Defining a ‘reasonable geographic framework’: Path Distance as native copper provenance in the Arctic, Subarctic, and Northwest Coast

https://doi.org/10.1016/j.jasrep.2019.101903Get rights and content

Highlights

  • Path Distance provides an alternative to geochemical provenance when sufficient geochemical data is lacking.

  • Path Distance models benefit from the explicit inclusion of multiple environmental and behavioral variables.

  • Path Distance can function as a proxy for ‘acquisition cost’ of native copper.

Abstract

The relative purity of geologic native copper, compared to other raw materials such as obsidian, creates challenges for conducting provenance research. Accurate provenance also depends upon a robust dataset of trace element variation within geological sources. These challenges combine in Northwest North America where three native copper source regions (Central Arctic, Western Subarctic, Northwest Coast) supplied multiple technological traditions in native copper through time. This paper utilizes Path Distance models of human mobility and interaction with the landscape as a proxy for geochemical provenance to estimate a ‘reasonable geographic framework’ (sensu Rapp et al., 2000:96) of the geologic origins of copper artifacts spread across Northwest North America. Generated Path Distance values often link archaeological sites with geologic sources different than when using simple Euclidean distances. Because of their generation using real-world variables to estimate mobility, these values can also be considered a proxy for ‘acquisition cost’ – the cost in time, energy, and social capital spent to acquire raw material – that can frame the analysis of variation in copper and other materials-based technologies in the future.

Introduction

The study presented here is part of a larger research project investigating native copper innovation among diverse hunter-gatherers in far northwestern North America. The Central Canadian Arctic and Subarctic, the Western Subarctic, and the Northwest Coast and Plateau (references to ‘Northwest Coast’ throughout the paper include this broader region), all contain source regions with geological formations capable of forming native copper (Fig. 1) and archaeological evidence for the use of native copper by Indigenous People prior to sustained contact with Europeans beginning in the 18th c. (Cooper, 2012; 2016; Franklin et al., 1981; Hunt, 2015). The presence of Northern Dené communities practicing copper metallurgy in both the Western and Central Canadian Subarctic, and the trade and exchange of copper between Western Subarctic Dené and people on the Northwest Coast documented ethnohistorically, raises questions regarding the potential movement of copper artifacts and knowledge between these three regions. But despite the rich oral history surrounding the use of copper in parts of greater northwest North America (e.g., Cooper, 2011; de Laguna 1972), the relationship, if any, between native copper innovation in these three regions remains unclear.

Due to the vast spatial scale of this study region and current lack of adequate sampling of copper sources across all three regions (see Rapp et al., 1990), provenance research is unlikely to answer questions regarding copper innovation and diffusion any time soon. In lieu of matching copper artifacts to material sources using trace elements or isotopic ratios, this paper uses Path Distance analysis to model the influence of mobility on the distribution of archaeological native copper relative to known sources. The resulting model serves as a proxy for the movement of copper across the landscape.

Native copper is copper that occurs geologically in pure metallic form, most frequently in amygdaloidal basalts (see Rapp et al., 2000 for a further discussion of North American native copper geology), rather than as an ore and often exceeds 99.9% purity (Wayman, 1989). This purity allows it to be used without melting or smelting. Copper's face-centered cubic molecular structure gives it malleability. As a raw material for the production of tools in the past, this is important because it allowed copper to be worked and shaped into a variety of forms using similar tools and techniques as were used for lithics and other materials, e.g. stone hammer and anvils. Repetitive hammering results in the copper becoming ‘work-hardened’ - a drastic decrease in malleability resulting from the deformation and flattening of the molecular structure. Annealing is a complementary production technique to hammering that counteracts the effects of work-hardening. Heating copper to temperatures as low as 200 °C (well within the range of a wood camp fire) (Vernon, 1990; Wayman, 1989) results in the re-crystallization of the deformed molecular structure of a hammered piece of copper and restores malleability, allowing for continued shaping through hammering (Ehrhardt, 2009). Although annealing copper may appear similar to the process of hot-forging employed by blacksmiths, it is achieved through lower temperatures and is functionally different than working metals that have been softened in a forge. Native copper technologies do not require forge technologies to achieve the annealing process. Repeated cycles of hammering and annealing allow for a much greater range of potential forms that can be produced from copper than would be possible through hammering alone (Cooper et al., 2015).

