis 1928:99). To my knowledge, the only putativel y Athapaskan compound-bow in a museum collection is a glueless-type specimen hel d in British Columbia, Canada, catalog number RBCM 6563. The catalog suggests this weapon is either Cree or (presumably Chipewyan) Athapaskan. Chipewyan were indeed known to use glueless cable-backed weapons, but compound weapons such as this one were not recorded (Curtis 1928:27). Past use of some form of cable-backed we apon by Western Woods Cree is tentatively suggested in a second-hand account of Cree-e lder Louis Bird, who recalls a peculiar type of Cree sinew backed bow (which he did not actually see) described to him by his father (Bird and Bohr 2001:15). Bird does no t mention any form of compound (or truecomposite) weapon. In cont rast, the Eskimo compound bow s from north of Chipewyan and Cree territory often employ a more dist inctive and complicated High Arctic-style cable-backing method (Mur doch 1885:310-313). The compound RCBM 6563 might have been collected at a village near Fort Churchill, on Hudsons Bay in northern Manitoba, where Cree and Chipewyan allies were known to congregate with Eskimos. Th e uncomplicated backing is reminiscent of 200

PAGE 201

southern Alaskan bows, secured by fifteen pairs of half-hitches, with one pair approximately every two inches of length. Most (if not a ll) of the Northern Athapaskan sinew-backed bows I have measured and document ed are similar in this regard. But the compound-nature of the stave is more typically Eskimo; RCBM 6563 has at least four stave segments of horn and wood held together with rivets. RCBM 6563, the one sample of this kind of bow on record, now only exists as a photograph; the specimen in question was lost sometime during the past twenty years (Figure 5-4). The photograph shows mostly the belly side of the bow, with only a few inches of the sinew backing visible at one end of the bow. A firm cla ssification of this weapon is unfortunately impossible on the basis of current evidence. The Old World Origins of the Composite Bow The oldest example of true composite bows were wooden bows reinforced with bone or antler plates (and presumably back ed with sinew), excavated in the midtwentieth-century from the Sayan uplands of Siberia, in the Altai region west of Lake Baikal (Rausing 1967), (Figur e 5-5). This region happens to be the probable Yeniseian language urheimat. These bow pa rts date to the late Neolithic or Early Bronze age slightly more than 4000 years ago but less than 5000 (Chard 1958:10). The sophisticated recurved composite weapons of later Eurasian steppe empires are careful refinements of the ancient technol ogy of the Altai region. There has long been a debate about whether t he Eskimo-style of glueless cablebacked weapon represents a northeast Siberian antecedent technology, which gave rise to the glued southwest Siberian composite weapon somewhat later on (Balfour 1890). Early anthropologists were convince d that Eskimos and American Indians were descendants of the forebears of all North Eurasians, and their possession of less201

PAGE 202

refined complex bows seemed to support this idea, with North American sinew-backed bows allegedly representing the s urvival of an early form in the direct line which has led to the perfected Asiatic bows (Balfour 1890:224). But this cannot be assumed based on archaeological data, as the earliest evidence of pre-Dorset Paleo-Eskimo complex archery (consisting of a few bow bracers) is about 1000 years younger than the Altai specimens, dating from r oughly 2900-3900 years ago (Meldgaar d 1960:74-75). This early North American complex ar chery did not survive to late Paleo-Eskimo times but had to be reintroduced from Siberia during the Common Era expansion of Neo-Eskimos (Meldgaard 1962:pl. 5). The sudden American appearance of true composite bows with the arrival of horse cultur e to the northern Plains and Basin is something of a conundrum, especially considering these bows st riking similarity to the original south Siberian prototypes. This pr ompted the foremost authority T.M. Hamilton to speculate that they were independently reinvented, on the basis of the fact that there are no archaeological precedents for tr ue composite bows in Pre-Co lumbian North America. But he nonetheless recognizes the importance of a close comparative study of New World and Old World examples of the weapons. [T]he true composite . suddenly appeared on the Americ an high plains along with the advent of the horse in the early 18th century. I am not trying to imply that there can be any c onnection between Baikal bows of around 2500 B.C. and the American horn bows of about A.D. 1700 other than similarity in design. However, I do think that some authority on horn bows should examine the Baikal specimens (Hamilton 1970:45). American true composite bows (consisti ng of laminations of horn, wood and sinew) are extremely rare. I examined only one such bow in my study, a Shoshone bow 38 inches long, NMAI 13/1513 (Figure 5-6). 202

PAGE 203

Athapaskan Guard Bows Eight of the bows in my study have Northern Athapaskan-st yle attached wrist guards, consisting of a thin trapezoidal wedge a couple of inches long, lashed to the belly of the bow near the grip. Of these ei ght bows, six are most likely Athapaskan, one is attributed to Eskimo, and one is attributed to Aleut. Two of them (both Athapaskan Denaina) are glueless sinew-backed bows, and six are self-bows. An additional sinewbacked Denaina bow of this type, PGM 2667-20, is in St. Petersburg, Russia where I was able to obtain some measurement s through direct correspondence with the museum (Figure 5-7). The attached-guard bow despite being popularly emblematic of Athapaskan archery technology, does not enc ompass the Athapaskan expansion as a whole. The technology is confined mostly to Alaska with limited distribution among Western Canadian Athapaskans, spanning roughly t he territory from the Cook Inlet to the Yukon headwaters (Mc Kennan 1959:52). Four of the guard-bows I examin ed were from the western extreme of this territory (western Alaska), while four of them were from the eas tern extreme (the southeastern Alaskan panhandle and the Yukon Territory). Additional data for the presence of this bow-type in the intervening territory is cu lled from other sources (Benson 1975). The primary difference between the easternmost and westernmost examples of Athapaskanstyle guard-bows is where the hole is drilled in the guard itsel f; that is where the sinew cords pass through the base of the guard to secure it to the center of the bow stave. There are four bows in the dat aset from the upper Yukon; three are from Whitehorse (Yukon Territory) and one is from Haines (A laska). All four have the hole for the attachment thong drilled through the large fa ce of the trapezoid (longitudinally with respect to the stave). In contrast, the Eskimo, Aleut, and Denaina Athapaskan guard 203

PAGE 204

bows from southwestern Alaska all have t he hole drilled in the small face of the trapezoid, transversely (laterally) with respec t to the stave (Figure 5-8; Appendix A). Northeast Alaska appears to follow the sout heast Alaska pattern; the typical Gwichin guard bows from the PAH coll ection illustrated by Benson (1975:38), (which I did not measure) also has the hole dr illed in the small face; Benson implied that this guard design was typical of all the Gwichin bows he examined. Likewise this western pattern is evident in the Gwichin guard bow MAC 1994-0970 (Grayson et al. 2007:167-168). The Inclusion of Self-Bows in a Study of Complex Bows Self-bows (as opposed to complex bows) were virtually ubiquitous throughout the Western Hemisphere in pre-contact timeframes; attempts to link their diffusion to particular phyletic expansion events are un likely to be productive. It is the limited distribution of complex bow s among northern and western peoples that makes their distribution potentially illuminating for questions of migration and diffusion. However, I did examine a number of simple or self-bows as well. In some regions (particularly in the Subarctic homeland of Northern Athapaskan speakers) the extreme rarity of sinew-backed bows ma y require that self-bows are examined (to some extent) in lieu of the elusive complex ones. While it is often the case t hat the two types differ markedly, there are other cases where thei r design features appear to merge (such as in southern Alaska). And a sample of se lf-bows helps to form a baseline for the difference between the two types (simple vs. co mplex). I examined a total of 29 self bows. Ten are Northern Athapaskan in orig in, and two are Eskimo or Aleut, but of a common Northern Athapaskan type extremely rare among non-Athapaskans. Nine selfbows are Southern Athapaskan (Nav ajo or Apache). Another five are California bows from the Hupa, Yurok and/or Pomo (r epresenting a common California bowmaking 204

PAGE 205

tradition; see Figure 5-9 for a sinew-backed bow of the same type), and one is a Ute beaded-bow which is apparently also indebted to this tradi tion. I examined a Comanche self bow with some similarity to Kiowa and Plains-Apache self bows in the literature. Finally, I examined one sinew-reinforced self -bow from a cave in Colorado bearing some resemblance to a similar cave-bow from Arizona (the latter having a sinew back) and to some northern bows and rein forced spears (Hanson 1994:181-182). Cross-sectional Shape Profiles for Bows The cross-sectional shape of the bow is an attribute relevant to the spatial patterning of bow morphology and its cultural implications. This is an attribute which most often must be directly observed in pers on, as it is rarely evident in photographs and not generally included in the text of catalog descriptions. Simple measures of length and width and thickness are more often found in such descriptions, but these dimensions may be practically constrained by t he availability of particular bow-woods or use contexts. By contrast, the shape of the cross-section of the bowlimb (and/or grip) appears to vary more an to be more directly link ed to variable cultural factors, and it is the feature which appe ars to most strongly differentia te regional and inter-ethnic bowmaking traditions. Intermediate Forms I created a system of classification for de scribing the variation in cross-sectional bow shape. Figure 3-2 shows the ten most common distinct cross-sectional staveshapes observed in the dataset. These are coded wit h the letters B, D, J, K, L, O, P, Q, R and T. Since ambiguous intermediate form s (with attributes of more than one shape are relatively common, occasionally it is nec essary to use a hyb rid classification scheme. A roughly ovate shape (code O) with four flattened, parallel sides like a 205

PAGE 206

rectangle (code R), may also be conceived as a roughly rectangular shape with heavily rounded corners. The subjecti vity of arbitrary assignment of an ambiguous to one or the other category could introduce bias to t he dataset. It is oft en true that a shape is much closer to one end of a continuum than another, but for truly intermediate specimens, I would prefer the classification O/R, O/L, etc. signifying that aspects of both shape profiles are present. The order is mere ly alphabetical, meaning that no priority is implied; it is meant to convey intermediacy. Many forms can be hybrid with O ovat e, or R rectangular shapes, and their defining features may be especia lly subtle. A typical D shape has two surfaces; a flat belly with a smooth, arcing back. A D/O shape is like a D without distinct tips; that is where both surfaces merge r oundly at the margins. Another possible hybrid shape form is D/L, D-Lenticular, where an otherwise lenticular shape has one surface bulging outward asymmetrically, to approach a letter-D shape. D/L has sharply defined margins (unlike D/O) but is without t he flat face of the D. A ty pical B shape is like a slice of bread, with three flat sides and one convex arch. A B/R is a shape where the bulge of the arced surface is particularly subtle (approaching rectangular) but still distinctly present. Rarely, one may see a B/D shape, where only one of the parallel sides required for a B designation, or where both parallel sides are short or subtly merge with the arced surface in a D-like fashion. P and Q are the mirro red (inverted) forms of B and D respectively (with the shape of the bows belly surface substituted for the back). P and Q hybridize with R and O respectively in the same manner as their mirrored counterparts. A hybrid with the mirrored form (B/P or D/Q) is impossible, because one 206

PAGE 207

would not be an ambiguous intermediate fo rm, but a common sym metrical one (D/Q = O; B/P = R or J, jarred, fl at sides, curved belly and back). Tri-local Classification System I have recorded cross-sectional shape at three different locations on the bow stave, beginning with the grip and moving outward along the bowlimb to ward the tip. As many as two different shape-codes may apply to any particular cross-sectional measurement. I transcribe crosssectional shape profiles from le ft to right in the grip-totip direction. A bow with a uniformly D-shaped cross-sectional profile throughout its length would be coded D-D-D. The shape at any one of these three locations can then be straightforwardly compared across mult iple bows in a population (or between populations) using normal statistical measures. The influence of ambiguous intermediate shape forms can be accounted for in any bow, by assigning a half-value to each component of an intermediate (hybrid) shape. So a hypothetical bow with a D(D/O)-O profile, would contribute one D to t he dataset for the grip locus, half a D and half an O to the dataset for the midlimb locus, and one O to the dataset for the tip locus. Morphological Data for North American Complex Bows I have grouped several illustrative portions of my dataset into tables based on cultural or geographic attribut ion (Tables 5-2, 5-3, 5-4 a nd 5-5). I have elected to reproduce only those variables which are pa rticularly illuminating for subsequent analysis. For the most part, the bows are grouped in tables bas ed on ethnolinguistic affiliation, and there are some clear correlations between ethnicity/language and bow morphology. However, in other cases t here have been several linguistically unrelated groups with particularly close historical ti es, such as (Athapaskan) Hupa and (Algic) 207

PAGE 208

Yurok in California, whose bow s are virtually indistinguis hable from each other. Likewise, Plains Siouan and some Plai ns Algonkian groups (like Arapaho) are especially close historically, and their bow mo rphology data may be safely pooled into one population system. Blackfeet bows, on the other hand, seem to be more closely related to Plateau bows, with Cheyenne bows diffi cult to classify (W issler 1910:160) For some Alaskan bows, a place of origin is all that can be det ermined with certainty, and so specific bows could be pooled with eit her Eskimo or Athapaskan populations. Several unprovenanced or misprovenanced bows have been reclassified based on my own comparative analysis (Table 5-1 and 5-2) Where possible, these items are pooled with the data corresponding to the corrected classification. The Abundance of Athapaskan Glue-Type Complex Bows The largest portion of my sample wa s glue-type sinew-backed bows, at 170 bows, and of this sample, 93 (54%) are mo st likely made by Athapaskan speakers (66 Southern Athapaskan bows and 27 Pacific Coas t Athapaskan Bows). At most of the museums where I measured these bows, I made no effort to single out Athapaskan bows in general. I worked my way through all of the glue-type sinew-backed bows I could find, meaning that the ratio of Athapaskan to non-Athapaskan glue-type bows in these institutions, pooled, is approxim ately mirrored in my dataset. To my knowledge, at eight of nine inst itutions where such bows are held (UFL, GMA, AMNH, YPM, FMNH, MPM, SDM and PAH) I examined all of the extant glue-type sinew-backed bows, save for those inaccessible because they were on permanent display or otherwise inaccessibly stored (see List of Abbreviations). The notable exception is one museum, NMAI, where time and circumstances prevented my examination of non-Athapask an bows; thus NMAI bows are over-represented by 208

PAGE 209

Apachean weapons in the glue-type dataset. A similar situation pertains for glueless (trussed) bows of Eskimo manufacture; because of time constraints in some museums I wasnt able to view Arctic weapons at all. The glue-type portion of the sinew-backed bow dataset, in contrast may be considered roughly representative of the population of glue type bows in these US museums, after correcting for the bias provided by the NMAI portion of the data. Because I was unable to exam ine any but a handful of non Athapaskan glue-type sinew-backed bows at NMAI, my inclusion of the NMAI glue-type bo ws in the dataset somewhat inflates the proportion of Apachean weapons to glue type weapons in general. To correct for this and get a bette r sense of the ratio of Athapaskan to nonAthapaskan glue-type sinew-backed bows in American museums, I simply subtracted the eighteen NMAI Athapaskan and four NMAI non-Athapa skan glue-type bows from the pool (also subtracting one Crow and one Apache weapon whose measurements were culled from the literature), I am left wit h a rough ratio of 74 of 147; almost 50% of glue-type sinew-backed bows in eight American museums are of Athapaskan manufacture. Because ethnological museum displays strive for equal representation of the cultures of interest, permanent museum displays do not accurately reflect the Athapaskan-bias of the collections; I have observed that glue-type bows on permanent display over-represent Non-Athapaskan exampl es of this technology by about 2:1. There are several possible factors that could help to explain the overrepresentation of Athapaskan bows in collections of glue-ty pe sinew-backed bows in North American Museums. 209

PAGE 210

Demographic Advantage If Athapaskan speakers in the past held a significant demogr aphic advantage over other tribes, then it is concei vable that their cultural arti facts could be overrepresented in museum collections. The Southern At hapaskans definitely have a demographic advantage over other indigenous linguistic groups today. The United States Census for the year 2000 showed the Navajo is the sec ond largest native ethnic group in the United States, second only to the Cherokee, with about 270,000 Navajos self-identifying as fullbloods, and about 300,000 when mixed-race individuals are included. Apache are the sixth largest indigenous ethnic group in the United States, with between 57,000 and 97,000 individuals depending on whether mix ed-race people are counted. The total Native American population of the US in 2000 was 2.5 million or 4.1 million depending on whether mixed-race are counted (Ogunwole 2002:1, 10). Full-blooded Navajos constitute about 10.8% of all full-blooded Na tive Americans; the population including mixed-race individuals constitutes about 7.3% of the adjusted total. When Apaches are factored in to account for all Souther n Athapaskans, these two numbers are about 13.1% and 9.7% respectively. The number of Athapaskans in the United States in 1894 (based on statistics published by the office of the US Commissio ner of Indian Affairs) was 16,102 Navajos, 5,742 Apaches, and about 22,616 Athapaskans tota l including the addition of less than 1000 Pacific Coast Athapaskans (Morice 1894:16) The 1890s constituted the all-time nadir of the aboriginal American population, with only about 250,000 indiviuals surviving in the US (Thornton 1987:43). It is not clear if this esti mate includes Alaska Natives (Alaska was then a US territory), but taken at face value this would mean that the total percentage of Navajos to natives in general was 6.4%, the total per centage of Southern 210

PAGE 211

Athapaskans (including Apaches) was 8.7% and the total including Pacific Coast Athapaskans was 9%. These numbers are indeed large in comparison with other tribes, but it remains to be seen if they are large enough to account for the radically disproportionate (1:1) cont ribution of Athapaskan-made sinew-backed bows to nonAthapaskan-made sinew-backed bows in the ethnographic record. The population of Navajos was admittedly relatively high in 1894, but the twentie th-century witnessed a Navajo population boom both in terms of ac tual numbers and as a proportion of Indians in general. Furthermore, if demography were the ke y factor in the equation, then one would expect to see far more Navajo bows than Apache and Pacific Coast Athapaskan bows put together. In reality, t here are more than three times as many Apache bows than Navajo bows in the dataset, and about twic e as many Pacific Coast Athapaskan bows as Navajo bows, despite the fact that t he Pacific Coast Athapaskan population in 1894 was less than 5% of the Navajo population, and the Apache population was barely more than one third of the Navajo. Demogr aphic advantage cannot hav e been the primary factor at play here. The Possibility of Deliberate Misidentification Deliberate misidentification in the ninet eenth-century could have skewed the data somewhat. Nineteenth-century curio dealers were liable to want to make a profit in any way they could, and glamorizing the provenanc e of a bow is one way to make the item worth more. One could speculate that the weapons of infamous Apache resistance fighters may have been coveted by collectors, and so the Apache designation could have found its way onto a number of generic I ndian bows. In museums, perhaps, this was less likely than in private collections, but many bows in museums originated from 211

PAGE 212

private collections of indi viduals who acquired their item s on the open market during the turn-of-the-century reservation period. Such deception would be more likely (and profitable) for crude self bows lacking di agnostic features and made with haste and efficiency for the tourist trade. Sinew-ba cked bows, by contrast, were more likely sought by more knowledgeable collectors for whom such bald-faced deception could be a liability. Also, the inherent value of a sinew-backed bow (a refined weapon requiring painstaking time and skill to make) is rela tively high to begin with, and the distinct stylistic features of the bows of different tribes are more obvious to the appraiser. Finally, all the the typological features of Southern and Pacific Coast Athapaskan sinew-backed bows, resepctively are among the most consistent in my dataset, strongly suggesting few if any false positives. If anything, the relative abundance of Apache bows would have facilitated deceptive misidentif ication of them as other, more highly coveted artifacts. There is some indicati on this may have taken place. For example there is a bow in the Grayson collection (MA C 1995-0027A) which is in all respects very much like an Apache weapon, but has a ni neteenth-century dealers sticker on it reading Little Big Horn Battle. The general similarity between bows of the Plains and Southwest is undeniable, but it is hard to gauge the precise cause of this similarity, whether it be trade of bow s between groups, horizontal language transmission and panethnic coalescence (or band conversion, e.g. from Apache to Sioux), or antiquities dealer fraud. But if the latter played a role, there is no evidence it could have inflated the ranks of putatively Athapaskan bows so drastically; if anything it could have reduced them somewhat in favor of more rare and unique provenances. 212

PAGE 213

Longer Retention of Trad itional Lifestyles The abandonment of traditional American Indian technologies is closely linked to assimilation and acculturation to Euro-Ame rican society; the earlier that contact and assimilation occurred, the earlier the transit ion to Yankee technological norms (firearms among other things). It is conceivable that much longer Southern and Pacific Athapaskan resistance to assimilation resulted in more retention of traditional arts like bowyery, in contrast to other groups. In aboriginal western North America, crafts and technologies associated with women persisted longer in the last-century period of intensive acculturation than did those associated with men. Women continued to practice traditional crafts such as basketry and pottery, and traditional foodways such as seed and root collecting, well into the historic period. Aspects of this culture still survive in some regions. . In contrast, the advent first of the fur trade and t hen of widespread mining and ranching enterprises brought employment opportunities and new technologies, including firearms. . In many regions these factors combined to significantly reduce big-game populatio ns. As a result, male-oriented technologies, such as flintworking and archery, and hunting as a regular subsistence pursuit, rapidly declined (Wilke 1988:4). This situation definitel y obtained when nineteenth-cent ury ethnographers first visited the southernmost Canadian Athapaskan groups (Tsuutina, Sekani, Chilcotin, etc.); their once-great tradi tion of complex glue-type sinew backed bowyery was already a lost-art, annihilated several decades ear lier by a flash-flood of British and French firearms on the open market. The manuf acture of complex bows could then be described in detail only by mature informants (Morice 1894:53-59). The abandonment of bows happened similarly early among Yukon Rive r-dwelling Athapaskans in Alaska and Blackfeet Indians in Alberta; among all of these groups, the firearm quickly replaced the bow and arrow in the second half of the nineteenth-century (Da ll 1897:147; Wissler 213

PAGE 214

1910:155-162). This is in stark contrast to the situation wit h Apache of the Southern Plains during the same era: Even in the second half of the nineteenthcentury, when firearms were more easily secured than they had been in Spanish times, the basic trusted weapon of the [Lipan Apache] warrior remained his bow and arrow. Lipan poverty was not the only reason for th is. A bow and arrow, in comparison with a single-barreled, muzzle-loading ri fle, was a very effective weapon when hunting buffalo or when making a charge (Newcomb 1961:117). Many aspects of traditional Apache and Na vajo society endured through the late nineteenth and early twentieth-century; the fact that the nuclei of the richest ethnographic collections were formed during this timeframe cannot be a coincidence; there were active bowyers in the communiti es when first visited by anthropologists. Likewise, Pacific Coast Athapaskan bowyery was maintained in relative roadless isolation from Euro-American society. At a time when Eu ro-American migration into northwest California had only recently begun, Brizard (1903:1) wrote of the complete isolation of the tribes, and the dirth of tourists to the country owing to the distance, expense and hardship of a trip here. Late nineteenthcentury Pacific Coast and Southern Athapaskan cultural persistence in the face of American imperialism could account, at least partially, for the overabundance of their bows in ethnographic collections of that era. The replacement of archery technology with firearms is not, stri ctly speaking, one of an inferior technology simply subsiding in the face of a superior one. The long retention of technologically superior bows by Apacheans is not an anomaly. As long as traditional Apache archery was a living tradition, it could more than hold its own in the face of the sound and fury of crude mu zzle-loading firearms. Lynn White notes: The acceptance or rejection of an invention, or the extent to which its implications are realized if it is accepted, depends quite as much upon the 214

PAGE 215

condition of society, and upon the imagi nation of its leaders, as upon the nature of the technological item itself (White 1962:28). Commenting on the replacement of the English longbow by firearms between the fifteenth and sixteenth centurie s, Thomas Esper (1965:393) noted the replacement of the longbow by firearms occurred at a ti me when the former was still a superior weapon; it was the decline in the social forces promoting archery as the national pastime that resulted in the decline of archery itself. Once the English no longer practiced the sport religiously, the decline in the penetration power of the bow rapidly followed the decline of ability of individual archers. The anachroni stic perception that the bow and arrow is somehow an inferior weapon to early firearms is rooted in the pervasive modern ignorance of good archery practice. The logic holding true for Renaissance Europe also holds true for the late nineteenth-centur y American West. Crude firearms inexpertly wielded were no match for sophisticated bows, and those cultures where archery practice was st rongest (like the Southern Athapaskans) were able to retain their traditional weapons for several decades longer than those for whom the social basis for the maintenance of archery practice was less powerful. Guns did not replace bows because they were superior weapons. It may be too simple to suggest that guns require less skill to use than bows, but the sound and fury of firearms provide an addi tional advantage in combat. Furthermore, pointing and shooting a gun may require somewhat less pr actice and skill than archery. Another factor to consider is equi pment cost. The supply of ammunition and gunpowder could be unreliable and variable in proximity to fr ontier trading posts; this could have been exacerbated for peoples like Apaches with highly mobile lifestyles. 215

PAGE 216

Entrepreneurism or Mercantile Opportunism Entrepreneurial forces could also be at play in the significant overrepresentation of Athapaskan bows in the sample (nearly one Athapaskan-made weapon for every nonAthapaskan one). Athapaskan mercantile oppor tunism was responsible for the long history of bowyers wares being produced fo r trade between neighboring tribes; this is even evident in oral traditions about early Navajo-Pueblo interactions. In Navajo chantway traditions of the origin times, the Ancestral Navajo traded bows, quivers and buckskin garments for Anasazi (ancestral Pueblo) basketry, pottery, and kitchen utensils (Warburton and Begay 2005:550). The transfer of this specialists art to the market of white collectors and museum curators may have been a natural transition during adjustment to reservation life. This is demonstrably the case for Hupa and Yurok bows especially (Brizard 1903:13). The Hupa and Yurok were already the most reput able bowyers in California when they began producing their refined craft in a cash economy. As a result, Pacific Coast Athapaskan style sinew backed bows (inc luding Athapaskan, Yurok and Pomo weapons of the same type) constitute over 75% of my sample of sinew-backed bows from Oregon and California. Entrepreneur ism is a likely culprit. In another example, what was (the last Yahi survivor) Is his practice of making bows in his new home (the U.C. anthropology mu seum) if not paying the rent? There were still Pacific Coast native bowyers profic iently making bows in the Hupa-Yurok style in the late twentieth-century, in marked contrast to the vast majority of tribes for whom this weapon was the norm, but the art is now lost. Similar entr epreneurial forces may have been at play in the Apac hean cultural sphere in the late nineteenth-century, although I am much less certain of the specif ics of this process in the Southwest. 216

PAGE 217

Discussion of Complex Archery Data Southern Athapaskan Complex Bows Opler (1971:83-92) suggested that the Mescalero Apaches were marginal to the Plains cultures, and adopted the practice of making sinew-backed bows from central Plains societies, via Jicarilla Apache intermediaries. My large Apachean bow dataset neither confirms nor refutes this interpreta tion. One of the two Mescalero bows I saw was similar to Jicarilla bows, while the ot her was more like Arizona Apache and Navajo bows, suggesting that both ar eas may have influenced Mescalero bowmaking traditions. In Tables 5-3, 5-4 and 5-5, the morphological consis tency of Southern Athapaskan complex bows is demonstrat ed. Several features ar e notable. Navajo bows and westerly (non-Jicarilla) Apache bows are remarkably similar to each other, and are rarely painted. They are exclusively paint ed red when they are painted. They have unusual hide-wrapped grips nearly as often as solid-field sinew-wrapped grips. They often have multiple transverse sinew seizings to help hold the backing in place. They are single curved almost as often as they are double curved only sometimes. Jicarilla Apache bows in particular are very consis tent in having a bread-shaped cross-sectional profile throughout, only rarely single curved, and rarely having midlimb transverse seizings. The grip is never hide or cloth, and only very rarely solid-field. Most commonly, the eastern Apachean grip is either multi-banded or barber-pole spiraled sinew (Figure 5-10). This barber-pole sinew spiral te chnique, quite common among Southern Athapaskan weapons, has never been found on any non-Apachean bow in my database. The only exception is a single Sek ani Athapaskan bow from British Columbia identified by Morice (1894:58). This lost Sekani weapon had a similar double-curved 217

PAGE 218

shape with an almost rectangular cross-section; almost rectangular is an accurate description for both the distinctive B and J cross-section codes which are peculiar (alonst exclusively) to Eastern Apache bow s (Figure 5-11). The only major difference from a typical Eastern Apac hean bow is the length which longer than almost any bow I saw from any region. Ther e is one notable exception howeveran extraordinary 73 inch double-curved Apache bow, JES 858, which is closer to the Sekani bow in many respects (Figure 5-12). These bows, A pache and Sekani, could represent a link between the Apache bowyers and the lost glue-type sinew-backed bow-making tradition of British Columbia. Another noteworthy fact is the extrem ely high percentage of Southern Athapaskan sinew-backed bows which have the classic Tur kish double-curved shape (at least 79%, and at least 88% including Apache only and excluding Navajo). The double curved shape is achieved through reflexing a bow with a strongly-curved grip -area, so that the handle retains the original curve and only the limbs are truly reflexed, into something like a flattened M shape (F igure 4-2). When American bows are unbraced, it can be difficult to precisely determine whether t hey are singleor double-curved, so my estimated frequency of the two main longitud inal shape-classes is imprecise (likely undercounting double curves). Of the twenty-one complete complex weapons made by Numic and Puebloan peoples, I estimated doubl e curvature was present about 50% of the time, similar to the frequency among Navajo bows. To reiterate, this may slightly underestimate the real frequency of occu rrence, as unbraced weapons are often ambiguous. 218

PAGE 219

Apache bows are significantly more likel y to have double-curved profiles than are non-Apache bows, and these shapes are less am biguous in nature, that is they are more extremely curved and easier to identify as such regardless of whether the weapon is braced or not. When only eastern Apac he bows are considered, the frequency of double curvature is at least 92%, far mo re often than the occurrence among NonApache bows. The double-curved shape is by no means limited either to sinew-backed bows or to Athapaskan bows; the form is widespread throughout the Great Plains (Bergmann et al. 1988:662). But it is not found among peoples un familiar with sinewbacked bows. I suspect that the reflexed d ouble-curved form occurs on some self-bows for purely aesthetic reasons, in imitation of (or as a vestige of) the shape of the technologically superior complex-bow. Ther e is no technological advantage to doublerecurvature when not other wise accompanied by sinew-backing (Bergmann et al. 1988:662-667). The double-curved shape appears limited to glue-type complex bows and (less frequently) self-bows. Glueless cable-backe d bows are never double-curved; this is understandable, as the free-fl oating cable-backing would not easily conform to the double curve shape, and would likely slip and become detached from the bow and cease to function. At least three of t he four Canadian glue-ty pe sinew-backed bows I know of (from the Columbia Plateau and nort hern Plains) are double curved. So are all of the (lost) Northern Athapaskan glue-type sinew-backed bows illu strated by Morice (1894:57-59; see Figure 5-11). The northernmost example of this shape I have seen is an unusual double-curved Northern-Athapaskan attached-guard self-bow from the upper Yukon region, J ES 365 (Figure 5-13). 219

PAGE 220

The Spanish chroniclers invariably call ed Apache recurved complex bows Turkish bows because of their strong resemblance to Turkish composite bows (LeBlanc 1999:102). This might not be a coincidence, if the linguistic affinity between the Xiongnu and the (Dene)-Yeniseians bears scrutiny. The so-called Turkish bow, a fully developed Eurasian composite bow of the early-Common Era, is actually just the culmination of a millennium of development from late Bronze Age and Iron Age Xiongnu prototype bows, exemplified by those excavated at burials along the Qum-Darya river in the eastern Tarim Basin in East Central Asia. Bows of this type have been spread all over northern Siberia by t he Neo-Siberian pastoralist mi grants (Rausing 1967:143-144; see also Hall and Farrell 2008). Simpler Hunni c prototypes of the Qum-Darya burials are remarkably similar in shape to Apache bows (and other similar Native American bows), moreso than are any other Old World weapons I am aware of (Figure 5-15). The average length of a Turkish bow is 44-46 inc hes (Rausing 1967:144). The mean length of my Southern Athapaskan dataset (67 bows) is 43.9 inches, and the median length is 45 inches. More conclusive proof of the c onnection between Northern and Southern Athapaskan bowyery is found in a rare tr ussed glueless-type sinew-backed bow, probably Apache or Western Navajo, from a cave in Norther n Arizona, now in a private collection in New Mexico (Figure 5-13, par t A). It falls near the median range of Southern Athapaskan bows for all measurements, but has an uncomplicated backing similar to the two Athapaskan sinew-back ed bows from the U pper and Lower Yukon River in my database, NMNH E5588-0 (from the Deg Hitan, see Figure 5-13, part B) 220

PAGE 221

and JES 2687 (from the Alaska Panhandle)it is also rather similar to Siberian bows JES 2649 and JES 437, suggesting possible c ontinuity with Siberian forebears. Additionally, the Arizona bow is unusual among cable-backed bows in that it has an ovate cross-section throughout its entire length. In my observations, such an ovate cross-section for a cable-backed weapon is found on only one other specimen; the very small Deg Hitan Athapaskan boys bow E5588-0 (Figure 5-13, part B). Hanson (1994:181-182) suggest the Ariz ona cave bow may be as old as 1150 CE, but this seems extremely implausible based on its excellent condition, one more in keeping with an ethnographic artifact than a weapon alleged to be one of the oldest complete bows in the entire hemisphere. Further 1150 CE is more than 100 years older than the oldest sinew-backed cave bows in the region, acco rding to LeBlanc (1999:103), dating to the late thirteenth century at the earliest. There is no clear indication of how this date was arrived at, but it appears to be a guess. The only Southwestern peoples known with certainty to use such Arctic-type cabl e-backed bows are Chiricahua and Western Apaches (Gifford 1940:29). Driver and Mass ey also suggest that westernmost Navajo and some western Pueblo groups used gluele ss sinew cabled bows as well, but the source of their data is obscure They comment perceptively: The tying on of sinew cords to reinforc e the bow, characteri stic of the entire Arctic, is also the rule in northeast Siberia. These two areas are actually one, and the sixty mile gap at Bering Straits was negotiated regularly by peoples on both sides. A single origin for the cord reinforced bow is therefore indicated. What about the occurrence of this trait in the southwestern United States? With out the other evidence on bows and arrows this would present a puzzle. However, when we note that two other traits of predominant ly northern provenience, the Mediterranean arrow release and tangential arrow feathering, also crop up again in the Southwest, the case for northern or igin and southward spread becomes stronger. When we add to this the pres ence in the Southwest of a number of Athapaskan-speaking peoples, generally believed to have migrated from 221

PAGE 222

the north, our case becomes still str onger. It looks as if these bow and arrow traits may have been carried s outhward by migrating Athapaskans (Driver and Massey 1957:355). If Driver and Massey are correct and the a rrival of these bow and arrow traits in the Southwest is attributable to the Southern Athapaskan migration, this raises the possibility that the migrations may hav e begun later, commenced much faster, and culminated earlier than is generally believed. Archaeologic al evidence for the spread of complex archery suggests the strong comp lex bow became firmly established in Southern Alaska quite late in the first m illennium of the Common Era, within only a few centuries of its ultimate arrival in Arizona between roughly 1200 and 1450 CE (Dixon et al. 2005; Hare et al. 2004; LeBlanc 1999: 103). More work needs to be done to establish the plausibility of such a rapid migration scenario but the data are suggestive. The scarcity of sinew-backed bows surviving among southern Canadian Athapaskans means that material for dire ct comparison of Canadian Athapaskan bows with Southern Athapaskan bows is extremel y difficult. There are some analogs between a few of the Canadian selfbows I have seen. The Sekani-Beaver-Tsuutina family is one which has been plausibly suggested as including collateral relatives of the Southern Athapaskans (Gunnerson and Gunnerson 1971). The handful of Beaver and Tsuutina self-bows I have seen appear to hav e very close analogs with particular Apache weapons. For example, the single Tsuut ina bow that I know of in a museum collection is AMNH 50/5985B. It has a reasonably close analog in the Oklahoma Apache self bow SNO E/1955/9/21A. Both are thinly-painted r oughly-made self-bows with similar dimensions, D-shaped cross-se ctional shapes, and single side-notched nock configurations. But really not much can be said about the significance of these 222

PAGE 223

similarities, as similar roughly-made self -bows are found throughout the intervening cultures of the Plains (A llely and Hamm 2002; Wissler 1910). More significant may be the similari ty between one unusual Apache bow, MPM 14748, and typical Subarctic self-bows, including Beaver, Slavey and Deg Hitan weapons like AMNH 50.1/7664-A (Beaver), AMNH 50.1/7619-A (Slavey), AMNH 50.1/8284 (Slavey), and FMNH 12477 (Deg Hitan). These five weapons (four Northern Athapaskan and one Apache) are am ong the most robust I have seen. All are thick and rectangular in cross-section across the ent ire length, vary between 60 and 67 inches long, and average 1.44 inches wide at the grip Such massive longbows are normal in the forested Subarctic, but ar e rather unusual in the desert Southwest. Could this also be an indication of some retained northern self-bowmaking tradition among Southern Athapaskans? Unlike the smaller self-bow s mentioned above, thes e simple longbows are not widespread in the intervening space of the Plains, so the anomalous Apache bow stands out among its immedi ate neighbors (Figure 5-16). But the simplicity of this self-bow means that this suggestion of simila rity is also inconclusive; there is not enough design complexity to preclude the possibility of chance resemblance. The similarity is intriguing nonetheless; on first inspection, I would have thought that this robust Apache weapon came fr om the western Subarctic. Pacific Coast Athapaskan and Algic Complex Bows Tables 5-6, 5-7, 5-8 and 5-9 provide summary information for the combined Pacific Coast Athapaskan, Yurok, and Pomo dataset. Table 5-8 shows five fewer bows, because the other tables were expanded slight ly (adding three Hupa bows, one Tolowa bow and one Yurok bow) using publically available information from the online database of the Grayson collection (UMO). This database does not include comprehensive 223

PAGE 224

cross-sectional shape data, and so these five bows are left out of Table 5-8. The HupaYurok cultural continuum in northwest California was the traditiona l center of complex bow manufacture for the entire region, se rving very distant peoples by trade. Reservation period oral traditions attribut ed the most skilled bowyery to the Algicspeaking Yurok (Smith and Barrett 1961). It is possible that Algic and Athapaskan speakers have shared one material culture fo r a very long time; they may have already been deeply culturally engaged in the Columbia Plateau region (where both ancestral stocks reside) before they arrived together in California during the thirteenthor fourteenth-centuries. All of these Pacfic Coast weapons (including one Pomo bow) form a single type-class, which is arguably the most distinctive of any geographic or cultural subgroup of glue-type sinew-back ed bows I have measured. I have previously alluded to the Hupa def erence to the superiority of Yurok bowcraft (conveyed by Bradley Marshall, also see Smith and Barret 1961). But regardless of the reputed difference in skill, it appears from my data that the Pacific Coast Athapaskans (Hupa plus Oregon At hapaskans) were more prolific bowyers than their Algic-speaking counterparts ; Athapaskan-made weapons outnumber Algic-made weapons in these museums by nearly three to one. It is admittedly possible that some Yurok bows have been misidentified as Hupa bow s in museum collections because they also came from the Hoopa reservation (B rizard 1903). But this cannot explain the noticeably different average bow length and width between t he otherwise identical Algicmade and Athapaskan-made bows. The mean measures for length and width appear significantly longer for Yurok weapons than for Athapaskan ones. Bows classified as Yurok are an average of three inches longe r and almost half an inch wider than Pacific 224

PAGE 225

Coast Athapaskan bows. Also, PCA bows were slightly more consistent in their crosssectional profile and nock formation than Yurok bows. Only L and D shapes were observed in the PCA dataset, while unusual h ybrid L/Q shapes appeared twice in the much smaller Yurok pool. Hupa nock-design appears more consistently faithful to the type, invariably including the distinctive shar ply recurved tips, typical for this class of weapons. In contrast, at least three of the Yurok bows adopted a much less distinctive and technologically simpler side-notched nock, lacking the sharp recurvature of the typical bow of this class. But the vast majority of Hupa and Yurok bows are more or less indistinguishable. It is only as populat ions that they appear to diverge slightly. Athapaskan-Numic Cultural Reticula tion and Complex Archery Diffusion Most of the vast intervening spac e between Southern and Pacific Coast Athapsaskan territory is the traditional reside nce of Numic (Uto-Aztecan) speakers. The Numic expansion is, along with the Athapaskan expansion, one of the two largest North American phyletic expansion events of the mid-Common Era. It is difficult to determine which began first. The Numic spread was sout h to north, in contrast to the north to south direction of the Athapa skan. But the two expansions were largely coterminous, with multiple temporal and geogr aphic interfaces. Before discussing the Numic data for complex archery distribution, we should be cognizant of Moores (1994) discussion of ethnogenesis as necessarily involving multip le antecedent societies. The NumicAthapaskan interface is a relevant case in point. Ambiguity surrounds the identification of early northern Numic versus early Sout hern Athapaskan cultural remains, including archery artifacts. Most of recorded hi story has seen varying degrees of mutual hostilities between Southern Athapaskans and va rious Numic groups (including Utes, Paiutes, Shoshones and Comanches). But ther e is reason to suspect that very early 225

PAGE 226

Apachean groups included lar ge contingents of bilingual Numic speakers who adopted Apachean lifestyles. The early Southern Athapaskan speakers of the Canadian River (OK, TX, NM), the ancestors of modern Ji carilla and Eastern Navajo peoples, were commonly referred to as Iyutta-jenne, a tr anscription of Ute-Dene, essentially NumicAthapaskans (Hyde 1959:39, 95). The DNA evidence suggests that this relationship was primarily one-way; Southern Athapaskans absorbed and assimilated large numbers of individuals from the Great Basin and Southwest, but local Southw estern groups do not show traces of significant DNA admixture with Athapaskans (Malhi et al. 2009:203). This is an understandable result of dem ographic imbalance of t he two founding groups, suggesting that Numic speakers may have alr eady been well established in the region when the Athapaskans arrived. If the hy pothetical proto-Apachean founding population was extremely small, and dependent on in-marri age to maintain demic integrity, then genetic drift would tend to erase the si gnature of out-marriage. But two-way technological exchanges would not be subject to the same constraints. It is possible that the technological impact of t he Athapaskans was much greater than the genetic impact. Tables 5-10 and 5-11 show the length dat a and cross-sectional shape profiles for the eleven complete Numic complex bows I measured. In star k contrast to the remarkable consistency of the much lar ger Southern and Pacific Coast Athapaskan data blocs, there is little evi dence of overarching typological unity of the Numic dataset. But this lack of consistency can be attributed to the fact that the samp le size (11) is too small to facilitatie meaningf ul interpretation. There is some consistency in Shoshone cross-sectional bow shape (which is often L or Q). The high frequency of the B shape 226

PAGE 227

among Ute bows (3 out of 3 Ut e midlimbs are B shaped) and the ubiquity of J, B, and R, is most similar to Jicarilla Apache bow s, supporting Hydes (1959:39) assertion that the Utes and early eastern Apaches formed one community in northeast New Mexico and the Texas panhandle. Numic speakers may well have been involved in the spread of complex archery, but Numic involvement in this process must have involved complicated back-and-forth relations with Calif ornia, Plateau, and/or Plains societies, because the general diffusion gradient of sinew -backed bows was assuredly upstream of the northward thrust of the Numic migration. Great Basin Archaeological Bow Fr agments: Numic or Athapaskan? Promontory Point, Utah Archaeological evidence is wo rth bringing up here. Rela tively little evidence of prehistoric sinew-backed bows has been uncover ed in the Great Basin. For what has been found, there is some question of which cultures are represented. The earliest examples were the bow fragments excavated by Julian Steward (1937) from a cave-site at Promontory Point, in northern Utah, on the margins of the Gr eat Basin. This site is located in modern Shoshone territory, but the late prehistoric Promontory peoples have very little in common with Shoshone, and mu ch more in common with northern hunting peoples like Athapaskans. The presence in particular at Promontor y Cave of northernstyle four-piece moccasins, mittens and sinew-ba cked bows is indicative of cultural ties with Subarctic Canada and/or the Columbia Plateau. Steward writes: it is a safe conclusion that the Promont ory culture is definitely not Shoshoni as the latter is now understood . t he impression one gains from studying the material is that the culture is basically one of a northern hunting people . This possibility, however, at once suggests that we may have remains left by one of the Athapascan speaking tribes who, during their southward migrations, acquired puebloan traits fr om the tribes they are presumed to have driven out of the Norther n Periphery (Steward 1937:86-87). 227

PAGE 228

Stewards suggestion of the At hapaskan identity of the la te prehistoric phase at Promontory was taken seriously and support ed by a number of workers through the 1960s, who further connected Promontory (a nd Athapaskans) with the Fremont culture of the Great Basin, and with the Dismal River culture of the Plains (Aikens 1966; Gunnerson 1956). However, alternative t heories of the Souther n Athapaskan migration came into vogue in the 1970s, and Stewards theory fell out of favor late in the century. In a typical statement of this view, Davi d R. Wilcox states: Promontory Point and Fremont are indigenous developments in Utah and that Stewards Promontory Point culture is merely a late regional manifestat ion of Fremont culture (Wilcox 1981:217). But more recent work has largely vindicat ed Stewards early claims of the Athapaskan identity of the Promontory people, particu larly through analysis of moccasin design and petroglyphs which suggest strong ties to Subarctic Canada. Ives and Rice note: Pictographs at Grotto Canyon in s outhern Alberta reveal dancers in unequivocally Southwestern styles. T he extraordinarily preserved remains of the Promontory Caves, situated near a zone of cult ural interaction at the northeastern periphery of the Great Basin, include dozens of soft-soled moccasins made in styles used by northern Athapaskans, as well as unique Subarctic scraping tools (chi-thos) (Ives and Rice 2006:1). Athropologists have long speculated t hat Athapaskans may have introduced northern footgear designs to various regi ons including the Plateau and Southwest (Salwen 1960; Teit 1900:211). Unfortunately Stewards fa mous Promontory sinewbacked bow fragment (Steward 1937:18) was never dated, was not evidently accessioned by the museum (UMNH), and has since been lost (John Ives, personal communication 27 Aug 2010). Stewards orig inal drawing of the bow fragment is reproduced as Figure 5-17. Another ear ly Promontory Cave bow fragment (UMNH 42Bo1 11602.2) has recently been 14C dated (along with the Canadian-style moccasins 228

PAGE 229

and other organic remains) using accelerator mass spectrometry. The results have not been formally reported yet, but in personal communication John Ives (25 Feb 2011) tells me: We have about 35 AMS dates for peris hables in the caves; with the single exception of a recent Shoshone winnowing ba sket, the perishables are coming in tightly focused on the AD 13th century. This would seem to indicate the earliest presence of Athapaskans in the Promontor y Caves was about 800 years ago, in perfect agreement with the earliest sinew-back ed bow dates in the southwest mentioned by LeBlanc (1999:103), although it is not clear where exactly LeBlancs dates come from. No sinew fibers survive on the specimen UMNH 42Bo1 11602.2, but there are two morphological features which indicate the lik elihood that it once had a sinew-backing. In common with ethnographic Shoshone sinew backed bows from the same region, there is evidence of a lateral (transverse) in cision in the belly side, less than ten inches from the tip (Figure 5-18). The belly app ears to have been thinned toward the grip, below this major incision, but the fragment ary nature of the bow means we cannot say how much thinning was done. Toward the ti p, above this major incision, there are numerous minor parallel incisions in the belly surface, lateral (transverse) and perpendicular to the long dimension of the bow. I have only ever seen this treatment on the tips of sinew-backed bows made by Numi c speakers. From my observations, I suggest that the tips are left thicker than the adjacent bowlimb to resist breakage under tension, and the bowlimb is thinned to increase flexibility and decrease overall weight. The lateral incisions are in place to help secure the transverse tip seizings. While such transverse sinew bands are pr esent on the tips of some self-bows, none require this deeply textured surface because they are simp ly present to strengthen the tips. In 229

PAGE 230

contrast, on sinew-backed bows these wrappings are there to secure the sinew-backing itself, under extreme tension a nd wrapped endwise around the tip, behind the lateral banding. This configuration is confi rmed when one examines a Shoshone bow with deterioration of the sinew on the tip, revealing the longitudinal backing strips secured by the transverse banding over these parallel lateral incisions (Figure 5-18). This Promontory bow fragment could be the technological antecedent for this aspect of Shoshone bowmaking, but I would st op short of saying that the Promontory bow is a Shoshone bow. It has a very flat belly su rface; it simply doesnt look like any of the Shoshone bows I have seen. Four out of five (80%) of the Shoshone bows I measured had deeply rounded belly surfaces (L, Q or J shape), as did seven of eleven, (64%) of the Numic bows. In contrast, at least 81% of the Southern Athapaskan bows in my sample had flat belly surfaces (B, D, or R shape), and thos e with rounded bellys appear often only very subtly rounded (e.g. hyb rid D/L or J/R shapes). In terms of cross-sectional shape, this Promontory bow is more like an Apachean bow than a Shoshone one, while in terms of tip-treatment, it is mo re like a Shoshone bow than an Apachean one. It is possible that the t he Promontory people prov ided ethnogenetic and technological antecedents to both the Southern Athapaskans and the northernmost Numic peoples, who where not simply ships passing in the night, but collaborators in each others formative cultural development. Lovelock Cave, Nevada Another important Great Basin archaeological site is Lovelock Cave in northwestern Nevada. At the Hearst Museum in Berkeley I observed a small sinewbacked bow tip-fragment from this site ( PAH 1/21418; Figure 5-19), with sinew fibers still attached (Loud and Harrington 1929: plate 47). There is no record of an 230

PAGE 231

Athapaskan presence in the r egion, but Steward (1937:85-86) thought that Lovelock and Promontory peoples could have been harbringers of t he same (Athapaskan?) northern hunting culture, indicated by similar hafted knives and similar textiles, including birdskin blankets and matting ma terials. Matting from Lovelo ck Cave, especially of the late period, is much like Promontory matti ng (Steward 1937:85). A further link between Lovelock quillwork textiles and similar Northern Athapaskan textiles is highlighted by Galina Dzeniskevich, who further suggests t hat this quillwork tradition may have been introduced from Siberian reindeer cultures around the middle of the Common Era: Noteworthy among the examples of possible early borrowing is the ornamentation of Athapaskan buckskin clothing with split and flattened porcupine quills. That t here are common elements in the technique of decorating with reindeer neck hair am ong the peoples of Siberia and with porcupine quills among t he Athapaskans is indisputable. As some researchers have pointed out, such an involved ornamentation technique could hardly have develope d independently, and since the reindeer hair ornamentation is probably the more ancient one, the borrowing in this particular instance came from Asia. It is still impossible to ascertain when porcupine quill ornamentation first appeared among the Athapaskan Indians. The only available evidence c onsists of some fragments of clothing with porcupine quills sewn onto them that were found in Lovelock cave, Nevada, in 1929 (Dzeniskevich 1994:56). A final element connecting Lovelo ck peoples with Apaches and Northern Athapaskans is the presence of longitudinally grooved arrowshafts. Called lightning grooves or blood grooves, these grooves are alleged to allow the blood to run out along the shaft, away from absorbient fur (or cl othing), creating a blood trail to assist in tracking the hunters (or warriors) wounded quarry (Figure 4-8). These grooves are atypical of Great Basin cultures, but typical of Northern and Southern Athapaskan cultures including Sekani Athapaskans and Ji carilla Apaches, and other Plains Indians (Loud and Harrington 1929:97-98; Mason 1907:92; Morice 1894:55). 231

PAGE 232

Northern Paiute oral tradition, as re corded by Sarah Winnemucca Hopkins in 1883, suggests that the late phase of Lovelo ck Cave culture was occupied by ferocious foreign barbarians during the lifetime of Hopkins third or fifth great grandfather (Hopkins 1883:74). These so-called barbari ans were unrivalled archers and warriors who migrated from far away, down the Humboldt River. T hey were despised by the Paiutes for their alleged cannibalism, and for re fusing to adopt Paiute lifestyles. After a three-year war, they were ultimately annihilated when the Paiutes in one evening surreptitiously filled the Lovelock cave ent rance with firewood and brush, and set their enemies dwelling ablaze while they slept (Hopkins 1883:73-75). Could these tales be describing an aborted Athapaskan effort to expand into the western Great Basin? The migration of At hapaskan speakers from the Plateau to the Pacific Coast of Oregon and California similarly may have involved westward progression down various river valleys (Jacobs 1937:61-62). The sinew-backed bow fragment (PAH 1/21418; Figure 5-19) is much too small to be definitively typed, but it shows a D-shaped cross-sectional shape and small size consistent with smaller Pacific Coast Athapaskan complex bows. Yet it is also perfectly consistent with one of the three Paiute bows I have seen. Once again, it proves very difficult to distinguish between Athapaskan and Numic material culture. Northern Athapaskan Complex Bows Denaina Bows The number of Northern Athapaskan comple x bows in museum collections is relatively few, and all of them are of the glueless type. But it may be possible to draw some conclusions about the spread of the te chnology from the ones that do exist. An outright majority of the surviving Norther n Athapaskan complex bo ws are Denaina in 232

PAGE 233

manufactureat least five such specim ens exist (YPM 15844, JES 2691, PAH 2/6361, PGM 2667-17 and PGM 2667-20). Furthermore one Alaskan bow of uncertain provenance (MPM 576) closely resembles the Denaina bow YPM 15844, and another similar one (JES 2681) is from the Kenai Pe ninsula and could be either Eskimo or Denaina. Denaina have been t he principal residents of Kenai in the post-contact era. Although the aboriginal Kenai Eskimo band (know n as Unixkugmiut) is thought to have dissolved before the Russian period, later synthetic bands of Chugach and Koniag Eskimos returned to the western tip of the peninsula to trade with the Russians there (Townsend 1979:161). Denaina bow YPM 15844 was made using Murdochs (1885) southern style of cable-backing (Osgood 1937:87). But JES 2681 (like most Denaina bows) uses the Kodiak style of backing. This style-name I have not found in the literature, but only on occasion listed in mus eum catalogs. Direct observation suggests that the Kodiak style is roughly inte rmediate between Murdochs southern and western styles. The western style is found often in Siberian Yuit bows, and the southern style is found more often on Alaskan Yupik bows (Figure 5-20). The general opinion of scholars has been t hat Denaina sinew backed bows were borrowed wholesale from their Eskimo nei ghbors. Osgood believ ed that the Denaina only adopted the sinew-backed bow in the last few centuries: That part of the Tanaina area closes t to the Kaniagmiut had an obviously Eskimo-like material culture. This is shown by the addition of sea-mammals to the food supply . the use of Eskimo type outer garments, the dance house, the kaiak and umiak, harpoons with floats, the sinew-backed bow, the stone lamp, and the absence of birch bark. These developments probably occurred within the last fe w hundred years (Osgood 1933:716). Likewise, Dzeniskevich (1981:124-125) argues that the typical Athapaskan self-bow with attached wrist guard was the original bow used by the Denaina, who moved into 233

PAGE 234

their present territory no earlier than the first few centuries of the second millennium CE. Dzeniskevich suggests tha that only with t heir eventual adaptation to maritime hunting around Cook Inlet (in the last few hundred y ears) did they adopt the sinew-backed bow from the Eskimos. These attitudes are em blematic of the perva sive anthropological stereotype, which says that Athapaskans are always promiscuous culture-borrowers and Eskimos are the opposite. As Richard K. Nelson put it: Eskimos occupying an inland boreal fore st environment, ma intaining a long history of peaceful contacts with nearby Koyukon Indians, . have remained overwhelmingly Eskimo. By contrast, whenever Athapaskans have experienced similar contacts they have tended to lose their identity, not just borrowing but undergoing full-scale acculturation (Nelson 1974:49) These views are conditioned by the assump tion that all of t he complex western Alaskan culture traits are essentially (or at least primarily) Eskimo in derivation while the simpler interior culture traits are essentially Athapaskan. But ethnographic reality is not so cut and dry. Joan Townsend is one anthropologist who has strenuously objected to the scholarly imposition of the Eskimo-Indian dichotomy in southwest Alaska: What hostilities and animosities that may exist today between Indians and Eskimos are couched in a Western Indian/Eskimo paradigm which has been imposed from outside within the last 150 years. If these hostilities reflect past attitudes, those, past atti tudes were not in terms of Eskimos versus Indians but were in terms of loca l, individual societal feelings at one particular time about another individual society. "Indianness" or "Eskimoness" was not within the concept ual framework. I want to stress that it is people that interact on a so cial and, more significant, on a local (usually village) level, not on a mor phological/ genetic or linguistic phylum level (Townsend 1979:178). Given that the Athapaskan expansion closely coincided with the Neo-Eskimo expansion, which in turn coincided closely to the reintroduction of sinew-backed bows to North America, we must now consider w hether the mixed Eskimo-Athapaskan cultures of Alaska were realy equal partners in the spread of th e newly introduced technologies. 234

PAGE 235

My data does not support the notion that Denaina have simply borrowed their sinewbacked bows from their Eski mo Neighbors. Three of t he five surviving Denaina complex bows which can be positively i dentified (JES 2691, PAH 2/6361, and PGM 2667-20) each posess quintessentially Athapask an attached wrist guards in addition to complex sinew backings expertl y applied. At the very le ast, this fact indicates a seamless adaptation of this technol ogy to Athapaskan cultural norms, not slavish imitiation. And I have documented weapons collected from the Aleut of St. Pauls Island and the Koniag of Kodiak Island, both utilizin g the Athapaskan-style attached wrist guard (YPM 10054 and JES 366). No evidence justifie s the assumption that the technological borrowing was entirely unidirectional. Iv es (2010) suggests t hat understanding the dynamic cultural interface along the vast Eskimo-Athapaskan frontier is crucial to understanding the Athapaskan expansio n. He points out that this is analogous to the situation in ancient Central Asia, where equestrian technologies were spread along the vast frontier of Scythian-Alta ic bilingualism in the Eurasi an steppes. For most of the technologies of the ste ppes, it is incorrect to claim t hat they were either Turkic or Persian; they were essentially both at the same time. Likewise this is the case with Indian and Eskimo complex archery technology in southwest Alaska. Other Northern Athapaskan Bows There are at least three examples of (non-Denaina) Northern Athapaskan complex bows in museum collections that I am aware of (RBCM 6571, NMNH E5588-0, and JES 2687). They are from vastly differ ent regions. JES 2687 was collected in the late nineteenth-century boom-t own of Haines, Alaska very near the upper Yukon River. This is a region of high Athapaskan language density, so it could have originated from 235

PAGE 236

one of several different Athapaskan speaking groups. The boys bow, NMNH E5588-0 is Athapaskan-attributed and was collected along the lower Yukon (Figure 5-14 B.). It was collected in 1867 by the naturalis t William Dall, and accessioned to the Smithsonian during the following year. It is functional, but at just 23 inches long, it would work for hunting only mice. I suspect it is of Deg Hitan manufacture, because Dall (1897:67) describes a scene of Deg Hitan children hunting mice with miniature bows and arrows, and bartering wit h the author for candy and trinkets in exchange for their quarry of rodents. Although Dall does not specifically mention acquiring one of these bows, I strongly suspect that this particular bow was purchased around this same time. Just like JES 2687, NMNH E5588-0 is backed with hide thongs rather than sinew. Although one is a tiny miniature, both bows have similar proportions (as though one is a scale model of the other); the configuration and manner of affixing the hide backing cables is similar enough to suggest that t hey both originate from the same Yukon Athapaskan bowmaking tradition. The last Northern Athapaskan complex bow (RBCM 6571) is probably Gwichin, acquired in the Mackenzie River region. Th is is another place, like western Alaska, where the Gwichin are generally assumed to have borrowed their complex bows from the Eskimos. This assessment owes lar gely to the dearth of complex bows in neighboring interior Athapaskan territory to the south and west, and in the northern Canadian Cordillera (McClellan and Dennist on 1981:378). This bow is backed with sinew (not hide, like the other two bows), and the backing design is not particularly distinctiveit is generally si milar to the configuration of the other two Athapaskan bows, but, not identical due to different material constriants. Ho wever some features do seem 236

PAGE 237

to link it to the southern Yukon bow (JES 268 7). The bows are similar lengths (one is 46 inches, the other 48.5 inc hes), and nearly identical width and thickness, maximum at the grip (both are near 1. 25 by 0.8 inches), and both have rounded recangular crosssections (lacking sharp edges). But most significantly, these two far-flung Athapaskan bows have a ~0.25 inch wide blackened trough-lik e channel running t he entire length of the belly surface of the bow, along the midline (Figure 5-21) I do not recall seeing this feature on any of the dozens of Eskimo complex bows I have examined, suggesting it may be a trait common to Athapaskan complex bows. In summary, these three Northern Athapaskan complex bows (RBCM 6571, NMNH E5588-0, and JES 2687) appear all to be typologically more similar to each other than any one of them is to common varietie s Eskimo bows. They are from far-flung extremes of the Athapask an range (the upper Yukon, th e lower Yukon, and the Mackenzie River, respectively). This suggests that the Athapaskans in all three sectors of the Northern Athapaskan bloc have inher ited bowmaking traditions from a common ancestral source rather than borrowing it peacemeal from neighboring Eskimo societies. This is not necessarily a direct descendant of proto-Athapaskan bowmaking tradition however, as the groups in question maintained long-distance contacts and could transfer technology to each other at various times. For exampl e, it is possible that the Deg Hitan lost their complex bowmaking tradition sometime in the past only to have it be subsequently reintroduced by Gwichin intermediaries. Osgood (1940:201-202) gives gi -te as the Deg Hitan word for bow, but only the second element in the compound (-handle) was intelligible to his informant. The first element is an obscure archaism (at leas t from the Deg Hitan per spective). To refer 237

PAGE 238

to a sinew-backed bow, one must add a modifier meaning to tie: gi -te a e en. But the Proto-Athapaskan root word for the compound gi -te already means sinew-backed bow; the Deg Hitan have forgotten the meaning of the element gi and thus added a redundant modifier to the term; th us the technical meaning of gi -te a e en is sinewbacked-sinew-backed bow. I asked Victor Golla if this possibly meant that the Deg Hitan may have lost the knowledge of si new-backing technology at some point, and then regained it, resulting in this lexical redundancy. He told me that the initial g-sound is irregular in Deg Hitan, but regular in Gwichin, so the term seems to have been borrowed down the Yukon River in recent times with the meaning bow of any kind (Victor Golla, personal communication 10 Dec 2010). This suggests indeed that the specific details of Northern Athapaskan bow technology have been subject to change through time, and that the Gwichin terms for bows (and maybe the bows themselves) could have been transferred to other Athapaskan groups throughout the Yukon Basin. This provides another interpretive framew ork for my observations about the remarkable consistency in Northern Athapaskan complex bow morphology among the far-flung Gwichin, Deg Hitan, and diverse upper Yu kon peoples. But a sample of only three weapons means the strength of these conclusions is somewhat limited. Summary Although the Norther n Athapaskan complex bow data pool is too small to draw firm conclusions, the few data that I have collected are still surprisingly suggestive considering their numerical limitations. Wi th the notable excepti on of the glue-type bow region of the southern Canadi an Cordillera where no sinewbacked bows have survived, my dataset includes representative samples from virtually every major ethnographically known center of Athapaskan complex bow pr oduction in the north. The lack of complex 238

PAGE 239

bows in vast stretches of Northern Athapask an territory has caused some to suggest that the technology was not originally Athapaskan, but me rely borrowed from Eskimos on the periphery of Athapaskan territory. The morphological consistency of non-Eskimo complex bows suggests this might not be the case; the existence of a distinct, early Athapaskan complex bowmaking tr adition remains a viable hypothesis, with extensive areas of the Athapaskan heartland subsequently losing this craft. It could simply be that hunting in the subarctic boreal forest environment does not demand the superior complex bow technology, and so the Athapaskans who settled t here lost the art of its manufacture. The use of a powerful complex bow in wooded terrain was often unnecessary, and generally not preferred by Northern Athapaskans (Dzeniskevich 1981:124-125). When a diversified subsistence strategy places specialized labor at a premium, and where self-bows are all that is required for effective subsistence hunting, then why would anyone bother to retain so complicated and time consuming a manufacturing process? Bowhunting in the far north can be extremely tedious on the best days, with only marginal returns on a cons iderable investment in labor (Stefnsson 1913:503). This would mean that the time invested in sinew-backing would be counterproductive, unless interpersonal strife was the root cause. As Roland Bohr writes: The subarctic is not an environment t hat favors archery. Primer raw materials for bowmaking . are not ava ilable in the region. . Native peoples of the subarctic di d not generally adopt sinew backing. Even if the tensile strength had been accommodated by some form of sinew backing, bows would still break in the extreme cold, because the wood cells on the belly would collapse when the bow was drawn far enough to launch the fairly long arrows that were co mmon throughout the region (Bohr 2005:193195). This point is further emphasized by Foley Benson: 239

PAGE 240

240 The bow was utilized by the Kutchin as only one of an array of hunting techniques applicable to any game resour ce. It is possible that given the other hunting techniques, the bow might have been dispensed with altogether. When the bow was used, it was often a supplemental tool. When it was incorporated as the primary hunting weapon, no demands were made on it for accuracy and traj ectory beyond about 30 yards. It appears that there was little need fo r a high trajectory weapon [like a complex bow] in the Kutc hin economy. In fact, such a tool may have been maladaptive in the long term (Benson 1975:40). Collectively, these assessments serve to af firm that the general lack of complex bows in the Northern Athapaskan heartland was not reflective of some original primitiveness on the part of Athapaskan speake rs, but was simply a natural outcome of their cultural ecological circumstances. In Chapter 6, I will address the implications of data for Athapaskan complex archery in the context of a proposed cross-disciplinary synthesis suggesting a substantially revised chronology for the Athapaskan expansion.