Fifty-four discrete sources of native copper have previously been recorded in the Western Subarctic, the majority being concentrated in the Wrangell and St. Elias Mountain ranges that span the modern southern Alaska-Yukon border (Fig. 1 Area 1) (Cooper, 2007; Cooper et al., 2008). Several different ethnolinguistic groups including Northern Dené-speakers, Eyak, Tlingit, and Sugpiat used native copper. However, the Northern Dené groups, especially the Ahtna and Tutchone, are best known for their copper-working skill and control of copper sources. Most copper artifacts from the Western Subarctic are from the traditional territories of these two groups. Copper was used for a variety of tools including knives and projectile points but awls are the most common copper tool type found archaeologically (Cooper, 2012). Native copper metallurgy appears during the Late Prehistoric Period beginning around 1000 CE, possibly a few hundred years earlier (Dixon et al., 2005; Le Blanc, 1984), peaked between 1000 and 1700 CE, and continued until after contact with Europeans in the 18th c. and the subsequent widespread availability of metal trade goods (Cooper, 2012). Copper from Western Subarctic sources was reported to have traveled down the Northwest Coast (Brooks, 1900) as far south as Haida Gwaii (Acheson, 2003; Brooks, 1900) and Vancouver Island (Emmons, 1991), and reached the northernmost Late Prehistoric Alaskan (Inupiat) contexts (Giddings, 1952; Irving, 1953). However, these regions could also have been supplied, respectively, from Northwest Coast and Central Arctic sources.

Native copper that occurs in the Central Arctic (Fig. 1 Area 2), and was utilized by Arctic and Subarctic peoples through history, is concentrated in three general regions – in the vicinity of the lower Coppermine River (including Dismal Lakes and the Coppermine Mountains), in the Natkusiak Formation on southern Victoria Island, and on the islands and coastline of Bathurst Inlet – and are documented both geologically and ethnographically by Inuit and Northern Dené informants (Bell, 1901; Burwash, 1930; Franklin, 1823; see Rapp et al., 1990 and Franklin et al., 1981 for source lists; Geological Survey of Canada - Open File 691, 1980; Hanbury, 1904; Hearne, 1958; Jefferson et al., 1988; Jenness, 1922; Kindle, 1972; Newbury, 1969; O'Neill, 1924; Steffanson, 1914; Williamson et al., 2013).

The earliest documented instances of copper use in the Arctic are near geologic source regions in the central Arctic where roughly 20 native copper artifacts have been attributed to Pre-Dorset and Arctic Small Tool tradition peoples (Clark, 1975; McGhee, 1972; Noble, 1981; Taylor, 1967), the earliest inhabitants of the region (by 2800 BCE, possibly as early as 3200 BCE) (Friesen, 2016; Savelle and Dyke, 2009). There is a sharp increase in the intensity and geographic spread of the use of central Arctic copper among Late Dorset (500–1300 CE) groups outside of geologic source regions (Appelt et al., 2016; LeMoine et al., 2003; Rowley and Rowley, 1997) that shows a marked difference from preceding Pre-Dorset, Early Dorset, and Middle Dorset copper use that are restricted to within several hundred kilometers of major outcrops in the Coppermine Mountains, Bathurst Inlet, and northern Victoria Island. This is coincident with both the increase in use of iron from the Cape York meteorite and with an increase in interaction and communication in an increasingly interconnected Late Dorset world (Appelt et al., 2016; McGhee, 1996). The evidence for use of Coronation Gulf copper continues to increase into the second millennium CE.

The influx of Thule migrants across the Arctic in the thirteenth century CE (Friesen, 2016; Friesen and Arnold, 2008) precedes the broadest dispersal of native copper through trade networks facilitated by Thule dogsled and watercraft (qayaq and umiaq) technologies (McCartney, 1991, McCartney, 1988; Rowley and Rowley, 1997). Because Thule people inhabited Central Arctic source regions around Coronation Gulf, much of the archaeological evidence from this time is from Thule and later Inuinnait (Copper Inuit) context. The transition from pioneering Thule groups to regional Inuit groups (Inuvialuit, Inuinnait, Inupiat, etc.) manifests as behavioral adaptations in subsistence strategies and residence in relation to the onset of the Little Ice Age (Arnold, 2016; Arnold and Hart, 1991; Dawson, 2016; Stevenson, 1997). This period of reorganization impacts the dispersal of copper through trade networks and the composition of patterns of copper tool types in regional technological toolkits (Pike, 2019).