PAGE 241

Table 5-1. Objects whose provenances were substantiall y revised or corrected as a result of this study. Museum code and catalog number Object description Cultural affiliation in museum catalog Corrected cultural affiliation Comparable objects as basis for correction Notes GMA 7336.209 painted selfbow Unknown California, Hupa-Yurok or Modoc MAC 2000-0214 (Hupa-Yurok, made in Modoc style) Paint similar to Hupa-Yurok sinewbacked bows. GMA 8426.1895b sinewbacked bow Unknown Apache, likely Numerous Many features consistent with typical Apache bows NMAI 9/8154 sinewbacked bow painted red Possibly Navajo AZ Apache, Isleta Pueblo or northwest Navajo YPM 14368 (Apache, AZ) NMAI 19/3088 (White Mountain Apache, AZ) JES 4455 (Isleta) Shape is atypical for Navajo. Red paint is rare on Navajo sinew-backed bows (except in northwest) NMAI 25/1667 sinewbacked bow Arctic Northern California, Hupa, likely YPM 17242 (Hupa) PAH 1/11617 (Hupa) Glue-type sinewbacked bows unattested in Arctic PAH 1/14608 painted sinewbacked bow Northern California Hupa, likely YPM-17243 (Hupa) Unusual paint and sinew application WFU 1984.E.0437 and WFU 1984.E.0668 plaited splint basket Athapaskan, Subarctic Northwest Coast, likely Salish UFL P-1593 through P-1599 (Salish) Plaited splint basketry unattested for Athapaskans, but known for Salishans and Wakashans 241

PAGE 242

Table 5-2. Objects whose provenances were substantially revised or corre cted as a result of this study. Museum code and catalog number Object description Cultural affiliation listed in museum catalog Corrected cultural affiliation Comparable objects as basis for correction Notes UFL P-530 sinewbacked bow Unknown Copper Eskimo MPM 33610 (Copper Eskimo) UFL P-530 is larger than MPM 33610, but is otherwise nearly identical YPM 15005 sinewbacked bow Mexico Central California, Miwok or Wintu AMNH 50/6439 (Miwok) PAH 1/71850 (Miwok or Wintu) Miwok bows are very unlike Mexican (southern Apache) sinew-backed bows YPM 30856 sinewbacked bow Seminole, Florida Plains or Southwest, possibly Navajo MAC 1995-0661A (Northern Navajo or Plains) Sinew-backed bows are unattested in the Southeast YPM 145233 sinewbacked bow Unknown Apache, likely Numerous Grip, tips, paint and cross-section all consistent with typical Apache bows YPM 145286 sinewbacked bow Southwest Apache MAC 1991-0863 (Apache) Rare red incised rectangles on sides 242

PAGE 243

Table 5-3. Southern Athapaskan bow morphology. Longitudinal shape Cross-sectional shape code (Figure 4-3). Ethnic Group sample number single curve double curve not sure B D J L O R Navajo 16 7 8 1 (0.5)-(0.5)-(0) (10.5)-(11)-(6.5) (0)-(0)-(0 ) (0)-(0)-(0) (5)-(4.5)-(9.5) (0)-(0)-(0) % 100 44 50 6 3-3-0 66-69-41 0-0-0 0-0-0 31-28-59 0-0-0 Jicarilla-Apache 12 1 11 0 (12)-(12)(11.5) (0)-(0)-(0) (0)-(0)-(0.5) (0)-(0)-(0) (0)-(0)-(0) (0)-(0)-(0) % 100 8 92 0 100-100-96 0-0-0 0-0-4 0-0-0 0-0-0 0-0-0 Non-Jicarilla Apache 8 2 5 1 (1.5)-(1.5)(0.5) (4.5)-(4.5)-(3.5) (1)-(1)-(0) (0)-(0)-(0) (1)-(1)-(3) (0)-(0)-(1) % 100 25 62.5 12.5 19-19-6 56-56-44 12.5-12.5-0 0-0-0 12.5-12.537.5 0-0-12.5 Apache: Subgroup not Specified 31 1 29 1 (12.5)-(17.5)(7) (6)-(5)-(3) (10)-(6)-(3) (0.5)-(0.5)(0) (1)-(0)-(3) (1)-(2)-(15) % 100 3 94 3 40-56-22.5 19-16-9.7 32-19-9.7 2-2-0 3-0-9.7 3-6-48 Total Apache 51 4 45 2 (26)-(31)-(19) (10.5)-(9.5)-(6.5) (11)-(7)-(3.5) (0.5)-(0.5)(0) (2)-(1)-(6) (1)-(2)-(16) % 100 8 88 4 51-61-37 20-19-13 22-14-7 1-1-0 4-2-12 2-4-31 Navajo + nonJicarilla Apache 24 9 13 2 (2)-(2)-(0.5) (15)-(15.5)-(10) (1)-(1)-(0) (0)-(0)-(0) (6)-(5.5)(12.5) (0)-(0)-(1) % 100 38 54 8 8-8-2 63-65-42 4-4-0 0-0-0 25-23-50 0-0-4 Total Southern Athapaskan 67 11 53 3 (26.5)-(31.5)(19) (21)-(20.5)-(13) (11)-(7)-(3.5) (0.5)-(0.5)(0) (7)-(5.5)(15.5) (1)-(2)-(16) % 100 16.5 79 4.5 40 47 28 31 31 19 16 10 5 1 1 0 10 8 23 2 3 24 243

PAGE 244

Table 5-4. Southern Athapask an bow material components. Ethnic Group number of bows sinew grip: solid sinew grip: multiple sinew grip: barber sinew grip: combo sinew grip: form obscure hide grip cloth grip unwrapped grip unknown grip Navajo 16 7 1 0 0 1 4 1 2 0 Jicarilla Apache 12 3 1 3 0 0 0 0 3 2 Non Jicarilla Apache 8 1 1 1 0 0 3 0 2 0 Apache: Subgroup not Specified 31 3 18 5 2 0 2 0 1 0 Total Apache 51 7 20 9 2 0 5 0 6 2 Total Navajo plus non Jicarilla Apache 24 8 2 1 0 1 7 1 4 0 Total Southern Athapaskan 67 14 21 9 2 1 9 1 8 2 Table 5-5. Southern Athapaskan bow length measurements (inches). Ethnic Group sample number mean length maximum length minimum length median length modal length Navajo 16 43.38 46 39 44 45 Jicarilla Apache 12 45.38 52 41 43, 45 45 Non Jicarilla Apache 8 43.09 46 39.5 43 46 Apache: Subgroup not Specified 31 44.62 73 39.5 43 45 Total Apache 51 44.1 73 39.5 43 45 Total Navajo plus non Jicarilla Apache 24 43.23 46 39 44 46 Total Southern Athapaskan 67 43.91 73 39 43 45 244

PAGE 245

245 Table 5-6. Pacific Coast Athapaskan, Yurok and Pomo bow length and maximum width (inches). Ethnic Group number of bows mean length maximum length minimum length median length modal length mean width at midlimb maximum width at midlimb minimum width at midlimb median width at midlimb Hupa 24 37.47 49.5 26.5 37.5 39 2.01 3.5 1.13 1.95 Tolowa 4 37.05 38.5 36 37 37 1.78 1.5 1.88 1.9 Tutuni (Rogue River) 1 32 32 1.25 1.25 Total Oregon Athapaskan 5 36.04 38.5 32 37 37 1.67 1.88 1.25 1.9 Total Pacific Coast Athapaskan 29 37.22 49.5 26.5 37 39 1.95 3.5 1.13 1.9 Yurok 10 40.13 53.25 27.5 40.5 41 2.34 3.19 1.69 2.3 Pomo 1 37 37 2.8 2.8 Total PCA type, non Athapaskan 11 39.84 53.25 27.5 40 41 2.39 3.19 1.69 2.4 Total PCA type 40 37.94 53.25 26.5 38 37 2.07 3.5 1.13 2 Table 5-7. Pacific Coast Athapaskan, Yurok and Pomo paint colors Ethnic Group paint present paint absent finished paint absent unfinished red and blue red and black orange and blue orange and black 3 color red, blue and black Spiral paint on both sides Hupa 19 4 1 12 3 3 0 1 2 Tolowa 4 0 0 3 0 0 0 1 0 Tutuni (Rogue River) 0 1 0 0 0 0 0 0 0 Total Oregon Athapaskan 4 1 0 3 0 0 0 1 0 Total Pacific Coast Athapaskan 23 5 1 15 3 3 0 2 2 Yurok 8 2 0 4 0 3 1 0 0 Pomo 1 0 0 1 0 0 0 0 0 Total PCA type non Athapaskan 9 2 0 5 0 3 1 0 0 Total PCA type 32 7 1 20 3 6 1 2 2

PAGE 246

Table 5-8. Pacific Coast Athapaskan, Yuro k and Pomo cross-sectional bow shape. See Figure 4-3 for shape code legend. Ethnic Group number of bows L D Q Hupa 21 (20.5) (20.5) (8) (0.5) (0.5) (13) (0) (0) (0) % 100 989838 2 2 62 0 0 0 Tolowa 3 (2.5) (3)(0) (0.5) (0)(3) (0) (0) (0) % 100 83100 0 170 100 0 0 0 Tutuni (Rogue River) 1 (1) (1) (0) (0) (0) (1) (0) (0) (0) % 100 100 100 0 0 0 100 0 0 0 Total Oregon Athapaskan 4 (3.5) (4)(0) (0.5) (0)(4) (0) (0) (0) % 100 87.5 100 0 12.5 0 100 0 0 0 Total Pacific Coast Athapaskan 25 (24) (24.5) (8) (1) (0.5) (17) (0) (0) (0) % 100 969832 4 2 68 0 0 0 Yurok 9 (7) (7.5) (2) (1) (0.5) (7) (1) (1) (0) % 100 788322 116 78 11110 Pomo 1 (1) (1) (0) (0) (0) (1) (0) (0) (0) % 100 100 100 0 0 0 100 0 0 0 Total PCA type non Athapaskan 10 (8) (8.5) (2) (1) (0.5) (8) (1) (1) (0) % 100 808520 105 80 10100 Total PCA type 35 (32) (33) (10) (2) (1) (25) (1) (1) (0) % 100 919429 6 3 71 3 3 0 Table 5-9. Pacific Coast Athapaskan, Yurok and Pomo bow grip treatments. Ethnic Group number of bows hide grip: solid hide grip: spiral cloth grip string grip unfinished/ unwrapped rabbit fur trim Hupa 24 2 17 3 0 2 3 Tolowa 4 0 4 0 0 0 0 Tutuni (Rogue River) 1 0 0 0 1 0 0 Total Oregon Ath. 5 0 4 0 1 0 0 Total PCA 29 2 21 3 1 2 3 Yurok 10 1 7 0 0 2 0 Pomo 1 0 1 0 0 0 ? Total PCA type non Athapaskan 11 1 8 0 0 2 0 Total PCA type 40 3 29 3 1 4 3? 246

PAGE 247

Table 5-10. Numic bow length and width distributions. Ethnic Group number of bows mean length maximum length minimum length median length constricted at grip N. Paiute, Bannock, ID 1 40.75 40.75 40.75 41 0 S. Paiute, Beaver, UT 1 38.25 38.25 38.25 38 0 Paiute, Unclassified 1 51 51 51 51 0 Total Paiute 3 43.33 51 38.25 41 0 Shoshone 5 41.7 48 38 42 2 Shoshone+Bannock 6 41.54 48 38 41.5 2 Ute (CO & UT) 3 40.67 45.5 33.5 43 0 Ute+S. Paiute 4 43.25 51 33.5 44.5 0 TOTAL NUMIC 11 41.86 51 33.5 42 2 Table 5-11. Numic bow cross-sectional shape profiles. See Figure 4-3 for shape code legend. Ethnic Group number L D Q O J B R N. Paiute, Bannock, ID 1 (0.5) (0.5) (0) (0) (0) (0) (0.5) (0.5) (1) (0) (0) (0) (0) (0) (0) (0) (0) (0) (0) (0) (0) % 100 50 50 0 0 0 050 50 100 0 0 0 0 0 0 0 0 0 0 0 0 S. Paiute, Beaver, UT 1 (0) (0) (0) (0) (0) (0) (0) (0) (0) (1) (1) (1) (0) (0) (0) (0) (0) (0) (0) (0) (0) % 100 0 0 0 0 0 0 0 0 0 100 100 100 0 0 0 0 0 0 0 0 0 Paiute, Unclassified 1 (0.5) (0) (0) (0) (1) (0.5) (0) (0) (0) (0.5) (0) (0.5) (0) (0) (0) (0) (0) (0) (0) (0) (0) % 100 50 0 0 0 100 50 0 0 0 50 0 50 0 0 0 0 0 0 0 0 0 Total Paiute 3 (1) (0.5) (0) (0) (1) (0.5) (0.5) (0.5) (1) (1.5) (1) (1.5) (0) (0) (0) (0) (0) (0) (0) (0) (0) % 100 33 17 0 0 33 17 17 17 33 50 33 50 0 0 0 0 0 0 0 0 0 Shoshone 5 (3) (3.5) (3) (0) (0) (0) (0.5) (0.5) (0) (0.5) (0) (1) (1) (1) (0) (0) (0) (0) (0) (0) (1) % 100 60 70 60 0 0 0 10 10 0 10 0 20 20 20 0 0 0 0 0 0 20 Shoshone+Bannock 6 (3.5) (4) (3) (0) (0) (0) (1) (1) (1) (0.5) (0) (1) (1) (1) (0) (0) (0) (0) (0) (0) (1) % 100 58 67 50 0 0 0 17 17 17 8 0 17 17 17 0 0 0 0 0 0 17 Ute (CO & UT) 3 (0) (0) (0) (0) (0) (0) (0) (0) (0) (0) (0) (0) (1) (0) (0) (2) (3) (1) (0) (0) (2) % 100 0 0 0 0 0 0 0 0 0 0 0 0 33 0 0 67 100 33 0 0 67 TOTAL NUMIC 11 (4) (4) (3) (0) (1) (0.5) (1) (1) (1) (2) (1) (2.5) (2) (1) (0) (2) (3) (1) (0) (0) (3) % 100 36 36 27 0 9 5 9 9 9 18 9 23 18 9 0 18 27 9 0 0 27 247

PAGE 248

A B C Figure 5-1. Three common surface treatments for the backing material of complex bows. Photos by author. A) AM NH 16/1341, Inland Salish or Nicola Athapaskan bow with snakeskin strip gl ued over sinew. Courtesy American Museum of Natural History. B) FMNH 14614, Hupa At hapaskan bow from California with heavy coating of paint mixed with fish-glue flux. C) FMNH 61036, Uncompahgre Ute bow from Colora do with heavy coating of bitumen and/or pitch on back and sides. Courtesy Field Museum of Natural History. 248

PAGE 249

A. B. C. Figure 5-2. Comparison of archaeological bow to regional ethnographic specimens. Photos by author. A) PAH 1/174972, Miwok-Costanoan (?) bow fragment from San Francisco Bay area showing possible glue residue. Courtesy of the Phoebe A. Hearst Museum of Anthropol ogy and the Regents of the University of California. B) YPM 15005; Miwok or Wintu sinew-backed bow tip. C.) Side-view of YPM 15005 tip, showing characteristic Miwok fiddlehead profile of sculpted sinew. Courtesy of the Yale Peabody Museum. 249

PAGE 250

Figure 5-3. Detail of Peter Ponds Map of 1785. The headwaters of the Peace River in the Canadian Cordillera correspond to Se kani territory, labeled Strong Bow Indians. Map reproduced in Ives (2003) and Gillespie (1975). Figure 5-4. RBCM 6563. Central C anadian Cree or Athapa skan sinew-backed compound bow, thirty-three inches long. Courtesy of the Royal British Columbia Museum. Figure 5-5. Spliced antler belly laths from co mposite segment bow. Neolithic, late third millennium BCE. Angara River, Yenisei Ba sin, Lake Baikal region, Siberia. Length not given (after Rausing 1967: Figure 60). 250

PAGE 251

Figure 5-6. NMAI 13/1513. Shoshone sinew-backed true co mposite bow, thirty-eight inches long. Photo by author. Courtesy of the National Museum of the American Indian Figure 5-7. PGM 2667-20. Denaina At hapaskan sinew-backed bow from Kenai Peninsula, Alaska, with hair fringe and a ttached wrist guard, fifty-nine inches long. Photo courtesy of the Peter t he Great Museum of Anthropology and Ethnography. A. B. Figure 5-8. Western versus eastern style attached wrist guards for Northern Athapaskan bows; note difference in placement of hole for sinew lashings. Photos by author. A) PAH 2/6361; Denaina Athapaskan sinew-backed bow from Kenai Peninsula, Alaska. Courte sy of the Phoebe A. H earst Museum of Anthropology and the Regents of the Univ ersity of California. B) FMNH 267485; Southern Tutchone Athapaskan se lf-bow from Whitehorse, Yukon Territory. Courtesy of the Fiel d Museum of Natural History. 251

PAGE 252

Figure 5-9. JES 141; Pomo sinew-back ed bow from west-central California, indistinguishable from northwest Cali fornia Hupa-Yurok bows. Photo by Author. Courtesy of the S an Diego Museum of Man. Figure 5-10. FMNH 17096. Apache double-curved sinew-backed bow; detail of barberpole sinew-spiral grip-wrap. Photo by author. Courtesy of the Field Museum of Natural History. 252

PAGE 253

Figure 5-11. Sixty-six inch double-curved Sekani Athapaskan sinew-backed bow from central British Columbia, with barber-spiral wrapping on bowlimbs and nearly rectangular in cross-section (Morice 1894:58). Figure 5-12. JES 858; Seventy-three inch double-curved Apache sinew-backed bow with nearly rectangular cross-section. Note normal-length Apache doublecurved bows in background for scale. P hoto by Author. Courtesy of the San Diego Museum of Man. 253

PAGE 254

A. B. Figure 5-13. Two Athapaskan-attributed trussed cable-backed short bows, ovate crosssection. A) Likely Apache or Wester n Navajo sinew-backed bow, Northern Arizona (Allely and Hamm 2002:220); drawing by Steve Allely reproduced by permission of Jim Hamm. B) NMNH E5588; Likely Deg Hitan Athapaskan boys bow from Yukon River, Alaska; by permission of Smithsonian Institution. 254

PAGE 255

Figure 5-14. JES 365; Unusual doublecurved Northern Athapaskan self-bow with attached wrist guard, upper Yukon River. Photo by Author. Courtesy of the San Diego Museum of Man. Figure 5-15. Outline of a Qum-Darya (Xiongnu) bow. Authors sketch (after Rausing 1967, Figure 5-f). A. B. Figure 5-16. Comparison of Apache and Slav ey longbows (detail of midlimbs). A) AMNH 50.1/7619A, Slavey longbow, rect angular cross-section, 66.1 inches long, 1.57 inches wide at grip. Phot o by author. Courtesy of American Museum of Natural History. B) MPM 14748, Apache longbow, rectangular cross-section, 67 inches long, 1.44 inc hes wide at grip. Photo by author. Courtesy of Milwaukee Public Museum. 255

PAGE 256

Figure 5-17. 2.5-inch long juniper sinew -back bow fragment excavated by Julian Steward from Promontory Cave No. 1. ; northern Utah (specimen now lost); (Steward 1937). A. B. Figure 5-18. Comparison of Promontory Cave bow fragment to decayed ethnographic Shoshone sinew-backed bow tip. A) UMNH 42Bo1 11602.2, belly-view of 800 year old bow fragment from Promontory Point, Utah. Courtesy of Utah Museum of Natural History. B) FMNH 60815, detail, belly -view of Eastern Shoshone sinew-backed bow, collected from Wind River, Wyoming, ca. 1900. Photo by Author. Courtesy of Fi eld Museum of Natural History. 256

PAGE 257

Figure 5-19. PAH 1/21418, back-view of 4. 3 inch Lovelock Cave sinew-backed bow fragment, with longitudinal sinew-strands still attached, visible on lower edge (Loud and Harrington 1929: pl. 47m). Photo by author. Courtesy of the Phoebe A. Hearst Museum of Anthropol ogy and the Regents of the University of California. A. B. C. Figure 5-20. Comparison of southern, Kod iak, and western style used to secure the cable backing to the bow (Murdoch 1885). Only the southern style (far left) uses the bows nocks to secure the cable backing. The other two utilize special parallel notches in the sides of the bow. A). YPM 15844, southern style Denaina bow, Iliamna, AK. Photo by Roger Colt en. Courtesy of the Yale Peabody Museum. B). JES 2681 Kodiak style bow, Kenai, AK. C). JES 2649, western style bow, Yuit, East Cape, Siber ia. Photo by author. Courtesy of the San Diego Museum of Man. 257

PAGE 258

A. B. Figure 5-21. Comparison of belly-faces of Yukon River and Mackenzie River Athapaskan complex bows. Note bla ckened longitudinal grooves. A). JES 2687, Southern Yukon babiche-backed bow. Photo by author. Courtesy of the San Diego Museum of Man. B). Mackenzie River sinew-backed bow, photo by Brian Seymour. Courtesy of the Royal British Columbia Museum. 258

PAGE 259

CHAPTER 6 CONCLUSION: PROTO-ATHAPASKAN AN D DENE-YENISEIAN MATERIAL CULTURE Cross-Disciplinary Synthesis In Chapter 5, I interpreted the data I collected for complex archery in North America as it relates to the problem of Athapaskan migrations and demographic expansion. In Chapter 6, I will address t he broader cross-disciplinary implications of these regional and continent al-scale ethnohistorical scenarios. The picture is complicated and subject to varying interp retations, but with a broad enough vision, it may be possible to utilize all the major subdisciplines of anthropology to construct a coordinated metanarrative around these numerous data. Genes, Languages, and Material Culture Figures 3-8 and 4-5 displayed the close c ontinental-scale patterning of the NaDene languages on the one hand, with molecular genetics (Haplogroup C3 Y-DNA) and material cultural data (complex bows) on the other hand. No rth American haplogroup C3 Y-DNA, is closely associated with Na-De ne speakers and is cladistically rooted in the Altai-Selkup-Ket population system of southwest Siberia (Zegura et al. 2004). The recent and rapid spread of Na-Dene C3 ma y have been the product of a male mediated migration from Asia (Wilson 2008). Another major male -mediated demographic expansion event took place in Eurasia during the middle Common Era, whereby the Genghis-Khanid Mongols conqu ered most of the Eurasian landmass during the thirteenth century CE. This event was fac ilitated by the Mongols mastery of complex bow technology originating in the Altai-Sayan uplands, and is closely associated with the rapid spread of low-to-moderate frequencies of C3 Y-chromosomes throughout the Eurasian landmass (Zerjal et al. 2003). In ot her words, the spread of Haplogroup C3 is 259

PAGE 260

linked to a particular material culture (com plex bow technology) of the middle Common Era in most of the Eurasian landmass. The Altaian region is the Mongols traditional homeland, and it is generally believed also to be the homeland of the Yeniseans, linguistic relatives of the Na-Dene. The rapid expansion (in low-to-moderat e frequencies) of Haplogroup C3 is associated with the spread of complex archery in the New Worl d, at precisely the same time (early second millennium CE ) as in the Old World. Fi gure 6-1 combines data from both Figures 3-8 and 4-5 to show that a strong case can be made that complex bows are geographically correlated with the C3 YDNA expansion. C3 has been christened both the Genghis-Khanid Y-Chromosome and the Na-Dene Y-Chromosome (Lell et al. 2002; Zerjal et al. 2003). Ne ither of these designations is a perfect fit, because ethnolinguistic identity is subject to change from one generation to the next (Moore 1994). I suggest tentatively that Y-DNA haplogroup C3 is mo re strongly correlated with the presence of complex archery than it is with any particular ethnic group The complex bow may have facilitated t he rapid southward expansion of the Athapaskan languages and t heir associated Y-Chromosomes in the New World. But oddly, neither Athapaskan speakers nor any glue-type complex bows ever made significant inroads into the eastern half of the North American conti nent, despite the fact that the technology and langua ges cross-cut major physiogra phic barriers from Alaska to Mexico. I suggest that these two factors ar e linked, and the primary constraint in both cases is ecology. In the arid western Unit ed States, complex bows can make especially effective tools for subsistence hunting and for raiding, but they are extremely vulnerable to moisture damage. They were similarly successful tools in the rainy Pacific Northwest 260

PAGE 261

only through the development of sophisticated waterproofing techniques (Figure 5-1). The eastward progression of the technology may have been halted by incremental change in precipitation and humid ity. The technology is exceedingly rare in those climates where annual rainfall exc eeds thirty inches per year. Historical Scenarios Accounting for Male Mediated Migration Scholars of Old World history are quite accustomed to imperial-scale ethnic expansions taking place within a few generations New World prehistorians, in contrast, are much more likely to model similarly sign ificant expansions over several centuries or even millennia. There are some notable recent exceptions to this general rule (Gullv and McGhee 2006; Sassaman 2010). I would lik e to encourage more collaboration by specialists working on both sides of Bering Strait. For my proposed synthesis of genetic, linguistic and material cultural ev idence to be successful, a number of challenges must be addressed. One major difficulty is the apparent sexual asymmetry of the Na-Dene founding populati on; very little maternal Mitochondrial DNA can be found to definitively link Athapskan peoples to late Holocene Eurasian populations. This fact must be better accounted fo r using historical models. At least three different scenarios could plausibly account for extreme sexual asymmetry among migrant parties in the Nort h Pacific Rim or Bering Sea. These are the (1) mercantile, (2) military and (3) maritime hypotheses. First, Central Asian male merchant-elite dominated the North Pacific fur trade during the Common Era (especially during th e last 1300 years) fr om headquarters at the trading hubs of the Amur De lta sea ports (Fitzhugh 199 4:36-41). Marriages between indigenous northeast Asian wo men and Central Asian men could have preserved common local Beringian mtDNA lineages, derived ultimately from America via reflux 261

PAGE 262

migrations (Tamm et al. 2007). Children of such mixed marriages, speaking paternallyderived Central Asian languages could nonetheless have possessed apparently Native American mtDNA lineages (Wils on 2008:272). This period of Central Asian dominance of trade in furs, metals and armaments corr esponds precisely to the introduction of the Asian warfare complex to Alaska, bet ween 800 and 1150 CE (Maschner and Jordan 2008), and the pre-Columbian apex of native metalworking in the Subarctic (Clark 1991). A mercantile hypothesis for Na-Dene orig ins has distinct resonance with the hypothesis of Gullv and McGhee (2006) regarding Neo-Eskimo origins; they postulate that aggressive Siberian Punuk/Thule metal-traders may have spurred the Neo-Eskimo expansion in the Common Era. If my interpretation of the material-cultural evidence is correct, the Athapaskan expans ion and the neo-Eskimo exp ansion are two different linguistic-cultural fronts of the same historical pr ocess (Ives 2010). The steady intensification of Bering Sea mercantilism throughout the Common Era seems a likely culprit. The most obvious weakness of this hypot hesis . is that many trappings of medieval Asian high civilization were never successfully transplanted to North America. To do so may have been difficult because the Asian immigrants did not constitute large co lonies of literate people with advanced metallurgical skills, but lived with a cult ural ecology similar to their Native American trade partners. It is plausible that individuals or small groups of highly mobile Asians, limited in their possessions, moved through existing trade networks where local cust oms prevailed (Wilson 2008:273). A second possible scenario is military expansionism in the North Pacific, e.g. by Xiongnu or affiliated regiments moving coastw ard from the northeaste rn borderlands of their empire. Imperial milit ary cadres were highly mobile groups of men, often multiethnic (including local and foreign soldie rs) but united by a lingua franca. This 262

PAGE 263

hypothesis could be linked to the hypothesis (1) above, if men from Yeniseian-speaking garrisons provided security for any of the mari time ports of the nort heast Asian fur trade, especially the Amur Delta Hub, which was a key center in the diffusion of Asian material culture to North America (Fitzhugh 1994). A period of Asian imperia l expansion set the stage for a series of rapid political devel opments in the Bering Straits region, as Maschner and Jordan have argued: the introduction of the Asian war comple x [to North America] includes the recurve or backed bow, armor, wr ist guards, defendable villages or the elaboration of fortificati ons, and the development of small, specialized arrow points. Most East Asian societies ar e experiencing a period of imperial expansion during this peri od (Maschner and Jordan 2008:104). Mason (1998) and Laufer (1914) provide addi tional discussion of the diffusion of Asiatic military culture and defensive a rmor through the Bering Sea and to the Northwest coast in the last two millennia. See Figure 6-3 for a comparison of Na-Dene and Central Asian armor designs. Mason (2009) has further developed a version of this hypothesis in terms of its implications for Neo-Eskimo origins, s uggesting that the NeoEskimo expansion was initiated by Siberian military cadres. Th e military hypothesis easily complements the mercantile hypothesis in this case, as war and trade were closely related phenomena in the region (Bur ch 1988). Merchants and warriors, males united by language, could easily have migrated t ogether. In the most extreme variant of this hypothesis, Ethel Stewart (1991) s uggested that Na-Dene founders were male refugee soldiers from the Altai region and Ta rim Basin, allies of the defeated Tanguts fleeing Mongol conquests during the early thirt eenth century, using riverboats to reach the Amur Delta, thence to Alaska. Her ve ry radical proposal has been dismissed on linguistic grounds (Campbell 1997:261), and adm ittedly relies upon highly questionable use of the linguistic data. Nonetheless, the historical plausibility of this scenario is 263