Most of the archaeological evidence of metallurgical practices in the Central Subarctic comes from the Taltheilei tradition. These are primarily Late Taltheilei (as early as 700 CE) but also include some instances from Middle Taltheilei contexts (as early as 200 CE) at Whitefish Lake in the eastern Barrenlands, as well as among more recent Late Precontact and Protohistoric Northern Dené groups (Andrews, 1996; Gordon, 1996; Moodie et al., 1992; Noble, 1981). These Late Precontact and Protohistoric groups descended from the Taltheilei tradition include Dënesųłıné from the eastern Barrenlands east of Great Slave Lake, Tłı̨chǫ from the Central District of Mackenzie between Great Slave Lake and Great Bear Lake, Sahtúgot'ıne from Great Bear Lake, and Wıìlıìdeh/Tetsǫ́t'ıné from North of Great Slave Lake and the Coppermine drainage. Wıìlıìdeh/Tetsǫ́t'ıné in particular were known for their use of copper and, similar to the Inuinnait (Copper Inuit), were labeled as ‘Yellowknives’ by early Europeans. The Gwich'in of northern interior Yukon, northwestern Northwest Territories, and northeastern Alaska also used copper during the Late Precontact and Protohistoric periods. However they are not directly related to the Taltheilei tradition and this region may have been supplied from Western Subarctic sources (Cooper, 2016, Cooper, 2007).

The influx of European metals beginning in the 17th c. alters but does not supplant the use of native copper by Dené and Inuit groups from Central Arctic geological sources, as evidenced in ethnographic accounts from the late 18th c. through the early 20th c. (Bell, 1901; Burwash, 1930; Franklin, 1823; Hanbury, 1904; Hearne, 1958; Jenness, 1922; O'Neill, 1924; Steffanson, 1914).

The Tlingit use of native copper on the northernmost Northwest Coast is attested to in both oral history and archaeology, and is clearly a result of their interaction with Northern Dené people (Ahtna and Tutchone) in the interior. As with nearby interior Alaska and Yukon, native copper metallurgy appears on the northern Northwest Coast during the Late Prehistoric Period (within the last 2000 years), though likely showed up later than in the interior (Cooper, 2006, Cooper, 2012; de Laguna 1972; de Laguna et al., 1964; McClellan, 1975). The Tlingit credited Northern Dené people with innovating copper metallurgy (de Laguna, 1972; McClellan, 1975). The Western Subarctic, specifically, southeast Alaska interior and southwest Yukon, has been suggested as the place of origin for Pre-contact copper found on the southern Northwest Coast, (e.g., Acheson, 2003; Blake, 2004). However, the earliest evidence for the use of copper on the Northwest Coast comes from further south in the Salish Sea, where copper has been recovered from contexts dating as early as 2000 years ago at the beginning of the Marpole phase (Hunt, 2015). This pre-dates archaeological evidence of copper innovation in the Western Subarctic by around 1000 years (Cooper, 2006; Cooper et al., 2008; de Laguna et al., 1964) making it an unlikely recipient of the diffusion of copper innovation from the Western Subarctic. Both Borden (1970) and Mitchell (1971) considered copper to be a defining feature of the Marpole phase, though Burley (1980) expressed doubts regarding this association.

The northern Tlingit archaeological site of Old Town on the Alaskan Coast has produced copper tools such as knives and projectile points, but 16 of the 48 copper artifacts recovered by de Laguna et al. (1964) from habitation and midden contexts are objects of personal adornment. This contrasts with the almost exclusive use of copper for tools by Northern Dené people. On the southern Northwest Coast copper artifacts are almost exclusively objects of bodily adornment associated with mortuary contexts (Hunt, 2015). As a result, archaeological copper has been associated with the development of cultural complexity and social inequality in the region and linked to the development of notions of material wealth and socially inherited status (Ames, 2005; Borden, 1951; Burley, 1980; Clark, 2013; Mitchell, 1971).

Although interior British Columbia (Fig. 1 Area 3) provides Canada with its most commercially productive copper ore deposits (Fredericks et al., 2009), these deeply buried deposits were likely not a viable option for use pre-contact. Native copper might have been collected in nugget form from a number of exposed surface sources (Hunt, 2015). However, while there is a great deal of oral history, linguistic, and place name information associated with the Indigenous use of copper in the western Subarctic (e.g., Cooper, 2011; de Laguna and McClellan, 1981; Kari, 1990; Kari and Fall, 2003) and Central Canadian Arctic (e.g. Collignon, 2006), including its acquisition from specific locations, Indigenous oral history from the Northwest Coast related to copper is different. Much of this oral history appears to be related to the manufacture and display of ‘Coppers’, large shield-shaped objects often associated with potlatches and believed to be largely a post-Contact phenomenon (Cooper, 2006; Emmons, 1991; Rickard, 1939). Though sites of copper working have been excavated by archaeologists in the Western Subarctic (Cooper, 2012), Central Subarctic (Andrews, 1996), and Central Canadian Arctic (McGhee, 1972), no such sites have yet been located along the Northwest Coast (Hunt, 2015).