PAGE 264

defensible (Wilson 2005). Asi anist John Krueger notes there is no reason why such a migration was not feasible. . Stewart has thus presented a challenge to Mongolian studies, one which must be considered and discussed, and refuted, not just dismissed (Krueger 1993:287-288). The third scenario for male-mediated migrat ion is the maritime hypothesis, which seems less likely based upon the fact that the bulk of Na-Dene an d Yeniseian lexical cognates seem to indicate a shared boreal forest (interior) adaptation rather than a coastal one (Vajda 2010a). Still, it bears mentioning that northeast Asian fishing vessels were crewed by mostly men, and a ccidental voyages and shipwrecks of stormtossed vessels were common throughout history. A portion of the many iron artifacts found in late pre-Columbian archaeological contexts doubtless came from such wayward vessels (Keddie 1990), and East Asian survivors of shipwrecks were periodically assimilated by Northwest Coast peoples, as were the three Japanese sailors found living among Am erican Indians in early ni neteenth-century British Columbia (Kakuyabashi 1981:516). A few such individuals scattered over the course of centuries would have helped co ntribute to gene flow from Asia to America, and the observed derived phenotypic traits among Native Americans. A few charismatic individuals could even spread Buddhistic eschatology among the Tlingit or introduce Taoist concepts into the American shamani sm complex. These scattered maritime migrations would subtly increase Y-chromoso me diversity but be virtually undetectable in terms of mtDNA. 264

PAGE 265

Linguistic Horizons: Proto-Dene-Yeni seian Words for Technologies of Late Holocene Acquisition The ongoing reconstruction of proto-lexe mes in the Dene-Yeniseian languages will ultimately yield much relevant data on the material culture utilized by these speech communities in antiquity. At this prelimi nary stage of research, Eric Vajda has found a surprisingly robust inventory of words relevant to material culture, apparently consistent with the long-chronology view of Dene-Ye niseian origins, as he points out: Cognates in the realm of material cu lture are also limit ed to items and practices present on both sides of Bering Strait already many thousands of years ago: snow-sled runner, canoe, holding hook, verb roots denoting specific types of striking motions (hit endwise with a long object, slash) or object deformations (bend into a hook shape, bend less than 180, twist into a spiral). Predictably, t he cognates do not include words for technologies of Late Holocene acquisi tion, such as the bow and arrow, elaborate storage techniques, or the characteristic North American snowshoes (Vajda 2010b:102). If there is indeed a lack of Dene-Yeniseian cognate terms for bow and arrow, this would appear to uphold the conventional wisdom that the two populations began diverge prior to the Late Holocene, and t he common late prehistoric Asian-Athapaskan culture traits are convergent developments or fo rtuitous coincidence. Yet Vajda avoids calling attention to his own identification of the verb to shoot (an arrow), in protoAthapaskan *t' q' which is cognate with proto-Yeniseian *dq (Vadja 2010a:81). Admittedly this verb alone might have once referr ed to the shooting of an atlatl dart. But Vajda has presumably also chosen to re ject the proposed DeneYeniseian bow and arrow cognates proposed by Ruhlen (1998:13995) Vajda has rejected them along with over 75% of Ruhlens proposed Dene-Yeniseian word list. The great bulk of Ruhlens work, according to Vajda (2010a), consists merely of coincidental look-alikes, unsupported by the rigorous phonological correspondences which are necessary to 265

PAGE 266

prove genetic links. But the basis for reject ion of Ruhlens bow/arrow cognate is not clear, because it is one of Ruhlens better-s upported words according to even Vajdas high standards. The elements of Ruhlens word list which pass muster (according to Vajdas inspection) include the Dene-Yeniseian words for birchbark, stone, and foot. These acceptab le cognates rely upon the same system of sound correspondences (particularly the rules r egarding the glottal st op). Ruhlens proposed word for bow/arrow also follows this rule. Ruhlen discusses the birchbark etymon (which Vajdas work specifically corroborates): The Ket word for birch bark is q y, . almost identical to the word reconstructed for birch tree [or birch bark] in Proto-Athabaskan: q y. [T]he glottalization in the Proto-At habaskan form (symbolized as ) has shifted after the vowel in Ket, where it is symbolized as the glottal stop . . The difference in phonetic form al so has a simple explanation because the different location of the glottal stop in Yeniseian and Na-Dene is not an idiosyncratic feature of the particular word birch bark, but is rather a recurrent sound correspondence connecting these two families. It affects not just the word for bir ch bark, but also the wo rds for stone, utensil, bow, and foot (Ruhlen 1998:13994). Ruhlen (1998:13995) suggests pl ausibly that the Yeniseian word for bow (* q Ket q t, Yugh q t ), became the Athapaskan word for arrow ( *-qa, Koyukon qo Chipewyan ka Hupa -qa Mattole ka Navajo k ). Given that this sound correspondence follows the same phonological rule of thumb as these other words, it is not altogether clear why Vajda rejects it when he accepts birchbark, stone and foot. It may be that a tautology is in effect here; Dene-Yeniseian bow/arrow terms cannot possibly be valid cognates because of their necessarily late-Holocene historical derivation (in contravention to what we k now about Athapaskan prehi story). This logic is suspect and should be reconsidered. 266

PAGE 267

More telling is the specifically proto-Athapaskan word for (sinew-backed) bow, *ts -t attested to in all three geographic blocs, e.g. Deg Hitan gi -te Denaina c -den Gwichin ki -tai Hupa tsi -ti Chipewyan -tn and Navajo -tn (Golla 2007:72; Kari 1978:63; Osgood 1940:201; 1937:213), (Victor Golla, personal communication 7 November 2009). Sinew-ba cked bow technology was a post-Neolithic Eurasian invention. The presence of a spec ific term for this technology among the undifferentiated proto-Athapaskans is suggestive. This etymon is absent in Yeniseian; it might be a proto-Athapaskan ne ologism, in keeping with the well known Athapaskan preference to use the creative powers of their languages to make new terms (Ives and Rice 2006). Either way, the possibility of Late Holocene Eurasian origins cannot be discounted. Proto-Athapaskan Terms for Metal and Knife There is an intriguing precedent in the presence of a documented Eurasian loanword for an edge weapon and associated raw material in proto-Athapaskan and Eyak. The German term wanderwort, wandering word, is used by linguists to refer to a widespread loanword in multiple languages whose etymylogical origins may be obscure. In Eurasia, wanderworts are typical in ca ses of technological diffusion and trade items. The proto-Athapaskan and Eyak words for met al/knife are cognate with a wide range of post-Bronze Age Eurasian languages, with the most striking resemblances found in Central Asia, including notably some in the Yenisei river drainage, but not in Yeniseian proper (Golla 1998:2-4; see Table 6-1). T here are numerous other American Indian (but not Eskimo-Aleut) languages which have some variation of this Late Holocene trans-Eurasian wanderwort for metal, however Athapaskan is unique among all the American variants in that it appeared early enough for the form to have worked its way 267

PAGE 268

into Athabaskan at the protolanguage level in contrast with its presence in Algonkian, Siouan, Iroquoian, etc. some time after their linguistic diversification had already commenced (Victor Golla, personal communica tion 18 Aug 2010). Also the Athapaskan variant is most consistent in referring to the metal used in an edge-tool or weapon, whereas many of the other American Indian versions of the lexeme are likely to emphasize other attributes of the metal (especially its co lor). When the narrower range of Athapaskan meanings and particular stem morphology is considered, then the range of possible Eurasian diffusional so urces may be narrowed as well. Northern Athapaskan fired-copperwork has probable Asian antecedents, and is regarded as the finest example of Native American metallurgy by Witthoft and Eyman (1969:22-23). This high esteem is reflect ed in the name of the Yellowknife band of Chipewyan. The recorded names for Ye llowknife in Athapaskan, Algonquian, and European languages all appear to refer ultimate ly to the copper found in their territory from which they made cutting tools for tr ade to other Indians (Gillespie and Goddard 1981:288). Interestingly, the specif ically Athapaskan terms (Chipewyan: tal ho in Yellowknife Indians; Dogrib: teh i Chipewyan Indians) reflect an obscure euphemism meaning pond-scum people, because of the similarity in color between blue-green algae and bluegreen cuprous oxide (289). T he Woods Cree (Algonkian) name for them, mi kohkoma n red metal, red knife, by contrast is remotely linked to the same Eurasian metal-lexeme from which the proto-Athapaskan fo rms (like Tsuutina mes and Chipewyan bes both meaning knife) are derived (Gillespie and Goddard 1981: 288; Golla 1998:2; Table 6-1). 268

PAGE 269

One striking similarity in Table 6-1 is the near perfect match between the protoform Athapaskan w metal, knife and North Caucasan (Ubykh) w a copper. The Dene-Caucasian hypothesis is a dire ct forerunner to the Dene-Yeniseian hypothesis. Blaek and Bengtson have presaged Vajdas conclusions (quoted on page 265 above) when they write of Dene-C aucasian (DC) material culuture: DC technological terms are all consist ent with the late Upper Paleolithic period: awl, thread, rope, cord sn are, strap, basket, wooden spoon, wooden bowl, trough, knife (originally of st one), arrow, fence. The house was possibly of a hurdle type (interlaced t wigs), supported by a sturdy frame. Starostin ([1991:]33) posits one word for metal, possibly copper, which would have been free copper and not an indication of metallurgy. So DC culture was a typical Upper Paleolithi c, hunter-gatherer culture (Blaek and Bengtson 1995:42). And yet, we must confront the stubborn fa ct that evidence of diffusion after a major technological breakthrough, probably annealing, is provided by the widespread late rise to popularity of copper-worki ng among Athapaskans and Eskimos during the second millennium A.D. (Clark 1991:116). Clark also not es that the conceptually related process of heat-treat ment of chert and quartzite is not demonstrated to be the impetus for this copper revolution, because heat treatment of lithic raw materials is not attested in the Subarctic. Thus the widespr ead linguistic evidence for Dene metal use is highly ambiguous and not at all conclusive of the Stone Age or igins of the metal/knife and bow/arrow lexemes in particular. T here are several independent diffusionalclusters of metal words, involving Indo-European, Turkic, and Mongolian languages; the North Caucasian form shown here is intermediate between different Indo-European forms (Golla 1998:6), and strongly suggesting late Holocene origins for the term, corresponding to the period of intensiv e interaction between Eastern and Western Eurasian languages in the crucible of meta l-age cosmopolitanism. If a historical 269

PAGE 270

connection between the Caucasia n and Athapaskan lexemes is indeed likely, it is not necessarily of the conventiona l phylogenetic type, but is more easily explained as owing to the historical extent of a frontier-interface between distinct Asian cultures (Wilson 2008:269). As Wolfgang Behr notes conc erning Dene-Sino-Caucasian and related multilateralist proposals: Whatever the merits of each of t hese proposals might be as far as the question of genealogical descent . it is likely that most of them may to a certain degree reflect ancient loan cont acts . during the late Neolithic and early bronze age (Behr 2004:176). A similar argument pertains for bow and arrow terms in North America. The profusion of certain archery words acro ss language family-boundaries does nothing to support the existence of large genetic units, whether they are hypothetical Pleistocene linguistic megafamilies like Amer ind or Dene-Caucasian, or ev en just the larger of the more generally accepted American phyla like Hokan. This is because the bow and arrow did not exist in North America prior to 4500 years ago at the absolute earliest, considerably postdating the time when each of these large and ancient units would have begun their long process of dive rsification. The widespread similarities of bow terms are rather more likely a result of wi despread borrowing betw een various distinct populations involved in the trans fer of technology in the Late Holocene, not evidence of descent from a remote common ancestor for a large stock with complex branching patterns and great time depth. Athapaskan is different however, in that it is a smaller family in terms of linguistic diversity (although it is demographically and terri torially immense) wit h little branching [and] low to moderate time depth (Kaufman and Golla 2000:52). It is entirely plausible to find protoforms for specific types of archery technology (sinew backed bows) 270

PAGE 271

transmitted throughout the daughter languages in a straightforward phylogenetic fashion. The quality of the linguistic ev idence for the spread of archery terms in Athapaskan daughter languages is altogether more lucid than in other American Indian languages. Even though Athapaskan is a very young stock, it is still hypothetically plausible that the recent, rapid continent-wide spread of achery technology and its associated lexicon has also diffused cognat e words throughout the family, which now fortuitously resemble protoforms related by common descent. This is very unlikely to be the case though, because the phonological regularity of the lexemes is too consistent and idiomatically integrated into the dial ects in question. Loanwords have many diagnostic features; these features are only rarely found in Athapaskan archery terms (the Deg Hitan example in Chapter 5 is one notable exception). In general, Athapaskans encountering unfamilia r items of foreign origin have strongly favored the creation of neologisms to describe them (Ives and Rice 2006). The overwhelming consistency of the Athapskan archery words (and the regularity of their localized phoenetic variation) is strongly indicative that they are not recent loanwords. They have been recognized as primary root s (that is lineal descendants of proto-Athapaskan) since at least Adrian Morices time (Morice 1894:55 ). The likelihood that multiple related languages would come up with the same neologism is improbable in the extreme. Archery terms in other Native American languages serve to illustrate contrast. Katherine Turner (1983:221) has used archery terms in cognat e sets to argue for the common Hokan genetic legacy of California lingu istic isolates, Chumash, Esselen, and Salinan, all which have cognate terms for bow similar to ax or xak To use this to argue for common descent is problematic for tw o reasons. First, Chumash is no longer 271

PAGE 272

considered a Hokan language by most linguists (Golla 2007:80). Second, if it were to be included in the stock along with accept ed members, it would be grouped among the half-dozen or so California Hokan isolat es, as the eroded remnants of formerly widespread language groups (Golla 2007:78). In other word s, even assuming that all these languages are in fact related, their kinship at or is near the maximum time depth possible within the Hokan sto ck, well before the introduction of archery to California,. Hokan is the oldest linguistic relationshi p among western North American languages that can be established by no rmal comparative linguistic me thods. The time depth of the relationship is on the order of 8,000 year s (Golla 2007:78). California Hokan cognate bow words must have diffused much la ter, probably along with archery itself. Furthermore, Chumash has separate and distinct words for self-bows versus complexbows (Turner 1983:221) raising the possibility that the two distinct forms of archery technology were adopted at different ti mes and from different sources. Joseph Greenberg (1987:146) also compiled a list of bow and arrow terms as a part of his massive and controve rsial Amerind superphylum, spec ifically in an attempt to show an archaic relationship between west coast languages (mostly Penutian) and east coast languages (Muskogean). His Natchez le xeme is errant as demonstrated by Kimball (1992:453), and his single proto-Mayan form, *la(:)h nettle, has no relation to archery and is therefore irrelevant. Tabl e 6-2 has the remainder of Greenbergs more plausible Amerind archery terms with so me additions and corrections from other sources. The possibility of cognate relati onship between the Muskogean, Wappo and Penutian forms is weakened by Kimballs su btle revision of the Wappo and Muskogean 272

PAGE 273

forms, particularly the proto-Muskogean form aki which lacks the nasal being compared across the languages (Kimball 1992:453). Furthermore, Wappo is neighbored by multiple California Penutian languages and consequentially has a very large number of documented Penutian loanw ords (Dixon and Kroeber 1919:115). Even if we accept that all the non-Penutian archer y terms are related, they cannot be related within the immense timefr ame required for Amerinds alleged post-Pleistocene diversification. Even the divergence between the various substocks of Penutian represented in Greenbergs list is on the or der of 6500 years, comparable to IndoEuropean (Golla 2007:75). The bow and arrow did not appear until several millennia later, as John H. Blitz notes. [A] continent-wide perspective reveal s a north to south chronological distribution for the initial adoption of the bow. Multiple episodes of independent invention or extensive mo vements of people are rejected as explanations in favor of a secondar y diffusion process (Blitz 1988:137). Looking at Figure 6-4 (fro m Blitz 1988:132), it appears that the bow was widely adopted in a very short time, appearing in most regions within a few centuries after 500 CE. Blitzs map suggests the apparent route by which it spread throughout the continent was through the Plateau and wester n Great Basin around 200 CE, thence to the Northwest Coast and California, Plains, Southwest and Northeast, all between 500 and 600 CE. Among the last regions to s ee archery technology was the Muskogean Southeast, between 700 and 800 CE (and also, as we shall see, the Subarctic). So it is at least fathomable that Plateau Penutia n speakers, who were among the earliest archers in the middle-lattitudes, were also involved in the transmission of archery technology to their linguistic relatives in California, and also later (possibly through 273

PAGE 274

intermediaries) to the Southeast. Plateau Penutian archery voc abulary may therefore have diffused in association with the weapon itself. Only one physiographic region or culture area on Blitzs m ap is entirely blank: the Subarctic. This is because archaeological information regarding the date of the adoption of archery was very poor there at t he time of his writi ng (contemporary data are discussed below). Even as late as the 1990s, there was little certainty when archery first appeared, particularly in the We stern Subarctic. Looking at the map of Northern Athapaskan territory (Figure 3-8A), the alleged proto-At hapaskan homeland is squarely in this vast zone of uncertainty (Figure 4-5). The southern margin of the Subarctic (bordering the Plat eau and Northern Plains) yields ample evidence for archery in the early Common Era. Yet Athapaskan in cursions into this frontier (Beaver and Chipewyan south and east of Lake Athabasca Chilcotin south into the Plateau borderlands, and Sarsi into the Northern Pl ains) all occurred sometime between the thirteenth century and historical times, long after the continent al ubiquity of archery was well established (Smith 1987; Wilmeth 1977). The question of the physical evidence of archery in the proto-Athapa skan homeland is discussed at length by Donald W. Clark: The first firm evidence that norther n Indians used the bow and arrow is relatively late. . Subarctic peoples began to use bows and arrows only a few centuries ago. That is later than when archery was adopted elsewhere on the continent. . Nevertheless, doubts remain as to whether the features of stone points ar e reliable guides for identifying them as either arrow or spear points. In late preh istoric times, arrowheads were usually made from bone and antler, and some lo cally were pounded out of native copper. . As bone and antler have a poor preservation record in the Subarctic, it may be premature to assu me that the absence in the earlier archaeological sites of points made of those materials indicates that the bow and arrow was not yet in use. . [I] t is difficult to determine how widely the bow and arrow was used in the Western Subarctic or when it was first adopted (Clark 1991:102-103). 274

PAGE 275

Archeological Horizons Work done in the late 1990s and early 2000s suddenly and radically changed this situation, as a large number of more-or-less complete proj ectiles with bone and antler points intact, sinew bindings, and wooden s hafts have been collected from permanent ice patches in Wrangell-St. Elias National Park, Southeast Alaska, yielding a plethora of ironclad 14C dates (Dixon et al. 2005; Hare et al 2004). The first unequivocal evidence of bow and arrow is three fragments of a maple bow dating around 700 CE and the last evidence of atlatls disappeared within about a century of that time (Hare et al. 2004:270). The ubiquity of bow-and-arrow technology was not established until 850 CE in most of Krauss and Gollas (1981) proposed area for the proto-Athapaskan homeland, the borderlands bet ween southeast interior Alaska, southwest Yukon Territory, and northwest British Columbia (I ves 2010:328). The tran sition from darts to arrows in the region is also clearly indicate d by a transition from stone spearpoints to barbed antler arrowheads and a transition from birch to spruce as the predominant shaft-wood for the projectiles (Hare et al. 2004: 270). The transition later culminated in the southern Yukon (like the western Subar ctic as a whole) with common Athapaskan use of annealed copper arrowheads and obsidi an-tipped antler arrowheads in the second millennium CE, particularly after circa 1200 CE, signaling a greater reliance on the bow and arrow (Clark 1991:67). This chronology is problematic from the standpoint of At hapaskan linguistics because a cognate bow terminology exis ts throughout Athapaskan, whose speech communities [presumably] must have begun diverging [from the region in question] long before the adoption of bows and arrows (Ive s 2010:328). This same problem is exacerbated by existence of Asian cognates for the metal/knife and bow/arrow terms 275

PAGE 276

already discussed. As Ives himself told me [i]f Dene ancestors were not living in that vast region for at least the last two millennia, I'd say we haven't the least idea about anything in Dene prehistory (John Ives, personal communication 8 Aug 2010). Consequently, Ives maintains, the protoAthapaskans must have possessed an ancient cultural frontier with Eskimoid peoples wh o did possess archery, and thus could have been cognizant of its use l ong before they (the Athapaskans ) themselves adopted it (Ives 2010:328-329). Ives offers an enterprising solution, but it is insufficient for two ma in reasons. First, while the early paleo-Eskimos (pre-Dorset) were indeed proficient archers 4000-5000 years ago, as suggested by Blitz (1988), their use of the weapon never caught on elsewhere, and completely vanished for mo re than 2000 years during later (Dorset) Paleo-Eskimo timeframes. Pre-Dorset archery artifacts have no counterparts in any Dorset stratum (Meldgaard 1962:pl. 5). The second problem is the hypothetical location of the proto-Athapaskan homeland in the region in question, which is really only (sort of) near the area of great divergence (Kari 2010:2 10). These borderlands were selected as the urheimat by default, precisely to avoid the problem of long term interaction with Eskimoid peoples, which seemingly would have left ample linguistic substratum influences absent in proto-At hapaskan (Krauss and Golla 1981:68). On internal linguistic evidence alone, the proto-Athapaskan homeland should be sought in central or western interior Alaska, nearer to Eskimos and also much nearer to the main beachhead for transoceanic migrations (Kari 1989:538-539). A more westerly Alaskan homeland would thus be closer to ground zero for the diffusion of Asiatic archery technology. 276

PAGE 277

In the last 2000 years, the Asiatic bow and arrow has radically transformed the entire Western hemisphere, both in terms of hunting economy and wa rfare. It seems counterintuitive that archery could have been introduced from Siberia, flourished for an extended period circa 5000 years ago, but have been completely extinguished in the north slightly less than 3000 years ago. Yet th is should come as no surprise to anyone familiar with the general arc of Arct ic prehistory. As Krupnik notes: the evolution of weapons and technology in the Eskimo hunting economy moved in a spiral, not in a linear, progr essive fashion. . [T]he entire traditional complex of Eskimo hunting equipment was essentially formed . some two thousand years agoseveral clear periods can be defined in those two millennia when it was simplifie d or even reverted to original, more primitive forms [in a] . develop mental model of vivid explosions alternating with regression to . long outmoded forms (Krupnik 1993:197198). The sudden reappearance of Asiatic archery te chnology in Alaska during the middle Common Era accords roughly with the s hortest linguistic chronology (circa 1300 years depth) proposed by Hoijer (1956), and also with the dram atic explosion in native copper-use. Proto-Athapaskan metal-terms and archery-terms both make sense in this context if one allows that the Athapaskan languages were linguistically undifferentiated and geographically limited during this time per iod. These are defensible positions on both archaeological and historic al-linguistic grounds. However, we have yet to address the issue of complex, sinew-backed bows, whose presence is not established by the recent archaeological data. Recall that proto-Athapaskan *ts -t specifically means (sinew-backed) bow. To date, none of the bows unearthed (or de-iced) in and around Wrangell-St. Elias are anything but selfbows (John Ives, personsal communication 8 Aug 2010), and furthermore, very few east-cent ral interior Alaskan Athapaskans were known to use sinew-backed bows in the ethnogr aphic record, in contrast to virtually all 277

PAGE 278

of the westernmost Alaskan Athapaskans an d Eyak who did in fact use sinew-backed bows (Driver and Massey 1957:352). It is difficult to precisely date the arrival of the sinew-backed bow in Alaska. There are some probable late Punuk or Thule ar chaeological specimens of compound bow parts on the Siberian side of Bering Strait, at Cape Chaplino on the east coast of the Chukchi Peninsula, but these have not been dated (Mason 2009:84). The ~3900 yearold (pre-Dorset) bow bracers found at Igloolik (Eastern Arctic) were indeed used in the manufacture of complex bows (Meldgaard 1960:74-75). However, this fact has no bearing on this question, because Dorset never adopted these technologies (Agger and Maschner 2009:322) and the pre-Dorset were long-extinct when the protoAthapaskan speech community exis ted. The lack of firm evidence for the reintroduction of the weapons to North America means that some guesswork is involved, but it seems reasonable to say that the str ong complex bow arrived in most regions of western North America within a few centurie s of the initial (re-)appearance of bows in general in North America. Maschner and Jordan make the case that, as a key element in the Asian warfare complex, the sinew-backed bow was a catalyst for the escalation of hemispheric violence in the second millennium CE. [T]he recurve bow and its accout rements changed warfare. It created powerful groups who could easily annihila te peoples who were not using it. As this technology spread through North America in the centuries just prior to AD 1150, it became the de facto m eans of military aggression in every region. We believe that by AD 1150, a threshold had been reached where every single group in the Americas was under the threat of a potential violent conflict and something, somewhere, snapped, causing a cascade of change across the hemisphere (Maschner and Jordan 2008:105). The bows found preserved in caves in the Southwest and northern Great Basin are exclusively self bows until at least the thirteenth century. Likewise the artistic 278

PAGE 279

renderings of archery in Puebl oan ceramics are exclusively self bows until at least the late twelfth century CE. In the Southwest the adoption of the sinew-backed, recurved bow dates to somewhere between A.D. 1200 and A.D. 1450 (LeBlanc 1999:103). The archaeological evidence for the arrival of th e sinew-backed bow in California is even more ambiguous because of poorer preservati on of organic remains. The linguistic evidence is unequivocal about the presence of archery technology in the proto-Pacific Coast Athapaskan (PCA) community, who were one of the last ethnic groups (if not the last) to arrive in prehistoric California. Refering to proto-PCA archery lexemes, Victor Golla writes: This seems to me incontrovertible ev idence that the linguistic ancestors of the Pacific Coast Athabaskans enter ed the area with the bow and arrow, and probably with knowled ge of the sinew-backing technique. Whether or not Athabaskans were responsible for the spread of bow-and-arrow technology into California, the firs t archaeological appearance here of projectile points that can be classi fied as arrowheads (apparently ca. 600 AD) undoubtedly marks the earliest possible date for an Athabaskan presence in the region (Victor Golla, personal commnication 7 Nov 2009). The arrival of the bow and arrow in California in 500-600 CE is no doubt a terminus post quem for the arrival of the At hapaskans. However, it would be unlikely that they came at this early date if they had original kn owledge of the sinew-backing technique, as is extremely likely based on the linguistic data. Whis tler (1979) favors a thirteenth century CE dat e for the entrance of Athapaskan speakers, based on a combination of linguistics, archaeology, and et hnographic reconstruction. This date is also in good accord with the earliest estimate s for the arrival of the sinew-backed bow in another region (the Southwest) as noted by LeBlanc (1999:103). Other linguistic candidates for early archery in California are the Penutian speakers, who I hypothesize may have adopted the bow-and-arrow circ a 500-600 CE through long-distance 279

PAGE 280

interactions with their PlateauPenutian linguistic brethren. Given the evidence that the closest linguistic relatives of PCA communi ties were Plateau Athapaskans (e.g. Nicola and Chilcotin) who also possessed similar si new-backed bow technology, it seems that there may have been an enduring north-to-south acculturat ion gradient involved in migration and the diffusion of weapons technol ogies from Asia, through the Plateau, to California. By this route, self bows c ould have bypassed the more remote subarctic tundra at an early date. The ultimate As ian origin of the bow and arrow might be traceable in the stunning similarity between Siberian Uralic and Central California Penutian archery and weapons terminology noted by Otto von Sadovszky (1996:26-27). Von Sadovszky (1996:270) sugge sted that the Penutian and Uralic languages were phylogenetically linked, and that the Penutian languages must have diverged from a Siberian source language after the protoUgrian period (later than 500 BCE). His work has not been well received by Americanists who rightfully note that the divergence between the Penutian languages t hemselves is much greater than this timeframe would allow. Nonetheless, his voluminous data are compelling even if he did not produce a plausible historical scenario to account for them. Late, rapid coas tal migrations from Siberia to the Pacific Northw est would certainly help to account for the appearance of archery in California and the Plateau several centuries earlier than interior Alaska. I suspect that non-genetic factors (that is, lingu istic borrowing) may be at play here more than von Sadovszky would admit. Enduring c ontacts between distantly related groups in the North Pacific Rim could result in an uneven patchwork of lexical innovation and loanwords following the spread of new technology. 280

PAGE 281

It is quite reasonable that bows and arro ws could spread from person to person through existing populations as is conventiona lly assumed, but the incredible speed of the diffusion (continent-wide in only a few cent uries) implies that lo ng-distance travel of individual culture-bearers must have been invo lved at some stage, because the craft of the bowyer is such that instantaneous transfe r of the required ski ll-set between ethnic groups seems unlikely. This is especially true for complex archery. The manufacture of strong complex bows is a discipline exclusiv ely associated with intensively trained fulltime craft specialists in the Old World. With some except ions, this is not generally the case in North America. Complex bowma king nonetheless required considerable time and expertise, and was not something undertaken casually (Bergmann and McEwen 1997:159). This seems to preclude the likelihood that advanced weapons technology would spread like wildfire in t he absence of a major migration. The physical presence of the Athapaskans remains the most plausible vector accounting for the lightning-fast dispersal of complex bows from Alaska as far as California and New Mexico. Scholars need not shy away from the possi ble historical implications of von Sadovszkys much-maligned Uralic-Penutian hy pothesis, nor should we be afraid to subject Vajdas widely acclaimed Dene-Yeni seian data to the same sort of broad historical critique. AsianAmerican prehistoric relationships may be subtle and manylayered, ultimately much more complic ated than a single phyletic branchpoint on the map of Siberia. It should be asked whether language families like Na-Dene and Penutian may in fact be polygenetic hybri ds with significant inheritance both from postPleistocence North-America and fr om post-Bronze Age Eurasia (cf. Dimmendaal 1995). 281

PAGE 282

Proto-Athapaskan Pottery Words The final piece in the linguistic puzzle to be addressed here is just as fraught with historical ambiguity as the prot o-Athapaskan words for metal/knife ( *we ), sinewbacked bow ( *ts -t ), and arrow ( *-qa ). The evidence of a proto-Athapaskan word for pottery ( *onsa, *asa or *usa:k ) is equally problematic, because the vast bulk of Northern Athapaskan territory never possessed ceramics until European contact (Driver and Massey 1957:230). This vast subarctic ceramic-free zone notably includes the entirety of Krauss and Gollas (1981) putative proto-Athapaskan homeland in eastern Interior Alaska and northwest Canada. Howeve r, pottery is well attested in western Interior Alaska, the region of the greatest Athapaskan linguistic and cultural diversity. James Kari (1989:566 n.3) notes the exist ence of aboriginal pottery traditions among the Lime Village Denaina, Yukon Deg Hitan, Upper Kuskokwim Athapaskans, and Tsuutina. At least nine different Nort hern and Southern Athapas kan languages have a variant of the same term for cla y pot or kettle (Table 6-4). The Tanana and Koyukon in west-central Alaska were also very proficient potters (Rainey 1939:376-377). A portion of Kaska territory in north-central British Columbia (a quadrant bordering the Tahltan) are also indicated by Driver and Massey (1957:340, Map 127) as a native source of ceramic potte ry according to ethnographic records, but I have not found primary sources confirming this. It is tempting to view Tsuutina pottery as a northern extension of Plains ceramic traditions. However, Edward Sapirs ethnographic work among the Tsuutina was strongly suggestive of northern/Alaskan origins for their pottery. T heir Blackfeet allies (from whom they learned most Plains culture traits) are less certainly associated with the ceramics craft than are the Tsuutina 282

PAGE 283

themselves, and Tsuutina pottery also shares with Alaskan ceramics a highly unusual low-fire animal-hair organic temper. Sapir writes: It is natural to look upon the pottery of the Blackfoot and Sarcee [Tsuutina] country as but a marginal outpost of t he more intensive pottery culture of the Mississippi Valley and the western Great Lakes. Is it not at least possible, however, that the old Sarcee pottery, of which the Indians retain such a clear tradition, is the survival of a northern type that is historically connected with the Eskimo ware or that it represents a compromise between northern and eastern streams of influence? . [It seems possible] that pottery was more extensivel y used among the Sarcee than among the Blackfoot or, at any rate, the Piegan, the southernmos t of the three Blackfoot tribes (Sapir 1923:252). The archaeological traces of cerami cs scattered near the lakes to the east and northeast of the Tsuutina are almost certai nly extensions of more easterly and southeasterly cultures, probably Algonkian (less than 1000 years old) and not ancestral Chipewyan (current residents of the regi on), who themselves possessed no ceramics tradition (Wintemberg 1942:131; Ives 2003:276). Nonetheless, the existence of protoAthapaskan pottery terminology is another majo r problem for the standard models of the Athapaskan expansion which locate the epic enter of this exp ansion deep within the aceramic northwestern continental interi or 2000 or more years ago. John Ives approaches the dilemma the same way that he does for the archery problem, by positing that the proto-Athapaskans must ha ve long known about pottery via contacts with Eskimo neighbors, without actually adopting it themselves (cf. Sassaman 1993). He notes that ceramics industries, t hough apparent on the Alaska Coast for 4,000 years, were not well established in the American Arctic until the Norton period after 500 BCE. A Neoeskimo tradition of pottery t hen extends into the recent past (influencing neighboring Dene like the D eg Hitan and Koyukon). None of these ancient ceramic traditions necessa rily occur within regions thought to contain the Dene homeland, but they do occur adjacent to Athapaskan 283

PAGE 284

homeland regions. Otherwise, ceramics are simply absent over vast regions of interior northwestern No rth America throughout the last 4000 years, in a time range when Dene ancestors must have been widespread in the western Subarctic. Despite this absence, a clay pottery term is found throughout Northern Athapaskan and Apachean (though not Pacific Athapaskan), as Sapir (1923) pointed out. Once again, it would appear that Athapaskan ancestors were aware of t he technology, but they did not adopt it for their own use until Apachean ancestors took up ceramics (Ives 2010:329). I am not convinced that this reasoning is sufficient to account for the data. Take the example of the Pacific Coast Athapaskans who Ives mentioned. They are also descended from the proto-Athapaskans who should have been aware of pottery at the very least. And yet they re tain no equivalent word for th e technology, having evidently lost it themselves at some point after a just few centuries in California and Oregon, a region completely lacking in cerami cs (Figure 6-3). As Sapir writes: The term is apparently absent in Pa cific Athabaskan. Presumably the Athabaskan term originally referred to a pail-like or kettle-like receptacle of bark, only secondarily to one of clay. And yet can we be sure that its primary meaning was not "clay pot"? If it was, we could understand why it was lost in the Pacific dialects, for a term for clay cooking vessel would not be readily used for one of twined basketry, while a term for bark vessel conceivably might be (Sapir 1923:253). Of the nine Athapaskan groups listed abov e as possessing the proto-Athapaskan pottery word, four of them (Deg Hitan, Denaina, Tsuutina, Navajo) were known manufacturers and users of ceramics in precontact timeframes. Two more (Beaver, Sekani) were close kin of the Tsuutina, with oral traditions of their common origin. If the Tsuutina were originally potters (as Sapir suggests) it is almost certain that Beaver and Sekani were also potters at one time, becaus e all three groups were undifferentiated in the recent past. The Carrier in turn sust ained close contact with this whole bloc long before it differentiated. The other two groups (Hare and Mountain) formed one bloc in direct contact with Mackenzie Eskimo potters of arctic coastal Ca nada. Alaska-derived 284

PAGE 285

Thule pottery was more abundant than soapst one in the late prehistoric Mackenzie Delta region (Morrison 1991:240-242). The extinct Hare and Mountain bands are so poorly studied that it may be impossible to determine the exte nt of their knowledge of any particular technologies. In summary, pre-contact direct k nowledge of ceramics can plausi bly be attributed to all of the abovementioned groups, whereas I am unawar e of any of the Athapaskans of the putative homeland region (southeast interior Alaska, adjacent tr acts of British Columbia and Yukon Territory) who maintain any variant of the proto-Athapaskan word for clay pot. Could it thus be possible, as Ives suggests, that this lexical knowledge was maintained for thousands of years in this vast aceramic heartland region, when the equally aceramic PCA bloc lost it through eros ion in less than 1000 years? It is more reasonable to suggest (with Kari 1989) that the proto-Athapaskan homeland is to be found in western Alaska, where ceramics industries were indeed present. Also noteworthy in this regard is t he Dene-Yeniseian cognate term for dish, proposed by Ruhlen, which also could include items of ceramic manufacture (1998:13996). The Yeniseian word *s k, attested in both Ket and Yugh, refers to a trough-like bowl for bread dough. It is comparabl e to many Na-Dene words for dish or plate, for example Tlingit six Ahtna ts k Denaina tsuk and Beaver ts This proposed cognate also obeys the phoenetic principles governing the shift of the glottal stop, like birchbark and bow/arrow. It thus appears to withstand Vajdas scrutiny, although Vajda does not comment upon it in his recent (2010a) essay. Hair-tempered Pottery in the New World The final point to be made about Athapaskan pottery concerns the extreme rarity of animal-hair temper in North America, in the context of late neo-Eskimo low-fired 285

PAGE 286

ceramics. What was once dismissed as neo-Eskimo crudeness (merely the attenuation of more refined ceramics of pal eo-Eskimo times) is more a ccurately characterized as Thule versatility, as Thule potters adjusted temper and the firing r egime to adapt their ceramics to a variety of resource areas in an environment badly suited to pottery manufacture (Stimmell and Stromberg 1986: 237; see also Harry et al. 2009). This adaptation permitted the expans ion of ceramics industries into more mobile groups, including the interior Athapaskans. Organic-tempered ceramics have super ior performance characteristics during manufacture, allowing for an expedient ceramic technology. This, along with reduced weight and greater portability, may explain the preference for organic-tempered vessels by groups that frequently shift their residence (Skibo et al. 1989:122). Notably, low-fired organic-te mper pots are also particularly susceptible to total decomposition in a moist environment when s ubjected to freeze-thaw cycles, meaning archaeologists doubtless have underestimated the ex tent of their use (Reid 1984). It may be that hair-tempers have greater anti quity and range than we are presently aware of, however it is still safe to say for the most part they are a very late neo-Eskimo technology in America. It is uncertain whether northeast Asian hair-tempered pots are beholden to Japanese fiber-tempered ceramics, but hair temper emerged in the Lena valley, Siberia at the terminal Neolithic, and spanned the Bronze Age. During this time, the technology spread eastward into the C hukchi Peninsula, first appearing on the Siberian side of Bering Strait during late paleo-Eskimo timeframes (circa 2000 years ago) and substantially later in Alaska. On the American side, in marked contras t, hair temper is relatively recent, the oldest examples dating from A.D. 1250 to 1400 on the Kobuk River (Giddings 1952:94-5). It is also ment ioned for the Tena [Deg Hitan] Indians of the middle Yukon (De Laguna 1947:141) and for the modern Alaskan Eskimolocality unspecified (Collins 1937:167), but is the least common 286

PAGE 287

organic material encountered in Ala skan pottery. Elsewhere in North America it is reported ethnographically from the Sarcee [Tsuutina] and Kutenai (Oswalt 1955:41), but otherwis e seems not to be known (Chard 1958:193). The Kutenai resided due south and nearly adjacent to the Tsuutina, and could have easily learned Tsuutina horsehair-temper ceramic technique as a result of marriage alliances formed in conjunction with long-term cooperative hunting arrangements. The hunt of t he buffalo led to certain al liances and unions for the season of the chase . a few of which may have become permanent. Thus the Kutenai [and] Sarsi . hunted together on the plains of the Saskatchewan and the upper Missouri (Chamberlain 1907:478). Hair temper is otherwise most closely identified with the Tanana, Kuskokwim and Koyukon Athapaskans of central Alaska, who used bear hair-temper in their pottery (Rainey 1939:376-377). Koyukon of Alaska also shared with Tsuutina of the northwest Plains some specific molding techniques (direct molding and bark molding) wh ich are not widespr ead elsewhere. Wendell Oswalt (1955:41) comments on these and other general similarities between northwest Plains and Alaskan pottery (e.g. cylindrical forms, oiling and watertesting of finished pots), that some of these features may be found among any potteryusing people, but the im portant fact is that these parti cular technics all are common to 2 geographically disconnected groups who are t he most proximal po ttery manufacturers in northwestern North America. He further addresses the discipline-wide implications of these facts, noting that If such a technological relationship did exist, it would be highly significant on distribut ional grounds and demonstrate t hat a 2000-mile spatial gap without pottery is insignificant to its spread (Oswalt 1955:41). 287

PAGE 288

The proto-Athapaskan linguistic evidence ( *ons a *as a *usa:k clay pot) would seem to bear Oswalt out, particularly when you look at the lexicostatistical evidence suggesting that western Alaska may hav e been the proto-Athapas kan homeland (Kari 1989; Sapir 1923). Particularly noteworthy is the fact that Tanana and its nearest Alaskan Athapaskan neighbors (Tanacross and Aht na) all share a mean retention ratio (average number of shared cognates) of ex actly 69.2%, consistent with the three languages having diverged from eac h other in situ. Tsuutina has the next-nearest ratio of 69.6% shared cognates, causing Fowler (1977:103-104) to specul ate that Tsuutina and Tanana were also geographical neighbors in this same region at around the same time, a few centuries ago. Tsuutina manufac ture of distinctive low-fired hair-tempered pottery similar to Tanana pottery, and use of the word as clay pot would seem to point to a departure from western Alaska du ring the early Neo-Eskimo period after circa 1000 CE, when hair-tempered pottery crossed Bering Strait and made inroads to the interior. Regarding this dating, Dzensikev ich (1981:123) says that Northern Athapaskan ceramics could not appear in Alaska earlier t han the 10th century, See Figure 6-5 for a visual comparison of Tsuutina and Alaskan ceramics. Corroborating Evidence for Late Dispersals from Western Alaska Several material cultural traits co me into focus here. The abovementioned timeframe is also roughly the period (after circa 1000 CE) when we can say with relative certainty that the Asian sinew-backed bow was reintroduced to western Alaska (Maschner and Jordan 2008:105). This is another Eurasian technology closely identified with a prot o-Athapaskan root-word, *ts -t This is the same period that unprecedented amount of annealed copper-work diffused rapidly through the interior western Subarctic along wit h a booming copper trade (Clark 1991:84). The proto288

PAGE 289

Athapaskan root-word *we knife, metal is linked to several Central Asian source languages, and may be a trans-Eurasian wanderwort (Table 6-1). Finally, it bears mentioning another cultural f eature with a similar geographic distribution to pottery and complex bows. Hard-soled leather footgear is ubiqui tous among the Southern and Pacific Coast Athapaskans, but is found in t he far north among only E skimos, Tsuutina, and western Alaskan Athapaskans, e.g. Deg Hitan (Driver and Massey 1957:326-328). Driver and Massey note: The distribution of the hard-soled moccasi n is more difficult to explain. Its Oasis, Great Basin, and Plains occurr ences favor derivation from the hide sandal. Some of the awkward kinds of Pueblo moccasins suggest the addition of an upper to a sandal. However that may be, a hard and stiff sole is good protection from the thorns and stones of the deserts and plains of the West. In the wooded East, the soft-soled type was sufficient. The hard and separate soled Eskimo boots appear to have developed independently of the hardsoled forms of footgear to the south. The Eskimo boot is apparently of Asiatic origin, and may ev en be historically related to the riding boots there (Driver and Massey 1957:328). Tsuutina hair-tempered pottery traditions suggest rapid long-distance dispersal from western Alaska within the last 1000 year s. The intervening Subarctic is dominated by soft-soled moccasins which are closely as sociated with the use of snowshoes. But nothing precludes the suggestion that the Tsuutina also brou ght their hard-soled shoes with them from western Alaska during thei r rapid expansion. The possible Asiatic derivation of Southern and Pacific Coas t Athapaskan hard-soled footgear must therefore also be considered, as first suggested by Downs (1972:6). The northern origin of Southwestern moccasins has long been suspected, as Bert Salwen notes: [t]he leather shoe did not come into general us e in the Pueblos until after AD 1300. . The Pueblo leather footgear trait might have been bo rrowed from the Athabaskan invaders (Salwen 1960:222). 289

PAGE 290

Maping Vestigial Traits: the Attenuation of Complex Bows Old-fashioned cladistic analysis of materi al culture and technology was concerned with the development of new adaptive features which improve on old designs. In other words, material cultural change is always c onceived in terms of progress towards some ultimate human achievement. But there is no reason to expect tools to be more sophisticated than necessary, and when circumstances change, tools can become less sophisticated. To maintain unnecessary sophistication can be maladaptive. In some ways, my present grand synthesis, successful or not, is indebted to nineteenth-century scholarship, even as I seek to account for twenty-first century data. But many nineteenth-century anthropologists felt the need to place less sophisticated cultures on the bottom rung of an evolutionary ladder. Thus Henry Balfour (1890:224) saw Native American complex bows as the direct histor ical forerunners of Eurasian bows, rather than simply as an attenuated side-branch in the dev elopment of archery traditions. In orienting the data this way, technology is portrayed as the supreme force for change in the world, capable of overcoming any obstacle through ever-increasing sophistication. I strenuously argue that the occasional failure of over-designed tools and the subsequent loss of sophistication in technology over time and space are both increasingly relevant topics (alt hough the concept of sophisticat ion is itself suspect). Yet social scientists have not fully appr eciated the implications of the adaptive significance of retrograde movements within the long arc of history. It may even be the case that a loss in technological comple xity is associated with an improvement in function, where circumstances are right. Far from representing a more primitive form of the complex bow constrained in its development by the absence of glue (as Ba lfour and others would suggest), the form of 290

PAGE 291

the complex bow found in the American (and Northeast Asian) Arctic represents a clear functional improvement on the Asian weapon through the abandonment of an unnecessary design feature. The gluelesstype of composite bow is perfectly adapted to circumpolar life. Fish and animal-based glue (perfectly adapted to life in the arid Eurasian steppes) is prone to catastrophic failure in the salt-spray of a kayak during a sea-otter hunt. By having a free cable backing, one can freel y adjust the tension in the truss for a variety of te mperature and humidity conditi ons, as Hamilton argues: Aside from the fact that the Eskimo bow did use driftwood and sinewor antler or even bone when driftwood was not availableits genius lay in using a free sinew back in which the tension could be readily adjusted to meet varying conditions of humidity. . When we take the Eskimo way of life and his environment into full consi deration, we cannot conceive of a more successful solution of an al most impossible problem (Hamilton 1970:50). For an illustration of a tabular tensioner in the cable backing of a Siberian bow, see Figure 6-6. It is not clear whether the 4000-yea r-old paleo-Eskimo complex bow bracers found by Meldgaard (1960, 1962) in the High Arctic were m eant to be used with gluetype weapons or with glueless-type weapons. If it were the former, then the ecological inadequacy of bow-glue could have been a c ontributing factor in the eventual abandonment of pre-Dorset arc hery technology in the American Arctic. But given the very slight margin for error in polar subs istence strategies, a number of intervening factors may have ensued. But even as the glueless complex bow adaptation may have been ultimately absolutely necessary to est ablish archery firmly in the extreme north, other highly derived features of Asian complex bows were retained as aesthetic details by traditional bowyers, even when their adapt ive significance had been forgotten. Not every technological trait is adaptive. The presence of vestigial siyahs (bow limb 291

PAGE 292

extensions) is taken as proof of the Asian origin of the technology. In Asia, bow limb extensions serve to arrest the flight of the string promoting uniform release of tension and even flight of the proj ectile. In North America, these distinctive features exist only in imitation of an Asian prototy pe; they do not usually make contact with the string and hence generally serve no function (B ergman and McEwan 1997:158; Hamilton 1970:50). See Figure 6-7 for a detail of a vestigia l siyah from a Copper Inuit bow. In this case it is the retention of a non-functional appendage in the east which has conclusively demonstrated the west-to-east transmission of the technology. Functional siyahs do occasionally appear on Arctic weapons, but the non-functional (vestigial) variety appears to be the norm. But siyahs are inva riably functional wherever they occur in Inner Asian bows, indicating that the siyah technology is almost certainly Asian in origin. The concept of the vestigial feature c an help to explain other elements of my complex bow dataset. I sugges t that the north-to-south transmission of complex-bow technology can be cladisticaly mapped by looking for the loss of original complexity as indicated by vestigial features. One way to do this is to look at the recent adaptation of the Tlingit to a maritime hunting lifestyle Ethnographers have long suspected Tlingit moved from an interior to a coastal envir onmental adaptation wit hing the last 500 to 1000 years (Moss 2004:184). In the process of adapting to sea mammal hunting, they have mostly retained the self bowTlingit co mplex bows are extrem ely rare. Kroeber writes about the recent transfe r of the Asian composite bow to North America, which seems to have bypassed the central regions of the Northwest Coast: Of specific traits common to the No rthwest Coast and Asia, armor is the outstanding one to rise to mind. The si new-backed bow, assuming it to be a modification of the Asiatic composite bow, extends into the Plains and 292

PAGE 293

Southwest, but is uncharacteristic of mo st of the Northwest Coast (Kroeber 1923:18). Complex bows are extremely rare in souther n Tlingit territory, around Sitka Alaska. This is the historical center of Tlingit society. By contrast, northern bands like the Yakutat and Inland Tlingit represent very late incursions into Athapaskan-Eyak territory, and it is among these northern Tlingit that complex bows are more common, having evidently adopted them from assimilated Eyak, Athapaskans, and Chugash Eskimos. The one Northwest Coast glueless-type bow I have seen is SDM 27257, and it is most likely Tlingit. It has functional siyahs, but it has another even more striking example of a vestigial feature; the entire southern-style backing is likely of minimally functional importance at best, because it is not m ade of an elastic material like sinew or conditioned babiche. It is backed using cotton thread (Figure 6-8). This bow is sharply recurved and appears to have been strung backwards. This is a feature held in common with the only other Tlingit bow I know of, from Sitka (NMNH E274444), a donation from the estate of the former gover nor of the Alaska territory in 1912. These two bows have very similar length, width and thickness meas urements at all points, supporting the view that they come from a common cultural s ource. But NMNH E274444 has an additional feature suggesting that it (like SDM 27257) also possesses a vestigial cable backing; it has a D-shaped cross-section, and the backing surf ace of of the stave is semi-circular. This is an extremely rare feature which is not conducive to a functional backing. Murdoch writes of a similar Smithsonian Sitka Tlingit bow (now lost): There is one bow in the Museum, not an Eskimo bow, which is interesting in the present connection. It comes from Sitka, where the Indians use a plain Spruce or cedar bow with a round back and flat belly. The bow in question is of the same shape as the other bow s from the same locality, but the maker, who has evidently had some acquaintance with the handiwork of the nearest Eskimos [or Athapaskan-Eyak speakers], has tried to improve it by 293

PAGE 294

putting on a typical "southern" backing of sinew. This, however, is of but little use, as the round back of the bow is not of the proper shape to receive it, and, in spite of the lashing round the handle, it slips off to one side as soon as the bow is bent (Murdoch 1885:315). The single example of a Southern-Athapska n attributed Arctic-type cable-backed bow that I know of (in a private collection) also has a rounded backing surface, and the cable backing also appears to be slipping off to one side, rather ineffectively (Figure 514A). This feature argues persuasively fo r the northern origin of the bow-design in question, and for the vestigial st atus of the cable backing in this context. The extreme rarity of this type of bow in the south (found only in the westernmost Apachean range) could owe something to the retention of the glueless backing design feature in deference to ancestral tradition, and not due to any thorough understanding of bow engineering and performance. The glueless method of c able-backing was developed for maritime hunting contexts. It is likely vestigial in the arid Southwest, and may demonstrate retained technological knowled ge in the absence of understanding of its underlying purpose. The aridity the desert Southwest is ideal for the use of the nearly ubiquitous glue-type complex bo ws with backing materials most similar to those of the prototypical complex bows of the arid Eurasian interior. Now the question of the southward spread of the glue-type complex bow remains. Experimental archery has demonstrated the outright superiority, in terms of projectile velocity and range, of the doubl e-curved Apache complex bow design over a North American self-bow of equivalent size and shape (Bergman et al. 1988). Bergman and McEwan (1997) admit the Asiati c origin of the Eskimo co mplex bow is irrefutable, but they suggest that independent invent ion and use of the technology of sinew reinforcement among other Native American groups remain possibilities (Bergman and 294

PAGE 295

McEwan 1997:158). I suggest that the double curve shape, like the siyah, may be a vestigial feature of Central Asian bow design, at least in certain contexts. As I have earlier argued, the double curved shape origi nated in Central Asia during the Bronze Age (Figure 5-15), and through the Xiongnu period it was developed into various Siberian manifestations culminating in the ubiquity of comp lex archery in Siberia within the last 2000 years (Figure 6-9). This is on the eve of the expansion of this technology into the New World. The double curved shape commonly occurs in self bows in the Plains and Southwest, and has even appeared in the Athapaskan Subarctic (Figure 5-14). It is my contention that this is a nonfunctional vestige of the recurved form of glue-type sinew backed bows. The performance characteristi cs of a Sioux double-curved self bow are significantly poorer than an Apache sinew backed bow of the same basic shape (Bergman et al. 1988:662-665). I contend that the double curved design of self bows was itself a vestige of comple xity (like the arctic siyah), retained for aesthetic reasons and in deference to tradition. It could also be that the shape was retained because in the hands of a warrior, it could be mistaken for a superior complex bow from a distance. But there are limits to this sort of adaptation, and recurving a selfbow needlessly has the drawbacks of increasing material stresses in the wood, breaking down the cellular structure without the compensat ion in tensile str ength provided by the sinew-backing. There is further reason to suggest that the double curved shape by itself could be maladaptive in self bows, even without accounting for the increased strain and fatigue it causes with the bowlimbs. Among those Western Apaches who lost the sinew backing technology, they continued the complicated bo wlimb conditioning process to make the 295

PAGE 296

double curved shape in deference to tradition, at least until someo ne pointed out that this was not a very practical approach. As one of Grenville Goodwins anonymous informants states: We used to make our bows in two shapes, the single arc bow and the double arc bow. The double arc bow you ge t into that shape when it is still green. The double arc bow we had before the single arc bow. But later we learned that the single arc bow was bette r because there was more room to string and draw an arrow on it, there being no hump in the middle (Goodwin and Basso 1971:224). In a precise analogy to the case of vestigial siyahs which demonstrate the Asian origin of the glueless (Eskimo) complex bow, we find prevalent knowledge of proper recurved bowmaking techniques in the absence of the understanding of the function of this technique as a complement to the co mplex backing principle. This is a strong indication that neither the double curved s hape, nor the glue-type sinew backing method was developed independently in New World contexts. The case that it is part of the Dene-Yeniseian cultural legacy remains strong. Final Thoughts My project is indebted to the pioneering work of Jordan Paper (1993), who has challenged conventional wisdom regading Na-Dene origins. In his earlier synthesis of archaeology, genetics, oral tradition and material culture, Paper first connected the dots between complex archery, D ene-Central Asian language links, and Athapaskan oral traditions which explicitly claim Central Asian origins. Notably, he also raised awareness of the unique Chiric ahua Apache use of Central Asian style thumb rings, whose presence implies the former prevalence of a Mongolian style arrow release (otherwise virtually absent in the New World) The nearest use of this thumb ring (found 296

PAGE 297

invariably in association with complex ar chery) is among NeoSiberian reindeer pastoralists in the Amur River region of Southeast Siberia. Paper writes: The archeological evidence suggests that Athapaskan speaking peoples did not settle in the interior of British Columbia until approximately a thousand years ago. When they arrived in this area, they and they alone were familiar with the bow and meta l arrow points. Apachean speakers then, according to linguistic and ethnohistorical data, would not have remained for any length of time, but continued their journey southwards to reach their present locale approx imately 600 years ago. . The Chiricahua migration myth accord s with a Bering Strait crossing interpretation, although this is not a necessary one. That this interpretation comes from the Chiricahua themselves, however, strengthens the possibility. . The loss of the tec hnologically complicated full recurve bow is not unexpected; only the Inuit mainta ined continued contact with Asia and hunted very large animals over long di stances. Also, the use of the bow tends to disappear with the introduction of firearms. Only those cultures that did not readily have firearms availabl e until the late eighteenth century would have fully retained their arc hery complex until the beginning of modem ethnology (Paper 1993:8-9). Hair-tempered pottery, forged copper bifaces and daggers, and sinew-backed bows were standard equipment in western inte rior Alaska during the second millennium CE (but not during the first millennium), and the Athapaskans residing there used words for these tools related by common descent (r ather than borrowing) to similar words for similar technologies in the far-flung peripherie s of the zones of the familys phyletic expansion; often these words are found to be related to Cent ral Asian words. Neither pottery nor sinew-backed bows have yet been found (archaeologically or ethnographically) in a vast sect or of western-interior Ala ska or adjacent regions of British Columbia or Yukon, a region tradi tionally regarded as the proto-Athapaskan homeland. Nor am I aware of any Athapaskan languages of this putative homeland region which retain the proto-Athapaskan ro otwords for pottery or sinew backed bows, suggesting these terms may have been lost ar ound the time of initial settlement. 297

PAGE 298

298 The collective implications of the linguistic evidence, in conjunction with ethnographic and archaeological data, are unanimous in favoring a revised timeframe for the Athapaskan expansion, commencing in the second millennium CE (slightly later than is generally recognized) from an epicenter slightly to the west of where it is commonly regarded by specialists today, and rapi dly resulting in a so uthern expansion, with the first Southwestern presence slightly earlier than most specialists would favor. The linguistic data come into focus with precis ion here. Consistent with the principle of Occams razor, this model is simpler and more efficient than other scenarios for the Athapaskan migrations, and it accounts for a broader range of linguistic, ethnological and biological evidence, in addition to helping to explain the distribution of complex archery in the American West. Most of the detailed work to support this model is still in its preliminary stages, but I suggest that a paradigm shift is now underway. As Vine Deloria has written: Samuel Eliot Morris on is now dead, and we need no longer cringe in fear that he will discredi t us for believing that som eone besides Columbus visited these shores (Deloria 1992:597).

PAGE 299

Table 6-1. Metal/Knife Terms in Prot o-Athapaskan, Eyak and Eurasian Languages. Adapted from (Golla 1998) with additi ons from (Alderson and Iz 1959; Clauson 1972; Figueiredo 2008; Greenfel d 1973; Kari 1978; Takata 2000). North America Proto-Athapaskan *we w knife, (metal, stone) for knife [= *me ] Upper Tanana basr stone Lower Tanana basr knife metal, tin Ahtna baes stone, la-baes large circular stone or copper knife Denaina vash-la round, womans knife; ulu, cha-vash large ulu Deg Hitan vaxa knife, adze, tool, tavasr womans knife, ulu Carrier -bes knife (in compounds) Hare bie knife Bearlake bes i arrowhead Chipewyan bs, bis knife Slavey nbh [mb ] metal Dogrib me [mb ] knife Tsuutina mas knife Plains Apache b metal White Mountian bee knife, metal San Carlos b knife, metal (also in Chiricahua) Navajo bsh ~ bee knife, flint, metal, iron Eyak we.g ( -g ) ulu; womans scraper knife. Eurasia Samoyed Enets (Yenisei) bese of metal Nganasan (Tavgi) bsa of metal Kamas baza of metal Nenets (Yurak) wese ~ yeese of metal; wiee iron North Caucasian Ubykh w a copper Turkic Uyghur (Buddhist) b ak knife; b ~ bi to cut Turkish b ak knife Azeri mis copper (also Uyghur, Kazakh, Turkmen, Uzbek) Kyrgyz jez copper Sino Tibetan Preclassic Tibetan pi-cag knife (loanword from Buddhist Uyghur) Mongolian mes edged tool or weapon, knife, sword es ( ed, dialectal) copper 299

PAGE 300

Table 6-2. Muskogean, Yukian and Penutian Archery Terms, ada pted from Greenberg (1987:146) with additions and corrections. Penutian California Maidu nok o arrow Wintu n t arrow (von Sadovszky 1996): 26) Yokuts (Yaudanchi) nuk on bow Yokuts (Gashowu) nek -, nuk bow Oregon Kalipuya enuk bow Yukian Wappo lka bow (corrected from luka, Kimball 1992:453) Muskogean Proto-Muskogean ak arrow (Kimball 1992:453) Choctaw naki arrow (Kimball 1992:453) Alabama aki arrow (corrected from nak Kimball 1992:453) Koasati aki arrow Table 6-3. Siberian Uralic and California Penutian Archery lexicons Adapted from von Sadovszky (1996:26-27). Siberian Uralic California Penutian jow-t bow (Vogul) jawe bow (Central Sierra Miwok) l, t arrow (Vogul, Ostyak, south) t arrow (Wintu) k libowstring (Vogul) k li bowstring (Patwin) tul quiver (Ostyak, north) tul-im quiver (Clear Lake Miwok) pos broad arrow head (Ostyak) bos arrowhead, knife, spearhead (Nisenan) kesi knife (Vogul) ki e knife (Central Sierra Miwok) Table 6-4. Athapaskan pottery terms (Kari 1989:566; 1978:86 and Sapir 1923:253). Proto-Athapaskan *onsa *asa *usa:k clay pot, kettle Deg Hitan (Yukon) e'o eo'x, e'ok ethok clay pot Denaina (Lime) *usa:k clay pot Carrier o sa kettle Sekani o sha kettle Hare o fwa kettle Mountain o fa, kettle Beaver sa pail Tsuutina as clay pot Navajo 's pot, native pottery 300

PAGE 301

Figure 6-1. Haplogroup C3 Y-DNA distribution overlaid wit h complex archery distribution. 301

PAGE 302

Figure 6-2. Comparison of complex bow di stribution in the cont iguous 48 states with mean annual precipitation. A. B. Figure 6-3. Comparison of Tlingit and Scyt hian scale armor. A) Tarku Inland TlingitAthapaskan scale armor using Chinese coins (Laufer 1914: pl. 13). B.) Scythian leather scale armor, sixt h century BCE. Courtesy of the Metropolitan Museum of Art, New York. 302

PAGE 303

Figure 6-4. Estimated chronology for the adoption of the bow in North America (Blitz 1988); 1988 Baywood Publishing Co mpany, used by permission. A. B. Figure 6-5. Comparison of Tsuutina and Al askan flat-bottomed pottery vessels. A). Crude model of Tsuutina hair-tempered pottery pail, made by wife of Two Guns (Sapir 1923:250). B). Yupik Eskimo pot collected by E.W. Nelson in the late nineteenth-century (Holmes 1910:298). 303

PAGE 304

Figure 6-6. JES 2649; Detail of tensioner (cable tension adjuster), late nineteenthcentury Yuit hide cable-backed bow, East Cape, Siberia. Photo by author. Courtesy of the San Diego Museum of Man. Figure 6-7. MPM 33610; detail of vestigial siyah on Copper Inuit bow. Note the lack of contact between the limb-tip extenston and the bow string. P hoto by author. Courtesy of the Milwaukee Public Museum. Figure 6-8. SDM 27257; detail of cotton cable backing near t he siyah of a likely Tlingit bow. Photo by author. Courtesy of the San Diego Museum of Man. 304

PAGE 305

305 Figure 6-9. Yukaghir composite bow, Northeast Siberia (Jochelson 1910:383). Figure 6-10. GMA 7336.284; Apache double-curved self bow. Photo by author. Courtesy of the Gilcrease Museum of the Americas.