Section snippets

Path Distance as copper provenance

Native copper provenance research is complicated by its purity; trace elements are typically limited and present in quantities at or near the detection limits of many measuring instruments. As a result, native copper sources in different locations exhibit low ‘between-source’ variability, which can, in turn, result in the assignment of objects to geographically highly unlikely sources (Rapp et al., 2000). This prompted Rapp et al. (2000, p. 96) to urge researchers to operate within a

Methods

Creating explicit models formalizes the implicit models that organize our understanding of the past. The goal is to simultaneously achieve three things – generality, realism, and precision (Winterhalder, 1981). Though impossible from a practical standpoint, a balance between these factors can be achieved by selecting appropriate variables in order to address specific questions. Every model, both implicit and explicit, is incomplete. But if they are informed by real-world phenomena, they have

Results

Path Distance was calculated for each scenario (Summer Travel, Winter Travel, and Winter Travel by Dogsled, all of which include the potential for travel by watercraft). The result is a raster with cumulative travel time values between every point on the landscape to the nearest source location. These values represent a minimum cumulative cost of movement as defined by the variables considered here. Travel cost values at the location of each archaeological site are considered to be a proxy for

Path Distance vs. Euclidean Distance for approximating provenance

Although many archaeological sites are linked to the same geologic source by both Euclidean and Path Distance, a large enough proportion of them show a change in source linkage when using Path Distance that we consider Path Distance to be a significant improvement over Euclidean Distance when approximating provenance, specifically because Path Distance considers variable travel behavior within a model based on real world geographic and cultural variation. Although these determinations are based

Conclusions

This analysis is only a first step in a larger process of reorganizing one way of understanding the distribution of archaeological copper across this region. By structuring the models as comparative between source regions, generalizations were necessary in spatial resolution and the particularities of travel behaviors that may be documented ethnographically. This is an acceptable level of generalization in this case because by organizing the distribution of native copper using travel time as a

Acknowledgements

Many organizations and individuals assisted with the compilation and analysis of the site databases used in this research, including Jeff Speakman of the Center for Applied Isotope Studies at UGA; the British Columbia Archaeology Branch; the Royal British Columbia Museum; the University of British Columbia; Simon Fraser University; Prince of Wales Northern Heritage Centre and NWT Cultural Places Division (Tom Andrews, Glen MacKay, Joanne Bird, Susan Irving); Yukon Heritage Resources Unit,

References (109)

  • M.S. Taliaferro et al.

    Obsidian procurement, least cost path analysis, and social interaction in the Mimbres area of southwestern New Mexico

    J. Archaeol. Sci.

    (2010)
  • S. Acheson

    The thin edge: evidence for precontact use and working of metal on the Northwest Coast

  • K.M. Ames

    Going by boat: the forager-collector continuum at sea

  • K.M. Ames

    The sites, analytical units, and chronology

  • T.D. Andrews

    Final Report: NWT Archaeologists Permit 93-761

    (1996)
  • T.D. Andrews et al.

    The Idaa trail: archaeology and the Dogrib cultural landscape, Northwest Territories, Canada (with John B. Zoe)

  • C. Aporta

    Routes, trails and tracks: trail breaking among the Inuit of Igloolik

    Etudes/Inuit/Studies

    (2004)
  • C. Aporta

    The trail as home: Inuit and their pan-Arctic network of routes

    Hum. Ecol.

    (2009)
  • Aporta, C., Bravo, M., Taylor, F., n.d. Pan Inuit Trails Atlas [WWW Document]. URL...
  • M. Appelt et al.

    Late Dorset

  • C.D. Arnold

    Development of Mackenzie Inuit culture

  • C.D. Arnold et al.

    Winter houses of the Mackenzie Inuit

    Bull. Soc. Study Archit. Canada

    (1991)
  • J.M. Bell

    Report on the Topography and Geology of Great Bear Lake and of a Chain of Lakes and Streams Thence to Great Slave Lake - Annual Report 1899

    (1901)
  • Michael Blake

    Fraser Valley trade and prestige as seen from Scowlitz

  • F. Boas

    The Central Eskimo

  • C.E. Borden

    Facts and problems of Northwest Coast prehistory

    Anthropol. Br. Columbia

    (1951)
  • Charles E. Borden

    Cultural History of the Fraser-Delta Region: An Outline

    BC Stud.