PAGE 306

APPENDIX A INTERNAL CLASSIFICATION OF THE ATHAPASKAN LANGUAGES (Adapted from Hoijer 1971; Jacobs 1937; Kr auss & Golla 1981; Whistler 1979; Pierce and Ryherd 1964; Shipley 1978). The numberi ng scheme approximates subdivisions based on mutual intellig ibility. Symbol indicates dialect-like mutual intelligibility between units which are often classified as separate languages. Forward slash / indicates alternate names for one group; dash - between true dialects of one language. I. Northern Athapaskan (Alaska, western Canada). The number of Northern Athapaskan languages is difficult to determine because language boundaries are generally soft with some degree of mutual intelligibility. There are probably 20 to 30 languages; the imprecise number reflects more or less constant intergroup communication. 1) Denaina 2) Ahtna 3) Deg Hitan/Ingalik Holikachuk/Innoko Koyukon 4) Upper Tanana Tanacross/Nabesna Lower Tanana Upper Kuskokwim/Kolchan; Upper Kuskokwim also has some notable linguistic affinities with (1) and (2), and is geographi cally and culturally linked to (3). 5) Tutchone; northern and southern dialects sometimes classed as separate languages. 6) Gwichin; with eastern and western dialects. 7) Han; closely related to (6). Han speakers often understand (6), but not the reverse. 8) Tahltan Kaska Tagish; (Tahltan + Kaska = Nahanni). 9) Tsetstaut; poorly documented in termediary between (8) and (10). 10) Sekani Beaver/Dunneza Tsuutina/Sarsi 11) Chipewyan; includes Yellowknif e dialect and Chipewyan proper. 12) Slavey Bearlake Mountain Hare 13) Dogrib 14) Carrier/Dakelh Babine/North Carrier 15) Chilcotin Nicola; relatives of (14). Some of these languages/dialects have been further grouped into larger genetic sub-families (i.e. family-tree branches) by various scholars, but these groupings are somewhat arbitrary as they d epend on which particular features are compareda heterogeneous patchwork of innovations makes for multiple distinct overlapping zones of aff iliation. Proposed subgroups include the following: A) Tanana-Slavey-Chipewyan (Hoijer 1963) B) Tanana-Koyukon (Dumond 1969) C) Tannacross-Tutchone (Hoijer 1963) D) Deg Hitan-Gwichin-Han (Hoijer 1963) E) Ahtna-Tahltan-Kaska-Sekani-Tsu utina-Beaver (Hoijer 1963) 306

PAGE 307

307 F) Hare-Dogrib (Hoijer 1963) G) Northeastern Athapaskan: Mounti an-Hare-Bearlake-Slavey-DogribChipewyan (Howren 1975) H) Canadian Athapaskan: all Athapaskan lan guages of Canada, plus Gwichin, Han, and Upper Tanana straddling the Alaska border (Dumond 1969). II. Pacific Coast Athapaskan (Washington, Oregon, California). A) Kwalhioqua-Clatskanie-Owilapsh (C olumbia basin, Washington and Oregon). Whistler (1979:14) writes: perhaps one language with 3 dialects . considered either as a (geographical) part of Pacifi c Coast Athapaskan or as a (linguistic) intermediate between Pa cific Coast Athapaskan and Northern Athapaskan. Nicola Athapaskan, a central Plateau dialect of Chilcotin located in southern British Columbia, is discontinuous with the rest of the northern bloc, and geographically in termediate between KwalhioquaClatskanie-Owilapsh and Chilcotin. Nicola might also be an intermediary with KCO, e.g. Chilcotin Nicola KCO; however Nicola and KCO are both too poorly documented to be certain of th is. Some specialists classify KCO as its own independent Athapaskan division, rather than as a subdivision of PCA. B) Oregon Athapaskan (Southern Or egon-Northern California) 1) Upper Umpqua [probably one language; dialect relationships unknown] 2) Rogue River; several languages, probably connected by a dialect continuum: Coquille Tututni Chastacosta-Chetco-Tolowa GaliceApplegate C) California Athapaskan 1) Hupa-Chilula-Whilkut [one language, three dialects] 2) Mattole Wailaki-Nongatl-Lassik-Sinkyone-Cahto [2 languages, probably connected by a dialect continuum] III. Southern Athapaskan (Apachean). Two languages. Some mutual intelligibility between Apachean dialects and Canad ian Athapaskan languages (e.g. Chipewyan, Tsuutina) suggests that the Apachean languages are still on the precipice of linguistic differentiation wit h those of the north, i.e. they are practically still dialects of Canadian Athapaskan. This general similarity makes it very difficult to determi ne which Canadian group is genetically closest to Apachean, or where the geographical divergence began. At the time of initial separation, it is unlikely that there was significant differentiation between the southernm ost Canadian languages. 1) Plains Apache/Kiowa Apache 2) Southwestern Apachean [one language, six closely related dialects: Navajo; San Carlos; Chiricahua; Mescalero; Jicarilla; Lipan].

PAGE 308

LIST OF REFERENCES Abel, K. 1993 Drum Songs: Glimpses of Dene History McGill-Queen's University Press, Montreal. Aberle, D. F. 1982 The Future of Navajo Religion In Navajo Religion and Cu lture: Selected Views: Papers in Honor of Leland C. Wyman edited by D. M. Brugge and C. J. Frisbie, pp. 219-231. Museum of New Mexico Papers in Anthropology. Museum of New Mexico Press, Santa Fe. Achilli, A., U. A. Perego, C. M. Bravi, M. D. Coble, Q.-P. Kong, S. R. Woodward, A. Salas, A. Torroni and H.-J. Bandelt 2008 The Phylogeny of the Four Pan-American MtDNA Haplogroups: Implications for Evolutionary and Disease Studies. PLoS ONE 3(3):e1764. Adachi, N., K.-i. Shinoda, K. Umetsu and H. Matsumura 2009 Mitochondrial DNA Analysis of Jomon Skeletons from the Funadomari Site, Hokkaido, and its Implication fo r the Origins of Native American. American Journal of Physical Anthropology 138(3):255-265. Adelung, J. C. and J. S. Vater 1816 Mithridates oder allgemeine Spra chenkunde mit dem Vater Unser als Sprachprobe n bey nahe fnf hundert Sprachen und Mundarten III. Vossischen Buchhandlung, Berlin. Agger, W. A. and H. Maschner 2009 Medieval Norse and the Bidire ctional Spread of Epidemic Disease Between Europe and Northeastern America: A New Hypothesis In The Northern World: AD 900-1400 edited by H. Maschner, O. Mason and R. McGhee, pp. 321-337. University of Utah Press, Salt Lake City. Aikens, C. M. 1966 Fremont-Promontory-Plains Relati onships: Including a Report of Excavations at the Injun Creek and Bear River Number 1 Sites, Northern Utah. University of Utah Anthropological Papers 82 University of Utah Press, Salt Lake City. Alderson, A. D. and F. Iz 1959 The Concise Oxford Turkish Dictionary Clarendon Press, London. Allely, S. and J. Hamm 2002 Encyclopedia of Native American Bows, Arrows & Quivers Volume 2: Plains & Southwest Bois D'Arc Press, Goldthwaite, TX 308

PAGE 309

Arnaiz-Villena, A., C. PargaLozano, E. Moreno, C. Ar eces, D. Rey and P. GomezPrieto 2010 The Origin of Amerindians and the Peopling of the Americas According to HLA Genes: Admixture with Asian and Pacific People. Current Genomics 11:103114. Baldwin, S. J. 1997 Apacheans Bearing Gifts: Prehispani c Influence on the Pueblo Indians The Arizona Archaeologist No. 29. Arizona Archaeological Society, Phoenix. Balfour, H. 1890 On the Structure and Affini ties of the Composite Bow. The Journal of the Anthropological Institute of Great Britain and Ireland 19:220-250. Balzer, M. M. 1981 Rituals of Gender Ident ity: Markers of Siberian Khanty Ethnicity, Status, and Belief. American Anthropologist 83(4):850-867. Bancroft, H. H. 1875 The Native Races of the Pacific States of North America, vol. 3, Myths and Languages D. Appleton & Co., New York. Barnett, H. G. 1937 Culture Element Distributions:VII: Oregon Coast. University of California Anthropological Records 1(3):155-204. 1938 Culture Element Distributions : IX: Gulf of Georgia Salish. University of California Anthropological Records 1(5):221-295 Basch, M. 2010 Smithsonian Facility Reopens After Police Search Electronic Document, page=164&nid=708&sid=1930771 accessed February 10, 2011. Basso, K. H. 1966 The Gift of Changing Woman. Bulletin of the Bureau of American Ethnology 196(76):113-159. Battilana, J., L. Cardoso-Silva, R. Barrantes, K. Hill, A. M. Hurtado, F. M. Salzano and S. L. Bonatto 2007 Molecular Variability of the 16p13.3 Region in Amerindians and its Anthropological Significance. Annals of Human Genetics 71(1):64-76. 309

PAGE 310

Behr, W. 2004 'To Translate' is 'To Exchange'Li nguistic Diversity and the Terms for Translation in Ancient China In Mapping Meanings: The Field of New Learning in Late Qing China edited by M. Lackner and N. Vittinghoff, pp. 173-209. Sinica Leidensia. Brill, Leiden. Bellezza, J. V. 1998 Thogchags: Talismans of Tibet. Arts of Asia 28(3):44-64. Bellwood, P. 1996 Phylogeny vs Reticulation in Prehistory. Antiquity 70(270):881-890. Bengtson, J. D. 1994 Edward Sapir and t he "Sino-Dene" Hypothesis. Anthropological Science: Journal of the Anthropol ogical Society of Nippon 102(3):207-230. Benson, F. C. 1975 The Northern Athapaskan Bow: An Analysis of the Role and Importance of Trajectory Weapons in the Hunting Economy of the Kutchin Masters thesis, Department of Anthropology, Washi ngton State University, Pullman. Berezkin, Y. 2005 Cosmic Hunt: Variants of Siberian-North American Myth. Folklore (Tartu) 31:79-100. Bergen, A. W., C. Y. W ang, J. Tsai, K. Jefferson, C. Dey, K. D. Smith, S. C. Park, S. J. Tsai and D. Goldman 1999 An AsianNative American Paternal Lineage Identified by RPS4Y Resequencing and by Micros atellite Haplotyping. Annals of Human Genetics 63(1):63-80. Bergman, C. A. and E. McEwen 1997 Sinew-Reinforced and Composite Bows: Technology, Function, and Social Implications. In Projectile Technology pp. 143-164. Plenum Press, New York. Bergman, C. A., E. McEwen and R. Miller 1988 Experimental Archery: Projectile Velocities and Comparison of Bow Performances. Antiquity 62(237):658-670. Bernard, H. R. 1995 Research Methods in Anthropolog y: Qualitative and Quantitative Approaches 2nd ed. AltaMira, Walnut Creek, CA. 310

PAGE 311

Bird, L. and R. Bohr 2001 Our Voices: Guns and Bows, Omushkego Oral History Project. Electronic document, accessed Feburary 10, 2011. Birket-Smith, K. and F. de Laguna 1938 The Eyak Indians of the Copper River Delta, Alaska Levin and Munksgaard, Copenhagen. Blaek, V. and J. D. Bengtson 1995 Lexica Dene-Caucasica. Central Asiatic Journal 39(1):11-50. Black, L. 1973 The Nivkh (Gilyak) of Sa khalin and the Lower Amur. Arctic Anthropology 10(1):1-110. Blitz, J. H. 1988 Adoption of the Bow in Prehistoric North America. North American Archaeologist 9(2):123-145. Boas, F. 1894 Classification of the Languag es of the North Pacific Coast In Memoirs of the International Congress of Anthropology edited by C. S. Wake, pp. 339-346. Schulte Publishing Company, Chicago. 1940 Race, Language and Culture. Macmillan, New York. Bohr, R. 2005 Introduction: Arrows and Thunder sticks: Technologies Old and New In Telling Our Stories: Omushkego Legends and Histories from Hudson Bay edited by L. Bird, pp. 189-196. Br oadview Press, Peterborough. Bolnick, D. A., D. I. Bolnick and D. G. Smith 2006 Asymmetric Male and Female Genetic Histories among Native Americans from Eastern North America. Molecular Biology and Evolution 23(11):2161-2174. Bolnick, D. A., B. A. Shook L. Campbell and I. Goddard 2004 Problematic Use of Greenberg's Linguistic Classific ation of the Americas in Studies of Native Am erican Genetic Variation. The American Journal of Human Genetics 75(3):519-522. Bonatto, S. L. and F. M. Salzano 1997 A Single and Early Migration for the Peopling of the Americas Supported by Mitochondrial DNA Sequence Data. Proceedings of the National Academy of Sciences 94(5):1866-1871. 311

PAGE 312

Bortolini, M.-C., F. M. Salzan o, M. G. Thomas, S. Stuart, S. P. K. Nasanen, C. H. D. Bau, M. H. Hutz, Z. Layrisse, M. L. Petzl-Erler, L. T. Tsunet o, K. Hill, A. M. Hurtado, D. Castro-de-Guerra, M. M. Torr es, H. Groot, R. Michalski, P. Nymadawa, G. Bedoya, N. Bradman, D. Labuda and A. Ruiz-Linares 2003 Y-Chromosome Evidence for Diffe ring Ancient Demographic Histories in the Americas. The American Journal of Human Genetics 73(3):524-539. Bosch, E., F. Calafell, Z. H. Rosser, S. Nrby, N. Lynner up, M. E. Hurles and M. A. Jobling 2003 High Level of Male-Biased Scand inavian Admixture in Greenlandic Inuit shown by Y-Chromosomal Analysis. Human Genetics 112(4):353-363. Bourke, J. G. 1892 The Medicine-Men of the Apache In Ninth Annual Report of the Bureau of Ethnology edited by J. W. Powell, pp. 443-603. Smithsonian Institution, Washington D.C. Brant, C. S. 1951 The Kiowa Apache Indians: A Study in Ethnology and Acculturation. Ph.D. Dissertation, Department of Anthr opology, Cornell University, Ithaca. Brasser, T. J. 1979 The Pedigree of the Hugging Be ar Tipi in the Blackfoot Camp. American Indian Art Magazine 5(1):32-39. Brizard, A. 1903 Indian Baskets: Hupa & Klamath Tribes Barnhart and Swasey, San Francisco. Brugge, D. M. 1972 The Navajo Exodus Archaeological Society of New Mexico Supplement No. 5. Archaeological Society of New Mexico, Las Cruces, NM. 1978 A Comparative Study of Navajo Mortuary Practices. American Indian Quarterly 4(4):309-328. Bulatova, N. I. 1997 The Evenk Alga Ritual of Blessing: Matri ona Kurbeltinova in Action In Shamanic Worlds: Rituals and Lore of Siberia and Central Asia edited by M. M. Balzer, pp. 237-244. M.E. Sharpe, Armonke, NY. Burch, E. C. 1988 War and Trade. In Crossroads of Continents: Cultures of Siberia and Alaska edited by W. W. Fitzhugh and A. Crowell, pp. 227-240. Smithsonian, Washington, DC. 312

PAGE 313

Burch, E. C. and T. C. Correll 1972 Alliance and Conflict: Inter Regional Relations in North Alaska In Alliance in Eskimo Society edited by L. Guemple, pp. 17-39. Proceedings of the American Ethnological Society. Univ ersity of Washington, Seattle. Byrne, J. 1885 General Principles of the Structure of Language 1. Trbner & Co., London. Campbell, L. 1997 American Indian Languages: The Historica l Linguistics of Native America Oxford University Press, New York. 1998 Historical Linguistics: An Introduction Edinburgh University Press, Edinburgh, UK. Carmean, K. 2002 Spider Woman Walks This Land: Tradi tional Cultural Properties and the Navajo Nation AltaMira Press, Walnut Creek, CA. Cattelain, P. 1997 Hunting During the Upper Paleo lithic: Bow, Spearthrower, or Both In Projectile Technology, edited by H. Knecht, pp. 213-240. Plenum Press, New York. Chamberlain, A. F. 1907 Fur Trade. Bulletin of the Bureau of American Ethnology 30(1):477-478. Chard, C. S. 1958a Organic Tempering in Northeast Asia and Alaska. American Antiquity 24(2):193-194. 1958b An Outline of the Prehistory of Siberia Part 1. The Pre-Metal Periods. Southwestern Journal of Anthropology 14(1):1-33. Chiao, C. 1982 Navajo Sandpainting and Tibetan Mandala: A Preliminary Comparison In Navajo Religion and Culture: Selected Views: Papers in Honor of Leland C. Wyman, edited by D. M. Brugge and C. J. Frisbie, pp. 21-27. Museum of New Mexico Papers in Anthropology. Museum of New Mexico Press, Santa Fe. Clark, D. W. 1991 Western Subarctic Prehistory. Canadian Prehistory Series. Canadian Museum of Civilization, Hull. 313

PAGE 314

2001 Proto-Athapaskan In Encyclopedia of Prehistory, Volume 2: Arctic and Subarctic edited by P. Peregrine and M. Em ber, pp. 169-178. Kluwer Academic Publishers, New York. Clauson, G. 1972 An Etymological Dictionary of Pre-Thirteenth-Century Turkish The Clarendon Press, Oxford. Cole, D. C. 1981 An Ethnohistory of the Chiricah ua Apache Indian Reservation, 1872-1876 PhD Dissertation, American Studies, University of New Mexico, Albuquerque. 1988 The Chiricahua Apache, 1846-1876: From War to Reservation University of New Mexico Press, Albuquerque. Collins, H. B. 1937 Archeology of St. Lawrence Island, Alaska. Smithsonian Miscellaneous Collections 96(1):1-431. 1954 Arctic Area: Indigenous Period Instituto Panamericano de Geografa Historia, Mexico City. Comrie, B. 2010 The Dene-Yeniseian Hy pothesis: An Introduction. Anthropological Papers of the University of Alaska 5(1/2):25-31. Copeland, J. M. 1998 Navajo Hero Twin Ceremonial Art in Dinetah In Din Bkyah: Papers in Honor of David M. Brugge edited by M. S. Duran and D. T. Kirkpatrick, pp. 5768. Archaeological Society of New Mexico, Albuquerque. Cordaux, R., R. Aunger, G. Bentley, I. Nasidze, S. M. Sirajuddin and M. Stoneking 2004 Independent Origins of Indian Cast e and Tribal Paternal Lineages. Current Biology 14(3):231-235. Crosby, A. W. 1976 Virgin Soil Epidemi cs as a Factor in the Aboriginal Depopulation in America. The William and Mary Quarterly 33(2):289-299. Cruikshank, J. 1979 Athapaskan Women: Lives and Legends. Mercury Series. National Museum of Man, Ottawa. Curtis, E. S. 1928 The North American Indian 18. The Plimpton Press, Norwood, MA. 314

PAGE 315

Czaplicka, M. A. 1914 Aboriginal Siberia: A Study in Social Anthropology The Clarendon Press, Oxford. Dalkesen, N. 2010 Gender Relations Regulated by Customary Law as Reflected through Legends and Historical Records in the Turkic Societies. Fe Dergi: Feminist Ele tiri 2(2):16-25. Dall, W. H. 1897 Alaska and Its Resources Lee and Shepherd Publishers, Boston. de Laguna, F. 1947 The Prehistory of Northern North America as Seen from the Yukon Memoir of the Society for American Archaeology No. 3. So ciety for American Archaeology, Menasha, WI. 1990 Eyak In Handbook of North American Indians edited by W. Suttles, pp. 189-196. vol. 10, W. Sturtevant, general editor. Sm ithsonian Institution, Washington, D.C. Deloria, V., Jr. 1992 Indians, Archaeologists, and the Future. American Antiquity 57(4):595598. Derry, D. E. 1975 Later Athapaskan Prehisto ry: A Migration Hypothesis. Western Canadian Journal of Anthropology 5(3/4):134-147. Derenko, M. V., T. Grzybowski, B. A. Malyarchuk, J. Czarny, D. Mi cickaliwka and I. A. Zakharov 2001 The Presence of Mitochondrial Haplogroup X in Altaians from South Siberia. American Journal of Human Genetics 69(1):237-241. Dimmendaal, G. J. 1995 Do Some Languages Have a Multi-G enetic or Non-Genetic Origin?: An Exercise in Taxonomy In Proceedings of the Fifth Nilo-Saharan Linguistics Colloquium, Nice, August 24-29 1992 edited by R. Nicola and F. Rottland, pp. 357-372. Kppe Verlag, Kln. Dixon, E. J., W. F. Manley and C. M. Lee 2005 The Emerging Archaeology of Glaciers and Ice Patches: Examples from Alaska's Wrangell-St. Elias National Park and Preserve. American Antiquity 70(1):129-143. 315

PAGE 316

Dixon, R. B. and A. L. Kroeber 1919 Linguistic Families of California. University of California Publications in American Archaeology and Ethnology 16(3):47-118. Downs, J. F. 1972 The Navajo Holt, Rinehart and Winston, New York. Driver, H. E. 1939 Culture Element Distributi ons: X: Northwest California. University of California Anthropological Records 1(6):297-433. 1941 Girls' Puberty Rites in Western North America. University of California Anthropological Records 6(2):21-90. 1969 Girls' Puberty Rites and Matrilocal Residence. American Anthropologist 71(5):905-908. Driver, H. E. and W. C. Massey 1957 Comparative Studies of North American Indians. Transactions of the American Philosophical Society 47(2):165-456. Drucker, P. 1950 Culture Element Distribut ions: XXVI: Northwest Coast. University of California Anthropological Records 9(3):157-294. 1963 Indians of the Northwest Coast Natural History Press, Garden City, NY. Dumond, D. E. 1969 Toward a Prehistory of the Na-Dene, with a General Comment on Population Movements am ong Nomadic Hunters. American Anthropologist 71(5):857-863. 1980 A Chronology of Native Alaskan Subsistence Systems. Senri Ethnological Studies 4:23-48. 2009 The "Arctic Maritime" Expans ion: A View from the South In The Northern World, AD 900-1400 edited by H. Maschner, O. Mason and R. McGhee, pp. 5875. University of Utah Press, Salt Lake City. Drr, M. and E. Renner 1995 The History of the Na -Dene Controversy: A Sketch In Language and Culture in North America: Studies in Honor of Heinz-Jrgen Pinnow edited by M. Drr, E. Renner and W. Oleschinski, pp. 3-18. Lincom Europa, Munich. 316

PAGE 317

Dzeniskevich, G. 1981 Ecology and Chronology of Athapascan Settlement on the Southern Coast of Alaska In North American Indian Studi es: European Contributions edited by P. Hovens, pp. 123-126. vol. 1. 2 vols. Edition Herodot, Gttingen. 1994 American-Asian Ties as Reflec ted in Athapaskan Material Culture In Anthropology of the North Pacific Rim edited by W. W. Fitzhugh and V. Chaussonnet, pp. 53-62. Sm ithsonian, Washington DC. Emmons, G. T. 1911 The Tahltan Indians. University of Pennsylvania Museum Anthropological Publications 4(1):1-120. Emmons, G. T. and F. de Laguna 1991 The Tlingit Indians University of Washi ngton Press, Seattle. Enrico, J. 2004 Toward Proto-Na-Dene. Anthropological Linguistics 46(3):229-302. Erickson, R. P. 2009 Autosomal Recessive Diseases among the Athabaskans of the Southwestern United States: Recent Adva nces and Implications for the Future. American Journal of Medical Genetics Part A 149A(11):2602-2611. Esper, T. 1965 The Replacement of the Longbow by Firearms in the English Army. Technology and Culture 6(3):382-393. Essene, F. 1942 Culture Element Distr ibutions: XXI: Round Valley. University of California Anthropological Records 8(1):1-97. Ettlinger, E. 1939 British Amulet s in London Museums. Folklore 50(2):148-175. Fagundes, N. J. R., R. Kanitz, R. Eckert, A. C. S. Valls, M. R. Bogo, F. M. Salzano, D. G. Smith, W. A. Silva, M. A. Zago, A. K. Ribeiro-dos-Santos S. E. B. Santos, M. L. Petzl-Erler and S. L. Bonatto 2008 Mitochondrial Population Genomics Supports a Single Pre-Clovis Origin with a Coastal Route for the Peopling of the Americas. American Journal of Human Genetics 82(3):583-592. Ferg, A. and W. B. Kessel 1987 Subsistence In Western Apache Material Cu lture: the Goodwin and Guenther Collections edited by A. Ferg, pp. 49-86. University of Arizona Press, Tucson. 317

PAGE 318

Figueiredo, R. B. 2008 Freelang Online Dictio nary. Electronic document, accessed January 27, 2011. Fitzhugh, W. W. 1994 Crossroads of Conti nents: Review and Prospect. In Anthropology of the North Pacific Rim edited by W. W. Fitzhugh and V. Chaussonnet, pp. 27-51. Smithsonian, Washington DC. Fortescue, M. 1998 Language Relations across Bering Strai t: Reappraising the Archaeological and Linguistic Evidence. Open Linguistics. Cassell, London. 2004 How Far West into Asia have Eskimo Languages Been Spoken, and Which Ones? tudes/Inuit/Studies 28(2):159-183. 2010 Yeniseian: Siberi an Intruder or Remnant. Anthropological Papers of the University of Alaska 5(1/2):310-315. Fowler, W. R., Jr 1977 Linguistic Evidence for Athapaskan Prehistory In Problems in the Prehistory of the North American Subarctic: The Athapaskan Question edited by J. W. Helmer, S. Van Dyke and F. J. Kense, pp. 102-105. Proceedings of the Ninth Annual Chacmool Conference. University of Calgary, Calgary, Alberta. Franklin, J. 1824 Narrative of a Journey to the Shores of the Polar Sea, in the Years 181920-21-22 3rd ed. 2 vols. John Murray, London. Franklin, S. G., S. I. Wolf, Y. zdemir, G. T. Yregir, T. sbir and B. S. Blumberg 1980 Albumin Naskapi Variant in No rth American Indians and Eti Turks. Proceedings of the Na tional Academy of Sciences of the United States of America 77(9):5480-5482. Franklin, U. M., E. Badone, R. Gotthardt and B. Yorga 1981 An Examination of Prehistoric Copper Technology and Copper Sources in Western Arctic and Subarctic North America Mercury Series. National Museums of Canada, Ottawa. Garfield, V. E. 1953 Possibilities of Genetic Relationship in Northern Pacific Moiety Structures. Memoirs of the Society for American Archaeology (9):58-61. 318

PAGE 319

Gat, A. 1999 The Pattern of Fighting in Simple Small-Scale, Prestate Societies. Journal of Anthropological Research 55(4):563-583. Giddings, J. L. 1952 The Arctic Woodland Cult ure of the Kobuk River University Museum, University of Pennsylvania, Philadelphia. Gifford, E. W. 1940 Culture Element Distributions: XII: Apache-Pueblo. University of California Anthropological Records 4(1):1-207. Gilbert, M. T. P., T. Kivisild B. Grnnow, P. K. Andersen, E. Metspalu, M. Reidla, E. Tamm, E. Axelsson, A. Gtherst rm, P. F. Campos, M. Rasm ussen, M. Metspalu, T. F. G. Higham, J.-L. Schwenninger, R. Nathan, C.-J De Hoog, A. Koch, L. N. Mller, C. Andreasen, M. Meldgaard, R. Villems C. Bendixen and E. Willerslev 2008 Paleo-Eskimo mtDNA Genome Reveals Matrilineal Discontinuity in Greenland. Science 320(5884):1787-1789. Gillespie, B. C. 1975 Territorial Expansion of t he Chipewyan in the 18th Century In Proceedings: Northern At hapaskan Conference, 1971 edited by A. M. Clark, pp. 350-388. National Museum of Man Mercur y Series: Canadian Ethnology Service Paper No. 27. vol. 2. Nationa l Museums of Canada, Ottawa. 1981 Yellowknife In Handbook of North American Indians edited by J. Helm, pp. 285-290. vol. 6. Smithsonian Inst itution Press, Washington D.C. Gillespie, B. C. and I. Goddard 1981 Synonymy for Yellowknife In Handbook of North American Indians edited by J. Helm, pp. 288-289. vol. 6. Smiths onian Institution Pre ss, Washington D.C. Goddard, P. E. 1904 Life and Cult ure of the Hupa. University of California Publications in American Archaeology and Ethnology 1(1):1-88. 1916 The Beaver Indians. Anthropological Papers of the American Museum of Natural History 10(4):201-293. 1920 Has Tlingit a Genet ic Relation to Athapascan? International Journal of American Linguistics 1(4):266-279. Goebel, T., M. R. Waters and D. H. O'Rourke 2008 The Late Pleistocene Dispersal of Modern Hum ans in the Americas. Science 319(5869):1497-1502. 319

PAGE 320

Goggin, J. M. 1949 Plaited Basketry in the New World. Southwestern Journal of Anthropology 5(2):165-168. Gold, P. 1994 Navajo and Tibetan Sacred Wisdom: Circle of the Spirit Inner Traditions, Rochester, VT. Golla, V. 1998 An Old Borrowing for Copper, Knife in North America Paper presented at the Society for the Study of the Indigenous Languages of the Am ericas Annual Meeting, New York. 2007 Linguistic Prehistory In California Prehistory: Colo nization, Culture, and Complexity edited by T. L. Jones and K. A. Klar, pp. 71-82. AltaMira, Lanham. Gonzlez-Jos, R., M. C. Bortolini, F. R. Santos and S. L. Bonatto 2008 The Peopling of Amer ica: Craniofacial Shape Va riation on a Continental Scale and its Interpretation fr om an Interdisciplinary View. American Journal of Physical Anthropology 137(2):175-187. Goodwin, G. and K. H. Basso 1971 Western Apache Ra iding and Warfare University of Arizona Press, Tucson. Gordon, B. C. 2012 The Ancestral Chipewyan Became the Navajo and Apache: New Support for a Northwest Plains-Mountain Rout e to the American Southwest. In From the Land of Ever Winter: Athapaska n Migrations from the Subarctic to the American Southwest edited by D. J. Seymour. University of Utah Press, Sa lt Lake City, in press. Grayson, C. E., M. Fr ench and M. J. O'Brien 2007 Traditional Archery from Six Continents: the Charles E. Grayson Collection University of Missouri Press, Columbia, MO. Greenberg, J. H. 1987 Language in the Americas Stanford University Press, Stanford, CA. Greenberg, J. H., C. G. Tu rner, III and S. L. Zegura 1986 The Settlement of t he Americas: A Comparison of the Linguistic, Dental, and Genetic Evidence. Current Anthropology 27(5):477-497. Greenfeld, P. J. 1973 Cultural Conservatism as an Inhibitor of Ling uistic Change: A Possible Apache Case. International Journal of American Linguistics 39(2):98-104. 320

PAGE 321

Grinnell, G. B. 1920 Who Were the Padouca? American Anthropologist 22(3):248-260. Gruhn, R. 1988 Linguistic Evidence in Support of the Coastal Route of Earliest Entry Into the New World. Man 23(1):77-100. Guenther, M. 1996 Hunter Gatherer Revisionism. In Encyclopedia of Cultural Anthropology edited by D. Levinson and M. Ember, pp. 62 2-624. vol. 2. Henr y Holt, New York. Gullv, H. C. and R. McGhee 2006 Did Bering Strait People Initiate the Thule Migration? Alaska Journal of Anthropology 4(1/2):54-63. Gunnerson, J. H. 1956 Plains-Promontory Relationships. American Antiquity 22(1):69-72. Gunnerson, J. H. and D. A. Gunnerson 1971 Apachean Culture: A St udy in Unity and Diversity In Apachean Culture History and Ethnology, edited by K. H. Basso and M. E. Opler, pp. 7-28. University of Arizona Press, Tucson. Haile, B. 1943 Soul Concept s of the Navaho. Annali Lateranensi 7:59-94. Hall, A. and J. Farrell 2008 Bows and Arrows from Miran, China. Journal of the Society of ArcherAntiquaries 51:89-98. Hall, E. S., Jr. 1969 Speculations on the Late Prehistory of t he Kutchin Athapaskans. Ethnohistory 16(4):317-333. Hamilton, T. M. 1970 The Eskimo Bow and the Asiatic Composite. Arctic Anthropology 6(2):4352. Hammer, M., T. Karafet, H. Park, K. Omoto, S. Harihara, M. Stoneking and S. Horai 2006 Dual Origins of the Japanese: Common Ground for Hunter-Gatherer and Farmer Y Chromosomes. Journal of Human Genetics 51(1):47-58. Hanson, J. A. 1994 Spirits in the Art: From the Pl ains and Southwest Indian Cultures Lowell Press, Kansas City, MO. 321

PAGE 322

Hare, P. G., S. Greer, R. Gott hardt, R. Farnell, V. Bowyer C. Schweger and D. Strand 2004 Ethnographic and Archaeol ogical Investigations of Alpine Ice Patches in Southwest Yukon, Canada. Arctic 57(3):260-272. Harrington, J. P. 1940 Southern Peripheral Athapaskawan Or igins, Divisions, and Migrations. Smithsonian Miscellaneous Collections 100:503-532. Harris, C. 1995 Towards a Geography of White Powe r in the Cordilleran Fur Ttrade. Canadian Geographer 39(2):131-140. Harry, K., L. Frink, B. OToole and A. Charest 2009 How to Make an Unfired Cla y Cooking Pot: Understanding the Technological Choices M ade by Arctic Potters. Journal of Archaeological Method and Theory 16(1):33-50. Headland, T. N. and L. A. Reid 1989 Hunter-Gatherers and Their Neighbor s from Prehistory to the Present. Current Anthropology 30(1):43-66. Helm, J. and N. O. Lurie 1961 The Subsistence Economy of the Dogrib Indians of Lac la Martre in the Mackenzie District of the Northwest Territories Northern Co-ordination and Research Centre, Department of Nort hern Affairs and National Resources, Ottawa. Hey, J. 2005 On the Number of New World Founder s: A Population Genetic Portrait of the Peopling of the Americas. PLoS Biology 3(6):965-975. Heyer, E., J. Puymirat, P. Die ltjes, E. Bakker and P. de Knijff 1997 Estimating Y Chromosome Specific Microsatellite Mutation Frequencies using Deep Rooting Pedigrees. Human Molecular Genetics 6(5):799-803. Heyne, F. G. 1999 The Social Significance of t he Shaman among the Ch inese ReindeerEvenki. Asian Folklore Studies 58(2):377-395. Hoijer, H. 1938 The Southern Athapaskan Languages. American Anthropologist 40(1):7587. 1956 The Chronology of the Athapaskan Languages. International Journal of American Linguistics 22(4):219-232. 322

PAGE 323

1963 The Athapaskan Languages In Studies in the Athapaskan Languages, edited by H. Hoijer, pp. 1. University of California Publicatio ns in Linguistics. University of California Press, Berkeley. 1971 The Position of the Apac hean Languages in the Athapaskan Stock In Apachean Culture History and Ethnology edited by K. H. Basso and M. E. Opler, pp. 3-6. Anthropological Papers of the Univ ersity of Arizona. University of Arizona Press, Tucson. Holmes, W. H. 1910 Pottery. Bulletin of the Bureau of American Ethnology 30(2):295-299. Honigmann, J. J. 1945 Northern and Souther n Athapaskan Eschatology. American Anthropologist 47(3):467-469. 1954 The Kaska Indians: An Ethographic Reconstruction Yale University Publications in Anthropology Number 51. Yale University Press, New Haven, CT. Hopkins, S. W. 1883 Life Among the Piutes: Their Wrongs and Claims Cupples, Upham & Co., Boston. Howard, O. O. 1907 My Life and Experiences among our Hostile Indians A.D. Worthington & Company, Hartford, CT. Howren, R. 1975 Some Isoglosses in McKenzie Drainage Athapaskan: First Steps toward a Subgroup In Proceedings of the Norther n Athapaskan Conference, 1971 pp. 577-618. vol. 2. 2 vols. National Museums of Canada, Ottawa. Hrdli ka, A. 1905 Notes on the San Carlos Apache. American Anthropologist 7(3):480-495. 1925 The Origin and Antiquit y of the American Indian. Annual Report of the Board of Regents of the Smithsonian Institiution 1923:481-494. Humphrey, C. 1994 Shamanic Practices and the State in Northern Asia: Views from the Center and Periphery In Shamanism, History, and the State edited by N. Thomas and C. Humphrey, pp. 191-228. University of Michigan Press, Ann Arbor. 323

PAGE 324

Hunley, K. and J. C. Long 2005 Gene Flow across Linguistic B oundaries in Native North American Populations. Proceedings of the Nationa l Academy of Sciences 102(5):13121317. Huntsinger, L. and S. McCaffrey 1995 A Forest for the Trees: Forest Management and the Yurok Environment, 1850 to 1994. American Indian Culture and Research Journal 19(4):155-192. Hyde, G. E. 1959 Indians of the High Plains: From the Prehistoric Period to the Coming of Europeans The Civilization of the American Indian. University of Oklahoma Press, Norman. Hymes, D. H. 1956 Na-Dn and Positional Analysis of Categories. American Anthropologist 58(4):624-638. 1957 A Note on Athapaskan Glottochronology. International Journal of American Linguistics 23(4):291-297. Igjug rjuk and K. Rasmussen 2001 Seeking Knowledge in the Solitude of Nature In Shamans Through Time: 500 Years on the Path to Knowledge edited by J. Narby and F. Huxley, pp. 8183. Putnam, New York. Ives, J. W. 1990 A Theory of Northern Athapaskan Prehistory Westview Press, Boulder, CO. 2003 Alberta, Athapaskans and Apachean Origins In Archaeology in Alberta: A View from the New Millennium edited by J. W. Brink and J. F. Dormaar, pp. 256289. Archaeological Society of Alberta, Medicine Hat. 2010 Dene-Yeniseian, Mi gration, and Prehistory. Anthropological Papers of the University of Alaska 5(1/2):324-334. Ives, J. W. and S. Rice 2006 Correspondences in Archaeological, Genetic and Linguistic Evidence for Apachean History. Paper presented at the Conference on Languages and Genes, Max Planck Institute, Department of Linguistics, University of California, Santa Barbara. Electronic document, accessed February 1, 2011. 324

PAGE 325

Jacobs, M. 1937 Historic Perspectives in Indian Languages of Oregon and Washington. Pacific Northwest Quarterly 28:55-74. Jschke, H. A. 1881 A Tibetan-English Dictionary Motilal Banarsidass, Delhi. Jenness, D. 1943 The Carrier Indians of the Bulkley River: Their Social and Religious Life. Bulletin of the Bureau of American Ethnology 133(25):469-586. Jett, S. C. 1978 The Origins of Navajo Settlement Patterns. Annals of the Association of American Geographers 68(3):351-362. 1983 Pre-Columbian Transoceanic Contacts In Ancient North Americans edited by J. D. Jennings, pp. 556. W.H. Freeman and Comp any, New York. 2006 Reconstructing the Itineraries of Navajo Chantway Stories: A Trial at Canyon de Chelly, Arizona In Southwestern Interludes: Papers in Honor Charlotte J. and Theodore R. Frisbie edited by R. N. Wise man, T. C. OLaughlin and C. T. Snow, pp. 75-86. vol. 32. Archaeological Societ y of New Mexico, Albuquerque. Jochelson, W. 1910 The Yukaghir and the Yukaghirized Tungus, Part 2. Memoirs of the American Museum of Natural History 13(2):135-342. Kakuyabashi, F. 1981 Japanese Drift Records and the Sharp Hypothesis. Journal of the Polynesian Society 90(4):515-524. Karafet, T., L. Xu, R. Du, W. Wang, S. Feng, R. S. Wells, A. J. Redd, S. L. Zegura and M. F. Hammer 2001 Paternal Population History of East Asia: Sources, Patterns, and Microevolutionary Processes. American Journal of Human Genetics 69(3):615628. Karafet, T. M., L. P. Osipova, M. A. G ubina, O. L. Posukh, S. L. Zegura and M. F. Hammer 2002 High Levels of Y-Chromosome Differentiation among Native Siberian Populations and the Genetic Signature of a Bor eal Hunter-Gatherer Way of Life. Human Biology 74(6):761-790. 325

PAGE 326

Karafet, T. M., S. L. Zegura, O. Posukh, L. Osipova, A. Bergen, J. Long, D. Goldman, W. Klitz, S. Harihara, P. de Kn ijff, V. Wiebe, R. C. Griffiths, A. R. Templeton and M. F. Hammer 1999 Ancestral Asian Source(s) of New World Y-Chromosome Founder Haplotypes. American Journal of Human Genetics 64(3):817-831. Karasulas, A. 2004 Mounted Archers of the Steppe 600 BC-AD 1300 Osprey, Oxford. Kari, J. 1978 Deg Xinag: Ingalik Noun Dictionary Alaska Native Language Center, Fairbanks. 1989 Some Linguistic Insigh ts Into Dena'ina Prehistory In Athapaskan Linguistics: Current Perspectives on a Language Family edited by E.-D. Cook and K. D. Rice, pp. 533-574. Mouton de Gruyter, Berlin. 2007 Dena'ina Topical Dictionary Alaska Native Language Center, Fairbanks. 2010 The Concept of Geolinguistic Conservatism in Na-Dene Prehistory. Anthropological Papers of the University of Alaska 5:194-222. Kari, J. and B. Potter 2010 The Dene Yeniseian Connection: Bridging Asia and North America. Anthropological Papers of the University of Alaska 5(1/2):1-24. Karnakova, A. 1913 Mongolian Sand-painting. American Anthropologist 15(2):370-371. Kaufman, T. and V. Golla 2000 Language Groupings in the New World: Their Reliability and Usability in Cross-Disciplinary Studies In America Past, America Present: Genes and Languages in the Americas and Beyond edited by C. Renfrew, pp. 47-57. McDonald Institute for Archaeol ogical Research, Cambridge. Kaye, A. S. 1992 Distant Genetic Relati onship and Edward Sapir. Semiotica 91(3/4):273300. Kayser, M., L. Roewer, M. Hedman, L. H enke, J. Henke, S. Brauer, C. Krger, M. Krawczak, M. Nagy, T. Dobosz, R. Szibor, P. de Knijff, M. Stonek ing and A. Sajantila 2000 Characteristics and Frequency of Germli ne Mutations at Micr osatellite Loci from the Human Y Chromosome, as Revealed by Direct Observation in Father/Son Pairs. American Journal of Human Genetics 66(5):1580-1588. 326

PAGE 327

Keddie, G. 1990 The Question of Asiatic Object s on the North Pacific Coast of America:Historic or Prehistoric? Contributions to Human History 3:1-26. Kelly, R. L. 1995 The Foraging Spectrum: Diversity in Hunter-Gatherer Lifeways. Smithsonian Institution Press, Washington D.C. Kehoe, A. B. 1981 Revisionist Anthropology: Aboriginal North America. Current Anthropology 22(5):503-517. 2003 The Fringe of Amer ican Archaeology: Transoceanic and Transcontinental Contacts in Prehistoric America. Journal of Scientific Exploration 17(1):19-36. 2010 Consensus and the Fringe in American Archaeology. Archaeologies 6(2):197-214. Keithahn, E. L. 1964 The Origin of the "Chiefs Copper" or Tinneh. Anthropological Papers of the University of Alaska 12(2):59-78. Kessel, W. B. 2005 White Mountain Apache Reflections. Journal of the Southwest 47(1):5769. Keyser-Tracqui, C., E. Crubezy and B. Ludes 2003 Nuclear and Mitochondrial DNA Analys is of a 2,000-Year-Old Necropolis in the Egyin Gol Valley of Mongolia. American Journal of Human Genetics 73(2):247-260. Keyser-Tracqui, C., E. Crubzy, H. Pamzsav, T. Varga and B. Ludes 2006 Population Origins in Mongolia: Genet ic Structure Analysis of Ancient and Modern DNA. American Journal of Physical Anthropology 131(2):272-281. Keyser, C., C. Bouakaze, E. Crubzy, V. Nikolaev, D. Montagnon, T. Reis and B. Ludes 2009 Ancient DNA Provides New Insights into the History of South Siberian Kurgan People. Human Genetics 126(3):395-410. Kharkov, V. N., V. A. Stepanov, O. F. Medvedeva, M. G. Sp iridonova, N. R. Maksimova, A. N. Nogovitsina and V. P. Puzyrev 2008 The Origin of Yakuts: Analysis of the Y-chromosome Haplotypes. Molecular Biology 42(2):198-208. 327

PAGE 328

Kharkov, V. N., V. A. Stepanov O. F. Medvedeva, M. G. Spir idonova, M. I. Voevoda, V. N. Tadinova and V. P. Puzyrev 2007 Gene Pool Differences Detween No rthern and Southern Altaians Inferred from the Data on Y-Chromosomal Haplogroups. Russian Journal of Genetics 43(5):551-562. Khitrinskaya, I. Y., V. A. St epanov, V. P. Puzyrev, M. G. Sp iridonova, K. V. Puzyrev, N. R. Maksimova and A. N. Nogovitsyna 2003 Genetic Peculiarity of the Yakut Population as Inferred from Autosomal Loci. Molecular Biology 37(2):205-209. Kim, K., C. H. Brenner, V. H. Mair, K.-H. Lee, J.-H. Kim, E. Gelegdorj, N. Batbold, Y.-C. Song, H.-W. Yun, E.-J. Chang, G. Lkhagvasuren, M. Bazarragchaa, A.-J. Park, I. Lim, Y.-P. Hong, W. Kim, S.-I. Chung, D.-J. Kim, Y.-H. Chung, S.-S. Ki m, W.-B. Lee and K.Y. Kim 2010 A Western Eurasian Male is F ound in 2000-Year-Old Elite Xiongnu Cemetery in Northeast Mongolia. American Journal of Physical Anthropology 142(3):429-440. Kimball, G. 1992 A Critique of Muskogean, "Gulf," and Yukian Material in "Language in the Americas". International Journal of American Linguistics 58(4):447-501. Klein, S. 1983 Analogy and Mysticism and the Structure of Culture. Current Anthropology 24(2):151-180. Kniffen, F. B., H. F. Gregory and G. A. Stokes 1994 Historic Indian Tribes of Louisi ana: From 1542 to the Present Louisiana State University Press, Baton Rouge. Kolman, C. J., N. Sam buughin and E. Bermingham 1996 Mitochondrial DNA Analysis of Mongol ian Populations and Implications for the Origin of New World Founders. Genetics 142(4):1321-1334. Kozintsev, A. G., A. Gromov and V. Moiseyev 1999 Collateral Relatives of American Indians among the Bronze Age Populations of Siberia? American Journal of Physical Anthropology 108(2):193204. Krantz, G. S. 1977 The Populating of Western North America. Occasional Papers in Method and Theory in California Archaeology 1:1-63. 328

PAGE 329

Krauss, M. E. 1964 Proto-Athapaskan-Ey ak and the Problem of Na-Dene: The Phonology. International Journal of American Linguistics 30(2):118-131. 1973 Na-Dene In Linguistics in North America edited by T. A. Sebeok, pp. 903978. Current Trends in Linguistics vol. 10. Mouton, The Hague. 2005a Athabaskan Tone In Athabaskan Prosody edited by S. Hargus and K. Rice, pp. 51. Current Issues in Linguistic Theory. John Benjamins, Amsterdam. 2005b Eskimo Languages in Asia, 1791 on, and the Wrangel Island-Point Hope Connection. tudes/Inuit/Studies 29(1/2):163-185. Krauss, M. E. and V. K. Golla 1981 Northern Athapaskan Languages In Handbook of North American Indians edited by J. Helm, pp. 67-85. vol. 6. Smithsonian In stitution Press, Washington D.C. Krippner, S. 1997 The Role Played by Mandalas in Navajo and Tibetan Rituals. Anthropology of Consciousness 8(1):22-31. Kroeber, A. L. 1923 American Culture an d the Northwest Coast. American Anthropologist 25(1):1-20. 1948 Anthropology: Race, Language, Culture, Psychology, Prehistory 2nd ed. Harcourt, New York. 1955 Linguistic Time Depth Results so Far and Their Meaning. International Journal of American Linguistics 21(2):91-104. 1959 Reflections and Tests on Athabascan Glottochronology. University of California Publications in American Archaeology and Ethnology 47(3):241-258. Krueger, J. R. 1993 The Mongolian Migrati on to America in 1233: Stew art's Hypothesis of the Escape From Chinggis Khan Paper presented at the International Symposium on Mongolian Culture, Taipei, R.O.C. Krupnik, I. 1993 Arctic Adaptations: Native Whaler s and Reindeer Herders of Northern Eurasia. University Press of New England, Hanover, NH. 329

PAGE 330

Landar, H. J. 1960 The Loss of Athapaskan Words for Fish in the Southwest. International Journal of American Linguistics 26(1):75-77. Laufer, B. 1914 Chinese Clay Figures Part 1: Prolegomena on the History of Defensive Armor. Field Museum of Natural History Anthropological Series 13(2):73-315. LeBlanc, S. A. 1999 Prehistoric Warfare in the American Southwest University of Utah Press, Salt Lake City. Lee, P. T. 1967 The Ket: A Contributio n to the Ethnography of a Central Siberian Tribe PhD Dissertation, Anthropology, Stanford University, Palo Alto. Lell, J. T., R. I. Sukernik, Y. B. Starikovskaya, B. Su, L. Jin, T. G. Schurr, P. A. Underhill and D. C. Wallace 2002 The Dual Origin and Siberian Affinities of Native American Y Chromosomes. American Journal of Human Genetics 70(1):192-206. Lell, J. T., R. I. Suke rnik and D. C. Wallace 2002 Reply to Tarazona-Santos and Santos. American Journal of Human Genetics 70(5):1380-1381. Lewis, D. and A. T. Jordan 2008 Creek Indian Medicine Ways: The E nduring Power of Mvskoke Religion University of New Mexico Press, Albuquerque. Li, C., H. Li, Y. Cui, C. Xie, D. Cai, W. Li, V. Mair, Z. Xu, Q. Zhang, I. Abuduresule, L. Jin, H. Zhu and H. Zhou 2010 Evidence that a West-East Admixed Population Lived in the Tarim Basin as Early as the Early Bronze Age. BMC Biology 8(1):15. Littleton, C. S. 1983 Some Possible Arthurian Them es in Japanese Mythology and Folklore. Journal of Folklore Research 20(1):67-81. 1995 Yamato-takeru: An "Arthurian" Hero in Japanese Tradition. Asian Folklore Studies 54(2):259-274. Longman, C. J. and H. Walrond 1894 Archery Longmans, Green, and Co., London. 330

PAGE 331

Lorenz, J. G. and D. G. Smith 1996 Distribution of Four Founding mtDNA Haplogroups among Native North Americans. American Journal of Physical Anthropology 101(3):307-323. Loud, L. L. and M. R. Harrington 1929 Lovelock Cave. University of California Publications in American Archaeology and Ethnology 25(1):1-183. Mackenzie, A. 1902 Voyages from Montreal through the C ontinent of North America to the Frozen and Pacific Oceans in 1789 and 179 3: With an Account of the Rise and State of the Fur Trade New Amsterdam Book Company, New York. Malhi, R. S., A. Gonzalez-Oliver, K. B. Sc hroeder, B. M. Kemp, J. A. Greenberg, S. Z. Dobrowski, D. G. Smith, A. Resendez, T. Kara fet, M. Hammer, S. Ze gura and T. Brovko 2008 Distribution of Y Chromosomes am ong Native North Am ericans: A Study of Athapaskan Population History. American Journal of Physical Anthropology 137(4):412-424. Malhi, R. S., B. M. Kemp, J. A. Eshleman, J. Cybulski, D. G. Smith, S. Cousins and H. Harry 2007 Mitochondrial Haplogroup M Discover ed in Prehistoric North Americans. Journal of Archaeological Science 34(4):642-648. Malhi, R. S., K. B. Schroeder and B. M. Kemp 2009 Uses and Limitations of Genetic Da ta Relating to Athapaskan Migrations: A Reply to Seymour. American Journal of Physical Anthropology 140(2):203-204. Mandelbaum, D. G. 1940 The Plains Cree. Anthropological Papers of the American Museum of Natural History 37(2):155-316. Martin, D. L. and A. H. Goodman 2002 Health Conditions Before Colum bus: Paleopathology of Native North Americans. Western Journal of Medicine 176(1):65-68. Maschner, H. D. G. and J. W. Jordan 2008 Catastrophic Events and Punctuated Culture Change: The Southern Bering Sea and North Pacific in a Dynamic Global System In Time and Change: Archaeological and Anthropological Pers pectives on the Long-Term in HunterGatherer Societies edited by D. Papag ianni, R. Layton and H. Maschner, pp. 95113. Oxbow Books, Oxford, UK. 331

PAGE 332

Mason, O. K. 1998 The Contest between the Ipiutak, Old Bering Sea, and Birnirk Polities and the Origin of Whaling during the Firs t Millennium A.D. along Bering Strait. Journal of Anthropological Archaeology 17(3):240-325. 2009 Flight from Bering Strait: Did Siberian Punuk/Thule Military Cadres Conquer Northwest Alaska? In The Northern World: AD 900-1400 edited by H. Maschner, O. Mason and R. McG hee, pp. 76-128. University of Utah Press, Salt Lake City. Mason, O. T. 1894 North American Bows, Arrows, and Quivers. Annual Report of the Board of Regents of the Smithsonian Institiution 1893:631-679. 1907 Arrows, Bows, and Quivers. Bulletin of the Bureau of American Ethnology 30(1):91-94. Matson, R. G. and M. P. R. Magne 2007 Athapaskan Migrations: The Archaeology of Eagle Lake, British Columbia. University of Arizona Press, Tucson. Matthews, W. 1897 Navaho Legends Memoirs of the American Fo lk-Lore Society, vol. 5. Houghton Mifflin and Company, Boston. Maxwell, M. S. 1980 Archaeology of the Arctic and Subarctic Zones. Annual Review of Anthropology 9:161-185. McClellan, C. 1981 Tagish. In Handbook of North American Indians edited by J. Helm, pp. 481-492. vol. 6. Smithsonian Instit ution Press, Washington D.C. McClellan, C. and G. Denniston 1981 Environment and Cult ure in the Cordillera In Handbook of North American Indians edited by J. Helm, pp. 372-386. vol. 6. Smithsonian In stitution Press, Washington D.C. McCormack, P. 1988 Northwind Dreaming: Fort Chipewyan 1788-1988. Provincial Museum of Alberta Special Publication No.6. Prov incial Museum of Alberta, Edmonton. McEwan, G. F. and D. B. Dickson 1978 Valdivia, Jomon Fishermen, and the Nature of the North Pacific: Some Nautical Problems with Meggers, Evans and Estrada's (1965) Transoceanic Contact Thesis. American Antiquity 43(3):362-371. 332

PAGE 333

McKennan, R. A. 1959 The Upper Tanana Indians Yale University Publications in Anthropology Number 55. Yale University Press, New Haven. Meggers, B. J. and C. Evans 1969 Speculations on Early Pottery Dif fusion Routes Between South and Middle America. Biotropica 1(1):20-27. Meggers, B. J., C. Evans and E. Estrada 1965 Early Formative Period of Coasta l Ecuador: the Valdivia and Machalilla Phases. Smithsonian Contributio ns to Anthropology 1. Meiklejohn, C. 1977 Genetic Differentiation and Deme Structure: Considerations for an Inderstanding of the Athapaskan/ Algonkian Continuum In Problems in the Prehistory of the North American Subarctic: The Athapaskan Question edited by J. W. Helmer, S. Van Dyke and F. J. Kense, pp. 106-110. The Archaeological Association of the Universi ty of Calgary, Calgary. Meldgaard, J. 1960 Origin and Evolution of Eskimo Cultures in the Eastern Arctic. Canadian Geographical Journal 60(2):64-75. 1962 On the Formative Peri od of the Dorset Culture In Prehistoric Cultural Relations Between the Arctic and Te mperate Zones of North America edited by J. M. Campbell, pp. 92-95. Arctic Institut e of North America, Montreal. Mellaart, J. 1980 Turkish Kilims. Anatolian Studies 30:91-99. Meredith, J. 2002 Mapping Identity: Tibetans in New Mexico and their Incorporation into the Tri-ethnic Myth. Journal of the Southwest 44(1):105-119. Merriwether, D. A., F. Rothhammer and R. E. Ferrell 1995 Distribution of the Four Founding Lineage Haplotypes in Native Americans Suggests a Single Wave of Mi gration for the New World. American Journal of Physical Anthropology 98(4):411-430. Michalove, P. A., S. Georg and A. M. Ramer 1998 Current Issues in Linguistic Taxonomy. Annual Review of Anthropology 27(1):451-472. 333

PAGE 334

Milke, W. 1959 Athapaskische Chronologie: Versuch einer Revision. International Journal of American Linguistics 25(3):182-188. Miller, W. R. 1986 Numic Languages In Handbook of North American Indians edited by W. L. D'Azevedo, pp. 98-106. vol. 11, W. Sturtevant, general edi tor. Smithsonian Institution, Washington, D.C. Mills, A. 1994 Reincarnation Belief among North American Indians and Inuit: Context, Distribution, and Variation In Amerindian Rebirth: Reincarnation Belief among North American Indians and Inuit edited by A. Mills and R. Slobodin, pp. 15-37. University of Toronto Press, Toronto. Mironov, N. D. and S. M. Shirokogoroff 1924 Sramana-Shaman: Etymol ogy of the Word "Shaman". Journal of the Royal Asiatic Society, North China Branch 55:105-130. Montgomery, R. E. 1974 A Cross-Cultural St udy of Menstruation, Mens trual Taboos, and Related Social Variables. Ethos 2(2):137-170. Mooder, K. P., A. W. Weber, F. J. Bamforth, A. R. Lieverse, T. G. Schurr, V. I. Bazaliiski and N. A. Savel'ev 2005 Matrilineal Affinities and Prehistor ic Siberian Mortuary Practices: A Case Study from Neolithic Lake Baikal. Journal of Archaeological Science 32(4):619634. Moodie, D. W., A. J. W. Catchpole and K. Abel 1992 Northern Athapaskan Oral Traditi ons and the White River Volcano. Ethnohistory 39(2):148-171. Moore, J. H. 1994 Putting Anthropology Back Together Again: The Ethnogenetic Critique of Cladistic Theory. American Anthropologist 96(4):925-948. Morice, A. G. 1889 The Western Denes, t heir Manners and Customs. Proceedings of the Canadian Institute (3rd series) 7(1):109-174. 1894 Notes Archological, Industrial and Sociological, on the Western Dns with an Ethnographical Sket ch of the Same. Transactions of the Canadian Institute 4(1):1-222. 334

PAGE 335

1914 Northwestern Dns and Northeastern Asiatics: A Study in the Origin of the Former. Transactions of the Ro yal Canadian Institute 10:131-193. 1925 About Cremation. American Anthropologist 27(4):576-577. Morlan, R. E. 1970a Symposium on Northern Athabaskan Prehistory: Introductory Remarks. Canadian Archaeological Association Bulletin 2:1-2. 1970b Toward the Definition of a Prehistoric Athaba skan Culture. Canadian Archaeological Association Bulletin 2:24-33 Morrison, D. 1991 The Copper Inuit Soapstone Trade. Arctic 44(3):239-246. Mortenson, E. 2006 Raven Augury from Tibet to Al aska: Dialects, Divine Agency, and the Bird's-Eye View. In A Communion of Subjects: Animals in Religion, Science, and Ethics, edited by P. Waldau and K. Patton, pp. 423-436. Columbia University Press, New York. Moss, M. L. 2004 The Status of Archaeology and Archaeological Practice in Southeast Alaska in Relation to the Larger Northwest Coast. Arctic Anthropology 41(2):177196. Moss, M. L., M. E. Jon and R. Stuckenrath 1989 The Antiquity of Tlingit Settlement on Admiralty Island, Southeast Alaska. American Antiquity 54(3):534-543. Mulder, M. B., C. L. Nunn and M. C. Towner 2006 Cultural Macroev olution and the Transmission of Traits. Evolutionary Anthropology: Issues, News, and Reviews 15(2):52-64. Mulligan, C. J., K. Hunley, S. Cole and J. C. Long 2004 Population Genetics, History, and Health Patterns in Native Americans. Annual Review of Geno mics and Human Genetics 5(1):295-315. Murdoch, J. 1885 A Study of the Eskimo Bows in the U. S. National Museum. Annual Report of the U. S. National Museum 1884(2):307-316. Murdock, G. P. 1955 North American Social Structure. Davidson Journal of Anthropology 1(2):85-97. 335

PAGE 336

1957 World Ethnographic Sample. American Anthropologist 59(4):664-687. Nater, H. 1994 The Athapaskan Component of Nuxalk. International Journal of American Linguistics 60(2):177-190. Neff, H. 1992 Ceramics and Evolution. Archaeological Method and Theory 4:141-193. Nelson, R. K. 1974 Relationships Between Eskimo and Athapaskan Cultures in Alaska: An Ethnographic Perspective. Arctic Anthropology 11:48-53. Nettle, D. 1999 Linguistic Diversity of the Amer icas can be Reconciled with a Recent Colonization. Proceedings of the Nationa l Academy of Sciences 96(6):33253329. Neumann, G. K. 1952 Archeology and Race in the American Indian In Archeology of Eastern United States edited by J. B. Griffin, pp. 1334. University of Chicago Press, Chicago. Newcomb, W. W., Jr. 1961 The Indians of Texas: From Prehistoric to Modern Times. University of Texas Press, Austin. Nichols, J. 2010 Proving Dene-Yenisei an Genealogical Relatedness. Anthropological Papers of the University of Alaska 5(1/2):299-309. Nikolaev, R. V. 1985 Fol'klor i voprosy tnichesko istorii ketov. Krasnoyarsk State University, Krasnoyarsk, USSR. 1989 Arkheologicheskie issl. Krasno arskogo kraevogo muze a v 1955-1963 gg. In Problemy izucheni a Sibiri v nauchno-issl. rabote muzeev (tezisy): pp. 70-72. Krasnoyarsk Regional Museum, Krasnoyarsk, USSR. Noble, W. C. 1975 Applications of t he Direct Historic Approach in Central District of Mackenzie, NWT In Proceedings: Northern Atha paskan Conference, 1971 edited by A. M. Clark, pp. 760-785. Mercur y Series. vol. 2. National Museums of Canada, Ottawa. 336

PAGE 337

1977 The Taltheilei Shal e Tradition: An Update In Problems in the Prehistory of the North American Subarct ic: The Athapaskan Question, edited by J. W. Helmer, S. Van Dyke and F. J. Kense, pp. 65-71. University of Calgary, Calgary. Norton-McBride, A. 2003 Commentary: Navajos in Tibet All Things Considered. October 9th. National Public Radio. Archived online at: story/story.php? storyId=1460615 accessed March 15, 2011. Obeyesekere, G. 1994 Foreword In Amerindian Rebirth: Reincarnation Belief among North American Indians and Inuit edited by A. Mills and R. Slobodin, pp. xi-xxiv. University of Toronto Press, Toronto. 2002 Imagining Karma: Ethical Transform ation in Amerindian, Buddhist, and Greek Rebirth University of Califor nia Press, Berkeley. O'Brien, M. and T. D. Holland 1990 Variation, Selection, and the Archaeological Record. Archaeological Method and Theory 2:31-79. 1992 The Role of Adaptation in Archaeological Explanation. American Antiquity 57(1):36-59. Ogunwole, S. U. 2002 The American Indian and Alaska Native Population: 2000 Census 2000 Brief. U.S. Census Bu reau, Washington D.C. Opler, M. K. 1971 Plains and Pueblo Influences in Mescalero Apache Culture In Themes in Culture: Essays in Honor of Morris E. Opler edited by M. D. Zamora, J. M. Mahar and O. Henry, pp. 73-112. Kayu manggi Publishe rs, Quezon City, Philippines. Opler, M. E. 1936a A Summary of Jicarilla Apache Culture. American Anthropologist 38(2):202-223. 1936b The Kinship Systems of the Southern Athabaskan-Speaking Tribes. American Anthropologist 38(4):620-633. 1983 The Apachean Culture Pattern and its Origins In Handbook of North American Indians edited by A. Ortiz, pp. 368-392. vol. 10, W. Sturtevant, general editor. Smithsonian Instit ution, Washington, D.C. 337