    (1970)
  • A.H. Brooks

    A Reconnaissance From Pyramid Harbor to Eagle City, Alaska, Including a Description of the Copper Deposits of the Upper White and Tanana Rivers

    (1900)
  • D.V. Burley

    Marpole: Anthropological Reconstructions of a Prehistoric Northwest Coast Culture Type

    (1980)
  • L.J. Burwash

    Coronation Gulf Copper Deposits

    (1930)
  • D.W. Clark

    Archaeological Reconnaissance in Northern Interior District of Mackenzie: 1969, 1970, 1972

    (1975)
  • T. Clark

    Rewriting Marpole: The Path to Cultural Complexity in the Gulf of Georgia

    (2013)
  • B. Collignon

    Knowing Places: The Inuinnait, Landscapes, and the Environment, Circumpola

    (2006)
  • H.K. Cooper

    Copper and social complexity: Frederica de Laguna's contribution to our understanding of the role of metals in Native Alaskan society

    Arctic Anthropol

    (2006)
  • H.K. Cooper

    The Anthropology of Native Copper Technology and Social Complexity in Alaska and the Yukon Territory: An Analysis Using Archaeology, Archaeometry, and Ethnohistory

    (2007)
  • H.K. Cooper

    The life (lives) and times of native copper in Northwest North America

    World Archaeol.

    (2011)
  • H.K. Cooper

    Innovation and Prestige Among Northern Hunter-Gatherers: Late Prehistoric Native Copper Use in Alaska and Yukon

    Am. Antiq.

    (2012)
  • H.K. Cooper

    Arctic archaeometallurgy

  • H.K. Cooper et al.

    Is that awl? Experimental insight into native copper working and innovation

    Ethnoarchaeology

    (2015)
  • P. Dawson

    The Thule-Inuit succession in the Central Arctic

  • E.J. Dixon et al.

    The Emerging Archaeology of Glaciers and Ice Patches: Examples from Alaska’s Wrangell-St. Elias National Park and Preserve

    Am. Antiq.

    (2005)
  • F. de Laguna

    Under Mount St. Elias: The History and Culture of the Yakutat Tlingit, Smithsonia

    (1972)
  • de Laguna, F., McClellan, C., 1981. Ahtna, in: Helm, J. (Ed.), Subarctic. William C. Sturtevant, general editor,...
  • Frederica de Laguna et al.

    Archaeology of the Yakutat Bay area, Alaska. Smithsonian Institution Bureau of American Ethnology Bulletin, 192

    (1964)
  • K.L. Ehrhardt

    Copper working technologies, contexts of use, and social complexity in the eastern woodlands of Native North America

    J. World Prehistory

    (2009)
  • G. Emmons

    The Tlingit Indians

    (1991)
  • J. Franklin

    Narrative of a Journey to the Shores of the Polar Sea in the Years 1819, 20, 21 and 22

    (1823)
  • U. Franklin et al.

    An examination of prehistoric copper technology and copper sources in Western Arctic and Subarctic North America, Archaeolog

  • J. Fredericks et al.

    British Columbian Mining and Mineral Exploration Overview 2009

    (2009)
  • T.M. Friesen

    Pan-Arctic population movements: The Early Paleo-Inuit and Thule Inuit migrations

  • Cited by (4)

    • Re-examining trade networks in Late Woodland Virginia (900–1600 CE): An LA-ICP-MS analysis of copper artifacts

      2019, Journal of Archaeological Science: Reports
      Citation Excerpt :

      As an extension of this project, a future project will examine the distribution and variety of the European smelted copper artifacts identified in the early stages of our analysis (after Hudgins, 2004, 2005, 2006; Legg et al., 2018; Michelaki et al., 2015). This research will use geospatial methods (including an analysis of least cost paths/surfaces) to refine understandings of how the archaeological sites, copper mines, and trails discussed in this study articulated in local and regional landscapes of movement (e.g., Bollwerk, 2015; Cross, 2012; Nash, 2009; Pike et al., 2019; Sampeck et al., 2015). Although outside the scope of this paper, we recommend that future research on copper trade in Late Woodland Virginia include an expanded sample of geological copper (in particular, additional copper source samples from the Southeast), as well as an expanded sample of archaeological copper (wherever access to/analysis of additional copper artifacts is both possible and desired by descendant communities).

    View full text