PAGE 338

Osgood, C. 1933 Tanaina Culture. American Anthropologist 35(4):695-717. 1937 The Ethnography of the Tanaina Yale University Publications in Anthropology Number 16. Yale University Press, New Haven. 1940 Ingalik Material Culture Yale University Publications in Anthropology Number 22. Yale University Press, New Haven. Oswalt, W. 1955 Alaskan Pottery: A Classifica tion and Historical Reconstruction. American Antiquity 21(1):32-43. Pakendorf, B., B. Mora r, L. Tarskaia, M. Kayser, H. Soodyall, A. Rodewald and M. Stoneking 2002 Y-chromosomal Evidence for a Str ong Reduction in Male Population size of Yakuts. Human Genetics 110(2):198-200. Pakendorf, B., I. Novgorodov, V. Osakovskij, A. b. Danilova, A. Protodjakonov and M. Stoneking 2006 Investigating the Effects of Prehi storic Migrations in Siberia: Genetic Variation and the Origins of Yakuts. Human Genetics 120(3):334-353. Pakendorf, B., V. Wiebe, L. A. Tarskaia, V. A. Spitsyn, H. Soodyall, A. Rodewald and M. Stoneking 2003 Mitochondrial DNA Evidence for Ad mixed Origins of Central Siberian Populations. American Journal of Physical Anthropology 120(3):211-224. Paper, J. 1993 A Material Case for a Late Beri ng Strait Crossing Coincident with PreColumbian Trans-Pacific Crossings. Sino-Platonic Papers 39:1-17. Pearson, R. 1968 Migration from Japan to Ecuador: the Japanese Evidence. American Anthropologist 70(1):85-86. Pearson, R. and A. Underhill 1987 The Chinese Neolithic: Recent Trends in Research. American Anthropologist 89(4):807-822. Perego, U. A., N. Angerhofer, M. Pala, A. Olivieri, H. Lanc ioni, B. H. Kashani, V. Carossa, J. E. Ekins, A. G mez-Carballa, G. Huber, B. Zimmermann, D. Corach, N. Babudri, F. Panara, N. M. Myre s, W. Parson, O. Semino, A. Salas, S. R. Woodward, A. Achilli and A. Torroni 2010 The Initial Peopling of the Americas: A Growing Number of Founding Mitochondrial Genomes from Beringia. Genome Research 20(9):1174-1179. 338

PAGE 339

Perry, R. J. 1980 The Apachean Transition from the Subarctic to the Southwest. Plains Anthropologist 25(90):279-296. 1983 Proto-Athapaskan Cu lture: The Use of Ethnographic Reconstruction. American Ethnologist 10(4):715-733. 1989 Matrilineal Descent in a H unting Context: The Athapaskan Case. Ethnology 28(1):33-51. Petitot, E. 1878 Monograph of the D n-Dindji Indians Translated by D. Brymner. Public Archives of Canada, Ottawa. Pierce, J. E. and J. M. Ryherd 1964 The Status of Athapaskan Research in Oregon. International Journal of American Linguistics 30(2):137-143. Pinnow, H.-J. 1985 Das Haida als Na-Dene-Sprache: Materialen zu den Wortfeldern und zur Komparation des Verbs Vlkerkundliche Arbeitsgemeinschaft, Nortorf. Pope, S. 1918 Yahi Archery. University of California Publications in American Archaeology and Ethnology 13(3):103-152. Powers, S. 1877 Tribes of California Contributions to North American Ethnology vol. 3. Government Printing Office, Washington D.C. Pulleyblank, E. G. 1962 The Consonantal System of Old Chinese: Part II. Asia Major 9(2):206-265. Puzyrev, V. P., V. A. Stepanov, M. V. Golubenk o, K. V. Puzyrev, N. R. Maximova, V. N. Kharkov, M. G. Spiri donova and A. N. Nogovitsina 2003 MtDNA and Y-Chromosome Lin eages in the Yakut Population. Russian Journal of Genetics 39(7):816-822. Radloff, L. 1857 ber die Sprache der Ugalachmut. Bulletin historico-philosophique de l'Acadmie Impriale des Sciences de St. Ptersbourg 15:468-524. Rainey, F. G. 1939 Archaeology in Central Alaska. Anthropological Papers of the American Museum of Natural History 36(4):351-405. 339

PAGE 340

Rausing, G. 1967 The Bow: Some Notes on its Origin and Development. Acta Archaeologica Lundensia 6. Verlag, Bonn. Ray, V. F. 1942 Culture Element Dis tributions: XXII Plateau. University of California Anthropological Records 8(2):99-258. Reid, K. C. 1984 Fire and Ice: New Evidence for the Production and Preservation of Late Archaic Fiber-Tempered Pottery in the Middle-Latitude Lowlands. American Antiquity 49(1):55-76. Reischauer, E. O. 1940 The Thunder-Weapon in Ancient Japan. Harvard Journal of Asiatic Studies 5(2):137-141. Renfrew, C. 2010 Archaeogenetics -Towards a 'New Synthesis'? Current Biology 20(4):R162-R165. Ricaut, F.-X., S. Kolodesnikov, C. Keyser-Tra cqui, A. N. Alekseev, E. Crubzy and B. Ludes 2006 Molecular Genetic Analysis of 400-Year-Old Human Remains Found in Two Yakut Burial Sites. American Journal of Physical Anthropology 129(1):55-63. Rogers, R. A. 1985 Glacial Geography and Na tive North American Languages. Quaternary Research 23(1):130-137. Rogers, R. A., L. D. Ma rtin and T. D. Nicklas 1990 Ice-Age Geography and the Distri bution of Native North American Languages. Journal of Biogeography 17(2):131-143. Rogers, S. L. 1940 The Aboriginal Bow and Arrow of North America and Eastern Asia. American Anthropologist 42(2):255-269. Romero, F. C. 1998 A Population Genetic Study of Athabascan-speak ing Populations in the American Southwest, Anthropology, University of New Mexico, Albuquerque. Rosedale, J. 1891 Inscription and History of the Sword of Miles Standish In Catalogue of the Cabinet and Pictures in Pilgrim Hall p. 13. Avery and Doten, Plymouth, MA. 340

PAGE 341

Rubicz, R., K. L. Melv in and M. H. Crawford 2002 Genetic Evidence for the Phyl ogenetic Relationship between Na-Dene and Yeniseian Speakers. Human Biology 74(6):743-760. Ruhlen, M. 1998 The Origin of the Na-Dene. Proceedings of the Na tional Academy of Sciences of the United States of America 95(23):13994-13996. Saillard, J., P. Forster, N. Lynnerup, H.-J. Bandelt and S. Nrby 2000 mtDNA Variation am ong Greenland Eskimos: T he Edge of the Beringian Expansion. American Journal of Human Genetics 67(3):718-726. Salwen, B. 1960 The Introduction of Leather Footgear in the Pueblo Area. Ethnohistory 7(3):206-238. Samuels, J. 1995 Healing Sand: A Com parative Investigation of Navajo Sandpaintings and Tibetan Sand Mandalas. Journal of Ritual Studies 9(1):101-126. Sapir, E. 1915 The Na-Dene Languages, a Preliminary Report. American Anthropologist 17(3):534-558. 1923 A Note on Sarcee Pottery. American Anthropologist 25(2):247-253. 1925 The Similarity of Chinese and Indian Languages. Science 62(1607):xii. 1936 Internal Linguistic Evidence Sugg estive of the Nort hern Origin of the Navaho. American Anthropologist 38(2):224-235. 1991 The Sino-Dene Hypothesis In American Indian Languages 2 edited by V. Golla, pp. 133-139. The Collected Works of Edward Sapir VI. Moutin de Gruyter, Berlin. Sassaman, K. E. 1993 Early Pottery in the Southeast: Tr adition and Innovation in Cooking Technology University of Alabama Press, Tuscaloosa. 2010 The Eastern Archaic, Historicized Altamira Press, Lanham, MD. Sattler, R. A. 1996 Remnants, Renegades, and Runaways: Seminole Ethnogenesis Reconsidered In History, Power, and Identity: Ethnogenesis in the Americas, 1492-1992 edited by J. D. Hill, pp. 36-69. Univer sity of Iowa Press, Iowa City. 341

PAGE 342

Schurr, T. G. 2004 The Peopling of the New World: Perspectives from Molecular Anthropology. Annual Review of Anthropology 33:551-583. Schwarz, M. T. 1997 Unraveling the Anchoring Cord: Navajo Relocation, 1974 to 1996. American Anthropologist 99(1):43-55. Scott, G. R. and D. O'Rourke 2010 Genes Across Beringia: A Physical Anthropological Perspective on the Dene-Yeniseian Hypothesis. Anthropological Papers of the University of Alaska 5(1/2):119-137. Sedlek, K. 2008 The Yeniseian Languages of the 18th Century and Ket and Sino-Tibetan Word Comparisons. Central Asiatic Journal 52(2):219-305. Seymour, D. J. 2008 Despoblado or Athapaskan Heartland: a Methodological Perspective on Ancestral Apache Landscape Use in the Safford Area In Crossroads of the Southwest: Culture, Ethnicity, and Mi gration in Arizonas Safford Basin edited by D. E. Purcell, pp. 119-161. Cambridge Scholars Press, Newcastle, UK. Shafer, R. 1952 Athapaskan and Sino-Tibetan. International Journal of American Linguistics 18(1):12-19. 1957 Note on Athapaskan and Sino-Tibetan. International Journal of American Linguistics 23(2):116-117. Shapiro, H. L. 1931 The Alaskan Eskimo: A Study of the Relationship between the Eskimo and the Chipewyan Indians of Central Canada. Anthropological Papers of the American Museum of Natural History 31(6):347-384. Shinkwin, A. D. 1977 The "Archaeological Visibility" of Northern Athapask ans in the Tanana River Area, Central Alaska: A Discussion In Problems in the Prehistory of the North American Subarctic: The Athapaskan Question edited by J. W. Helmer, S. Van Dyke and F. J. Kense, pp. 4045. Proceedings of the Ninth Annual Chacmool Conference. University of Calgary, Calgary, Alberta. Shipley, W. F. 1978 Native Languages of California In Handbook of North American Indians edited by R. F. Heizer, pp. 80-90. vol. 8. Smithsonian, Washington DC. 342

PAGE 343

Shirokogoroff, S. M. 1933 Social Organization of the Northern Tungus Commercial Press, Shanghai. Sternberg, L. Y. 1904 Bemerkungen ueber die Beziehungen zwischen der Morphologie der giliakischen und amerikanischen Sprachen. Internationaler Amerikanisten Kongress, Stuttgart 1904:137-140. Skeat, W. W. 1912 "Snakestones" and Stone Thunderbolts as Subjects for Systematic Investigation. Folklore 23(1):45-80. Skibo, J. M., M. B. Schiffer and K. C. Reid 1989 Organic-Tempered Pottery : An Experimental Study. American Antiquity 54(1):122-146. Slobodin, R. 1975 Without Fire: A Kutchin Tale of Warfare, Survival and Vengeance In Proceedings of the Northern Athapask an Conference, 1971, pp. 259-301. vol. 1. 2 vols. National Museums of Canada, Ottawa. Smith, C. B. and S. A. Barrett 1961 Sinew-Backed Bow and Its Arrows : As Made by the Yurok Indians of Northwestern California In American Indian Films Series in Ethnology edited by S. A. Barrett, pp. 24 minutes. Translat ed by H. Cooper, C. B. Smith, general editor. University of California Berkeley Department of Anthropology, United States of America. Smith, D. G., J. Lorenz, B. K. Rolfs, R. L. Bettinger, B. Gr een, J. Eshleman, B. Schultz and R. Malhi 2000 Implications of the Distribution of Albumin Nask api and Albumin Mexico for New World Prehistory. American Journal of Physical Anthropology 111(4):557572. Smith, J. G. E. 1987 The Western Woods Cr ee: Anthropological Myth and Historical Reality. American Ethnologist 14(3):434-448. Snow, J. H. 1981 Ingalik In Handbook of North American Indians edited by J. Helm, pp. 602-617. vol. 6. Smithsonian Instit ution Press, Washington D.C. 343

PAGE 344

Starikovskaya, E. B., R. I. Sukernik, O. A. De rbeneva, N. V. Volodko, E. Ruiz-Pesini, A. Torroni, M. D. Brown, M. T. Lott, S. H. Hosseini, K. Huoponen and D. C. Wallace 2005 Mitochondrial DNA Diversity in I ndigenous Populations of the Southern Extent of Siberia, and the Origins of Native American Haplogroups. Annals of Human Genetics 69(1):67-89. Starostin, S. A. 1991 On the Hypothesis of a Genet ic Connection Between the Sino-Tibetan languages and the Yeniseian and North-Caucasian languages In Dene-SinoCaucasian Languages edited by V. Shevoroshkin, pp. 12. Brockmeyer, Bochum. Stefnsson, V. 1913 My Life With the Eskimo Macmillan, New York. Steward, J. H. 1937 Ancient Caves of the Great Salt Lake Region Bulletin of the Bureau of American Ethnology 116. Smithsonian Institution, Washington DC. 1941 Culture Element Distributions: XIII: Nevada Shoshone. University of California Anthropological Records 4(2):209-359. Stewart, E. G. 1991 The Dene and Na-Dene Indian Migrat ion 1233 AD: Escape from Genghis Khan to America ISAC Press, Columbus GA. Stevens, J. L. 2004 The Spirit of Water in the World of Shamanism In The Healing Power of Water edited by M. Emoto, pp. 185-201. Hay House, Inc., Carlsbad, CA. Stevenson, I. 1974 Twenty Cases Suggestiv e of Reincarnation 2nd ed. University Press of Virginia, Charlottesville. Stimmell, C. and R. L. Stromberg 1986 A Reassessment of Thule Eskimo Ceramic Technology In Technology and Style edited by W. D. Kingery, pp. 23 7-250. Ceramics and Civilization. The American Ceramics Society, Columbus, OH. Stone, A. C. and M. Stoneking 1998 mtDNA Analysis of a Prehistoric O neota Population: Implications for the Peopling of the New World. American Journal of Human Genetics 62(5):11531170. 344

PAGE 345

Sutherland, P. D. 2009 The Question of Contact Betw een Dorset Paleo-Eskimos and Early Europeans in the Eastern Arctic In The Northern World: AD 900-1400 edited by H. Maschner, O. Mason and R. McGhee, pp. 279-299. University of Utah Press, Salt Lake City. Sutton, M. Q. 2008 An Introduction to Native North America 3rd ed. Pearson, Boston. Swadesh, M. 1951 Diffusional Cumulation and Archaic Residue as Historical Explanations. Southwestern Journal of Anthropology 7(1):1-21. 1952 Reviews and Notes: Athapaskan and Sino-Tibetan. International Journal of American Linguistics 18(3):178-181. 1954 Time Depths of Am erican Linguistic Groupings. American Anthropologist 56(3):361-364. Swanton, J. R. and R. B. Dixon 1914 Primitive Am erican History. American Anthropologist 16(3):376-412. Szathmary, E. J. E. and N. S. Ossenberg 1978 Are the Biological Differences Between North American Indians and Eskimos Truly Profound? Current Anthropology 19(4):673-701. Takata, T. 2000 Multilingualism in Tun-huang. Acta Asiatica: Bulletin of the Institute of Eastern Culture 78:49-70. Tambets, K., S. Rootsi, T. Kivisild, H. He lp, P. Serk, E.-L. Loogvli, H.-V. Tolk, M. Reidla, E. Metspalu, L. Pliss, O. Bala novsky, A. Pshenichnov, E. Balanovska, M. Gubina, S. Zhadanov, L. Osipova, L. Damba, M. Voevoda, I. Kutuev, M. Bermisheva, E. Khusnutdinova, V. Gusar, E. Grechanina, J. Parik, E. Pennarun, C. Richard, A. Chaventre, J.-P. Moisan, L. Bara, M. Peri i, P. Rudan, R. Terz i, I. Mikerezi, A. Krumina, V. Baumanis, S. Koziel, O. Rickards, G. F. De Stefano, N. Anagno u, K. I. Pappa, E. Michalodimitrakis, V. Ferk, S. Fredi R. Komel, L. Beckman and R. Villems 2004 The Western and Eastern Roots of the Saami: the Story of Genetic Outliers Told by Mitochondrial DNA and Y Chromosomes. American Journal of Human Genetics 74(4):661-682. 345

PAGE 346

Tamm, E., T. Kivisild, M. Reid la, M. Metspalu, D. G. Smith, C. J. Mulligan, C. M. Bravi, O. Rickards, C. Martinez-Labarga, E. K. Khusnutdinova, S. A. Fedorova, M. V. Golubenko, V. A. Stepanov, M. A. Gubina, S. I. Zhadanov, L. P. Ossipova, L. Damba, M. I. Voevoda, J. E. Dipierri R. Villems and R. S. Malhi 2007 Beringian Standstill and Spr ead of Native American Founders. PLoS ONE 2(9):e829. Tarazona-Santos, E. and F. R. Santos 2002 The Peopling of the Amer icas: A Second Major Migration? American Journal of Human Genetics 70(5):1377-1380. Tarskaia, L. A., S. B. Makarov, L. S. Bychkovskaia, G. V. Pai, B. Pakendorf, G. I. Elchinova, B. E. Deriabin and V. A. Spitsyn 2002 Ethnogenetics of Yakuts from Th ree Regions of Republic' of Sakha (Yakutia) Inferred from the Frequencie s of Biochemical Gene Markers. Russian Journal of Genetics 38(9):1088-1097. Teit, J. A. 1900 The Thompson Indians of British Columbia. Memoirs of the American Museum of Natural History 2(4):163-392. 1906 Notes on the Tahltan I ndians of British Columbia. In Boas Anniversary Volume: Anthropological Papers Wr itten in Honor of Franz Boas edited by B. Laufer, pp. 337-349. G. E. Stec hert & Co., New York. Thomas, D. H. 1978 Arrowheads and Atlatl Darts: How the Stones Got the Shaft. American Antiquity 43(3):461-472. Thornton, R. 1986 History, Structure, and Survival: A Comparison of the Yuki (Ukomno'm) and Tolowa (Hush) Indians of Northern California. Ethnology 25(2):119-130. 1987 American Indian Holocaust and Survival: A Population History since 1942 Civilization of the Am erican Indian. University of Oklahoma Press, Norman, OK. Towner, R. H. 2003 Defending the Dintah: Pueblitos in the Ancestral Navajo Homeland University of Utah Press, Salt Lake City. Townsend, J. B. 1979 Indian or Eskimo? Interaction and Identity in Southern Alaska. Arctic Anthropology 16(2):160-182. 346

PAGE 347

Tsuchiyama, T. 1947 A Comparison of the Folklore of the Northern, Southern and Pacific Athabaskans: A Study in Stability of Fo lklore within a Linguistic Stock Ph.D., Anthropology, University of California, Berkeley. Turner, C. G. and G. R. Scott 2007 The Dentition of Am erican Indians: Evolutionary Results and Demographic Implications Following Colonization from Siberia In Handbook of Paleoanthropology edited by W. Henke and I. Tattersall, pp. 1901-1941. Springer, Berlin. Turner, K. 1983 Areal and Genetic Linguistic Affiliati ons of the Salinan. Kansas Working Papers in Linguistics 8(2):215-246. Uinuk-Ool, T. S., N. Takezaki and J. Klein 2003 Ancestry and Kinships of Native Siberian Populations: The HLA Evidence. Evolutionary Anthropology: Issues, News, and Reviews 12(5):231-245. Underhill, P. A., L. Jin, R. Zemans, P. J. Oefner and L. L. Cavalli-Sforza 1996 A Pre-Columbian Y Chromosome-Spec ific Transition and its Implications for Human Evolutionary History. Proceedings of the National Academy of Sciences 93(1):196-200. Underhill, P. A., N. M. Myres, S. Rootsi, M. Metspalu, L. A. Zhivotovsky, R. J. King, A. A. Lin, C.-E. T. Chow, O. Semino, V. Battag lia, I. Kutuev, M. Jarve, G. Chaubey, Q. Ayub, A. Mohyuddin, S. Q. Mehdi, S. Sengupt a, E. I. Rogaev, E. K. Khusnutdinova, A. Pshenichnov, O. Balanovsky, E. Balanovska, N. Jeran, D. H. Augustin, M. Baldovic, R. J. Herrera, K. Thangaraj, V. Si ngh, L. Singh, P. Majumder, P. Rudan, D. Primorac, R. Villems and T. Kivisild 2010 Separating the post-Glacial C oancestry of European and Asian Y Chromosomes within Haplogroup R1a. European Journal of Human Genetics 18(4):479-484. Upham, S. 1982 Politics and Power: A Social and Econo mic History of the Western Pueblo Academic Press, New York. 1984 Adaptive Diversity and Southwestern Abandonment. Journal of Anthropological Research 40(2):235-256. Vajda, E. 2001 Yeniseian Peoples and Languag es: A History of Yeniseian Studies with an Annotated Bibliography and a Source Guide Curzon Press, Richmond, UK. 347

PAGE 348

2009 The Languages of Siberia. Language and Linguistics Compass 3(1):424440. 2010a A Siberian Link with the Na-Dene. Anthropological Papers of the University of Alaska 5:31-99. 2010b Yeniseian, Na-Dene, and Historical Linguistics. Anthropological Papers of the University of Alaska 5:100-118. VanStone, J. W. 1974 Athapaskan Adaptations: Hunters and Fishermen of the Subarctic Forests. Aldine Publishing Company, Chicago. 1996 The Cherry Collection of Deg Hit'an (Ingalik) Material Culture Fieldiana Anthropology (N.S.) 27. Field Museum of Natural History, Chicago. Villaseor, D. V. 1963 Tapestries in Sand: The Spir it of Indian Sandpainting Naturegraph Company, Healdsburg, CA. Volodko, N. V., E. B. Starikovskaya, I. O. Mazunin, N. P. Eltsov, P. V. Naidenko, D. C. Wallace and R. I. Sukernik 2008 Mitochondrial Genome Diversity in Arctic Siberians, with Particular Reference to the Evolutionary History of Beringia and Pleistocenic Peopling of the Americas. The American Journal of Human Genetics 82(5):1084-1100. von Sadovszky, O. J. 1996 The Discovery of California: a Cal-Ugrian Comparative Study Budapest. Akadmiai Kiad. Vovin, A. 2000 Did the Xiong-nu Speak a Yeniseian Language? Central Asiatic journal 44(1):87-104. 2003 Did the Xiongnu Speak a Yeni seian Language? Part 2: Vocabulary In Altaica Budapestinensia MMII: Proceedings of the 45th Permane nt International Altaistic Conference, edited by A. Srkzy and A. Rkos, pp. 389-394. Hungarian Academy of Sciences, Budapest. Wagner, M. R. and R. W. Travis 1979 In Search of the Navajo's Canadian Connection. Annals of the Association of American Geographers 69(3):480-482. Walsh, E. H. C. 1904 A Note on Stone Implements fo und in the Darjeeling District. Journal of the Asiatic Society of Bengal 73(3):20-24. 348

PAGE 349

Wang, S., C. M. Lewis, Jr., M. Jakobsson, S. Ramachandran N. Ray, G. Bedoya, W. Rojas, M. V. Parra, J. A. Moli na, C. Gallo, G. Mazzotti, G. Po letti, K. Hill, A. M. Hurtado, D. Labuda, W. Klitz, R. Barrantes, M. C. Bortolini, F. M. Salzan o, M. L. Petzl-Erler, L. T. Tsuneto, E. Llop, F. Rothhammer, L. Excoffier M. W. Feldman, N. A. Rosenberg and A. Ruiz-Linares 2007 Genetic Variation and Population Structure in Native Americans. PLoS Genetics 3(11):2049-2067. Warburton, M. and R. M. Begay 2005 An Exploration of Navajo-Anasazi Relationships. Ethnohistory 52(3):533561. Washburn, W. E. 1989 Anthropological Advocacy in the Hopi-Navajo Land Dispute. American Anthropologist 91(3):738-743. Weisiger, M. 2004 The Origins of Navajo Pastoralism. Journal of the Southwest 46(2):253282. Weiss, K. and F. Hayashida 2002 KulturCrisis! Cultural Evol ution Going Round in Circles. Evolutionary Anthropology: Issues, News, and Reviews 11(4):136-141. Weiss, K. M. and E. Woolford 1986 Comment: The Settlement of t he Americas: A Comparison of the Linguistic, Dental, and Genetic Evidence. Current Anthropology 27(5):491-492. Wells, S. 2006 Deep Ancestry: Inside the Genographic Project National Geographic Society, Washington D.C. Whistler, K. A. 1979 Linguistic Prehistory of the Northwest California Coastal Area In A Study of Cultural Resources in Redwood National Park edited by P. M. Quick, pp. 1126. U.S. Department of t he Interior, National Park Service, Denver Service Center, Denver. White, L. T., Jr. 1962 Medieval Technology and Social Change Oxford University Press, Oxford. 349

PAGE 350

Wilcox, D. R. 1981 The Entry of Athapaskans into the American Southwest: The Problem Today In The Protohistoric Period in the No rth American Southwest, AD 1450 1700 edited by D. R. Wilcox and W. B. Massey, pp. 213-256. Anthropological Research Papers. Arizona State University, Tempe. Wilke, P. J. 1988 Bow Staves Harvested from Juniper Trees by Indians of Nevada. Journal of California and Great Basin Anthropology 10(1):3-31. Wilkins, J. F. 2006 Unraveling Male and Female His tories from Human Genetic Data. Current Opinion in Genetics & Development 16(6):611-617. Wilmeth, R. 1977 Chilcotin Archaeology: The Direct Historic Approach In Problems in the Prehistory of the North American Subarctic: The Athapaskan Question edited by J. W. Helmer, S. Van Dyke and F. J. Kense, pp. 33-40. vol. 97-101. The Archaeological Association of the University of Calgary, Calgary. Wilson, J. A. P. 2005 Relatives Halfway Round the World: Southern Athabascans and Southern Tarim Fugitives. Limina 11:67-78. 2008 A New Perspective on Later Migrat ion(s): The Possible Recent Origin of Some Native American Haplotypes. Critique of Anthropology 28(3):267-278. Wintemberg, W. J. 1942 The Geographical Distribution of Aboriginal Pottery in Canada. American Antiquity 8(2):129-141. Wissler, C. 1910 Material Culture of the Blackfoot Indians. Anthropological Papers of the American Museum of Natural History 5(1):1-176. 1914 Material Cultures of the North American Indians. American Anthropologist 16(3):447-505. 1920 North American Indians of the Plains 2nd ed. Handbook Series. American Museum of Natural History, New York. Witthoft, J. and F. Eyman 1969 Metallurgy of the Tlingit, Dene, and Eskimo. Expedition 11(3):12-23. 350

PAGE 351

Wixman, R. 1984 The Peoples of the U SSR: An Ethnographic Handbook M.E. Sharpe, Armonk, NY. Woodbury, A. C. 1984 Eskimo and Aleut Languages. In Handbook of North American Indians edited by D. Damas, pp. 49. vol. 5. Smithsonian Institution Press, Washington D.C. Workman, W. B. 1974 The Cultural Significance of a Vo lcanic Ash Which Fell in the Upper Yukon about 1400 Years Ago In International Conference on the Prehistory and Palaeoecology of Western North American Arctic and Subarctic edited by S. Raymond and P. Schledermann, pp. 239-261. University of Calgary, Calgary. Wrangell, F. P. 1839 Statistische und ethnographische Nachrichten ber die Russische Besitzungen an der Norwestkste von Amerika In Beitrge Zur Kenntniss Des Russischen Reiches und Der Angrnzenden Lnder Asiens edited by K. E. von Baer and G. von Helmerson, pp. 1-332. vol. 1. Im Verlage der Kaiserlichen Akademie der Wissenschaften, St. Petersburg. Yao, Y.-G., Q.-P. Kong, C.-Y. Wang, C.-L. Zhu and Y.-P. Zhang 2004 Different Matrilineal Contributions to Genetic Structure of Ethnic Groups in the Silk Road Region in China. Molecular Biology and Evolution 21(12):22652280. Zakharov, I. A., M. V. Derenk o, B. A. Maliarchuk, I. K. Dambueva, C. M. Dorzhu and S. Y. Rychkov 2004 Mitochondrial DNA Variation in the Aboriginal Populations of the AltaiBaikal Region: Implications for the Genet ic History of North Asia and America. Annals of the New York Academy of Sciences 1011(1):21-35. Zegura, S. L., T. M. Karafet, L. A. Zhivotovsky and M. F. Hammer 2004 High-Resolution SNPs and Microsatellite Haplotypes Point to a Single, Recent Entry of Native American Y Chromosomes into the Americas. Molecular Biology and Evolution 21(1):164-175. Zieme, P. 2008 A Note on the Word 'Shaman' in Old Turkic. Shaman 16(1/2):137-142. 351

PAGE 352

352 Zerjal, T., Y. Xue, G. Bertore lle, R. S. Wells, W. Bao, S. Zhu, R. Qamar, Q. Ayub, A. Mohyuddin, S. Fu, P. Li, N. Yu ldasheva, R. Ruzibakiev, J. Xu Q. Shu, R. Du, H. Yang, M. E. Hurles, E. Robinson, T. Gerelsaikhan, B. Dashnyam, S. Q. Mehdi and C. TylerSmith 2003 The Genetic Legacy of the Mongols. The American Journal of Human Genetics 72(3):717-721. Zhong, H., H. Shi, X.-B. Qi, C.-J. Xi ao, L. Jin, R. Z. Ma and B. Su 2010 Global Distribution of Y-Chro mosome Haplogroup C Reveals the Prehistoric Migration Rout es of African Exodus and Early Settlement in East Asia. Journal of Human Genetics 55(7):428-435. Zlojutro, M. 2008 Mitochondrial DNA and Y-Chromo some Variation of Eastern Aleut Populations: Implications for the Genetic Structure and Peopling of the Aleutian Archipelago, Anthropology, Univ ersity of Kansas, Lawrence. Zlojutro, M., L. A. Tarskaia, M. Sorensen, J. J. Snodgrass, W. R. Leonard and M. H. Crawford 2009 Coalescent Simulations of Ya kut mtDNA Variation Suggest Small Founding Population. American Journal of Physical Anthropology 139(4):474482.

PAGE 353

BIOGRAPHICAL SKETCH Joseph A.P. Wilson became interested in Athapaskan communities in 1997 during a six-month stint as a conservation aide wit h the National Park Service near the Navajo Reservation in Arizona. He earned a Bachelor of Science from Kent State University, in Ohio, Summa cum Laude, with Phi Beta Kappa and University Honors in anthropology and religious studies. Joe studied abroad at the University of Leicester, and the University of London, School of Oriental and African Studies, where he earned a Master of Arts in Oriental and African Religions in 2002. His London thesis addresses the similarity between Navajo ceremonialism and medieval Vajray ana Buddhism. After returning to the US, he earned a Master of Science in Industrial Archaeology at Michigan Technological University in 2004, with a thesis addressing the history and archaeology of a small Michigan copper mine. While at Michigan he met and married fellow graduate student Michelle D. Trim. They moved to South Carolina where Joseph worked as a genetic technologist at Gr eenwood Genetic Center and taught at Lander University until 2007, when he began his doctoral studies at the University of Florida. He finished his Ph.D. and graduated from the University of Florida in 2011. Joe and Michelle have two sons: Oscar, born in 2009 and Arthur, born in 2011. 353